Académique Documents
Professionnel Documents
Culture Documents
I N T E R N A T I O N A L M O N E T A R Y F U N D
2017 International Monetary Fund
Cataloging-in-Publication Data
HC10.80
The World Economic Outlook (WEO) is a survey by the IMF staff published twice a
year, in the spring and fall. The WEO is prepared by the IMF staff and has benefited
from comments and suggestions by Executive Directors following their discussion of the
report on September 21, 2017. The views expressed in this publication are those of the
IMF staff and do not necessarily represent the views of the IMFs Executive Directors
or their national authorities.
ERRATUM
Correction to Figure 3.6, panel 6 in Chapter 3 (page 125): The data in panel 6 (Wildfire) are
incorrect in the printed version of the report and the initial PDF posted online. The corrected
figure panel appears in this version of the report posted online on Oct. 11, 2017.
CONTENTS
Preface xii
Foreword xiii
Executive Summary xv
Chapter 3: The Effects of Weather Shocks on Economic Activity: How Can Low-Income Countries Cope? 117
Introduction 117
Temperature and Precipitation: Historical Patterns and Projections 121
The Macroeconomic Impact of Weather Shocks 125
Coping with Weather Shocks and Climate Change 131
Long-Term Effects of Temperature IncreaseA Model-Based Approach 137
Summary and Policy Implications 139
Tables
Table 1.1. Overview of the World Economic Outlook Projections 14
Scenario Table 1. Assumed Policy Actions Relative to the WEO Baseline 36
Table 1.3.1. Correlates of Growth Projections, EMDEs, 201722 46
Annex Table 1.1.1. European Economies: Real GDP, Consumer Prices, Current Account Balance, and
Unemployment 63
Annex Table 1.1.2. Asian and Pacific Economies: Real GDP, Consumer Prices, Current Account Balance, and
Unemployment 64
Annex Table 1.1.3. Western Hemisphere Economies: Real GDP, Consumer Prices, Current Account Balance,
and Unemployment 65
Annex Table 1.1.4. Commonwealth of Independent States Economies: Real GDP, Consumer Prices, Current
Account Balance, and Unemployment 66
Annex Table 1.1.5. Middle East, North African Economies, Afghanistan, and Pakistan: Real GDP, Consumer
Prices, Current Account Balance, and Unemployment 67
Annex Table 1.1.6. Sub-Saharan African Economies: Real GDP, Consumer Prices, Current Account Balance,
and Unemployment 68
Annex Table 1.1.7. Summary of World Real per Capita Output 69
Table 2.3.1. Precrisis Financial Vulnerabilities and Postcrisis Labor Adjustments 97
Annex Table 2.1.1. Country Coverage 98
Annex Table 2.1.2. Data Sources 98
Annex Table 2.2.1. Aggregate Forces and Sectoral Exposures 102
Annex Table 2.3.1. Estimates of Wage Phillips Curves 103
Annex Table 2.3.2. Estimates of Wage Phillips Curves with Alternative Measures 104
Annex Table 2.3.3. Estimation of Wage Phillips Curve Augmented with Involuntary
Part-Time Employment Share by Country Group 106
Annex Table 2.3.4. Estimation of Wage Phillips Curve Augmented with Involuntary
Part-Time Employment Share: Full Sample and Countries with Unemployment Rates
Lower than 200007 Average 106
Annex Table 2.3.5. Estimation of Wage Phillips Curve Augmented with Involuntary
Part-Time Employment Share: Countries with Unemployment Rates Moderately Higher and
Appreciably Higher than 200007 Average 107
Annex Table 2.3.6. Estimation of Wage Phillips Curve Augmented with Temporary Contract
Employment Share: Full Sample and Countries with Unemployment Rates Lower than
200007 Average 107
Annex Table 2.3.7. Estimation of Wage Phillips Curve Augmented with Temporary Contract
Employment Share: Countries with Unemployment Rates Moderately Higher and
Appreciably Higher than 200007 Average 108
Annex Table 2.3.8. Estimation of Wage Phillips Curve Augmented with Structural Variables 109
Annex Table 2.3.9. Estimation of Wage Phillips Curve Augmented with Structural Variables:
Excluding 2008 and 2009 110
Annex Table 2.3.10. Drivers of Involuntary Part-Time Employment Share, Aggregate Analysis 110
Annex Table 2.3.11. Drivers of Sectoral Nominal Wage Growth 111
Annex Table 2.3.12. Drivers of Sectoral Part-Time Employment Shares 111
Annex Table 2.3.13. Drivers of Nominal Wage Growth, Employment Growth, and
Part-Time Employment 113
Table 3.1.1. Characteristics of the Average Tropical Cyclone by Country Group 142
Table 3.1.2. Effect of Weather and Wind Shocks on Economic Activity 143
Table 3.5.1. Effect of Historical Climate on Current Real Output 154
Annex Table 3.1.1. Data Sources 159
Annex Table 3.1.2. Country and Territory Groups 160
Annex Table 3.2.1. Effect of Weather Shocks on Natural Disasters, 19902014 162
Annex Table 3.3.1. Effect of Weather Shocks on Output 164
Annex Table 3.3.2. Effect of Weather Shocks on Sectoral Output 167
Annex Table 3.3.3. Effect of Weather Shocks on Productivity, Capital, and Labor 168
Annex Table 3.3.4. Role of Policy Buffers 170
Annex Table 3.3.5. Role of Structural Policies and Institutions 171
Annex Table 3.3.6. Role of Development: Evidence from Subnational Data 172
Annex Table 3.4.1. Effect of Weather Shocks and Natural Disasters on Emigration, 19802015 173
Annex Table 3.5.1. Parameterization of the Debt, Investment, and Growth Model 175
Annex Table 4.1.1. Data Sources for Quarterly Fiscal Data by Source Country 207
Annex Table 4.1.2. Data Sources for Recipient Countries 208
Annex Table 4.1.3. Recipient Countries in Sample 208
Online Tables
Table B1. Advanced Economies: Unemployment, Employment, and Real GDP per Capita
Table B2. Emerging Market and Developing Economies: Real GDP
Table B3. Advanced Economies: Hourly Earnings, Productivity, and Unit Labor Costs in Manufacturing
Table B4. Emerging Market and Developing Economies: Consumer Prices
Table B5. Summary of Fiscal and Financial Indicators
Table B6. Advanced Economies: General and Central Government Net Lending/Borrowing and
General Government Net Lending/Borrowing Excluding Social Security Schemes
Table B7. Advanced Economies: General Government Structural Balances
Table B8. Emerging Market and Developing Economies: General Government Net Lending/Borrowing
and Overall Fiscal Balance
Table B9. Emerging Market and Developing Economies: General Government Net Lending/Borrowing
Table B10. Selected Advanced Economies: Exchange Rates
Table B11. Emerging Market and Developing Economies: Broad Money Aggregates
Table B12. Advanced Economies: Export Volumes, Import Volumes, and Terms of Trade in
Goods and Services
Table B13. Emerging Market and Developing Economies by Region: Total Trade in Goods
Table B14. Emerging Market and Developing Economies by Source of Export Earnings:
Total Trade in Goods
Table B15. Summary of Current Account Transactions
Table B16. Emerging Market and Developing Economies: Summary of External Debt and Debt Service
Table B17. Emerging Market and Developing Economies by Region: External Debt by Maturity
Table B18. Emerging Market and Developing Economies by Analytical Criteria:
External Debt by Maturity
Table B19. Emerging Market and Developing Economies: Ratio of External Debt to GDP
Table B20. Emerging Market and Developing Economies: Debt-Service Ratios
Table B21. Emerging Market and Developing Economies, Medium-Term Baseline Scenario:
Selected Economic Indicators
Figures
Figure 1.1. Global Activity Indicators 2
Figure 1.2. Global Fixed Investment and Trade 3
Figure 1.3. Commodity Prices 3
Figure 1.4. Global Inflation 4
vi International Monetary Fund | October 2017
CONTENTS
Figure 3.9. Effect of Temperature Increase on Sectoral Output Estimated at the Temperature of the
Median Low-Income Developing Country 128
Figure 3.10. Effect of Temperature Increase on Productivity, Capital, and Labor Input Estimated
at the Temperature of the Median Low-Income Developing Country 130
Figure 3.11. Effect of Temperature Increase on Real per Capita Output Estimated at the Temperature
of the Median Low-Income Developing Country over Time 131
Figure 3.12. Coping with Weather Shocks and Climate Change: A Toolkit 132
Figure 3.13. Role of Policy Buffers 134
Figure 3.14. Role of Structural Policies and Institutions 134
Figure 3.15. Role of Development: Evidence from Subnational Data 135
Figure 3.16. Effect of Temperature and Natural Disasters on International Migration 136
Figure 3.17. Long-Term Impact of Temperature Increase for a Representative Low-Income
Developing Country: Model Simulations 138
Figure 3.18. Vulnerability to Temperature Increase and Adaption Prospects 141
Figure 3.1.1. Effect of Tropical Cyclone Exposure on Real GDP per Capita 144
Figure 3.1.2. Cumulative Effect of Average Tropical Cyclone on Real GDP per Capita after Seven Years 144
Figure 3.2.1. Role of Policies: A Model-Based Analysis 145
Figure 3.4.1. Insurance Penetration: Non-Life Insurance Premium 151
Figure 3.4.2. Catastrophe Bond Market 152
Figure 3.4.3. Temperature Shocks and Stock Price Predictability: Food and Beverages Sector 152
Figure 3.6.1. Effectiveness of Mitigation Policies in China 157
Annex Figure 3.3.1. Effect of Temperature Increase on Real per Capita Output across the Globe,
with Countries Rescaled in Proportion to Their Projected Population as of 2100 165
Annex Figure 3.6.1. The Long-Term Impact of Temperature Increase on Real per Capita Output
across the Globe 177
Figure 4.1. Output Gap in Selected Countries 186
Figure 4.2. The Transmission of a Fiscal Shock 188
Figure 4.3. Tracking Tax Shocks in the United States 191
Figure 4.4. Dynamic Responses of Recipient Countries Output to Fiscal Shocks 192
Figure 4.5. Spillovers of Fiscal Shocks on Recipient Countries Output 193
Figure 4.6. Dynamic Responses of Components of Recipient Countries Output to a Fiscal Shock 193
Figure 4.7. Spillovers under Various Economic and Policy Conditions 194
Figure 4.8. Dynamic Responses of Components of Recipient Countries Output under
Normal Times and Effective Lower Bound in Recipient Countries 195
Figure 4.9. Dynamic Responses of Recipient Countries Output to US Spending Shock under
Various Exchange Rate Regimes 196
Figure 4.10. Impact of Fiscal Shocks on Global GDP Based on Various Instruments 197
Figure 4.11. Spillovers from US Fiscal Shocks with and without Monetary Accommodation 198
Figure 4.12. Regional GDP Impact of Government Spending Shocks from the United States,
Europe, and China 199
Figure 4.13. Dynamic Responses to a US Government Spending Shock 199
Figure 4.14. Spillovers from US Spending Shock with and without a US Term-Premium Increase 200
Figure 4.15. Spillovers from Corporate Income Tax Reduction Financed by an Offsetting
Increase in Value-Added Tax 201
Figure 4.16. Spillovers from Increase in Government Investment in Five Major Economies 203
Figure 4.1.1. Response of Recipient Countries Trade Balance and Real Exchange Rate vis--vis US Dollar 205
Annex Figure 4.3.1. Effects of Spending Shock and Tax Shock on Recipient Countries Output:
Comparison with Panel Vector Autoregression 211
Annex Figure 4.3.2. Effects of Spending and Tax Shock on Recipient Countries Output: Forecast Errors 212
Annex Figure 4.3.3. Effects of US Tax Shock on Recipient Countries Output: Comparison with US
Narrative Tax Shock, 19952007 213
A number of assumptions have been adopted for the projections presented in the World Economic Outlook (WEO).
It has been assumed that real effective exchange rates remained constant at their average levels during July 20 to August
17, 2017, except for those for the currencies participating in the European exchange rate mechanism II (ERM II),
which are assumed to have remained constant in nominal terms relative to the euro; that established policies of national
authorities will be maintained (for specific assumptions about fiscal and monetary policies for selected economies, see
Box A1 in the Statistical Appendix); that the average price of oil will be $50.28 a barrel in 2017 and $50.17 a barrel in
2018 and will remain unchanged in real terms over the medium term; that the six-month London interbank offered
rate (LIBOR) on US dollar deposits will average 1.4 percent in 2017 and 1.9 percent in 2018; that the three-month
euro deposit rate will average 0.3 percent in 2017 and 2018; and that the six-month Japanese yen deposit rate will
yield on average 0.1 percent in 2017 and 0.2 percent in 2018. These are, of course, working hypotheses rather than
forecasts, and the uncertainties surrounding them add to the margin of error that would in any event be involved in the
projections. The estimates and projections are based on statistical information available through September 22, 2017.
The following conventions are used throughout the WEO:
. . . to indicate that data are not available or not applicable;
between years or months (for example, 201617 or JanuaryJune) to indicate the years or months
covered, including the beginning and ending years or months; and
/ between years or months (for example, 2016/17) to indicate a fiscal or financial year.
Billion means a thousand million; trillion means a thousand billion.
Basis points refers to hundredths of 1 percentage point (for example, 25 basis points are equivalent to of 1
percentage point).
Data refer to calendar years, except in the case of a few countries that use fiscal years. Please refer to Table F in
the Statistical Appendix, which lists the economies with exceptional reporting periods for national accounts and
government finance data for each country.
For some countries, the figures for 2016 and earlier are based on estimates rather than actual outturns. Please
refer to Table G in the Statistical Appendix, which lists the latest actual outturns for the indicators in the national
accounts, prices, government finance, and balance of payments indicators for each country.
What is new in this publication:
Data for Somalia are included in the emerging market and developing economies group composites. Somalia is
classified as a member of the Middle East and North Africa region.
Starting with the October 2017 WEO, the real GDP per capita data in Statistical Tables A1, B1, and B2 are
shown at purchasing power parity. This differs from the treatment of these data in the April 2017 WEO and
earlier issues, in which the data were shown in local national currency.
In the tables and figures, the following conventions apply:
If no source is listed on tables and figures, data are drawn from the WEO database.
When countries are not listed alphabetically, they are ordered on the basis of economic size.
Minor discrepancies between sums of constituent figures and totals shown reflect rounding.
As used in this report, the terms country and economy do not in all cases refer to a territorial entity that is
a state as understood by international law and practice. As used here, the term also covers some territorial entities
that are not states but for which statistical data are maintained on a separate and independent basis.
Composite data are provided for various groups of countries organized according to economic characteristics or
region. Unless noted otherwise, country group composites represent calculations based on 90 percent or more of
the weighted group data.
The boundaries, colors, denominations, and any other information shown on the maps do not imply, on the
part of the International Monetary Fund, any judgment on the legal status of any territory or any endorsement or
acceptance of such boundaries.
This version of the World Economic Outlook (WEO) is available in full through the IMF eLibrary (www.elibrary.
imf.org) and the IMF website (www.imf.org). Accompanying the publication on the IMF website is a larger compila-
tion of data from the WEO database than is included in the report itself, including files containing the series most
frequently requested by readers. These files may be downloaded for use in a variety of software packages.
The data appearing in the WEO are compiled by the IMF staff at the time of the WEO exercises. The histori-
cal data and projections are based on the information gathered by the IMF country desk officers in the context
of their missions to IMF member countries and through their ongoing analysis of the evolving situation in each
country. Historical data are updated on a continual basis as more information becomes available, and structural
breaks in data are often adjusted to produce smooth series with the use of splicing and other techniques. IMF
staff estimates continue to serve as proxies for historical series when complete information is unavailable. As a
result, WEO data can differ from those in other sources with official data, including the IMFs International
Financial Statistics.
The WEO data and metadata provided are as is and as available, and every effort is made to ensure their
timeliness, accuracy, and completeness, but it cannot be guaranteed. When errors are discovered, there is a con-
certed effort to correct them as appropriate and feasible. Corrections and revisions made after publication are
incorporated into the electronic editions available from the IMF eLibrary (www.elibrary.imf.org) and on the IMF
website (www.imf.org). All substantive changes are listed in detail in the online tables of contents.
For details on the terms and conditions for usage of the WEO database, please refer to the IMF Copyright and
Usage website (www.imf.org/external/terms.htm).
Inquiries about the content of the WEO and the WEO database should be sent by mail, fax, or online forum
(telephone inquiries cannot be accepted):
World Economic Studies Division
Research Department
International Monetary Fund
700 19th Street, NW
Washington, DC 20431, USA
Fax: (202) 623-6343
Online Forum: www.imf.org/weoforum
The analysis and projections contained in the World Economic Outlook are integral elements of the IMFs surveil-
lance of economic developments and policies in its member countries, of developments in international financial
markets, and of the global economic system. The survey of prospects and policies is the product of a comprehen-
sive interdepartmental review of world economic developments, which draws primarily on information the IMF
staff gathers through its consultations with member countries. These consultations are carried out in particular by
the IMFs area departmentsnamely, the African Department, Asia and Pacific Department, European Depart-
ment, Middle East and Central Asia Department, and Western Hemisphere Departmenttogether with the
Strategy, Policy, and Review Department, the Monetary and Capital Markets Department, and the Fiscal Affairs
Department.
The analysis in this report was coordinated in the Research Department under the general direction of Maurice
Obstfeld, Economic Counsellor and Director of Research. The project was directed by Gian Maria Milesi-Ferretti,
Deputy Director, Research Department; Oya Celasun, Division Chief, Research Department; and Helge Berger,
Assistant Director, Research Department and Head of the IMFs Spillover Task Force.
The primary contributors to this report were Sebastian Acevedo, Patrick Blagrave, Christian Bogmans,
Francesco Grigoli, Giang Ho, Gee Hee Hong, Zska Kczn, Ksenia Koloskova, Toh Kuan, Weicheng Lian,
Akito Matsumoto, Mico Mrkaic, Malhar Nabar, Natalija Novta, Marcos Poplawski-Ribeiro, Evgenia Pugacheva,
Petia Topalova, and Esteban Vesperoni.
Other contributors include JaeBin Ahn, Gavin Asdorian, Manoj Atolia, Claudio Baccianti, Kimberly Beaton,
Jared Bebee, Felicia Belostecinic, John C. Bluedorn, Luis Cato, Eugenio Cerutti, Angela Espiritu, Rachel Yuting
Fan, Alan Xiaochen Feng, Bertrand Gruss, Meron Haile, Mahnaz Hemmati, Benjamin Hilgenstock, Ava Yeabin
Hong, Benjamin Hunt, Hao Jiang, Christopher Johns, Sung Eun Jung, Alimata Kini-Kabor, Lama Kiyasseh,
Robin Koepke, Jungjin Lee, Yiqun Li, Ricardo Marto, Trevor Charles Meadows, Joannes Mongardini, Susanna
Mursula, Cynthia Nyanchama Nyakeri, Emory Oakes, Ian Parry, Adina Popescu, Daniel Rivera Greenwood, Ippei
Shibata, Kadir Tanyeri, Nicholas Tong, Jilun Xing, Yuan Zeng, Qiaoqiao Zhang, and Huiyuan Zhao.
Joseph Procopio from the Communications Department led the editorial team for the report, with production
and editorial support from Michael Harrup, Christine Ebrahimzadeh, and Linda Kean and editorial assistance
from James Unwin, Lucy Scott Morales, Sherrie Brown, and Vector Talent Resources.
The analysis has benefited from comments and suggestions by staff members from other IMF departments, as
well as by Executive Directors following their discussion of the report on September 21,2017. However, both pro-
jections and policy considerations are those of the IMF staff and should not be attributed to Executive Directors or
to their national authorities.
T
he global cyclical upswing that began at which central banks can no longer lower interest
midway through 2016 continues to gather rates, more likely. Chapter 2 of this World Economic
strength. Only a year and a half ago, the Outlook report studies the recent surprisingly slow
world economy faced stalling growth and growth of nominal wages, which reinforces a lon-
financial market turbulence. The picture now is very ger trend of stagnant median wages, rising income
different, with accelerating growth in Europe, Japan, inequality, and job polarization such that middle-skill
China, and the United States. Financial conditions but well-paying jobs have become increasingly scarce.
remain buoyant across the world, and financial Those developments have stoked considerable popular
markets seem to be expecting little turbulence going anti-globalization backlashone significant threat to
forward, even as the Federal Reserve continues its the world economyalthough technological develop-
monetary normalization process and the European ments and government policies together have played
Central Bank inches up to its own. larger roles in increasing income inequality, and fears
These positive developments give good cause for of faster automation are a current cause of anxiety.
greater confidence, but neither policymakers nor mar- Emerging markets have faced similar pressures in the
kets should be lulled into complacency. A closer look face of trade liberalization and technological change,
suggests that the global recovery may not be sustain- but growth has in many cases lifted all deciles of their
ablenot all countries are participating, inflation often income distributions and attitudes toward trades
remains below target with weak wage growth, and the effects on labor markets remain largely optimistic.
medium-term outlook still disappoints in many parts Across countries. The current upswing reaches more
of the world. The recovery is also vulnerable to seri- broadly than any in a decaderoughly 75 percent of
ous risks. Financial markets that ignore these risks are the world economy, measured by GDP at purchasing
susceptible to disruptive repricing, and are sending a power parityis sharing in the acceleration. But that
misleading message to policymakers. Policymakers, in means the glass is 25 percent empty, implying a drag
turn, need to maintain a longer-term vision and seize on world growth and a potential source of destabilizing
the current opportunity to implement the structural political shocks. Emerging and low-income commod-
and fiscal reforms needed for greater resilience, produc- ity exporters, especially energy exporters, continue to
tivity, and investment. The possibility that they dont struggle, as do several countries experiencing civil or
governments far too often wait for crises to push them political unrest, mostly in the Middle East, North and
into decisive actionis itself a source of risks to the sub-Saharan Africa, and Latin America. And many of
outlook, as well as a barrier to more inclusive and these same countries are the ones that are also most
sustainable growth. Recent economic progress provides exposed to the negative impacts of climate change
a global environment of opportunity, and policymakers already being felt via more frequent extreme weather
should not let their chance go to waste. events in some regions, such as heat waves and heavy
The current recovery is incomplete in some impor- precipitation. Chapter 3 focuses on the economic costs
tant ways: within countries, across countries, and over of climate change and the need for adaptation invest-
time. ments in low-income countries. Advanced economies
Within countries. Even as negative output gaps close will not be immune to future climate developments,
across the advanced economies, growth in nominal howeverthrough either direct impacts in some
and real wages remains weak compared with past advanced regions, such as the coastal United States, or
recoveries. Weak wage growth is one source of the sur- the spillovers from mass migrations and geopolitical
prisingly weak inflation that itself is a source of con- instability emanating from poorer countries.
cern, as it leaves nominal interest rates low and makes Over time. Behind recent positive growth develop-
encounters with the effective lower bound, the point ments, longer-term trend per capita growth rates in
many economies will be lower than trend growth rates through much-needed productive infrastructure
of the past. In particular, most advanced economies investment or through fiscal spending that supports
face medium-term growth rates significantly lower structural reforms. This global fiscal package can also
than in the decade before the global financial crisis help reduce excess global imbalances.
of 200709. The reasons behind these slowdowns Critically important to inclusive and sustainable
differ across countries. For some economies, notably growth is investment in people at all life-cycle stages,
Chinas, declining long-term growth is a natural result but especially the young. Better education, training,
of rebalancing and convergence. For emerging com- and retraining can both ease labor market adjustment
modity exporters, which benefited from Chinas own to secular economic transformationcoming from all
rapid manufacturing growth in years past, perma- sources, not only tradeand raise productivity. In the
nently lower export prices call for new growth models. short term, the excessive youth unemployment that
For advanced economies, expected slow productivity afflicts many countries urgently deserves attention.
growth and aging workforces play major roles. Lower Investing in human capital should help to push labors
trend per capita growth rates can be problematic for income share upward, contrary to the broad trend of
several reasons: they make it harder for the poor to recent decadesbut governments should also consider
raise their living standards; they raise the pain of real- correcting distortions that may have reduced workers
locating resources in the face of economic changes; bargaining power excessively. In sum, policy should
they deter productivity-enhancing investment; they promote an environment conducive to sustainable real
harm the sustainability of publicly funded social safety wage growth.
nets; and they feed political resentment by undermin- Numerous global problems require multilateral
ing hopes for the future and beliefs about the fairness action. Priorities for mutually beneficial coopera-
of economic outcomes. In turn, these forces could tion include strengthening the global trading system,
derail the baseline forecast. further improving financial regulation, enhancing the
The preceding gaps in the recovery challenge global financial safety net, reducing international tax
policymakers to actionaction that should take place avoidance, fighting famine and infectious diseases,
now, while times are good. Needed structural reforms mitigating greenhouse gas emissions before they create
differ across countries, but all countries have ample more irreversible damage, and helping poorer coun-
room for measures that would raise economic resil- tries, which are not themselves substantial emitters,
ience along with potential output. For some countries adapt to climate change. If the strength of the cur-
where output gaps have closed, the time has come to rent upswing makes the moment ideal for domestic
think about gradual fiscal consolidation, to reduce reforms, its breadth makes multilateral cooperation
swollen public debt levels and create buffers to be opportune. Policymakers should act while the window
used in the next recession. Such actions could entail of opportunity is open.
adverse spillovers abroad, as discussed in Chapter
4. Countries with more fiscal space can, however, Maurice Obstfeld
offset the reduction in global demandfor example, Economic Counsellor
The global upswing in economic activity is strengthening. gaps shrink (leaving less scope for cyclical improve-
Global growth, which in 2016 was the weakest since ment) and demographic factors and weak productivity
the global financial crisis at 3.2 percent, is projected weigh on potential growth.
to rise to 3.6 percent in 2017 and to 3.7 percent in Changes to aggregate growth forecasts relative
2018. The growth forecasts for both 2017 and 2018 to the April 2017 WEO are generally positive but
are 0.1 percentage point stronger compared with the modest, with some meaningful changes for specific
April 2017 World Economic Outlook (WEO) forecast. country groups and individual countries.
Broad-based upward revisions in the euro area, Japan, In line with stronger-than-expected momentum in
emerging Asia, emerging Europe, and Russiawhere the first half of 2017, the forecast sees a stronger
growth outcomes in the first half of 2017 were better rebound in advanced economies in 2017 (to 2.2
than expectedmore than offset downward revisions percent versus 2.0 percent foreseen in April), driven
for the United States and the United Kingdom. by stronger growth in the euro area, Japan, and
But the recovery is not complete: while the baseline Canada. In contrast, compared with the April 2017
outlook is strengthening, growth remains weak in many WEO forecast, growth has been marked down
countries, and inflation is below target in most advanced for 2017 in the United Kingdom and for both
economies. Commodity exporters, especially of fuel, are 2017 and 2018 in the United States, implying a
particularly hard hit as their adjustment to a sharp 0.1 percentage-point aggregate growth downgrade
stepdown in foreign earnings continues. And while short- for advanced economies in 2018. Activity in the
term risks are broadly balanced, medium-term risks are United Kingdom slowed more than anticipated
still tilted to the downside. The welcome cyclical pickup in the first half of 2017; as for the United States,
in global activity thus provides an ideal window of given the significant policy uncertainty, the forecast
opportunity to tackle the key policy challengesnamely now uses a baseline assumption of unchanged poli-
to boost potential output while ensuring its benefits are cies, whereas in April it assumed a fiscal stimulus
broadly shared, and to build resilience against down- driven by then-anticipated tax cuts.
side risks. A renewed multilateral effort is also needed Growth prospects for emerging and developing
to tackle the common challenges of an integrated global economies are marked up by 0.1 percentage point
economy. for both 2017 and 2018 relative to April, primarily
The global pickup in activity that started in the owing to a stronger growth projection for China.
second half of 2016 gained further momentum in The countrys 2017 forecast (6.8 percent, against
the first half of 2017. Growth is projected to rise over 6.6 percent in April) reflects stronger growth
this year and next in emerging market and devel- outturns in the first half of 2017 as well as more
oping economies, supported by improved external buoyant external demand. For 2018, the revision
factorsa benign global financial environment and mainly reflects an expectation that the authorities
a recovery in advanced economies. Growth in China will maintain a sufficiently expansionary policy mix
and other parts of emerging Asia remains strong, and to meet their target of doubling real GDP between
the still-difficult conditions faced by several commod- 2010 and 2020. Growth forecasts have also been
ity exporters in Latin America, the Commonwealth marked up for emerging Europe for 2017, reflect-
of Independent States, and sub-Saharan Africa show ing stronger growth in Turkey and other countries
some signs of improvement. In advanced economies, in the region, for Russia for 2017 and 2018, and
the notable 2017 growth pickup is broad based, with Brazil in 2017.
stronger activity in the United States and Canada,
the euro area, and Japan. Prospects for medium-term Financial market sentiment has generally been
growth are more subdued, however, as negative output strong, with continued gains in equity markets in
both advanced and emerging market economies. which would take a toll on macroeconomic activity
Given current expectations of a more gradual pace of through weaker confidence, lower asset valuations,
monetary policy normalization compared with March, and wider risk premia.
US long-term interest rates have declined by some 25 Financial turmoil in emerging market economies. The
basis points since then, and the dollar has depreciated upward revision to Chinas growth forecasts reflects
by more than 5 percent in real effective terms, with a a slower rebalancing of activity toward services and
commensurate real appreciation of the euro. Despite consumption, a higher projected debt trajectory, and
expectations of more robust global demand going diminished fiscal space. Unless the Chinese authori-
forward, commodity prices have remained low, with oil ties counter the associated risks by accelerating their
prices reflecting stronger-than-anticipated supply. recent encouraging efforts to curb the expansion of
Headline consumer price inflation has softened credit, these factors imply a heightened probability
since the spring, as the boost to prices from the oil of a sharp growth slowdown in China, with adverse
price recovery of 2016 has faded and the decline in oil international repercussions. Following a period of
prices in recent months has started to exert downward abundant credit supply, a sudden tightening of
pressure. Despite stronger growth in domestic demand, global financial conditions (and an associated US
core inflation has generally remained muted across dollar appreciation) could expose financial fragili-
advanced economies, reflecting still-weak wage growth ties in some emerging markets, imposing strains on
(Chapter 2). Inflation is likely to rise only gradually economies with US dollar pegs, high leverage, and
toward central bank targets. Across emerging and balance sheet mismatches.
developing economies, the waning of pass-through Persistently low inflation in advanced economies. If
effects from earlier currency depreciations against the domestic demand were to falter, it could lead to
US dollar, and in some cases recent appreciations, have a decline in medium-term inflation expectations,
helped moderate core inflation rates. prolonging and reinforcing the weakness in infla-
Short-term risks are broadly balanced. On the tion. Low inflation and nominal interest rates would
positive side, the recovery could strengthen further, in turn reduce central banks capacity to lower real
supported by strong consumer and business confidence interest rates to restore full employment in an eco-
and benign financial conditions. At the same time, in nomic downturn.
an environment of high policy uncertainty and geo- A broad rollback of the improvements in financial
political tensions, policy misstepswhich the baseline regulation and oversight achieved since the global
assumes will be avoidedcould take a toll on market financial crisis. Such a rollback could lower capital
confidence, resulting in tighter financial conditions and and liquidity buffers or weaken supervisory effective-
weaker asset prices. ness, with negative repercussions for global financial
Risks to growth in the medium term are still skewed stability.
to the downside, owing to several potential hazards: An inward shift in policies. A shift toward protection-
A more rapid and sizable tightening of global financial ism would reduce trade and cross-border investment
conditions. This could take the form of higher long- flows, harming global growth.
term interest rates in the United States and else- Noneconomic factors. These would include geopoliti-
where, triggered by faster-than-expected monetary cal tensions, domestic political discord, risks from
policy normalization or a decompression of term weak governance and corruption, extreme weather
premia, with adverse repercussions for vulnerable events, and terrorism and security concerns, which
economies. Monetary policy tightening in the euro could derail growth.
area, if it had to comewhile the recovery in prices
and growth is still lagging in highly indebted mem- These risks are closely interconnected and can be
ber economies, could pose risks for these econo- mutually reinforcing. For example, an inward turn in
mies if they have not undertaken the needed fiscal policies could be associated with increased geopoliti-
adjustment and implemented structural reforms cal tensions as well as with rising global risk aversion;
to boost supply potential. Tighter global financial noneconomic shocks can weigh directly on economic
conditions could also result from a sharp decrease in activity while shaking confidence and market senti-
global risk appetite from its currently strong levels, ment; and a faster-than-anticipated tightening of global
financial conditions or a shift toward protectionism in to other countries, especially if it involves economies
advanced economies could exacerbate capital outflow with slack and monetary accommodation. Indeed,
pressures on emerging markets. adopting these policy recommendations would help
The welcome cyclical pickup in global economic reduce external imbalances, notably for advanced
activity after disappointing growth over the past few economies with excess surpluses, where stronger
years provides an ideal window of opportunity to domestic demand would offset negative demand
undertake key reforms designed to boost potential effects from the needed rebalancing by deficit
output and ensure that its benefits are broadly shared countries. In many emerging market and developing
and to build resilience against downside risks. With economies, fiscal space to support demand is lim-
countries still facing differences in cyclical conditions, ited, especially in commodity exporters. But mon-
varied stances of monetary and fiscal policy remain etary policy can generally be supportive, as inflation
appropriate. Completing the economic recovery and appears to have peaked in many countries. Exchange
adopting strategies to ensure fiscal sustainability remain rate flexibility helps with the adjustment to com-
important goals in many economies. modity price shocks. Efforts to improve governance
Important areas of strategic focus include: and the investment climate would also strengthen
Raising potential output: Structural reforms and growth prospects. In low-income countries, many of
growth-friendly fiscal policy are needed to boost which need to undertake durable fiscal adjustment
productivity and labor supply, with differing pri- efforts and reduce financial vulnerabilities, growth-
orities across countries. Looking ahead, ongoing enhancing reforms would help make the best use of
structural transformation (labor-saving technologi- the coming demographic dividend by spurring job
cal change and cross-border competition) demands creation.
comprehensive policy approaches, including policies Strengthening international cooperation: For many of
that reduce the pain of adjustment and provide the challenges that the global economy confronts,
opportunities for all. individual country actions can be more effective if
Securing the recovery and building resilience: In supported by multilateral cooperation. Preserving
advanced economies, monetary policy settings the global economic expansion will require policy-
should remain accommodative until there are firm makers to avoid protectionist measures and to do
signs of inflation returning to targets. As docu- more to ensure that gains from growth are shared
mented in Chapter 2, still-subdued wage pressures more widely. In addition to preserving an open trad-
mostly reflect remaining slack, not fully captured ing system, key areas for collective action include:
by headline unemployment rates. At the same time, safeguarding global financial stability; achieving
stretched asset valuations and increasing leverage in equitable tax systems and avoiding a race to the bot-
some parts of the financial sector require close mon- tom; continuing to support low-income countries as
itoring, with proactive micro- and macroprudential they pursue their development goals; and mitigating
supervision, as necessary. The stance of fiscal policy and adapting to climate change. As Chapter 3 illus-
should be aligned with structural reform efforts, trates, many of the economies suffering the worst
taking advantage of favorable cyclical conditions consequences of higher temperatures and changed
to place public debt on a sustainable path while weather patterns are those with the fewest resources
supporting demand where still needed and feasible. to deal with these challenges. Richer countries will
As Chapter 4 emphasizes, higher public spending increasingly feel direct negative effects from unmiti-
designed to boost potential output can result in gated climate change, however, and will not be
both domestic benefits as well as positive spillovers immune to spillovers from the rest of the world.
The pickup in growth projected in the April 2017 World pression of risk premiums and higher long-term interest
Economic Outlook (WEO) is strengthening. The global rates would expose fragilities, including by worsening
growth forecast for 2017 and 20183.6percent and public debt dynamics. Although progress has been made
3.7percent, respectivelyis 0.1percentage point higher in addressing European banking sector issues, remaining
in both years than in the April and July forecasts. Notable problems need to be addressed forcefully to avoid weak-
pickups in investment, trade, and industrial production, ening confidence and fears of adverse feedback loops
coupled with strengthening business and consumer confi- between low demand, prices, and balance sheets in parts
dence, are supporting the recovery. With growth outcomes of the euro area. Persistently low inflation in advanced
in the first half of 2017 generally stronger than expected, economies, which could ensue if domestic demand were
upward revisions to growth are broad based, including to falter, also carries significant risks, as it could lead to
for the euro area, Japan, China, emerging Europe, and lower medium-term inflation expectations and interest
Russia. These more than offset downward revisions for rates, reducing central banks capacity to cut real interest
the United States, the United Kingdom, and India. rates in an economic downturn. Although the chances
However, the recovery is not complete: although of advanced economy policies turning inward appear to
the baseline outlook is better, growth remains weak in have diminished in the near term, pressures for increased
many countries. The outlook for advanced economies protectionism have not disappeared and ought to be
has improved, notably for the euro area, but in many resisted. A host of noneconomic risks, including intensified
countries inflation remains weak, indicating that slack conflict and geopolitical tensions, also remain salient.
has yet to be eliminated, and prospects for growth in The welcome cyclical upturn after disappointing
GDP per capita are held back by weak productivity growth over the past few years provides an ideal window
growth and rising old-age dependency ratios. Prospects of opportunity to undertake critical reforms, thereby
for many emerging market and developing economies in staving off downside risks and raising potential output
sub-Saharan Africa, the Middle East, and Latin Amer- and standards of living more broadly. Structural reforms
ica are lackluster, with several experiencing stagnant per and growth-friendly fiscal policy measures are needed to
capita incomes. Fuel exporters are particularly hard hit by boost productivity and labor supply, with varying prior-
the protracted adjustment to lower commodity revenues. ities across countries. In advanced economies, monetary
Risks to the baseline are broadly balanced in the short policy should remain accommodative until there are
term but skewed to the downside in the medium term. firm signs of inflation returning to targets. At the same
Short-term growth could increase further, as stronger time, stretched asset valuations and increasing leverage in
confidence and favorable market conditions unleash some market segments bear close monitoring, including
pent-up demand, but setbacks are also possible. With high through proactive micro- and macroprudential supervi-
policy uncertainty, misstepswhich the baseline assumes sion, as necessary. Fiscal policy should be aligned with
will be avoidedor other shocks could materialize, structural reform efforts, taking advantage of favorable
taking a toll on market confidence and asset valuations, cyclical conditions to place public debt on a sustain-
and tightening financial conditions. Over the medium able path while supporting demand where still needed
term, dealing with financial sector challenges will be and feasible. In many emerging market and developing
essential. Minimizing the risk of a sharp slowdown in economies, fiscal space to support demand is limited,
China will require the Chinese authorities to intensify especially in commodity exporters. But monetary policy
their efforts to rein in the credit expansion. Many other can generally be supportive because inflation appears to
economies need to guard against a buildup of financial have peaked in many countries. Exchange rate flexibil-
stability risks in a global environment of easy finance and ity helps the adjustment to external shocks. Efforts to
monitor the risks from volatility as advanced economies improve governance and the investment climate would
central banks gradually withdraw stimulus. A decom- also strengthen growth prospects. Growth-enhancing
Figure 1.1. Global Activity Indicators reforms would help low-income countriesmany of which
Global activity strengthened in the rst half of 2017, reecting rmer domestic need to undertake durable fiscal adjustment efforts and
demand growth in advanced economies and China and improved performance in reduce financial vulnerabilitiesmake the best use of the
other large emerging market economies. Global manufacturing purchasing
managers indices indicate strong momentum continued into the third quarter.
coming demographic dividend by spurring job creation.
(CC only), Israel, Japan, Korea, New Zealand (PMI only), Norway (CC only), mance and a diminished pace of contraction in domes-
Singapore (PMI only), Sweden (CC only), Switzerland, Taiwan Province of China, tic demand allowed the economy to return to positive
United Kingdom, United States. growth in the first quarter of 2017, after eight quarters
2Argentina (CC only), Brazil, China, Colombia (CC only), Hungary, India (PMI only),
Indonesia, Latvia (CC only), Malaysia (PMI only), Mexico (PMI only), Philippines (CC of decline. Mexico maintained growth momentum,
only), Poland, Russia, South Africa, Thailand (CC only), Turkey, Ukraine (CC only). despite uncertainty related to the renegotiation of the
North American Free Trade Agreement and significant
Figure 1.2. Global Fixed Investment and Trade Figure 1.3. Commodity Prices
(Deated using US consumer price index; index, 2014 = 100)
Investment began to pick up in the third quarter of 2016. Global trade accelerated
as well, before moderating more recently. Commodity prices softened during the rst half of 2017.
0 Sources: IMF, Primary Commodity Price System; and IMF staff estimates.
China
3 Other countries2
World
6
2015:Q1 15:Q2 15:Q3 15:Q4 16:Q1 16:Q2 16:Q3 16:Q4 17:Q1 17:Q2 first quarter of 2018. The main drivers of lower
prices were higher-than-expected US shale produc-
Source: IMF staff calculations. tion and stronger-than-expected production recov-
1Data for 2017:Q2 are based on preliminary estimates for Russia.
2Other countries include Brazil, Canada, India, Korea, Mexico, Russia, South Africa, eries in Libya and Nigeria. In addition, exports
Taiwan, Turkey, and the United Kingdom. from OPEC countries remained at relatively high
levels, even with lower production. Following some
strengthening in recent weeks, oil prices stood at
tightening of monetary policy over the past two years. about $50 a barrel as of late August, still lower
Recovering domestic and external demand supported than in the spring.
rebounding growth in Russia and Turkey. Internal and The natural gas price indexan average for Europe,
cross-border conflict in parts of the Middle East still Japan, and the United Statesdecreased by 9.6per-
weighed on economic activity, while Venezuela faced cent from February to August 2017. The decline
a political and humanitarian crisis amid a deepen- was mostly tied to seasonal factors and robust supply
ing recession. from the United States and Russia, and lower oil
prices, which some natural gas prices are indexed
Softer Commodity Prices to. The diplomatic rift between Qatar, the worlds
The IMFs Primary Commodities Price Index largest exporter of liquefied natural gas, and several
declined by 5percent between Februaryand August other countries in the region, including Saudi Ara-
2017that is, between the reference periods for the bia, has not affected liquefied natural gas markets, as
April2017 WEO and the current report (Figure1.3). Qatars exports have continued.
Some of the biggest price drops were among fuels: The coal price indexan average of Australian and
Oil prices fell by 8.1percent between Febru- South African pricesincreased by 16.5percent
ary and August, even as the Organization of the between February and August 2017. Following
Petroleum Exporting Countries (OPEC) and some the end of the disruption to coal transportation in
non-OPEC oil exporters announced in May that Australia caused by Cyclone Debbie in late March,
they would extend oil production cuts through the coal prices declined until June. Strong demand from
year. Core inflation in the euro area has been stuck prices should, in turn, spur nominal wage growth in a
in low gear at about 1.2percent since April (after self-reinforcing dynamic.
hovering at just below 1percent for a couple of years), In many emerging market and developing econo-
while in Japan it remained slightly negative for six mies, the waning of pass-through effects from earlier
months through July. In the United Stateswhere exchange rate depreciations and, in some cases, recent
core inflation is higherthe annual change in the appreciations against the US dollar, have helped
core personal consumption expenditure deflator (the moderate core inflation rates. However, much of the
Federal Reserves preferred measure) declined from softening of core inflation in emerging market econo-
just below 2percent in early 2017 to 1.4percent in mies in recent months can be attributed to India and
August. This decline in part reflected one-off factors Brazil, where a one-off drop in food price inflation in
(including a reduction in prices of cell phone plans June and high excess capacity in the economy after two
and prescription drug prices). Many other advanced years of recession, respectively, have also contributed
economies, including Australia, Canada, Denmark, to weaker inflation. In China, core inflation remained
Korea, Norway, and especially Taiwan Province of broadly stable at about 2percent in July. In con-
China, are also experiencing weak inflation pressure. trast, some other countries in the Commonwealth of
The United Kingdom, where the strong depreciation Independent States and the Middle East, North Africa,
of the pound since last summer has passed through Afghanistan, and Pakistan region are experiencing
into higher consumer prices, is an exception to continued inflationary pressures in 2017 as a result of
this pattern. exchange rate depreciations, the removal of subsidies,
Sluggishness in core inflation in advanced or increases in excise or value-added taxes.
economiesa surprise in view of stronger-than-
expected activityhas coincided with slow trans- Supportive Financial Conditions
mission of declining unemployment rates into faster Market sentiment has remained strong and vol-
wage growth. Real wages in most large advanced atility low since the publication of the most recent
economies have moved broadly with labor productiv- (April 2017) WEO, even as expectations of US fiscal
ity in recent years, as indicated by flat labor income easing have dimmed. On the monetary policy front,
shares (Figure1.4, panel 6). As shown in Chapter2, the USFederal Reserve raised short-term interest rates
muted growth in nominal wages in recent years partly in June to 11.25percent, as expected. Following
reflects sluggishness in labor productivity.1 However, the Federal Open Market Committee announcement
the analysis also reveals continued spare capacity in of September 20, markets priced in a 70percent
labor markets as a key drag: wage growth has been probability of one additional rate increase by the
particularly soft where unemployment and the share of end of2017. In most other advanced economies, the
workers involuntarily working part time remain high. monetary policy stance remained broadly unchanged,
The corollary of this finding is that, once firms and except for Canada, which raised its policy rate by of
workers become more confident in the outlook, and a percentage point in July and September.
labor markets tighten, wages should accelerate. In the With markets pricing in a slightly more gradual nor-
short term, higher wages should feed into higher unit malization of USmonetary policy than anticipated in
labor costs (unless productivity picks up), and higher the spring, given diminished expectations about fiscal
stimulus, nominal yields on 10-year US Treasury bonds
1The part of the wage-inflation weakening attributable to lower as of mid-September have declined by about 20 basis
productivity growth would likely have little or no pass-through into points from their March 2017 average (Figure1.5).
weaker price inflation, given that the changes would have no net effect
on conventionally measured unit labor costs. A broad slowdown in
Long-term sovereign bond yields have remained
total factor productivity and an interrelated decline in capital accu- broadly stable in Japan and Germany, risen by some
mulation have been the drivers of the slowdown in labor productivity 10 basis points in the United Kingdom, and declined
(Adler and others 2017). Shifts in the composition of the labor force
by 2030 basis points in France, Italy, and Spain, as
since the global financial crisis may also have exerted downward
pressure on productivity and wages. These shifts include the expanded spreads relative to German bund yields compressed
shares of female and older workers, whose participation rates have sharply, particularly in the aftermath of the French
generally risen (Box1.1). New entrants tend to be paid less than exist- presidential election. Equity markets in advanced econ-
ing workers (Daly, Hobijn, and Pedtke 2017). A larger share of older
workers has also been linked to slower productivity growth (Feyrer omies have continued to rise in recent months amid
2007; Aiyar, Ebeke, and Shao 2016; Adler and others 2017). strong earnings, further improvements in consumer
Figure 1.5. Advanced Economies: Monetary and Financial and business confidence, and favorable macroeconomic
Market Conditions data. Market volatility indicators remain low.
(Percent, unless noted otherwise)
With narrowing interest differentials, the USdollar
Market sentiment has been strong in advanced economies. Compared with the weakened in real effective terms by over 7percent from
spring, a more gradual normalization of US monetary policy is anticipated and March to mid-September 2017 (Figure1.6, panel 1),
credit spreads remain compressed.
more than reversing its gains after the US election,
4.0 1. US Policy Rate 2. Policy Rate Expectations 1 4.5 whereas the euro and the Canadian dollar appreciated
Expectations1 (Percent; dashed lines 4.0 by 6percent on stronger growth prospects and higher
3.5 are from the April
Sep. 16, 2015 3.5 policy rates in Canada. Among other currencies, the
3.0 2017 WEO)
Sep. 16, 2016 3.0
2.5 United States yen depreciated by about 3percent and the Swiss franc
Nov. 8, 2016 2.5
Euro area and Korean won by 4percent.
2.0 Apr. 3, 2017 2.0
United Kingdom
Sep. 15, 2017 In emerging market economies, financial conditions
1.5 1.5
1.0 since March generally have been supportive of a pickup
1.0
0.5 in economic activity. Equity markets have strengthened
0.5 0.0 (Figure1.7); long-term interest rates on local-currency
0.0 0.5
2015 16 17 18 19 Sep. 2017 18 19 Sep. bonds have generally declined (Figure1.8), China being
20 20 the exception; and spreads on the Global Emerging
3. Key Interest Rates 2 4. Credit Spreads3 Markets Bond Index have fallen slightly. As search
6 1,200
US average 30-year (Basis points) for yield continues (Chapter1 of the October Global
5 1,000
xed-rate mortgage US high grade Financial Stability Report [GFSR]), emerging market
4 US high yield
Euro high grade 800 currencies have generally strengthened relative to the US
3 United States Euro high dollar. As of August 2017, changes since March in real
yield 600
2 effective terms have generally been moderate (Figure1.6,
1 Germany 400 panel 2). The Mexican peso appreciated by 10percent
Japan 200
on tighter monetary policy and declining concerns about
0
trade-related frictions with the United States, while the
1 0 South African rand depreciated by 7percent on domes-
2013 14 15 16 Sep. 2013 14 15 16 Sep.
17 17 tic political uncertainty, the Brazilian real depreciated by
over 4percent on monetary policy easing and concerns
200 5. Equity Markets 6. Price-to-Earnings Ratios3 35
180 (Index, 2007 = 100) about the reform agenda, and the Russian ruble depreci-
United States
160 S&P 500 Japan 30 ated by a similar amount on weakening oil prices.
140 Germany Capital flows to emerging market economies have
120 Italy 25 remained resilient in recent months, continuing their
100
recovery after a sharp decline in late 2015 and early
80 20
60 2016. As discussed in Box1.2, this pattern reflects a
MSCI Emerging Market pickup in capital flows to China and a strong global
40 15
Euro Stoxx
20 TOPIX recovery in nonresident portfolio inflows in the first
0 10 half of 2017 (Figure1.9, panel 1) as investor optimism
2013 14 15 16 Aug. 2013 14 15 16 Sep.
17 17 about the global economic outlook improved and
financial conditions eased.
Sources: Bloomberg L.P.; Thomson Reuters Datastream; and IMF staff calculations.
Note: MSCI = Morgan Stanley Capital International; S&P = Standard & Poors;
TOPIX = Tokyo Stock Price Index; WEO = World Economic Outlook.
1Expectations are based on the federal funds rate futures for the United States, the
Key Forces Shaping the Outlook
sterling overnight interbank average rate for the United Kingdom, and the euro Continued Cyclical Recovery in Advanced Economies
interbank offered forward rate for the euro area; updated September 15, 2017.
2Interest rates are 10-year government bond yields, unless noted otherwise. Data
(and Revisions to Potential Output)
are through September 15, 2017. In advanced economies, the ongoing cyclical
3Data are through September 15, 2017.
recovery is stronger than previously projected. Indeed,
positive surprises in growth in the first half of 2017
typically occurred in countries where estimates for
output were below potential in 2016 (Figure1.10,
Figure 1.6. Real Effective Exchange Rate Changes, Figure 1.7. Emerging Market Economies: Equity Markets and
November 2016September 2017 Credit
(Percent)
Equity indices in emerging market economies have risen since the spring and
In real effective terms, the US dollar weakened by about 7 percent and the euro credit growth remains supportive of a pickup in activity.
strengthened by 6 percent from March to August 2017. Changes in most emerging
market currencies have been moderate.
200 1. Equity Markets Emerging Asia Latin America
(Index, 2007 = 100) excluding China
180
Latest relative to August 2017
August 2017 relative to March 2017 160
March 2017 relative to November 8, 2016 140
120
8 1. Advanced Economies 1
100
6
80
4
60 Emerging Europe China
2
0 40
2012 13 14 15 16 Aug.
2 17
4
Real Credit Growth1
6 (Year-over-year percent change)
8
10 25 2. 3. 40
BRA CHN COL IDN
USA EA JPN GBR SWE CHE KOR TWN SGP CAN NOR AUS NZL IND MEX MYS RUS
20
TUR 30
15
16 2. Emerging Market Economies1
10 20
12
5
8 0 10
4 5
0
0 10
4 15 10
2012 13 14 15 16 Jul. 2012 13 14 15 16 Jun
8 17 17
12 Credit-to-GDP Ratio1
ZAF CHN IND IDN MYS PHL THA HUN POL RUS TUR BRA CHL COL MEX PER (Percent)
85 4. 230 5. 30
Source: IMF staff calculations. MEX (right scale)
Note: EA = euro area. Data labels in the gure use International Organization for 75 BRA COL 210 CHN 25
Standardization (ISO) country codes. IDN IND MYS
1Latest data available are for September 15, 2017. 65 190
RUS TUR 20
55 170
15
45 150
10
panel1). With growth generally above potential out- 35 130
put, economic slack is gradually being reduced. 25 110 5
Positive revisions to growth have also come with some
15 90 0
upward revisions to the estimated path of potential out- 2006 08 10 12 14 17: 2006 08 10 12 14 17:
Q2 Q2
put. Indeed, despite an upward revision to the cumula-
tive growth rate over 201618 relative to the October Sources: Bloomberg L.P.; Haver Analytics; IMF, International Financial Statistics
2016 WEO forecast of about 0.7percentage point, the (IFS) database; and IMF staff calculations.
forecast of the output gap for 2018 has been revised in Note: Data labels in the gure use International Organization for Standardization
(ISO) country codes.
absolute terms by only half as much. As Figure1.10, 1
Credit is other depository corporations claims on the private sector (from IFS),
panel 2 shows, the upward revision to growth exceeds except in the case of Brazil, for which private sector credit is from the Monetary
Policy and Financial System Credit Operations published by Banco Central do
the decline in the output gap for most individual Brasil, and China, for which credit is total social nancing after adjusting for local
countries. The difference is explained by slightly higher government debt swaps.
projected potential growth during this period (about
Figure 1.8. Emerging Market Economies: Interest Rates Figure 1.9. Emerging Market Economies: Capital Flows
Long maturity yields on local currency debt have generally declined. Capital ows to emerging market economies continued to recover.
0.1percentage point a year), driven by higher projected Figure 1.10. Revisions to 2017 Growth and 2016 Output Gaps
(Percent)
investment, which boosts productive capacity.
With output in 2017 remaining slightly below The ongoing cyclical recovery is stronger than previously projected, with positive
potential for the advanced economies group, the cycli- growth surprises in the rst half of 2017 typically occurring in countries with
output below estimated potential in 2016.
cal recovery still has some room to run. This assess-
ment is consistent with still elevated unemployment
2.0 1. 2016 Output Gaps and 2017 Growth Surprises
rates in a few countries and relatively high shares of LVA
workers who would prefer to work full time but can 1.6
EST SWE
largest countries, China and India, which account for discussed in the April 2017 Fiscal Monitor. So far,
more than 40percent of GDP (whether measured at exchange rate flexibility has helped the adjustment
purchasing power parity or market rates) and more than countries that allowed greater exchange rate flexibility
40percent of the population of emerging market and have drawn less on their buffers (Box1.4).
developing economies.2 Indeed, the forecast for growth Looking ahead, growth in commodity exporters is
in GDP per capita falls below the groups aggregate fig- forecast to recover further, contributing significantly to
ure of 3.5percent for about of emerging market and the projected pickup in global growth between 2016
developing economies. And for 43 economies (28per- and 2022 (the last year of the WEO forecast hori-
cent of the total), per capita growth rates are projected zon) (Figure1.12, panels 23). Nevertheless, growth
to be lower than for advanced economies, implying a in commodity exporters is projected to remain well
decline in relative living standards rather than conver- below its historical average and will account for only
gence. Box1.3 also shows that very small economies a modest share of total growth for emerging market
(with populations of less than 500,000 people) and fuel and developing economies as a group (Figure1.12,
exporters are overrepresented among the economies with panel 1). In contrast, growth is projected to remain
weak projected growth. high for the group of commodity-importing countries,
The challenges faced by very small economies, which account for the lions share of global growth,
related to such factors as diseconomies of scale, lack of with higher growth in India and other commodity
diversification, and the frequency of natural disasters, importers more than offsetting a slowdown in China.
are well documented.3 As also highlighted in previous A similar pattern is at play for low-income developing
WEOs, many commodity exportersespecially fuel countries, where growth in commodity importers is
exportersare still struggling to adjust to sharply forecast to exceed that in commodity exporters (Fig-
lower commodity prices relative to those prevailing ear- ure1.12, panel 4).
lier in the decade.
macroeconomic implications of changes in weather patterns for 4The fiscal impulse is defined as the change in the structural fiscal
Figure 1.11. Emerging Markets: Terms-of-Trade Windfall Figure 1.12. GDP Growth, 19992022
Gains and Losses (Percent)
Commodity terms-of-trade shifts imply relatively small projected gains and losses While commodity exporters are projected to grow at rates well below their
in disposable income when compared with the very large losses for commodity historical averages, they are nevertheless expected to contribute signicantly to
exporters during 201516. the projected global growth pickup between 2016 and 2022.
4
40 1999 2001 03 05 07 09 11 13 15 17 19 22
SAU DZA IRN COL AUS MEX ARG
KAZ NGA RUS MYS CAN BRA IDN
Advanced economies Noncommodity exporters excluding China,
India India, and Brazil
8 2. Terms-of-Trade
Terms Windfall Gains1
China Commodity exporters
(Percent of GDP)
2 3 3
0 2 2
2 1 1
USA FRA DEU ITA JPN IND PAK
EGY TUR POL ESP CHN THA KOR
0 0
2016 2022 2016 2022
Source: IMF staff estimates.
Note: Data labels in the gure use International Organization for Standardization
(ISO) country codes. WEO = World Economic Outlook. 18 4. Growth in Low-Income Developing Countries
1
Gains (losses) for 201718 are simple averages of annual incremental gains 16 Nigeria
(losses) for 2017 and 2018. The windfall is an estimate of the change in 14 Others
disposable income arising from commodity price changes. The windfall gain in Commodity exporters excluding Nigeria and Yemen
12
year t for a country exporting x US dollars of commodity A and importing m US
10
dollars of commodity B in year t 1 is dened as (ptAxt 1 ptBmt 1) / Yt 1, in
which ptA and ptB are the percentage changes in the prices of A and B between 8
year t 1 and year t, and Y is GDP in year t 1 in US dollars. See also Gruss (2014). 6
4
2
0
then-anticipated corporate and personal income tax 2
1999 2001 03 05 07 09 11 13 15 17 19 22
reductions). In emerging market and developing econ-
omies, fiscal policy is expected to be broadly neutral in Source: IMF staff estimates.
both 2017 and 2018. (The projected looser fiscal pol- Note: Commodity exporters includes fuel and nonfuel primary products exporters,
icy for the group in 2018 relative to the assumptions as indicated in Table D of the Statistical Appendix, plus Brazil and Peru. EMDEs =
emerging market and developing economies; PPP = purchasing power parity.
in April primarily reflects downward revisions for the
structural fiscal balances of Brazil and China).
On monetary policy, the forecast assumes a some-
what more gradual normalization of the policy interest
rate in the United States than projected in the April
2017 WEO. With US fiscal policy now set to be
Figure 1.13. Fiscal Indicators broadly neutral in 2017 and projected to tighten in
(Percent of GDP, unless noted otherwise)
2018, monetary policy is projected to be moderately
The projected overall neutral scal policy stance for 2017 and 2018 masks
more accommodative than previously expected, given
variation across countries. weaker projected demand and diminished inflation
pressure. The USpolicy interest rate is projected to
2 1. Change in the Structural Fiscal Balance remain broadly unchanged at 100125 basis points
(Percentage points) 2013 2014 2015 for the rest of 2017 and rise by about 75 basis points
1 2016 2017 2018
April 2017 WEO in 2018, reaching a long-term equilibrium rate of
slightly less than 3percent in2020. In the euro
0
area and Japan, the forecast assumes that monetary
1 policy will remain very accommodative. Short-term
rates are projected to remain negative in the euro
2 area through2018 and close to zero in Japan over
Advanced Emerging market and
economies developing economies the forecast horizon. The assumed monetary pol-
icy stances across emerging market economies vary,
3 2. Change in the Structural Fiscal Balance reflecting these economies diverse cyclical positions.
(Percentage points)
2013 2014 2015 Given faster-than-expected declines in inflation rates
2 2016 2017 2018 in many larger economies, such as Brazil, India, and
April 2017 WEO
Russia, the projected level of monetary policy inter-
1
est rates for the group is somewhat lower than in the
0 April 2017 WEO.
Global financial conditions are assumed to remain
1
United States Japan1 France, Germany, Greece, Ireland,
accommodative, in line with the April projections. As
United Kingdom Italy, Portugal, Spain discussed in Chapter1 of the October 2017 GFSR,
an easing of lending conditions in major economies
2 3. Fiscal Balance is expected to offset the anticipated gradual rise in
0 long-term interest rates, while the normalization of
2
monetary policy in the United States and the United
Kingdom is expected to proceed smoothly, without
4
triggering large and protracted increases in financial
6
World market volatility. Except for several vulnerable econ-
Emerging market and
8 Advanced economies omies, most emerging markets are expected to face
developing economies
10
2001 04 07 10 13 16 19 22
generally accommodative financial conditions, with
higher policy rates partially offset by a recovery in risk
180 4. Gross Public Debt
appetite, as reflected in generally contained sovereign
160 World Advanced economies 2 bond spreads and the uptick in most equity markets.
Major advanced Emerging and Latin America Despite the recent decline in commodity prices, the
140
economies2,3 developing Asia and the
120 Other emerging market and Caribbean IMFs commodity price index is expected to increase
100 developing economies by 12.3percent in 2017 from its average in 2016, and
80
then fall slightly again in 2018, by 0.1percent. After
60
40
averaging $43 a barrel in 2016, oil prices are expected
20 to average $50.3 a barrel in2017 (down from $55.2 a
1950 60 70 80 90 2000 10 22 barrel in the April 2017 WEO), and stay at about that
level in 2018. Nonfuel commodity prices are expected
Source: IMF staff estimates.
Note: WEO = World Economic Outlook. to strengthen in201718 from their2016 averages
1Japan's latest gures reect comprehensive methodological revisions adopted in because of stronger demand for metals from China,
December 2016. tight supply conditions for food, and a general pickup
2Data through 2000 exclude the United States.
3Canada, France, Germany, Italy, Japan, United Kingdom, United States. in global demand.
Looking further ahead, futures markets point
toward a slight rise in commodity prices by 2022.
While energy prices are expected to increase modestly across emerging market and developing economies
because of growing demand in emerging markets, food remain heterogeneous, with emerging Asian countries
prices are expected to fall moderately as some supply generally growing at a fast pace, but many countries
disruptions wane. in Latin America, sub-Saharan Africa, and the Middle
Finally, against a backdrop of elevated policy uncer- East struggling with subpar performance.
tainty, the forecast rests on the assumption that major
policy missteps are avoided. For instance, negotiations
Growth Outlook for the Medium Term
on the future economic relations between the United
Kingdom and the European Union (EU) are assumed Global growth is forecast to increase marginally
to proceed without raising excessive uncertainty, and beyond2018, reaching 3.8percent by2021. With
the arrangements are expected to eventually settle in growth in advanced economies projected to gradually
a manner that avoids a very large increase in eco- decline toward potential growth rates of about 1.7per-
nomic barriers. cent once economic slack is eliminated, this further
pickup in global activity is entirely driven by emerging
market and developing economies. In these countries,
Global Outlook for201718 growth is projected to increase to 5percent by the
World growth is projected to increase from 3.2per- end of the forecast period, with their impact on global
cent in 2016 to 3.6percent in2017 and 3.7percent activity boosted by their rising world economic weight.
in2018an upward revision of 0.1percentage point This forecast assumes some strengthening of growth
forboth 2017 and 2018 relative to April. Economic in commodity exporters, though to rates much more
activity is projected to pick up speed in all country modest than in200015; a gradual increase in Indias
groups except for the Middle East, and forecasts of growth rate resulting from implementation of import-
the strength of the outlook by region have changed ant structural reforms; continued strong growth in
only modestly (Table 1.1). other commodity importers; and a lower but still high
In line with a stronger-than-expected rise in growth trend growth rate in China (Figure1.12, panels 13).
in advanced economies so far in 2017 (especially in the
euro area), their projected growth rate has been revised Growth Outlook for Individual Countries and Regions
upward to 2.2percent for 2017 (from 2percent pro-
jected in April)a notable increase from 1.7percent Advanced Economies
in 2016. The advanced economy forecast for 2018 is The USeconomy is projected to expand at 2.2per-
unchanged, with lower projected US growth (under cent in 2017 and2.3percent in 2018. The projec-
the assumption that fiscal policy will not provide tion of a continuation of near-term growth that is
the previously envisaged boost to demand) offsetting moderately above potential reflects very supportive
higher projected growth in the euro area. financial conditions and strong business and con-
Growth is forecast to increase strongly in emerging sumer confidence. The downward revision relative
market and developing economies, from an upwardly to the April WEO forecasts (of 2.3 and 2.5percent
revised 4.3percent in 2016 to 4.6percent in 2017 and for 2017 and 2018, respectively) reflects a major
4.9percent in 2018, a 0.1percentage point increase correction in US fiscal policy assumptions. Given
for 2017 and 2018 relative to the April forecast. The the significant policy uncertainty, IMF staff s macro-
upward revisions to the growth forecast primarily economic forecast now uses a baseline assumption of
reflect stronger projected activity in China and in unchanged policies, whereas the April 2017 WEO
emerging Europe for 2017 and 2018. built in a fiscal stimulus from anticipated tax cuts.
As discussed earlier, although commodity importers Over a longer horizon, US growth is expected to
account for the lions share of growth in emerging mar- moderate. Potential growth is estimated at 1.8per-
ket and developing economies, the projected increase cent, reflecting the assumption of continued sluggish
in growth from 2016 is driven primarily by stronger growth in total factor productivity and diminished
projected growth for commodity exporters, most growth of the workforce due to population aging.
notably Brazil and Russia, that experienced severe mac- The euro area recovery is expected to gather strength
roeconomic strains during 201516. As emphasized this year, with growth projected to rise to 2.1per-
in previous WEO reports and in Box1.3, prospects cent in 2017, before moderating to 1.9percent in
switch from the System of National Accounts 1993 to the System of National Accounts 2008 and the updating of the benchmark year from 2005 to 2011.
3Excludes the Group of Seven (Canada, France, Germany, Italy, Japan, United Kingdom, United States) and euro area countries.
4For India, data and forecasts are presented on a fiscal year basis and GDP from 2011 onward is based on GDP at market prices with fiscal year 2011/12 as a
base year.
14 International Monetary Fund | October 2017
CHAPTER 1 Global Prospects and Policies
Year-over-Year Q4-over-Q48
Projections Projections
2015 2016 2017 2018 2015 2016 2017 2018
World Output 3.4 3.2 3.6 3.7 3.2 3.2 3.7 3.7
Advanced Economies 2.2 1.7 2.2 2.0 1.9 2.0 2.2 1.9
United States 2.9 1.5 2.2 2.3 2.0 1.8 2.3 2.3
Euro Area 2.0 1.8 2.1 1.9 1.9 1.9 2.2 1.7
Germany 1.5 1.9 2.0 1.8 1.3 1.9 2.2 1.8
France 1.1 1.2 1.6 1.8 1.0 1.2 2.1 1.4
Italy 0.8 0.9 1.5 1.1 1.0 1.2 1.5 1.0
Spain 3.2 3.2 3.1 2.5 3.5 3.0 3.1 2.1
Japan2 1.1 1.0 1.5 0.7 1.1 1.7 1.4 0.5
United Kingdom 2.2 1.8 1.7 1.5 1.7 1.9 1.3 1.5
Canada 0.9 1.5 3.0 2.1 0.4 2.0 3.0 2.0
Other Advanced Economies3 2.1 2.2 2.6 2.5 2.0 2.5 2.5 2.6
Emerging Market and Developing Economies 4.3 4.3 4.6 4.9 4.4 4.2 5.0 5.2
Commonwealth of Independent States 2.2 0.4 2.1 2.1 2.8 0.6 1.9 2.2
Russia 2.8 0.2 1.8 1.6 3.3 0.3 1.9 2.0
Excluding Russia 0.6 1.9 2.9 3.3 ... ... ... ...
Emerging and Developing Asia 6.8 6.4 6.5 6.5 6.9 6.2 6.6 6.5
China 6.9 6.7 6.8 6.5 6.8 6.8 6.5 6.5
India4 8.0 7.1 6.7 7.4 8.9 5.6 7.9 7.4
ASEAN-55 4.9 4.9 5.2 5.2 4.9 4.8 5.3 5.2
Emerging and Developing Europe 4.7 3.1 4.5 3.5 4.8 3.8 2.6 4.7
Latin America and the Caribbean 0.1 0.9 1.2 1.9 1.3 1.1 1.7 2.0
Brazil 3.8 3.6 0.7 1.5 5.8 2.5 1.9 1.8
Mexico 2.6 2.3 2.1 1.9 2.5 2.3 1.0 3.2
Middle East, North Africa, Afghanistan, and Pakistan 2.7 5.0 2.6 3.5 ... ... ... ...
Saudi Arabia 4.1 1.7 0.1 1.1 4.3 2.2 0.6 1.4
Sub-Saharan Africa 3.4 1.4 2.6 3.4 ... ... ... ...
Nigeria 2.7 1.6 0.8 1.9 ... ... ... ...
South Africa 1.3 0.3 0.7 1.1 0.3 0.4 1.1 0.8
Memorandum
European Union 2.3 2.0 2.3 2.1 2.3 2.1 2.2 2.0
Low-Income Developing Countries 4.7 3.6 4.6 5.2 ... ... ... ...
Middle East and North Africa 2.6 5.1 2.2 3.2 ... ... ... ...
World Growth Based on Market Exchange Rates 2.7 2.5 3.0 3.1 2.4 2.6 3.1 3.0
World Trade Volume (goods and services) 2.8 2.4 4.2 4.0 ... ... ... ...
Imports
Advanced Economies 4.6 2.7 4.0 3.8 ... ... ... ...
Emerging Market and Developing Economies 0.9 2.0 4.4 4.9 ... ... ... ...
Exports
Advanced Economies 3.8 2.2 3.8 3.6 ... ... ... ...
Emerging Market and Developing Economies 1.8 2.5 4.8 4.5 ... ... ... ...
Commodity Prices (US dollars)
Oil6 47.2 15.7 17.4 0.2 43.4 16.2 1.4 1.1
Nonfuel (average based on world commodity export
weights) 17.5 1.8 7.1 0.5 19.1 9.9 3.1 0.6
Consumer Prices
Advanced Economies 0.3 0.8 1.7 1.7 0.4 1.2 1.5 1.9
Emerging Market and Developing Economies7 4.7 4.3 4.2 4.4 4.6 3.7 3.9 3.7
London Interbank Offered Rate (percent)
On US Dollar Deposits (six month) 0.5 1.1 1.4 1.9 ... ... ... ...
On Euro Deposits (three month) 0.0 0.3 0.3 0.3 ... ... ... ...
On Japanese Yen Deposits (six month) 0.1 0.0 0.1 0.2 ... ... ... ...
5Indonesia, Malaysia, Philippines, Thailand, Vietnam.
6Simple average of prices of UK Brent, Dubai Fateh, and West Texas Intermediate crude oil. The average price of oil in US dollars a barrel was $42.84 in
2016; the assumed price based on futures markets is $50.28 in 2017 and $50.17 in 2018.
7Excludes Argentina and Venezuela. See country-specific notes for Argentina and Venezuela in the Country Notes section of the Statistical Appendix.
8For World Output, the quarterly estimates and projections account for approximately 90 percent of annual world output at purchasing-power-parity weights.
For Emerging Market and Developing Economies, the quarterly estimates and projections account for approximately 80 percent of annual emerging market
and developing economies output at purchasing-power-parity weights.
International Monetary Fund | October 2017 15
WORLD ECONOMIC OUTLOOK: Seeking Sustainable GrowthShort-Term Recovery, Long-Term Challenges
2018 (slightly stronger than the 1.8percent growth In most other advanced economies, the pace of
estimated for 2016). The forecast is 0.4percentage activity is expected to accelerate.
point and 0.3percentage point higher for 2017 and oo Growth in oil-exporting advanced economies is
2018, respectively, relative to April. The increase projected to recover. In2017, it is forecast to rise
in growth in 2017 mostly reflects an acceleration to 1.4percent in Norway, and increase (by about
in exports in the context of the broader pickup 1 percentage points) to 3.0percent in Canada.
in global trade and continued strength in domes- This growth pickup reflects reduced drag from
tic demand growth supported by accommodative the adjustment to lower oil and gas prices and
financial conditions amid diminished political risk accommodative fiscal and monetary policies. By
and policy uncertainty. Growth is forecast to pick contrast, growth is expected to soften temporarily
up this year and moderate next year in Germany to 2.2percent in Australia, where housing invest-
(2.0percent in2017 and 1.8percent in2018), hold ment and mining exports in the first half of the
steady this year and moderate next year in Spain year were undermined by bad weather.
(3.1percent in2017 and 2.5percent in2018), rise oo A pickup in growth for2017 is projected in Korea
this year and next in France (1.6percent in 2017 (to 3.0percent), Hong Kong Special Administrative
and 1.8percent in 2018), and increase this year Region (to 3.5percent), Taiwan Province of China
and soften next year in Italy (1.5percent in2017 (to 2.0percent), and Singapore (to 2.5percent).
and1.1percent in 2018). The medium-term A common driver behind this projected pickup
outlook for the euro area remains subdued because (which is generally stronger than projected in the
projected potential growth is held back by weak April 2017 WEO) is the recovery in global trade
productivity, adverse demographics and in some and Chinas import demand.
countries, a public and private debt overhang.
Growth in the United Kingdom is projected to sub- Emerging Market and Developing Economies
side to 1.7percent in2017 and 1.5percent in2018. In China, growth is projected to notch up to
The 0.3percentage point downward revision to 6.8percent in 2017, and to slow to 6.5percent
the2017 forecast relative to the April 2017 WEO in 2018. The upward revision to the 2017 fore-
is driven by weaker-than-expected growth outturns cast0.2percentage point relative to the April
for the first two quarters of the year. The slowdown 2017 WEOreflects the stronger-than-expected
is driven by softer growth in private consumption outturn in the first half of the year underpinned by
as the pounds depreciation weighed on household previous policy easing and supply-side reforms. For
real income. The medium-term growth outlook is 2018, the upward revision of 0.3percentage point
highly uncertain and will depend in part on the new mainly reflects an expectation that the authorities
economic relationship with the EU and the extent will maintain a sufficiently expansionary policy mix
of the increase in barriers to trade, migration, and (especially through high public investment) to meet
cross-border financial activity. their target of doubling real GDP between 2010
In Japan, momentum is driven by the strengthening and 2020. Growth rates for 201922 have similarly
of global demand and policy actions to sustain a been revised upward by 0.2percentage point, on
supportive fiscal stance, and is expected to con- average, reflecting the assumed delay in withdrawing
tinue in2017, with growth forecast at 1.5percent. stimulus. Delay comes at the cost of further large
The pace of expansion is expected to weaken increases in debt, however, so downside risks around
thereafter (to 0.7percent in 2018), based on the this baseline have also increased.
assumption that fiscal support fades as currently In the rest of emerging market and developing Asia,
scheduled, private consumption growth moderates, growth is expected to be vigorous and marginally
and the boost from 2020 Olympics-related private higher than in the April 2017 WEO. Strong gov-
investment is offset by higher imports and slower ernment spending and data revisions in India led to
projected growth in foreign demand. Over the an upward revision of 2016 growth to 7.1percent
medium term, a shrinking Japanese labor force will (6.8percent in April), with upward revisions of
curtail GDP growth although, in per capita income about 0.2percentage point, on average, for 2014
terms, Japans growth is projected to remain close to and 2015. However, the growth projection for 2017
recent averages. has been revised down to 6.7percent (7.2percent in
April), reflecting still lingering disruptions associated ical and policy uncertainty led to a downward
with the currency exchange initiative introduced in revision of the 2018 forecast of 0.2percentage
November 2016, as well as transition costs related to point. A gradual restoration of confidenceas
the launch of the national Goods and Services Tax key reforms to ensure fiscal sustainability are
in July 2017. The latter move, which promises the implemented over timeis projected to raise
unification of Indias vast domestic market, is among growth to 2percent in the medium term.
several key structural reforms under implementa- oo In Argentina, growth is projected to rebound to
tion that are expected to help push growth above 2.5percent in 2017 from last years recession as
8percent in the medium term. In the ASEAN-5 higher real wages boost consumption; investment
economies (Indonesia, Malaysia, Philippines, Thai- picks up, supported by public works; and exports
land, Vietnam), growth is expected to strengthen in benefit from stronger external demand. Growth
2017 to 5.2percent (from 5percent in April), partly is expected to remain about 2 percent in 2018,
because of stronger-than-expected external demand as private domestic demand continues to improve
from China and Europe. Specifically, economic gradually against the backdrop of tight macro-
activity in 2017 is projected to expand by 5.2per- economic policy settings (high real interest rates
cent in Indonesia, 5.4percent in Malaysia, 6.6per- required by the disinflation process and the start
cent in the Philippines, 3.7percent in Thailand, and of the fiscal consolidation). The intensification
6.3percent in Vietnam. of the political crisis in Venezuela weighs heav-
In Latin America and the Caribbean, where GDP ily on economic activity, which is expected to
contracted by almost 1percent in 2016, real GDP contract by more than 10percent in 2017 as oil
is projected to increase by 1.2percent in 2017 and production declines and uncertainty rises further.
1.9percent in 2018broadly as in the April 2017 In Chile, growth is projected to be 1.4percent in
WEO. Although growth is holding up well in Cen- 2017 amid weakness in private fixed investment,
tral America and strengthening, on average, in the mining output, and public consumption, and
Caribbean, domestic demand continues to under- to recover to 2.5percent in 2018 amid growing
perform in much of the rest of the region, and some confidence, higher copper prices, and interest
idiosyncratic factors are playing a key role in shaping rate cuts implemented over the past few months.
substantially different outlooks across countries. In Colombia, growth is projected to be 1.7per-
oo In Mexico, growth is expected to soften to cent in 2017, amid continued adjustment to
2.1percent in 2017 and 1.9percent in 2018. lower revenues. Higher infrastructure spending,
Despite the uncertainty related to renegotiation investment-friendly tax reform, and the boost in
of the North American Free Trade Agreement and confidence from the peace agreement are expected
a downward revision to economic activity in the to raise growth to about 3.5percent in the
United States, growth for 2017 has been revised medium term.
upward by 0.4percent since the April 2017 The outlook for the Commonwealth of Indepen-
WEO, reflecting better-than-expected growth dent States continues to improve, following a deep
outturns for the first two quarters of the year and recession in 2015 and very shallow growth in 2016,
a recovery in financial market confidence. In the with growth projected at 2.1percent in 2017 and
medium term, the assumed full implementation 2018an upward revision of 0.4percentage point
of the structural reform agenda is projected to lift for 2017 relative to the April 2017 WEO. After two
growth to 2.7percent. years of recession, economic activity in Russia is pro-
oo After entering positive territory in the first half jected to expand by 1.8percent in 2017, helped by
of 2017, growth in Brazil is expected to reach stabilizing oil prices, easing financial conditions, and
0.7percent for the year and 1.5percent in 2018. improved confidence. Over the medium term, how-
A bumper crop and a boost to consumption, ever, growth is expected to remain about 1.5per-
including from allowing workers to draw on sav- cent, constrained by moderate oil prices, adverse
ings accumulated in their severance accounts, led demographics, and other structural impediments.
to an upward revision of half a percentage point Among other oil exporters, growth in Kazakhstan is
in 2017 relative to the April forecast, but ongoing projected to rise to 3.3percent in 2017 on the back
weakness in investment and an increase in polit- of strong oil production.
In emerging and developing Europe, short-term on the back of a slowdown in the Islamic Republic
growth has been revised upward to 4.5percent of Irans economy after very fast growth in 2016
(from 3.0percent in the April 2017 WEO). and cuts in oil production in oil exporters through
This change is driven to an important extent March 2018 under the extended OPEC agreement.
by the revision to Turkeys growth in 2017 to In 2018, growth is expected to increase to 3.5per-
5.1percent (2.5percent in April), reflecting a cent, mostly reflecting stronger domestic demand
stronger-than-expected outturn in the first quarter in oil importers and a rebound of oil production
of the year, driven in part by a recovery in exports in oil exporters. However, regional insecurity and
after several quarters of contraction and a more geopolitical risks still weigh on the outlook. In
expansionary fiscal stance. The outlook was also Saudi Arabia, although non-oil growth is expected
revised up for Poland (to 3.8percent in 2017 and to strengthen somewhat this year, overall output is
3.3percent in 2018), reflecting better-than-expected expected to be broadly flat as real oil GDP declines
growth in the first half of 2017 and the expected as a result of the commitments under the extended
pickup in EU-funded projects. OPEC agreement. In 2018, growth is projected to
Economic growth in sub-Saharan Africa is pro- increase to 1.1percent, reflecting an increase in oil
jected to reach 2.6percent in 2017 and 3.4percent output associated with the expiration of the OPEC
in 2018 (broadly in line with the April forecast), agreement. Economic prospects in Pakistan have
with sizable differences across countries. Downside improved, with growth expected to reach 5.3per-
risks have risen because of idiosyncratic factors in cent in 2017 and 5.6percent in 2018, benefiting
the regions largest economies and delays in imple- from investment in the China-Pakistan Economic
menting policy adjustments. Beyond the near term, Corridor and strong private sector credit. In Egypt,
growth is expected to rise gradually, but barely above growth was 4.1percent in fiscal year 2017 according
population growth, as large consolidation needs to preliminary estimates, and is forecast to reach
weigh on public spending. Nigeria is expected to 4.5percent in 2018, supported by reforms aimed at
emerge from the 2016 recession caused by low oil correcting fiscal and external imbalances, restoring
prices and the disruption of oil production. Growth competitiveness, and creating jobs.
in 2017 is projected at 0.8percent, owing to recov-
ering oil production and ongoing strength in the
agricultural sector. However, concerns about policy Inflation Outlook for201718
implementation, market segmentation in a foreign Headline inflation rates are projected to increase
exchange market that remains dependent on central in both advanced and emerging market and devel-
bank interventions (despite initial steps to liberalize oping economies, though somewhat less briskly than
the foreign exchange market), and banking-system anticipated in the April 2017 WEO, partly reflecting
fragilities are expected to weigh on activity in the weaker-than-expected oil prices. In advanced econo-
medium term. In South Africa, growth is pro- mies, inflation is forecast to pick up from 0.8percent
jected to remain subdued at 0.7percent in 2017 in 2016 to 1.7percent in 2017, reflecting the contin-
and 1.1percent in 2018, despite more favorable ued cyclical recovery in demand and the increase in
commodity export prices and strong agricultural commodity prices in the second half of 2016. Head-
production, as heightened political uncertainty line inflation is expected to stay at 1.7percent in 2018
saps consumer and business confidence.In Angola, before converging to 2percent over the medium term.
growth in 2017 has been revised upward to 1.5per- Inflation in emerging market and developing econo-
cent (1.3percent in April) because a downward mies (excluding Argentina and Venezuela) is projected
revision to oil production in 2016 has raised the to remain roughly stable in 2017 and 2018 (at 4.2per-
extent of the expected rebound.The outlook for cent and 4.4percent, respectivelyclose to the 2016
fuel-importing countries is generally brighter, with estimate of 4.3percent).
an aggregate growth rate of 3.9percent in 2017, Because of weaker fuel prices and negative shocks
rising to 4.4percent in 2018. linked to cell phone prices and prescription drugs,
In the Middle East, North Africa, Afghanistan, and headline inflation in the United States is expected
Pakistan, growth is projected to slow significantly to increase by less than envisioned in the April
in 2017 to 2.6percent (from 5.0percent in 2016) 2017 WEO, though it will still increase signifi-
cantly. Consumer price inflation is projected to pass-through from the pesos depreciation through
reach 2.1percent in 2017 (2.7percent in the January 2017, and to fall within Banco de Mxicos
April WEO), up from 1.3percent in 2016. Core tolerance band of 24percent in 2018. In Argen-
personal consumer expenditure inflation remains tina, annual consumer price index inflation is pro-
subdued and is projected to rise more slowly, slightly jected to decline sharply during 2017 and 2018 as
exceeding 2percent in 2019 before returning to the the impacts of the large exchange rate depreciation
medium-term objective of 2percent targeted by the and tariff adjustment in 2016 fade, the central bank
Federal Reserve. maintains a tight monetary policy stance, and wage
Inflation is also projected to pick up in the euro negotiations become more forward looking. After
area, from 0.2percent in 2016 to 1.5percent this rising to 6.3percent in 2016, headline inflation in
year, mostly reflecting higher energy prices and the South Africa is forecast to decline to 5.4percent in
ongoing cyclical recovery in demand. But under- 2017, which is within the target band; slowing wage
lying inflation remains stubbornly low and wage growth, a widening output gap, and the easing of
growth subdued amid still-high unemployment in drought conditions are expected to more than offset
some countries. Headline inflation is projected to the effect of higher oil prices and an increase in
converge to core inflation as energy price effects excise taxes. The inflation rate in Turkey has spiked,
dissipate and gradually approach the European Cen- following the liras depreciation, and is expected to
tral Banks objective of below but close to 2percent remain above the 5percent target throughout the
over the next few years, reaching 1.9percent only forecast horizon. Inflation in 201718 is expected
in 2021. In the United Kingdom, the headline to remain elevated at two-digit levels in Angola
inflation rate is projected to peak at 2.6percent this and Nigeria, reflecting the persistent effects of past
year, up from 0.7percent in 2016, before gradually inflationary shocks coming from sharp currency
declining to the Bank of Englands target of 2per- depreciations (including of the parallel exchange
cent as the temporary effect of the pounds depre- rate) as well as higher electricity and fuel prices and,
ciation wanes and inflation expectations remain in the case of Nigeria, reflecting the assumption
well anchored. that monetary policy will remain accommodative
Headline inflation rates are expected to return to going forward.
positive territory in all advanced economies that
experienced deflation in 2016. In particular, head-
line inflation in Japan, after being slightly negative External Sector Outlook
in 2016, is expected to increase to 0.4percent in Global trade is estimated to have grown by 2.4per-
2017 on the back of higher energy prices on a cent in2016 in volume terms, the slowest pace
year-over-year basis and a narrowing output gap. since2009, with weak growth in both advanced
But inflation rates are projected to remain below economies and emerging market and developing
the Bank of Japans target throughout the fore- economies. In the former, weaker trade growth was
cast horizon. related to an investment slowdown and inventory
The modest increase in inflation rates projected adjustment, especially during the first part of the
for emerging market and developing economies as year. In the latter, persistent weakness in trade growth
a group conceals sizable cross-country differences. was related to a protracted trade slowdown in China
Headline inflation in China is expected to remain and a sharp import contraction in some commodity
tame at 1.8percent in 2017, reflecting weakening exporters facing macroeconomic strains, notably Latin
food prices in recent months, and to pick up gradu- America, sub-Saharan Africa, and the Common-
ally to 2.6percent over the medium term. Inflation wealth of Independent States. As discussed earlier,
rates in Brazil and Russia are forecast to decline global trade growth picked up meaningfully in late
faster than projected in the April 2017 WEO, 2016 and early 2017, reflecting a recovery in global
reflecting stronger effects from negative output demand and especially capital spending. Consequently,
gaps, currency appreciations, and favorable supply global trade growth is projected to rebound to about
shocks to food prices. In Mexico, headline inflation 4percent in 2017 and into the medium term, about
is expected to rise to 5.9percent this year because 1percentage point higher than GDP growth at market
of the liberalization of domestic fuel prices and exchange rates.
Figure 1.14. Global Current Account Balances in China as imports recover. Among debtor countries,
current account deficits are expected to moderate in coun-
Global current account imbalances narrowed marginally in 2016 and are expected
to further compress slightly in 2017. tries in the other advanced economies group, including
Australia and especially the United Kingdom.
1. Global Current Account Balance Although there is no normative presumption that
(Percent of World GDP) current account deficits and surpluses should be com-
4 Afr. and ME Japan China
3 Eur. creditors Adv. Asia Oil exporters pressed, the IMFs2017 External Sector Report highlights
2
how in2016 current account imbalances in some of the
worlds largest economies were too large in relation to
1
country-specific norms consistent with underlying funda-
0
mentals and desirable policies. Current account balances
1
are expected to move in a direction consistent with a
2
United States Other adv. Em. Asia Discrepancy narrowing of these excess imbalances, even under the
3
Eur. debtors Lat. Am. CEE assumption of constant real exchange rates underpinning
4
2002 04 06 08 10 12 14 16 18 20 22 the projections. The first panel of Figure1.15 depicts
on the horizontal axis the gap between the 2016 current
2. International Investment Position 2016 and Change in Current
Account Balance (Percent of GDP) account balance and its norm and, on the vertical axis,
2.5
Change in current-account-to-GDP
Figure 1.15. Real Exchange Rates and Current Account Figure 1.16. Net International Investment Positions
Balances in Relation to Economic Fundamentals
Creditor and debtor international investment positions widened in 2016 and are
Current account balances are expected to narrow their gaps relative to the 2016 projected to continue widening into the medium term.
current account norm.
1. Global International Investment Position
1. 2016 Current Account Gaps and Change in Current Account (Percent of world GDP)
40
6 Balances, 201617 Afr. and ME Japan China
30 Eur. creditors Adv. Asia Oil exporters
5
Change in current account-to-
SAU 20
4 AUS
GDP ratio, 201617
RUS MEX 10
3 GBR BRA
JPN 0
2 ZAF MYS
IDN IND 10
1 ITA
BEL SGP
NLD 20
0
FRA DEU 30 United States Other adv. Em. Asia
1 USA EA CHE SWE CHN Eur. debtors Lat. Am. CEE
ESP CAN HKG POL KOR THA 40
TUR
2 2005 07 09 11 13 15 17 19 21 22
10 8 6 4 2 0 2 4 6 8
Current account gap, 2016 2. Net International Investment Position, 2016, and Projected
Changes, 201622 (Percent of GDP)
40
decline from peak levels in some economies, reflecting, ening of financial regulation and oversight achieved
in part, a downturn in capital expenditure in extractive since the global financial crisisboth nationally and
industries. Against this backdrop, net financial flows internationallycould lower capital and liquidity buf-
to emerging market and developing economies have fers or weaken supervisory effectiveness, with negative
picked up over the past year, as the current account repercussions for global financial stability.
balances of commodity exporters have shrunk and
global risk appetite has recovered. Following a period A Retreat from Cross-Border Economic Integration
of abundant credit supply, a sudden tightening of Slow growth in median incomes since the global
global financial conditions could expose financial financial crisis and a longer-term trend of worsening
fragilities, especially where buffers may be wearing thin income distributions have contributed to disillusion-
after a period of macroeconomic strains and financial ment with globalization in advanced economies
volatility. For instance, faster-than-expected monetary notably in the United States and parts of Europe. Over
policy normalization in the United States could cause the longer term, a failure to lift potential growth and
reversals in capital flows to emerging markets and an make growth more inclusive in advanced economies
appreciation of the US dollar, imposing strains on could exacerbate the risk of a retreat from cross-border
economies with high leverage, balance sheet mis- integration and hinder the political consensus for
matches, or exchange rates pegged to the US dollar. At necessary market-friendly reforms. Greater protec-
the same time, to the extent that such monetary policy tionism could disrupt global supply chains (Yi 2003;
tightening reflects a stronger outlook for the US econ- Bems, Johnson, and Yi 2010; Koopman, Wang, and
omy, US trading partners would benefit from positive Wei 2014), reduce global productivity, and make
demand spillovers. tradable consumer goods less affordable, harming
Challenges facing euro area banks: The euro area low-income households disproportionately (Fajgelbaum
banking sector has made further progress with balance and Khandelwal 2016). Similarly, indiscriminate curbs
sheet cleanup since the spring, and bank credit growth on immigration would hinder a channel for alleviating
to the nonfinancial private sector has been positive labor force constraints in aging societies and reduce
since mid-2015 (though below GDP growth). None- opportunities for skills specialization and productivity
theless, nonperforming loan (NPL) ratios were still growth over the long term.6
high in the first quarter of 2017, at about 5.7per-
cent for the euro area, and greater than 10percent Persistently Low Inflation in Advanced Economies
in six countries (including Italy, which accounts In many advanced economies, steady progress
for about 30percent of the euro areas NPL stock). toward central bank inflation targets has been elusive,
Profitability also remains a challenge, with stubbornly reflecting in part the slow reduction of spare capac-
high cost-to-asset ratios, especially for medium- and ity in labor markets. An environment of persistently
small-size banks. As discussed in Chapter1 of the subdued inflation (which could ensue if domestic
October 2017 GFSR, about one-third of global sys- demand were to falter) can carry significant risks
temically important banks (mostly European banks) by leading to a belief that central banks are willing
are not expected by analysts to generate sustainable to accept below-target inflation, thereby reducing
returns, even by 2019. Low earnings hinder banks medium-term inflation expectations.7 Low inflation
ability to build cushions against unexpected losses and and interest rates would reduce central banks capac-
to raise capital in markets. Without a more concerted ity to lower real interest rates to restore full employ-
effort to clean up balance sheets and improve banks ment in an economic downturn. Real wages would
cost efficiency, financial stability concerns and fears of also be less flexible, and when demand falters, firms
adverse feedback loops among weak demand, prices, would be more likely to resort to laying off workers
and balance sheets could be reignited in parts of to reduce costs, amplifying the recessionary impulse.
the euro area. If political risks were to reemerge, for In sum, prolonged below-target inflation deepens the
instance, an accompanying rise in long-term interest
rates would worsen public debt dynamics, especially if
6Chapter4 of the October2016 WEO analyzes the impact of
inflation were to remain weak.
immigration flows on productivity growth in recipient countries.
Financial deregulation: As discussed in Box1.2 of 7Chapter3 of the October 2016 WEO provides a fuller
Figure 1.17. Geopolitical Risk Index and southern Africa. If these factors intensify, the hard-
(Index) ship in directly affected countries, especially smaller
developing economies, would rise commensurately.
Geopolitical risks have risen in recent months.
The risks discussed above are interdependent
and can be mutually reinforcing. For example, a
600
Iraq Russian shift toward inward-looking policy approaches to
invasion annexation
of Crimea Paris cross-border trade, investment, and migration can
500 London attacks increase geopolitical tensions and global risk aver-
bombing ISIS sion. In addition, noneconomic shocks can weigh
Syrian
400 civil war escalation directly on near-term economic activity and hurt
Madrid
9/11 escalation
bombing Iran nuclear longer-term confidence and market sentiment. Also,
tensions Arab Spring:
300 Syrian and faster-than-anticipated tightening of global financial
Libyan wars conditions or a shift toward protectionism in advanced
economies could create capital outflow pressures from
200
emerging markets.
100
Fan Chart
0
A fan chart analysisbased on equity and com-
2000 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 Aug. modity market data as well as the dispersion of
17
inflation and term spread projections of private sector
forecastersyields a balance of risks that remains
Source: Caldara and Iacoviello (2017).
Note: ISIS = Islamic State. slightly tilted to the downside for 2017 and 2018 (Fig-
ure1.18). Despite the broadly unchanged balance of
risks around the global growth forecast, the contribu-
tions of selected risk factors have changed. Relative to
downside risks to advanced economies medium-term the estimates made in October 2016, the distribution
growth prospects. of term premiums forecasts and the prices of S&P 500
Index options now imply more upside risk to growth
Noneconomic Factors
in 2017 and less upside risk to growth in 2018, likely
Rising geopolitical tensions and domestic polit- reflecting less upbeat views for US fiscal stimulus over
ical discord can hurt global market sentiment and the medium term and optimistic valuations in the
confidence, burdening economic activity. For many US stock marketboth of which leave less room for
countries severely affected by such factors, the base- upward surprises. At the same time, the distribution of
line scenario assumes a gradual easing of tensions. inflation forecasts and oil price options imply some-
However, these episodes may turn out to be more what more downside risk than a year ago, suggesting
protracted, delaying recovery in these economies. Mea- that analysts see greater scope for inflation and oil
sures of geopolitical risk have risen in recent months prices to surprise on the upside and dampen growth
(Figure1.17), and recent research shows that higher (an upward surprise in inflation could lead central
geopolitical tensions can weigh on global activity.8 banks to tighten monetary policy earlier than markets
Weak governance and large-scale corruption can also currently predict, while higher-than-expected oil prices
undermine confidence and popular support, taking a would subtract from consumer disposable income).
heavy toll on domestic activity. Other noneconomic The probability of a recession over a four-quarter
factors weighing on growth in certain regions include horizon has declined relative to the probability com-
the damaging effect of weather-related disasters, puted in March 2017 in the euro area, Japan, and
including the persistent effects of drought in eastern the Latin America 5 group (Brazil, Chile, Colombia,
Mexico, and Peru), consistent with higher projected
8Caldara and Iacoviello (2017) construct an index of geopoliti-
growth rates. Recession probabilities are broadly
cal risk and document how increases in the index have historically
been associated with negative effects on a broad set of economic unchanged for the United States and other regions
activity indicators. (Figure1.19). Deflation risksas measured by the esti-
mated probability of a decline in the price level four Figure 1.18. Risks to the Global Outlook
quarters aheadhave declined for the euro area and The balance of risks implied by the fan chart analysis is slightly tilted to the
Japan, reflecting stronger projected growth in domestic downside in 2017 and 2018.
demand. Deflation probabilities have increased slightly 5 1. Prospects for World GDP Growth
1
from low levels in the East Asia region, where inflation (Percent change)
has softened in several economies in recent months, 4
and for the Latin America 5 group, where inflation is
3 WEO baseline
projected to decline further over the coming year (as
pass-through from earlier currency depreciations fade 50 percent condence interval
2
70 percent condence interval
and negative output gaps continue to exert downward 90 percent condence interval
pressure on inflation in some economies). 1 90 percent condence interval from October 2016 WEO
0
2014 15 16 17 18
Policy Priorities 1.5 2. Balance of Risks Associated with Selected Risk Factors 2
(Coefcient of skewness expressed in units of the underlying
The main cross-cutting policy challenges are to 1.0 variables)
boost potential output and ensure that its benefits are
0.5
broadly shared, and to build resilience against down-
side risks. With countries now facing divergent cyclical 0.0
conditions, varied monetary and fiscal policy stances 0.5 Current year (October 2016 WEO)
remain appropriateand completing the economic Current year (October 2017 WEO)
recovery and adopting strategies to ensure fiscal sus- 1.0 Balance of risks for Next year (October 2016 WEO)
Next year (October 2017 WEO)
tainability are still imperatives for many economies. 1.5
Term spread S&P 500 Ination risk Oil market risks
The urgency for structural reform is particularly
high in advanced economies, where crisis legacies, Dispersion of Forecasts and Implied Volatility 3
demographic shifts, and continued weak productivity 80 3. 1.2 125 4. 0.5
GDP (right scale) Term spread
trends are restraining potential growth; but also in 70 (right scale)
VIX (left scale) 1.0 100
many emerging market and developing economies, 60 Oil (left scale) 0.4
many of which need to activate new sources of growth. 50 0.8 75
The cyclical upswing opens an ideal window of 40 0.3
opportunity for making progress with reforms, espe- 30 0.6 50
cially those that have more powerful economic benefits 20 0.2
when implemented in times of strong demand (such as 0.4 25
10
reforms to job protection and unemployment benefits,
0 0.2 0 0.1
as discussed in Chapter2 of the April 2016 WEO). By 2006 08 10 12 14 Aug. 2006 08 10 12 14 Aug.
17 17
the same token, where aggregate demand is still weak,
macroeconomic policy needs to be supportive to foster Sources: Bloomberg L.P.; Chicago Board Options Exchange (CBOE); Consensus
reform implementation. Economics; Haver Analytics; and IMF staff estimates.
1The fan chart shows the uncertainty around the October 2017 World Economic
By acting together, policymakers could amplify the Outlook (WEO) central forecast with 50, 70, and 90 percent condence intervals. As
beneficial effects of reforms and help reduce down- shown, the 70 percent condence interval includes the 50 percent interval, and the
side risks to the outlook. The model simulations in 90 percent condence interval includes the 50 and 70 percent intervals. See
Appendix 1.2 of the April 2009 WEO for details. The 90 percent intervals for the
Scenario Box1 show that the IMFs macroeconomic current-year and one-year-ahead forecasts from the October 2016 WEO are shown.
policy advice for the Group of Twenty economies 2The bars depict the coefcient of skewness expressed in units of the underlying
variables. The values for ination risks and oil market risks enter with the opposite
(in addition to what is already assumed in the WEO sign since they represent downside risks to growth.
baseline) would have key global benefits, especially if 3GDP measures the purchasing-power-parity-weighted average dispersion of GDP
implemented at the same time. The policy stimulus in growth forecasts for the Group of Seven economies (Canada, France, Germany,
Italy, Japan, United Kingdom, United States), Brazil, China, India, and Mexico. VIX is
countries with fiscal space would strengthen external the CBOE Standard & Poors (S&P) 500 Implied Volatility Index. Term spread
demand for countries needing fiscal consolidation, measures the average dispersion of term spreads implicit in interest rate forecasts
for Germany, Japan, the United Kingdom, and the United States. Oil is the CBOE
buffering the near-term drag on activity; in advanced crude oil volatility index. Forecasts are from Consensus Economics surveys. Dashed
economies, tightening policy, the net effect on output lines represent the average values from 2000 to the present.
of spillovers from abroad and domestic policy tight-
Fiscal policy should, in principle, also be calibrated Figure 1.20. Advanced Economy Output Gaps, 2017
(Percent of potential GDP)
with cyclical conditions but, in many advanced
economies with remaining slack, it is constrained Most large advanced economies are estimated to be operating below potential.
by the need to avoid potentially destabilizing public
debt dynamics or to rebuild buffers. Given the need 1.0
to secure the recovery and bolster inclusiveness, the
composition of spending and revenues and any con-
0.5
solidation measures should be made as growth- and
distribution-friendly as possible.
In the United States, where output is approaching 0.0
potential, the consolidation should start in 2018. And PPP GDP-weighted average
in the short term, avoiding political brinkmanship over 0.5
appropriations and promptly raising the debt ceiling
are essential. In the euro area, countries with very low
1.0
deficits and relatively low debt should use fiscal space to
support structural reforms and boost public investment
to raise potential growth. For instance, a more expan- 1.5
sionary stance in Germany, where tax buoyancy amid
an economic recovery is adding to fiscal space, would 2.0
AUS FRA ITA KOR JPN CHE ESP TWN CAN GBR BEL USA NLD DEU SWE
permit a much-needed increase in public investment
while generating positive spillovers to countries with
Source: IMF staff estimates.
deficient demand. Avoiding a re-emergence of fiscal sur- Note: Data labels in the gure use International Organization for Standardization
pluses would also help correct the external imbalances (ISO) country codes. PPP = purchasing power parity.
of Germany. Indeed, as Chapter4 emphasizes, higher
public spending designed to boost potential output can
have both domestic benefits as well as positive spillovers management to smaller banks, faster modernization
to other economies, especially those with economic and harmonization of insolvency regimes, and stimu-
slack and monetary accommodation. By contrast, grad- lating distressed debt markets by facilitating national
ual fiscal adjustment accompanied by growth-friendly asset-management firms. To raise bank profitability
measures is appropriate for Italy and France. In view sustainably, further business-model upgrading, cost
of remaining economic slack and exceptionally weak rationalization, and consolidation remain critical;
core inflation, Japan should withdraw fiscal support a proactive approach to bank resolution could help
very gradually, including through a gradual increase provide incentives for action in these areas. Faster
in the consumption tax rate over several years to bring progress is also needed for completing the Banking
the primary balance to a debt-stabilizing level, while Union (with a common, effective deposit insurance
prioritizing demand-friendly structural reforms. In the scheme and common fiscal backstop) and advancing
United Kingdom, where uncertainty about the out- the Capital Markets Union plan.
come of negotiations with the EU weighs on sentiment
Bolstering Medium-Term Potential Output and
and investment, a gradual consolidation path remains
Inclusiveness
appropriate.
Strengthening resilience and securing the recov- A cyclical upswing provides a golden opportunity
ery in the euro area will also require accelerating the for adopting structural reforms and will amplify and
repair of bank balance sheets and durably improving accelerate their beneficial effects. Policymakers can
banking system profitability. Only a comprehensive safeguard and improve prospects for potential output
and proactive approach to reducing NPLs can lift through measures to expand labor supply and create
the drag on credit growth and eliminate risks of an an environment conducive to stronger productivity
adverse feedback mechanism among weak inflation, growth. Many of these reforms would also help raise
balance sheets, investment, and productivity. Mea- the inclusiveness of income gains, and some would
sures to accelerate the reduction of NPLs can include broaden economic opportunities across the skills spec-
broadening European Central Bank guidance on NPL trum. Reform priorities vary across countries, depend-
ing on the key impediments to potential output, but Investment in physical infrastructure: Empirical
generally fall into the following areas: evidence from advanced economies suggests that,
Distribution-friendly fiscal policies: As discussed in if done right, infrastructure investment brings
depth in the October 2017 Fiscal Monitor, govern- both short- and long-term benefits: an increase in
ments seeking to improve equity in incomes and public investment of 1percent of GDP can raise
opportunities can rely on fiscal policy as a powerful the level of output by 1 percent over the medium
redistributive tool. For many advanced economies term (Abiad, Furceri, and Topalova 2016). After
with high public debt, limited fiscal space, and high three decades of almost continuous decline, public
tax and spending levels, fiscal and redistributive investment in infrastructure and the stock of public
objectives should be achieved through revenue-neutral capital as a share of output are near historic lows
increases in tax progressivity, spending reallocations, in advanced economies. Many countries could take
and improved spending efficiency. In advanced econ- advantage of the favorable funding environment to
omies where tax progressivity has declined in the past improve the quality of the existing infrastructure
few decades, raising the top marginal tax rates and stock and implement new projects (see Chapter3 of
reducing opportunities for tax avoidance and evasion, the October 2014 WEO). Countries with deficits
especially for high-income earners, could improve the in infrastructure include Australia, Canada, Ger-
distribution of income. Many advanced economies many, the United Kingdom, and the United States.
also have room to significantly increase the taxation Priorities vary but, in most cases, include upgrading
of immobile capital and wealth. surface transportation and improving infrastructure
Investment in human capital: Ensuring broad-based technologies (in high-speed rail, ports, telecommu-
access to high-quality education promotes produc- nications, broadband), as well as green investments.
tivity and a more equitable distribution of income Fostering greater labor supply: Population aging will
over the long term. It also raises the adaptability of exert downward pressure on labor force participation
the workforce to structural transformation, includ- rates in most advanced economies in the coming
ing a persistent shift in work and employment years, with growth in the workforce projected to
relations (with a greater incidence of part-time work decline from about 0.8percent a year in 19952015
in many advanced economies and a greater share of to about half that rate by 2022 (based on October
workers on temporary contracts), as highlighted in 2017 WEO forecasts). To counter this decline,
Chapter2. Short-term measures to help households policymakers could raise the statutory retirement age
through economic downturns or technology- and (where doing so would help close funding gaps in
trade-related displacement include active labor mar- pension systems) and take measures to accelerate the
ket policies (that help workers find jobs in expand- narrowing of gender gaps in labor force participa-
ing sectors) and social safety nets (to smooth the tion. Gender gaps could be narrowed by eliminating
effects of temporary income loss and keep workers tax provisions that discourage second earners in
attached to the labor force). In the longer term, households (Italy, Japan, United States), ensuring
attaining inclusive and sustainable growth amid the availability of affordable child care (Canada,
continued structural change will require adequate Germany, Italy, Japan, United Kingdom, United
education, skills building and retraining, and pol- States), fostering flexible work arrangements (Can-
icies (such as credit access) to facilitate geographic ada, Japan), and offering family-friendly benefits
mobility. In the United States, policy priorities such as parental leave (Canada, United States). In
include supporting early childhood education and aging societies, ensuring the affordability of elderly
science, technology, engineering, and mathematics care is also crucial, given that, if care is too expen-
programs, and rethinking the financing model for sive, it would typically be the secondary earners
public schools and funding for tertiary education in householdstypically womenwho shoulder
to improve outcomes for youth from lower- and the burden of unpaid work at home. Immigration
middle-income households. Apprenticeship and reform could also help expand the labor force, limit
vocational programs have worked well in some the increase in dependency ratios, and raise pro-
countries to offer attractive careers (for example, in ductivity and labor force growth in some countries
Germany) and can be upgraded in many countries, (through, for example, skills-based immigration
for instance, in France and the United States. reform in the United States, continued targeted
immigration policy in Canada, and allowing Figure 1.21. Emerging Market and Developing Economy Output
Gaps, 2017
more use of foreign workers in Japan). In Europe,
(Percent of potential GDP)
integration of refugees into the workforce should
be facilitated through swift processing of asylum Cyclical conditions are diverse in emerging market and developing economies.
applications, language training, job search assistance,
better recognition of migrants skills through creden- 2
tial systems, and support for entrepreneurship.
Product and labor market reforms: Persistently sluggish
1
productivity in some countries has led to greater
emphasis on product and labor market reforms,
especially given the scarcity of fiscal space. These 0
reforms have been found to raise productivity and PPP GDP-weighted average
employment and to improve resilience to shocks.10 1
Priorities include lower barriers to entry into profes-
sional services, certain network industries, or retail
2
trade (for example, Australia, Greece, Italy, Japan,
Spain); employment protection legislation reforms to
reduce labor market duality, such as easing hiring and 3
In China, the composition of fiscal policy should to persist so long as monetary policy settings remain
favor the rebalancing of the economy, and the aug- broadly accommodative and equilibrium real interest
mented deficit should be gradually lowered to a rates remain low in advanced economies. Countries
debt-stabilizing level. The pace of deficit reduction receiving buoyant capital inflows may need to step up
planned in Russia would appropriately entail a steady efforts in financial sector supervision and regulation to
adjustment to lower oil prices, but should be built on manage vulnerabilities, deter excessive borrowing, and
more permanent and better-targeted measures than help ensure that financing flows to projects that contrib-
currently envisaged. ute to raising aggregate productivity.
In Saudi Arabia, a gradual but sustained fiscal con- Where an important share of external borrowing is
solidation to eliminate the budget deficit over several undertaken directly by the corporate sector, curtailing
years would strike the right balance between safeguard- any tax preferences for debt (over equity finance) could
ing activity and preserving fiscal buffers. help keep the risk of overborrowing in check. Ensuring
As currencies have stabilized or gained against the efficient corporate insolvency and restructuring frame-
US dollar since the spring, inflation has continued to works would also help achieve faster and less costly
decline in many emerging market economies, more resolution of problems should repayment difficulties
recently helped by the decline in oil prices. Disin- arise as global financing conditions gradually become
flation has been more rapid than expected in some less accommodative.
countries, such as Brazil, India, and Russia, which
has allowed monetary policy easing in recent months. Bolstering Medium-Term Potential Output and
Monetary policy will need to stay tight in countries Inclusiveness
where inflation rates remain well above central bank Safeguarding and furthering past gains in per capita
targets, such as in Argentina and Turkey. In China, incomes and living standards is imperative across
where monetary accommodation should be gradually emerging market and developing economies in light of
reduced, the monetary policy framework could be the sizable development needs of most countries. Some
made more effective by phasing out monetary targets, countries that are projected to maintain strong growth
resuming progress toward a more flexible exchange rates in the baseline forecast will need to keep the main
rate, and improving communications. downside risks in check (for instance, in China, where
Exchange rate flexibility has served many emerging it would be advisable to deemphasize near-term growth
market and developing economies well in recent years. targets and focus on reforms that would enhance
It has helped support capital inflows where domestic the sustainability of growth). Countries with modest
and external financial conditions have tightened, and medium-term growth prospects will urgently need to
helped safeguard growth and limit the drawdown tackle the most binding structural impediments to
of fiscal and reserve buffers following terms-of-trade growth. Priorities vary, but, in many countries, include
declines in commodity exporters. Wherever possible, improving the quality of infrastructure and education,
exchange rates should be used as the main buffer strengthening governance, enhancing the business
against external shocks. climate, and facilitating greater female labor market
Strengthening financial resilience is an overarch- participation, as well as a host of product and labor
ing priority for emerging market and developing market reforms and further trade integration.
economies. In China, minimizing the risk of a sharp Inclusiveness: As discussed in the Fiscal Monitor, emerg-
economic slowdown will require intensification of the ing market and developing economies generally have
authorities current efforts to tighten supervision, rein higher levels of inequality than advanced economies
in the expansion of credit, and tackle the underlying but, in many cases, their lower administrative capacity
stock of bad assets. and limited fiscal space restrict the fiscal tools available
Many other emerging market and developing econ- for redistribution. For countries with low adminis-
omies with open capital accounts need to be mindful trative capacity and larger informal sectors, setting a
of a possible buildup of financial stability risks in an relatively high tax-exempt threshold for the personal
environment of easy global monetary conditions, and income tax and gradually decreasing it as adminis-
be aware of the risks from volatility as the US Federal trative capacity improves would help increase com-
Reserve gradually withdraws stimulus. Net capital inflow pliance as well as progressivity over time. Reducing
pressures for emerging market economies are likely opportunities for tax avoidance and evasion, especially
for high-income earners, is also important. Indirect those for men in emerging market and developing
taxation (such as a value-added tax or a consumption economies (the average gap is close to 30percentage
tax) has still the potential to be progressive, if revenues points for emerging market economies of the Group
are used to finance progressive spending and if com- of Twenty). Gender gaps in labor force participa-
plemented by excise taxes on luxury goods. Improving tion not only hold back potential output, but also
access to quality education and health care for the limit womens economic and social opportunities,
disadvantaged is also crucial for improving equity. In harming inclusiveness. Priority reforms include
education, efforts should be focused on eliminating eliminating legal barriers that prevent women from
enrollment gaps in primary and secondary education, working, improving infrastructure, and enhancing
especially for the disadvantaged, and expanding the gender equality in accessing social services, finance,
role of private financing and student loans for higher and education (for example, India).
education. In the area of health care, the priority is to Product and labor market regulations and trade
achieve universal health coverage with a broad package policies: Fostering greater competition in domes-
of essential health services. Improving efficiency of tic product and service markets, simplifying labor
social spending is also crucial. market regulations, and removing barriers to trade
Infrastructure: In emerging market economies and are also important broad reform areas for many
low-income countries, infrastructure provision per economies, and involve a varied set of priorities. In
capita is still a fraction of that in advanced econo- South Africa, for example, further progress is needed
mies. Inadequate infrastructure is widely judged a to facilitate entry by new firms into power genera-
key barrier to growth and development, especially tion, transportation, and telecommunications, which
in Latin America and sub-Saharan Africa. Selecting would reduce the cost of key business inputs and
public infrastructure projects with diffused pro- thereby foster growth and job creation. The recent
ductivity gains and raising the efficiency of public agreement to introduce a national minimum wage,
infrastructure spending are principal challenges for combined with a code of good practice for collective
many economies. In Brazil, ongoing efforts to make bargaining, has the potential to raise living standards
the infrastructure concessions program more attrac- for those below the poverty line. At the same time,
tive to investors while improving the standards of its employment impact will need to be carefully
governance and program design would help alleviate monitored, with the government standing ready to
key supply-side bottlenecks and support near-term introduce complementary measures for vulnerable
demand. In Colombia, implementation of the sectors, such as small and medium-sized enter-
authorities infrastructure agenda would help reduce prises. Further labor market reforms are advisable to
a historical infrastructure gap, foster private invest- ensure that wages are determined by firm-specific
ment, and help exporters access markets. conditions. In India, simplifying and easing labor
Institutions: Many emerging market and developing market regulations and land acquisition procedures
economies have substantial scope to improve the cli- are long-standing requirements for improving the
mate for business and investment. Decisive actions business climate. Expanding the role of market
to enhance governance and the rule of law would forces in the economy is a priority in China and will
help rein in corruption, strengthening business entail removing barriers to entry in the highly closed
confidence and providing a boost to investment in services sector and allowing state-owned enterprises
some countries (for example, Brazil, Mexico, Peru). to face harder budget constraints. Productivity
Strengthening institutions can also help reduce could be fostered by reducing tariff and nontariff
country risk perceptions and act as a countervailing barriers to international trade (for instance, Brazil,
force against a possible tightening of global financial China, and India).
conditions. Many countries could simplify regula-
tions and administrative procedures for starting a
business, increase the efficiency of the legal system, PoliciesLow-Income Developing Countries
and reduce regulatory uncertainty (for example, As with the broader group of emerging market
Turkey, South Africa). and developing economies, low-income developing
Unleashing greater labor supply: Labor force par- countries dependent on commodity exports continue
ticipation rates for women are much lower than to face weaker economic prospects than those countries
Figure 1.22. Per Capita Real GDP Growth across Low-Income distress or already in debt distress, and one-third
Developing Countries
at moderate risk.12 Many low-income developing
(Percent)
countries continue to experience conflict and secu-
Low-income developing economies dependent on commodity exports continue to rity disruptions (Afghanistan, Chad, Somalia, South
face weaker economic prospects than those with more diversied export bases. Sudan, Yemen, a few parts of Nigeria), whereas parts
of sub-Saharan Africa face food insecurity related to
6 droughts (The Gambia, South Sudan, Somalia).
With divergent prospects, policy priorities continue
4 to differ across low-income developing countries.
Prospects for commodity exporters are heavily influ-
enced by the process of adjustment to lower com-
2
modity prices. The adjustment needs to continue
and, in some cases, accelerate, based on comprehen-
0 sive and internally consistent sets of policies. Fiscal
policy needs to be better calibrated to contain debt
2
19952005
accumulation while protecting outlays key to growth
200616 prospects, such as priority capital expenditures and
201722 social spending. In many countries, improvements
4
in domestic revenue mobilization and continued
rationalization of spending needs, along with
6
Diversied exporters Oil exporters Non-oil commodity concessional financing, are necessary to underpin
exporters successful adjustment processes. Allowing greater
exchange rate flexibilitywhere an optioncould
Source: IMF staff estimates. act as a shock absorber and facilitate adjustment,
Note: Bars denote PPP GDP weighted averages; red markers indicate the medians;
and black markers denote the top and bottom deciles of per capita GDP growth in supported by monetary policy settings to contain
the country groups. Country groups are dened in IMF (2015). the inflation pressures that may result from cur-
rency depreciations. Financial stability needs to be
maintained through enhanced financial sector regu-
with more diversified export bases (Figure1.22).11 lation and supervision and by addressing emerging
With policy adjustments to lower oil revenues financial sector vulnerabilities, including increased
delayed, fiscal deficits in some commodity-exporting domestic arrears and NPLs. Countries in or at high
low-income countries remain large, external posi- risk of debt distress need to accelerate the adjust-
tions are weaker, and financial sector vulnerabilities ment and limit nonconcessional external borrowing.
are emerging. Although GDP is set to grow in most Policy priorities for diversified low-income develop-
commodity-exporting low-income countries in2017, ing countries vary. However, an overarching goal for
fuel exporters are projected to do worse than nonfuel these economies should be to strike a better balance
commodity exporters. By contrast, countries with between spending for development and social needs
more diversified export bases have recorded relatively and improving public debt sustainability by rebuild-
strong growth, which is expected to continue at a ing fiscal positions and foreign reserves holdings
rapid clip, in part, with the benefit of lower oil bills. while growth is strong.
Robust growth, however, has not always translated into
improved fiscal and external current account positions, Across all low-income countries, better debt manage-
reflecting limited progress in adopting countercyclical ment would also help those exposed to global financial
policies and higher public sector spending. markets cope with volatility in capital inflows, balance
Total public debt and debt service have therefore sheet currency exposures, and the prospect of monetary
risen sharply across low-income developing countries, policy normalization in the United States. Over the
with about one-third at high risk of external debt long term, the 2030 Agenda for Sustainable Develop-
11Classifications of low-income countries according to commodity 12Based on the Debt Sustainability Framework for Low-Income
ment identifies a broad range of issues that will require lions were lifted out of poverty in emerging market
action to deliver durable and inclusive growth. Within and developing economies during a period of rapid
this framework, generating sustainable and resilient cross-border integration, helping reduce global income
growth will require steps to promote diversification and inequality. However, global trade has slowed dramat-
structural transformation and bridge infrastructure gaps. ically in recent years, mostly reflecting weakness in
In particular, efforts to boost domestic revenue mobi- aggregate demand, but also the slower pace of new
lization, strengthen debt management, and ensure that trade reforms and an uptick in protectionist measures.
public spending is efficient and well targeted would con- And trade rules have not kept pace with the evolving
tribute to scaling up infrastructure investment without global economy; for example, integrated global pro-
endangering public debt sustainability. To make growth duction structures require more coherent rules across
more inclusive and resilient, policies should be oriented several policy areas, such as goods trade, services trade,
toward creating jobs and encouraging gender equality, investment policy, and intellectual property.
promoting environmental sustainability, boosting access Rolling back temporary barriers to trade introduced
to financial services, and strengthening the redistributive since the global financial crisis and reducing trade
role of fiscal policy to protect the most vulnerable. costs would support the nascent recovery in trade,
reigniting an important driver of global productivity
Multilateral Policies growth. To that end, pressing ahead with an ambitious
trade agenda is crucial. A global trading systemwith
Strong, sustainable, balanced, and inclusive growth strong, well-enforced rules that continue to adjust
requires a well-functioning, cooperative, multilat- to promote competition and a level playing field
eral framework for international economic relations. remains critical (IMF, World Bank, and WTO 2017).
Because national policies create spillovers across Addressing tariff barriers in sectors where they remain
countries, all countries are better served when policy- high, such as agriculture, and implementing commit-
makers engage in regular dialogue and work within ments under the Trade Facilitation Agreement, which
agreed mechanisms to resolve disagreements. At the came into effect in February2017, can significantly
same time, the international community continuously reduce trade costs in traditional areas. Advancing trade
needs to adapt the multilateral system to the changing reforms in services and in other areas, such as digital
global economy. Active dialogue and cooperation will trade, and improving cooperation in investment pol-
help improve and modernize rules while addressing icies can make positive contributions to cross-border
individual countries valid concerns. This process will trade flows and global growth; although progress is
ensure continued mutual benefits and evenhanded- best made at the global level, ambitious, broad-based
ness and, together with strong domestic policies, help regional agreements that address these frontier areas
avoid a broad withdrawal from multilateralism, either of trade policy can also be helpful. As discussed in
through widespread protectionism or competitive races Chapter1 of the April 2017 WEO, open trade policies
to the bottom in taxation and financial and regula- should be complemented by comprehensive policy
tory oversight. Multilateral cooperation is also vital approaches at national levels to reduce adjustment
for addressing important longer-term challenges in pains and provide opportunities for all.
the global economy, including providing support to
low-income countries for meeting development goals Cooperation for Maintaining Global
and mitigating and adapting to climate change. Financial Stability
Maintaining robust national financial regulatory
Maintaining Rules-Based, Open Multilateral Trade
regimes, including in countries and regions with
with Broadly Shared Gains
systemic financial systems, such as China, Europe,
Cross-border economic integration through trade and the United States, and recapitalizing institutions
openness has been a critical source of productivity and cleaning up balance sheets where necessary pro-
growth and resilience over the past several decades duces positive spillovers for global financial stability.
for countries at all income levels.13Hundreds ofmil-
of global resources, boosted incomes, and expanded access to goods
and services. For a recent summary, see Baldwin (2016). See also
13A body of research has documented that economic integration Wacziarg and Welch (2008); Costinot and Rodrguez-Clare (2013);
together with technological progress has increased the efficient use and Fajgelbaum and Khandelwal (2016).
In addition, there is an urgent need to finalize the and investment integration. Policymakers can make
international financial regulatory reform agenda by more meaningful progress toward equitable tax systems
tackling outstanding challenges, such as the regulation (that prevent an increasing share of after-tax income
and oversight of financial institutions, including non- from accruing to owners of capital) if their national
banks; ensuring regulators can resolve globally systemic efforts to safeguard revenues are backed by multilateral
financial institutions effectively; and strengthening the cooperation.
resilience of central counterparty clearing for deriva-
tives. Coordinated and collective action is needed to Noneconomic Challenges
manage risks to financial stability from cyberattacks, Multilateral cooperation is also indispensable for
money laundering, and terrorism financing. Closer addressing important medium-term global challenges,
cross-border regulatory cooperation is also needed to such as meeting the2030 Sustainable Development
address the pressures that several countries have expe- Goals, and providing financial support to vulnerable
rienced in correspondent banking relationships, which economies and fragile states that face the greatest
play a key role in facilitating global trade, remittances, development needs and, in many cases, deep eco-
and economic activity. As shown in Box1.5, remit- nomic and security challenges. The international
tances have grown in global importance and are a key community will have a key role to play in fostering
mechanism for sustaining consumption in the face of and coordinating financial and other types of support
income shocks. for countries most vulnerable to climate change. As
Last, the high degree of international financial discussed in Chapter3, increases in temperature have
interconnectedness and vulnerability in some regions vastly unequal effects across the world, with the brunt
calls for a closely coordinated and adequately resourced of adverse consequences borne by those who can least
global financial safety net as well as stronger frame- afford it and those who have contributed the least
works for the prevention and resolution of debt crises. to the rising threat of climate change. Low-income
countries will likely suffer disproportionately from
Cooperation on International Taxation Issues further global warming, which is expected to trigger
As increased capital mobility across borders has more severe droughts, storms, and epidemics. Coupled
fueled international tax competition, governments have with rising sea levels, these effects could feed social
found it more challenging to finance their budgets unrest and refugee flows, with important cross-border
without increasing taxes on labor income or imposing implications. A concerted multilateral effort to help
regressive consumption taxes. International corporate vulnerable economies cope with the consequences
income tax evasion and avoidance through, for exam- of climate change and stem the man-made causes of
ple, profit shifting to lower tax jurisdictions, could global warming is amply justified from both equity
further erode popular support for international trade and efficiency perspectives.
2000
2007
2016
2000
2007
2016
2000
2007
2016
2000
2007
2016
lower the overall participation rate.
But beneath the headline figures, the variations in 1524 2554 5564 65+
how participation rates within various age and gender
groups have changed are striking, with remarkable 4 2. Changes in Population Shares, 200716
(Percentage points)
gains in the participation rates of women in some 3
countries. If such gains continue and broaden, the 2
demographic transition may not immediately translate 1
into a slowdown in the growth of the labor force. This 0
heterogeneity (as well as some evidence of convergence
1
in participation rates) also suggests that there is scope
2
for policies to postpone the adverse effects of the demo-
graphic transition on the growth rate of the workforce. 3
4
Age Groups 1524 2554 5564 65+
For the adult population of advanced economies Sources: Organisation for Economic Co-operation and
as a whole, labor force participation rates declined by Development; and IMF staff calculations.
0.8percentage point since 2007.2 Participation rates Note: The gure shows population-weighted averages across
declined for the young (age 1524the group with 31 advanced economies.
the largest cross-country dispersion in participation
rates), in part because more people stay in school for
longer.3 For the 2554 age group, where participation
remains the highest, rates have been mostly flat in
Prepared by Zska Kczn, with research assistance
total, though with starkly divergent paths for men and
from Ava Hong.
1Unless stated otherwise, the figures for advanced economies women, with mens participation rate declining and
in this box refer to the combined workforces and working-age womens increasing. Participation rates of both men
populations of 31 advanced economies, which account for about and women in the 5564 age group showed a sharp
95percent of the total population of countries classified as rise, and the 65+ participation rates also rose for both
advanced economies in the World Economic Outlook (WEO).
2The total labor force participation rate can be written as the
genders, especially after 2007 (Figure1.1.2).4
population-share weighted average of the participation rates of
different age groups: 3As discussed, for instance, by Balleer, Gmez-Salvador, and
4LFPR i ____p opti Turunen (2009); Aaronson and others (2014); Council of Eco-
LFPRt =i=1 t popt
.
nomic Advisors (2014); Canon, Kudlyak, and Liu (2015); and
Here, i refers to the following age groups: 1524, 2554, 5564, Dvorkin and Shell (2015).
65+. Results are robust to using a finer breakdown of age groups 4Declining participation rates of the young and prime-age
into five- or 10-year intervals. men are highlighted by Balleer, Gmez-Salvador, and Turunen
16
2000
07
16
07
16
2000
07
16
2000
07
16
2000
07
Shifting population shares have tended to push +LFPR0i PSti LFPRti PSti )
overall participation rates down, while rising partic- pop i
where PSti =____t
popt is the population share and t=0
ipation rates within some age groups have tended to
refers to year 2007, the initial year. The contribution
increase them. This effect can be documented using
of the interaction term (combining changes in partic-
a shift-share decomposition, as illustrated in Fig-
ipation rates and changes in group sizes) is typically
ure1.1.3. The figure decomposes changes in overall
very small and is included in the between change in
Figure1.1.3.
(2009), Dvorkin and Shell (2015), Council of Economic
Advisors (2016), and Krause and Sawhill (2017). In European This decomposition suggests that the decline in
economies, this stands in contrast with rising female labor overall participation rates was driven by aging
force participation, which has been declining in the United captured by between changeswhile within
States (for example, Krause and Sawhill 2017). Balleer and changes would have acted to increase participation
others (2009) examine the drivers of the increase in labor force
rates: the contribution of the decline in the partic-
participation rates during the precrisis period in the euro area
and predict a fall in participation rates over the following years ipation rates of the young is more than offset by
based on an age and cohort analysis. the increase in participation rates of the age 25 and
2
FRA DEU ITA JPN ESP GBR USA Outlook and Policy Implications
Looking ahead, demographics are likely to con-
Source: Organisation for Economic Co-operation and
Development.
tinue to play a prominent role in determining the
Note: Labels in the gure use International Organization for path of the aggregate labor force participation rate.
Standardization (ISO) country codes. Over the longer term, the downward influence of
aging on the aggregate labor force participation rate
is likely to dominate. This will restrain growth in the
tion rates: countries where participation rates were potential labor force (affected by the size and age
lower in 2000 tended to see larger increases, while composition of the working-age population and the
those with the highest rates saw smaller increases or participation rates of the demographic groups) and
outright declines (Figure1.1.6).9 hence potential output, as noted in Chapter3 of the
April 2015 WEO.
9Blau and Kahn (2013) examine the drivers of this conver-
Policies to raise participation would help slow
gence and find that the expansion of family-friendly policies
the decline in the labor force growth rate, in turn
(including parental leave and part-time work entitlements) in slowing the rise of the dependency ratio and thereby
other Organisation for Economic Co-operation and Develop- supporting fiscal sustainability. Eliminating poli-
ment countries can explain close to 30percent of the relative cies that discourage second earners in households,
decrease in US womens labor force participation. However, they ensuring the availability of affordable child care
note that these policies also appear to encourage part-time work
and employment in lower-level positions: in the United States,
and elderly care, fostering flexible work arrange-
women are more likely than in other countries to have full-time ments, and offering family-friendly benefits, such
jobs and to work as managers or professionals. as parental leave, would generally be beneficial.
Box 1.2. Will the Revival in Capital Flows to Emerging Markets Be Sustained?
Capital flows to emerging markets slumped to a Figure 1.2.1. Capital Flows to Emerging
multidecade low in 2015, prompting concerns that Market and Developing Economies
outflow pressures could trigger a broader economic
downturn and lead to crises in those economies (see 12 1. EMDEs Capital Flows
Chapter2 of the April 2016 World Economic Outlook). (Percent of GDP)
10
A useful measure for illustrating the unusual downturn
Nonresident
is nonresident capital inflows, which are defined as the 8 inows
net acquisition of emerging market assets by foreign 6 Net ows
investors (also referred to as gross inflows). As a share
of emerging market GDP, nonresident inflows fell 4
to 1.6percent in 2015, the lowest level since 1990 2
(Figure1.2.1, panel 1). Another useful measure is net 0
capital flows, which is defined as nonresident inflows US recessions
less net outward investment by emerging market econ- 2
1990 94 98 2002 06 10 14 16
omy residents excluding official reserves accumulation.
Net capital flows turned negative in 2015 for the 2. Nonresident Capital Inows to EMDEs
first time in at least 35 years, reaching 1.0percent 600 (Billions of US dollars)
China
of emerging market GDP, and remained negative the 500 Em. Asia excluding China
following year. Lat. Am.
400
In recent quarters, however, capital inflows to Em. Eur.
emerging markets have revived. Total nonresident 300 Afr. and ME
Total
capital inflows to emerging markets are estimated to 200
have averaged $200billion in the first two quarters of
100
2017, up from a quarterly average of $120billion in
201516 (Figure1.2.1, panel 2). Net capital flows have 0
also turned up in recent quarters, reaching $115billion 100
in the first half of 2017. The sharp downturn and the 2010 11 12 13 14 15 16 17:
Q1
recent revival in both measures of capital flows can be
attributed to two main developmentsthe evolution
of Chinas financial account and a rollercoaster ride in Sources: Haver Analytics; and IMF staff estimates.
Note: Afr. and ME = Africa and the Middle East; Em. Asia =
portfolio flows to emerging markets. emerging Asia; Em. Eur. = emerging Europe; EMDEs =
emerging market and developing economies; Lat. Am. =
Stabilization of External Pressures in China Latin America.
China experienced a sharp decline in nonresident
capital inflows between the third quarter of 2015 and
the first quarter of 2016. During this period, concerns
about the possibility of a sharp depreciation of the central bank, which kept renminbi depreciation in
Chinese renminbi prompted the repayment of dollar check (Figure1.2.2, panel 1).
debt by Chinese firms. In addition, foreign investors Initially, the capital flows reversal was driven primar-
sought to reduce their exposures to renminbi assets, ily by a reduction in Chinese liabilities to the rest of the
especially offshore bank deposits. Because those funds world, while resident outward investment continued to
had been on-lent by Chinese banks foreign affiliates to grow broadly in line with previous trends (Figure1.2.2,
banks domiciled on the mainland, the mainland banks panel 2). Nonresident inflows recovered in the second
had to repay those loans, thus further reducing total quarter of 2016, but at that point domestic investors
external debt (see McCaulay and Shu 2016). External began to move more and more money out of the coun-
pressures prompted large reserves interventions by the try by acquiring foreign assets. Since the beginning of
2017, resident outflow pressures have abated follow-
ing tighter enforcement of capital flow management
The author of this box is Robin Koepke, with research assis- measures, weakening in the US dollar, and a pickup in
tance from Gavin Asdorian. growth momentum. Net capital outflows (including
250 1. China: Accumulation of Ofcial Reserves 160 1. Nonresident Portfolio Flows to EMDEs
200 (Billions of US dollars) 140 (Billions of US dollars; three-month rolling sum)
120
150 100 Portfolio debt
100 80 Portfolio equity
50 60
40
0 20
50 0
100 20
40 Taper RMB
150 60 tantrum shock US election
200 80
2000 02 04 06 08 10 12 14 17: 2013 14 15 16 Aug.
Q2 17
300 2. China: Breakdown of Quarterly Capital Flows 12 2. EMDEs Capital Flows and GDP Growth
(Billions of US dollars) Nonresident inows
200 Nonresident inows 10 (percent of GDP)
Resident outows EMDEs growth
100 8 (percent)
0 6
100 4
errors and omissions) eased to about $20billion in the In mid-2015, portfolio equity and debt inflows
second quarter of 2017 (from a peak of $210billion in again came under significant pressure when concerns
the third quarter of 2016), which also marked the first about possible renminbi devaluation intensified. From
quarter of central bank reserves accumulation in China the third quarter of 2015 to the first quarter of 2016
since the second quarter of 2015. global investors sold a net $52billion in emerging
market stocks and bonds, exceeding outflows of an
A Rollercoaster Ride in Emerging Market estimated $32billion during the taper tantrum. The
Portfolio Flows episode was a stark illustration of Chinas growing
The second development behind the recent slump importance for global financial markets and the world
and revival of capital flows to emerging markets was a economy, and for other emerging market economies
rollercoaster ride in portfolio inflows that began with in particular.
the taper tantrum in mid-2013 (Figure1.2.3, panel1). After a modest recovery in 2016, portfolio flows
During that episode, investors reacted strongly to were hit by renewed repricing of US bonds after the
signals from the US Federal Reserve that it would start US election in November 2016. This time, the jump
tapering purchases of bonds sooner than previously in US bond yields was driven by expectations of fiscal
expected. Rising US market interest rates weighed on expansion and deregulation that would support growth
emerging market asset prices as foreign investors began and prompt faster monetary tightening. Similar to
to pare their emerging market exposures. the taper tantrum episode, investors responded by
Box 1.3. Emerging Market and Developing Economy Growth: Heterogeneity and Income Convergence
over the Forecast Horizon
Per capita real GDP growth in emerging mar- across countries.1 Zooming in on countries growth
ket and developing economies is projected to pick prospects reveals that they are not as favorable for
up from 3.2percent in 2017 to 3.6percent in some economies in the group as the headline figures
2019 and stay at about 3.7percent in 202022 would suggest.
(Figure1.3.1). The growth differential relative to
advanced economies, where real per capita growth Heterogeneity
is projected to average 1.4percent between 2017 In general, there are sizable differences in emerg-
and 2022, suggests some catching up between the ing market and developing economy growth rates
two groups. However, the headline growth figures
for emerging market and developing economies 1Per capita real income for each group is calculated by sum-
are heavily influenced by the largest economies ming real GDP at purchasing power parities and dividing by
in the group and conceal substantial differences total population for the group.
Figure 1.3.1. Per Capita Real GDP Growth Figure 1.3.2. Per Capita Real GDP Growth,
across Country Groups Emerging Market and Developing Economies,
(Percent) by Region
(Percent)
10
8
8 7
6
6
5
4 4
3
2
2
0 1
0
2 1
19952005
200616 2 19952005
4 201722 3 200616
201722
6 4
AEs EMDEs China Fuel Nonfuel
exporters exporters 5
LAC MENAP EMDE EMDE SSA CIS
excluding Asia Europe
China excluding
China
Source: IMF staff estimates.
Note: Bars denote PPP (purchasing power parity) GDP Source: IMF staff estimates.
weighted averages; red markers indicate the medians; and Note: Bars denote PPP (purchasing power parity) GDP
black markers denote the top and bottom deciles of per weighted averages; red markers indicate the medians;
capita GDP growth in the country groups. The fuel and nonfuel and black markers denote the top and bottom deciles of
exporter subgroups are dened in Table D of the Statistical per capita GDP growth in the country groups. CIS =
Appendix and cover EMDEs only. AEs = advanced economies; Commonwealth of Independent States; EMDE = emerging
EMDEs = emerging market and developing economies. market and developing economies; LAC = Latin America
and the Caribbean; MENAP = Middle East, North Africa,
Afghanistan, and Pakistan; SSA = sub-Saharan Africa.
China (as suggested by Figure1.3.1) as well as India. calculated over different periods (as opposed to 201216 as in
Differences in median growth rates across regions are the regression presented in Table1.3.1), as well as to estimating
more modest. the regressions by weighted least squares. Running the same
regressions with October 2016 WEO data yields similar results
An even starker difference in per capita growth for the fuel-exporter dummy, albeit with a smaller coefficient.
rates exists between fuel-exporting and fuel-importing Dropping large countries, such as China and India, does not
emerging market and developing economies. The affect the results.
15 35 1. 19952016
Zero growth differential
30 with respect to AEs,
Per capita real GDP growth differential 201722
25 19952016
10 Historical Continuing (1.5 percent)
divergence convergence 20
and projected
convergence 15
5
10
5
0 0
5 [4,3] [2,1] [0,1] [2,3] [4,5]
Historical [5,4] [3,2] [1,0] [1,2] [3,4] 5
No
convergence convergence Growth differential with respect to AEs
5 and current (percentage points)
divergence
45 2. 201722
10 40 Zero growth differential
15 10 5 0 5 10 15 20 with respect to AEs,
35
Per capita real GDP growth differential 19952016 201722
30 (1.4 percent)
25
Source: IMF staff estimates.
Note: The gure depicts countries per capita real GDP 20
growth rates averaged over 19952016 (x-axis) against their 15
projected growth rates averaged over 201722 (y-axis), in 10
both cases expressed as a deviation from the per capita real
GDP growth rate for advanced economies averaged over the 5
same period. EMDEs = emerging market and developing 0
economies. 5 [4,3] [2,1] [0,1] [2,3] [4,5]
[5,4] [3,2] [1,0] [1,2] [3,4] 5
Growth differential with AEs
(percentage points)
4For an analysis of emerging market and developing econ- 5The existence of convergence groups or clubs has been widely
omies growth performance compared with that of advanced discussed and tested in the literature on income convergence
economies over the past four decades, see Chapter2 of the (Durlauf and Johnson 1995; Desdoigts 1999; Durlauf and Quah
April 2017 WEO. 1999; Canova 2004).
level) between levels of 2011 per capita real GDP and projected
growth rates holds even when countries growing more slowly
than advanced economies are excluded from the sample.
7Based on data from the World Panel Income Distribution
110
10 1. Commodity Fuel Exporters
105
8
100
6
95
4 90
2 85 Fixed dollar
Fixed nondollar
80 Flexible
0
Fixed dollar Fixed Flexible Regime Regime adjustment
nondollar adjustment 75
70
6 2. Commodity Metal Exporters
65
2010 11 12 13 14 15 16 Jun.
17
4
Source: Gruss 2014.
Note: PPP = purchasing power parity.
2
0
Fixed dollar Fixed Flexible Regime In response to terms-of-trade shocks that directly
nondollar adjustment
affect the external balance, net export volume could
adjust, partly offsetting the initial impacts of the
6 3. Commodity Food Exporters
shocks. The change in real exchange rates in response
to the terms-of-trade shock facilitates this exter-
nal adjustment through the expenditure-switching
4
channel. Such real effective exchange rate adjust-
ment and the associated switch in expenditures are
expected to be more pronounced in countries with
2
a flexible exchange rate regime (Adler, Magud, and
Werner 2017; IMF 2017b). Panel 1 of Figure1.4.5
confirms this notion and shows that, despite facing
0
Fixed dollar Fixed Flexible Regime bigger terms-of-trade shocks, countries with fixed
nondollar adjustment exchange rates experienced the smallest adjustment
in net exports, whereas those with flexible exchange
Source: IMF staff calculations. rate regimes saw strong net export adjustments, which
Note: Regime adjustment covers xed exchange rate
regimes that devalued their parity or changed the exchange
more than offset their terms-of-trade shocks. Export
rate regime toward more exibility during 201317. volumes did not react much, on average, across the
different exchange rate regimes, likely reflecting the
insensitivity of commodity exports to the exchange
rate as well as these countries limited export diver-
100 15 Regime
90 adjustment
10
80 5
70 0
60 5
50 10
40 15
2010 11 12 13 14 15 16 17 40 30 20 10 0 10 20 30 40
REER appreciation
Sources: IMF, Information Notice System; and IMF staff (percent)
calculations.
Note: Yearly average for 2010 16; as of June for 2017.
PPP = purchasing power parity. Source: IMF staff calculations.
Note: REER = real effective exchange rate; ToT = terms of
trade.
ing economies. As such, remittances have the potential (2009); Chami, Hakura, and Montiel (2009); Combes and
Ebeke (2011); De and others (2016); and Beaton and others
to be an increasingly important mechanism for sharing
(2017) consider the importance of remittances as a risk-sharing
income risks on a global scale. arrangement to smooth consumption in developing countries
Although remittances play a positive long-term generally. Beaton, Cevik, and Yousefi (2017) explicitly consider
role in economic and social development, this box the importance of remittances in smoothing consumption under
focuses on an arguably no-less-critical rolethat of fiscal shocks. Few studies have focused on the role of remittances
in smoothing commodity price shocks.
mitigating cyclical risks to household consumption 3Kose, Prasad, and Terrones (2009) define consumption
Less than 1 5 to 10
1 to 0 More than 10
0 to 5 No data
Sources: IMF, World Economic Outlook database; World Bank, Migration and Remittances database; and IMF staff calculations.
of payments flows (Figure1.5.3, panel 1). Their vola- correlations in levels and in first differences in a broad
tility is even lower than that of foreign direct invest- cross-country panel spanning 19902015. Looking at bilateral
ment flows, which are well known to be less volatile remittance flows, Frankel (2011) finds that remittances are
mostly countercyclical for the recipient country. Yet, in some
than equity and portfolio financial flows. Remittances cases, remittances sent primarily for investment motives can be
are also significantly less positively correlated with procyclical, even if to a lesser extent than portfolio or foreign
GDP than foreign portfolio investment and foreign direct investment flows.
Noncommodity
Noncommodity exporter
high remittances
High remittances:
Commodity prices have decreased since the release of the Figure 1.SF.1. Commodity Market Developments
April2017 World Economic Outlook (WEO). Despite
the extension of the production agreement by the Orga- 300 1. Commodity Price Indices All commodities Energy
(2005 = 100) Food Metals
nization of the Petroleum Exporting Countries (OPEC), 260
this year, metal prices have bounced back since June, in 140
Since then, oil prices have rebounded, to about average annual prices of $50.3 a barrel in2017an
$50 a barrel as of late August, in response to signs of increase of 17.4percent fromthe2016 averageand
a slowdown in US production growth. US inventories $50.2 a barrel in2018 (Figure1.SF.1, panel 3).
increased dramatically in June 2017, but declined Uncertainty remains around the baseline assump-
sharply in July and August. The US Energy Informa- tions for oil prices, although risks are balanced. Upside
tion Administration expects US crude production in risks include unscheduled outages and geopolitical
2018 to reach 9.9 mbd, exceeding the previous high of events, especially in the Middle East and Latin Amer-
9.6 mbd recorded in 1970. The International Energy ica as the United States put additional sanctions on
Agency expects demand growth to increase from 1.3 Venezuela. Although these development could cause oil
mbd in 2016 to 1.6 mbd in 2017 and then to soften market disruptions, high inventoriesincluding drilled
to 1.4 mbd in 2018. Hurricane Harvey impacted but uncompleted wellsand the rapid response by
US refinery capacity in late August and spot gasoline shale producers should prevent sharp price rises in the
prices increased sharply. However, crude oil prices near future. As oil markets focus on the US produc-
and medium-term gasoline futures reacted much less, tion/inventory figure, Hurricane Harvey may influence
partially because crude inventories were large, and crude markets significantly if it turns out that physical
reduced production of refined oil translates into weaker damages to infrastructure or labor force dislocation
demand for crude oil. are larger than initially assessed. Natural gas markets
The natural gas price indexan average for Europe, face additional uncertainty due to the Qatar crisis
Japan, and the United Statesdecreased by 9.6per- and renewed tensions between Russia and the United
cent between February 2017 and August 2017, States after the United States approved new sanctions
reflecting seasonal factors and firm supply from the against Russia.
United States and Russia.1 Lower oil prices add extra
downward pressures in countries where oil-linked pric-
ing is more common. Markets were relatively unfazed Metals: China in the Mix
when Saudi Arabia and a coalition of countries severed Metal prices have increased by 0.8percent between
diplomatic ties with Qatar, the worlds largest LNG February and August 2017, with considerable variation
exporter, as exports from Qatar continue. across commodities. By June the metal price index had
The coal price indexan average of Australian reached its lowest point in eight months due to slower
and South African pricesincreased by 16.5percent demand growth in China and the United States. How-
from February 2017 to August 2017. This increase ever, prices rebounded since and continued to do so
follows an initial decline caused by the end of the into August with the improvement in macroeconomic
disruption to coal transportation in Australia due to sentiment, especially in China.
Cyclone Debbie on March 28, 2017. However, strong Iron ore prices dropped by 35percent between
demand from China helped prices recover. In addition, February and June 2017, mainly driven by expansion
sporadic labor disputes in Australian mines provided of production by big producers in Australia and Brazil
additional support, while import restrictions by China attempting to increase market share. Iron ore invento-
put downward pressure on prices, especially for lower- ries at Chinese ports reached an all-time high of more
quality coals. than 140million tons by late June, up 40percent
Oil futures contracts point to a gradual increase of from the year before, according to data from Thomson
prices to about $53 a barrel in 2022 (Figure1.SF.1, Reuters Datastream. With steel prices in China soaring
panel 2). Baseline assumptions for the IMFs average again, however, Chinas steel producers increased
petroleum spot prices, based on futures prices, suggest output to a record high of 74million tons in July.
This, in turn, drove up demand for the key ingredient
in steelmaking, especially for higher-grade ores that
1The IMFs natural gas price index is a weighted average of US
increase the efficiency of steel mills and help lower air
Henry Hub prices, Netherlands Title Transfer Facility prices, and
Argus Northeast Asia liquefied natural gas (LNG) prices. Up to pollution. As a result, the price of iron ore rallied by
December 2016, the index is the average of US Henry Hub, Ger- 29percent from its low in June, reaching an average of
man border prices from Russia (long-term contract), and Japanese
$74.6 per ton in August.
LNG import prices from Indonesia (Japanese Custom-cleared
Crude indexed). The update reflects the increased importance of Copper prices tumbled between February and early
spot markets. May, after strikes at major mines in Chile and Peru
ended, and the export ban in Indonesia was tempo- Price Swings in Agricultural Markets
rarily lifted. However, with supply from Chile again The IMFs agricultural price index decreased by
disrupted and larger-than-expected demand, cop- 4.9percent from February 2017 to August 2017, with
per prices rebounded since June. In August, further the sub-indices of food, beverages, and agricultural raw
boosted by Chinas possible ban by the end of 2018 materials decreasing by 4.3percent, 4.3percent, and
on imports of scrap metals, copper stood 9.2percent 6.9percent, respectively. The decline has been fairly
higher than in February, reaching its highest level since uniform across different food groups as well; the indi-
November 2014. The partial resumption of ore exports ces for cereals lost 4.0percent, sugar 27.5percent, veg-
from Indonesia had also put downward pressure on etable oils 6.5percent, and beverages 4.3percent, with
nickel prices in the first half of 2017. Then, buoyed by only the index for meat seeing gains, of 6.3percent.
solid demand for stainless steel, particularly in Chinas Wheat prices decreased by 5.6percent from Febru-
construction sector, the price of nickel experienced a ary 2017 to August 2017. As hot, dry weather on the
strong recovery through July and was up by 2.3per- US Great Plains and in France raised doubts about
cent in August compared with February. yields in the Northern Hemisphere, prices increased
Aluminum prices increased by 9.1percent from sharply in June. The price rally was followed, how-
February 2017 to August 2017, supported by a global ever, by a 20.3percent decline, month-on-month,
shortage outside of China that, according to data in August, after the United States Department of
from the World Bureau of Metal Statistics, began in Agriculture unexpectedly raised its forecast of grain
the fall of 2016. By mid-August 2017, London Metal stocks at the end of the 201718 season for reasons
Exchange warehouse inventories of aluminum were that include prospects of a record upcoming Black Sea
44percent lower than in mid-January, hitting their harvest of wheat.
lowest point since 2008. On top of the increase so far, Maize prices declined too, by 8.8percent. Weather
futures prices are pointing to a sharp rise in prices, in the corn-growing regions of the United States did
likely fueled by expectations that China will cut its not affect prices much, and corn supplies, includ-
production capacity because of environmental con- ing from other major producing countries in South
cerns. Zinc rallied by 4.8percent between February America, remain abundant. Soybean prices trended
and August to a near 10-year high, following stock downward from February because supply from South
reductions, tight supplies and strong demand for steel America remains plentiful following a record harvest in
galvanization, especially from Chinese infrastructure Brazil, even though a stronger real discourages farmers
development. from selling their produce. Prospects of a relatively
The IMF metal price index is projected to rise large upcoming US soybean crop increased on good
briefly in the second half of 2017, followed by a gentle weather conditions in the critical growing month of
decline. The annual index for 2017 is expected to August, also putting downward pressure on prices.
increase by 20.6percent from its 2016 level, reflecting Palm oil prices fell by 12.0percent from February
the earlier surge this year, while futures are pointing 2017 to August 2017, as production in Malaysia and
to a slight decline throughout 2018, with the current Indonesia continued to rebound from the 201516
projection for the fourth quarter of 2018 0.4percent El Nio, and are expected to increase further, partly
below the level for the third quarter of 2017. because of seasonal factors. Indeed, palm oil future
Downside risks to the outlook for metal prices curves remain in backwardation, indicating that sup-
include credit tightening and a slowing down of Chi- ply is expected to be relatively more abundant in the
nas property market, which consumes more than half future. With China continuing to sell off its reserves,
of the worlds metal production. However, the Caixin and the upcoming US crop not severely affected by
Manufacturing Purchasing Managers Index increased Hurricane Harvey, cotton prices declined by 6.8per-
to 51.6 in August, indicating further expansion of cent between February 2017 and August 2017. Fur-
the worlds biggest manufacturing sector in the near thermore, output in the 201718 season is expected
term. Upside risks also include vigorous capacity cuts to be buoyant in major producers, including China,
in China and the possibility of greater restrictions on India, Pakistan, and the United States.
international trade, such as those potentially aris- Pork prices increased substantially up to July amid
ing from the US Section232 investigations for steel stronger demand and tighter supplies. Following
and aluminum. increases in global supplies, prices have slumped,
although they still stood 10.1percent higher in August the end of the 201718 season increased in August.
compared to February this year (based on monthly Annual food prices are now expected to increase
averages). While supplies are expected to increase by 3.6percent in2017 and an additional 1.1per-
further in the second half of 2017, strong global cent in 2018. Food prices are expected to decline
demand implies that markets are expected to clear at slightly again for the years thereafter for reasons that
higher year-over-year prices. Similarly, the price of include potentially better supply conditions for some
beef climbed steadily, by 2.4percent, because export commodities.
demand for red meat was stronger than expected and Weather disruptions and variability are an upside
leaner cattle contributed to weaker US supply growth. risk to the forecast for agricultural prices. As of
As the number of cattle on US feedlots has increased September 2017, there is an increasing chance (about
unexpectedly during summer, prices are expected to 55percent to 60percent) of a La Nia onset during
soften in the second half of this year. the Northern Hemisphere fall and winter of 201718.
Projections for grain prices have been revised sub- The increased use by governments of agricultural
stantially downward because concerns over hot, dry support policies is another upside risk. Downside risks
weather that sparked a rally in grain markets in June may arise if China sells more than anticipated from its
this year have waned, and forecasts for grain stocks at large reserves of grains, sugar, and cotton.
Annex Table 1.1.1. European Economies: Real GDP, Consumer Prices, Current Account Balance, and Unemployment
(Annual percent change, unless noted otherwise)
Real GDP Consumer Prices1 Current Account Balance2 Unemployment3
Projections Projections Projections Projections
2016 2017 2018 2016 2017 2018 2016 2017 2018 2016 2017 2018
Europe 2.1 2.5 2.2 0.9 2.5 2.4 2.2 2.4 2.3 ... ... ...
Advanced Europe 1.8 2.1 1.9 0.4 1.6 1.6 2.7 2.9 2.9 8.7 7.9 7.6
Euro Area4,5 1.8 2.1 1.9 0.2 1.5 1.4 3.5 3.1 3.0 10.0 9.2 8.7
Germany 1.9 2.0 1.8 0.4 1.6 1.5 8.3 8.1 7.7 4.2 3.8 3.7
France 1.2 1.6 1.8 0.3 1.2 1.3 1.0 1.1 0.8 10.0 9.5 9.0
Italy 0.9 1.5 1.1 0.1 1.4 1.2 2.6 2.7 2.3 11.7 11.4 11.0
Spain 3.2 3.1 2.5 0.2 2.0 1.5 1.9 1.9 2.0 19.6 17.1 15.6
Netherlands 2.2 3.1 2.6 0.1 1.3 1.4 8.5 10.0 10.0 5.9 5.1 4.9
Belgium 1.2 1.6 1.6 1.8 2.2 1.5 0.4 0.3 0.0 7.9 7.5 7.3
Austria 1.5 2.3 1.9 1.0 1.6 1.8 1.7 2.1 2.2 6.0 5.4 5.3
Greece 0.0 1.8 2.6 0.0 1.2 1.3 0.6 0.2 0.1 23.6 22.3 20.7
Portugal 1.4 2.5 2.0 0.6 1.6 2.0 0.7 0.4 0.3 11.1 9.7 9.0
Ireland 5.1 4.1 3.4 0.2 0.4 1.5 3.3 3.4 3.5 7.9 6.4 5.9
Finland 1.9 2.8 2.3 0.4 0.8 1.2 1.1 0.4 0.4 8.8 8.7 8.1
Slovak Republic 3.3 3.3 3.7 0.5 1.2 1.4 0.7 0.3 0.2 9.6 8.1 7.5
Lithuania 2.3 3.5 3.5 0.7 3.5 2.0 0.9 1.6 1.4 7.9 7.0 6.5
Slovenia 3.1 4.0 2.5 0.1 1.6 1.8 5.2 5.0 4.9 8.0 6.8 6.4
Luxembourg 4.2 3.9 3.6 0.0 1.2 1.3 4.7 4.7 4.9 6.4 5.9 5.5
Latvia 2.0 3.8 3.9 0.1 3.0 3.0 1.5 0.3 1.5 9.6 9.0 8.7
Estonia 2.1 4.0 3.7 0.8 3.8 3.4 1.9 1.8 1.4 6.8 8.4 9.0
Cyprus 2.8 3.4 2.6 1.2 0.8 0.7 5.3 3.8 2.7 13.0 11.8 10.7
Malta 5.5 5.1 4.4 0.9 1.3 1.6 7.9 8.9 8.8 4.7 4.4 4.5
United Kingdom5 1.8 1.7 1.5 0.7 2.6 2.6 4.4 3.6 3.3 4.9 4.4 4.4
Switzerland 1.4 1.0 1.3 0.4 0.5 0.6 10.5 9.9 9.4 3.3 3.0 3.0
Sweden 3.2 3.1 2.4 1.1 1.6 1.6 4.5 3.9 3.7 7.0 6.6 6.3
Norway 1.1 1.4 1.6 3.6 2.1 2.0 5.0 5.5 5.7 4.7 4.0 3.8
Czech Republic 2.6 3.5 2.6 0.7 2.3 1.8 1.1 0.6 0.1 4.0 2.8 3.0
Denmark 1.7 1.9 1.8 0.3 1.0 1.4 7.9 7.3 7.0 6.2 5.8 5.8
Iceland 7.2 5.5 3.3 1.7 1.8 2.6 7.9 6.2 6.1 3.0 2.8 3.2
San Marino 1.0 1.2 1.3 0.6 0.9 1.0 ... ... ... 8.6 8.0 7.4
Emerging and Developing Europe6 3.1 4.5 3.5 3.3 6.0 5.7 1.8 2.4 2.5 ... ... ...
Turkey 3.2 5.1 3.5 7.8 10.9 9.3 3.8 4.6 4.6 10.9 11.2 10.7
Poland 2.6 3.8 3.3 0.6 1.9 2.3 0.2 1.0 1.2 6.2 4.8 4.0
Romania 4.8 5.5 4.4 1.6 1.1 3.3 2.3 3.0 2.9 5.9 5.3 5.2
Hungary 2.0 3.2 3.4 0.4 2.5 3.2 5.5 4.8 4.2 5.1 4.4 4.3
Bulgaria5 3.4 3.6 3.2 1.3 1.1 1.4 4.2 2.5 1.9 7.7 6.6 6.4
Serbia 2.8 3.0 3.5 1.1 3.4 3.0 4.0 4.0 3.9 15.9 16.0 15.6
Croatia 3.0 2.9 2.7 1.1 1.1 1.2 2.6 3.8 3.0 15.0 13.9 13.5
Note: Data for some countries are based on fiscal years. Refer to Table F in the Statistical Appendix for a list of economies with exceptional reporting periods.
1Movements in consumer prices are shown as annual averages. Year-end to year-end changes can be found in Tables A6 and A7 in the Statistical Appendix.
2Percent of GDP.
3Percent. National definitions of unemployment may differ.
4Current account position corrected for reporting discrepancies in intra-area transactions.
5Based on Eurostats harmonized index of consumer prices except for Slovenia.
6Includes Albania, Bosnia and Herzegovina, Kosovo, FYR Macedonia, and Montenegro.
Annex Table 1.1.2. Asian and Pacific Economies: Real GDP, Consumer Prices, Current Account Balance, and Unemployment
(Annual percent change, unless noted otherwise)
Real GDP Consumer Prices1 Current Account Balance2 Unemployment3
Projections Projections Projections Projections
2016 2017 2018 2016 2017 2018 2016 2017 2018 2016 2017 2018
Asia 5.4 5.6 5.5 2.3 2.3 2.8 2.5 2.1 1.9 ... ... ...
Advanced Asia 1.7 2.2 1.7 0.5 1.0 1.2 4.5 4.3 4.2 3.6 3.4 3.4
Japan 1.0 1.5 0.7 0.1 0.4 0.5 3.8 3.6 3.8 3.1 2.9 2.9
Korea 2.8 3.0 3.0 1.0 1.9 1.9 7.0 5.6 5.4 3.7 3.8 3.6
Australia 2.5 2.2 2.9 1.3 2.0 2.2 2.6 1.6 2.4 5.7 5.6 5.4
Taiwan Province of China 1.5 2.0 1.9 1.4 1.0 1.4 14.0 13.8 13.9 3.9 3.8 3.8
Singapore 2.0 2.5 2.6 0.5 0.9 1.3 19.0 19.6 19.5 2.1 2.2 2.1
Hong Kong SAR 2.0 3.5 2.7 2.6 2.0 2.2 4.6 3.0 3.1 2.7 2.6 2.6
New Zealand 3.6 3.5 3.0 0.6 2.2 2.0 2.8 3.6 3.8 5.1 4.9 4.6
Macao SAR 2.1 13.4 7.0 2.4 1.5 2.2 27.4 33.0 34.5 1.9 2.0 2.0
Emerging and Developing Asia 6.4 6.5 6.5 2.8 2.6 3.2 1.4 0.9 0.7 ... ... ...
China 6.7 6.8 6.5 2.0 1.8 2.4 1.7 1.4 1.2 4.0 4.0 4.0
India4 7.1 6.7 7.4 4.5 3.8 4.9 0.7 1.4 1.5 ... ... ...
ASEAN-5 4.9 5.2 5.2 2.4 3.3 3.1 2.1 1.6 1.1 ... ... ...
Indonesia 5.0 5.2 5.3 3.5 4.0 3.9 1.8 1.7 1.8 5.6 5.4 5.2
Thailand 3.2 3.7 3.5 0.2 0.6 1.0 11.5 10.1 8.1 0.8 0.7 0.7
Malaysia 4.2 5.4 4.8 2.1 3.8 2.9 2.4 2.4 2.2 3.5 3.4 3.2
Philippines 6.9 6.6 6.7 1.8 3.1 3.0 0.2 0.1 0.3 5.5 6.0 5.5
Vietnam 6.2 6.3 6.3 2.7 4.4 4.0 4.1 1.3 1.4 2.3 2.3 2.3
Other Emerging and Developing
Asia5 5.6 6.3 6.3 5.2 5.5 5.4 0.9 1.9 2.5 ... ... ...
Memorandum
Emerging Asia6 6.5 6.5 6.5 2.7 2.5 3.1 1.5 1.0 0.8 ... ... ...
Note: Data for some countries are based on fiscal years. Refer to Table F in the Statistical Appendix for a list of economies with exceptional reporting periods.
1Movements in consumer prices are shown as annual averages. Year-end to year-end changes can be found in Tables A6 and A7 in the Statistical Appendix.
2Percent of GDP.
3Percent. National definitions of unemployment may differ.
4See country-specific notes for India in the Country Notes section of the Statistical Appendix.
5Other Emerging and Developing Asia comprises Bangladesh, Bhutan, Brunei Darussalam, Cambodia, Fiji, Kiribati, Lao P.D.R., Maldives, Marshall Islands, Micronesia, Mon-
golia, Myanmar, Nauru, Nepal, Palau, Papua New Guinea, Samoa, Solomon Islands, Sri Lanka, Timor-Leste, Tonga, Tuvalu, and Vanuatu.
6Emerging Asia comprises the ASEAN-5 (Indonesia, Malaysia, Philippines, Thailand, Vietnam) economies, China, and India.
Annex Table 1.1.3. Western Hemisphere Economies: Real GDP, Consumer Prices, Current Account Balance, and Unemployment
(Annual percent change, unless noted otherwise)
Real GDP Consumer Prices1 Current Account Balance2 Unemployment3
Projections Projections Projections Projections
2016 2017 2018 2016 2017 2018 2016 2017 2018 2016 2017 2018
North America 1.5 2.2 2.2 1.4 2.4 2.3 2.5 2.4 2.6 ... ... ...
United States 1.5 2.2 2.3 1.3 2.1 2.1 2.4 2.4 2.6 4.9 4.4 4.1
Canada 1.5 3.0 2.1 1.4 1.6 1.8 3.3 3.4 2.9 7.0 6.5 6.3
Mexico 2.3 2.1 1.9 2.8 5.9 3.8 2.2 1.7 2.0 3.9 3.6 3.7
Puerto Rico4 2.6 2.8 2.5 0.3 1.1 0.9 ... ... ... 11.8 11.5 11.6
South America5 2.6 0.6 1.6 ... ... ... 1.8 1.9 2.3 ... ... ...
Brazil 3.6 0.7 1.5 8.7 3.7 4.0 1.3 1.4 1.8 11.3 13.1 11.8
Argentina 2.2 2.5 2.5 ... 26.9 17.8 2.7 3.6 3.7 8.5 8.1 7.7
Colombia 2.0 1.7 2.8 7.5 4.3 3.3 4.3 3.8 3.6 9.2 9.3 9.2
Venezuela 16.5 12.0 6.0 254.4 652.7 2,349.3 1.6 0.4 1.3 20.6 26.4 29.8
Chile 1.6 1.4 2.5 3.8 2.3 2.7 1.4 2.3 2.8 6.5 7.0 6.8
Peru 4.0 2.7 3.8 3.6 3.2 2.3 2.7 1.5 1.6 6.7 6.7 6.7
Ecuador 1.5 0.2 0.6 1.7 0.7 0.7 1.4 0.7 1.6 5.2 5.1 5.3
Bolivia 4.3 4.2 4.0 3.6 3.2 5.1 5.7 4.7 4.8 4.0 4.0 4.0
Uruguay 1.5 3.5 3.1 9.6 6.1 6.3 0.1 0.4 0.8 7.9 7.3 7.3
Paraguay 4.1 3.9 4.0 4.1 3.5 4.0 1.7 1.1 0.4 6.0 6.5 6.2
Central America6 3.7 3.8 3.9 2.1 2.8 3.2 2.9 2.9 2.8 ... ... ...
Caribbean7 3.4 2.8 4.4 2.6 3.8 3.8 4.1 4.1 4.3 ... ... ...
Memorandum
Latin America and the Caribbean8 0.9 1.2 1.9 5.6 4.2 3.6 2.0 2.0 2.3 ... ... ...
East Caribbean Currency Union9 2.6 2.6 2.8 0.7 1.3 1.4 5.4 6.6 7.4 ... ... ...
Note: Data for some countries are based on fiscal years. Refer to Table F in the Statistical Appendix for a list of economies with exceptional reporting periods.
1Movements in consumer prices are shown as annual averages. Year-end to year-end changes can be found in Tables A6 and A7 in the Statistical Appendix.
2Percent of GDP.
3Percent. National definitions of unemployment may differ.
4Puerto Rico is a territory of the United States but its statistical data are maintained on a separate and independent basis.
5Includes Guyana and Suriname. Data for Argentinas and Venezuelas consumer prices are excluded. See country-specific notes for Argentina and Venezuela in the
consumer prices are excluded. See country-specific notes for Argentina and Venezuela in the Country Notes section of the Statistical Appendix.
9Eastern Caribbean Currency Union comprises Antigua and Barbuda, Dominica, Grenada, St. Kitts and Nevis, St. Lucia, and St. Vincent and the Grenadines as well as
Annex Table 1.1.4. Commonwealth of Independent States Economies: Real GDP, Consumer Prices, Current Account
Balance, and Unemployment
(Annual percent change, unless noted otherwise)
Real GDP Consumer Prices1 Current Account Balance2 Unemployment3
Projections Projections Projections Projections
2016 2017 2018 2016 2017 2018 2016 2017 2018 2016 2017 2018
Commonwealth of Independent States4 0.4 2.1 2.1 8.3 5.8 5.2 0.0 0.9 1.3 ... ... ...
Net Energy Exporters 0.3 2.1 2.0 7.9 5.2 4.7 0.5 1.6 2.0 ... ... ...
Russia 0.2 1.8 1.6 7.0 4.2 3.9 2.0 2.8 3.2 5.5 5.5 5.5
Kazakhstan 1.1 3.3 2.8 14.6 7.3 6.5 6.4 5.3 3.8 5.0 5.0 5.0
Uzbekistan 7.8 6.0 6.0 8.0 13.0 12.7 0.7 0.9 0.3 ... ... ...
Azerbaijan 3.1 1.0 1.3 12.4 12.0 8.0 3.6 1.9 2.5 6.0 6.0 6.0
Turkmenistan 6.2 6.5 6.3 3.6 6.0 6.2 21.0 15.4 14.3 ... ... ...
Net Energy Importers 1.2 2.1 2.7 11.0 10.0 8.3 4.7 4.9 4.5 ... ... ...
Ukraine 2.3 2.0 3.2 13.9 12.8 10.0 4.1 3.3 3.0 9.3 9.5 9.3
Belarus 2.6 0.7 0.7 11.8 8.0 7.5 3.6 5.3 4.6 1.0 1.0 1.0
Georgia 2.7 4.0 4.2 2.1 6.0 3.0 13.3 11.9 10.7 11.8 ... ...
Armenia 0.2 3.5 2.9 1.4 1.9 3.5 2.3 3.6 3.2 18.8 18.9 18.9
Tajikistan 6.9 4.5 4.0 5.9 8.9 8.0 3.8 6.3 6.2 ... ... ...
Kyrgyz Republic 3.8 3.5 3.8 0.4 3.8 5.1 9.7 11.6 12.0 7.5 7.4 7.3
Moldova 4.3 4.0 3.7 6.4 6.5 5.3 3.8 4.0 4.0 4.2 4.3 4.2
Memorandum
Caucasus and Central Asia5 2.5 3.6 3.7 10.4 8.8 7.8 6.4 4.9 4.2 ... ... ...
Low-Income CIS Countries6 6.1 5.2 5.2 5.8 10.0 9.6 2.5 2.7 3.1 ... ... ...
Net Energy Exporters Excluding Russia 2.4 3.5 3.7 11.6 9.3 8.2 6.2 4.4 3.6 ... ... ...
Note: Data for some countries are based on fiscal years. Refer to Table F in the Statistical Appendix for a list of economies with exceptional reporting periods.
1Movements in consumer prices are shown as annual averages. Year-end to year-end changes can be found in Table A7 in the Statistical Appendix.
2Percent of GDP.
3Percent. National definitions of unemployment may differ.
4Georgia, Turkmenistan, and Ukraine, which are not members of the Commonwealth of Independent States (CIS), are included in this group for reasons of geography and
Annex Table 1.1.5. Middle East, North African Economies, Afghanistan, and Pakistan: Real GDP, Consumer Prices, Current
Account Balance, and Unemployment
(Annual percent change, unless noted otherwise)
Real GDP Consumer Prices1 Current Account Balance2 Unemployment3
Projections Projections Projections Projections
2016 2017 2018 2016 2017 2018 2016 2017 2018 2016 2017 2018
Middle East, North Africa, Afghanistan,
and Pakistan 5.0 2.6 3.5 5.1 6.8 7.7 4.1 1.9 1.6 ... ... ...
Oil Exporters4 5.6 1.7 3.0 4.6 4.3 6.0 3.6 0.4 0.2 ... ... ...
Saudi Arabia 1.7 0.1 1.1 3.5 0.2 5.0 4.3 0.6 0.4 5.6 ... ...
Iran 12.5 3.5 3.8 9.0 10.5 10.1 4.1 5.1 5.9 12.5 12.4 12.4
United Arab Emirates 3.0 1.3 3.4 1.8 2.1 2.9 2.4 2.1 2.1 ... ... ...
Algeria 3.3 1.5 0.8 6.4 5.5 4.4 16.5 13.0 10.8 10.5 11.7 13.2
Iraq 11.0 0.4 2.9 0.4 2.0 2.0 8.7 6.3 6.7 ... ... ...
Qatar 2.2 2.5 3.1 2.7 0.9 4.8 4.9 2.3 1.0 ... ... ...
Kuwait 2.5 2.1 4.1 3.5 2.5 2.7 4.5 0.6 1.4 2.1 2.1 2.1
Oil Importers5 3.6 4.3 4.4 6.2 12.1 11.2 5.3 5.3 4.8 ... ... ...
Egypt 4.3 4.1 4.5 10.2 23.5 21.3 6.0 5.9 3.8 12.7 12.2 11.5
Pakistan 4.5 5.3 5.6 2.9 4.1 4.8 1.7 4.0 4.9 6.0 6.0 6.1
Morocco 1.2 4.8 3.0 1.6 0.9 1.6 4.4 4.0 2.9 9.4 9.3 9.5
Sudan 3.0 3.7 3.6 17.8 26.9 19.0 5.6 1.9 2.0 20.6 19.6 18.6
Tunisia 1.0 2.3 3.0 3.7 4.5 4.4 9.0 8.7 8.4 14.0 13.0 12.0
Lebanon 1.0 1.5 2.0 0.8 3.1 2.5 18.6 18.0 16.8 ... ... ...
Jordan 2.0 2.3 2.5 0.8 3.3 1.5 9.3 8.4 8.3 15.3 ... ...
Memorandum
Middle East and North Africa 5.1 2.2 3.2 5.4 7.1 8.1 4.4 1.7 1.3 ... ... ...
Israel6 4.0 3.1 3.4 0.5 0.2 0.5 3.6 4.1 3.1 4.8 4.3 4.5
Maghreb7 2.2 5.4 3.8 5.4 5.4 5.4 12.1 8.5 5.6 ... ... ...
Mashreq8 3.9 3.8 4.2 8.7 20.7 18.7 7.8 8.2 6.4 ... ... ...
Note: Data for some countries are based on fiscal years. Refer to Table F in the Statistical Appendix for a list of economies with exceptional reporting periods.
1Movements in consumer prices are shown as annual averages. Year-end to year-end changes can be found in Tables A6 and A7 in the Statistical Appendix.
2Percent of GDP.
3Percent. National definitions of unemployment may differ.
4Includes Bahrain, Libya, Oman, and Yemen.
5Includes Afghanistan, Djibouti, Mauritania, and Somalia. Excludes Syria because of the uncertain political situation.
6Israel, which is not a member of the economic region, is included for reasons of geography but is not included in the regional aggregates.
7The Maghreb comprises Algeria, Libya, Mauritania, Morocco, and Tunisia.
8The Mashreq comprises Egypt, Jordan, and Lebanon. Syria is excluded because of the uncertain political situation.
Annex Table 1.1.6. Sub-Saharan African Economies: Real GDP, Consumer Prices, Current Account Balance, and Unemployment
(Annual percent change, unless noted otherwise)
Real GDP Consumer Prices1 Current Account Balance2 Unemployment3
Projections Projections Projections Projections
2016 2017 2018 2016 2017 2018 2016 2017 2018 2016 2017 2018
Sub-Saharan Africa 1.4 2.6 3.4 11.3 11.0 9.5 4.2 3.4 3.6 ... ... ...
Oil Exporters4 1.9 0.6 1.6 18.8 18.1 14.7 2.0 0.3 0.6 ... ... ...
Nigeria 1.6 0.8 1.9 15.7 16.3 14.8 0.7 1.9 1.0 13.4 ... ...
Angola 0.7 1.5 1.6 32.4 30.9 20.6 5.1 4.8 4.5 ... ... ...
Gabon 2.1 1.0 2.7 2.1 2.5 2.5 10.2 9.3 6.7 ... ... ...
Chad 6.4 0.6 2.4 1.1 0.2 1.9 9.2 2.0 2.8 ... ... ...
Republic of Congo 2.8 3.6 2.8 3.6 0.4 1.1 70.1 15.9 2.5 ... ... ...
Middle-Income Countries5 2.0 2.5 3.2 6.8 5.3 5.1 3.4 3.2 3.5 ... ... ...
South Africa 0.3 0.7 1.1 6.3 5.4 5.3 3.3 2.9 3.3 26.7 27.6 28.3
Ghana 3.5 5.9 8.9 17.5 11.8 9.0 6.7 5.8 5.4 ... ... ...
Cte dIvoire 7.7 7.6 7.3 0.7 1.0 2.0 1.1 2.9 2.8 ... ... ...
Cameroon 4.7 4.0 4.6 0.9 0.7 1.1 3.6 3.6 3.5 ... ... ...
Zambia 3.4 4.0 4.5 17.9 6.8 7.4 4.4 3.6 2.8 ... ... ...
Senegal 6.7 6.8 7.0 0.9 2.1 2.2 5.3 5.1 5.2 ... ... ...
Low-Income Countries6 5.3 5.6 5.9 6.6 8.8 8.2 8.3 7.9 8.3 ... ... ...
Ethiopia 8.0 8.5 8.5 7.3 8.1 8.0 9.9 8.3 7.4 ... ... ...
Kenya 5.8 5.0 5.5 6.3 8.0 5.2 5.2 6.1 7.0 ... ... ...
Tanzania 7.0 6.5 6.8 5.2 5.4 5.0 5.6 5.6 6.5 ... ... ...
Uganda 2.3 4.4 5.2 5.5 5.8 5.6 4.3 5.6 7.2 ... ... ...
Madagascar 4.2 4.3 5.3 6.7 7.8 6.8 0.8 4.7 5.3 ... ... ...
Democratic Republic of the Congo 2.4 2.8 3.0 18.2 41.7 44.0 3.4 4.6 2.1 ... ... ...
Memorandum
Sub-Saharan Africa Excluding South
Sudan 1.5 2.7 3.4 10.4 10.5 9.3 4.2 3.4 3.6 ... ... ...
Note: Data for some countries are based on fiscal years. Refer to Table F in the Statistical Appendix for a list of economies with exceptional reporting periods.
1Movements in consumer prices are shown as annual averages. Year-end to year-end changes can be found in Table A7 in the Statistical Appendix.
2Percent of GDP.
3Percent. National definitions of unemployment may differ.
4Includes Equatorial Guinea and South Sudan.
5Includes Botswana, Cabo Verde, Lesotho, Mauritius, Namibia, Seychelles, and Swaziland.
6Includes Benin, Burkina Faso, Burundi, the Central African Republic, Comoros, Eritrea, The Gambia, Guinea, Guinea-Bissau, Liberia, Malawi, Mali, Mozambique, Niger,
Feyrer, James. 2007. Demographics and Productivity. Review of IIF Working Paper, Institute of International Finance,
Economics and Statistics 89 (1): 10009. Washington, DC.
Frankel, Jeffrey. 2011. Are Bilateral Remittances Countercycli- Koopman, Robert, Zhi Wang, and Shang-Jin Wei. 2014. Trac-
cal? Open Economies Review 22 (1):116. ing Value-Added and Double Counting in Gross Exports.
Giuliano, Paola, and Marta Ruiz-Arranz. 2009. Remittances, American Economic Review 104 (2): 45994.
Financial Development, and Growth. Journal of Development Kose, M. Ayhan, Eswar S. Prasad, and Marco E. Terrones. 2009.
Economics 90 (1): 14452. Does Financial Globalization Promote Risk Sharing? Jour-
Gruss, Bertrand. 2014. After the BoomCommodity Prices nal of Development Economics 89 (2): 25870.
and Economic Growth in Latin America and the Caribbean. Krause, Eleanor, and Isabel Sawhill. 2017. What We Know and
IMF Working Paper 14/154, International Monetary Fund, Dont Know about Declining Labor Force Participation: A
Washington, DC. Review. Center on Children and Families at Brookings, The
Hadzi-Vaskov, Metodij. 2006. Workers Remittances and Brookings Institution, Washington, DC.
International Risk-Sharing. Tjalling C. Koopmans Laeven, Luc, and Fabian Valencia. 2008. Systemic Banking
Research Institute Discussion Paper 0619, Univer- Crises: A New Database. IMF Working Paper 08/224, Inter-
sity of Utrecht. national Monetary Fund, Washington, DC.
International Monetary Fund (IMF). 2013a. Macroeconomic . 2010. Resolution of Banking Crises: The Good, the
Issues in Small States: Implications for Fund Engagement. Bad, and the Ugly. IMF Working Paper 10/44, International
Washington, DC. Monetary Fund, Washington, DC.
. 2013b. Staff Guidance Note on the Application of . 2012. Systemic Banking Crises Database: An Update.
the Joint Bank-Fund Debt Sustainability Framework for IMF Working Paper 12/163, International Monetary Fund,
Low-Income Countries. Washington, DC. Washington, DC.
. 2014. Is It Time for an Infrastructure Push? The Mac- Lakner, Christoph, and Branko Milanovic. 2015. Global
roeconomic Effects of Public Investment. Chapter3 of the Income Distribution: From the Fall of the Berlin Wall
October World Economic Outlook, Washington, DC. to the Great Recession. World Bank Economic Review 30
. 2014. Macroeconomic Developments in Low- (2): 20332.
Income Developing Countries. IMF Policy Paper, Lane, Philip R., and Gian Maria Milesi-Ferretti. 2017. Inter-
Washington, DC. national Financial Integration in the Aftermath of the Global
. 2016a. External Sector Report. Washington, DC. Financial Crisis. IMF Working Paper 17/115, International
. 2016b. Small States Resilience to Natural Disasters Monetary Fund, Washington, DC.
and Climate ChangeRole of the IMF. IMF Policy Paper, Lewis, Karen. 1996. What Can Explain the Apparent Lack of
Washington, DC. International Consumption Risk Sharing? Journal of Political
. 2017a. Article IV Consultation with the United States Economy 104 (2): 26797.
of America. Concluding Statement of the IMF Mission, McCauley, Robert N., and Chang Shu. 2016. Dollars and
Washington, DC. Renminbi Flowed out of China. Bank for International Settle-
. 2017b. External Adjustment to Terms-of-Trade Shifts. ments Quarterly Review, March 2016.
Chapter3 of the Western Hemisphere Regional Economic Out- Obstfeld, Maurice. 1993. Are Industrial-Country Consumption
look. April, Washington, DC. Risks Globally Diversified? NBER Working Paper 4308,
. 2017c. External Sector Report. Washington, DC. National Bureau of Economic Research, Cambridge, MA.
. 2017d. Forthcoming. G20 Report on Strong, Sustain- Prasad, Eswar, Kenneth Rogoff, Shang-Jin Wei, and M.
able, and Balanced Growth. Washington, DC. Ayhan Kose. 2003. Effects of Financial Globalization on
, World Bank, and World Trade Organization. 2017. Developing Countries: Some Empirical Evidence. IMF
Making Trade an Engine of Growth for All. Policy Papers, Occasional Paper 220, International Monetary Fund,
Washington, DC. Washington, DC.
Jongwanich, Juthathip. 2007. Workers Remittances, Eco- Rajan, Raghuram, and Arvind Subramanian. 2005. What
nomic Growth and Poverty in Developing Asia and the Undermines Aids Impact on Growth? IMF Working Paper
Pacific Countries. UNESCAP Working Paper 07/01, United 05/126, International Monetary Fund, Washington, DC.
Nations Economic and Social Commission for Asia and the Ratha, Dilip. 2003. Workers Remittances: An Important and
Pacific, Bangkok. Stable Source of External Development Finance. In Global
Kalemli-Ozcan, Sebnem, Bent E. Sorensen, and Oved Yosha. Development Finance 2003: Striving for Stability in Develop-
2003. Risk Sharing and Industrial Specialization: Regional ment Finance. Washington, DC: World Bank.
and International Evidence. American Economic Review 93 Sorensen, Bent, Yi-Tsung Wu, Oved Yosha, and Yu Zhu. 2005.
(3): 90318. Home Bias and International Risk Sharing: Twin Puzzles
Koepke, Robin. 2015. What Drives Capital Flows to Separated at Birth. CEPR Discussion Paper 5113, Centre for
Emerging Markets? A Survey of the Empirical Literature. Economic Policy Research, London.
Wacziarg, Romain, and Karen Horn Welch. 2008. Trade Yi, Kei-Mu. 2003. Can Vertical Specialization Explain the Growth
Liberalization and Growth: New Evidence. NBER Work- in World Trade? Journal of Political Economy 111 (1): 52102.
ing Paper 10152, National Bureau of Economic Research, . 2010. Can Multistage Production Explain the Home
Cambridge, MA. Bias in Trade? American Economic Review 100 (1): 36493.
Nominal wage growth in most advanced economies unsustainably high wage growth prior to the Great
remains markedly lower than it was before the Great Recession. The pattern, however, is more widespread.
Recession of 200809. This chapter finds that the bulk Nominal wage dynamics, in general, are related to
of the wage slowdown can be explained by labor market underlying changes in a real componentphysical
slack (both headline unemployment and underutili- output created by labor together with other inputs
zation of labor in the form of involuntary part-time into productionas well as inflation pressure in the
employment), inflation expectations, and trend produc- economy. Viewed through this lens, subdued nominal
tivity growth. While involuntary part-time employment wage growth is, in principle, consistent with a widely
may have helped support labor force participation and recognized slowdown in labor productivity, which can
facilitated stronger engagement with the workplace weigh on underlying real wage dynamics, and generally
than the alternative of unemployment, it also appears low inflation across advanced economies.1
to have weakened wage growth. This is the case even Subdued nominal wage growth has also generally
in economies where measured slack appears low (that coincided with a reduction in hours per worker and,
is, headline unemployment rates are now at, or below, in some cases, a higher rate of involuntary part-time
their averages in the years leading up to the recession). employment and an increased share of temporary
Common factorsbeyond slack, productivity, and price employment contracts. Headline unemployment
inflationhave also exerted downward pressure on wages measures are therefore not as indicative of labor market
in recent years, suggesting that the synchronized nature slack, given this increase in part-time employment and
of excess capacity across countries may have amplified its temporary contracts. These developments may also
effects. While accommodative policies can help lift demand point to persistent changes in the nature of employ-
and lower headline unemployment rates, wage growth ment relationships between firms and workers in
may continue to remain subdued until involuntary response to technological change and remaining labor
part-time employment diminishes or trend productivity market rigidities in some countries that deter employ-
growth picks up. Inflation rates will also likely remain ers from hiring on standard, full-time contracts.2
low unless wage growth accelerates beyond productiv- From a macroeconomic perspective, shedding light
ity growth in a sustained manner. Assessing the true on the forces shaping nominal wage developments
degree of slack beyond measured headline unemployment could inform the debate on the extent of slack in
rates will be important when judging the appropriate the economy and the appropriate pace of exit from
pace of exit from accommodative monetary policies. accommodative monetary policies. As noted in Chap-
ter1, core inflation rates in most advanced economies
remain below targets and have not shown a steady
upswing even as growth has generally picked up over
Introduction the past year. With wages being the largest compo-
Close to a decade after the Great Recession of 200809, nent of most firms production costs, the upswing in
nominal wage growth in most advanced economies wages in response to falling unemployment is the main
remains markedly lower than it was before the recession. reason core inflation typically picks up as aggregate
This is the case even in countries where unemployment demand strengthens and excess capacity in the econ-
rates are now at, or even below, their averages in the
years leading up to the recession. In some instances,
1On the productivity slowdown, see Fernald (2014); Byrne, Fer-
recent wage dynamics may reflect a correction from nald, and Reinsdorf (2016); and Adler and others (2017). On weak
inflation rates in advanced economies, see Chapter3 of the October
The authors of this chapter are Gee Hee Hong, Zska Kczn, 2016 World Economic Outlook (WEO).
Weicheng Lian, and Malhar Nabar (team leader), with support from 2See Bentolila and others (2012) for a discussion of labor market
Benjamin Hilgenstock and Jungjin Lee. rigidities and the use of temporary contracts.
omy shrinks.3 Core inflation in advanced economies is suggests that common factors have been exerting
thus unlikely to recover in a sustained manner before increasing downward pressure on wage inflation
labor market tightening spurs higher wage inflation. In in the aftermath of the global financial crisis and
sum, a better understanding of the forces that weigh especially during 201416. For a number of euro
on wage growth is important for assessing the appro- area economies with large precrisis current account
priate course of monetary policy. deficits, this may reflect policy measures to slow
Insights into the drivers of wage dynamics and the wage growth and improve competitiveness in the
role of part-time employment and temporary contracts aftermath of the global financial crisis and euro area
may also offer perspective on prospects for income sovereign debt crisis.4 More broadly, the finding of
inequality and possible policy actions to address the sizable common factors behind wage weakness could
income security of workers with part-time jobs or indicate the growing effect on wage setting in any
temporary contracts. The latter could include tackling given economy of labor market conditions in other
slack, supporting retraining and reskilling, addressing countries (in the context of stronger cross-border
remaining labor market and structural rigidities, and economic integration). It could also point to the
ensuring fairness of treatment across employees under role of broad-based and synchronized demand weak-
various types of contracts. ness across many countries and heightened concern
Accordingly, the chapter addresses the following about job losses, which may have hindered wage
main questions: growth in the aftermath of the global financial crisis
Drivers: How well do aggregate macroeconomic and the euro area sovereign debt crisis.
factors such as labor market slack, inflation expec- The relative roles of labor market slack and produc-
tations, and trend labor productivity growth tivity growth vary across countries. In economies
account for nominal wage dynamics observed across where unemployment rates are still appreciably
advanced economies since the Great Recession? How above their averages before the Great Recession,
has the evolving mix of full-time versus involuntary conventional measures of labor market slack can
part-time employment and open-ended versus tem- explain about half of the slowdown in nominal
porary work contracts affected labor market slack wage growth since 2007, with involuntary part-time
and hence wage dynamics? employment acting as a further significant drag on
Underlying changes: How have changes in firms wages. Productivity growth is in turn relatively less
incentives and constraints in recent years (for important because these economies had generally
example, related to changing expectations about lower productivity growth to begin with, and less
medium-term growth prospects, technology, and of a slowdown.
global production processes) affected nominal wage In economies where unemployment rates are below
setting and part-time employment? What impact their averages before the Great Recession, slow
have shifts in bargaining power (arising, for exam- productivity growth can account for mostabout
ple, from changes in employment regulations, two-thirdsof the slowdown in nominal wage
unionization, and degree of import competition) growth since 2007. However, even here, involun-
had on wages and part-time employment? tary part-time employment appears to be weighing
on wage growth, suggesting greater slack in the
These are the main findings of the chapter: labor market than captured by headline unem-
Macroeconomic factors such as labor market slack ployment rates.
(both headline unemployment and underutiliza- Involuntary part-time employment has risen more
tion of labor in the form of involuntary part-time in countries where output is estimated to fall short
employment), inflation expectations, and trend of its potential. Once the influence of slack is taken
productivity growth can account for the bulk into account, involuntary part-time employment has
of the variation in nominal wage growth at the increased more where medium-term growth expec-
country level in recent years. The analysis also tations have fallen more, automation has progressed
faster, and the importance of services in the econ-
3As noted in Chapter1, the part of the wage-inflation weakening
omy has increased.
attributable to lower productivity growth would not translate into
weaker price inflation, given that the changes would have no net
effect on cost pressures (proxied by unit labor costs). 4Also see Kang and Shambaugh (2014).
The analysis suggests that while accommodative lay off workers. Average wage growth would then
policies can help lift demand and lower headline also weaken, and weakening inflation pressure would
unemployment rates, wage growth may continue to transmit back to weaker nominal wage growth. Thus,
remain subdued until involuntary part-time employ- two key cyclical factors associated with wages are
ment diminishes or trend productivity growth picks the degree of slack in the economy and inflation
up. Inflation rates will also likely remain low unless expectations.
wage growth accelerates beyond productivity growth During the past decadewith a deep and pro-
in a sustained manner. Assessing the true degree longed recession, and fewer and fewer workers working
of slack beyond measured headline unemployment full-timeother dimensions of labor underutilization
rates will be important when judging the appro- beyond the standard slack measure of the unemploy-
priate pace of exit from accommodative mone- ment rate also appear to have had a bearing on wages.5
tary policies. Recent studies have found, for example, evidence of
a negative impact of discouraged workers, or a rising
The next section presents a primer on the deter- share of part-time employment, on wages (Blanch-
minants of wage growth to help set the stage for the flower and Posen 2014; Smith 2014).6
empirical analysis. The chapter then takes stock of In addition to the business cycle, a key force shap-
changes in the labor markets of advanced economies ing average wage growth is trend labor productivity
over recent years. In subsequent sections, the forces growthincreases in the output produced by each
shaping nominal wage dynamics and employment hour of labor input in combination with other factors
outcomes at the aggregate level are assessed. The of production. From a firms perspective, as trend labor
chapter concludes with a discussion of the main policy productivity growth accelerates, the value of hiring addi-
implications to be drawn from the analysis. tional workers increases relative to the cost of expanding
the payroll.7 Greater demand for labor translates into
rising vacancies relative to jobseekers, and therefore
Wage DeterminationA Primer rising pressure on wages. Conversely, as productivity
Nominal wages are determined by the interaction growth weakens, all else equal, profitability declines,
between labor demand and supply, which are both along with firms ability to accommodate wage increases
subject to multiple, interrelated influences. It is useful for their existing workers or their willingness to attract
to categorize these as influences related to the business new workers with high wages. Thus, wage growth tends
cycle and forces that are slower moving (secular). to weaken as productivity growth slows. Wage rigidities
Over the business cycle, aggregate demand for
final output translates into labor demand. In the 5See Trigari (2014).
expansionary phase, employers increase labor input 6Altig and Higgins (2014) note the negative impact on wages of
to meet rising final demand. Rising demand for labor people working part-time for economic reasons. Other studies look
at whether the long-term unemployed affect wage dynamics as much
can result in a combination of more hours (including as the short-term unemployed (Stock 2011; Gordon 2013; Council
overtime), a decline in involuntary part-time employ- of Economic Advisers 2014; Krueger, Cramer, and Cho 2014;
ment, and an increase in the number of employed Rudebusch and Williams 2014; Watson 2014), partly motivated
by the fact that both price and wage inflation rates in the early
workers. Eventually, as demand continues to rise, the
aftermath of the Great Recession appeared more robust than would
pool of jobseekers (a combination of unemployed be predicted based on conventional price and wage Phillips curves.
plus currently employed workers who are searching These studies have generally noted a greater impact of short-term
for more attractive employment) shrinks relative to than of long-term unemployment. Others have noted, however, that
in the United States, for example, the long- and short-term unem-
vacancies, and employers pay more to attract work- ployment rates evolved closely together in the few decades preceding
ers or to retain those on the payroll. To the extent the Great Recession, and hence it can be difficult to disentangle their
that nominal wages are indexed to consumer price impacts (Kiley 2014; Smith 2014).
7The acceleration in labor productivity growth can occur through
inflation and influenced by the expected path of infla- a combination of capital deepening (or an increase in the machinery
tion, rising price pressures in the expansionary phase and equipment each worker operates), improvements in human cap-
of the cycle can also boost average nominal wage ital and the average skill composition of the workforce, and a faster
pace of technology diffusion that complements the skills of a typical
growth. The opposite happens when final demand
worker. The effects on particular types of workers may vary, depend-
weakens and the business cycle turns. Firms may ing on the complementarity of technological change with their skills
initially hoard labor and, once the slump deepens, and the tasks they perform, as discussed further below.
(Hall 2005; Taylor 2016) mean that changes in labor would weaken the bargaining power of such workers
productivity may not translate one-for-one into wages and lead to less attractive terms of employment, pos-
immediately; wage growth is thus linked more to the sibly in lower-skill occupations (for example, weaker
trend of productivity growth (Dew-Becker and Gordon wage growth, fewer hours, or an increase in the share
2005; Yellen 2005).8,9 of part-time employment). At the other extreme,
As long as workers are able to bargain for a stable advances in design technology can be highly comple-
share of the economys value added, wage growth is mentary for high-skilled workers, such as engineers
generally in line with trend labor productivity growth and architects whose jobs call for complex problem
(Mortensen and Pissarides 1999; Hall 2005). But solving, boosting their productivity and ability to
the strength of the association may waver.10 When command higher wages. Workers bargaining power
workers bargaining power improves over the medium can also be influenced by exposure to international
term, more trend productivity growth increments are competition. This may arise through trade and firms
transmitted to wage growth. participation in global supply chains, but it could
Workers bargaining power is a function of inter- also stem from the threat of production facilities
related drivers.11 These include institutional factors, relocating to economies where costs overall are lower.
such as union density, the coverage of collective Automation and increased competition can in turn
bargaining agreements, and the degree of centraliza- weaken unionization.
tion of such agreements (for example, sectoral versus From a firms perspective, uncertainty about growth
firm-level). Labor laws and employment regulations over the medium term can also influence hiring
that circumscribe firms flexibility in laying off workers decisions and the resulting wage dynamics. At times of
can have an impact on hiring, wage setting, and terms greater optimism and certainty about future revenue,
of employment.12 firms may be more willing to hire full-time workers,
As mentioned earlier, technological changes can create jobs with open-ended contracts, and pay better
also have varying impacts on bargaining power, wages to retain workers or improve the quality of the
depending on the complementarity between new match in the labor market. During times of dimin-
technologies and the mix of tasks performed on the ished growth expectations, perceptions of downside
job and workers skills. At one extreme, automation risks, or uncertainty about the future, firms may be
can substitute for some low- or middle-skilled work- less willing to lock themselves into potentially costly
ers whose jobs mostly call for routine inputs imple- employment arrangements and prefer instead to hire
mented under precise instructions (Autor and Dorn labor part time, or on temporary contracts, with less
2013; Goos, Manning, and Salomons 2014). This favorable wages and benefits. Such growth expectations
could incorporate both demand and supply compo-
8Aone-for-one relationship between real wages and average labor nents, including future demand and expected produc-
productivity over the long term would require an elasticity of substi- tivity growth.
tution between capital and labor of one. The elasticity of substitution
The next section examines the evolution of key labor
between capital and labor is important in determining how the labor
share in national income responds to changes in the relative costs of market indicators in recent years.
labor and capital.
9Of course, this link between wages and productivity may
not strictly hold at the sectoral level (as illustrated by the Advanced Economy Labor Markets:
Balassa-Samuelson effect). Surface Healing Masks Deeper Changes
10During the two decades before the Great Recession, for
Figure 2.1. Distribution of Labor Market Indicators shows, labor force participation has risen in more than
(Percentage-point difference relative to 2007) half of advanced economies relative to 2007 levels,
generally reflecting higher participation by workers
Unemployment rates have been generally declining since 2013 but remain
elevated in about three-quarters of advanced economies relative to their 2007 older than 54 and women in these countries (analyzed
levels. These declines are mostly reective of job creation and not a result of in detail in Box1.1).13,14 Higher unemployment rates,
working-age members of the population dropping out of the labor force. In fact,
labor force participation has risen in more than half of advanced economies.
combined with higher labor force participation rates,
leave employment ratios (employed workers as a share
12 1. Unemployment Rate of the age 15+ population) very close to or above their
10 preGreat Recession peak (2007) in about half of
8 advanced economies.15
6
Wages
4
Panel 1 of Figure2.2 shows that for virtually all
2
advanced economies, nominal wage growth (measured
0
as nominal compensation per hour, and comparable
2
across countries) remains below preGreat Recession
4
2000 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 ranges.16 This is particularly notable for economies
where unemployment rates have declined relatively
4 2. Labor Force Participation Rate rapidly and are now close to or below preGreat
3 Recession averages (Figure2.2, panel 2). Even in
2
economies where nominal wage growth in 2016 was
1
higher than before the Great Recession, such as Ger-
0
many and Japan, the gains have been from low bases:
1
a period of wage moderation in Germany intensified
2
3
by the Hartz labor market reforms and in the midst
4 of Japans decade-long deflation and shrinking nomi-
5 nal wages.17
2000 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16
lower edges of the box show the top and bottom quartiles, and the red markers economies are below what they were before the Great Reces-
denote the top and bottom deciles. sion,whether viewed as consumption real wages (that is, nominal
wages deflated by headline consumer price inflation, which influ-
ences living standards and labor supply decisions) or as product real
wages (that is, nominal wages deflated by the GDP deflator, which
influences firms profitability and hiring decisions). See Annex2.1
for more details on the wage measures and Annex Figure2.2.1 on
the dynamics of real wages.
17See Burda and Seele (2016) for a discussion of the effects of the
Hartz reforms on the German labor market and Aoyagi and Ganelli
(2015) on Japans labor market outcomes during the 2000s.
Figure 2.2. Distribution of Nominal Wage Growth and increased across virtually the entire sample in 2009 and
Correlation with Changes in the Unemployment Rate remains above the 2007 level in more than three-quarters
of countries. In the United States, the share increased
Despite improved headline employment indicators, nominal wages in virtually all
advanced economies are growing at a slower pace than before the Great
from 0.8percent in 2007 to 1.3percent in 2016,
Recession. This is particularly notable for economies where unemployment rates while in the United Kingdom it rose from 2.4per-
are now close to or below pre-Great Recession averages. cent to 3.9percent, and in France from 5.3percent to
7.8percent. Germany is an exception, although its 2016
8 1. Nominal Wage Growth
(Percentage-point difference relative to 2007)
involuntary part-time employment share (3.1percent)
6
was above the 2.7percent average for 200007.
4 As panel 2 of Figure2.3 shows, the largest increases
2 in involuntary part-time employment occurred in econ-
0 omies with unemployment rates above their 200007
2 averages. But even for economies with rates now close
4 to their 200007 averages (points clustered around the
6 vertical axis), the involuntary part-time share of employ-
8 ment is higher than it was before the crisis.
2000 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16
Temporary Contracts
6 2. Nominal Wage Growth and Unemployment Rate1
(Percentage points) Along with involuntary part-time employment,
4
JPN ISL the incidence of temporary contractual arrangements
compensation per hour
2 ISR DEU has attracted attention in recent years (see Aoyagi and
Growth of nominal
AUS
0
CAN FRA Ganelli 2015; Brainard 2016). These contracts can
2 USA ESP
help reduce unemployment spells, allow workers to
4 KOR
ITA avoid gaps in their employment history, and maintain
GBR
6 SVK GRC
CZE
their engagement in the labor force. However, they
8 EST typically offer briefer employment than do open-ended
10 contracts, less opportunity for workers to develop skills
10 5 0 5 10 15
Change in unemployment rate and expand responsibilities, and sometimes weaker
benefits. By 2016, in just over half the economies,
Sources: Eurostat; national authorities; Organisation for Economic Co-operation the temporary contract share was higher than in 2007
and Development; and IMF staff calculations.
Note: The sample in panel 1 excludes Baltic countries. The wage variable used is (Figure2.4, panel 1). Temporary contracts are more
compensation per hour of workers excluding the self-employed. The horizontal line common now than in 200007 for most advanced
inside each box represents the median, the upper and lower edges of the box
show the top and bottom quartiles, and the red markers denote the top and bottom
economies (Figure2.4, panel 2).18
deciles. Data labels in panel 2 use International Organization for Standardization
(ISO) country codes. Outliers and the 10 largest advanced economies (by 2016 Hours
nominal GDP in US dollars) are labeled.
1
Changes shown are 2016 values relative to the 200007 average. A third category of job attributes, which in part
reflects worker preferences, is hours worked per worker.
In more than half of the economies, hours per worker
Involuntary Part-Time Employment, Temporary are at least 2percent below 2007 levels (Figure2.5,
Contracts, Hours panel 1). However, hours had been declining before
that, and the pattern has continued.19
A more complete picture of the labor market
emerges by considering additional indicators that sug- 18In the case of Japan, the figure shows that the share of temporary
gest greater slack in the labor market than captured by contract workers has dropped by close to 6percentage points compared
headline unemployment rates, and possibly weaker job with the 200007 average. But as noted in IMF (2016), the wider
category of nonregular workersthose who either (1) are not hired
security than prior to the Great Recession.
directly by the employer, (2) work part-time, or (3) do not have an
open-ended contractactually increased as a share of overall employ-
Involuntary Part-Time Employment ment during this period. See also Aoyagi and Ganelli (2015). There are
Panel 1 of Figure2.3 documents that involuntary no comparable cross-country data on regular versus nonregular workers.
19The measure may understate the decline in hours per job if an
part-time employment (workers employed fewer than 30 individual now accumulates hours across multiple jobs more often
hours a week who report they would like longer hours) than in the past.
Figure 2.3. Job Attributes: Involuntary Part-Time Figure 2.4. Job Attributes: Temporary Contracts
Employment
The temporary contract share in 2016 is above its 2007 level in over half of
Involuntary part-time employment shares increased across virtually the entire advanced economies. Temporary contracts are more common now than in 2000
sample in 2009 and remain above the 2007 level in more than three-quarters of 07, primarily in economies where the unemployment rate remains above its
the economies. The largest increases occurred in economies with unemployment pre-Great Recession average.
rates above their 200007 averages.
4 1. Share of Temporary Contracts
6 1. Involuntary Part-Time Employment, Share of Total Employment (Percentage-point difference relative to 2007)
2
5 (Percentage-point difference relative to 2007)
4 0
3 2
2
4
1
0 6
1
8
2
3 10
2000 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16
4
2000 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16
8 2. Temporary Contracts and Unemployment1
10 2. Involuntary Part-Time Employment and Unemployment1 (Percentage points)
6
(Percentage points) DNK LUX PRT
8
part-time employment share
temporary contracts
4 SVK FRA
Change in share of
ITA
NLD ITA
Change in involuntary
ESP
6 GRC 2 CHE
IRL DEU IRL
4 FRA 0
SVK PRT GBR
2 NLD KOR NOR
EST 2
CZE LTU LVA
0 ESP
NOR 4
DEU LTU JPN JPN
2 BEL 6
ISR
4 10 5 0 5 10 15
10 5 0 5 10 15 Change in unemployment rate
Change in unemployment rate
Sources: National authorities; Organisation for Economic Co-operation and
Sources: National authorities; Organisation for Economic Co-operation and Development; and IMF staff calculations.
Development; and IMF staff calculations. Note: Temporary workers are those with work contracts of limited duration;
Note: Involuntary part-time workers are those working less than 30 hours a week thresholds are country specic. The share of temporary contracts is calculated as
because they could not nd a full-time position. The involuntary part-time the number of temporary workers divided by total employment. In panel 1, the
employment share is calculated as the total number of involuntary part-time horizontal line inside each box represents the median, the upper and lower edges of
workers divided by total employment. In panel 1, the horizontal line inside each the box show the top and bottom quartiles, and the red markers denote the top and
box represents the median, the upper and lower edges of the box show the top and bottom deciles. In panel 2, countries in gold are those with decreases in the share
bottom quartiles, and the red markers denote the top and bottom deciles. In panel of temporary contracts; countries in red are those with pronounced increases. Data
2, countries in gold are those with decreases in the involuntary part-time labels in the gure use International Organization for Standardization (ISO) country
employment share; countries in red are those with pronounced increases. Data codes.
1
labels in the gure use International Organization for Standardization (ISO) country Changes shown are 2016 values relative to the 200007 average.
codes.
1
Changes shown are 2016 values relative to the 200007 average.
The decline in hours could reflect worker preferences are referred to as zero-hours contracts in the United
for greater flexibility and willingness to work fewer Kingdom; similar agreements govern employment
hours (for example for elderly workers or students who relationships elsewhere, including in Australia and
previously may not have been in the labor force). But Canada.20 As Box2.1 documents, hours declined more
it could also reflect firms preference for hiring work- in sectors with higher shares of low- and middle-skilled
ers for fewer hours or on an as-needed basis. These workers, suggesting that factors beyond worker pref-
just-in-time matches are often governed by agreements
between firms and workers. The firm need not guar- 20In the United Kingdom for example, workers on zero-hours
antee minimum hours, and workers are not obligated contracts as a share of employed workers rose from 0.6percent in
to accept an offer made by the firm. These contracts 2010 to 3percent in 2016 (Haldane 2017).
Figure 2.5. Job Attributes: Hours per Worker erences were at play. A concurrent rise in involuntary
part-time employment also suggests that the decline in
In more than half of advanced economies in 2016, hours per worker were at least hours per worker was driven by reduced demand for
2 percent below 2007 levels. However, this appears to be a continuation of the
pre-2007 pattern. Hours per worker have fallen from their 200007 averages, hours of work by firms, rather than reduced supply of
regardless of whether unemployment rates are now higher or lower than before hours by workers. However, it is still difficult to sepa-
the Great Recession. Declining hours also tend to be associated with higher shares
of involuntary part-time employment. rate workers preferences that shape labor supply from
the binding constraints of weak labor demand.
8 1. Hours per Worker Hours per worker have fallen from their 200007
6 (Log level difference relative to 2007) averages, regardless of whether unemployment rates are
4 higher or lower than they were (Figure2.5, panel 2).
2 Declining hours also tend to be associated with higher
0 shares of involuntary part-time employment (panel 3).
2
4
6 Separating Compositional Shifts from Common Patterns
8 across Sectors
10 The previous sections point to a widespread change
2000 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16
in labor market outcomes (subdued wage growth,
4 2. Hours per Worker and Unemployment
1
larger involuntary part-time employment, higher
Change in log (hours per worker)
0.0
cate that the actual change and the imputed change are
0.2
identical; it is therefore within-sector developments,
0.4
rather than compositional change across sectors, that
0.6
drive aggregate dynamics. Conversely, points off the
0.8
45-degree line indicate that compositional change
1.0
contributed to the overall development. Points marked
1.2 in red are those for which the indicator deteriorated
0.4 0.2 0.0 0.2 0.4 0.6 0.8 1.0
Average change in involuntary part-time during 200916 and compositional change in sectoral
employment share (percentage points) employment shares made a quantitatively import-
ant contribution to that decline (that is, a shift in
Sources: National authorities; Organisation for Economic Co-operation and
Development; and IMF staff calculations. employment toward sectors where the deterioration
Note: In panel 1, the horizontal line inside each box represents the median, the was deeper). The figures indicate that compositional
upper and lower edges of the box show the top and bottom quartiles, and the red
markers denote the top and bottom deciles. In panel 2, countries in gold are those
changes seem to play greater roles for part-time
with increases in hours per worker; countries in red are those with pronounced employment shares, temporary contracts, and hours
decreases. In panel 3, countries in red display (on average) falling hours per per worker than for growth in nominal wages.21
worker and (on average) an increase in the involuntary part-time employment
share for 200916. Data labels in the gure use International Organization for
21Labor mobility across sectors could cause wage growth to be
Standardization (ISO) country codes.
1
Changes shown are 2016 values relative to the 200007 average. broadly synchronized across sectors such that aggregate wage devel-
opments appear to reflect mostly within-sector developments.
Figure 2.6. Average Nominal Wage Growth, 200916, Actual Figure 2.7. Changes in Labor Market Indicators, Actual
versus Imputed Using 2008 Sectoral Employment Shares versus Imputed Using 2008 Sectoral Employment Shares
(Percent)
Compositional change played an important role for changes in job attributes.
Compositional changes do not appear to have an important role in recent nominal Shifts in employment shares across sectors can explain about 22 percent of the
wage growth dynamics. All advanced economies are close to the 45-degree line, increase in the part-time employment share, 18 percent of the increase in the
indicating that aggregate wage growth is driven by within-sector developments. share of temporary contracts, and 23 percent of the reduction in hours per
worker.
4.5
0.75 1. Average Change in Part-Time Employment Share, 200916
4.0 (Percentage points) CHE AUT
ITA SVK
IRL
3.5 0.50 ESP AUS
DNK EST
MLT ISL
3.0 NLD DEU BEL LVA
Actual
0.25 FIN
USA GBR
2.5 CZE FRA
CAN LUX
Actual
2.0 0.00
LTU
1.5
0.25
1.0 0.25 0.00 0.25 0.50 0.75
Imputed using 2008 employment shares
0.5
0.75 2. Average Change in Share of Temporary Contracts, 200916
0.0 SVK
(Percentage points) DNK
0.5 0.50 CZE ISL LUX
0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 ITA
Imputed using 2008 employment shares PRT BEL
0.25 NLD
Actual
SWE MLT
CAN
Sources: Eurostat; national authorities; and IMF staff calculations. 0.00
Note: The wage variable used is annual wage per worker excluding the self- GBR
AUT
employed.
0.25
0.50
0.50 0.25 0.00 0.25 0.50 0.75
Imputed using 2008 employment shares
In the case of part-time employment, 26 countries
in the sample experienced an increased share of
0.50 3. Average Change in Hours per Worker, 200916
part-time workers. In 12 of the 26 countries, com- (Log level difference)
positional change accounted for more than 25per- 0.25
cent of the increase (and more than half the increase JPN
in four countries). 0.00
Actual
LTU IRL
EST SVK
Regarding the temporary contract share of employ- SVN DEU GBR
0.25 LVA
ment, 19 of the 26 countries experienced an CZE
CAN MLT CHE
ITA FIN
increase. Compositional change accounted for more 0.50 AUS
ESP ISL DNK
than 25percent of the increase in seven of those AUT
countries (and more than half in three countries). 0.75
0.75 0.50 0.25 0.00 0.25 0.50
Declines in hours per worker were seen in 25 Imputed using 2008 employment shares
countries, with compositional change accounting for
more than 25percent of this decrease in 10 coun- Sources: Eurostat; national authorities; and IMF staff calculations.
tries (and more than half in five countries). Note: The part-time employment share is calculated as the number of part-time
workers in a sector divided by total employment in the sector. Temporary workers
are people with work contracts of limited duration; thresholds are country
Panels 1 and 2 of Figure 2.8 show that, during specic. The share of temporary contracts is calculated as the number of
temporary workers in a sector divided by total employment in the sector.
200816, declining employment shares in sectors with Countries in red represent cases in which compositional changes amplied
low part-time employment and temporary contracts within-sector increases (panels 1 and 2) or decreases (panel 3). Data labels in the
gure use International Organization for Standardization (ISO) country codes.
(mining and manufacturing), together with faster
increases in employment in sectors with higher shares
Figure 2.8. Job Attributes and Changes in Sectoral of these attributes (services), contributed to rising
Employment Shares, 200816 overall shares of part-time employment and temporary
Compositional shifts in employment toward sectors with relatively high shares of
contracts. Panel 3 of Figure 2.8 shows that shifts in
part-time employment and temporary employment and relatively low hours per employment toward sectors with relatively low hours
worker contributed to the overall changes in these job attributes. per worker contributed to the aggregate change in this
6 1. Part-Time Employment
job attribute.
In sum, sectors that tend to have traditional
Change in employment share
PRF HEA ADM ART contracts and part-time employment, longer hours per
2 INF REA
worker) have seen outright declines or weaker growth
WAT EDU ACC OTH
0
TRA FIN in employment than sectors where arrangements are
ELC PUB more flexible. All in all, shifts in employment shares
MNF TRD
MNG
2 CON AGR across sectors can explain about 22percent of the
increase in part-time employment, 18percent of the
4 increase in temporary contracts, and 23percent of the
25 50 75 100 125 150 175
Average part-time employment share relative to country mean reduction in hours per worker.
6 2. Temporary Contracts
4
(percentage points)
ART HEA ADM PRF growth across advanced economies in recent years. The
2 ACC REA empirical approach is guided by the sequence outlined
INF
EDU WAT
0
OTH TRA in the primer on wage determination. It first explores
TRD PUB FIN MNG the role of cyclical factors, such as headline unemploy-
ELC
2 AGR ment and inflation expectations and medium-term
MNF CON
factors (trend productivity growth), before examin-
4 ing how the changing nature of employment affects
80 90 100 110 120
Average hours per worker relative to country mean wage dynamics. Finally, it explores the influence of
slower-moving factors on wage dynamics and involun-
Sources: Eurostat; national authorities; Organisation for Economic Co-operation
and Development; and IMF staff calculations. tary part-time employment.
Note: Markers to the right of 100 represent sectors with relatively high values
(relative to country mean); markers to the left of 100 represent sectors with
relatively low values. ACC = accommodation and food service activities; ADM =
administrative and support service activities; AGR = agriculture, forestry, and
Aggregate AnalysisCross-Country Evidence
shing; ART = arts, entertainment, and recreation; CON = construction; EDU = The baseline approach is a panel variant of the wage
education; ELC = electricity, gas, steam, and air-conditioning supply; FIN =
nancial and insurance activities; HEA = human health and social work activities; Phillips curve estimated in Gali (2011), in which
INF = information and communication; MNF = manufacturing; MNG = mining and wage growth is regressed on expected inflation, lagged
quarrying; OTH = other services; PRF = professional, scientic, and technical
activities; PUB = public administration and defense; REA = real estate activities; inflation, and the unemployment rate.22 The analysis
TRA = transportation and storage; TRD = wholesale and retail trade; WAT = water
supply, sewerage, waste management, and remediation activities.
22The baseline wage measure is compensation per hour, excluding
focuses on nominal wage growth, examining the influ- in part because employers did not cut wages imme-
ence of past inflation and inflation expectations explic- diately after the financial crisis (Yellen 2014; Daly
itly, alongside drivers that could be acting through real and Hobijn 2015), or the idea that wage growth
wage dynamics. may have been inhibited by a decline in the entry of
These cyclical drivers can be rationalized as follows. new firms, a reduction in labor market churn, and
Nominal wage growth depends on expected inflation fewer job-to-job transitionsand thus fewer discrete
(if wage setting is forward looking) or on lagged increases in wages that often occur with these transi-
inflation (if backward indexation occurs); in aggre- tions.24 While testing these country-specific hypotheses
gate, it is likely to depend on a combination of the in detail is beyond the scope of this chapter, two boxes
two. Given that the benchmark model assumes a supplement the cross-country analysis by shedding
constant natural rate of unemployment and constant light on particular mechanisms that apply in certain
hours per worker, the unemployment rate proxies advanced economy contexts. Box2.2 examines the
for labor market slack. In other models (described incidence of nominal wage freezes and cuts using
in Annexes 2.2 and 2.3), the output gap is used as firm-level data from Europe. Box2.3 studies how wage
an alternative measure of labor market slack. Greater growth in a broad sample of advanced economies may
slack in the labor market is expected to slow wage have been affected by firm-level balance sheet health
growth. Furthermore, at any given labor market slack after the financial crisis.
and inflation expectations, wage growth can vary,
depending on whether the economy is entering or Slack and Inflation
exiting recession. The wage Phillips curves therefore The analysis indicates that slack and past inflation
also control for changes in unemployment (Manning are statistically significantly associated with nominal
1993; Gali 2011). As described in the primer on wage growth, with expected signs (Annex Table2.3.1,
wage determination, a key influence on wage growth column 1). A 1percentage point increase in the
is trend labor productivity growth. The benchmark unemployment rate is associated with a 0.3 to 0.4per-
model controls for this factor as well.23 centage point decline in nominal wage growth, while a
The panel structure allows for the examination of 1percentage point increase in lagged inflation is associ-
wage dynamics across advanced economies, exploiting ated with a 0.2percentage point increase in nominal
variation in the determinants of wage growth over time wage growth.
and across countries. Robustness tests are conducted
by allowing the relationships between wage growth Trend Labor Productivity Growth
and labor market slack, changes in the unemployment Firms profitability and ability to accommodate
rate, and inflation expectations to be country specific. wage increases are linked to changes in trend labor
Allowing coefficients to be country specific can help productivity growth, as discussed in the primer on
capture particular features of individual contextsfor wage determination. The empirical evidence sug-
instance, the hypothesis that nominal wage growth gests that nominal wage growth indeed appears to
in the United States has been subdued in recent years move broadly in line with trend productivity growth
(Annex Table2.3.1, column 2). A 1percentage point
rately determine the shares of value added captured by labor versus increase in trend productivity growth is associated
capital for the self-employed, the baseline measure does not consider
the wages of the self-employed. Results are broadly robust to using with a 0.7percentage point increase in nominal wage
alternative wage measures. growth.25
23The inclusion of trend productivity growth in wage equations
that examine the role of cyclical factors, such as slack and inflation
expectations, is argued for by Ball and Moffitt (2001), Dew-Becker 24Danninger (2016), for example, finds that job-to-job transitions
and Gordon (2005), Hall (2005), and Yellen (2005). The theoretical in the United States have slowed for all skill and age groups in recent
motivation for including productivity growth in wage Phillips curves years. These developments are not necessarily a legacy of the Great
is shown, for example, in Blanchard and Katz (1997), although the Recession. Davis and Haltiwanger (2014) show that worker reallo-
authors note that the empirical estimates for US Phillips curves cation rates declined by 25percent after 2000, suggesting that the
estimated up to the time of writing do not strongly argue for its labor market had begun to turn less fluid before the Great Recession.
inclusion in the specification. The pass-through from labor produc- 25The impact of trend productivity growth on wage growth is
tivity to real wages depends on the bargaining power of workers and consistent with other studies. These results suggest that a 1percent-
the elasticity of substitution between capital and labor (Chapter3 of age point increase in trend productivity growth rate is associated
the April 2017 WEO). Annex Figure2.2.3 illustrates the dynamics with 0.4 to 0.9percentage point higher wage growth, a range
of trend productivity growth. that includes the impact of about 0.8percentage point implied in
Figure 2.9. Effects on Growth of Compensation per Hour: Figure 2.10. Effects on Growth of Compensation per Hour:
Panel Estimation Country-by-Country Estimation, Cross-Country Dispersion
(Percentage points) (Percentage points)
Slack, past ination, and trend labor productivity growth are statistically A country-by-country exploration of the inuences of slack, past ination, and
signicantly associated with nominal wage growth, with expected signs. trend labor productivity growth points to country-specic estimates that are
broadly consistent with the coefcients obtained from the country panel
estimation.
1.0
1.5
1.0
0.5
0.5
0.0
0.0
0.5
0.5
Unemployment Change in Lagged ination Trend labor
rate unemployment productivity 1.0
rate growth Unemployment Change in Lagged ination Trend labor
rate unemployment productivity
rate growth
Source: IMF staff calculations.
Note: The wage variable used is compensation per hour of workers excluding the
self-employed. Markers show estimated coefcients, and lines display 90 percent Source: IMF staff calculations.
condence intervals. Sample excludes Baltic countries. Oil price is used as an Note: The wage variable used is compensation per hour of workers excluding the
instrument for lagged ination. Figure is based on column (7) of Annex Table self-employed. Markers show means of country-by-country estimation
2.3.1. coefcients, and lines display corresponding interquartile ranges. Sample
excludes Baltic countries. Figure is based on column (8) of Annex Table 2.3.1.
earnings). See Chapter3 of the April 2017 WEO for a more exten- however, less precisely estimated than the panel coefficients due to
sive discussion. smaller samples.
The Changing Nature of Employment and previously. Across all countries, on average, a 1per-
Latent Slack centage point increase in the involuntary part-time
Recent studies have argued that measured unem- employment share is associated with a 0.3percentage
ployment rates may not accurately capture slack in point decline in nominal wage growth. To allow for the
the United States (with a resulting focus on U-6 as a possibility that coefficients might vary across countries
broader measure of slack) and some parts of the euro that have had different degrees of labor market tight-
area (ECB2017).27,28 Furthermore, to the extent that ening since the crisis, the regressions are also estimated
declining unemployment rates partly reflect workers separately for three subgroups. The coefficient is larger
forced into part-time jobs, increases in such types of for the sample of countries where the unemployment
employment may overstate the tightening of the labor rate is below preGreat Recession averages. Within
market. Specifically, these workers may be willing this group of countries, a 1percentage point increase
to accept slower increases in wages and, at the same in the involuntary part-time employment share is
time, may continue to seek full-time employment and associated with a 0.7percentage point decline in wage
open-ended contracts. By doing so, they compete with growth. The estimated effect is only 0.2percentage
workers employed under more traditional arrange- point for countries with unemployment appreciably
ments and, so, weigh on their wage growth as well. above the preGreat Recession averages. Though the
True labor market slack may therefore be larger than point estimates are different for these subsamples, these
suggested by headline unemployment rates.29 differences are not statistically significant (Figure2.11
Extensions of the baseline approach examine whether depicts the coefficients shown in Annex Table2.3.3,
the changing nature of employment (as documented in columns 58).
the section Surface Healing Masks Deeper Changes) In contrast to the finding that involuntary part-time
may have contributed to latent slack in the economy employment has weighed on nominal wage growth,
that is not picked up in headline unemployment the analysis does not detect a role for temporary
numbers (Annex Tables 2.3.32.3.7). The analysis contracts in affecting wage dynamics. In general, the
augments the baseline approach by including the shares temporary contract share of employment does not have
of involuntary part-time employment and temporary a statistically significant effect on aggregate wages for
contracts.30 the whole sample or different subgroups (Annex Tables
A higher share of involuntary part-time employ- 2.3.6 and 2.3.7).31
ment is associated with lower wage growth, even after Contributions to Changes in Nominal Wage Growth
controlling for the influence of the variables discussed
Putting the influences of slack, past inflation,
27U-6 includes the total unemployed, plus all marginally attached and trend productivity growth together, Figure2.12
workers and total employed part-time for economic reasons as a per- examines the contributions of these factors to changes
cent of the civilian labor force plus all marginally attached workers. in average nominal wage growth since 2008 relative to
28The evidence for the United States appears mixed. Krueger
nonregular employment in Japan in recent years. Katz and Krueger tional labor market indicatorsthat is, the unemploy-
(2016) discuss the rise of alternative, flexible work arrangements ment rate and its change) would have acted to increase
temping, contracting, freelancing through short-term gigsin the nominal wage growth since 2014. However, involuntary
United States. They estimate that workers in such arrangements now
comprise 16percent of the US workforce. See also Brainard (2016).
part-time employment continues to weigh on nominal
30These could be seen as signs of binding constraints on workers
(possibly stemming from weak labor demand since the Great 31This could in part reflect measurement problems for this
Recession), reflecting in part structural developments, though with variable; to ensure cross-country comparability, the analysis uses a
an important cyclical component. Given that hours per worker also measure that does not contain information on regular versus non-
reflect worker preferences, this attribute is not considered here as a regular contracts, but rather one that adheres to a legal definition of
measure of latent slack. temporariness. See also note 18.
Figure 2.11. Effects of Involuntary Part-Time Employment on Figure 2.12. Decomposition of Wage Dynamics, 200016
Growth of Compensation per Hour, 200016 (Percentage-point change relative to 200007 average)
(Percentage points)
For countries with unemployment rates below 200007 averages, a large part of
A higher share of involuntary part-time employment is associated with lower the decline in nominal wage growth can be explained by slower trend labor
wage growth, even after controlling for the inuence of other variables. The effect productivity growth, while lower slack would have acted to increase nominal wage
is more pronounced in countries where the unemployment rate is below pre-Great growth. In contrast, in countries with unemployment rates still above what they
Recession averages. were before the crisis, both conventional labor market slack measures and
involuntary part-time employment weigh on nominal wage growth.
0.2
Actual wage growth Conventional labor market indicators
Lagged ination Involuntary part-time employment share
0.0 Trend productivity growth Residual
0.4
0
0.6
1
0.8
1.0 2
200007 0812 13 14 15 16
1.2
All countries Unemployment Unemployment Unemployment 1 2. Unemployment Rates Moderately Higher than 200007 Average
rates lower than rates moderately rates appreciably
200007 average higher than higher than
200007 average 200007 average 0
3
wage growth (Figure2.12, panel 1). In contrast, in
countries with unemployment rates still above what they 4
200007 0812 13 14 15 16
were before the crisis, conventional measures of labor
market slack can explain about half of the slowdown Source: IMF staff calculations.
in nominal wage growth since 2007, with involuntary Note: The wage variable used is compensation per hour of workers excluding the
self-employed. Involuntary part-time workers are those working less than 30
part-time employment further weighing on wages hours a week because they could not nd a full-time position. The involuntary
(although part-time employment, even if involuntary, part-time employment share is calculated as the total number of involuntary part-
time workers divided by total employment. Groups are as dened in Figure 2.11.
may have supported labor force participation and facili- The decomposition is based on the coefcients reported in column (5) of Annex
tated stronger engagement with the workplace than the Table 2.3.3 and is weighted by GDP at market exchange rates across countries.
alternative of unemployment). Productivity growth plays
a smaller role, possibly as it was already slow in the years Figure 2.13. Year Fixed Effects and Common Drivers, 200016
before the crisis (Figure2.12, panels 2 and 3). (Index)
The domestic conditions driving wages (such as
The estimated year xed effects tend to be correlated with advanced economy
unemployment) could have a significant common averages of lagged ination, trend productivity growth, unemployment, and
component, given economic linkages between coun- involuntary part-time employment. However, even beyond these factors, there is a
tries as well as the common influence of global factors. negative residual after 2009, and especially during 201416. This could be
picking up the effects of increased integration as well as downward pressure on
In addition, domestic conditions in one country could wage demands as a result of synchronized recessions.
have direct spillover effects on wage setting in others.
For instance, relative wage weakness in one country 1.2
Residuals Predicted values Year xed effects
could put downward pressure on wages in other coun- 1.0
tries, given the threat of production relocation toward
0.8
lower-cost destinations. These common factors would
be picked up by statistically significant time effects in 0.6
Figure 2.14. Changes in Growth Expectations and Labor Changes in regulations related to individual and collec-
Market Institutions tive dismissals (a measure of employment protection;
see Annex2.3.1 for details) do not have a statistically
Subdued nominal wage growth and changes to the nature of employment have significant effect on nominal wage growth. Because
taken place in an environment of declining potential growth and weakening worker
bargaining power.
these factors may be interrelated (an increase in global
value chain participation and offshoring of production
can, for example, contribute to lower unionization),
200007 201216
ascribing precise contributions to each factors influ-
4.5 1. Expected Growth ence on recent wage dynamics is inherently difficult.
4.0 (Percent, average across advanced economies) Nevertheless, as seen in Figure2.15, the limited decline
3.5 in the relative price of investment goods in recent years
3.0 compared with the earlier downward trend suggests
2.5 that automation (as proxied by this measure) may not
2.0 have made a large contribution to the subdued wage
1.5 dynamics following the Great Recession.34
1.0
Such slower-moving forces may have also played a
0.5
role in the increase in involuntary part-time employ-
0.0
Expected growth Expected growth ment, beyond the influence of cyclical factors (Annex
(one year ahead) (ve years ahead) Table2.3.10). While a more negative output gap
(shortfall of actual output relative to the economys
60 2. Labor Market Institutions
(Percent, average across advanced economies) potential) is associated with an increase in the involun-
50 tary part-time employment share, other factors, such
40 as medium-term growth expectations and automation,
also appear to have had an influence (Figure2.16).
30
With declining medium-term growth expectations,
20 firms may have preferred to hire workers part-time.
Automation of work processes could have also led to
10
structurally lower demand for labor. A higher services
0 sector share of employment is also associated with an
Union density Share of countries Share of countries
rate with bi- or tripartite with rm-level bargaining increase in involuntary part-time employment, consis-
agreements tent with the compositional shifts documented in the
section on how surface healing masks deeper changes
3.0 3. Strictness of Employment Protection
(Index, average across advanced economies) in advanced economy labor markets.
2.5
Summary and Policy Implications
2.0 Recent labor market developments in advanced
economies point to a possible disconnect between
1.5 unemployment and wages. Whereas in many econ-
omies headline unemployment is approaching ratios
1.0 seen before the Great Recession, or has even dipped
below those levels, nominal wage growth rates con-
Sources: Institutional Characteristics of Trade Unions, Wage Setting, State
Intervention, and Social Pacts database; Organisation for Economic Co-operation tinue to grow at a distinctly slower pace. For some
and Development; and IMF staff calculations. economies, this may reflect policy measures to slow
Note: Union density rate refers to net union membership as a proportion of wage
earners in employment (simple average across countries); bi- or tripartite wage growth and improve competitiveness in the
agreements refers to the existence of a bipartite council of a central union and
employers and/or the existence of a tripartite council with government
participation. Firm-level bargaining denotes whether bargaining takes place 34Studies focusing on long-term effects of automation tend to
predominantly at the local/company level. Strictness of employment protection
refers to individual and collective dismissals (regular contracts). The sample find larger effects on the wages of particular groups, for example
consists of 2633 advanced economies. middle-skilled workers (see Autor and Dorn 2013 and Chapter3 of
the April 2017 WEO).
Figure 2.15. Long-Term Drivers of Labor Market Dynamics Figure 2.16. Effects on Involuntary Part-Time Employment
Share, Aggregate Analysis
Technological advancements, captured by a declining relative price of investment, (Percentage points)
and falling union density rates could act as additional drivers of labor market
dynamics.
Larger declines in the relative price of investment, lower expected growth, and a
higher share of service workers are associated with a higher share of involuntary
2.0 1. Evolution of Relative Price of Investment in Advanced Economies part-time employment.
(Index)
0.2
1.5
1.0 0.0
0.5
2000 02 04 06 08 10 12 14 0.2
25
Source: IMF staff calculations.
Note: Markers show estimated coefcients, and lines display 90 percent
24
condence intervals. Figure is based on columns (2) to (6) of Annex Table 2.3.10.
GVC = global value chain.
23
22
2000 02 04 06 08 10 12 13
50 4. Union Density Rates by Sector aftermath of the global financial crisis and euro area
(Percent) sovereign debt crisis. Moreover, wage weakness appears
40 200007
200912 to have a common component across advanced econo-
30 mies, which could reflect larger cross-border spillovers
20 of weak labor market conditions since the Great Reces-
sion. Subdued nominal wage growth has also occurred
10
in a context of a higher rate of involuntary part-time
0 employment, an increased share of temporary employ-
MNG MNF CON TRA ACC INF FIN PUB
ment contracts, and a reduction in hours per worker.
The analysis finds that aggregate developments
Sources: Institutional Characteristics of Trade Unions, Wage Setting, State
Intervention, and Social Pacts database; Penn World Tables Capital Detail; World in part-time employment, temporary contracts, and
Bank, World Development Indicators database; and IMF staff calculations. hours, in part, reflect compositional shifts in employ-
Note: Numbers for advanced economies are calculated by rst aggregating over
sectors to the country level using sectoral value added as weight, and ment away from sectors that tend to have traditional
subsequently aggregating over countries using nominal GDP as weight. Sectoral employment arrangements (smaller shares of part-time
numbers are calculated by aggregating over countries using sectoral value added
as weight. Sector abbreviations are as dened in Figure 2.8. employment, a smaller proportion of temporary con-
tracts, longer hours per worker) toward sectors where
more flexible arrangements dominate.
However, there is less evidence that sectoral shifts policies can help stimulate demand and lower headline
in employment account for subdued wage growth. unemployment rates, but overall wage growth (and
Rather, the analysis finds that, at the country level, hence inflation) may continue to remain subdued
labor market slack, together with weak productivity until involuntary part-time employment diminishes or
growth and low inflation expectations, are the main trend productivity growth picks up. Assessing the true
forces weighing on wage growth. Automation (prox- degree of slack beyond measured headline unemploy-
ied by the relative price of investment goods) appears ment rates will be important when determining the
to have made a small contribution to subdued wage appropriate pace of exit from accommodative mone-
dynamics following the Great Recession due to a lim- tary policies.
ited decline in the relative price of investment goods The evidence also suggests that involuntary
in recent years compared with the previous downward part-time employment is in turn associated with
trend. The analysis suggests that automation could both cyclical factors and slower-moving drivers, such
weigh on wage growth more substantially in the future as automation, diminished medium-term growth
if the decline in the relative price of investment goods expectations, and the growing importance of the
were to pick up again. However, inferences about the services sector. Some of these developments point to a
impact of automation are not straightforward given persistent shift in the nature of work and employment
that, as noted previously, the relative price of invest- relations. Policymakers may therefore need to enhance
ment goods is just one channel through which its efforts to address the vulnerabilities that part-time
influence on wage growth may play out. workers face. Examples of possible initiatives in that
Comparing the years since 2008 with 200007, regard include strengthening secondary and tertiary
the chapter finds that in economies where unemploy- education to upgrade skills over the longer term;
ment rates are still appreciably above their averages broadening minimum wage coverage where it does not
before the Great Recession, conventional measures of currently include part-time workers; offering prorated
labor market slack can account for about half of the paid annual, family, and sick leave to secure parity with
slowdown, with involuntary part-time employment full-time workers; and providing subsidized training for
acting as a further significant drag on wages. In these part-time workers for reskilling and retooling (see also
economies, wage growth is unlikely to pick up unless the October 2017 Fiscal Monitor and Golden 2016
slack diminishes meaningfullyan outcome that will for a summary of measures taken by various cities in
require continued accommodative policies to boost the United States, for example). However, any policy
aggregate demand. actions to address the income security of workers that
In economies where unemployment rates are now hold part-time jobs or temporary contracts should be
below their averages before the Great Recession and designed to minimize possible adverse impacts on the
measured slack appears low, slow productivity growth flexibility of labor markets and job creation.
can account for about two-thirds of the slowdown in More generally, the rise of part-time employment
nominal wage growth since 2007. Even in these econo- and temporary contracts challenges the current
mies, involuntary part-time employment, while it may structure of social insurance systemsinstituted in
have helped labor force participation and continued many advanced economies in the aftermath of the
engagement with the workplace, appears to be weigh- Great Depression and World War IIwhich may be
ing on wage growth, alongside slower-moving drivers. better equipped to handle binary employment status
The evidence further indicates that countries expe- (people in the labor force are either employed full-time
riencing a slowdown in trend productivity will face or unemployed). To the extent that changes in the
headwinds to wage growth, even if unemployment nature of employment are not purely cyclical, but also
rates decline. Inflation rates will also remain low unless related to longer-term shifts in structural factors, a
wage growth accelerates beyond productivity growth broader rethinking of the nature of social insurance
in a sustained manner. In such cases, accommodative may be needed.
advanced economies (in particular the United States and the tively), the past decade brought relative gains for these
United Kingdom) in the 1980s, and to a lesser extent the 1990s, groups in terms of hourly wages.
to trade liberalization (Wood 1991, 1994, 1995; Leamer 1992,
1996; Burtless 1995), more intensive trade and migration (Borjas Shrinking Wage Dispersion
and Ramey 1995), outsourcing (Feenstra and Hanson 1996,
2001), or skill-biased technological change (Katz and Murphy The skill premium declined in European economies
1992; Berman, Bound, and Griliches 1994; Autor, Katz, and between 2006 and 2014 (Figure2.1.2); this is true in
Krueger 1998; Katz and Autor 1999; DiNardo and Card 2002; the case of the ratio of wages for high- to low-skilled
Autor, Katz, and Kearney 2008). Autor and Dorn (2013) analyze workers as well as for high- to middle-skilled work-
the polarization of employment and earnings in the United
States between 1980 and 2005 and emphasized the role of auto-
ers. In the United States, the former also declined
mation of routine tasks. over this period, however, the latter showed a
2Parteka (2010) notes the increasing wage gap for low-skilled small increase, pointing to relative wage losses of
workers in the EU15 (Austria, Belgium, Denmark, Finland, middle-skilled workers.
France, Germany, Greece, Ireland, Italy, Luxembourg, Neth- Examining variation across sectors reveals that sectors
erlands, Portugal, Spain, Sweden, United Kingdom) during
19952005 in most sectors, and EU (2015) finds that earnings
with a higher share of low-skilled workers saw higher
inequality increased from 2006 to 2011 in two-thirds of the nominal wage growth. Naturally (given that the shares
members of the European Union. Cho and Daz (2016), add up to 1) the opposite is the case for sectors with a
however, note that the skill premium fell in 200008 in the higher share of high-skilled workers (Figure2.1.3).
Baltic countries.
3Low-skilled workers are defined as those with up to
Hollowing Out of Employment
lower-secondary education, middle-skilled as those with
upper-secondary or postsecondary nontertiary education, and Changes in employment point to hollowing out in
high-skilled as those with tertiary education. European economies as wellin line with the liter-
(percentage points)
ACC
OTH
TRD
1.0 2.0 MNG
CON
HEA
0.5 1.5 FIN EDU
0.0 1.0
High/low-skilled High/middle- Middle/low- 0 10 20 30 40 50 60 70
skilled skilled Employment share of high-skilled
workers, 2010 (percent)
10 2. Change in Skill Premiums by Sector, 200614
0 3.0 2. Low-Skilled
10 Nominal wage growth, 201215 MNF
2.5 PUB
20 (percentage points)
MNG OTH ACC
30
2.0 TRD CON
40 HEA
EDU
50 High/low-skilled High/middle-skilled 1.5 FIN
60
MNG MNF CON TRD ACC FIN PUB EDU HEA
1.0
0 5 10 15 20 25 30
Sources: Eurostat; and IMF staff calculations. Employment share of low-skilled
Note: Low-skilled refers to workers with less than a high workers, 2010 (percent)
school diploma; middle-skilled refers to high school
graduates with no college education; high-skilled refers to
Sources: Eurostat; and IMF staff calculations.
those with at least a bachelors degree. The gure shows
Note: Low-skilled refers to workers with less than a high
simple averages across sectors and economies. ACC =
school diploma; high-skilled refers to those with at least a
accommodation and food service activities; CON =
bachelors degree. ACC = accommodation and food service
construction; EDU = education; FIN = nancial and
activities; CON = construction; EDU = education; FIN =
insurance activities; HEA = human health and social work
nancial and insurance activities; HEA = human health and
activities; MNF = manufacturing; MNG = mining and
social work activities; MNF = manufacturing; MNG = mining
quarrying; PUB = public administration and defense; TRD =
and quarrying; OTH = other services; PUB = public
wholesale and retail trade.
administration and defense; TRD = wholesale and retail
trade.
ature on the United States.4 The employment shares hollowing out declines in the employment shares of
of middle-skilled workers fell, while those of low- and middle-skilled labor.5
high-skilled workers increased (Figure2.1.4). This
pattern can be observed in all sectors, however, during Falling Hours among Low-Skilled Workers
this period it was starkest in services (finance, public Middle-skilled workers lost out in terms of employ-
administration, health, education). While sectoral data ment shares, but low-skilled workers appear to have
on the price of investment is limited, there are some experienced a larger decline in hours than other skill
evidence that sectors more exposed to technological groups. Country-sector-level data on hours by skill
change (that experienced larger declines in their price level are unfortunately not readily available. However,
of investment goods) also saw more pronounced
5Chapter3 of the April 2017 World Economic Outlook high-
sectors with larger shares of low-skilled workers have ings than inequality measures for monthly and hourly
seen larger declines in hours (Figure2.1.5). This agrees wages. Number of months and, to a lesser extent,
with the findings of EU (2015), which highlights hours worked in the year appear to be significant
significantly higher inequality levels for annual earn- sources of variation.
Box 2.2. Worker Contracts and Nominal Wage Rigidities in Europe: Firm-Level Evidence
This box examines the evolving nature of worker Figure 2.2.1. Changes in Employment Shares
contract types and their potential implications for (Percentage points)
wage dynamics in Europe during the postcrisis period.
The data set used in the analysis is from the Wage
Dynamics Network (WDN), constructed to capture 6
determinants of nominal wage dynamics for a large
4
sample of European firms (see Izquierdo and others
2017 for further details on the data set).1 The data set 2
is generated by three waves of surveys conducted in
2007, 2010, and 2014. 0
6
4 80
2
60
0
2
40
4
6 20
8
10 0
FIN MNF TRD MRK ENR CON 40 50 60 70 80 90 100
Share of permanent full-time workers (percent)
8 3. Temporary Workers
6 100 2. Workers with Temporary Contracts
Likelihood of wage cuts and freezes
4
2 80
0
60
2
4
40
6
8 20
10
FIN MNF TRD MRK ENR CON
0
0 10 20 30 40
Sources: Wage Dynamics Network, 2007, 2009, and 2014 Share of workers on temporary contracts (percent)
waves; and IMF staff calculations.
Note: CON = construction; ENR = energy; FIN = nancial
intermediation; MNF = manufacturing; MRK = market Sources: Wage Dynamics Network, 2007, 2009, and 2014
services; TRD = trade. waves; and IMF staff calculations.
Note: Each marker in the gure represents a country sector.
Box2.3. Wage and Employment Adjustment after the Global Financial Crisis: Firm-Level Evidence
How have revenue growth performance and and employment decisions, the box further explores
volatility affected firms labor-related decisions in the whether firms with different degrees of ex ante financial
postcrisis period? What role has firm-level financial vulnerability exhibit different wage and/or employment
vulnerability at the outset of the crisis played when it adjustment patterns in the postcrisis period.
comes to postcrisis firm-level labor market choices? The evidence suggests that firms with stronger recent
This box looks at these questions using the ORBIS growth performance (and thus arguably more optimistic
data set compiled by Bureau van Dijk. It is a rich, growth expectations) and low volatility exhibit higher
cross-country, firm-level data set that contains firms wage and employment growth. Moreover, firms with
balance sheet variables as well as total wage bill and weaker balance sheets before the crisis experience lower
total employment information.1 The box first explores growth in wages and employment following the crisis,
the association between recent growth (which arguably which highlights the potential role of crisis-related legacies
influences firm-level growth expectations) and uncer- in firms labor-related decisions in the postcrisis period.
tainty, and firms wages and employment growth follow-
ing the global financial crisis. To assess the potential Growth Expectations and Uncertainty as
effect of financial-crisis-related factors on firms wage Determinants of Wage and Employment Growth
To the extent that recent growth influences expectations
The author of this box is Gee Hee Hong. about future growth (for example, if firms form adaptive
1Comparability of the variables across countries and over time expectations), trailing five-year average revenue growth
is ensured as described in Duval, Hong, and Timmer (2017), can be considered a proxy for firm-level medium-term
following the methodology of Gal and Hijzen (2016). growth expectations. Moreover, the standard deviation of
2.5 1
2.0 0
Nominal wage growth
Employment growth 1
1.5
2
1.0
3
0.5 4
0.0 5
6
0.5
7
1.0
Volatility Volatility Growth 8
(revenue based) (output based) expectations Change in wage per worker Change in employment
Sources: ORBIS; and IMF staff calculations. Sources: ORBIS; and IMF staff calculations.
Note: Wage is dened as total wage bill divided by total Note: The left bar represents the estimated difference in
employment for each rm. The blue bars show the estimated postcrisis wage growth minus precrisis wage growth
wage growth differences between rms with high between a rm with a high ratio of debt maturing in 2008
uncertainty/growth expectations (75th percentile) and those (75th percentile) and a rm with a low ratio of debt maturing
with low uncertainty/growth expectations (25th percentile). in 2008 (25th percentile). The right bar represents the
The red bars show the corresponding employment growth estimated difference in postcrisis employment growth
differences. minus precrisis employment growth between the two types
of rms.
Box2.3 (continued)
revenue growth (volatility)or its ratio to average revenue Financial Frictions and Labor-Related Decisions
growth over the trailing five-year interval (coefficient Firms whose financial vulnerability was higher before
of variation)can be considered a proxy for firm-level the crisis appear to exhibit weaker wage and employment
uncertainty about the operating environment. growth in its aftermath, which highlights the potential
The evidence suggests that firms with more optimistic role of financial frictions or crisis-related legacies in wage
growth expectations or lower volatility show stronger and employment adjustments following the crisis.
wage and employment growth in the postcrisis period.2 Adopting the difference-in-differences methodology
Figure2.3.1 compares the differences in average wage that compares the averages of precrisis and postcrisis
and employment growth rates since 2008 between firms wage and employment growth following Duval, Hong,
whose volatility and growth expectations are in the 25th and Timmer (2017), firms with ex ante more vulner-
and 75th percentiles. Wage growth is 0.3 to 0.6per- able balance sheetshigher leverage and rollover risk
centage point lower for firms with higher volatility than entering the financial crisisexhibit lower wage and
for their counterparts with lower volatility (depending employment growth in the postcrisis years. The results
on the measure used to construct volatility). In addi- are robust to controlling for labor productivity and mul-
tion, firms whose growth expectations are more optimis- tifactor productivity, following Wooldridge (2009).3
tic show 2percentage points stronger wage growth than Table2.3.1 reports the results. Controlling for
their less optimistic counterparts. Similarly, for employ- different measures of productivity, a 10percentage
ment growth, firms with higher volatility experience 0.5 point higher leverage ratio before the crisis is associated
to 0.8percentage point lower employment growth than with 0.1percentage point weaker growth in wages and
those with lower volatility. Optimism in growth expec- employment after the crisis. Similarly, firms with higher
tations contributes positively to employment growth as precrisis rollover risk show about 0.3 to 0.4percentage
well: firms with more optimistic expectations experience point weaker growth in wages and employment.
nearly 1.5percentage points higher employment growth
than those that are less optimistic.
3Rollover risk, measured as the ratio of current liabilities (that is,
debt maturing within a year) to total sales in the 2007 balance sheet,
2The two main dependent variables are the annual growth rate allows for a causal interpretation. Firms debt structure in 2007 is
of total employment for each firm and the annual growth rate of unlikely to be associated with other unobserved firm characteristics
wage per employee, in which the wage per employee is calculated affecting wage and employment decisions given that the timing of
as the total wage bill divided by the total number of employees the global financial crisis was not foreseen from the vantage point in
for each firm. 2007 (Almeida and others 2012; Duval, Hong, and Timmer 2017).
Co-operation and Development (OECD), and periods unemployment rate given that the unemployment rate in
national authorities. Key sources for other variables the United States follows an autoregressive (2) process, in which the
expected unemployment rate is a function of current and previous
used in this chapter are the Eora Multi-Region unemployment rates. The analysis in this chapter uses a similar
Input-Output database; Database on Institutional argument for controlling for the change in unemployment rate: it
Characteristics of Trade Unions, Wage Setting, State captures the expectation of the evolution of unemployment rates
beyond the current rate. Intuitively, this captures the importance
Intervention and Social Pacts (ICTWSS); IMF World of whether a country is entering a recession (rising unemployment
Economic Outlook database; and the OECD. rates) or recovering from one (falling unemployment rates).
Annex Figure 2.2.1. Distribution of Real Compensation Annex Figure 2.2.2. Growth of Real Compensation per Hour
Growth Measures and Unemployment Rates
(Percentage-point difference relative to 2007) (Percentage-point change, 2016 relative to 200007 average)
5 1. Deated Using Consumer Price Index 8 1. Deated Using Consumer Price Index
4
3 6
Growth of real
0 DEU JPN
2
1 AUS ESP
2 0 KOR
3 USA FRA ITA
4 2
CAN
5 4 GBR
6
7 6
2000 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 10 5 0 5 10 15
Unemployment rate
6 2. Deated Using GDP Deator
5 8 2. Deated Using GDP Deator
4
3 6
2
Growth of real
0 2
1 CAN ESP
JPN AUS
2 0
3 DEU USA
4 2 FRA ITA
5 GBR
4 KOR
6
2000 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16
6
10 5 0 5 10 15
Unemployment rate
Sources: Eurostat; national authorities; Organisation for Economic Co-operation
and Development; and IMF staff calculations.
Note: The sample excludes Baltic countries. The wage variable used is Sources: National authorities; Organisation for Economic Co-operation and
compensation per hour of workers excluding the self-employed. The horizontal Development; and IMF staff calculations.
line inside each box represents the median, the upper and lower edges of the box Note: The wage variable used is compensation per hour of workers excluding the
show the top and bottom quartiles, and the red markers denote the top and self-employed. Data labels in the gure use International Organization for
bottom deciles. Standardization (ISO) country codes. The 10 largest advanced economies (by 2016
nominal GDP in US dollars) are labeled.
w
i,t = i + i,t1 + 1ui,t + 2 ui,t + i,t, (2.1) employed workers who take part-time jobs invol-
in which, for country iand time t, the nominal w is
i,t untarily, with part-time jobs defined as less than 30
wage growth, i,t1is lagged year-over-year inflation, hours a week, and the share of employed workers who
ui,tis the unemployment rate, and ui,t is the change have temporary work contracts. The primer earlier
in the unemployment rate. in this chapter explains why these drivers matter for
To explore how productivity growth and labor wage growth. As noted there, the analysis focuses on
underutilization may affect aggregate wage growth, nominal wage growth; Annex Figures 2.2.1 and 2.2.2
equation (2.1) is augmented with two sets of vari- illustrate real wage dynamics for reference.
ables: trend productivity growth and labor mar- Annex Figure2.2.3 shows the dynamics of two key
ket underutilization measures. Equation (2.2) drivers in equation (2.2): trend productivity growth
is estimated: and lagged inflation (a proxy for inflation indexation).
The analysis examines several robustness tests:
w
i,t = i + i,t1 + 1ui,t + 2 ui,t
Data frequency: The labor market underutilization
i,t + Z i,t + i,t, (2.2)
+ g Y/H measures (involuntary part-time and temporary
in which gY/H
i,t is the trend of the growth rate of real contract employment shares) are not available
output per hour, and Z i,t are labor underutiliza- at quarterly frequencyhence the analysis of
tion measures. These measures include the share of their impact on aggregate wage growth in Annex
Annnex Figure 2.2.3. Factors Associated with Nominal Wage Annex Figure 2.2.4. Effects of Involuntary Part-Time
Growth Employment on Compensation and Wages, 200016
(Percentage point difference relative to 2007) (Percentage points)
tion divided by total employees (compensation Annex Figure 2.2.5. Correlations between Aggregate Wage
per employee), aggregate wage bill divided by total Growth and Two-Quarter-Lagged Public Wage Growth
(Percent)
employees (wage per employee), aggregate com-
pensation divided by total hours (compensation
70
per hour), and aggregate wage bill divided by total
hours (wage per hour, which includes aggregate 60 200108
social contributions of employers). Annex Fig- 50 200916
upward. Similar logic applies to the labor underuti- growth may cause upward bias in the effect of trend productivity
lization measure. The main resultthat involuntary growth through employment growth. However, estimated coefficients
of the labor productivity trend are often lower than what is implied
from other studies in the literature (for instance, Karabarbounis and
37Addressing this reverse causality could be expected to reduce Neiman 2014), especially if the sample is restricted to the postGreat
the coefficient of lagged inflation. Annex Table2.3.3 suggests that Recession period. Together, these suggest that downward attenuation
this is indeed the case for groups A and C. There could be some bias may dominate reverse causality, causing an underestimation of
idiosyncratic reasons biasing the ordinary least squares estimate of the role of trend productivity growth. Results are broadly unchanged
lagged inflation downward for group B (the coefficient is negative when imposing the coefficient of trend productivity growth to be 1
and insignificant). or the value implied from other studies.
in the relative price of investment), growth expecta- accompanied subdued wage growth, even as headline
tions, and global value chain integration (proxied by employment has fallen. These developments have
the five-year change in foreign value added as a percent occurred in the context of falling growth expectations
of exports). and declines in worker bargaining power, as shown in
Figures 2.14 and 2.15.42
The sectoral analysis examines the effects of increas-
Sectoral Analysis ing trade openness, automation (captured by the
As a complement to the aggregate analysis, drivers declining relative price of investment), and slowing
of nominal wage growth and part-time employment sectoral growth rates (used to construct a measure of
shares are examined at the sectoral level.39 Following adaptive growth expectations at the sectoral level) on
the structure of the aggregate analysis, sector-level nominal wage growth and part-time employment as
regressions explore the roles of slack, medium-term a share of total employment.43 It does so by exploit-
growth expectations, technological progress, increased ing sectoral variations in exposure to aggregate forces
trade integration, and changes in labor market (Annex Table2.2.1). For instance, country-level slack
institutions.40 These are examined as possible drivers could be expected to matter more for labor market
of nominal wage growth and part-time employment dynamics in sectors that are more correlated with the
to determine their effect on different margins of aggregate economy, and the effects of a decline in the
adjustment. The analysis exploits variation in sectoral aggregate relative price of investment could vary by the
exposure to aggregate forces to shed light on mecha- initial capital intensity of the sector.
nisms that operate within countries.41 The analysis relies on annual data for a sample of
As noted earlier, across several advanced econo- 20 advanced economies starting in 2000, and relates
mies, a rise in involuntary part-time employment has changes in nominal wage growth to the same cyclical
and secular drivers used in the aggregate analysis, con-
39Estimates of involuntary part-time employment are not available
trolling for country, sector, and year fixed effects:
at the sectoral level, so the focus here is on total part-time employ-
ment, including both voluntary and involuntary.
40Control variables are in line with those used in ECB(2009) 42Panel 4 of Figure2.15 shows the decline in union density rates
and EC(2003), as well as in the literature on interindustry wage occurring in most sectors, with the notable exception of public
differentials and wage dispersion (for example, Erdil and Yetki- administration; coverage of sectoral union density rates is unfortu-
ner2001; Koeniger, Leonardi, and Nunziata2007; and Du Caju nately too limited to be included in the regression analysis.
and others2010). Wage regressions also control for inflation and 43Sectoral expected growth is measured as the five-year trailing
(sectoral) trend productivity growth. average of sectoral gross output growth rates. As noted above,
41The regressions also control for country, sector, and year this could be capturing expected productivity growth as well as
fixed effects. demand conditions.
yijt = i + j + t + Xijt
+ Zjt , (2.3) change in the relative price of investment with sectoral
in which yijtis nominal wage growth, Xijt
includes mea- capital intensity.
sures that vary at the country-sector level, such as the As in the aggregate regressions, the sectoral analysis
share of part-time employment, how correlated a sec- relates the share of part-time employment to slack
tors gross output growth is with the overall economy, (captured using the output gap and how correlated a
sectoral trend productivity growth (measured again sector is with the aggregate economy and the inter-
using a five-year trailing average), sectoral expected action between these two variables) and to secular
growth (an adaptive measure based on a five-year drivers: expected growth, change in final imports as a
trailing average of sectoral gross output growth), and share of gross output, change in the relative price of
the five-year change in final imports as a share of investment (also interacted with capital intensity), and
gross output. change in the union density rate.
Zjt includes measures that vary only at the country
level, such as the aggregate output gap and (lagged) Annex2.3. Empirical Results
inflation, the change in the relative price of invest-
ment, and measures of worker bargaining power Aggregate Analysis
(proxied again using the five-year change in the union Annex Tables 2.3.1 and 2.3.2 show estimates of
density rate). To exploit sectoral variation in exposure wage Phillips curves using ordinary least squares and
to aggregate forces, these are interacted with sec- instrumental variables estimations, for the full sample
toral characteristics, looking at the interaction of the as well as a sample excluding the Baltic countries, and
aggregate output gap with the correlation of the sector for alternative measures of the dependent and explana-
and the aggregate economy and the interaction of the tory variables.
Oil price changes are common across countries, so examine robustness to using different measures of
instrumental variables results do not control for year wages, as well as exploring differences across countries
fixed effects. The main results are not sensitive to the with unemployment rates below, moderately above,
choice of estimation methodordinary least squares and appreciably above 200007 averages.
including year fixed effects or instrumental variables Annex Tables 2.3.6 and 2.3.7 conduct a similar
without year fixed effects (Annex Figure2.3.1 com- exercise for the temporary contract employment
pared with Figure2.12). The share of variation in wage share instead of involuntary part-time employment
growth explained by inflation drivers is broadly similar share. Results are very similar if both the involuntary
across the two approaches.44 part-time employment share and temporary contract
Annex Tables 2.3.32.3.5 augment the wage Phillips employment share are controlled for simultaneously.
curve specification in Annex Table2.3.1 further with These labor market underutilization measures do
the share of involuntary part-time employment, and not appear to affect the sensitivity of wage growth
to unemployment ratesthey are thus included
44Further analysis relating the residuals from the wage Phillips
additively.
curve analysis to the global output gap (weighted by dollar GDP)
suggests that the global output gap is not significant in explaining
As described above, Annex Table2.3.8 augments the
such residuals. wage Phillips curve with secular drivers. Because wage
growth rates were volatile during the Great Recession, Annex Figure 2.3.1. Decomposition of Wage Dynamics,
Annex Table2.3.9 examines robustness to excluding 200016
(Percentage-point change relative to 200007 average)
the years 2008 and 2009.
Annex Table2.3.10 zooms in on the determinants
Lagged ination Year xed effects
of job attributes and examines the drivers of involun- Trend productivity growth Residual
tary part-time employment, linking it to the output Conventional labor market indicators Actual wage growth
Involuntary part-time employment share
gap and the secular drivers explored above.
2 1. Unemployment Rates Lower than 200007 Average
Sectoral Analysis 1
Annex Table2.3.3. Estimation of Wage Phillips Curve Augmented with Involuntary Part-Time Employment
Share by Country Group
(1) (2) (3) (4) (5) (6) (7) (8)
Full Sample Group A Group B Group C Full Sample Group A Group B Group C
OLS OLS OLS OLS IV1 IV1 IV1 IV1
Involuntary Part-Time 0.177** 0.503* 0.336** 0.0159 0.275*** 0.653** 0.291* 0.186*
Employment Share (0.0830) (0.274) (0.139) (0.124) (0.0829) (0.294) (0.154) (0.101)
Unemployment Rate 0.187*** 0.0178 0.00699 0.280*** 0.182*** 0.0855 0.284 0.395***
(0.0445) (0.128) (0.186) (0.0686) (0.0438) (0.146) (0.186) (0.0722)
Change in Unemployment Rate 0.349*** 0.690*** 0.609** 0.128 0.263*** 0.449** 0.830*** 0.0821
(0.0960) (0.244) (0.271) (0.129) (0.0887) (0.181) (0.247) (0.117)
Lagged Inflation 0.193*** 0.378*** 0.183 0.156 0.300* 0.287 0.397 0.279
(0.0728) (0.129) (0.124) (0.206) (0.164) (0.282) (0.248) (0.292)
Trend Productivity Growth Rate2 0.456*** 0.634* 0.131 0.699*** 0.624*** 0.763*** 0.00955 0.986***
(0.112) (0.348) (0.189) (0.170) (0.106) (0.223) (0.176) (0.170)
First-Stage F-statistics above 10 yes yes yes yes yes yes yes yes
Country Fixed Effects yes yes yes yes yes yes yes yes
Year Fixed Effects yes yes yes yes no no no no
Number of Observations 411 117 146 148 411 117 146 148
R2 0.610 0.709 0.649 0.723 0.577 0.652 0.458 0.660
Source: IMF staff calculations.
Note: Dependent variable = annual growth rates of compensation per hour of workers excluding the self-employed. Sample is of annual frequency from
2000 to 2016. See Annex Table2.1.1. for countries in the full sample. A few countries are not in the sample due to missing data on involuntary part-time
employment share. Country groups are divided by comparing unemployment rate in 2016 with 200007 average. Group A (2016 unemployment lower than
200007): Czech Republic, Germany, Japan, Israel, Slovak Republic, United Kingdom, and United States. Group B (2016 unemployment moderately higher
than 200007): Australia, Austria, Belgium, Canada, Switzerland, Finland, Iceland, Norway, and Sweden. Group C (2016 unemployment appreciably higher
than 200007): Denmark, Spain, France, Greece, Ireland, Italy, Netherlands, Portugal, and Slovenia. IV = instrumental variable. OLS = ordinary least squares.
Standard errors in parentheses. * p < .10; ** p < .05; *** p < .01.
1The instrumental variable for lagged inflation is the two-quarter-lagged change in oil price.
2Five-year trailing average of the labor productivity growth rate.
Annex Table2.3.4. Estimation of Wage Phillips Curve Augmented with Involuntary Part-Time Employment
Share: Full Sample and Countries with Unemployment Rates Lower than 200007 Average
(1) (2) (3) (4) (5) (6)
Countries with Unemployment Rate Lower
Full Sample than 200007 Average (Group A)
Compensation Compensation Wage Compensation Compensation Wage
per Employee1 per Hour1 per Hour1 per Employee1 per Hour1 per Hour1
IV2 IV2 IV2 IV2 IV2 IV2
Involuntary Part-Time 0.203** 0.275*** 0.242*** 0.535** 0.653** 0.705**
Employment Share (0.0803) (0.0829) (0.0805) (0.261) (0.294) (0.292)
Unemployment Rate 0.167*** 0.182*** 0.177*** 0.0174 0.0855 0.103
(0.0424) (0.0438) (0.0422) (0.130) (0.146) (0.145)
Change in Unemployment Rate 0.473*** 0.263*** 0.321*** 0.574*** 0.449** 0.567***
(0.0859) (0.0887) (0.0853) (0.161) (0.181) (0.180)
Lagged Inflation 0.509*** 0.300* 0.309* 0.491* 0.287 0.253
(0.159) (0.164) (0.162) (0.250) (0.282) (0.279)
Trend Productivity Growth Rate3 0.413*** 0.624*** 0.701*** 0.659*** 0.763*** 0.760***
(0.103) (0.106) (0.102) (0.198) (0.223) (0.222)
First-Stage F-statistics above 10 yes yes yes yes yes yes
Country Fixed Effects yes yes yes yes yes yes
Year Fixed Effects no no no no no no
Annex Table2.3.5. Estimation of Wage Phillips Curve Augmented with Involuntary Part-Time Employment
Share: Countries with Unemployment Rates Moderately Higher and Appreciably Higher than 200007 Average
(1) (2) (3) (4) (5) (6)
Countries with Unemployment Rates Moderately Countries with Unemployment Rates Appreciably
Higher than 200007 Average (Group B) Higher than 200007 Average (Group C)
Compensation Compensation Wage Compensation Compensation Wage
per Employee1 per Hour1 per Hour1 per Employee1 per Hour1 per Hour1
IV2 IV2 IV2 IV2 IV2 IV2
Involuntary Part-Time 0.221 0.291* 0.110 0.157* 0.186* 0.235**
Employment Share (0.147) (0.154) (0.147) (0.0923) (0.101) (0.105)
Unemployment Rate 0.203 0.284 0.147 0.358*** 0.395*** 0.375***
(0.177) (0.186) (0.187) (0.0663) (0.0722) (0.0751)
Change in Unemployment Rate 1.429*** 0.830*** 0.743*** 0.0369 0.0821 0.0381
(0.235) (0.247) (0.241) (0.107) (0.117) (0.121)
Lagged Inflation 0.522** 0.397 0.780*** 0.126 0.279 0.369
(0.236) (0.248) (0.259) (0.268) (0.292) (0.304)
Trend Productivity Growth Rate3 0.183 0.00955 0.0518 0.834*** 0.986*** 1.082***
(0.168) (0.176) (0.167) (0.156) (0.170) (0.177)
First-Stage F-statistics above 10 yes yes yes yes yes yes
Country Fixed Effects yes yes yes yes yes yes
Year Fixed Effects no no no no no no
Annex Table2.3.6. Estimation of Wage Phillips Curve Augmented with Temporary Contract Employment
Share: Full Sample and Countries with Unemployment Rates Lower than 200007 Average
(1) (2) (3) (4) (5) (6)
Countries with Unemployment Rate Lower than
Full Sample 200007 Average (Group A)
Compensation Compensation Wage Compensation Compensation Wage
per Employee1 per Hour1 per Hour1 per Employee1 per Hour1 per Hour1
IV2 IV2 IV2 IV2 IV2 IV2
Temporary Contract 0.0274 0.0866 0.0861 0.0498 0.115 0.146
Employment Share (0.0566) (0.0584) (0.0561) (0.135) (0.174) (0.176)
Unemployment Rate 0.244*** 0.297*** 0.277*** 0.0666 0.262 0.308
(0.0428) (0.0441) (0.0427) (0.219) (0.281) (0.285)
Change in Unemployment Rate 0.428*** 0.181* 0.249*** 0.392* 0.291 0.375
(0.0974) (0.100) (0.0960) (0.203) (0.261) (0.265)
Lagged Inflation 0.556*** 0.259 0.281 0.431 0.167 0.249
(0.182) (0.188) (0.183) (0.430) (0.553) (0.561)
Trend Productivity Growth Rate3 0.503*** 0.736*** 0.806*** 0.987*** 1.130*** 1.133***
(0.118) (0.122) (0.116) (0.195) (0.251) (0.254)
First-Stage F-statistics above 10 yes yes yes yes yes yes
Country Fixed Effects yes yes yes yes yes yes
Year Fixed Effects no no no no no no
Annex Table2.3.7. Estimation of Wage Phillips Curve Augmented with Temporary Contract
Employment Share: Countries with Unemployment Rates Moderately Higher and Appreciably
Higher than 200007 Average
(1) (2) (3) (4) (5) (6)
Countries with Unemployment Rate Moderately Countries with Unemployment Rate Appreciably
Higher than 200007 Average (Group B) Higher than 200007 Average (Group C)
Compensation Compensation Wage Compensation Compensation Wage
per Employee1 per Hour1 per Hour1 per Employee1 per Hour1 per Hour1
IV2 IV2 IV2 IV2 IV2 IV2
Temporary Contract Employment 0.0416 0.158 0.138 0.106 0.107 0.101
Share (0.0987) (0.102) (0.0975) (0.0818) (0.0875) (0.0919)
Unemployment Rate 0.489*** 0.446*** 0.383** 0.383*** 0.426*** 0.411***
(0.153) (0.158) (0.153) (0.0699) (0.0748) (0.0786)
Change in Unemployment Rate 1.227*** 0.636** 0.610** 0.0615 0.0538 0.0717
(0.249) (0.257) (0.250) (0.117) (0.126) (0.132)
Lagged Inflation 0.384 0.128 0.563* 0.0161 0.104 0.132
(0.274) (0.283) (0.293) (0.272) (0.291) (0.306)
Trend Productivity Growth Rate3 0.0832 0.303* 0.277* 0.862*** 1.000*** 1.097***
(0.158) (0.163) (0.155) (0.190) (0.204) (0.214)
First-Stage F-statistics above 10 yes yes yes yes yes yes
Country Fixed Effects yes yes yes yes yes yes
Year Fixed Effects no no no no no no
and the associations are broadly similar in size to While skill composition is not included in the baseline
those found in the aggregate analysis for involun- specifications due to data limitations, the results are
tary part-time employment (Figure2.3.3; Annex robust to including it as an additional control.
Table2.3.12). Further robustness tests have explored alternative
Annex Table2.3.13 reports the robustness test in trade measures, such as exports and intermediate
which growth in nominal wages, employment, and exports as a share of gross output and global value
part-time employment are treated as jointly deter- chain participation, aggregate expected growth (one
mined and estimates the system using three-stage and five years ahead) interacted with sectoral correla-
least squares, which treats the dependent variables as tion instead of sectoral expected growth, and further
endogenous, instruments them using the exogenous measures of worker bargaining power. Such further
variables, and allows them to be correlated with distur- measures include whether the country has a bi- or
bances in the systems equations. tripartite agreement, whether bargaining is done
The results are also robust to looking at three-year predominantly by firms (as opposed to at the sector or
nonoverlapping averages of the dependent and explan- country level), the ease of hiring and firing, and the
atory variables instead of annual data. Furthermore, as strictness of employment protection regulation. Results
in the aggregate analysis, results are robust to omitting on other variables are broadly comparable to those in
smaller advanced economies (the Baltic countries). the baseline regressions.
Annex Table2.3.8. Estimation of Wage Phillips Curve Augmented with Structural Variables
(1) (2) (3) (4) (5) (6) (7) (8) (9)
IV1 IV1 IV1 IV1 IV1 IV1 IV1 IV1 IV1
Involuntary Part-Time 0.275*** 0.306*** 0.192** 0.200** 0.166* 0.225*** 0.272*** 0.0840 0.0570
Employment Share (0.0829) (0.0947) (0.0845) (0.0976) (0.0977) (0.0794) (0.0830) (0.133) (0.125)
Unemployment Rate 0.182*** 0.226*** 0.211*** 0.293*** 0.365*** 0.199*** 0.177*** 0.333*** 0.362***
(0.0438) (0.0556) (0.0492) (0.0688) (0.0590) (0.0444) (0.0446) (0.0948) (0.0902)
Change in Unemployment Rate 0.263*** 0.225** 0.137 0.284*** 0.0325 0.247*** 0.267*** 0.295** 0.334***
(0.0887) (0.0969) (0.0833) (0.109) (0.0887) (0.0893) (0.0887) (0.130) (0.123)
Lagged Inflation 0.300* 0.0452 0.00644 0.380 0.236 0.199 0.308* 0.432 0.540
(0.164) (0.280) (0.197) (0.311) (0.206) (0.186) (0.164) (0.332) (0.327)
Trend Productivity Growth Rate 0.624*** 0.720*** 0.845*** 0.497*** 0.594*** 0.570*** 0.628*** 0.231 0.325**
(0.106) (0.118) (0.109) (0.123) (0.117) (0.101) (0.107) (0.168) (0.156)
Change in Foreign Value Added as 0.0944**
a Share of Exports2 (0.0424)
Change in the Relative Price of 0.114***
Investment2 (0.0302)
Change in the Union Density 0.330*** 0.340***
Rate2 (0.0774) (0.0774)
Change in Individual and Collective 0.259
Dismissal Regulation2 (0.918)
Expected Growth 0.459**
(0.180)
Change in the Share of Service 0.0194
Sector Workers2 (0.0327)
Union Density Rate (Level) 0.322*** 0.186***
(0.0836) (0.0678)
First-Stage F-statistics above 10 yes yes yes yes yes yes yes yes yes
Country Fixed Effects yes yes yes yes yes yes yes yes yes
Year Fixed Effects no no no no no no no no no
Number of Observations 411 361 316 288 247 411 411 267 264
R2 0.577 0.561 0.596 0.590 0.603 0.589 0.578 0.501 0.567
Source: IMF staff calculations.
Note: Dependent variable = annual growth rates of compensation per hour of workers excluding the self-employed. Sample is of annual frequency from 2000 to 2016.
See Annex Table2.1.1 for countries in the sample. IV = instrumental variable. Standard errors in parentheses. * p < .10; ** p < .05; *** p < .01.
1The instrumental variable for lagged inflation is the two-quarter-lagged change in oil price.
2Relative to five years ago.
Annex Table2.3.9. Estimation of Wage Phillips Curve Augmented with Structural Variables: Excluding 2008
and 2009
(1) (2) (3) (4) (5) (6) (7) (8) (9)
IV1 IV1 IV1 IV1 IV1 IV1 IV1 IV1 IV1
Involuntary Part-Time 0.213** 0.227 0.168* 0.215* 0.193 0.173** 0.205** 0.212 0.169
Employment Share (0.0912) (0.139) (0.102) (0.123) (0.131) (0.0862) (0.0913) (0.198) (0.178)
Unemployment Rate 0.174*** 0.205*** 0.186*** 0.301*** 0.319*** 0.196*** 0.162*** 0.367 0.380*
(0.0428) (0.0657) (0.0549) (0.105) (0.0768) (0.0446) (0.0435) (0.230) (0.227)
Change in Unemployment Rate 0.400*** 0.321* 0.280** 0.308* 0.129 0.352*** 0.406*** 0.495** 0.507**
(0.118) (0.183) (0.131) (0.168) (0.152) (0.127) (0.118) (0.229) (0.213)
Lagged Inflation 0.502** 0.351 0.180 0.583 0.254 0.354 0.520** 1.289 1.417
(0.208) (0.598) (0.346) (1.107) (0.677) (0.251) (0.207) (2.101) (2.053)
Trend Productivity Growth Rate 0.768*** 0.826*** 0.891*** 0.471*** 0.662*** 0.721*** 0.779*** 0.0674 0.151
(0.101) (0.118) (0.120) (0.154) (0.130) (0.0968) (0.101) (0.662) (0.466)
Change in Foreign Value Added as 0.0262
a Share of Exports2 (0.0452)
Change in Relative Price of 0.0911***
Investment2 (0.0338)
Change in Union Density Rate2 0.390* 0.483
(0.234) (0.373)
Change in Individual and Collective 0.390
Dismissal Regulation2 (1.653)
Expected Growth 0.414**
(0.197)
Change in Share of Service Sector 0.0424
Workers2 (0.0308)
Union Density Rate (Level) 0.542 0.302
(0.510) (0.302)
First-Stage F-statistics above 10 yes yes yes yes yes yes yes yes yes
Country Fixed Effects yes yes yes yes yes yes yes yes yes
Year Fixed Effects no no no no no no no no no
Number of Observations 361 311 274 241 203 361 361 221 219
R2 0.678 0.676 0.654 0.612 0.632 0.682 0.680 0.264 0.369
Source: IMF staff calculations.
Note: Dependent variable = annual growth rates of compensation per hour of workers excluding the self-employed. Sample is of annual frequency from 2000 to 2016.
See Annex Table2.1.1 for countries in the sample. IV = instrumental variable. Standard errors in parentheses. * p < .10; ** p < .05; *** p < .01.
1The instrumental variable for lagged inflation is the two-quarter-lagged change in oil price.
2Relative to five years ago.
Annex Figure 2.3.2. Decomposition of Sectoral Wage Annex Figure 2.3.3. Effects on Part-Time Employment Share,
Dynamics, 200015 Sectoral Analysis
(Percentage-point change relative to 200007 average) (Percentage points)
2 0.6
4 0.8
6 1.0
200007 0812 13 14 15
1.2
2 2. Unemployment Rates Moderately Higher than 200007 Average
1.4
Five-year Five-year Five-year Percent Percent
1 percentage- percentage- percentage- increase in increase in
point change point change point change expected expected
0 in nal imports in price of in union growth growth
(sectoral) investment density rate (aggregate) (sectoral)
1 (aggregate) (aggregate)
2.5
0.0
2.5
5.0
7.5
200007 0812 13 14 15
Annex Table2.3.13. Drivers of Nominal Wage Growth, Employment Growth, and Part-Time Employment
(1) (2) (3) (4) (5)
Part-Time
Nominal Wage Growth1 Employment Growth1 Employment1
3SLS2 3SLS3 3SLS2 3SLS3 3SLS3
Aggregate Output gap 0.284*** 0.241*** 0.0682 0.0513 0.0907
(0.0334) (0.0322) (0.0786) (0.0641) (0.0816)
Correlation of Sectoral and Aggregate 0.388** 0.412** 1.001** 0.226 0.698*
Output Growth (0.172) (0.163) (0.404) (0.324) (0.413)
Aggregate Output Gap Correlation 0.321*** 0.269*** 0.606*** 0.644*** 0.000166
(0.0548) (0.0531) (0.129) (0.106) (0.135)
Lagged Inflation 0.207*** 0.210*** 0.0552 0.0967 0.131
(0.0573) (0.0578) (0.135) (0.115) (0.147)
Expected Growth (Sectoral) 0.0205 0.0226* 0.0700** 0.0337 0.106***
(0.0138) (0.0137) (0.0324) (0.0272) (0.0347)
Change in Final Imports as a Share of 0.0103 0.00520 0.0280 0.0218 0.0588**
Gross Output4 (0.0110) (0.00973) (0.0258) (0.0194) (0.0247)
Change in the Relative Price of Investment4 0.102*** 0.115*** 0.0294 0.0133 0.0683
(0.0228) (0.0213) (0.0537) (0.0424) (0.0540)
Country Fixed Effects yes yes yes yes yes
Sector Fixed Effects yes yes yes yes yes
Year Fixed Effects no no no no no
Number of Observations 1,833 1,526
Source: IMF staff calculations.
Note: Sample is of annual frequency from 2000 to 2015. See Annex Table2.1.1 for countries in the sample. 3SLS = three-stage least squares. Standard errors
in parentheses. * p < .10; ** p < .05; *** p < .01.
1The dependent variable of the regression, defined as annual growth rates and share of total employment.
2System estimated using 3SLS for nominal wage growth and employment growth as the endogenous dependent variables.
3System estimated using 3SLS for nominal wage growth, employment growth, and the share of part-time employment as the endogenous dependent
variables.
4Relative to five years ago.
Great Recession: France versus Spain. The Economic Journal, Paper 20479, National Bureau of Economic Research,
122, F155F187. Cambridge, MA.
Berman Eli, John Bound, and Zvi Griliches. 1994. Changes Dew-Becker, Ian, and Robert J. Gordon. 2005. Where Did
in the Demand for Skilled Labor within US Manufacturing: Productivity Growth Go? Inflation Dynamics and the Distri-
Evidence from the Annual Survey of Manufacturers. Quar- bution of Income. Brookings Papers on Economic Activity 36
terly Journal of Economics 109 (2): 36797. (2): 67150.
Blanchard, Olivier, and Francesco Giavazzi. 2003. Macroeco- DiNardo, John, and David Card. 2002. Skill Biased Technolog-
nomic Effects of Regulation and Deregulation in Goods ical Change and Rising Wage Inequality: Some Problems and
and Labor Markets. Quarterly Journal of Economics 118 Puzzles. Journal of Labor Economics 20 (4): 73383.
(3): 879907. Du Caju, Philip, Gbor Ktay, Ana Lamo, Daphne Nicolit-
Blanchard, Olivier, and Lawrence F. Katz, 1997. Wage Dynam- sas, and Steven Poelhekke.2010. Inter-Industry Wage
ics: Reconciling Theory and Evidence. American Economic Differentials in EU Countries: What Do Cross-Country
Review 89 (2): 6974. Time-Varying Data Add to the Picture? European Central
Blanchflower, David G., and Adam S. Posen. 2014. Wages and Bank Working Paper189, Frankfurt.
Labor Market Slack: Making the Dual Mandate Operational. Duval, Romain, Gee Hee Hong, and Yannick Timmer. 2017.
PIIE Policy Brief 1410. Financial Frictions and the Great Productivity Slowdown.
Borjas, George J., and Valerie A. Ramey. 1995. Foreign Compe- IMF Working Paper 17/129, International Monetary Fund,
tition, Market Power, and Wage Inequality. Quarterly Journal Washington DC.
of Economics 110 (4): 1075110. Erdil, Erkan, and I. Hakan Yetkiner.2001. A Comparative
Borowczyk-Martins, Daniel, Gregory Jolivet, and Fabien Analysis of Inter-Industry Wage Differentials: Industrial-
Postel-Vinay. 2013. Accounting for Endogeneity in Match- ized versus Developing Countries. Applied Economics 33
ing Function Estimation. Review of Economic Dynamics 16 (13): 163948.
(3): 44051. European Central Bank (ECB).2009. Wage Dynamics in
Brainard, Lael.2016. The Gig Economy: Implications of the Europe: Final Report of the Wage Dynamics Network.
Growth of Contingent Work. Speech at the Evolution of Directorate General Research, Frankfurt.
Work conference, November17. https://www.federalreserve . 2017. Assessing Labor Market Slack. Economic
.gov/newsevents/speech/brainard20161117a.htm. Bulletin 3.
Burda, Michael C., and Stefanie Seele. 2016. No Role for the European Commission (EC).2003. Wage Structures and
Hartz Reforms? Demand and Supply Factors in the German Determinants in an Enlarged Europe. In Employment in
Labor Market, 19932014. SFB 649 Discussion Paper Europe2003: Recent Trends and Prospects. Directorate-General
2016010, Humboldt University, Berlin. for Employment and Social Affairs, Frankfurt.
Burtless, Gary. 1995. International Trade and the Rise in Earn- European Union (EU). 2015. Wage and Income Inequality
ings Inequality. Journal of Economic Literature 33 (2): 80016. in the European Union. Directorate-General for Internal
Byrne, David M., John G. Fernald, and Marshall B. Reinsdorf. Policies, Policy Department Economic and Scientific Pol-
2016. Does the United States Have a Productivity Slowdown icy, Frankfurt.
or a Measurement Problem? Working Paper Series 20163, Feenstra, Robert, and Gordon Hanson. 1996. Foreign Invest-
Federal Reserve Bank of San Francisco. ment, Outsourcing and Relative Wages. In The Political
Cho, Sang-Wook (Stanley), and Julan P. Daz. 2016. Skill Economy of Trade Policy: Paper in Honour of Jagdish Bhagwati,
Premium Divergence: The Roles of Trade, Capital, and edited by Robert Feenstra, Gene Grossman, and Douglas
Demographics. The Future of Europe: Central and Eastern Irwin, 89127. Cambridge, MA: MIT Press.
Europe in a Comparative Perspective conference. . 2001. Global Production Sharing and Rising
Council of Economic Advisers. 2014. Economic Report of the Inequality: A Survey of Trade and Wages. NBER Work-
President. Washington, DC. ing Paper 8372, National Bureau of Economic Research,
Daly, Mary, and Bart Hobijn.2015. Why Is Wage Growth Cambridge, MA.
So Slow? Federal Reserve Bank of San Francisco Economic Fernald, John G. 2014. Productivity and Potential Output
Letter201501, January. before, during, and after the Great Recession. NBER Work-
Danninger, Stephan.2016. Whats Up with USWage Growth ing Paper 20248, National Bureau of Economic Research,
and Job Mobility? IMF Working Paper 16/122, Interna- Cambridge, MA.
tional Monetary Fund, Washington DC. Gal, Peter, and Alexander Hijzen. 2016. The Short-Term
Das, Mitali, and Benjamin Hilgenstock. Forthcoming. Labor Impact of Product Market Reforms: A Cross-Country
Market Consequences of Routinization in Developed and Firm-Level Analysis. IMF Working Paper 16/116, Interna-
Developing Economies. Unpublished. tional Monetary Fund, Washington DC.
Davis, Steven J., and John Haltiwanger. 2014. Labor Mar- Gali, Jordi.2011. The Return of the Wage Phillips Curve.
ket Fluidity and Economic Performance. NBER Working Journal of the European Economic Association 9 (3): 43661.
Golden, Lonnie. 2016. Still Falling Short on Hours and In The Factory Free Economy, edited by Lionel Fontagne and
Pay. Economic Policy Institute, Washington, DC. Ann Harrison. https://www.researchgate.net/publication/
http://www.epi.org/publication/still-falling-short-on-hours 303374155_The_Factory_Free_Economy.
-and-pay-part-time-work-becoming-new-normal/. Krueger, Alan.2015. How Tight Is the Labor Market? Martin
Goos, Maarten, Alan Manning, and Anna Salomons. 2014. Feldstein Lecture, National Bureau of Economic Research.
Explaining Job Polarization: Routine-Biased Technological , Judd Cramer, and David Cho. 2014. Are the
Change and Offshoring. American Economic Review 104 Long-Term Unemployed on the Margins of the Labor
(8): 250926. Market? Brookings Papers on Economic Activity, Brookings
Gordon, Robert. 2013. The Phillips Curve Is Alive and Well: Institution, Washington, DC.
Inflation and the NAIRU during the Slow Recovery. NBER Kumar, Anil, and Pia Orrenius. 2016. A Closer Look at the
Working Paper 19390, National Bureau of Economic Phillips Curve Using State-Level Data. Journal of Macroeco-
Research, Cambridge, MA. nomics 47: 84102.
Haldane, Andrew G. 2017. Work, Wages, and Monetary Policy. Leamer, Edward. 1992. Wage Effects of a US-Mexican Free
Speech given at the National Science and Media Museum, Trade Agreement. NBER Working Paper 3991, National
Bradford, United Kingdom, June 20. http://www.bankofengland Bureau of Economic Research, Washington, DC.
.co.uk/publications/Pages/speeches/2017/984.aspx. . 1996. Wage Inequality from International Com-
Hall, Robert E. 2005. Employment Fluctuations with Equilibrium petition and Technological Change: Theory and Country
Wage Stickiness. American Economic Review 95 (1): 5065. Experience. American Economic Review Papers and Proceedings
International Monetary Fund (IMF). 2016. Japan Article IV 86 (2): 30914.
Consultation Report. IMF Staff Country Report 16/267, Manning, Alan. 1993. Wage Bargaining and the Phillips Curve:
Washington, DC. The Identification and Specification of Aggregate Wage Equa-
Izquierdo, Mario, Juan Francisco Jimeno, Theodora Kosma, Ana tions. CEP Discussion Paper dp0062, Centre for Economic
Lamo, Stephen Millard, Tairi Rm, and Eliana Viviano. Performance, London School of Economic, London.
2017. Labour Market Adjustment in Europe During the Mortensen, Dale T., and Christopher A. Pissarides. 1999. Job
Crisis: Microeconomic Evidence From the Wage Dynamics Reallocation, Employment Fluctuations and Unemployment.
Network Survey. European Central Bank Occasional Paper Handbook of Macroeconomics 1, 1171228. Amsterdam: Elsevier.
Series 192. European Central Bank, Frankfurt. Organisation for Economic Co-operation and Development
Kang, Joong Shik, and Jay C. Shambaugh. 2014. Progress (OECD). 2017. Economic Policy Reforms 2017: Going for
towards External Adjustment in the Euro Area Periphery Growth. Paris.
and the Baltics. IMF Working Paper 14/131, International Parteka, Aleksandra. 2010. Skilled-Unskilled Wage Gap in
Monetary Fund, Washington, DC. the Enlarged EU: Sectoral Analysis Based on the Experi-
Karabarbounis, Loukas, and Brent Neiman. 2014. The Global ence of EU-15 Countries and Five New Member States.
Decline of the Labor Share. Quarterly Journal of Economics 12th European Trade Study group conference, Lausanne,
129 (1): 61103. Switzerland.
Katz, Lawrence, and David Autor. 1999. Changes in the Wage Rudebusch, Glenn, and John Williams. 2014. A Wedge in the
Structure and Earnings Inequality. In Handbook of Labor Dual Mandate: Monetary Policy and Long-Term Unem-
Economics (3), edited by Orley Ashenfelter and David Card. ployment. Federal Reserve Bank of San Francisco Working
Amsterdam: Elsevier. Paper 201414.
Katz, Lawrence, and Alan Krueger.2016. The Rise and Smith, Christopher L. 2014. The Effect of Labor Slack on
Nature of Alternative Work Arrangements in the United Wages: Evidence from State-Level Relationships. FEDS
States,19952015. NBER Working Paper 22667, National Notes, Federal Reserve Board, Washington, DC.
Bureau of Economic Research, Cambridge, MA. Stock, James H. 2011. Discussion of Ball and Mazumder,
Katz, Lawrence, and Kevin Murphy. 1992. Changes in Relative Inflation Dynamics and the Great Recession. Brookings
Wages, 196387: Supply and Demand Factors. Quarterly Papers on Economic Activity 42 (1): 387402.
Journal of Economics 107 (1): 3578. Taylor, John.2016. The Staying Power of Staggered Wage and
Kiley, Michael. 2014. An Evaluation of the Inflationary Price Setting Models in Macroeconomics. NBER Work-
Pressure Associated with Short- and Long-term Unem- ing Paper22356, National Bureau of Economic Research,
ployment. FEDS Notes 201428, Federal Reserve Board, Cambridge, MA.
Washington, DC. Trigari, Antonella. 2014. Commentary: Long-Term Non-
Koeniger, Winfried, Marco Leonardi, and Luca Nunziata.2007. employment and Job Displacement. Paper prepared for
Labor Market Institutions and Wage Inequality. Industrial Re-Evaluating Labor Market Dynamics, August 2123,
and Labor Relations Review 60 (3): 34056. Jackson Hole, Wyoming.
Kramarz, Francis. 2017. Offshoring, Wages, and Employment: Valletta, Robert G. 2016. Recent Flattening in the Higher
Evidence from Data Matching Imports, Firms, and Workers. Education Wage Premium: Polarization, Skill Downgrading,
or Both? NBER Working Paper 22935, National Bureau of .1995. How Trade Hurt Unskilled Workers. Journal of
Economic Research, Cambridge, MA. Economic Perspectives 9 (3): 5780.
Watson, Mark W. 2014. Inflation Persistence, the NAIRU, and Wooldridge, Jeffrey. 2009. On Estimating Firm-Level Produc-
the Great Recession, American Economic Review: Papers and tion Functions Using Proxy Variables to Control for Unob-
Proceedings 104 (5): 3136. servables. Economic Letters 104 (3): 11214.
Wood, Adrian. 1991. How Much Does Trade with the South Yellen, Janet L. 2005. Productivity and Inflation. Federal
Affect Workers in the North? World Bank Research Observer Reserve Bank of San Francisco Economic Letter.
6: 1936, Washington, DC. . 2014. Labor Market Dynamics and Monetary
. 1994. North-South Trade, Employment and Inequal- Policy. Speech at Federal Reserve Bank of Kansas City
ity: Changing Fortunes in a Skill-Driven World. Oxford: Economic Symposium, Jackson Hole, Wyoming. https://www
Clarendon Press. .federalreserve.gov/newsevents/speech/yellen20140822a.htm.
Global temperatures have increased at an unprecedented over long periods, such as fluctuations in and out of
pace over the past 40 years, and significant further the Ice Ages. However, the speed at which the climate
warming could occur, depending on our ability to restrain has changed over the past 3040 years appears to be
greenhouse gas emissions. This chapter finds that increases unprecedented in the past 20,000 years (Figure3.1).1
in temperature have uneven macroeconomic effects, with Most scientists agree that global temperatures are set to
adverse consequences concentrated in countries with rise further, at a scale and pace very much dependent
relatively hot climates, such as most low-income coun- on our ability to restrain greenhouse gas emissions,
tries. In these countries, a rise in temperature lowers per the central cause of global warming (IPCC2013).
capita output, in both the short and medium term, by Extreme weather events, such as heat waves, droughts,
reducing agricultural output, suppressing the produc- and floods, are likely to become more frequent, and
tivity of workers exposed to heat, slowing investment, sea levels will rise. Although considerable uncertainty
and damaging health. To some extent, sound domestic surrounds temperature projections, the scientific
policies and development, in general, alongside invest- consensus predicts that without further action to tackle
ment in specific adaptation strategies, could help reduce climate change, average temperatures could rise by 4C
the adverse consequences of weather shocks. But given the or more by the end of the 21st century. Very substan-
constraints faced by low-income countries, the interna- tial cuts to current emissions will be needed to limit
tional community must play a key role in supporting warming to less than 2C. Will climate change have
these countries efforts to cope with climate changea significant macroeconomic consequences, especially in
global threat to which they have contributed little. And low-income developing countries that tend to be more
while the analysis of the chapter focuses on the impact exposed to the vagaries of the weather? And how can
of weather shocks in low-income countries, most coun- these countries cope with the rises in temperature they
tries will increasingly feel direct negative effects from are set to experience over the coming decades?
unmitigated climate change through warming above Pinning down the economic consequences of cli-
optimal levels in currently cooler countries, more frequent mate change is difficult. Temperature increases of the
natural disasters, rising sea levels, loss of biodiversity, magnitude that could potentially occur over the next
and adverse spillovers from vulnerable countries. Looking centuryand many other aspects of climate change,
ahead, only continued international cooperation and a such as rapid rise in sea levels, ocean acidification, and
concerted effort to stem the man-made causes of global the likesit well outside recent (and relevant) his-
warming can limit the long-term risks of climate change. torical experience and could affect a large number of
countries. Extrapolating from the historically observed
relationship between activity and weather patterns
could also be problematic as populations adapt to per-
Introduction
sistent changes in climate. Yet studying the macroeco-
Since the turn of the 20th century, the Earths nomic effects of annual variation in weather patterns
average surface temperature has increased significantly.
Sizable swings in global temperatures used to happen
1Climate refers to a distribution of weather outcomes for a
Figure 3.1. Average Global Temperature tures. Since low-income countries are concentrated
(Degrees Celsius) in geographic areas with hotter climates, the Burke,
Hsiang, and Miguel (2015a) findings suggest that a
The average global temperature has risen at an extraordinary pace over the past
century, and signicant further warming could occur. rise in temperature would be particularly harmful for
this set of economies.
20 1. From 20,000 BCE to Present Countries negatively affected by climate change will
Early farming
begins
need to increase their resilience to rising temperatures
19 and extreme weather events, both by enhancing their
Most mammoths ability to smooth out shocks, which could become
18 go extinct Industrial more frequent, and by investing in adaptation strat-
revolution
17 begins, ca. egies, such as activity diversification, infrastructure
The Great 1760 investment, and technology innovation, that reduce the
Pyramid
16 harm they do. Populations may also respond to chang-
Last glacial completed 1920
period ends ing climatic conditions by relocating geographically,
15
20,000 16,000 12,000 8,000 4,000 0 2014 which could have important cross-border ramifications.
But the evidence on which policies may help countries
7 2. Recent and 21st Century Projections and individuals cope with weather shocks is limited.
(Deviations from 18801910 average)
6 Understanding the macroeconomic effects of
5 Observed weather shocks and the scope for policy actions to
RCP 8.5 mean prediction
4 RCP 4.5 mean prediction moderate them will be crucial for low-income devel-
3 oping countries to achieve durable growth in the long
2 terma precondition for convergence and imple-
1 mentation of the United Nations Sustainable Devel-
0 opment Goals.
1 Drawing from and building on the existing litera-
1880 1900 20 40 60 80 2000 20 40 60 80 2100 ture, this chapter contributes to the policy debate by
examining the following questions:
Sources: Intergovernmental Panel on Climate Change (IPCC) Coupled Model
Intercomparison Project Phase Five AR5 Atlas subset; Marcott and others (2013);
What has been the historical relationship between
Matsuura and Willmott (2007); National Aeronautics and Space Administration temperature and precipitation shocks and economic
(NASA) Goddard Institute for Space Studies; Royal Netherlands Meteorological activity in both the short and the medium term?
Institute Climate Change Atlas; Shakun and others (2012); and IMF staff
Are low-income countries particularly vulnerable?
calculations.
Note: In panel 2, the thin lines represent each of the 40 models in the IPCC WG1 Through what channels do weather fluctuations
AR5 Annex I Atlas, where a model with different parametrization is treated as a affect the economy? And has the sensitivity of
separate model. The thick lines represent the multimodel mean. Representative
Concentration Pathways (RCP) are scenarios of greenhouse gas concentrations,
growth to weather shocks changed over time?
constructed by the IPCC. RCP 4.5 is an intermediate scenario, which assumes How can countries, particularly low-income ones,
increased attention to the environment, with emissions peaking around 2050 and cope with weather shocks? Can policies and other
declining thereafter. RCP 8.5 is an unmitigated scenario in which emissions
continue to rise throughout the 21st century. country characteristics mitigate the macroeconomic
response to weather fluctuations?
Given the projected path of temperature by the end
of the 21st century, what might be the impact of
could produce useful insights.2 In an influential
climate change on low-income countries?
study, Dell, Jones, and Olken (2012) find that higher
temperatures significantly reduce economic growth
To address these questions, the chapter starts by
in low-income countries. Burke, Hsiang, and Miguel
documenting the historical evolution and projected
(2015a) provide evidence that productivity peaks at
change in temperature and precipitation patterns across
about 13C and declines strongly at higher tempera-
broad country groups according to leading climate
change models, as well as these groups contributions
2Dell, Jones, and Olken (2014); Carleton and Hsiang (2016); and
to greenhouse gas emissions. It then examines the
Heal and Park (2016) provide surveys of the new climate literature,
which explores the impact of weather fluctuations on a broad range historical evidence on the macroeconomic effects of
of economic variables. annual variation in temperature and precipitation
across a large sample of economies, highlighting the of temperature shocks. Higher temperatures also
channels through which climatic conditions affect the constrain growth in hot regions of emerging market
macroeconomy. The chapter offers evidence on how and developing economies significantly more than in
various policies and country characteristics influence hot regions of advanced economies, which corrob-
the sensitivity of growth to weather variations, using orates the importance of development in reducing
both empirical analysis and model simulations, and vulnerability.
presents case studies of various climate change adap- The temperature increase projected by 2100 under
tation strategies. Finally, the chapter incorporates the a scenario of unmitigated climate change implies
empirical estimates of economic loss from weather significant economic losses for most low-income
shocks and projected changes in temperature into a countries. Under the conservative assumption that
dynamic general equilibrium model to trace the poten- weather shocks have permanent effects on the level,
tial long-term effects of climate change. rather than the growth rate, of per capita output,
The chapters main findings are as follows: model simulations suggest that the per capita GDP
The rise in temperature over the past century has of a representative low-income country would be
been broad based. No country has been spared from 9percent lower in 2100 than it would have been in
the warming of the Earths surface, and no coun- the absence of temperature increases, with the pres-
try is projected to be spared further temperature ent value of output losses amounting to more than
increases, with the largest increases in temperature 100percent of current GDP when discounted at the
expected in countries with relatively colder cli- growth-adjusted rate of 1.4percent.
mates. The contribution of low-income developing
countrieswhich tend to be situated in some of the Taken together, these findings paint a worrisome
hottest geographic areas on the planetto atmo- picture. Rising temperatures would have vastly unequal
spheric greenhouse gas concentrations is negligible, effects across the world, with the brunt of adverse
both in absolute terms and on a per capita basis. consequences borne by those who can least afford it. In
The macroeconomic effect of temperature shocks all likelihood, most countries will increasingly feel the
is uneven across countries. Confirming the global direct impact of unmitigated climate change, through
nonlinear relationship between annual tempera- warming above optimal temperatures, more frequent
ture and growth uncovered by Burke, Hsiang, and (and more damaging) natural disasters, rising sea levels,
Miguel (2015a) using an expanded data set, the loss of biodiversity, and many other hard-to-quantify
empirical analysis suggests that rising temperatures effects. In addition, climate change is likely to cre-
lower per capita output in countries with relatively ate economic winners and losers at both individual
high annual average temperature, such as most and sectoral levels, even in countries where the effect
low-income countries. In these economies, the might be moderate or positive on average. However,
adverse effect is long-lasting and operates through low-income countries will suffer disproportionately
several channels: lower agricultural output, depressed from further temperature increasesa global threat
labor productivity in sectors more exposed to the to which they have contributed little. And within
weather, reduced capital accumulation, and poorer low-income countries, the poor would likely be the
human health. Moreover, data indicate that macro- most heavily affected by climate change (Hallegatte
economic outcomes have not become any less sensi- and Rozenberg 2017). Having little influence on the
tive to temperature shocks in recent years, pointing future course of climate, how can these countries cope
to significant adaptation constraints. with the challenges they face as temperatures rise?
To some extent, sound policies and institutional The findings of this chapter suggest that domestic
frameworks, investment in infrastructure, and other policies can partially dampen the adverse effects of
adaptation strategies can reduce the damage from weather shocks. Improving buffers and strengthening
temperature shocks in hot countries. Although well-targeted social safety nets that can deliver sup-
causal interpretation is difficult, empirical evidence port when needed would help countries smooth some
suggests that countries with better-regulated capital of the instantaneous effects of weather shocks, while
markets, higher availability of infrastructure, flexible policies and institutions that make capital and labor
exchange rates, and more democratic institutions markets more flexible and foster structural economic
recover somewhat faster from the negative impacts transformation could help countries recover somewhat
faster and reduce their vulnerability to future shocks. It is important to highlight from the outset the
Adaptation strategies that reduce specific climate inherent difficulty of quantifying the potential mac-
change effects and risks, such as targeted infrastruc- roeconomic consequences of climate change. Extrap-
ture projects, adoption of appropriate technologies, olating from historically observed weather responses
and mechanisms to transfer and share these risks of GDP to the long-term effect of global warming is
through financial markets, could also be part of the challenging for several reasons.3 On one hand, such
toolkit for reducing the economic damage caused by an extrapolation may overstate the impact as govern-
climate change. ments and other economic agents take ameliorative
But putting in place the right policies will be par- actions, make investments, or develop new technolo-
ticularly difficult in low-income countries, which have gies that help populations adapt to persistent changes
huge spending needs and limited ability to mobilize in climate. On the other hand, the actual impact could
the resources necessary for adaptation in a challeng- be larger if there are nonlinearities in the response as
ing economic environment. In some cases, political the climate shifts to conditions beyond recent expe-
uncertainty and security issues exacerbate the chal- rience.4 Moreover, the chapter does not separately
lenge. Moreover, even when in place, domestic policies quantify the effects of natural disasters, whose higher
alone cannot fully insulate low-income countries from projected frequency may amplify the damages they
the adverse consequences of climate change, as higher cause; it does not analyze distributional impacts across
temperatures push the biophysical limits of these coun- sectors and households within countries, which may
tries ecosystems, potentially triggering more frequent be quite sizable; nor does it shed light on the conse-
epidemics, famines, and other natural disasters, along quences of many aspects of climate change, such as a
with armed conflict and refugee flows. The interna- rapid rise in sea levels, ocean acidification, and the like,
tional spillovers from these difficult-to-predict effects that have no historical precedent but could have very
of climate change could be very considerable. large macroeconomic consequences.5 Nevertheless, as
Climate change is a negative global externality of long as the Earth continues to warm over the rest of
potentially catastrophic proportions, and only collec- the 21st century in the same pattern as over the past
tive action and multilateral cooperation can effec- 50 yearsa stochastic series of annual shocks along
tively address its causes and consequences. Mitigating an upward trendthis chapter may provide valuable
climate change requires radically transforming the guidance on climate change vulnerabilities and adapta-
global energy system, including through the use of tion needs under the current production technologies
fiscal instruments to better reflect environmental costs and geographic distribution of populations (Dell,
in energy prices and promote cleaner technologies as Jones, and Olken 2012).
discussed in Box3.6. Adapting to the consequences of
climate change necessitates vast investments, including
3Dell, Jones, and Olken (2014); Carleton and Hsiang (2016);
in boosting infrastructure, reinforcing coastal zones,
Hsiang (2016); and Lemoine (2017) provide discussions of the
and strengthening water supply and flood protec-
conditions under which empirical estimates of the effect of weather
tion (Margulis and Narain 2010; UNEP 2016). The shocks based on historical data can shed light on the consequences
international community will have a key role to play of climate change.
4For example, the historically observed natural year-to-year
in fostering and coordinating financial and other
temperature variability for countries located in the tropics is roughly
types of support for affected low-income countries. 0.5C. The projected increase in temperature for these countries
With advanced and emerging market economies between 2005 and 2100 under the extreme unmitigated climate
contributing the lions share to the warming that has change scenario is 4.1Cin other words, more than 8.5 times
larger than the current natural variability, implying a totally new
occurred so far and is projected to continue, helping climatic regime (see also World Bank 2013).
low-income countries cope with its consequences is 5A large body of literature studies the macroeconomic impact of
a humanitarian imperative and sound global eco- natural disasters (see, for example, Noy 2009; Cavallo and others
2013; Acevedo2014; Felbermayr and Grschl 2014; Cabezon and
nomic policy. In the future, only continued interna-
others2015; IMF2016a; IMF2016b; Gerling, forthcoming; and
tional cooperation and a concerted effort to stem the Gerling, Moreno Badia, and Toffano, forthcoming). The chapter
man-made causes of global warming can limit the focuses on direct measures of the weather because natural disaster
long-term risks of climate change (IPCC 2014; IMF data may suffer from reporting and mismeasurement issues. Mismea-
surement could be a particular problem in low-income countries,
2015; Stern 2015; Farid and others 2016; Hallegatte which typically have lower capacity to accurately evaluate, record,
and others 2016). and report damage (Jennings 2011).
Temperature and Precipitation: Historical Figure 3.2. Increase in Average Global Temperature and
Patterns and Projections Contributions of Key Factors
(Deviations from 18801910 average, degrees Celsius)
This section sets the context for the rest of the
chapter by summarizing the scientific consensus on According to the Intergovernmental Panel on Climate Change, most of the
how climate and one of its key man-made drivers increase in temperature since 1950 can be attributed to human factors.
greenhouse gas emissionshave evolved over the past
century. The section then presents scientists projected Temperature GHG Human Natural
changes for the rest of the 21st century and discusses
the link between temperature, precipitation, and 1.2
weather-related disasters.
1.0
0.8
Historical Patterns
Global temperatures have increased by roughly 1C 0.6
compared with the 18801910 average (Figure3.2).
0.4
The rise started in earnest in the 1970s, following a
large increase in carbon dioxide (CO2) emissions.6 0.2
Although natural factors explain some of the warming
over the past century, according to the Intergovern- 0.0
mental Panel on Climate Change (IPCC), more than
half of the temperature increase since 1950 can be 0.2
attributed to human activity (IPCC 2014).
0.4
The increase in temperature has occurred in all 1880 1900 20 40 60 80 2000 16
regions, with the same accelerating trend, starting in
the 1970s (Figure3.3).7 The median temperature over Sources: Carbon Dioxide Information Analysis Center; National Aeronautics and
the first 15 years of this century, compared with the Space Administration (NASA) Goddard Institute for Space Studies; Roston and
Migliozzi (2015); and IMF staff calculations.
first 15 years of the past century, was 1.4C higher in
Note: The lines present the actual increase in land and ocean surface air
advanced economies, 1.3C higher in emerging market temperature relative to 18801910 and the increase predicted by different
economies, and 0.7C higher in low-income devel- factors. Human factors include land use, ozone emissions, aerosol emissions, and
GHG emissions. Natural factors include orbital changes, solar output, and volcanic
oping countries. Even though most of the warming activity. The contribution of each factor is estimated by ModelE2 by NASA
occurred in advanced economies, by 2015 the tem- Goddard Institute for Space Studies. GHG = greenhouse gases.
perature in the median low-income developing country
(25C) was more than twice that of the median
advanced economy (11C).
CO2 emissions have grown rapidly since the 1950s
Other aspects of the climate have also changed
across all income groups, along with rising incomes
appreciably. Since 1900, the global mean sea level has
and populations (Figure3.4). However, emissions in
risen by 1721 centimeters. As with temperature, there
low-income developing countries are still a fraction of
has been an increase in the pace at which the sea level
those in advanced and emerging market economies,
is rising: from 0.17 centimeter a year throughout most
in both aggregate and per capita terms. And although
of the 20th century to 0.32 centimeter a year over the
advanced economies have managed to contain their
past 20 years (IPCC 2014).
overall emissions over the past decade, in per capita
terms they still contribute vastly more than the rest of
the world.
6The three most important greenhouse gases, which are regulated
Figure 3.3. Temperature and Precipitation across Broad Figure 3.4. Annual CO 2 Emissions across Broad Country
Country Groups Groups
(Billion metric tons, unless noted otherwise)
Temperature has risen across all country groups, while precipitation does not
exhibit a clear pattern. CO2 emissions have grown rapidly since the 1950s across all income groups, but
emissions by low-income developing countries are negligible in both absolute and
Advanced Economies per capita terms.
8 600 0 0
1900 20 40 60 80 2000 15 1900 20 40 60 80 2000 15 1880 1990 20 40 60 80 2000 1880 1990 20 40 60 80 2000
20 600
1900 20 40 60 80 2000 15 1900 20 40 60 80 2000 15 Sources: Carbon Dioxide Information Analysis Center; and IMF staff calculations.
Note: AEs= advanced economies; CO 2 = carbon dioxide; EMs = emerging markets;
LIDCs = low-income developing countries.
Low-Income Developing Countries
change, Cook and others (2013) find that 97percent of the studies
expressing a position on the reasons behind global warming agree
that it is influenced by man-made causes. See also Cook and
others (2016).
of temperatures over the 21st century. The rest of the of fat tailsthe probability that catastrophic climate
chapter focuses on two of these scenarios: an interme- change can occurthat is behind calls for strong mit-
diate path (RCP 4.5) and an unmitigated path (RCP igation actions to reduce emissions and for adaptation
8.5), as shown in Figure3.1, panel 2.9 to prepare for significant shocks (Weitzman 2011).
Under the RCP 8.5 scenario of unmitigated climate
change, the average global temperature by 20812100
could rise by 3.7C (with a projected range of Weather-Related Disasters
2.6C4.8C).10 Warming would occur all over the As temperatures rise, the risks of extreme weather
globe, with larger increases over the northern hemi- events, such as floods, droughts, and heat waves, will
sphere, where some regions could experience tempera- increase (IPCC 2014). New statistical analysis suggests
tures almost 12C higher than in 2005 (Figure3.5). that projected climate change will likely bring more
Between 2005 and 2100, the increase for the median frequent weather-related disastersevents that cause
advanced economy is projected to be 4.4C, and 4.5C great damage or loss of life.11 This likelihood is partic-
for the median emerging market economy and median ularly important for low-income developing countries
low-income developing country. Increases are projected and small states, which historically have been much
to be smaller in absolute terms closer to the equator, more likely, relative to their land area, to experience
but are very significant when set against the historical natural disasters than advanced and emerging market
year-to-year and intrayear variability in temperature economies (Figure3.6, panel 1).12
observed in those locations. Change in precipitation Using monthly data from 1990 to 2014 on 8,000
will vary by region, with dry areas generally expected weather-related disasters, a statistical analysis uncovers
to become drier and wet regions expected to experience the historical relationship between the occurrence of a
an increase in rainfall. disaster and temperature and precipitation.13 It then
Under this scenario, the global mean sea level is combines the estimated elasticities and the projected
projected to rise by almost 0.8 meter by the end monthly temperature and precipitation in 2050 and
of the 21st century, exposing coastal areas, includ- 2100 under the RCP 8.5 scenario to forecast the
ing some large population centers, to higher risk likelihood of natural disasters. The results indicate
of flooding and erosion. Sea level rise will not be that most disaster types will be more common by the
uniform across regionsit is projected to be higher end of the century, across all country income levels.
than the global mean closer to the equator and less As depicted in Figure3.6, the frequency of disasters
than the global mean at high latitudes (IPCC 2014; caused by heat waves or tropical cyclones will increase
World Bank 2013). considerably (see Box3.1, which explores the effect
It is important once again to stress the large uncer- of tropical cyclones on economic activity).14 Similarly,
tainty surrounding climate change projections. Future
emissions depend on many factors that are difficult
11The International Disaster Database (EM-DAT) defines a natu-
to predict and, even for the same emission scenario, ral disaster as an event in which at least one of the following criteria
climate models differ widely in their temperature and is met: 10 or more people are reported killed, 100 or more people
precipitation projections (Figure3.1, panel 2). How- are reported affected, and either a declaration of a state of emergency
or a call for international assistance is made (Guha-Sapir, Below, and
ever, it is precisely this uncertainty and the possibility
Hoyois 2015).
12Low-income developing countries and small states, respectively,
9The Paris Agreement aims to contain the rise in temperature to are five and 200 times more likely to be hit by a weather-related
less than 2C (ideally to less than 1.5C) relative to the preindustrial natural disaster than the rest of the world, after controlling for
average, which would require policy efforts beyond those assumed country size.
under the RCP 4.5 scenario. Under the RCP 4.5 scenario, there is 13The probability of each disaster type (flood, tropical cyclone,
increased attention to the environment. CO2 emissions peak around and so on) is estimated using a panel logit with country fixed effects,
2050 and decline thereafter, with a resulting temperature increase of in which temperature and precipitation are the main explanatory
1.8C by 20812100 relative to 19862005 (a likely range of 1.1C variables. The analysis expands on Thomas and Lopez (2015) by
to 2.6C and a greater than 50percent chance of an increase exceed- modeling each disaster type separately and relying on monthly rather
ing 2C by 2100). Under the RCP 8.5 scenario, CO2 emissions grow than annual data. See Annex3.2 for further details.
throughout the 21st century. 14Scientists project that the frequency of tropical cyclone storms
10Under this scenario, the average increase in population-weighted will decrease, but their strength and intensity will rise in a warmer
temperature between 2005 and 2100 across the countries in the sam- world (Knutson and others 2010). This could lead to more natural
ple is projected to be 4.4C, with the median country experiencing disasters caused by more intense tropical cyclones despite the overall
warming of 4.5C. lower frequency of storms.
Figure 3.5. Temperature and Precipitation Projections under the RCP 8.5 Scenario
Under the scenario of continued increase in greenhouse gas emissions, temperatures across the globe are projected to rise signicantly.
12
10
8
6
4
2
1,500 +
1,000
500
0
500
1,000
Sources: National Aeronautics and Space Administration (NASA) Earth Exchange Global Daily Downscaled Projections (NEX-GDDP); World Bank Group Cartography
Unit; and IMF staff calculations.
Note: The NEX-GDDP data set comprises downscaled climate scenarios for the globe that are derived from the General Circulation Model (GCM) runs conducted under
the Coupled Model Intercomparison Project Phase 5 (CMIP5) and for two Representative Concentration Pathways (RCP) greenhouse gas emissions scenarios (4.5 and
8.5). The CMIP5 GCM runs were developed for the Intergovernmental Panel on Climate Change Fifth Assessment Report. The data set includes downscaled projections
from the 21 models and scenarios for daily maximum temperature, minimum temperature, and precipitation for 19502100. The spatial resolution of the data set is
0.25 degrees (~25 km x 25 km). mm = millimeter.
Figure 3.6. Natural Disasters: Historical and Projected floods and epidemics, which mainly affect low-income
Monthly Probability of Occurrence developing countries, will also become more common.
More frequent weather-related disasters, without a
Natural disasters, which have historically occurred with greater frequency in low-
corresponding increase in reconstruction capabilities,
income developing countries relative to their land area, could become more
common by the end of the 21st century under the scenario of continued increase in could amplify the damages they cause because econ-
greenhouse gas emissions. omies may have insufficient time to recover between
events (Hallegatte, Hourcade, and Dumas 2007).
201014 2050 2100
1. Natural Disasters by Type, 2. Tropical Cyclone 0.07 The Macroeconomic Impact of Weather Shocks
19902014
0.06 The design of appropriate policies to cope with
LIDCs Flood
Other 0.05 climate change requires an understanding of its poten-
8,476 0.04 tial macroeconomic consequences. In the absence of
disasters, historical experience with climate change that may be
in LIDCs 0.03
LIDCs relevant for countries today, the analysis in this section
2,606 LIDCs 0.02
Drought LIDCs builds on existing literature and identifies how annual
LIDCs 0.01 fluctuations in temperature and precipitation affect
Tropical cyclone Epidemic
0.00 macroeconomic performance in the short and medium
AEs EMs LIDCs
term. The channels through which macroeconomic
effects occur and the changes in the sensitivity of
0.10 3. Flood 4. Heat Wave 0.20
growth to weather shocks are explored, motivated by
0.08 0.16 evidence that higher temperatures constrain per capita
GDP growth in countries with hot climates.
0.06 0.12
Figure 3.7. Effect of Temperature Increase on Real per and growth, as demonstrated by Burke, Hsiang, and
Capita Output Miguel (2015a).15
(Percent)
The analysis confirms the existence of a statistically
In relatively hot countries, such as most low-income developing countries, an significant nonlinear effect of temperature on per capita
increase in temperature has a negative, statistically signicant, and long-lasting economic growth, first established by Burke, Hsiang,
effect on per capita output.
and Miguel (2015a), in this chapters substantially larger
Estimate 90 percent condence interval sample. In countries with high average temperatures,
Percent of countries (right scale) Median temperature an increase in temperature dampens economic activ-
Contemporaneous Effect Effect over Time ity, whereas it has the opposite effect in much colder
(Temperature on x-axis) (Years on x-axis) climates. The threshold temperature is estimated to be
1. Advanced Economies
about 13C to 15C (see Annex Table3.3.1).16 These
2 18 2. Advanced Economies 2
16 (T = 11C) results suggest highly uneven effects of warming across
1 1 the globe (Figures 3.7 and 3.8).
14
0 11 0 Because most advanced economies are in colder
9 locations, with annual average temperatures close to
1 7 1
the threshold, a marginal temperature increase does
5
2 2 not materially affect their contemporaneous growth
2
(Figure3.7, panel 1).17 Emerging market economies
3 0 3
0 6 10 14 18 22 26 30 1 0 1 2 3 4 5 6 7 and particularly low-income developing countries tend
2 3. Emerging Market 18 4. Emerging Market 2
15Average annual temperature and precipitation are constructed
Economies 16 Economies (T = 22C)
1 14 1 by aggregating weather data at the grid-cell level to the level of the
12 country using the population in each cell as weights to account
0 0 for differences in population density within countries and capture
10
8 the average weather experienced by a person in the country (see
1 1 Annexes3.1 and 3.3). The empirical approach consists of regressing
6
contemporaneous and future output growth on temperature and
2 4 2
precipitation and the squared terms to estimate an impulse response
2
function at various horizons, controlling for country fixed effects,
3 0 3
0 6 10 14 18 22 26 30 1 0 1 2 3 4 5 6 7 region-year fixed effects, lags and forwards of weather shocks, and
lagged growth. See Annex3.3 for further details.
2 5. Low-Income 18 6. Low-Income Developing 2 16The finding is robust to, among other things: (1) using alterna-
Developing Countries 15 Countries (T = 25C) tive sources of raw grid-level weather data, (2) aggregating grid-level
1 1
13 weather data to country averages with population weights from
0 0 different decades, (3) estimation through an autoregressive distribu-
10 tive lag specification instead of a local projection method, (4)using
1 8 1 country-specific linear and quadratic time trends as opposed to
5 region-year fixed effects, and (5) controlling for the occurrence of
2 2 natural disasters. The analysis does not find a consistently significant
3
relationship between precipitation and per capita GDP growth,
3 0 3 although it uncovers an effect of precipitation on agricultural output
0 6 10 14 18 22 26 30 1 0 1 2 3 4 5 6 7
(Annex Tables 3.3.1 and 3.3.2).
17Even if the effects on overall GDP in these countries are negli-
Source: IMF staff calculations.
Note: Left-hand-side panels superimpose the contemporaneous effect of a 1C gible, this may mask large losses and gains, with some sectors facing
increase in temperature on per capita output at different temperature levels large investment needs to cope with higher temperatures, rising sea
computed as per equation (3.3) over the distribution of average annual levels, or more damaging disasters. Moreover, the analysis focuses on
temperatures recorded in 2015 in advanced economies (panel 1), emerging the macroeconomic effects of a limited set of weather characteristics,
markets (panel 3), and low-income developing countries (panel 5). The blue lines namely temperature and precipitation. The negative impact of other
show the point estimates and 90 percent condence intervals, while the light blue aspects of the climate, such as the rise in sea levels or the occurrence
bars denote the percent of countries at each temperature level. The vertical red of extreme weather events, may be less unequal across broad income
line is the median temperature for the country group. Right-hand-side panels groups, as demonstrated in Box3.1, which documents similar out-
depict the impulse response of per capita output to a 1C increase in temperature
put losses from tropical cyclones across advanced and emerging mar-
estimated at the median temperature of advanced economies (panel 2), emerging
markets (panel 4), and low-income developing countries (panel 6). Horizon 0 is the ket economies. The estimates also abstract from potential spillovers
year of the shock. T = temperature. to advanced economies from famines, epidemics, social conflicts,
and other difficult-to-predict effects of weather shocks in vulner-
able economies. Moreover, under the scenario of unconstrained
CO2 emissions, most advanced economies will cross the threshold
temperature and would start suffering the negative effects of higher
temperatures on economic output (Annex Figure3.6.1).
Figure 3.8. Effect of Temperature Increase on Real per Capita Output across the Globe
(Percent)
An increase in temperature has a highly uneven effect across the globe, with adverse consequences concentrated in the parts of the world where the majority of
the worlds population lives.
1. Effect of a 1C Increase in Temperature on Real per Capita Output at the Grid Level
4.23
3.58
2.93
2.28
1.63
0.98
0.33
0.32
0.97
1.62
2.27
2. Effect of a 1C Increase in Temperature on Real per Capita Output at the Country Level, with Countries Rescaled in Proportion to Their Population
0.80 to 1.63
0.43 to 0.80
0.33 to 0.43
0.50 to 0.32
0.97 to 0.50
1.30 to 0.97
1.68 to 1.30
Insignicant effect
No data
Sources: Natural Earth; ScapeToad; United Nations World Population Prospects Database: the 2015 Revision; World Bank Group Cartography Unit;
and IMF staff calculations.
Note: The maps depict the contemporaneous effect of a 1C increase in temperature on per capita output computed as per equation (3.3). Panel 1 uses 2005 grid-
level temperature, while panel 2 uses the recent 10-year average country-level temperature together with estimated coefcients in Annex Table 3.3.1, column (5). In
the cartogram in panel 2, each country is rescaled in proportion to its 2015 population. Gray areas indicate the estimated impact is not statistically signicant.
to have much hotter climates, and a rise in tempera- Figure 3.9. Effect of Temperature Increase on Sectoral Output
ture significantly lowers per capita GDP growth. For Estimated at the Temperature of the Median Low-Income
the median emerging market economy, a 1C increase Developing Country
(Percent; years on x-axis)
from a temperature of 22C lowers growth in the
same year by 0.9percentage point. For the median
An increase in temperature lowers agricultural output, but also has adverse effects
low-income developing country, with a temperature on manufacturing value added in hot countries.
of 25C, the effect of a 1C increase in temperature is
even larger: growth falls by 1.2percentage points (Fig- Estimate 90 percent condence interval
ure3.7, panels 3 and 5).18 And even though countries
projected to be significantly affected by an increase in 2 1. Agriculture Value Added 2. Crop Production 2
temperature produced only about one-fifth of global 1 1
GDP in 2016, they are home to close to 60percent of 0 0
current global population and more than 75percent 1 1
of the projected global population at the end of the 2 2
century (Figure3.8 and Annex Figure3.3.1). 3 3
Does economic activity in countries with warmer 4 4
climates recover quickly after a rise in temperature? The 5 5
analysis suggests not. Even seven years after a weather 6 6
1 0 1 2 3 4 5 6 7 1 0 1 2 3 4 5 6 7
shock, per capita output is 1percent lower for the median
emerging market economy and 1.5percent lower for the
2 3. Manufacturing Value 4. Services Value Added 2
median low-income country (Figure3.7, panels 2, 4, and Added
1 1
6).19 A deepening in the shape of the estimated impulse
0 0
response of output to a temperature shock hints at the
1 1
possibility of a growth effect (and consequently much
2 2
larger economic losses from higher temperatures). How-
3 3
ever, statistically, it is not possible to reject the hypothesis
that the contemporaneous and medium-term effects of a 4 4
18There
so far has often studied these effects within a specific
are also substantial differences in the estimated effects
of temperature increases within each broad country group, which country or through laboratory experiments; this chapter
reflect the wide distribution of average temperature across countries examines whether these channels are also at work in a
(Figure3.7, panels 1, 3, and 5; Figure3.8). cross-country setting. Box3.1 extends the analysis in
19The persistence of the estimated effects may reflect the relatively
in Figure3.9, at the temperatures prevailing in the for sectors in which workers are directly exposed to
median low-income developing country, agricultural the weather.24
value added and crop production drop with higher Analysis of sectoral data on value added per worker
temperatures, recover somewhat in subsequent years, reveals that, at the temperatures prevailing in the
and generally remain depressed over the medium median low-income developing country, produc-
termmuch as expected and as documented in a large tivity of workers in heat-exposed industries falls
body of work.22 significantly after a rise in temperature (Figure3.10,
However, the analysis also confirms findings that panels 1 and 2). However, labor productivity is
industrial output is similarly hurt as temperatures rise unaffected in industries in which work is performed
in countries with hot climates, although the estimates mostly indoors.
are more imprecise (see also Dell, Jones, and Olken Overall productivity may also decline if weather
2012; Burke, Hsiang, and Miguel 2015a). Only shocks provoke political instability, incite conflict,
services sector output appears to be sheltered from or undermine governing institutions in other ways.
the weather. Although a more detailed analysis would be beyond
To shed light on the reasons weather shocks affect the scope of this chapter, numerous studies docu-
sectors besides agriculture, the analysis concentrates ment a strong link between weather shocks and these
on how key elements of the aggregate production outcomes.25 Since conflict is one of the key triggers of
functionproductivity and labor and capital inputs refugee flows, as discussed in Chapter1 of the April
respond to weather shocks. As in other studies, the 2017 World Economic Outlook (WEO), weather shocks
analysis aims to capture the net reduced-form effects of could result in substantial spillovers to neighboring
weather on various outcomes rather than uncover the countries and ultimately to advanced economies
potentially complex structural relationships that may through this channel.
exist among these variables.
Capital Accumulation
Productivity Temperature increases are largely supply-side shocks,
Evidence from surveys and other sources shows but they could lead to persistent output losses and
that exposure to heat above a certain point reduces affect growth if they influence the rate of factor accu-
peoples performance on both cognitive and physi- mulation.26 Using national accounts data, the analysis
cal tasks.23 The analysis therefore examines whether examines the response of the main components of
higher temperatures in parts of the world that are aggregate demandgross capital formation, consump-
hot decrease labor productivity. If productivity is a tion, exports, and importsto weather shocks within
channel through which weather shocks affect aggre- the empirical framework described above. At the tem-
gate GDP, the effect should be significantly larger
24The analysis follows Graff Zivin and Neidell (2014) and uses the
(2017) for evidence from the United States. Unlike per capita out- links climate to conflict. Forcible removal of rulers has also been
put, agricultural value added and crop production respond to precip- linked to fluctuations in climate (Burke and Leigh 2010; Dell, Jones,
itation, in addition to temperature shocks, with more precipitation and Olken 2012; Chaney 2013; Kim 2014), and several historical
generally boosting production. See Annex Table3.3.2. cases of societal collapse have been compellingly attributed to climate
23Seppnen, Fisk, and Faulkner (2003) report a productivity change (Cullen and others 2000; Haug and others 2003; Buckley
loss of about 2percent for every 1C increase in temperature and others 2010; Bntgen and others 2011).
above 25C, based on a survey of laboratory experiments. See also 26Investment may fall in response to temperature shocks because
Seppnen, Fisk, and Lei (2006) for a meta-analysis of the litera- there are fewer resources to invest, because the rate of return on
ture, Deryugina and Hsiang (2014) for evidence from the United capital is lower, and/or because the temporary negative shock to
States, and Somanathan and others (2017) for recent evidence on income raises the cost of financing investment in an environment
labor productivity from India. Heat stress may also reduce cognitive of imperfect capital markets (see, for example, Fankhauser and Tol
function, as captured in student performance (Wargocki and Wyon 2005). When access to formal savings, credit, or insurance is limited,
2007; Graff Zivin, Hsiang, and Neidell 2015; Garg, Jagnani, and households may also sell productive assets to smooth consumption in
Taraz 2017; Park 2017). response to weather shocks.
Figure 3.10. Effect of Temperature Increase on Productivity, perature of the median low-income country, all four
Capital, and Labor Input Estimated at the Temperature of the components respond negatively to a 1C increase in
Median Low-Income Developing Country temperature. However, in the medium term, the effect
(Percent; years on x-axis)
is most pronounced for investment, which is estimated
In hot countries, an increase in temperature dampens labor productivity in heat-
to be 6percent lower seven years after the shock
exposed industries, depresses investment and imports, and has damaging health (Figure3.10, panel 3). Imports, which are typically
effects. closely tied to investment, also exhibit a significant and
long-lasting drop as temperature rises (Chapter2 of
Estimate 90 percent condence interval the October 2016 WEO).27
Source: IMF staff calculations. hensive reviews of the literature on the link between temperature and
Note: The panels depict the effect of a 1C increase in temperature estimated at mortality and health. See, for example, Deschnes and Greenstone
the median low-income developing country temperature (25C). Horizon 0 is the (2011), Barreca (2012), and Barreca and others (2016) for evidence
year of the shock. Heat-exposed industries include agriculture, forestry, shing, and from the United States; Kudamatsu, Persson, and Strmberg (2012)
hunting, construction, mining, transportation, utilities, and manufacturing, for evidence from a subset of African countries; and Burgess and
following Graff Zivin and Neidell (2014). others (2014) for evidence from India. Carleton (2017) documents
a significant increase in suicide rates when higher temperatures
threaten agricultural yields in India. Deryugina and Hsiang (2014),
Graff Zivin and Neidell (2014), Park (2016), and Somanathan and
others (2017) find a direct effect of higher temperature on labor
supply and productivity.
to temperature shocks over rolling 20-year periods Figure 3.11. Effect of Temperature Increase on Real per
suggest that the relationship between the two variables Capita Output Estimated at the Temperature of the Median
has remained constant (Figure3.11).29 The reasons Low-Income Developing Country over Time
(Percent; years on x-axis)
behind this apparent lack of adaptation are not well
understood, but high costs, limited access to credit for The contemporaneous effect of temperature shocks on per capita output has
financing adaptation, insufficient information about remained relatively constant over time.
the benefits of adaptation, limited rationality in plan-
ning for future risks, and inadequate access to technol- 1.0
ogy are likely constraints, as discussed in Carleton and Estimate 90 percent condence interval
0.5
Hsiang (2016).
0.0
Coping with Weather Shocks and
0.5
Climate Change
This section examines how policies, institutions, 1.0
and other country characteristics can mitigate the
1.5
adverse consequences of temperature shocks and
climate change. It begins by discussing the toolkit
2.0
available to policymakers and private agents with
which to cope with weather shocks. It then presents 2.5
illustrative evidence of the extent to which, historically,
some policies (along with the overall level of develop- 3.0
1965 70 75 80 85 90 95
ment) have shaped the link between macroeconomic
performance and temperature shocks. The empirical Source: IMF staff calculations.
evidence is complemented in Box3.2 by dynamic Note: The gure depicts the effect of a 1C increase in temperature at horizon 0
estimated at the median low-income developing country temperature (25C), over
general equilibrium model scenarios of the response a 20-year rolling window. Each point estimate is for a period (t, t + 20).
of macroeconomic aggregates to weather shocks under
various proxies for relevant policies. Case studies of
specific adaptation strategies occupy Boxes 3.3 and 3.4. from weather shocks and from the risks that accom-
The section also examines migration as a response to pany climate change (Figure3.12).
persistent changes in climate as adaptation strategies Fluctuations in weather can be viewed as one of
reach their limits. Finally, the role of international many shocks that affect macroeconomic performance.
cooperation in supporting countries efforts to cope As such, their consequences could be attenuated by
with weather shocks and climate change is discussed. general macroeconomic and structural policies and
institutions that enhance countries ex ante and ex
A Toolkit post resilience to shocks. While priorities will vary
To structure the discussion, this subsection lays out depending on each countrys specific circumstances and
a toolkit of possible domestic policy actions and pri- weather-related threats, policies may include those that
vate choices that may help insulate economic activity seek to limit the short-term impact when shocks occur,
help the economy recover faster, and reduce vulnera-
29Studies reveal large differences in the ability of individual
bility to future shocks. Policies reinforce each other to
sectors to adapt to specific weather shocks. For example, Hsiang
achieve these goals. For example, countries with buffers
and Narita (2012) and Hsiang and Jina (2014) find that countries (fiscal and monetary space, large international reserves,
more frequently exposed to tropical cyclones experience less access to foreign aid) and well-targeted social safety
damage, which suggests that they have learned to cope with these
nets may be better placed to deliver support to people
extreme events. Mortality caused by high temperatures has declined
significantly over time with the introduction of air-conditioning affected by weather shocks, thus smoothing consump-
in the United States (Barreca and others 2016). But there is little tion in the short term. Adjusting to weather shocks
evidence of declining sensitivity of agricultural yields (Burke and and climate change will likely require reallocating peo-
Emerick 2016) or overall output (Dell, Jones, and Olken 2012;
Deryugina and Hsiang 2014; Burke, Hsiang, and Miguel 2015a) ple and capital across sectors and regions as production
to temperature fluctuations. and trade patterns shift. Policies and institutions that
Figure 3.12. Coping with Weather Shocks and Climate Change: A Toolkit
Policy buffers Labor market policies Public information provision Private and sovereign
To enable policy response To facilitate labor about climate-related risks insurance (for example,
movement across parametric insurance,
Well-targeted social production sectors and crop insurance,
regions Early warning systems and
safety nets catastrophe bonds)
evacuation schemes
To effectively support
those affected Education and health
policies
To strengthen human Stronger building laws, land use Multilateral risk-sharing
Exchange rate exibility capital, facilitate lifelong planning, and zoning rule; and better mechanisms
To cushion some of the learning, and develop a regulation of the use of common
economic cost of the exible and resilient labor resources (for example, water)
shock force
To reduce vulnerability
Fiscal incentives and appropriate
pricing for the development and
Financial sector policies adoption of appropriate
To ensure access to credit, insurance, and other technologies (for example,
nancial services needed by households to smooth resistant crops, air-conditioning,
consumption housing improvements)
To enable rms to invest, develop new technologies,
and so forth
Climate-smart infrastructure
Infrastructure investment investment (for example, irrigation,
drainage, seawalls)
facilitate the needed reallocation, such as those that public information provision, building codes, and
ensure access to finance, labor market flexibility, and land use and zoning laws, and devising warning and
investment in human capital and infrastructure, could evacuation systems, along with targeted incentives for
speed up recovery and foster the structural transforma- climate-related technologies (such as air-conditioning)
tion necessary to reduce vulnerability.30 and transferring and sharing risks related to weather
Mitigating the risks associated with climate change events (such as natural disasters, which may increase in
will also require some very specific adaptation policies frequency) through financial markets. Hard measures
to help countries reduce their exposure and vulnera- may include investment in climate-smart infrastruc-
bility to climatic events. Once the key climate change ture, such as retrofitting properties and building (or
risks are identified for a particular location, both soft upgrading) irrigation or drainage systems, building
and hard adaptation measures can be applied (Halle- seawalls, and the like.31 Appropriate adaptation mea-
gatte 2009). Soft measures may include strengthening sures are highly specific to the climate-related risks in
30The classification of policies presented in Figure3.12 is rather 31See Hallegatte (2009); Hallegatte, Lecocq, and de Perthuis
loose. Greater financial access could help farmers both smooth (2011); IPCC (2014); Cabezon and others (2015); OECD (2015a);
consumption when higher temperatures damage crops and invest Farid and others (2016); Hallegatte and others (2016); IMF (2016a);
in the technology needed to prevent future damage (such as buying and IMF (2016b) for a comprehensive discussion of various climate
heat-resistant seeds). change adaptation strategies.
each location and national circumstances; the infra- The results suggest that having the right policies
structure requirements for a flood-prone area would be and institutions in place may help attenuate the effects
vastly different from those of an area that is frequently of temperature shocks, to some extent. The instanta-
exposed to droughts. This specificity, together with lack neous effect of a temperature shock is slightly smaller
of comparable data on adaptation measures, precludes in countries with lower public debt, higher inflows of
cross-country empirical analysis. Case studies of foreign aid, and greater exchange rate flexibility. The
adaptation strategies, however, could prove insightful presence of monetary buffers (proxied by having below
and are presented in Box3.3. Box3.4 discusses the double-digit inflation) or international reserves makes
role of financial markets in sharing and transferring no notable difference (Figure3.13). However, the
weather-related risks. extent of attenuation that buffers provide is estimated
Important synergies exist between general macroeco- to be small and short lived.
nomic and structural policies and specific adaptation The evidence is somewhat more compelling for struc-
strategies: economic and institutional development tural policies and country characteristics that are typi-
will likely strengthen a countrys capacity to cope with cally deemed important for easing sectoral reallocation
climate change and to invest in specific adaptation of factors of production and structural transformation
strategies. For example, stronger institutions will make in general. Although the uncertainty surrounding the
enforcement of soft measures more effective, while fiscal empirical estimates is often very large, the medium-term
space will enable the investment in needed infrastruc- adverse effect of a temperature increase appears to fade
ture. Conversely, some adaptation strategies, such as when domestic and international financial markets are
efficient water use, climate-resilient housing, or activity better regulated, the exchange rate is flexible, infra-
diversification could facilitate development even in the structure is widely available, democratic institutions are
absence of climate change (Farid and others 2016). strong, and the distribution of income is fairly even
Finally, as adaptation strategies reach their limits, that is, in more-developed economies (Figure3.14).
economic agents could respond to persistent changes Patterns uncovered in the data broadly mirror sim-
in climate and the associated loss in income by relocat- ulations of a dynamic structural general equilibrium
ing geographically. model, which can properly isolate the causal effects of
the availability of buffers, costs of capital adjustment,
quality of institutions, and investment in adaptation
The Role of Domestic Policies and Institutions: strategies (Box3.2). They are also in line with the
Empirical Evidence empirical findings that show less damage from extreme
To study the extent to which macroeconomic and weather events and natural disasters in countries where
structural policies and country characteristics mute exchange rates are flexible, financial services are readily
the effect of weather shocks, the analysis extends the available, and institutions are strong.33,34
empirical approach described above. It does so by
allowing the response of per capita output to weather 33See Kahn (2005); Noy (2009); McDermott, Barry, and Tol
shocks to vary with various proxies for these policy and (2013); Burgess and others (2014); and Felbermayr and Grschl
institutional settings, which are included one at a time (2014) for the role of financial development, and Von Peter, Dahlen,
and Saxena (2012); Breckner and others (2016); and Lee, Villaruel,
in the analysis.32 It is important to emphasize that,
and Gaspar (2016) for the role of insurance penetration. Kahn
whereas fluctuations in temperature and precipitation (2005), Noy (2009), and Felbermayr and Grschl (2014) find evi-
are truly exogenous, which allows their causal impact dence for the role of institutions, and Ramcharan (2009) examines
to be identified, variations in policies and institutions the role of exchange rates in reducing damage from extreme weather
events and natural disasters.
across countries and over time are not. Accordingly, 34Two studies make a compelling case for the importance of
estimated correlations should be interpreted as being sectoral reallocation in alleviating output losses from climate change.
merely suggestive of causal impact. When quantifying the effects of climate change on agricultural
markets using micro data from 1.7million fields around the world,
Costinot, Donaldson, and Smith (2016) find that the welfare losses
32More specifically, the estimated specification augments equation would be three times larger if farmers were unable to switch produc-
(3.2) to include an interaction term between the weather shock and tion in response to changing climatic conditions and comparative
the policy variable. For simplicity, the sample is restricted to coun- advantage. In an empirical study, Colmer (2016) establishes that
tries with average temperature exceeding 15C, in which an increase labor movements from agriculture into manufacturing in India can
in temperature has a statistically significant linear negative impact on significantly offset the aggregate economic losses associated with
economic activity. See Annex3.3 for further details. weather-driven changes in agricultural productivity.
Figure 3.13. Role of Policy Buffers Figure 3.14. Role of Structural Policies and Institutions
(Percent; years on x-axis) (Percent; years on x-axis)
There is some suggestive evidence that the contemporaneous effect of There is some suggestive evidence that the medium-term effect of an increase in
temperature on per capita output is marginally lower in countries with lower public temperature on per capita output is marginally lower in countries with better-
debt, greater foreign aid inows, and exible exchange rates. regulated nancial markets, greater physical capital, more democratic institutions,
and lower income inequality.
0.5 1. Public Debt 2. Ination 0.5
0.5 1. Domestic Financial 2. International Finance 0.5
0.0 0.0 Sector Reform
Low debt to GDP Below double digit 0.0 0.0
High nancial sector Low restrictions on
0.5 High debt to GDP Above double digit 0.5 liberalization capital account
0.5 0.5
1.0 1.0
1.0 1.0
1.5 1.5
1.5 1.5
2.0 2.0 High
2.0 Low nancial sector restrictions on 2.0
2.5 2.5 liberalization capital account
1 0 1 2 3 4 5 6 7 1 0 1 2 3 4 5 6 7 2.5 2.5
1 0 1 2 3 4 5 6 7 1 0 1 2 3 4 5 6 7
0.5 3. International Reserves 4. Foreign Aid 0.5
0.5 3. Human Capital 4. Physical Capital 0.5
0.0 0.0
More than four High foreign aid 0.0 0.0
0.5 months of imports Low foreign aid 0.5 High human capital
0.5 Low human capital 0.5
1.0 1.0
1.0 1.0
1.5 1.5 High prevalence
Less than four 1.5 of paved roads 1.5
2.0 months of imports 2.0
2.0 Low prevalence
2.0
2.5 2.5 of paved roads
1 0 1 2 3 4 5 6 7 1 0 1 2 3 4 5 6 7 2.5 2.5
1 0 1 2 3 4 5 6 7 1 0 1 2 3 4 5 6 7
0.5 5. Remittances 6. Exchange Rate Flexibility 0.5
0.5 5. Political Regime 6. Inequality 0.5
0.0 0.0
High remittances 0.0 0.0
0.5 Low remittances 0.5 High polity score Low Gini
0.5 Low polity score High Gini 0.5
1.0 1.0
1.0 1.0
1.5 1.5
1.5 1.5
2.0 Not pegged 2.0
Pegged 2.0 2.0
2.5 2.5
1 0 1 2 3 4 5 6 7 1 0 1 2 3 4 5 6 7 2.5 2.5
1 0 1 2 3 4 5 6 7 1 0 1 2 3 4 5 6 7
Source: IMF staff calculations.
Note: The panels depict how the effect of a 1C increase in temperature on per Source: IMF staff calculations.
capita output in the sample of countries with average temperature exceeding 15C Note: The panels depict how the effect of a 1C increase in temperature on per
varies with the empirical proxy of a policy buffer. Horizon 0 is the year of the shock. capita output in the sample of countries with average temperature exceeding 15C
Gray areas indicate that the blue and red lines are signicantly different from each varies with the empirical proxies of structural policies and institutional settings.
other at the 15 percent level. See Annex 3.3 for the exact denition of policy Horizon 0 is the year of the shock. Gray areas indicate that the blue and red lines
variables. are signicantly different from each other at the 15 percent level. See Annex 3.3
for the exact denition of policy variables.
An alternative approach to assessing whether devel- Figure 3.15. Role of Development: Evidence from
opment more broadly reduces vulnerability to weather Subnational Data
(Percent; years on x-axis)
shocks takes advantage of subnational cross-country
data. It is difficult to establish definitively whether
The adverse effect of an increase in temperature on output is more pronounced in
advanced economies experience a smaller marginal non-advanced economies.
effect of heat on macroeconomic performance,
because so few of them have hot climates. However, Advanced economies Non-advanced economies
some of the larger advanced economies, such as the
United States, span several climate zones.35 This 1.5
within-country geographic heterogeneity makes it
possible to compare whether economic activity in 1.0
2.5
1 0 1 2 3 4 5 6 7
The Role of Migration
Migration is another possible adaptation strategy Source: IMF staff calculations.
for households hurt by weather shocks and persistent Note: The gure depicts how the effect of a 1C increase in temperature in the
changes in climateone with important cross-border sample of states or provinces with average temperature exceeding 15C varies
with an indicator of whether the state or province is located in an advanced
spillovers. Theoretically, the impact of weather shocks economy. Horizon 0 is the year of the shock. Gray area indicates that the blue and
on migration is ambiguous (see Dell, Jones, and Olken red lines are signicantly different from each other at the 15 percent level.
2014). Although lower incomes, safety concerns, and
physiological discomfort are powerful incentives to
relocate, the adverse income effect of weather shocks within country borders.38 Evidence of international
may undermine households ability to pay for transport migration responses is scarcer and typically focuses on
and other relocation expenses (Bryan, Chowdhury, and flows from individual countries.39
Mobarak 2014; Carleton and Hsiang 2016).37 Several The analysis builds on Cattaneo and Peri (2016)
empirical studies have documented adaptation to and examines whether weather shocks and natural
weather shocks and natural disasters through migration
38See Gray and Mueller (2012b) for evidence from Bangladesh;
35Average annual temperatures in the US states of Maine and and Boustan, Kahn, and Rhode (2012); Feng, Oppenheimer, and
Texas are about 7C and 21C, respectively. Schlenker (2012); Hornbeck (2012); and Hornbeck and Naidu
36Data constraints prevent the identification of the precise chan- (2014), among others, for evidence from the United States. Der-
nels through which development attenuates the link between weather yugina (2011), on the other hand, finds no population response in
and overall economic performance. Economic activity in hot areas in the 10 years following a hurricane landfall in the United States, but
advanced economies may be more insulated from temperature shocks documents a substantial increase in government transfer payments.
given that households exposed to these shocks have better access 39Munshi (2003), for example, finds that more migrants move
to ex post coping mechanisms (such as social protection) or have from Mexico to the United States when rainfall is lower in a given
reduced their vulnerability to shocks through ex ante adaptation Mexican communitya pattern also confirmed by Feng, Krueger,
strategies (such as activity diversification, adoption of air-condition- and Oppenheimer (2010). Country-specific evidence also includes
ing, and the like). Ethiopia (Gray and Mueller 2012a), Indonesia (Bohra-Mishra,
37Lack of knowledge and uncertainty about the risks caused by Oppenheimer, and Hsiang 2014), Pakistan (Mueller, Gray, and
slowly changing climate conditions (Lee and others 2015) as well as Kosec 2014), and Syria (Kelley and others 2015). Barrios, Bertinelli,
the provision of government assistance to disaster-prone areas may and Strobl (2006) and Marchiori, Maystadt, and Schumacher (2012)
also result in minimal behavioral change (Baez and others 2017). provide evidence from several countries in sub-Saharan Africa.
Figure 3.16. Effect of Temperature and Natural Disasters on low-income countries tend to be the most exposed
International Migration and vulnerable to climate change. These are also
(Percentage points of origin countrys total population) precisely the households with the fewest resources
available to finance relocation.
Among the sample of countries with average temperature exceeding 15C, an
increase in temperature and greater incidence of natural disasters induce Substantial migration flows, potentially spilling
migration, but only from non-low-income developing countries. across country borders, could arise if climate change
leads to a significant rise in sea levels. Hundreds of
18.0 1. Temperature 2. Number of Natural 1.2 millions of people in low-lying areas could become
Disasters
16.0 vulnerable to flooding, forcing them to abandon their
1.0 homes and relocate (Usery, Choi, and Finn 2007,
14.0
2009). In the United States alone, more than 4million
12.0 0.8 people living in coastal areas could be affected if oceans
rise the 80 centimeters the IPCC projects by 2100
10.0
0.6 under the unmitigated climate change scenario. If the
8.0 rise in sea levels is twice as much, the affected pop-
0.4 ulation would exceed 13million (Hauer, Evans, and
6.0
Mishra 2016).
4.0 0.2
2.0
0.0 International Support
0.0
Climate change is a global externality, and countries
2.0 0.2 will not be able to deal with its causes or its conse-
Emerging Low-income Emerging Low-income
market developing market developing quences on their own. Both equity and efficiency
economies countries economies countries
arguments call for active support from the interna-
tional community in helping low-income countries
Source: IMF staff calculations.
Note: Estimates from a panel regression of the effects of a 1C increase in 10-year plan, fund, and implement adaptation measures to
average temperature and number of natural disasters on the share of emigrants. cope with the consequences of climate change without
See Annex 3.4 for further details on the data, specication, and estimation.
Vertical lines denote 90 percent condence intervals.
compromising developmental objectives. On equity
grounds, low-income countries have contributed only
marginally to greenhouse gas emissions, yet they are
the most vulnerable to their harmful consequences,
disasters trigger emigration.40 The findings suggest
as this chapter demonstrates. On efficiency grounds,
that a rise in temperature and greater incidence of
requiring countries that have and/or are currently con-
weather-related disasters induce emigration, but only
tributing substantially to the atmospheric greenhouse
from countries where people can generally afford to
gas concentration to bear some of the adaptation costs
leave, which confirms Cattaneo and Peris (2016)
of low-income countries will help offset polluters fail-
results (Figure3.16; Annex Table3.4.1). Households
ure to fully internalize the cost of greenhouse gas emis-
in low-income developing countries, which tend
sions. And while the benefits of adaptation are largely
to have limited access to savings and credit, appear
domestic, successfully coping with weather shocks and
trapped by weather-induced income shocks (see Black
climate change could avert significant cross-border
and others 2011; Chen and others 2017). This inter-
spillovers, for example by stemming climate-induced
pretation is consistent with the findings of Hallegatte
population migration.
and others (2016) that the poorest households in
Support from the international community in the
form of concessional climate finance will be crucial to
40Focusing on the sample of countries with average annual
is an important step in that regard.41 In addition to public debt for a representative small open low-income
financial assistance, the transfer of appropriate adap- country. The model also highlights the role that struc-
tation and clean technologies to low-income countries tural transformation of low-income countries (that
can further enhance their efforts to cope with climate is, making the transition from agriculture to a more
change by improving access to state-of-the-art technol- services-based economy) could play in attenuating the
ogy, skills, and knowledge. Several initiatives under the impact of climate change. Box3.5 complements the
United Nations Framework Convention on Climate analysis by reviewing the evidence on the long-term
Change have promoted the international exchange effects of historical climate on economic performance.
of knowledge related to good practices in adaptation Simulations are based on the Debt, Investment, and
(such as the Adaptation Learning Mechanism), which Growth (DIG) model of Buffie and others (2012),
can be integrated into national and local plans. Multi- which captures aspects pertinent to low-income
lateral risk-sharing mechanisms, such as the Caribbean countriessuch as low public investment efficiency
Catastrophic Risk Insurance Facility and the African and high capital adjustment costsand can be
Risk Capacity, can also help countries with emergency extended easily to incorporate the structural transfor-
response in the immediate aftermath of a disaster, as mation process.43 These aspects of the DIG model
discussed in Boxes 3.3 and 3.4. make it preferable for studying the impact of climate
Cognizant of the challenges posed by climate change, change in low-income countries relative to the Inte-
the IMF, among other international financial institutions, grated Assessment Models (IAMs) more commonly
offers direct technical and financial support to small used to assess climate change effects.44
states and other countries that are vulnerable to weather In the DIG model, firms combine labor, private
conditions. To foster adaptation, it provides policy advice capital, and infrastructure to produce output. Consumers
and capacity building on how to enhance macroeco- supply labor and derive utility from consuming traded
nomic and risk management frameworks, determine and nontraded goods, while the government collects rev-
the appropriate balance between self-insurance and risk enue, redistributes income, and invests in infrastructure,
transfer, and strengthen investment and growth to build which it funds through domestic and external borrowing,
resilience.42 The IMF has also increased vulnerable coun- grants, and remittances. Based on the empirical results,
tries annual access limits under the Rapid Credit Facility changes in the exogenously-given sector-specific total
and Rapid Financing Instrument to provide rapid assis- factor productivity (TFP) levels are modeled as quadratic
tance to countries with urgent payment needs, including functions of temperature, while all other parameters are
as a result of natural disasters (IMF 2016b). calibrated broadly as in Buffie and others (2012).45
Figure 3.17. Long-Term Impact of Temperature Increase for The effects of climate change are examined through
a Representative Low-Income Developing Country: Model simulations of the macroeconomic response of output,
Simulations the public-debt-to-GDP ratio, and private investment
to the temperature increases projected under two
Model simulations suggest that the increase in temperature projected under the
intermediate and the unmitigated climate change scenarios could have signicant
of the scenarios prepared by the IPCC, as discussed
economic consequences for a representative low-income developing country, with in the Projections subsection of this chapter. The
sizable downside risks. simulations suggest that under both scenarios, the
representative low-income country will experience
Intermediate Scenario Unmitigated Climate Change sizable economic losses relative to a baseline of no
(RCP 4.5) Scenario (RCP 8.5)
changes in temperature, with significant downside risks
(Figure3.17).
Midpoint 95 percent condence interval
Under the milder scenario, the increase in tem-
6 1. GDP 2. GDP 6
perature will lower output by 4percent by 2100
4 (Percent deviation (Percent deviation 4 and depress private investment by 5percent as
2 from trend) from trend) 2 firms respond to lower productivity from rising
0 0
2 2
temperatures by cutting back capital spending. The
4 4 relative decline in output implies an increase in the
6 6 public-debt-to-GDP ratio of 2percentage points by
8 8
2100. Under the unmitigated climate change scenario,
10 10
12 12 the macroeconomic effect would be much larger.
14 14 Output would fall short by close to 9percent relative
16 16
2017 40 60 80 2100 2017 40 60 80 2100 to no climate change, private investment would fall by
11percent, and the public-debt-to-GDP ratio would
56 3. Public Debt 4. Public Debt 56 rise by 5percentage points by 2100.46
(Percent of GDP) (Percent of GDP) Conversely, the adverse effect would be significantly
54 54
52 52 smaller if the rise in temperature is successfully con-
50 50 tained to less than 2C, as stipulated in the 2015 Paris
48 48 Agreement, underscoring the critical importance of
46 46 mitigation efforts in limiting climate change damage.
44 44
Box3.6 discusses recent developments in climate miti-
42 42
gation efforts.
There is great uncertainty surrounding these central
40 40
2017 40 60 80 2100 2017 40 60 80 2100 projections because empirical estimates of the effect
of temperature shocks are imprecise and temperature
9 5. Investment 6. Investment 9 projections are uncertain. As a result, wide confidence
6 (Percent deviation from (Percent deviation from 6
trend investment rate) trend investment rate) intervals surround this chapters central projections.47
3 3
0 0 There is a 2.5percent chance of output declining more
3 3 than 8percent below the trend under the milder sce-
6 6 nario and more than 16percent under the unmitigated
9 9 climate change scenario. In line with lower output,
12 12
public debt would increase significantly relative to
15 15
18 18 output (about 10percent of GDP under the worst-case
21 21 scenario), and the private-investment-to-GDP
2017 40 60 80 2100 2017 40 60 80 2100
ratio could plummet by as much as 20percent strated in Box3.1 in the case of tropical cyclones, and
below the trend. could increase in frequency, potentially amplifying the
An alternative way to quantify climate change damage they cause. Certain expected or possible events
damage for a representative low-income country is to (such as rising sea levels) have no historic precedents
compute the present value of the shortfall in economic from which to draw inference but may have very sig-
output relative to the baseline of no climate change nificant economic consequences for many low-income
and to express this present value as a share of current countries, which are also not quantified in the simu-
output.48 Using a moderate growth-adjusted discount lations. Moreover, the long-term projections do not
rate of 1.4percent, the present value of output losses incorporate several of the channels through which
is large, at 48percent and 100percent of current temperature increases, and climate change in general,
output under the RCP 4.5 and RCP 8.5 scenarios, could affect economic activity, such as declining labor
respectively. supply from higher mortality and migration.
The above simulations assume a static economic Even abstracting from these difficulties, considerable
structure. However, as seen in the Channels of uncertainty exists about how to incorporate the empir-
Impact subsection, rising temperatures affect some ical estimates of economic losses into the dynamic
economic sectors more than others. For example, com- general equilibrium model. The analysis in this chapter
pared with agriculture, the services sector is relatively has taken a very conservative approach and assumes
sheltered from the adverse effects of higher tempera- that weather shocks have a permanent effect on the
ture. Hence, structural economic transformation from level of output. However, several studies have argued
a mostly agrarian to a more services-based economy that the empirical evidence is not inconsistent with
could lower the economic cost of climate change. The a persistent effect on the growth rate of output (Dell,
analysis extends the baseline DIG model to include an Jones, and Olken 2012; Burke, Hsiang, and Miguel
exogenous process of reallocating labor from agri- 2015a). Because even a small growth effect would
culture and manufacturing to services. The pace of ultimately dwarf a level effect, the adverse consequence
structural transformation is assumed to be moderate of temperature increases for the median low-income
and replicates past trends for low-income countries: country would be many times larger if rising tempera-
in the absence of shocks, the employment share of the tures were incorporated into the model as affecting the
services sector rises by 2.5percentage points a decade. growth path of output.49
Simulations in this extended model indicate that over
the long term, for the median low-income country,
structural transformation can reduce the cost of climate Summary and Policy Implications
change by about 25percent and 30percent under the Coping with climate change is one of the fun-
RCP 4.5 and RCP 8.5 scenarios, respectively. damental challenges of the 21st century, and this
The potential impact of climate change quantified challenge looms particularly large for low-income
in this section is subject to important caveats. First, developing economies. This chapter documents the
extrapolating from the short- to medium-term causal extraordinarily fast rise in temperature over the past
effects of weather shocks estimated from historical data century across advanced, emerging market, and
to the long-term impact of potential global warming low-income developing economies and the significant
may overstate the case if persistent changes in climate warming that could occur by the end of this century,
induce agents to adapt their economic activity to the depending on the international communitys ability
new environment. Conversely, permanent changes in
climate may have consequences that fluctuations in 49Burke, Hsiang, and Miguel (2015a) estimate much larger dam-
annual weather do not. Moreover, the model does not ages from climate change for hot countries: they model temperature
capture the effects of extreme weather events, which increases as having a persistent effect on the growth rate, rather than
the level of output. Permanent growth effects could arise if weather
inflict long-lasting macroeconomic damage, as demon-
shocks scar productivity growth through their effects on institutions,
innovation, or human capital accumulation. Several studies have
48In line with Nordhaus (2010), the real interest rate is assumed found evidence of effects of weather shocks on outcomes that could
to be 4.25percent, giving a growth-adjusted discount rate of plausibly shape productivity growth (for example, the link between
1.4percent. A more extreme discount rate of 0.1percent, proposed weather and conflict or weather and educational attainment), but
by Stern (2007), would increase the present value of damage by an it is difficult to establish empirically how long the growth damage
order of magnitude. through this channel lasts.
to contain greenhouse gas emissions. Low-income How can low-income countries cope with the
developing countries, which tend to be in some of rise in temperatures they are set to experience over
the hottest parts of the planet and are projected to the coming decades? Although causal interpretation
experience sizable increases in temperature, have con- is difficult, the chapter finds that the sensitivity of
tributed very little to the atmospheric concentration of per capita output to temperature shocks varies with
greenhouse gases. several mediating factors, and these factors are fun-
Yet the analysis suggests that rising temperatures damental to teasing out the chapters policy implica-
have highly uneven macroeconomic effects, with the tions. Sound domestic policies and institutions, and
adverse consequences borne disproportionately by development in general, could play a role in partially
countries with hot climates, such as most low-income reducing the adverse effects of weather shocks. Having
developing countries. The chapter finds that a rise policy buffers in place can help cushion some of the
in temperature lowers per capita output in countries negative effects of weather shocks by helping sustain
with high average temperatures, in both the short and public investment at adequate levels. Policies and
medium term, through a wide array of channels. In institutional settings that facilitate the reallocation
areas with hot climates, higher temperatures reduce of factors of production across economic sectors and
agricultural output, lower productivity of workers geographic regions and that foster developmentsuch
exposed to the heat, slow the rate of capital accumula- as better access to domestic and international finan-
tion, and damage health. These findings reflect impacts cial markets, high-quality infrastructure, and stronger
of weather shocks on average country outcomes. But institutionscan increase resilience to weather shocks
weather shocks could also have sizable unfavorable to some extent. These policies and institutional settings
distributional consequences within a country. Poor enable countries to recover faster from the negative
households tend to be more vulnerable to weather consequences of temperature increases and reduce their
fluctuations as a result of their heavy reliance on agri- exposure and vulnerability in the future. Investment in
cultural income, higher proportion of income devoted adaptation strategies and projectssuch as, for exam-
to food items, and limited access to savings and credit ple, well-targeted social safety nets that can promptly
(Hallegatte and others 2016; Hallegatte and Rozenberg deliver support where needed, climate-smart infrastruc-
2017; IMF 2016b). Despite the significant warming ture, and appropriate technologycould also reduce
that has occurred over the past century, the sensitivity some of the damage from climate change, as illustrated
of per capita output to temperature shocks has not by selected case studies.
changed materially, pointing to significant constraints But low-income countries have huge spending needs
to adaptation. and scarce resources to undertake the investments
The negative effects of projected climate change necessary to cope with climate change. According to
for low-income countries could be large. Focusing on United Nations estimates, attaining the Sustainable
one particular aspect of climate changenamely, the Development Goals would require low-income coun-
projected rise in temperatureand under the conser- tries to increase public spending by up to 30percent
vative assumption that temperature increases affect the of GDPan amount that likely exceeds the fiscal
level rather than the growth path of output, model space available in most countries (Baum and others
simulations suggest that, absent efforts to reduce global 2017; Schmidt-Traub 2015). Low-income countries
emissions, the output of a representative low-income also often lack the institutional setting, administrative
country could be 9percent lower than without an capacity, or political stability to implement appropri-
increase in temperature, with considerable downside ate macroeconomic policies or adaptation strategies
risks.50 The significant uncertainty about the mag- (Figure3.18). Moreover, domestic policies alone
nitude and effects of climate changenot only how cannot fully insulate low-income countries from the
much temperatures will rise, but also how the environ- consequences of climate change as higher temperatures
ment will reactcalls for careful consideration of these push the biophysical limits of these countries ecosys-
downside risks. tems, potentially triggering more frequent epidemics,
famines, and other natural disasters, at the same time
50Moreover, the negative welfare consequences of changing climate
fueling migration pressure and conflict risk. The
conditions will likely exceed output losses. Uncomfortably high
temperatures could spur investment as households adapt, but the international spillovers from these impacts of climate
increase in economic activity may not improve welfare. change in vulnerable countries could be very sizable.
Figure 3.18. Vulnerability to Temperature Increase and Given that low-income countries potential to
Adaptation Prospects address the climate change challenge by themselves
is limited, the international community must play a
Low-income developing countries, where the effect of temperature increase is
estimated to have the most pernicious effect, tend to have much lower climate
key role in providing and coordinating financial and
change adaptation capacity and readiness. nonfinancial support to these countries (see Box3.6).
Advanced and emerging market economies have con-
Low-income developing countries tributed the lions share to actual and projected climate
Emerging market economies change. Hence, helping low-income developing coun-
Advanced economies
tries cope with the consequences of climate change
2.0 1. Adaptation Capacity Index (x-axis)
is both a humanitarian imperative and sound global
economic policy that helps offset countries failure to
temperature on per capita output
1.5
fully internalize the costs of greenhouse gas emissions.
Effect of a 1C increase in
1.0
While the analysis in this chapter focused on the
0.5
impact of global warming in low-income countries, it
0.0
is important to note that all countries will increasingly
0.5
feel direct negative effects from unmitigated climate
1.0
change, through more frequent (and more damaging)
1.5
natural disasters (see Box3.1), rising sea levels, loss
2.0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 of biodiversity, and many other difficult-to-quantify
consequences. Warming will also begin to weigh on
2.0 2. Adaptation Readiness Index (x-axis) growth in many advanced economies as their tempera-
tures rise above optimal levels (see Annex Figure3.6.1).
temperature on per capita output
1.5
Effect of a 1C increase in
1.0
And even in countries where the effect might be
0.5
moderate or positive on average, climate change will
create winners and losers at both the individual and
0.0
sectoral levels. Moreover, the international spillovers
0.5
from the most vulnerable countries, through depressed
1.0
economic activity and potentially higher conflict and
1.5
migration flows, could be considerable. Going forward,
2.0
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 only a global effort to contain carbon emissions to lev-
els consistent with an acceptable increase in tempera-
Sources: Notre Dame Global Adaptation Index; and IMF staff calculations. ture can limit the long-term risks of climate change
Note: The gure depicts the estimated effect of a 1C increase in temperature on (Farid and others 2016; Hallegatte and others 2016;
per capita output at horizon 0 against countries score for adaptation readiness
and adaptation capacity. A higher score indicates better adaptation capacity and IMF 2015; Stern 2015; IPCC 2014).
more readiness.
edo (2014) use data from the EM-DAT to estimate the effects of 4The International Best Track Archive for Climate Steward-
different types of natural disasters (including storms) on growth, ship contains data on 7,140 tropical cyclones, with information
while a parallel body of literature (Strobl 2012; Bertinelli and on maximum sustained wind speed between 1950 and 2016
Strobl 2013; Hsiang and Jina 2014) uses wind-field models to (Knapp, Applequist, and others 2010; Knapp, Kruk, and others
estimate the effects of storm winds on growth. Bakkensen and 2010). These data are combined with the CIESIN (2016) settle-
Barrage (2016) use maximum wind speed at landfall, which is ments population in 2000, which contains data for 67,682 cities
closer to the approach used here. that range in population from one person to 18.5million people.
wind variable weighted by the share of population and islands that are generally more exposed to this type of
time exposed. The specification estimated is as follows: storm (Figure3.1.1).5 By income group, advanced econ-
omies are the hardest hit by tropical cyclones because
yi,t+h yi,t1 = h1( Wind i,tPi,t
TEi,t)
they tend to be exposed to higher wind speeds.
+ h2( Wind i,t1Pi,t1
TEi,t1) The estimates are not only statistically, but also
economically, significant. Seven years after an average
+ j=1 3 ( TEi,t+hj)
h1 h Wind
i,t+hjPi,t+hj
storm strikes, per capita output is almost 1percent
+ h1 ci,t + h2ci,t1 + j=1
h1 h c lower than if the storm had not happened, with
3 i,t+hj
2.5 times larger losses experienced by small states
+ h1yi,t1+ hi + hr,t + hi,t, (3.1.1) (Figure3.1.2).6 The effects of storms are very per-
sistent: even after 20 years, the economy has not fully
in which h indexes the estimation horizon, hi are recovered from the shock.7 Notably, the effect of
country fixed effects, hr,t are region-year fixed effects, tropical cyclones on economic activity is separate and
yi,t is the log of GDP per capita, and ci,t refers to in addition to the effects of temperature (Table3.1.2).
average annual temperature and precipitation and their Introducing the wind variable does not materially
squared terms. change the coefficients on temperature and precipita-
The results presented in Table3.1.2 indicate that tion for the same sample of countries.
if the wind speed increased by one knot throughout
the entire country (that is, the entire population is Climate Change and Tropical Cyclones
exposed), and for an entire year, real GDP per capita Climate scientists predict that, with climate change,
would decline by 26.7percent the year the storm there will be fewer tropical cyclones that form, but the
strikes. This, of course, is not a very useful indicator
of the effect of a typical storm on a country; a better
measure is the marginal effect of increasing wind speed 5For a discussion of small states vulnerability to natural disas-
as captured by Pi,t
TEi,t. ters and climate change, see IMF (2016b).
6A storm strike includes any tropical cyclone that passed
Box 3.2. The Role of Policies in Coping with Weather Shocks: A Model-Based Analysis
To illustrate how policies can help moderate the Figure 3.2.1. Role of Policies: A Model-
consequences of weather shocks in low-income coun- Based Analysis
tries, this box uses the Debt, Investment, and Growth (Real GDP, deviation from steady state;
(DIG) model developed by Buffie and others (2012) years on x-axis)
and simulates the macroeconomic effects of tempera-
ture increases under various assumptions for key policy 1 1. Baseline 2. Role of Fiscal 1
Space
variables.1 As demonstrated empirically in the chapter,
in hot countries, an increase in temperature reduces 0 0
productivity. Moreover, a temperature increase could
precipitate the loss of productive land. Consequently, 1 1
the analysis calibrates the weather damage to total fac-
tor productivity and private capital to broadly match Baseline
2 2
the estimated response of GDP to a 1C increase in Low grants
temperature in a representative low-income country High grants
with a baseline temperature of 25C and examines 3 3
1 0 1 2 3 4 5 6 7 1 0 1 2 3 4 5 6 7
how this damage can be shaped by macroeconomic
and structural policies (Figure3.2.1).2
1 3. Role of Public 4. Role of Capital 1
Sector Efciency Adjustment
Policy Space and the Role of Institutions Costs
0 0
Weather shocks can weigh significantly on the
Baseline
public purse of low-income countries. Government
revenues can be adversely affected by the reduction in 1 1
agricultural and industry output at the same time that
spending may need to be ramped up to deliver support Perfect efciency Baseline
2 2
Typical efciency High cost
to affected households if weather shocks compromise Medium cost
Low efciency
food security, to rebuild transport or communication
3 3
infrastructure if they are damaged by natural disasters, 1 0 1 2 3 4 5 6 7 1 0 1 2 3 4 5 6 7
and potentially to retrain the workforce. Because fiscal
space is often tight in many low-income countries, 1 5. Role of 6. Role of Fiscal 1
expanding transfers from advanced economiesfor Hysteresis Incentives for
Adaptation
instance, through the transfers agreed to under the 0 0
Paris Agreementcould strengthen countries ability Baseline
to reduce the impact of weather shocks. Model simula-
1 1
tions suggest that receiving additional transfers used to
build up public investment for three years, starting a
2 Baseline 2
Medium hysteresis
With adaptation
The authors of this box are Manoj Atolia, Claudio Baccianti, High hysteresis
Ricardo Marto, and Mico Mrkaic. 3 3
1The DIG model is a real, neoclassical, dynamic open econ- 1 0 1 2 3 4 5 6 7 1 0 1 2 3 4 5 6 7
omy framework with two production sectors that use public and
private capital as input and many features that are pertinent to Source: IMF staff calculations.
low-income countries, such as low public investment efficiency, Note: The baseline assumes no additional grants in panels 2
limited fiscal space, and capital adjustment costs. The model is and 3, low adjustment cost in panel 4, no hysteresis in panel
also used to simulate the long-term effects of climate change in 5, and no adaptation in panel 6. In panel 2, additional grants
the section of the chapter titled Long-Term Effects of Tempera- amount to 0.5 percent of GDP in low grants scenario, and 1
ture IncreaseA Model-Based Approach. percent of GDP in high grants scenario. In panel 3, all
2For simplicity, the traded and nontraded sectors are assumed simulations, except the baseline, assume high additional
grants.
to react equally to weather shocks. The findings are robust to this
modeling choice. Most other parameters are calibrated as in Buf-
fie and others (2012), except the real interest rate on public debt,
which is lower than in the original paper because of the decline
in global interest rates. See Annex3.5 for further details.
Box 3.3. Strategies for Coping with Weather Shocks and Climate Change: Selected Case Studies
Adverse effects of weather shocks and climate change food security of vulnerable householdsbeneficiaries
have motivated local communities and countries to of the PNSP experienced a 25percent smaller drop in
adapt and counter these unfavorable consequences. As consumption relative to those that were not covered
demonstrated in Figure3.12, a wide range of strategies by the program in the aftermath of droughts (Porter
could dampen the negative impacts of weather shocks and White 2016). The PSNP has also reduced the
and natural disasters by reducing exposure and vulner- number of people in need of humanitarian interven-
ability or by transferring and sharing weather-related tion and the cost of such intervention. Finally, the
risks. The purpose of this box is to showcase some PSNP has increased savings of vulnerable households
examples of successful coping strategies. and has facilitated improved access to educational and
health services.
Social Safety Nets
Approximately 85percent of the Ethiopian pop- Technology Adoption
ulation is employed in agriculture, mostly on small High temperatures significantly lower labor produc-
family-owned farms. Climate change and associated tivity and could lead to adverse health outcomessuch
droughts, delayed rains, and flooding weigh on agri- as increased incidence of hyperthermia and worsening
cultural productivity and food security. Furthermore, chronic cardiovascular or respiratory diseasesand
in some areas, the land has become degraded due to mortality, as demonstrated in a large body of work and
overuse. Consequently, approximately 10percent of the analysis in this chapter. Governments and individ-
the rural population is chronically food insecure. uals have various options for reducing these adverse
To assist the at-risk population, the Ethiopian economic and health impacts, such as green infrastruc-
government and international partners instituted the ture (to increase the presence of vegetation in cities)
Productive Safety Net Program (PSNP) in 2006. The and specific construction technologies (for example,
PSNP provides cash or food to households unable to roofs that are highly solar reflective). Among all options,
feed themselves all year, particularly in the lean season modern air-conditioning, invented at the turn of the
(JuneAugust). The aid is contingent on active partici- 20th century, is the most common solution adopted by
pation in local productivity-enhancing or environmental households and firms to deal with excessive heat.
programsfor example, land rehabilitation, improve- The benefits of climate control, both in the work-
ment of water sources, and construction of infrastruc- place and for health outcomes, are well documented.
ture such as roads and hospitals. A complementary In a 1957 survey, 90percent of American firms named
program, the Household Asset Building Program, which cooled air as the single biggest boost to their produc-
targets the same households as the PSNP, helps house- tivity (Cooper 2002), and Singapores founding father,
holds diversify their income sources and increase pro- Lee Kuan Yew, credited air-conditioning as the most
ductive assets, including by offering technical assistance, important factor in his countrys development success.
with the goal of achieving lasting food security. The dramatic decline in heat-related mortality over
With more than 7.6million participants (or almost the 20th century in the United States has also been
8percent of the Ethiopian population) and 47,000 attributed to the adoption of residential air condition-
small community projects every year, the PSNP is the ing (Barreca and others 2016).
largest climate change adaptation program in Africa. Nevertheless, the negative effects of air-conditioning
The community projects, which are mostly devoted cannot be ignored. Increased adoption of indoor cli-
to environmental restoration, are offering measurably mate control increases energy consumption and green-
positive results. The PSNP has reduced soil loss by more house gas emissions. Exhaust from air-conditioning
than 40percent and improved the quality and quantity machines and facilities can give rise to local pockets of
of available water. Studies suggest that land productiv- hot air, which can present significant negative exter-
ity has consequently increased by up to 400percent. nalities for nearby populations. High up-front costs
In addition, the PSNP has reduced the damage from and infrastructure requirements make this technology
seasonal flooding. The program has also improved the out of reach for poor and vulnerable populations,
especially in low-income developing countries.1
The authors of this box are Claudio Baccianti and 1As of 2012, slightly more than one-third of households had
level is closed and the tunnel is used as an ordinary 1.3million tons of crops.
Box 3.4. Coping with Weather Shocks: The Role of Financial Markets
Financial markets can reduce the adverse conse- Figure 3.4.1. Insurance Penetration:
quences of weather shocks by reallocating the costs Non-Life Insurance Premium
and risks of such shocks to those most willing and (Percent of GDP)
able to bear them. Insurance products, such as weather
derivatives, can help households and firms vulnerable 6
North America Japan
to short-term fluctuations in temperature and pre- Western Europe Africa
cipitation hedge their idiosyncratic weather exposure.
5
Catastrophe (Cat) bonds can help disperse catastrophic
weather risk to capital markets. However, the degree
to which financial markets can mitigate the impacts 4
Latin America and the Caribbean
of weather shocks hinges on the level of insurance Eastern Europe
penetration and on the capacity to correctly price South and East Asia
3
weather-related risks. This box reviews recent develop-
ments in the market for weather-related financial prod-
ucts and provides new evidence on the extent to which 2
stock markets efficiently price weather-related risks.
Insurance 1
ucts, parametric insurance can leave a fair amount of residual times, but absorb losses if a predefined catastrophe occurs. They
risk uncovered (basis risk), given that the actual loss may differ were first introduced in the mid-1990s, in the aftermath of
from the payment received by contract holders. Hurricane Andrew.
15
0
10
5 1
0
1997 2000 03 06 09 12 15 17
2
2.0 1.5 1.0 0.5 0.0 0.5 1.0 1.5 2.0
Source: Artemis Insurance-Linked Securities and Catastrophe Temperature, annual deviation from country average
Bond Market Report (www.artemis.bm).
Note: Years ending June 30.
Sources: Datastream; Peng and Feng (forthcoming); and IMF
staff calculations.
Note: One-year-ahead food and beverages sector returns are
the global financial crisis, as well as new regulations regressed on annual average temperature (deviation from the
country average, degrees Celsius). Sample is restricted to
that recognize the relief of capital through Cat bond countries with an average annual temperature above 15C.
issuance, have potentially contributed to the growth
of the Cat bond market. Cat bonds have become an
increasingly popular tool for private insurance and
reinsurance companies in Europe, Japan, and the of extreme weather conditions on member Afri-
United States to transfer away their risk exposures to can countries.
earthquakes, storms, and hurricanes.
As discussed in the chapter, low-income developing Do Financial Markets Correctly Price
countries and small states are especially vulnerable to Weather-Related Risks?
catastrophic risks. Mexico, in 2006, was the first coun- The optimal level of insurance against abnormal
try to issue Cat bonds; since then, several low-income weather conditions requires accurate assessments of
developing countries have issued Cat bonds covering weather-related risk. There is growing evidence that
hurricanes, earthquakes, and other extreme events. The investors in financial markets do not fully understand,
World Bank issued its first-ever Cat bond in 2014 to at least immediately, the impact of weather shocks on
provide reinsurance to the Caribbean Catastrophe Risk output and productivity. Hong, Li, and Xu (2016)
Insurance Facility, a risk-pooling facility designed to show that the stock indices of the food industry
limit the financial impact on 16 Caribbean country in the United States and in a few other advanced
governments after possible earthquakes and hurri- economies respond to changes in drought indices only
canes (see also Box3.3). A similar arrangementthe with a delay. This finding suggests that markets do
Extreme Climate Facilityis being developed by not incorporate weather information into prices until
the African Risk Capacity (see Box 3.3) to allow for several months later, perhaps after the losses incurred
the issuance of Cat bonds to alleviate the impact are reflected in food companies annual reports. The
Box 3.5. Historical Climate, Economic Development, and World Income Distribution
As argued in the chapter, climate change may have countrys temperature over different periods from 1730
very long-lived effects on economic performance, to 2000 and its modern income per capita, uncovering
although the exact magnitudes depend on many some striking patterns. A simple bivariate regression
factors, including economic agents adaptability and confirms the strong negative correlation between
the ability of the economy to structurally adjust. income in 2000 and the average temperature during
Empirically, it is very difficult to disentangle whether 197099 (Table3.5.1, regression 1). However, after
weather shocks have permanent level or growth effects controlling for historical average temperature in the
on output based on recent data (since 1950); if they 18th and 19th centuries, a time-varying and non-
reflect permanent growth rather than level effects, then monotonic effect of temperature on current country
the consequences may be many times larger than the incomes is revealed, with 18th century temperature
initial effects, but this impact would manifest only exhibiting a positive and large effect while 19th cen-
over a very long time. tury temperature shows an even larger negative effect
This box reviews a relatively new and growing liter- (Table3.5.1, regression 2). Interestingly, once histor-
ature that attempts to directly assess whether historical ical climate is introduced, 20th century temperature
climate can have a large and permanent effect on no longer shows a strong, negative association with
economic performance. Enabled by the rising avail- current income, suggesting that it may be serving as
ability and granularity of historical data, the literature a proxy for the combined effects of historical climate,
examines the relationship between modern outcomes rather than capturing a direct impact of the current
and historical climate, starting from the hypothesis temperature level in the simple regression.
that historical events (potentially in the very deep What might account for the estimated nonmono-
past) interact with the physical environment and can tonic relationship between temperature and income?
have permanent effects on economic development and Bluedorn, Valentinyi, and Vlassopoulos (2009)
performance.1 postulate that it could reflect interactions between
Leveraging the exogeneity of historical climate, temperature and historical events across centuries.
Bluedorn, Valentinyi, and Vlassopoulos (2009) For example, the large negative effect of 19th century
estimate the reduced-form relationship between a temperature on current incomes could be linked to
a slower diffusion of technologies from the United
Kingdom and Europe, which were at the technological
The author of this box is John C. Bluedorn.
1Nunn (2014) provides an excellent exposition of the idea, frontier then, and generally at the cooler end of the
which is central to recent empirical research on historical global temperature distribution. If the technologies
development. these countries developed were more suitable for
On equity grounds, there is some appeal in linking linearly over time to meet countries Paris Agreement mitigation
climate finance donations from advanced economies to pledges. Carbon charges for international aviation and maritime
fuels are another promising source of climate financea $30 a
their contribution to climate change. If the Group of ton CO2 charge on these fuels could raise revenues of $25billion
Twenty economies, excluding the five members with in 2020, even with full compensation for developing economies
lowest per capita income, donated $5 for each ton (Farid and others 2016).
Annex3.1. Data Sources and Country Groupings For real GDP per capita, investment, and imports,
Data Sources the sources are listed in the order in which they are
spliced (which entails extending the level of a primary
The primary data sources for this chapter are the series using the growth rate of a secondary series).
IMF World Economic Outlook database and the World
Bank World Development Indicators database. The
main data sources on temperature and precipitation are Data Definitions
the University of East Anglias Climate Research Unit The main historical temperature and precipitation
(historical data, 19012015) and National Aeronautics series used in the chapters analysis are constructed by
and Space Administration (NASA) Earth Exchange aggregating grid cell data at 0.5 0.5 degree resolu-
Global Daily Downscaled Projections data set (forecast, tion (approximately 56 kilometers 56 kilometers
present2100). All data sources used in the chapters at the equator) to the level of individual countries or
analysis are listed in Annex Table3.1.1. subnational regions at annual or monthly frequency.
The estimates are weighted by grid-level population range of 1.1C2.6C, with a greater than 50percent
(exploring three alternatives: population distribu- chance of an increase exceeding 2C by 2100). In the
tion as of 1950, 1990, and 2010) to account for RCP 8.5 scenario, CO2 emissions continue to grow
differences in population density (Dell, Jones, and unconstrained, and the average 20812100 tempera-
Olken 2014). ture is expected to be 3.7C higher (in a range of
Temperature and precipitation projections are 2.6C4.8C) relative to 19862005. The chapter uses
from two of the four scenarios, called Representative the average of the maximum and minimum daily tem-
Concentration Pathways (RCP), constructed by the perature and total daily precipitation data from 2005
Intergovernmental Panel on Climate Change. The and projections for 2050 and 2100 at the 0.25 x 0.25
RCP 4.5 scenario assumes increased attention to the degree resolution, averaged across the 21 models of the
environment with slow growth of carbon dioxide Coupled Model Intercomparison Project Phase 5 for
(CO2) emissions until 2050 and a decline of emissions each scenario. Annual temperatures are computed as
thereafter, resulting in a mean temperature increase the average of the daily temperature; annual precipita-
of 1.8C by 20812100 relative to 19862005 (in a tion is the sum of daily precipitation.
Country Groupings
Annex Table3.1.2. Country and Territory Groups
Advanced Australia, Austria, Belgium, Canada, Cyprus, Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Hong Kong SAR,*
Economies Iceland, Ireland, Israel, Italy, Japan, Korea, Latvia, Lithuania, Luxembourg, Macao SAR,* Malta, Netherlands, New Zealand, Norway,
Portugal, Puerto Rico, San Marino,* Singapore, Slovak Republic, Slovenia, Spain, Sweden, Switzerland, Taiwan Province of China,*
United Kingdom, United States
Emerging Market Albania, Algeria, Angola, Antigua and Barbuda, Argentina, Armenia, Azerbaijan, The Bahamas,* Bahrain, Barbados, Belarus, Belize, Bosnia
Economies and Herzegovina, Botswana, Brazil, Brunei Darussalam, Bulgaria, Cabo Verde, Chile, China, Colombia, Costa Rica, Croatia, Dominica,
Dominican Republic, Ecuador, Egypt, El Salvador, Equatorial Guinea, Fiji, Gabon, Georgia, Grenada, Guatemala, Guyana, Hungary, India,
Indonesia, Iran, Iraq, Jamaica, Jordan, Kazakhstan, Kosovo,* Kuwait, Lebanon, Libya, Macedonia FYR, Malaysia, Maldives,* Marshall
Islands,* Mauritius, Mexico, Micronesia,* Montenegro, Morocco, Namibia, Nauru,* Oman, Pakistan, Palau,* Panama, Paraguay, Peru,
Philippines, Poland, Qatar, Romania, Russia, Samoa, Saudi Arabia, Serbia, Seychelles,* South Africa, Sri Lanka, St. Kitts and Nevis, St.
Lucia, St. Vincent and the Grenadines, Suriname, Swaziland, Syria, Thailand, Timor-Leste, Tonga, Trinidad and Tobago, Tunisia, Turkey,
Turkmenistan, Tuvalu,* Ukraine, United Arab Emirates, Uruguay, Vanuatu, Venezuela
Low-Income Afghanistan, Bangladesh, Benin, Bhutan, Bolivia, Burkina Faso, Burundi, Cambodia, Cameroon, Central African Republic, Chad, Comoros,
Developing Democratic Republic of the Congo, Republic of Congo, Cte dIvoire, Djibouti, Eritrea, Ethiopia, The Gambia, Ghana, Guinea, Guinea-Bissau,
Countries Haiti, Honduras, Kenya, Kiribati,* Kyrgyz Republic, Lao P.D.R., Lesotho, Liberia, Madagascar, Malawi, Mali, Mauritania, Moldova, Mongolia,
Mozambique, Myanmar, Nepal, Nicaragua, Niger, Nigeria, Papua New Guinea, Rwanda, Senegal, Sierra Leone, Solomon Islands, Somalia,*
South Sudan, Sudan, So Tom and Prncipe, Tajikistan, Tanzania, Togo, Uganda, Uzbekistan, Vietnam, Yemen, Zambia, Zimbabwe
Countries and Algeria, American Samoa, Angola, Anguilla, Antigua and Barbuda, Argentina, Australia, Bahrain, Bangladesh, Barbados, Belize, Benin, Bhutan,
Territories with Botswana, Brazil, Brunei Darussalam, Burkina Faso, Burundi, Cabo Verde, Cambodia, Cameroon, Central African Republic, Chad, Colombia,
Average Annual Comoros, Democratic Republic of the Congo, Republic of Congo, Costa Rica, Cuba, Curaao,* Cyprus, Cte dIvoire, Djibouti, Dominica,
Temperature Dominican Republic, Ecuador, Egypt, El Salvador, Equatorial Guinea, Eritrea, Ethiopia, Fiji, Gabon, The Gambia, Ghana, Grenada, Guadeloupe,*
above 15C Guatemala, French Guiana,* Guinea, Guinea-Bissau, Guyana, Haiti, Honduras, India, Indonesia, Iraq, Israel, Jamaica, Jordan, Kenya, Kuwait,
Lao P.D.R., Lebanon, Liberia, Libya, Madagascar, Malawi, Malaysia, Mali, Malta, Martinique,* Mauritania, Mauritius, Mexico, Montserrat,
Morocco, Mozambique, Myanmar, Namibia, Nepal, New Caledonia, Nicaragua, Niger, Nigeria, Oman, Pakistan, Panama, Papua New Guinea,
Paraguay, Philippines, Puerto Rico, Qatar, Reunion,* Rwanda, Samoa, Saudi Arabia, Senegal, Sierra Leone, Singapore, Solomon Islands,
Somalia, South Africa, South Sudan, Sri Lanka, St. Kitts and Nevis, St. Lucia, St. Vincent and the Grenadines, Sudan, Suriname, Swaziland,
Syria, So Tom and Prncipe, Tanzania, Thailand, Timor-Leste, Togo, Tonga, Trinidad and Tobago, Tunisia, Turkmenistan, Turks and Caicos,*
Uganda, United Arab Emirates, Uruguay, Vanuatu, Venezuela, Vietnam, Virgin Islands (US), West Bank and Gaza, Yemen, Zambia, Zimbabwe
Countries with Albania, Argentina, Australia, Austria, Bangladesh, Belgium, Benin, Bolivia, Bosnia and Herzegovina, Brazil, Bulgaria, Canada, Chile,
Province-Level China, Colombia, Croatia, Czech Republic, Denmark, Ecuador, Egypt, El Salvador, Estonia, Finland, France, Germany, Greece, Guatemala,
Data Honduras, Hungary, India, Indonesia, Iran, Ireland, Italy, Japan, Jordan, Kazakhstan, Kenya, Korea, Kyrgyz Republic, Latvia, Lesotho,
Lithuania, Macedonia FYR, Malaysia, Mexico, Mongolia, Morocco, Mozambique, Nepal, Netherlands, Nicaragua, Nigeria, Norway, Pakistan,
Panama, Paraguay, Peru, Philippines, Poland, Portugal, Romania, Russia, Serbia, Slovak Republic, Slovenia, South Africa, Spain, Sri
Lanka, Sweden, Switzerland, Tanzania, Thailand, Turkey, Ukraine, United Arab Emirates, United Kingdom, United States, Uruguay,
Uzbekistan, Venezuela, Vietnam
Countries with Argentina, Bolivia, Botswana, Brazil, Chile, China, Colombia, Costa Rica, Denmark, Egypt, Ethiopia, France, Germany, Ghana, Hong Kong SAR,*
Sectoral-Level India, Indonesia, Italy, Japan, Kenya, Korea, Malawi, Malaysia, Mauritius, Mexico, Morocco, Netherlands, Nigeria, Peru, Philippines, Senegal,
Data Singapore, South Africa, Spain, Sweden, Taiwan Province of China,* Tanzania, Thailand, United Kingdom, United States, Venezuela, Zambia
Source: IMF staff compilation.
* Not included in the main regression analysis.
Annex3.2. Weather Shocks and very strong impact on the occurrence of disasters.
Natural Disasters More precipitation reduces the occurrence of disas-
Although there is a clear link between weather con- ters caused by droughts, wildfires, and heat waves,
ditions and the occurrence of extreme weather events, but increases the probability of disasters triggered by
the relationship between weather shocks and natural floods, landslides, cold waves, tropical cyclones, and
disastersextreme events associated with significant other storms. The effects of temperature are also as
economic damage and loss of lifehas not been stud- expected, with higher temperatures resulting in more
ied in detail. The analysis in this section examines how disasters caused by droughts, wildfires, heat waves,
weather conditions influence the frequency of various tropical cyclones, and other storms, but reducing the
types of weather-related natural disasters. probability of cold waves. The results also show that
A logit panel specification with country fixed effects precipitation has nonlinear effects on the probability
is used to estimate the effect of the weather variables of most disasters.
ci,t (temperature and precipitation) on the probabil- Interestingly, the estimations suggest that the
ity of a natural disaster taking place in country i in a weather conditions over the preceding 12 months
given month t. have a significant effect on the occurrence of most
types of disasters. Weather anomalies during the
Pr(disaster i,t = 1) = ( 1ci,t + 2ci,t2 + 1Devi,tT previous year, as captured in the cumulative deviation
of temperature and precipitation from its monthly
+ 2 Devi,tP + 3 Devi,tOcean
+ 1 ln(GDP) i,t12
10-year average, are important determinants of all
+ 2ln(Pop) i,t12+ i + i,t), (3.1) types of disasters, except those caused by landslides
or tropical cyclones, which are entirely a function of
in which the nonlinear function () = exp()/ short-term weather patterns. Epidemics, however, are
(1+exp()) captures the effect of the regressors on the not affected by short-term weather conditions, but
probability of a natural disaster. Country fixed effects respond to temperature deviations in the year before
(i) capture time-invariant country characteristics, the event is triggered.
such as the size and geographic location of the country To quantify the likely impact of climate change, the
and its topology, that may influence the exposure and analysis combines the estimation results and projected
vulnerability of countries to different types of disas- temperature and precipitation in 2050 and 2100 under
ters.51 The specification controls for the level of real Representative Concentration Pathway 8.5 to predict
GDP per capita and population, as well as for global the likelihood of each type of natural disaster. These
weather conditionsspecifically the deviation in predicted probabilities in 2050 and 2100 are compared
global ocean surface temperature from the 19012000 with the predicted incidence of natural disasters over
averagethat might affect the incidence of disasters. 201014 in Figure3.6.
The sample includes monthly data during 19902014
for 228 countries and territories on more than 8,000
weather-related disasters. Equation (3.1) is estimated Annex3.3. Empirical Analysis of the
separately for each type of natural disaster, improv- Macroeconomic Effects of Weather Shocks and
ing on Thomas and Lopez (2015), who perform a the Role of Policies
similar exercise on annual data, but group together This annex provides further details on the empiri-
all disasters. cal model used to quantify short- and medium-term
Annex Table3.2.1 presents the estimation results effects of weather on economic activity to identify
for each disaster type. Weather conditions have a the channels through which these effects occur,
investigate evidence or lack thereof of adapta-
tion over time, and study the role of various
51Given the large time dimension of the sample (each country
policy measures in attenuating the effects of tem-
has about 300 observations), a panel logit specification is preferred
to conditional logit models because it allows for the estimation of perature shocks.
predicted and marginal effects accounting for country fixed effects. The baseline analysis uses Jords (2005) local
The results are robust to the use of conditional logit regression projection method to trace out the impulse response
models developed by Chamberlain (1980) to avoid the incidental
parameters problem that may arise from estimating fixed effects with functions of various outcomes to weather shocks based
a small time sample. on the following equation:
Number of Observations 29,976 35,772 43,632 19,620 18,732 17,844 12,924 20,652 33,684
Number of Countries 101 120 147 66 63 61 44 69 114
Source: IMF staff calculations.
Note: The dependent variable is an indicator that takes the value of 1 if a natural disaster of a particular type is taking place. All specifications control for country fixed effects. Standard errors are clustered at the country level.
* p < 0.1; ** p < 0.05; *** p < 0.01.
CHAPTER 3 The Effects of Weather Shocks on Economic Activity: How Can Low-Income Countries Cope?
Annex Figure 3.3.1. Effect of Temperature Increase on Real per Capita Output across the Globe, with Countries Rescaled
in Proportion to Their Projected Population as of 2100
(Percent)
0.80 to 1.63
0.43 to 0.80
0.33 to 0.43
0.50 to 0.32
0.97 to 0.50
1.30 to 0.97
1.68 to 1.30
Insignicant effect
No data
Sources: Natural Earth; ScapeToad; United Nations World Population Prospects database: the 2015 revision; and IMF staff calculations.
Note: The map depicts the contemporaneous effect of a 1C increase in temperature on per capita output computed as per equation (3.3) using recent 10-year
average country-level temperature together with estimated coefcients in Annex Table 3.3.1, column (5). Each country is rescaled in proportion to the projected
population as of 2100. Using projected population as of 2100, 76 percent of world population will live in countries that experience a negative impact from 1C
increase. Gray areas indicate the estimated impact is not statistically signicant.
present results for the baseline specification with in the dynamic response. Column (8), however, tests the
region-year fixed effects instead of country-specific robustness of the findings to using the autodistributed lag
time trends. Column (6) limits the sample to countries model with seven lags of the weather variables and their
with at least 20 years of data. squared terms, as in Dell, Jones, and Olken (2012), who
Column (7) controls separately for the occurrence test different models from no lags up to 10 lags and find
of natural disasters given that temperature and pre- that, across different lag specifications, results are broadly
cipitation fluctuations might affect economic activ- consistent in magnitude and statistical significance.
ity through their effect on the incidence of natural Across all specifications, the estimated coefficient on
disasters, as discussed in Annex3.2. Controlling for temperature is positive, and the coefficient on tem-
natural disasters does not materially alter the estimated perature squared is negative, confirming the nonlinear
coefficients on temperature and precipitation.52 relationship between growth and temperature shocks
In columns (1)(7), impulse responses were estimated uncovered by Burke, Hsiang, and Miguel (2015a). At
using Jords (2005) local projection method. This low temperatures, an increase in temperature boosts
approach is advocated by Stock and Watson (2007), growth, whereas at high temperatures, an increase in
among others, as a flexible alternative that does not temperature hurts growth, with the threshold average
impose the dynamic restrictions embedded in vector annual temperature estimated to be about 13C15C.
autoregressions (autoregressive distributed lag) specifica- As an additional robustness check, column (9) pres-
tions and is particularly suited to estimating nonlinearities ents results of a linear regression without the squared
terms of the weather variables in which the sample is
52To further explore the robustness of these results, weather
limited to countries with average annual temperature
variables were transformed using natural logarithms or normalized by
above 15C. Indeed, within the sample of relatively hot
subtracting the country mean and dividing by the country standard
deviation. Availability of data on subnational per capita GDP and countries, the coefficient on temperature is negative
annual average temperature and precipitation allows us to estimate and statistically significant. The effect of temperature
the same regression at a subnational level using province fixed effects. increase across the globe is shown in Figure3.8 panel
Through all three specifications the main finding persists: there is a
nonlinear relationship between temperature and economic perfor- 1 at grid level; in panel 2, where countries are rescaled
mance (results available on request). in proportion to their 2015 population; and in Annex
Figure3.3.1, where countries are rescaled in proportion not affected by a precipitation shock today, which is
to projected 2100 population. different from the effect of temperature.
There is no consistently significant relationship
between precipitation and per capita GDP growth across Channels
the various specifications. The lack of robust relationship The chapter examines the potential channels through
could reflect potentially larger measurement error in which temperature shocks affect the macroeconomy
the precipitation variable, as discussed in Auffhammer in a broad and long-lasting manner by studying the
and others (2011), which could be further amplified by relationship between temperature and each of the main
temporal aggregation. For example, if the only channel components of the aggregate production function.
through which precipitation affects aggregate outcomes
is through its effect on agriculture, then only precipita- Investment
tion during crops growing periodpoorly proxied by
As hypothesized by Fankhauser and Tol (2005),
annual precipitationmay be relevant.
weather shocks could have long-lasting effects on
Annex Table3.3.1 also reveals the very persistent
output if they influence investment decisions, and
effects of temperature shocks. The lower half of panel
hence capital input. Equation (3.2) is estimated using
B presents the cumulative effects of a 1C increase in
real gross fixed capital formation as the outcome of
temperature estimated at the median temperature of
interest. The analysis also examines weathers impacts
advanced, emerging market, and low-income devel-
on imports, given the tight link between imports and
oping countries seven years after the shock. All but
investment. Results, presented in Annex Table3.3.3,
one specification show evidence of a long-lasting and
columns (1)(2), confirm the idea that temperature
potentially deepening adverse impact of temperature
shocks suppress investment. Although the uncertainty
shocks on per capita output at the temperatures experi-
surrounding the estimated contemporaneous effects
enced by the median low-income developing country.
is large, seven years after a temperature increase, both
To examine how widespread the effects of tempera-
investment and imports are significantly lower in coun-
ture may be, equation (3.2) is estimated using sectoral
tries with relatively hot climates (see also Figure3.10).
value added and agricultural production as the out-
comes of interest. Real value added of the agricultural,
Labor Input
manufacturing, and services sectors from the World
Banks World Development Indicators database is com- The analysis also examines whether labor supply
plemented with an index of crop production volume may be affected by temperature increases. Using infant
compiled by the United Nations Food and Agricul- mortality as the outcome of interest, equation (3.2) is
ture Organization. Results are presented in Annex estimated, uncovering a convex relationship between
Table3.3.2. There is a concave relationship between temperature and current (or future) labor supply
temperature and output in both the agricultural and (Annex Table3.3.3, column [3]). In hot countries, an
manufacturing sectors, whereas services value added increase in temperature raises infant mortality instan-
appears to be relatively protected from the effects of taneously, with the effect growing over time. In these
higher temperature. In other words, at the median countries, higher temperatures also have a negative
temperature of low-income countries, an increase in effect on a broader measure of human well-beingthe
temperature significantly reduces agricultural value Human Development Index, a weighted average of
added and crop production and lowers manufactur- per capita income, educational achievement, and life
ing output. expectancy (column [4]).
It is important to note that, unlike aggregate output,
agricultural production is significantly affected by pre- Productivity
cipitation in addition to temperature shocks. Although Motivated by the body of evidence of reduced
the results suggest a concave relationship between agri- human cognitive and physical performance at
cultural output and precipitation, at the typical levels high temperatures from laboratory experiments
of precipitation of all three country groups, an increase and country-specific studies, the analysis examines
in precipitation unambiguously improves agricultural whether reduced labor productivity may underpin
productivity. The effects of precipitation are also short the negative temperatureaggregate output relation-
lived; agricultural output seven years down the line is ship in countries with hot climates. If this is indeed
the case, sectors where workers are more exposed those that are heat-exposed and others to estimate
to heat should see a bigger decrease in labor pro- the following specification:53
ductivity when temperatures rise in relatively hot
53According to Graff Zivin and Neidell (2014), who follow
countries. The analysis uses the Groningen Growth
definitions from the National Institute for Occupational Safety and
and Development Centre 10-sector database, which Health, heat-exposed industries include agriculture, forestry, fishing
provides sectoral real value added and employment and hunting, construction, mining, transportation, and utilities
in 40 countries over 19502012, and Graff Zivin as well as manufacturing, in which facilities may not be climate
controlled in low-income countries and production processes often
and Neidells (2014) classification of sectors into
generate considerable heat.
C. Impact of a 100 mm per Year Increase in Precipitation on Dependent Variable Level at Horizon 0
AE (P=800 mm per year) 0.329 0.558*** 0.008 0.007 0.009 0.148
(0.262) (0.180) (0.015) (0.013) (0.133) (0.136)
EM (P=900 mm per year) 0.323 0.547*** 0.009 0.008 0.016 0.132
(0.246) (0.170) (0.015) (0.012) (0.125) (0.130)
LIDC (P=1,100 mm per year) 0.311 0.523*** 0.011 0.010 0.030 0.101
(0.216) (0.151) (0.013) (0.011) (0.109) (0.118)
Impact of a 100 mm per Year Increase in Precipitation on Dependent Variable Level at Horizon 7
AE (P=800 mm per year) 0.478 0.984** 0.071 0.102* 0.295 0.072
(0.689) (0.498) (0.163) (0.061) (0.832) (0.554)
EM (P=900 mm per year) 0.423 0.961** 0.074 0.097* 0.265 0.041
(0.649) (0.472) (0.149) (0.057) (0.776) (0.524)
LIDC (P=1,100 mm per year) 0.313 0.914** 0.080 0.087* 0.206 0.022
(0.573) (0.422) (0.123) (0.050) (0.666) (0.467)
Source: IMF staff calculations.
Note: Columns (14) present results from estimating equation (3.2) using the same specification as in Annex Table3.3.1, column (5), for different dependent
variables. Specification in column (5) presents results from estimating equation (3.4) where an indicator for heat exposed sectors is interacted with tempera-
ture and precipitation, their squared terms, and their lags and forwards; also controlling for country-sector and region-year fixed effects, and lag of growth.
Separate regressions are estimated for each horizon. In all specifications, standard errors are clustered at the country level. Panel A reports the estimated
coefficients on the weather variables for horizon 0. Panels B and C show the marginal impact of a change in temperature and precipitation computed as per
equation (3.3) at the median temperature (T) and median precipitation (P) of advanced economies (AE), emerging markets (EM), and low-income developing
countries (LIDC), contemporaneously (horizon 0) and cumulatively seven years after the shock. HDI = Human Development Index; mm = millimeter.
* p < 0.1; ** p < 0.05; *** p < 0.01.
yi,s,t+h yi,s,t1 = h1ci,t + h2ci,t2 + h1ci,t1 tions across countries and over time is not random.
Policies and institutions could also be correlated with
+ h2 ci,t1
2+ h1 h c
j=1 1 i,t+hj
relevant country attributes that are not controlled for in
+ j=1
h1 h c 2
2 i,t+hj + 1 ci,t Hs
h
the regression. Moreover, policy data availability varies
significantly in both temporal and country coverage,
+ h2 ci,t2 Hs + h1 ci,t1 Hs
resulting in sizable differences in the estimation sample.
+ h2 ci,t1 Hs +j=1
2 h1 h c
1 i,t+hj Hs For ease of interpretation, in the baseline results, each
+j=1
h1 h c 2 policy variable is transformed into an indicator variable
2 i,t+hjHs
depending on whether, in year t, the country is above or
+ h1
yi,s,t1 + hi,s + hr,t + hi,s,t
, (3.4) below the median value of this particular policy in the
in which yi,s,t is the log of real sectoral value added estimation sample.54 An exception to this approach is
per worker, Hsis an indicator for sectors that are the measurement of buffers. A country is considered to
heat-exposed, i,sh are country-sector fixed effects, and have (1) fiscal buffers if public debt as a share of GDP
hr,t are region-year fixed effects. Standard errors are is less than the 75th percentile, (2) monetary buffers if
clustered at the country level. annual inflation is less than 10percent, (3) high inter-
Annex Table3.3.3, specification (5) summarizes national reserves if international reserves minus gold can
the results of this estimation. At higher temperatures, cover at least four months of imports, (4) high foreign
an increase in temperature significantly lowers labor aid if foreign aid inflows as a share of GDP are in the
productivity in heat-exposed industries. Temperature 75th percentile, and (5) high remittances if per capita
increases, however, have no discernible effect on the remittances in real dollars received are greater than the
productivity of workers in non-heat-exposed sectors, 75th percentile. For exchange rate policy, the analysis
even in countries with hot climates. uses an indicator if the de facto exchange rate regime of
a country is not pegged based on the coarse classification
of Reinhart and Rogoff (2004).
The Role of Policies and Institutional Settings
Annex Tables 3.3.4 and 3.3.5 present the main find-
To study the extent to which macroeconomic and ings. For each policy, the tables report the estimated
structural policies and country characteristics mediate effect of a 1C increase in temperature on per capita
the effect of weather shocks, the analysis extends the output at horizons 0 through 7, where the policy is not
empirical approach described above by allowing the in place and where the policy is in place. The tables
response of per capita output to weather shocks to vary also report the p-value of a statistical test of the dif-
with various proxies for these policies. The estimated ference between the effect of temperature in different
specification augments equation (3.2) to include an policy scenarios.
interaction term between the weather shock and the The short-term negative effects of temperature
policy variable: shocks tend to be larger in countries with lower
yi,t + h yi,t 1 = h1 ci,t + h1(ci,t pi,t 1) + h1 pi,t 1 buffers, as evidenced by the larger estimated responses
in columns (2), (5), and (8) in Annex Table3.3.4.
+ h2c i,t1+ h2( ci,t1pi,t2)+ h2 pi,t2
However, the differences are typically not statistically
+h1
j=1 3c i,t + h j + 1
h h y
i,t 1 significant, and in the few cases in which they are
+ hi + hr,t + hi,t. (3.5) (fiscal buffers, foreign aid, and remittances), they tend
to be very short lived. Exchange rate regime, however,
The sample is restricted to countries with average
seems to be significantly associated with the extent
annual temperature exceeding 15C, in which an
of damage caused by weather shocks. Countries with
increase in temperature has a statistically significant
nonpegged exchange rates tend to recover faster from
linear negative impact on economic activity, as in Annex
these shocks. A similar pattern was documented by
Table3.3.1, column (9). Consequently, the weather
Ramcharan (2009), who finds that exchange rate flexi-
shock ci,t refers to average annual temperature and
bility helps economies adjust better in the aftermath of
precipitation. Most of the policy variables pi,t are lagged
windstorms and earthquakes.
to minimize reverse causality concerns and are included
one at a time. As emphasized in the chapter, it is difficult 54Results from an alternative specification in which the policy
to interpret causally the coefficients on the interaction variables are used in their continuous forms rather than transformed
terms, given that the variation in policies and institu- into indicators are available on request.
Impact of a 1C Increase
in Temperature on per Physical Capital Political Regime Index Inequality
Capita Output High Low P-value High Low P-value Low High P-value
Horizon 0 0.773*** 0.861*** 0.66 1.370*** 1.452*** 0.73 1.336*** 1.559*** 0.07
(0.294) (0.302) (0.328) (0.293) (0.431) (0.390)
Horizon 1 0.782* 0.777* 0.99 1.132*** 1.392*** 0.27 1.034* 1.240** 0.26
(0.405) (0.423) (0.393) (0.367) (0.580) (0.588)
Horizon 2 0.550 0.690 0.69 1.110*** 1.729*** 0.01 0.814 1.024* 0.35
(0.442) (0.459) (0.416) (0.433) (0.584) (0.591)
Horizon 3 0.430 0.820 0.30 1.374*** 1.929*** 0.03 0.947 1.386* 0.09
(0.411) (0.497) (0.466) (0.464) (0.714) (0.738)
Horizon 4 0.543 1.175** 0.15 1.599*** 2.095*** 0.09 0.819 1.391* 0.06
(0.464) (0.573) (0.566) (0.601) (0.827) (0.820)
Horizon 5 0.953 1.677** 0.17 1.587** 2.044*** 0.15 0.699 1.634* 0.01
(0.625) (0.755) (0.671) (0.705) (0.899) (0.877)
Horizon 6 0.381 1.546** 0.09 1.416** 2.128*** 0.06 1.061 2.067** 0.01
(0.586) (0.691) (0.679) (0.704) (0.930) (0.913)
Horizon 7 0.548 1.610* 0.14 1.325* 2.320*** 0.02 0.233 1.320 0.01
(0.645) (0.815) (0.751) (0.788) (1.060) (0.998)
The medium-term negative effects of temperature cross-country data. Combining subnational growth
shocks tend to be smaller in countries with better data from roughly 1,460 provinces and states across 79
structural policies and institutions (Annex Table3.3.5). countries from Gennaioli and others (2014) and annual
Standard errors are again quite large, and it is often temperature and precipitation data at the same level of
difficult to reject the hypothesis that policies do not aggregation, the analysis confirms that there is a nonlinear
have an effect, but the point estimates of the effect of relationship between subnational growth and temperature
temperature shocks in the outer horizons are substan- by estimating equation (3.2). It then zooms in on the set
tially larger in columns (2), (5), and (8). This evidence of provinces and states with average temperature greater
is in line with findings in the literature on the role of than 15C to examine whether economic activity in the
policies in attenuating the effects of natural disasters. hot states or provinces of advanced economies responds
See, among others, Kahn (2005); Noy (2009); Cavallo to a temperature increase in the same way as in states or
and others (2013); Felbermayr and Grschl (2014); provinces of emerging market and developing economies
and Breckner and others (2016) for the role of institu- with a similar average temperature. Equation (3.5) is
tional strength and democracy; Noy (2009); Von Peter, estimated with pi,t taking the value of 1 for states or prov-
Dahlen, and Saxena (2012); McDermott, Barry, and Tol inces located in advanced economies. pi,t is also interacted
(2013); Felbermayr and Grschl (2014); and Breckner with lag of growth, ih denote state or province fixed
and others (2016) for the role of financial markets; and effects, and region-year fixed effects, hr,t, are allowed
Noy (2009); Raddatz (2009); and Von Peter, Dahlen, to vary across advanced and non-advanced economies.
and Saxena (2012) for the role of development status. Standard errors are clustered at the province level.
Annex Table3.3.6 presents the estimated effects
of a 1C increase in temperature at horizons 0 to 7
The Role of Development in all subnational regions with temperature greater
The chapter examines whether the overall level of than 15C in column (1). The subsequent columns
development attenuates the negative effects of tem- present the estimated effects for subnational regions in
perature shocks in hot countries, using subnational advanced and non-advanced economies, as well as the
Annex Table3.4.1. Effect of Weather Shocks and Natural Disasters on Emigration, 19802015
Percent of Emigrants in
Total Population (1) (2) (3) (4) (5) (6)
Temperature 3.963 8.008* 8.067* 8.134* 8.127* 8.074*
(2.522) (4.477) (4.476) (4.357) (4.480) (4.287)
Precipitation 0.206 0.477 0.484 0.484 0.491 0.492
(0.710) (0.880) (0.878) (0.881) (0.878) (0.880)
Temperature LIDC 7.475* 7.672* 7.788* 7.571* 7.634*
(4.253) (4.255) (4.092) (4.249) (4.088)
Precipitation LIDC 0.935 0.918 0.929 0.972 0.992
(1.022) (1.018) (1.024) (1.039) (1.033)
Number of Natural Disasters 0.228* 0.228* 0.458 0.465*
(0.138) (0.136) (0.281) (0.269)
War 0.409 0.418
(2.283) (3.771)
Number of Natural Disasters LIDC 0.358 0.359
(0.309) (0.296)
War LIDC 1.216
(4.034)
Adjusted R2 0.04 0.06 0.06 0.06 0.06 0.05
Number of Observations 337 337 337 337 337 337
Source: IMF staff calculations.
Note: All specifications include country-of-origin fixed effects, decade-region fixed effects, and decade fixed effects interacted with a dummy for
low-income developing country (LIDC). Standard errors are clustered at the country level.
* p < 0.1; ** p < 0.05; *** p < 0.01.
p-value of a test of their difference. The negative effects country-decade. The latter three variables are further
of temperature shocks are felt much more heavily in interacted with a dummy identifying low-income devel-
non-advanced economies. oping countries (LIDC) to capture potential differences
in the emigration response to the weather fluctuations
Annex3.4. The Impact of Weather Changes and and natural disasters. As in Cattaneo and Peri (2016),
Natural Disasters on International Migration the regression further controls for country fixed effects
This annex provides additional details on the empirical ( i), region-decade fixed effects ( r,d), and decade fixed
analysis of the effect of temperature shocks and natural effects interacted with the LIDC dummy. The random
disasters on international migration. The analysis relies error term i,dis clustered at the country level.56 The
on data from zden and others (2011) on emigrant specification is purposefully parsimonious. Controls
stocks for 117 economies with average temperature typically included as determinants of migrations, such
greater than 15C between 1980 and 2015. Migrant as population size, sociopolitical environment, and
stocks, which are available at 10-year intervals, are differ- others, could themselves be affected by weather fluctu-
enced to compute net emigrant flows in each decade. ations and natural disasters. In a robustness check, the
Building on Cattaneo and Peri (2016), the analysis exercise controls for the incidence of war, an important
estimates the following specification: push factor for emigration, although arguably this
could be yet another channel through which weather
Emigranti,d =+Ti,d + Ti,d
LIDCi +Pi,d fluctuations trigger movements of people (see Burke,
Pi,d
+ LIDCi +Disasteri,d Hsiang, and Miguel 2015b).
+ Disasteri,d LIDCi + i Annex Table3.4.1 reports the main findings from
estimating equation (3.6). Higher average temperatures
+ r,d + d LIDCi + i,d, (3.6)
in which i indexes countries, d indexes decades,55 56Following Dell, Jones, and Olken (2012), the specifica-
Emigrant is the net flow of emigrants over the decade tion includes only fixed effects as controls, since other potential
as a percentage of the total population of the origin controls, such as population size or sociopolitical environment,
(source) country, T is the average temperature and P may themselves be affected by agricultural productivitya key
channel through which weather shocks may influence emigration
the average precipitation for the decade, and Disaster
potentially producing a bias in the estimation by introducing an
is the average number of natural disasters for each overcontrolling problem. The only exception is a dummy for wars
(see Beaton and others 2017), which is included in some of the
55The 2010 decade includes data up to 2015. specifications and confirms the robustness of the findings.
over a decade do not have a significant effect on emi- These goods are combined into a constant elasticity of
gration in the full sample of countries (column [1]). substitution basket, and savers maximize the pres-
However, once the response is allowed to vary across ent value of their lifetime utility. The model breaks
broad groups of countries, the results suggest that in Ricardian equivalence by including both savers and
countries that are not classified as low income, higher hand-to-mouth consumers.
temperature is indeed associated with greater emigra- The government spends on transfers, debt service,
tion flows (column [2]). A 1C increase in average and (partially inefficient) infrastructure investment. It
decadal temperature leads to an increase in the share collects revenue from the consumption value-added
of net emigrants of about 8percentage points (which tax and from user fees for infrastructure services. The
is equivalent to one standard deviation in the sample deficit is financed through domestic borrowing, exter-
investigated).57 Similarly, more natural disasters over a nal concessional borrowing, or external commercial
decade also increase net emigration flows, especially in borrowing. Policymakers accept all concessional loans
countries not classified as low income.58 offered by official creditors. The borrowing and amor-
tization schedule for these loans is fixed exogenously.
Debt sustainability requires that the value-added tax
Annex3.5. Model-Based Analysis and transfers eventually adjust to cover the entire defi-
The model used to analyze the long-term impact cit, given the exogenously determined upper limit on
of climate change and simulate the effects of policies taxes and lower limit on transfers. The model incor-
in Box 3.2 is developed and presented in Buffie and porates shocks to the government external debt risk
others (2012). It is commonly known as the Debt, premium (or world interest rates).
Investment, and Growth (DIG) model and has served The majority of the model parameters are set to
as a workhorse in many IMF studies of low-income the same values as in Buffie and others (2012), with
countries. The DIG is an optimizing intertemporal few exceptions, mostly to reflect the decline in global
model with perfect foresight. It describes a two-sector interest rates, the projection of trend GDP growth
small open economy model with private and pub- in low-income countries, and the sample median of
lic capital, learning by doing, and endogenous fiscal public-debt-to-GDP ratios. The parameters that differ
policies. Public capital is productive and is used in from the ones in Buffie and others (2012) are pre-
the production function in both sectors. Government sented in Annex Table3.5.1.
spending can raise output directly by augmenting the
stock of public capital and can crowd in and crowd
out private investment. Simulating the Long-Term Impact of Climate Change
Firms operate Cobb-Douglas technologies to com- To trace the long-term impact of climate change,
bine labor, private capital, and public capital (infra- the model incorporates the estimated relationship
structure) into output in the traded and nontraded between temperature and per capita output discussed
sectors. The evolution of total factor productivity in Annex3.3 and presented in Annex Table3.3.1,
(TFP) is exogenous in both sectors. Firms face separate column (5). The effect is assumed to occur through
prices for exports, and imports and are assumed to be temperatures effect on TFP; therefore, the estimated
profit maximizing. parameters are rescaled so that the model matches the
Consumers supply labor and derive utility from empirically estimated decline of GDP if temperature
consuming the domestic traded good, the foreign increases by 1C.59
traded good, and the domestic nontraded good. The temperature during 20172100 is assumed to
follow one of two alternative scenarios: Representative
57The flow of emigrants as a share of population in countries Concentration Pathway (RCP) 4.5 or RCP 8.5. The
that are not classified as low income in this sample is 2.5percent, temperature increases during 20172100 are calculated
on average, with a standard deviation of 8.1percentage points. For for the median low-income country in the sample and
low-income countries, these statistics are 0.6percent and 2.2per-
centage points, respectively. are equal to 2.0C and 3.9C for RCP 4.5 and RCP
58Results (not shown here and available on request) are robust to 8.5, respectively.
the use of other proxies for low-income countries, such as a dummy
identifying the countries in the bottom quartile of the average GDP
per capita distribution of the country sample during the full sample 59Estimates of the damage to GDP cannot be used directly given
There are two sources of uncertainty in the temperatures, the employment share of nontraded
simulationthe uncertainty of RCP projections goods increases from the baseline value of 42.27per-
and the uncertainty of the effect of temperature on cent to 65percent over 90 years.
TFP. Both sources of uncertainty are combined in
the analysis as follows. The upper-bound scenario is
simulated assuming that the temperature increase is Modeling Optimal Adaptation
equal to the lowest 5th percentile for each RCP.60 Box3.2 extends the original DIG model to incor-
To account for the uncertainty of estimated param- porate direct investment in adaptation strategies. The
eters, the TFP parameters are set to the conditional main addition is the inclusion of private adaptation
expected value for the upper 50percent of the TFP and public subsidies to private adaptation, whereas
distribution. The worst lower-bound scenario is sim- damages are modeled as before. In the absence of any
ulated analogously. adaptation measure, increased temperature causes
gross damage, denoted by GDjt , at time t in sector j.
The gross damage is expressed as a fraction of sec-
Modeling Structural Transformation toral output:
Structural transformation is generated in the DIG
GDjt
model by introducing diverging trends in sectoral TFP gdjt =_ (T).
q jt=f
growth, along the lines of Ngai and Pissarides (2007).
In their model, faster productivity growth in the Gross damage can be reduced by investing in adap-
traded goods sector goes along with a decline in the tation. Firm is capacity to adapt to climate change is
relative price of traded versus nontraded goods. Given denoted by Oi,jt. It is increasing in firm is protection
complementarity in final demand, production in the expenditures ADi,jt as well as in the total sectoral pro-
former sector relative to the latter does not increase tection expenditures AD 1
jt = di.61 The residual
0 ADi,jt
in the same proportion. The value share of the traded damage for firm i in sector j is
goods sector eventually shrinks, even in the presence gdjt
of international trade. While this approach relies on i,jt =_____________
,
Oi,jt (ADi,jt
,AD jt)
only one potential driver of structural transformation,
it generates the desired increase in employment and in which the marginal damage reduction from adap-
nominal-value-added shares of the nontraded goods tation spending is decreasing. The positive parameter
sector, which is mostly composed of services. The gap is the elasticity of damage reduction to the level
in sectoral TFP growth rates is set to replicate the of adaptation.
average increase in the service share of value added in If the cost of a unit of protection is equal to PAD,t
low-income developing countries in 19902015, which and the functional form for the capacity to adapt is
has risen at the rate of 2.5percentage points a decade. Oi,jt( ADi,jt
,AD jt; ) = ADi,jt
AD (with 0 1
jt ),
Given this calibration, in the scenario without rising then cost minimization by firms in the symmetric
increases are 1.2C to 2.8C and 2.8C to 5.1C for RCP 4.5 and 61Many adaptation measures have the nature of public goods;
RCP 8.5, respectively. hence, firms benefit from total sectoral protection spending.
equilibrium ADi,jt
=AD jt determines the optimal Annex3.6. Reduced Form Approach to
level of adaptation expenditure for each firm Estimating Potential Long-Term Effects of
1
_________
Climate Change
( PAD,t )
GDjt 1+(1+)
= ____
ADi,jt Indicative evidence of the potential impacts of climate
change and their distribution across the globe could
The optimal level of firm-specific residual also be gleaned by combining the estimated sensitivity
damage is then of per capita output to temperature increase (Annex
Table3.3.1, column [5]), baseline annual temperatures,
gdjt
jt =_______
(1+)
, and projected temperature changes for each geographic
ADjt
location. As in the modeling exercise, this analysis
which can be shown to be socially suboptimal. takes the most conservative approach and assumes
The social planners cost function, TotDi,jt, differs temperature increases have a permanent level, rather
from that of individual firms than growth, effect on per capita output. The estimated
cumulative impact on 2100 per capita GDP under the
SP = GDjt (ADjt SP )
(1+)
TotDi,jt ADjt SP .
+PAD,t Representative Concentration Pathway (RCP) 4.5 and
Minimizing the social cost gives socially optimal RCP 8.5 scenarios are presented in Annex Figure3.6.1.
adaptation expenditures It is important to note that this exercise captures the
likely impact of one particular aspect of climate change,
1
_________ namely temperature increases. The macroeconomic
[ ( ) PAD,t ]
GDjt 1+(1+)
ADjt SP = 1 + ____ effects of many expected or possible events (such as
higher incidence of natural disasters, rising sea levels,
It can be shown that private agents invest less than
ocean acidification, and the like) are not quantified in
the socially optimal amount. The adaptation spending
this exercise. Furthermore, the analysis abstracts from
gap (as a fraction of the socially optimal adaptation
cross-border spillovers that may arise if climate change
spending) is equal to
triggers more frequent epidemics, famines, and other
1
_________
natural disasters along with social unrest, armed conflict,
1 (____
1+ )
1 1+(1+)
. and associated refugee flows.
It can also be shown that the socially optimal The analysis suggests that the projected warming
amount of adaptation expenditures can be achieved if will have uneven effects across the globe. However, the
subsidies in the amount of ,jtper unit cost of protec- increase in temperature, especially under the RCP 8.5
tion are paid by the government to the firms scenario, will push many advanced economies beyond
the threshold temperature level, thus triggering direct
,jt =_____ . economic losses for these countries as well.
( 1 + )
Annex Figure 3.6.1. The Long-Term Impact of Temperature Increase on Real per Capita Output across the Globe
(Percent)
1. RCP 4.5 Scenario
12 +
8
4
0
4
18 +
12
6
0
6
12
Sources: National Aeronautics and Space Administration (NASA) Earth Exchange Global Daily Downscaled Projections (NEX-GDDP); World Bank Group Cartography
Unit; and IMF staff calculations.
Note: The maps depict the effect of the projected increase in temperature between 2005 and 2100 under RCP 4.5 and RCP 8.5 scenarios on real per capita output in
2100. Gray areas indicate the estimated impact is not statistically signicant. RCP = Representative Concentration Pathways.
Cabezon, Ezequiel, Leni Hunter, Patrizia Tumbarello, Kazuaki Cole, Shawn, Xavier Gine, Jeremy Tobacman, Petia Topalova,
Washimi, and Yigun Wu.2015. Enhancing Macroeconomic Robert Townsend, and James Vickery. 2013. Barriers to
Resilience to Natural Disasters and Climate Change in the Household Risk Management: Evidence from India. Ameri-
Small States of the Pacific. IMF Working Paper 15/125, can Economic Journal: Applied Economics 5: 10435.
International Monetary Fund, Washington, DC. Colmer, Jonathan. 2016. Weather Labor Reallocation and
Caldern, Csar, Enrique Moral-Benito, and Luis Servn. 2015. Industrial Production: Evidence from India. Unpublished,
Is Infrastructure Capital Productive? A Dynamic Het- Massachusetts Institute of Technology, Cambridge, MA.
erogeneous Approach. Journal of Applied Econometrics 30 Cook, John, Dana Nuccitelli, Sarah A. Green, Mark Richardson,
(2): 17798. Brbel Winkler, Rob Painting, Robert Way, Peter Jacobs,
Carleton, Tamma A. 2017. Crop-Damaging Temperatures and Andrew Skuce. 2013. Quantifying the Consensus on
Increase Suicide Rates in India. Proceedings of the National Anthropogenic Global Warming in the Scientific Literature.
Academy of Sciences 114 (33): 874651. Environmental Research Letters 8 (2).
, and Solomon M. Hsiang.2016. Social and Economic Cook, John, Naomi Oreskes, Peter T. Doran, William Anderegg,
Impacts of Climate. Science 353 (6304). Bart Verheggen, Ed W. Maibach, J. Stuart Carlton, Stephan
Cashin, Paul, Kamiar Mohaddes, and Mehdi Raissi.2017. Fair Lewandowsky, Andrew G. Skuce, Sarah A. Green, Dana Nuc-
Weather or Foul? The Macroeconomic Effects of El Nio. citelli, Peter Jacobs, Mark Richardson, Brbel Winkler, Rob
Journal of International Economics 106: 3754. Painting, and Ken Rice. 2016. Consensus on Consensus: A
Cattaneo, Cristina, and Giovanni Peri. 2016. The Migration Synthesis of Consensus Estimates on Human-Caused Global
Response to Increasing Temperatures. Journal of Development Warming. Environmental Research Letters 11 (4).
Economics 122: 12746. Cooper, Gail. 2002. Air-Conditioning America: Engineers and
Cavallo, Edurardo, Sebastian Galiani, Ilan Noy, and Juan Pan- the Controlled Environment, 19001960. Johns Hopkins
tano. 2013. Catastrophic Natural Disasters and Economic Studies in the History of Technology. Baltimore, MD: Johns
Growth. Review of Economics and Statistics 95 (5): 154961. Hopkins University Press.
Center for International Earth Science Information Network Costinot, Arnaud, Dave Donaldson, and Cory Smith.2016.
(CIESIN) Columbia University. 2016. Gridded Population Evolving Comparative Advantage and the Impact of Climate
of the World, Version 4 (GPWv4): Population Count. NASA Change in Agricultural Markets: Evidence from 1.7Million
Socioeconomic Data and Applications Center (SEDAC), Pal- Fields around the World. Journal of Political Economy 124
isades, NY. http://dx.doi.org/10.7927/H4X63JVC. Accessed (1):20548.
March 15, 2017. Cullen, H., P. B. deMenocal, S. Hemming, G. Hemming, F.
, United Nations Food and Agriculture Programme H. Brown, T. Guilderson, and F. Sirocko. 2000. Climate
(FAO), and Centro Internacional de Agricultura Tropical Change and the Collapse of the Akkadian Empire: Evidence
(CIAT). 2005. Gridded Population of the World, Version from the Deep Sea. Geology 28 (4): 37982.
3 (GPWv3): Population Count Grid. NASA Socioeco- Dell, Melissa, Benjamin F. Jones, and Benjamin A. Olken.2012.
nomic Data and Applications Center (SEDAC), Palisades, Temperature Shocks and Economic Growth: Evidence from
NY. http://dx.doi.org/10.7927/H4639MPP. Accessed the Last Half Century. American Economic Journal: Macroeco-
March 15, 2017. nomics 4 (3): 6695.
Chamberlain, Gary. 1980. Analysis of Covariance with Qualita- .2014. What Do We Learn from the Weather? The
tive Data. Review of Economic Studies 47: 22538. New ClimateEconomy Literature. Journal of Economic Liter-
Chaney, Eric. 2013. Revolt on the Nile: Economic Shocks, ature 52 (3): 74098.
Religion, and Political Power. Econometrica 81: 203353. Deryugina, Tatyana. 2011. The Role of Transfer Payments in
Chen, Chen, Ian Noble, Jessica Hellmann, Joyce Coffee, Martin Mitigating Shocks: Evidence from the Impact of Hurricanes.
Murillo, and Nitesh Chawla. 2015. University of Notre Dame MunichPersonal RePEc Archive (MPRA) Paper 53307, Munich.
Global Adaptation Index. Country Index Technical Report. , and Solomon M. Hsiang. 2014. Does the Environ-
Chen, Joyce J., Valerie Mueller, Yuanyuan Jia, and Steven ment Still Matter? Daily Temperature and Income in the
Kuo-Hsin Tseng. 2017. Validating Migration Responses to United States. NBER Working Paper 20750, National
Flooding Using Satellite and Vital Registration Data. Ameri- Bureau of Economic Research, Cambridge, MA.
can Economic Review: Papers and Proceedings 107 (5): 44145. Deschnes, Olivier.2012. Temperature, Human Health and
Cole, Shawn, Gautam Bastian, Sangita Vyas, Carina Wendel, Adaptation: A Review of the Empirical Literature. NBER
and Daniel Stein. 2012. The Effectiveness of Index-Based Working Paper 18345, National Bureau of Economic
Micro-Insurance in Helping Smallholders Manage Research, Cambridge, MA.
Weather-Related Risks. London: Evidence for Policy and , and Michael Greenstone.2011. Climate Change,
Practice Information and Co-ordinating (EPPI) Centre, Mortality, and Adaptation: Evidence from Annual Fluctu-
Social Science Research Unit, Institute of Education, Univer- ations in Weather in the US. American Economic Journal:
sity of London. Applied Economics 3 (4): 15285.
Fankhauser, Samuel, and Richard Tol. 2005. On Climate .2012b. Natural Disasters and Population Mobility in
Change and Economic Growth. Resource and Energy Econom- Bangladesh. Proceedings of the National Academy of Sciences
ics 27: 117. 109 (16): 600005.
Farid, Mai, Michael Keen, Michael Papaioannou, Ian Parry, Guha-Sapir, Debarati, Regina Below, and Philippe Hoyois. 2015.
Catherine Pattillo, and Anna Ter-Martirosyan.2016. After EM-DAT: The CRED/OFDA International Disaster Data-
Paris: Fiscal, Macroeconomic, and Financial Implications of base. Universit Catholique de Louvain, Brussels, Belgium.
Climate Change. IMF Staff Discussion Note 16/01, Interna- www.emdat.be.
tional Monetary Fund, Washington, DC. Guo, Y., A. Gasparrini, B. Armstrong, S. Li, B. Tawatsupa, A.
Felbermayr, Gabriel, and Jasmin Grschl. 2014. Naturally Tobias, E. Lavigne, M. de Sousa Zanotti Stagliorio Coelho,
Negative: The Growth Effects of Natural Disasters. Journal of M. Leone, X. Pan, S. Tong, L. Tian, H. Kim, M. Hashizume,
Development Economics 111: 92106. Y. Honda, YL Guo, CF Wu, K. Punnasiri, SM Yi, P. Mich-
Feng, Shuaizhang, Alan B. Krueger, and Michael Oppenheimer. elozzi, PH Saldiva, and G. Williams. 2014. Global Variation
2010. Linkages among Climate Change, Crop Yields, and in the Effects of Ambient Temperature on Mortality: A
MexicoUS Cross-Border Migration. Proceedings of the Systematic Evaluation. Epidemiology 25 (6): 78189.
National Academy of Sciences 107 (32): 1425762. Hallegatte, Stphane. 2009. Strategies to Adapt to an Uncer-
Feng, Shuaizhang, Michael Oppenheimer, and Wolfram tain Climate Change. Global Environmental Change 19
Schlenker. 2012. Climate Change, Crop Yields, and Internal (2): 24047.
Migration in the United States. NBER Working Paper 17734, , Mook Bangalore, Laura Bonzanigo, Marianne Fay,
National Bureau of Economic Research, Cambridge, MA. Tamaro Kane, Ulf Narloch, Julie Rozenberg, David Treguer,
Fenske, James, and Namrata Kala. 2015. Climate and the Slave and Adrien Vogt-Schilb. 2016. Shock Waves: Managing the
Trade. Journal of Development Economics 112: 1932. Impacts of Climate Change on Poverty. Climate Change and
Fomby, Thomas, Yuki Ikeda, and Norman Loayza. 2013. The Development Series. Washington, DC: World Bank.
Growth Aftermath of Natural Disasters. Journal of Applied Hallegatte, Stphane, Jean-Charles Hourcade, Patrice Dumas.
Econometrics 28 (3): 41234. 2007. Why Economic Dynamics Matter in Assessing
Foster, Vivien, and Cecilia Briceno-Garmendia. 2010. Climate Change Damages: Illustration on Extreme Events.
Africas Infrastructure: A Time for Transformation. Africa Ecological Economics 62: 33040.
Development Forum. Washington, DC: World Bank. Hallegatte, Stphane, Franck Lecocq, and Christian de Perthuis.
https://openknowledge.worldbank.org/handle/10986/2692. 2011. Designing Climate Change Adaptation Policies: An
Galor, Oded. 2011. Unified Growth Theory. Princeton, NJ: Economic Framework. Policy Research Working Paper 5568,
Princeton University Press. World Bank, Washington, DC.
Garg, Teevrat, Maulik Jagnani, and Viz Taraz. 2017. Human Hallegatte, Stphane, and Julie Rozenberg. 2017. Climate Change
Capital Costs of Climate Change: Evidence from Test Scores through a Poverty Lens. Nature Climate Change 4: 25056.
in India. Unpublished. Hauer, Mathew E., Jason M. Evans, and Deepak R. Mishra.
Gennaioli, Nicola, Rafael La Porta, Florencio Lopez De Silanes, 2016. Millions Projected to Be at Risk from Sea-Level
and Andrei Shleifer. 2014. Growth in Regions. Journal of Rise in the Continental United States. Nature Climate
Economic Growth 19 (3): 259309. Change 6: 69195.
Gerling, Kerstin. Forthcoming. The Macro-Fiscal Aftermath Haug, Gerald H., Detlef Gnther, Larry C. Peterson, Daniel M.
of Weather-Related Disasters in Developing Asia: Do Loss Sigman, Konrad A. Hughen, and Beat Aeschlimann. 2003.
Dimensions Matter? IMF Working Paper, International Climate and the Collapse of Maya Civilization. Science 299
Monetary Fund, Washington, DC. (5613): 173135.
, Maria Moreno Badia, and Priscilla Toffano. Forth- Heal, Geoffrey, and Jisung Park. 2016. Temperature Stress
coming. A Primer on the Macro-Fiscal Impact of Natural and the Direct Impact of Climate Change: A Review of an
Disasters. IMF Working Paper, International Monetary Emerging Literature. Review of Environmental Economics and
Fund, Washington, DC. Policy 10 (2): 34762.
Graff Zivin, J. S., Solomon M. Hsiang, and Matthew J. Neidell. Hong, Harrison, Frank Li, and Jiangmin Xu. 2016. Climate
2015. Temperature and Human Capital in the Short- and Risks and Market Efficiency. NBER Working Paper 22890,
Long-Run. NBER Working Paper 21157, National Bureau National Bureau of Economic Research, Cambridge, MA.
of Economic Research, Cambridge, MA. Hope, Chris. 2011. The PAGE09 Integrated Assessment Model:
Graff Zivin, J. S., and Matthew J. Neidell.2014. Tempera- A Technical Description. Cambridge Judge Business School
ture and the Allocation of Time: Implications for Climate Working Papers 4/11.
Change. Journal of Labor Economics 32 (1): 126. Hornbeck, Richard.2012. The Enduring Impact of the
Gray, Clark L., and Valerie Mueller.2012a. Drought and American Dust Bowl: Short- and Long-Run Adjustments to
Population Mobility in Rural Ethiopia. World Development Environmental Catastrophe. American Economic Review 102
40 (1): 13445. (4): 1477507.
, and Suresh Naidu. 2014. When the Levee Breaks: Credit and Risk Constraints. Quarterly Journal of Economics
Black Migration and Economic Development in the Ameri- 129 (2): 597652.
can South. American Economic Review 104 (3): 96390. Kelley, Colin P., Shahrzad Mohtadi, Mark A. Cane, Richard
Hsiang, Solomon M. 2016. Climate Econometrics. Annual Seager, and Yochanan Kushnir. 2015. Climate Change in
Review of Resource Economics 8: 4375. the Fertile Crescent and Implications of the Recent Syrian
, and Amir Jina. 2014. The Causal Effect of Environ- Drought. Proceedings of the National Academy of Sciences 112
mental Catastrophe on Long-Run Economic Growth: Evi- (11): 324146.
dence from 6,700 Cyclones. NBER Working Paper 20352, Kim, Nam Kyu. 2014. Revisiting Economic Shocks and
National Bureau of Economic Research, Cambridge, MA. Coups. Journal of Conflict Resolution 60: 331.
Hsiang, Solomon M., and Daiju Narita. 2012. Adaptation Klein Goldewijk, Kees, Arthur Beusen, Jonathan Doelman, and
to Cyclone Risk: Evidence from the Global Cross-Section. Elke Stehfest. 2016. New Anthropogenic Land Use Estimates
Climate Change Economics 3 (2). for the Holocene; HYDE 3.2. Earth System Science Data.
Hulten, Charles R. 1996. Infrastructure Capital and Economic Knapp, Kenneth, Scott Applequist, Howard Diamond, James
Growth: How Well You Use It May Be More Important Kossin, Michael Kruk, and Carl Schreck. 2010. NCDC
Than How Much You Have. NBER Working Paper 5847, International Best Track Archive for Climate Stewardship
National Bureau of Economic Research, Cambridge, MA. (IBTrACS) Project, Version 3. Revision 09. NOAA, National
Ilzetzki, Ethan, Carmen M. Reinhart, and Kenneth S. Rogoff. Centers for Environmental Information.
2008. Exchange Rate Arrangements Entering the 21st Century: Knapp, Kenneth, Michael Kruk, David Levinson, Howard Dia-
Which Anchor Will Hold? NBER Working Paper 23134, mond, and Charles Neumann. 2010. The International Best
National Bureau of Economic Research, Cambridge, MA. Track Archive for Climate Stewardship (IBTrACS): Unifying
Intergovernmental Panel on Climate Change (IPCC).2013. Tropical Cyclone Best Track Data. Bulletin of the American
Climate Change 2013: The Physical Science Basis. Sum- Meteorological Society 91: 36376.
mary for Policymakers. Contribution of Working Group I to Knutson, Thomas, John McBride, Johnny Chan, Kerry Eman-
the Fifth Assessment Report of the Intergovernmental Panel uel, Greg Holland, Chris Landsea, Isaac Held, James Kossin,
on Climate Change. Cambridge, United Kingdom and New A. Srivastava, and Masato Sugi. 2010. Tropical Cyclones and
York, NY: Cambridge University Press. Climate Change. Nature Geoscience 3: 15763.
.2014. Climate Change2014: Impacts, Adaptation, Knutson, Thomas, and Robert Tuleya. 2004. Impact of
and Vulnerability. Part A: Global and Sectoral Aspects. CO2-Induced Warming on Simulated Hurricane Intensity
Contribution of Working Group II to the Fifth Assessment and Precipitation: Sensitivity to the Choice of Climate
Report of the Intergovernmental Panel on Climate Change. Model and Convective Parameterization. Journal of Climate
Cambridge, United Kingdom and New York, NY: Cambridge 17: 347795.
University Press. Kudamatsu, Masayuki, Torsten Persson, and David Strmberg.2012.
International Monetary Fund (IMF).2015. Macroeconomic Weather and Infant Mortality in Africa. CEPR Discussion
Developments and Prospects in Low-Income Developing Paper 9222, Centre for Economic Policy Research, London.
Countries. Policy Paper, Washington, DC. Landsea, Christopher. 2004. Hurricane FAQ: Why Do Tropical
.2016a. Enhancing Resilience to Natural Disasters Cyclones Require 80F (26.5C) Ocean Temperatures to
in Sub-Saharan Africa. In Regional Economic Outlook: Form? NOAA, Hurricane Research Division of the Atlantic
Sub-Saharan Africa. Washington DC, October. Oceanographic and Meteorological Laboratory, Miami, FL.
.2016b. Small States Resilience to Natural Disas- http://www.aoml.noaa.gov/hrd/tcfaq/A16.html.
ters and Climate Change: Role for the IMF. Board Paper, Lane, Philip R., and Gian Maria Milesi-Ferretti. 2017. Inter-
Washington, DC. national Financial Integration in the Aftermath of the Global
.2017. Seychelles: Climate Change Policy Assessment. Financial Crisis. IMF Working Paper 17/115, International
IMF Country Report 17/162, Washington, DC. Monetary Fund, Washington, DC.
Jennings, Steve.2011. Times Bitter Flood: Trends in the Num- Lanzafame, Matteo. 2014. Temperature, Rainfall, and Eco-
ber of Reported Natural Disasters. Oxfam Policy and Practice: nomic Growth in Africa. Empirical Economics 46 (1): 118.
Climate Change and Resilience 7 (1): 11547. Lee, Minsoo, Mai Lin Villaruel, and Raymond Gaspar. 2016.
Jord, scar.2005. Estimation and Inference of Impulse Effects of Temperature Shocks on Economic Growth and
Responses by Local Projections. American Economic Review Welfare in Asia. ADB Economics Working Paper 501, Asian
95 (1): 16182. Development Bank, Mandaluyong City, Philippines.
Kahn, Matthew. 2005. The Death Toll from Natural Disasters: Lee, Tien Ming, Ezra M. Markowitz, Peter D. Howe, Chia-Ying
The Role of Income, Geography and Institutions. Review of Ko, and Anthony A. Leiserowitz. 2015. Predictors of Public
Economics and Statistics 87 (2): 27184. Climate Change Awareness and Risk Perception around the
Karlan, Dean S., Robert Osei, Isaac Osei-Akoto, and Chris- World. Nature Climate Change 5: 101420.
topher Udry. 2014. Agricultural Decisions after Relaxing
Lemoine, Derek. 2017. Expect above Average Temperatures: Noy, Ilan.2009. The Macroeconomic Consequences of Disas-
Identifying the Economic Impact of Climate Change. ters. Journal of Development Economics 88 (2): 22131.
NBER Working Paper 23549, National Bureau of Economic Nunn, Nathan. 2008. The Long-Term Effects of Africas Slave
Research, Cambridge, MA. Trades. Quarterly Journal of Economics 123 (1): 13976.
Lloyds. 2012. Global Underinsurance Report. London: . 2014. Historical Development. In Handbook of
October 2012. Economic Growth (2), edited by Philippe Aghion and Steven
Lobell, David B., Wolfram Schlenker, and Justin Costa-Roberts. Durlauf, 347402. Amsterdam: North-Holland.
2011. Climate Trends and Global Crop Production since Organisation for Economic Co-operation and Development
1980. Science 333 (6042): 61620. (OECD). 2015a. Climate Change Risks and Adaptation: Link-
Marchiori, Luca, Jean-Francois Maystadt, and Ingmar Schum- ing Policy and Economics. Paris: OECD Publishing.
acher. 2012. The Impact of Weather Anomalies on . 2015b. Climate Finance in 201314 and the USD
Migration in Sub-Saharan Africa. Journal of Environmental 100 Billion Goal. A report by the OECD in collaboration
Economics and Management 63 (3): 35574. with Climate Policy Initiative, Paris.
Marcott, Shaun A., Jeremy D. Shakun, Peter U. Clark, and zden, alar, Christopher Parsons, Maurice Schiff, and Terrie
Alan C. Mix. 2013. A Reconstruction of Regional and L. Walmsley. 2011. Where on Earth Is Everybody? The
Global Temperature for the Past 11,300 Years. Science 339 Evolution of Global Bilateral Migration, 19602000. World
(6124): 1198201. Bank Economic Review 25 (1): 1256.
Margulis, Sergio, and Urvashi Narain. 2010. The Costs to Devel- Park, Jisung. 2016. Will We Adapt? Temperature Shocks,
oping Countries of Adapting to Climate Change: New Methods Labor Productivity, and Adaptation to Climate Change
and EstimatesThe Global Report of the Economics of Adapta- in the United States. Unpublished. Harvard University,
tion to Climate Change Study. Washington: World Bank. Cambridge, MA.
Matsuura, Kenji, and Cort J. Willmott. 2007. Terrestrial Air . 2017. Heat Stress and Human Capital Production.
Temperature: 19002006 Gridded Monthly Time Series (Ver- Unpublished, Harvard University, Cambridge, MA.
sion 1.01) 2009. http://climate.geog.udel.edu/~climate/html Parry, Ian W. H., Ruud de Mooij, and Michael Keen. 2012.
_pages/Global2_Ts_2009/README.global_t_ts_2009.html. Fiscal Policy to Mitigate Climate Change: A Guide for Policy-
McDermott, Thomas K. J., Frank Barry, and Richard S. J. Tol. makers. Washington, DC: International Monetary Fund.
2013. Disasters and Development: Natural Disasters, Credit Parry, Ian W. H., Adele Morris, and Roberton Williams, editors.
Constraints, and Economic Growth. Oxford Economic Papers 2015. Implementing a US Carbon Tax: Challenges and Debates.
66 (3): 75073. London: Routledge.
Mueller, Valerie, Clark Gray, and Katrina Kosec. 2014. Heat Parry, Ian W. H., Victor Mylonas, and Nate Vernon. 2017.
Stress Increases Long-Term Human Migration in Rural Paki- Reforming Energy Policy in India: Assessing the Options.
stan. Nature Climate Change 4: 18285. IMF Working Paper 17/103, International Monetary Fund,
Munshi, Kaivan.2003. Networks in the Modern Economy: Washington, DC.
Mexican Migrants in the US Labor Market. Quarterly Jour- Parry, Ian W. H., Baoping Shang, Philippe Wingender, Nate
nal of Economics 118 (2): 54999. Vernon, and Tarun Narasimhan. 2016. Climate Mitigation
National Oceanic and Atmospheric Administration (NOAA). in China: Which Policies Are Most Effective? IMF Working
2017a. Climate at a Glance: Global Time Series. National Paper 16/148, International Monetary Fund, Washington, DC.
Centers for Environmental Information, NOAA, Department of Peng, Wei, and Alan Feng. Forthcoming. Underreacting to Climate
Commerce, Washington, DC.http://www.ncdc.noaa.gov/cag/. Shocks: Evidence from Global Stock Markets.
. 2017b. Tropical Cyclone Climatology. National Hur- Porter, Catherine, and Emily White. 2016. Potential for
ricane Center, NOAA, Department of Commerce, Washing- Application of a Probabilistic Catastrophe Risk Modelling
ton, DC. http://www.nhc.noaa.gov/climo/. Framework to Poverty Outcomes: General Form Vulnerability
Ngai, L. Rachel, and Christopher A. Pissarides. 2007. Struc- Functions Relating Household Poverty Outcomes to Hazard
tural Change in a Multi-Sector Model of Growth. American Intensity in Ethiopia. Policy Research Working Paper 7717,
Economic Review 97 (1): 42943. World Bank, Washington, DC.
Nordhaus, William D. 2010. Economic Aspects of Global Pritchett, Lant. 2000. The Tyranny of Concepts: CUDIE
Warming in a Post-Copenhagen Environment. Proceedings of (Cumulated, Depreciated, Investment Effort) Is Not Capital.
the National Academy of Sciences 107 (26): 1172126. Journal of Economic Growth 5 (4): 36184.
Nordhaus, William D., and Andrew Moffat. 2017. A Survey of Quinn, Dennis P. 1997. The Correlates of Change in Interna-
Global Impacts of Climate Change: Replication, Survey Meth- tional Financial Regulation. American Political Science Review
ods and a Statistical Analysis. NBER Working Paper 23646, 91 (3): 53151.
National Bureau of Economic Research, Cambridge, MA. , and A. Maria Toyoda. 2008. Does Capital Account
Nordhaus, William D., and Paul Sztorc. 2013. DICE 2013R: Liberalization Lead to Growth? Review of Financial Studies
Introduction and Users Manual. www.dicemodel.net. 21 (3): 140349.
Raddatz, Claudio. 2009. The Wrath of God: Macroeconomic Strikes in the Central American and Caribbean Regions.
Costs of Natural Disasters. Policy Research Working Paper Journal of Development Economics 97 (1): 13041.
5039, World Bank, Washington, DC. Teulings, Coen N., and Nikolay Zubanov.2014. Is Economic
Ramcharan, Rodney. 2009. Does the Exchange Rate Regime Recovery a Myth? Robust Estimation of Impulse Responses.
Matter for Real Shocks? Evidence from Windstorms and Earth- Journal of Applied Econometrics 29 (3): 497514.
quakes. Journal of International Economics 73 (1): 3147. Thomas, Vinod, and Ramon Lopez. 2015. Global Increase in
Reinhart, Carmen, and Kenneth S. Rogoff. 2004. The Modern Climate-Related Disasters. ADB Economics Working Paper
History of Exchange Rate Arrangements: A Reinterpretation. 466, Asian Development Bank, Manila.
Quarterly Journal of Economics 119 (1): 148. Timmer, Marcel P., Gaaitzen de Vries, and Klaas de Vries. 2015.
Roston, Eric, and Blacki Migliozzi. 2015. Whats Really Warm- Patterns of Structural Change in Developing Countries. In
ing the World? Bloomberg Business Week, June 24. Routledge Handbook of Industry and Development, edited by
Schlenker, Wolfram, and David B. Lobell. 2010. Robust John Weiss and Michael Tribe, 6583. London: Routledge.
Negative Impacts of Climate Change on African Agriculture. Tol, Richard S. 2009. The Economic Effects of Climate
Environmental Research Letters 5 (1). Change. Journal of Economic Perspectives 23 (2): 2951.
Schlenker, Wolfram, and Michael J. Roberts. 2009. Nonlinear . 2014. Correction and Update: The Economic Effects
Temperature Effects Indicate Severe Damages to US Crop of Climate Change. Journal of Economic Perspectives28
Yields under Climate Change. Proceedings of the National (November): 22126.
Academy of Sciences 106 (37): 1559498. United Nations Environment Programme (UNEP). 2016. The
Schmidt-Traub, Guido. 2015. Investment Needs to Achieve Adaptation Finance Gap Report 2016. United National Envi-
the Sustainable Development Goals. SDSN Working Paper, ronment Programme (UNEP), Nairobi, Kenya.
United Nations, New York. Usery, E. Lynn, Jinmu Choi, and Michael P. Finn. 2007.
Seppnen, Olli, William J. Fisk, and David Faulkner.2003. Modeling Sea-Level Rise Effects on Population Using Global
Cost Benefit Analysis of the Night-Time Ventila- Elevation and Land-Cover Data. US Geological Survey, Cen-
tive Cooling in Office Building. Lawrence Berkeley ter of Excellence for Geospatial Information Science (CEGIS)
National Laboratory, Berkeley, CA. Working Paper.
Seppnen, Olli, William J. Fisk, and Q. H. Lei. 2006. Room . 2009. Modeling Sea-Level Rise and Surge in Low-Lying
Temperature and Productivity in Office Work. LBNL-60952. Urban Areas Using Spatial Data, Geographic Information
Lawrence Berkeley National Laboratory, Berkeley, CA. Systems, and Animation Methods. In Geospatial Techniques in
Shakun, Jeremy D., Peter U. Clark, Feng He, Shaun A. Marcott, Urban Hazard and Disaster Analysis, edited by Pamela Showal-
Alan C. Mix, Zhengyu Liu, Bette Otto-Bliesner, Andreas ter and Yongmei Lu, 1130. Dordrecht: Springer.
Schmittner, and Edouard Bard. 2012. Global Warming Pre- Von Peter, Goetz, Sebastian V. Dahlen, and Sweta Saxena. 2012.
ceded by Increasing Carbon Dioxide Concentrations during Unmitigated Disasters? New Evidence on the Macroeco-
the Last Deglaciation. Nature 484: 4954. nomic Cost of Natural Catastrophes. BIS Working Paper
Somanathan, E., Rohini Somanathan, Anant Sudarshan, and 394, Bank for International Settlements, Basel.
Meenu Tewari. 2017. The Impact of Temperature on Pro- Wang, Sun Ling, Eldon Ball, Richard Nehring, Ryan Williams,
ductivity and Labor Supply: Evidence from Indian Manufac- and Truong Chau. 2017. Impacts of Climate Change and
turing. Unpublished. Extreme Weather on US Agricultural Productivity. NBER
Stern, Nicholas Herbert. 2007. The Economics of Climate Change: Working Paper 23533, National Bureau of Economic
The Stern Review. Cambridge University Press. Research, Cambridge, MA.
. 2015. Why Are We Waiting? The Logic, Urgency, Wargocki, Pawel, and David P. Wyon. 2007. The Effects of
and Promise of Tackling Climate Change. Cambridge, Moderately Raised Classroom Temperatures and Classroom
MA: MIT Press. Ventilation Rate on the Performance of Schoolwork by
, and Joseph Stiglitz. 2017. Report of Children (RP-1257). HVAC&R Research 13 (2): 193220.
the High-Level Commission on Carbon Pric- Weitzman, Martin. 2011. Fat-Tailed Uncertainty in the Eco-
ing. https://static1.squarespace.com/static/ nomics of Catastrophic Climate Change. Review of Environ-
54ff9c5ce4b0a53decccfb4c/t/59244eed17bffc0ac256cf16/ mental Economics and Policy 5 (2): 27592.
1495551740633/CarbonPricing_Final_May29.pdf. World Bank.2013. Turn Down the Heat: Climate Extremes, Regional
Stock, James H., and Mark W. Watson.2007. Why Has Impacts, and the Case for Resilience. A report for the World Bank
USInflation Become Harder to Forecast? Journal of Money, by the Potsdam Institute for Climate Impact Research and
Credit and Banking 39 (S1): 333. Climate Analytics. Washington, DC: World Bank.
Strobl, Eric. 2012. The Economic Growth Impact of Natural , Ecofys, Vivid Economics. 2016. State and Trends of
Disasters in Developing Countries: Evidence from Hurricane Carbon Pricing 2016. Washington, DC: World Bank.
Positive cross-country spillovers from collective fiscal space. This was not least because excess capacity and
action by the worlds largest economies helped speed the low interest rates would help limit crowding out
recovery from the global financial crisis nearly a decade of private spending and the expected positive spill-
ago. But do fiscal spillovers still matter today? The answer overs would make collective efforts to boost activity
is yesbut the extent depends on circumstances in both more effective.
the countries that generate fiscal shocks and in those that More recently, the global effects of fiscal policy have
are recipients of the shocks. This chapter combines new been discussed amid possible changes in the macro-
empirical research and model-based simulations to show economic policy mix in Japan and the United States.
that fiscal spillovers tend to be low when a fiscal shock Debate is also ongoing about the role of fiscal policies
originates from a country without output gaps, but the in addressing excess external imbalances, includ-
impact intensifies when a source or recipient country is in ing whether euro area countries with excess current
recession and/or benefiting from accommodative mone- account surpluses should raise fiscal spending, which
tary policywhich suggests that spillovers are large when could also support growth in the currency union.
domestic multipliers are also large. The chapter also finds Recent improvements in economic conditions in
that spillovers from government spending shocks are larger many countries and their implications for monetary
than those associated with tax shocks, that the transmis- policy raise questions about the size of potential
sion of fiscal shocks may be stronger among countries with spillovers from fiscal stimulus today. Cyclical positions
fixed exchange rates, and that fiscal spillovers impact the have improved across the board over the past few years,
external positions of source and recipient countries alike. although with differences across countries (Figure4.1).
Model-based simulations suggest that the cross-border For example, the United States is operating at close to
effects of budget-neutral fiscal reforms are generally mod- full employment and, as a result, the Federal Reserve
est, though large reforms can trigger spillovers, especially if has begun to normalize monetary policy conditions.
they affect cross-border investment decisions. Overall, this At the same time, although euro area economies
evidence draws attention to the cross-border repercussions and Japan are experiencing an encouraging cyclical
of corporate tax reform in the United States, for example, recovery, output gaps remain negative in many of these
or of an increase in public investment in Germany. countries and core inflation is stubbornly low, prompt-
ing monetary authorities to commit to accommoda-
tive policies for an extended period. As the chapter
discusses, cyclical conditions and the associated ability
Introduction or willingness of monetary policy to act, both in coun-
What is the potential for fiscal policy to affect tries emitting and receiving the fiscal shock, are key
macroeconomic outcomes in other economies through determinants of the magnitude of its impact.1 Consid-
cross-border spillovers? This question has been at the erations regarding fiscal space in source countries are
center of the policy debate, especially in the aftermath also relevantif term premiums increase and financial
of the global financial crisis, when many countries conditions tighten following a fiscal stimulus, spillovers
experienced persistent economic slack, and monetary could be smaller.
policy interest rates approached the effective lower Against this backdrop, the chapter aims to answer
bound. Fiscal stimulus was then advocated widely, the following questions:
especially in major economies with sufficient fiscal
The authors of this chapter are Patrick Blagrave, Giang Ho, Ksenia 1Throughout the chapter, countries from which fiscal shocks
Koloskova, and Esteban Vesperoni (lead author), with support from originate are referred to as source or shock-emitting; coun-
Sung Eun Jung and contributions from Jared Bebee, Ben Hunt, tries affected by these shocks are referred to as recipient or
Adina Popescu, and Ippei Shibata. shock-receiving.
Figure 4.1. Output Gap in Selected Countries ology of Blanchard and Perotti (2002). Information
(Percent)
from the five source-country shocks is combined using
2009 2017 the strength of trade links with a range of advanced
and emerging market recipient countries to assess
1 global spillovers.
0 To analyze the role that economic slack, constraints
on monetary policy, and exchange rate regimes play in
1 transmission, the chapter uses an econometric frame-
2
work that can flexibly test for the presence of nonlinear
effects. Model-based simulations then help to illustrate
3 the complex cross-border transmission channels of
fiscal shocks. This approach offers insights into poten-
4
tial changes in the external positions of source and
5 recipient countries, as well as the dynamic behavior of
key macroeconomic variables, and elucidates spillovers
6
from different types of fiscal reforms.
7 The chapters findings add to the existing empirical
literature on fiscal spillovers by expanding the scope
8
of the analysis. Previous empirical studies focus on a
JPN
DEU
CAN
GBR
KOR
ESP
AUS
TUR
ARG
MEX
RUS
ZAF
IND
IDN
CHN
USA
ITA
FRA
BRA
Relatively weak cyclical positions imply larger spillovers. especially if stimulus takes the form of tax policy mea-
Although modest in normal times, spillovers are sures. In the euro areawhere there is fiscal space in
larger when cyclical conditions are weak, likely due some countriesstimulus could have larger spillovers.
to the reduced crowding-out effects of public spend- This is in the context of prospects for continued mone-
ing on private sector activity. tary policy accommodation and still-significant slack in
Monetary policy constraints also increase spillovers. some recipient countries.
When monetary policy in either source or recipient The impact on external imbalances would also
countries does not counteract fiscal shocksfor depend on the source of fiscal stimulus, as stimulus
example, because the effective lower bound is in the United States is likely to increase imbalances,
bindingspillovers are much larger than during whereas stimulus in some surplus euro area coun-
normal times. tries could reduce them. Where countries are con-
Currency pegs between source and recipient coun- sidering significant reductions in corporate income
tries may amplify spillovers. There is some evidence tax rates, the analysis suggests associated changes in
suggesting that fiscal shocks tend to have larger spill- investment-location and profit-reporting decisions
overs on recipient countries with currencies pegged by multinational corporations could have significant
to the source countrys currency than on those with negative spillovers on activity and the fiscal position of
flexible exchange rates. nonreforming countries.
Fiscal policy can change external positions in source
and recipient countries. Trade balances deteriorate in
source countries following a fiscal expansion, with Spillovers from Fiscal PolicyA
a consequent improvement in recipients exter- Conceptual Framework
nal positions. The cross-border impact of fiscal policy changes
An increase in term premiums may dampen spillovers. in a given country depends on their initial domestic
If fiscal stimulus at the source increases the term effects and the transmission mechanisms of shocks.
premiumfor instance, because of concerns about This means that factors affecting the sources domestic
debt sustainabilityspillovers are somewhat lower fiscal multiplier are relevant for determining spillovers
compared with a constant term premium scenario. on recipient countries. The fiscal shock is propagated
Under some circumstances, fiscal reforms come through different channelsprimarily associated with
with spillovers as well. Most budget-neutral fiscal trade linkswith the final impact also depending on
reforms have limited cross-border effects, although the economic and policy conditions in the recipient
large reforms can generate significant spillovers. countries (Figure4.2). This section provides a brief
For example, a reform that substantially reduces overview of the domestic impact of fiscal shocks, out-
corporate income tax rates and is offset by higher lines their possible transmission channels, and discusses
consumption taxes in major economies can have the factors affecting transmission.
repercussions in the rest of the world, including
through higher global interest rates and cross-border
reallocation of investment and profits. Domestic Impact of a Fiscal Shock
A large body of literature on domestic fiscal multi-
These results point to several important policy pliers suggests that cyclical and policy conditions play
lessons that are relevant now. Although fiscal space is a role in the response of a domestic economy to fiscal
currently more limited, and improved cyclical condi- shocks. In general, multiplier estimates vary signifi-
tions in many countries mean that spillovers from fis- cantly across countries, sample periods, and method-
cal policy are likely to be lower than during the global ologies. While a comprehensive summary is beyond
financial crisis, the analysis suggests that fiscal stimulus the scope of this chapter (see, for example, Batini
in major economies can nonetheless be important in and others 2014), dynamic stochastic general equi-
lifting economic activity abroad, although not every- librium and structural vector autoregression models
where. For example, given the cyclical position and developed since the early 1990s suggest that the size
gradually less accommodative monetary policy condi- of multipliers tends to be modest (between zero and
tions in the United States, a US fiscal stimulus would one over the first year) in normal timesgenerally
likely have relatively modest cross-border spillovers, understood as circumstances in which the economy
Transmission
Channels
Trade
Expenditure shifting
Expenditure switching
SOURCE
RECIPIENT
Fiscal Shock
Interest rates Term premium
Exchange rates Equity prices
Financial
does not have a significant output gapand depends revenue measuresalso influences the size of the
on a number of structural characteristics, including a domestic multiplier. Many studies have found that, for
countrys trade openness, exchange rate regime, labor advanced economies, short-term spending multipliers
market rigidities, and size of public debt.2 Outside tend to be larger than revenue multipliers (for exam-
normal times, multipliers can vary with the state of the ple, see a survey in Mineshima, Poplawski-Ribeiro,
business cycle (generally larger in a downturn than in and Weber 2014). This has been explained using
an expansion, although the empirical evidence is not traditional Keynesian theoryfor example, while an
conclusive) or the degree of monetary accommodation additional dollar of government spending contrib-
(larger when monetary policy is unresponsive, such as utes directly to higher aggregate demand, a dollar of
at the effective lower bound).3 All else equal, a larger tax cuts can be either spent or saved by firms and/
domestic multiplier should be associated with greater or households (that is, the marginal propensity to
cross-border spillovers. consume can be less than one). Recent empirical evi-
The composition of the fiscal intervention dence using the narrative approach has found some-
whether it is based on government spending or what larger tax multipliers than spending multipliers,
although narrative-based evidence on the latter is pri-
2For example, see Cole and Ohanian (2004); Kirchner, Cima- marily limited to defense-related spending.4 Yet other
domo, and Hauptmeier (2010); Corsetti, Meier, and Mller (2012); studies suggest that the relative magnitude of the
Gorodnichenko, Mendoza, and Tesar (2012); Born, Juessen,
and Mller (2013); and Ilzetzki, Mendoza, and Vegh (2013). A
multiplier of one would suggest that a change in the fiscal balance 4The narrative method, pioneered by Romer and Romer (2010),
translatesdollar for dollarinto a similar change in GDP. makes use of narrative records, such as budget documents and
3For example, see Erceg and Lind (2010); Christiano, Eichen- speeches, to identify the size, timing, and principal motivation for
baum, and Rebelo (2011); Eggertsson (2011); Woodford (2011); fiscal actions. The Romer and Romer (2010) data set also divides fis-
Auerbach and Gorodnichenko (2012a, 2012b); Owyang, Ramey, and cal policy changes into those made for reasons related to prospective
Zubairy (2013); Nakamura and Steinsson (2014); Riera-Crichton, economic conditions and discretionary actions (for example, actions
Vegh, and Vuletin (2015); Blanchard, Erceg, and Lind (2016); and aimed at reducing public debt), thereby allowing for a causal analysis
Canzoneri and others (2016). However, Ramey and Zubairy (forth- of the impact of fiscal policy on output. See also Ramey (2011);
coming) found little evidence of state dependence of the government Cloyne (2013); Mertens and Ravn (2013); and Guajardo, Leigh, and
spending multiplier based on historical data from the United States. Pescatori (2014).
spending and revenue multipliers may differ between through changes in global financial conditions. A fiscal
consolidation and expansion episodes and among policy change in a large economy can impact global
different degrees of monetary accommodation.5 interest rates, exchange rates, and the slope of the
yield curvethe latter stemming from any perceived
or actual impact of the policy change on long-term
Channels of Cross-Border Transmission fiscal sustainability in the source country. The financial
In standard open-economy macroeconomic models, channel can work in the opposite direction to the trade
a fiscal shock is transmitted abroad primarily through channel. For example, the higher interest rates and
the trade channel, which consists of two effects:6 exchange rate appreciation associated with an expan-
Expenditure shifting (sometimes referred to as sionary fiscal shock in the source country can increase
leakages) refers to the direct impact of a fis- the cost of foreign currency borrowing and worsen
cal policy change on the source countrys import the balance sheets of corporations and households in
demand through changes in domestic consumption recipient countries if there are currency mismatches,
and investment, which affects trading partners. generating negative spillovers. Equity prices may also
Here, the marginal propensity to import by both adjust, with cross-border repercussions.
the public and private sectors plays a key roleif Overall, the relative strength of each transmission
most spending changes are in nontradable sectors channel will depend on the extent of trade and finan-
and do not translate into a higher or lower level of cial linkages between the source and recipient coun-
imports, spillovers from expenditure shifting may be tries. Thus, the net spillover impact of a fiscal shock is
smaller. Larger and more open economies tend to an empirical question.
import more, suggesting that fiscal policy changes in
these countries will have larger spillovers on others
through the expenditure shifting channel. Factors Affecting the Transmission
Expenditure switching refers to the impact of a Like the domestic fiscal multiplier, cross-border
fiscal shock operating through changes in the spillovers from fiscal actions tend to vary with eco-
real exchange rate, which can trigger substitution nomic circumstances. Two factors play particularly
between domestic and foreign goods consumption. important roles:
For example, in a Mundell-Fleming-Dornbusch Cyclical position: The domestic multiplierand
framework, fiscal expansion puts upward pressure on hence spillovers through expenditure shifting
interest rates, the nominal exchange rate appreciates may be larger when the source country has more
in the source country, and domestic prices increase.7 economic slack. For example, a fiscal stimulus that
The resulting real appreciation boosts import boosts public employment would be more likely to
demand as foreign goods become cheaper. This crowd out private employment when labor markets
effect will be more significant, especially in the short are tight (Michaillat 2014), resulting in smaller
term, when the nominal exchange rate is fully flexi- domestic and spillover impacts; the same logic
ble; where nominal exchange rates are fixed, relative applies to the case of fiscal tightening. Another
priceand hence real exchange rateadjustments possibility is that a fiscal stimulus relaxes borrow-
can take longer. Either way, expenditure switching ing constraints (which tend to be tighter during
effects imply that a fiscal shock can have nontrivial a downturn), for example, by raising the value of
cross-border spillovers, even if its domestic impact collateralizable assets along with demand, helping
is muted, because the boost to import demand can to increase credit and investment (Canzoneri and
occur without an increase in domestic income. others 2016). Somewhat similarly, if the recipient
country is operating close to full capacity when an
In addition to the trade channel, the response of external fiscal shock hits, greater demand in tradable
financial variables to a fiscal shock can trigger spillovers sectors may crowd out activity in the rest of the
economy, resulting in a more muted impact on
5For example, see Eggertsson (2011); and Erceg and Lind (2013). overall economic activity.
6For example, see Fleming (1962); Mundell (1963); Dornbusch Monetary policy constraints: Whether monetary
(1976); and Obstfeld and Rogoff (1995).
7Notice that other frameworks can deliver different exchange rate policy accommodates the fiscal shock matters, and
predictions (see Obstfeld and Rogoff 1995). it is relevant for both source and recipient countries.
Figure 4.3. Tracking Tax Shocks in the United States links between the source and the recipient.13 The
(Percent of GDP) baseline specification controls for factors that affect the
normal short-term dynamics of output in the recipient
Blanchard-Perotti method Narrative method (Romer-Romer)
country, such as past growth rates and external demand
developments. The specification is estimated using
1.0 (1) Tax hikes to put social security
system on a sounder footing quarterly data from the first quarter of 2000 through
the second quarter of 2016, and the sample of 55
(4) 1993 Omnibus Budget
Reconciliation Act advanced and emerging market economies represents
0.5
almost 85percent of world output. Thus, the panel
estimation gives spillover estimates for an average
0.0 country in the sample.14 For the panel estimation, the
shocks are expressed as a share of recipient countries
output to facilitate aggregation across sources. For ease
0.5 (3) 1986 Tax (6) Expiration of 2003 tax of interpretation of the economic magnitude, results
Reform Act relief are presented with shocks normalized to an average
1percent of GDP change in the fiscal position across
1.0 (5) Bush tax cuts source countries (see details in Annex4.2, which shows
(2) Reagan tax cuts how panel results are rescaled using relative GDP levels
and trade links).
1.5
1980 85 90 95 2000 05 10 15
Sources: Romer and Romer (2010); and IMF staff calculations. Spillovers on Economic Activity
The results point to significant spillover effects from
fiscal policy, especially from government spending
were put in place following the Greenspan Commis- shocks. Figure4.4 shows the estimated response to
sions recommendations to shore up financing of the a foreign fiscal shock of an average recipient coun-
social security system (Figure4.3).11 trys output over eight quarters. A shock to the fiscal
The structural shocks also have a statistically and balancehenceforth referred to as the overall fiscal
economically significant domestic impact. Consistent shockis constructed as a shock to government spend-
with traditional Keynesian theory and previous empir- ing minus a shock to tax revenues, such that a positive
ical work that uses a similar methodology, estimates of shock implies a reduction in the source countrys fiscal
domestic multipliers using the structural shocks tend balance (or an increase in the deficit). An overall fiscal
to be larger for spending instruments (slightly above shock would increase recipient output on impact,
one) than for tax instruments (slightly below one). reaching a peak around the third quarter after the
Some differences are seen in the size of domestic tax shock before starting to dissipate (Figure4.4, panel
multipliers across the five source countries, with the 1). Estimations for specific fiscal instruments show
multiplier of the United States being larger than that that spillovers from a government spending shock are
of European peers or Japan, possibly reflecting different larger, more persistent, and more precisely estimated
tax structures and the specific tax instruments used
13The use of trade links to weight the shock is instrumental to
(Blagrave and others, forthcoming).
obtaining country-specific external fiscal shocks, but it does not
The spillover effects from the fiscal shocks are preclude spillovers through channels other than trade given that the
estimated using the local projections method.12 The estimates capture the overall response of recipient-country GDP
econometric specification relates an economic outcome regardless of the channel of transmission. Combining shocks from
several source countries is important to use the variability emanating
in a recipient country, such as the level of output,
from different sources, given that trade patterns differ. In particu-
to a fiscal shock from the five source countries lar, while some source countriessuch as the United Statescan
constructed by pooling together shocks from source have a global impact, the impact of others is more regional; for
countries and weighting them by the strength of trade example, Germanys and Frances trading partners are more concen-
trated in Europe.
14More details about the data and empirical methodology are
11See Blagrave and others (forthcoming) for more examples. provided in Annexes 4.1 and 4.2, respectively, as well as in Blagrave
12See Jord (2005). and others, forthcoming.
Figure 4.4. Dynamic Responses of Recipient Countries examination of spillovers from specific spending or
Output to Fiscal Shocks tax instruments, such as government consumption
(Impact on output level, percent; quarters on x-axis)
or investmentan issue assessed later in the chapter
through model-based simulations.
0.2 1. Overall Fiscal Shock
Spillovers are economically significant and in line
with earlier estimates. For example, a 1percent of
GDP overall fiscal shock in an average major advanced
0.1
economy would raise output in the average recipient
country by about 0.08percent over the first year. For
0.0 a government spending increase of the same magni-
tude, the average spillover impact in recipient countries
increases to 0.15percent over the first year; for a tax
0.1 hike of similar size, output falls by about 0.05percent
0 1 2 3 4 5 6 7 8
(Figure4.5). As expected, spillovers from fiscal shocks
2. Spending Shock
are substantially lower than domestic fiscal multipliers
0.6
in source countries, but still relevant.16 These are of the
same order of magnitude as those found in previous
0.4
workfor example, Beetsma, Klaassen, and Wieland
(2006)although differences in country and time
0.2
samples as well as shock identification make a direct
comparison challenging.17 While the spillover estimates
0.0
in this section are averages across different economic and
policy conditions, subsequent analysis also shows that
0.2 there is a large difference between estimates in normal
0 1 2 3 4 5 6 7 8
times and those in times of economic slack, for example.
0.2 3. Tax Shock Further analysis of components of recipient-country
output corroborates the importance of trade for the
0.1 transmission of fiscal shocks (Figure4.6), consistent
with the conceptual framework outlined above. In par-
0.0 ticular, a positive fiscal shock from abroad is estimated
to raise recipient-country bilateral exports to the source
0.1
countries. With higher export demand, firms expand
investment to build production capacity, generating a
second-round effect on recipient-country investment,
0.2
0 1 2 3 4 5 6 7 8 whereas the impact on consumption appears negligible.
The boost to exports and investment increases imports,
Source: IMF staff calculations. some of which come from source countries. With
Note: t = 0 is the quarter of the respective shocks. Solid lines denote point
estimates and dashed lines denote 90 percent condence bands. Shocks are bilateral imports rising by much less than bilateral
normalized to an average of 1 percent of GDP across source countries. exports, however, the recipients trade balance with the
source countries improves following the fiscal shock.
than those from a tax shock of equal size (Figure4.4, 16As discussed earlier, fiscal shocks in the chapter yield domestic
panels 2 and 3).15 This is consistent with the evidence spending multipliers slightly above one and tax multipliers slightly
below one, on average, across the source countries.
pointing to larger domestic spending multipliers 17Beetsma and others (2006) find that a 1percent of German
than domestic tax multipliersas discussed earlier. (French) GDP shock to government spending results in a European
Data constraints prevent a more detailed empirical GDP response of about 0.14 (0.08) percent after two years. For a
tax shock, spillovers are about 0.05 (0.03) percent. Compared
with studies that express shocks in units of recipient-country GDP
15These effects are assumed to be symmetric during fiscal expan- (Auerbach and Gorodnichenko 2013; Goujard 2017), estimates are
sions and consolidationsthe panel analysis cannot disentangle a also broadly similar. A detailed comparison to the literature is
potential asymmetry from different policy actions. provided in Blagrave and others, forthcoming.
Figure 4.5. Spillovers of Fiscal Shocks on Recipient Figure 4.6. Dynamic Responses of Components of Recipient
Countries Output Countries Output to a Fiscal Shock
(One-year average impact on output; percent) (Percent of output; quarters on x-axis)
0.2 0.2
0.15
0.1 0.1
0.10
0.0 0.0
0.1 0.1
0.05
0.2 0.2
0 1 2 3 4 5 6 7 8 0 1 2 3 4 5 6 7 8
0.00
0.0 0.0
Source: IMF staff calculations.
Note: Shocks are normalized to an average of 1 percent of GDP across source
countries. 0.1 0.1
0.2 0.2
0 1 2 3 4 5 6 7 8 0 1 2 3 4 5 6 7 8
Figure 4.7. Spillovers under Various Economic and Policy they pertain to the source countryas well as their
Conditions cross-border transmission. In general, a larger impact
(One-year average impact on output; percent)
in the source country is expected to give rise to more
significant spillovers.
0.20 1. Conditions in Recipient Countries
Cyclical Position and Monetary Policy Constraints
0.15 To test how cyclical positions and monetary policy
affect the impact of fiscal shocks, the baseline econometric
0.10 framework is augmented to allow for potential regime
dependence (see Annex4.2 for details). The definitions of
0.05 regimes are based on the prevailing output gap or the level
of the short-term interest rate in either source or recipient
0.00 countries. Specifically, a negative output gap is assumed
Baseline No slack Slack No ELB ELB
to represent economic slack, and a short-term interest rate
below the 25th percentile of the relevant cross-country
0.20 2. Conditions in Source Countries distribution is a proxy for monetary policy constrained by
the effective lower bound.20 Results are robust to using
0.15 alternative definitions of slack, including the unemploy-
ment gap or smooth-transition probability as in Auerbach
0.10 and Gorodnichenko (2013). For the effective lower
bound, results are also robust to using an absolute interest
0.05 rate threshold that is common to all countries.
Consistent with theory and empirical findings in the
0.00 domestic multiplier literature, spillovers are estimated to
Baseline No slack Slack No ELB ELB be larger during episodes of economic slack than in nor-
mal times. For example, if the recipient country has slack
0.20 3. Summary when the external fiscal shock hits, its output would rise
by 0.11percent over the first year in response to a 1per-
0.15 cent of GDP overall fiscal shock in an average major
advanced economy. By contrast, the response to the same
0.10 shock would be almost halvedto 0.06percentwhen
there is no economic slack (Figure4.7, panel 1). Differ-
0.05 ential effects are also observed when the source econ-
omy has slack, compared with when it does notwith
0.00 estimates varying between 0.09percent and 0.03percent,
Baseline Normal times1 Slack2 ELB2
respectively (Figure4.7, panel 2).
Spillovers can be even larger when monetary policy
Source: IMF staff calculations.
Note: ELB = effective lower bound. Slack is dened as output gap below zero; and
is constrained by the effective lower bound, either in
ELB corresponds to short-term interest rates in the bottom 25 percent of cross- the source or the recipient country (Figure4.7, panels
country historical distribution. Responses to an overall scal shock are presented. 1 and 2). For example, subject to a 1percent of GDP
Shocks are normalized to an average of 1 percent of GDP across source countries.
1
Normal times refer to average of no slack and no ELB in both source and recipient overall fiscal shock in an average major advanced econ-
countries. omy, the response of recipient-country output can be
2
Average estimates for conditions in source countries and conditions in recipient
countries. more than four times greater when its interest rate is
exceptionally lower than in normal times.21 Monetary
market economies.
21These resultsfor both slack and effective lower bound cases,
policy constraints in source countries have a similar Figure 4.8. Dynamic Reponses of Components of Recipient
effect on spillovers, as they can amplify the domestic Countries Output under Normal Times and Effective Lower
impact of fiscal shocks. Although slack and the effec- Bound in Recipient Countries
(Percent of output; quarters on x-axis)
tive lower bound have distinct mechanisms to amplify
spillovers, it is often difficult to clearly distinguish the
0.3 1. Consumption under 2. Consumption under 0.3
two states in empirical estimation because they can Normal Times Effective Lower
coincide in practice, as has occurred in recent years.22 0.2 Bound 0.2
This caveat should be kept in mind when interpreting 0.1 0.1
the results.
0.0 0.0
The response of GDP components under monetary
policy constraints offers further insights into how 0.1 0.1
a fiscal shock is transmitted to recipient countries 0.2 0.2
(Figure4.8). Faced with a positive fiscal shock from 0 2 4 6 8 0 2 4 6 8
abroad, consumptionand particularly investment
in a recipient country responds much more strongly 0.5 3. Investment under 4. Investment under Effective 0.5
Normal Times Lower Bound
when the domestic nominal interest rate is close to 0.4 0.4
the effective lower bound, likely reflecting declining 0.3 0.3
real interest rates associated with higher expected 0.2 0.2
inflation. This is consistent with results from the-
0.1 0.1
oretical models (see section on factors affecting
0.0 0.0
transmission) and is confirmed by the results of
the model-based simulations presented in the next 0.1 0.1
0 2 4 6 8 0 2 4 6 8
section. The responses of exports to and imports from
the source countries are also stronger when monetary 0.8 5. Bilateral Exports under 6. Bilateral Exports under 0.8
policy accommodates the fiscal shock, in line with the Normal Times Effective Lower Bound
domestic response of investment. 0.6 0.6
0.4 0.4
Exchange Rate Regime
0.2 0.2
As discussed in the section on factors affecting
transmission, the exchange rate regime can also 0.0 0.0
impact the size of fiscal spillovers. To investigate this 0.2 0.2
question, this section analyzes whether the impact 0 2 4 6 8 0 2 4 6 8
of a fiscal shock in the United States varies for
recipient countries with fixed and flexible exchange 0.5 7. Bilateral Imports under 8. Bilateral Imports under 0.5
0.4 Normal Times Effective Lower Bound 0.4
rate regimes vis--vis the US dollar. The United
0.3 0.3
Stateswith its global currency and systemic trade
0.2 0.2
importanceis a suitable source country for this 0.1 0.1
exercise. Countries do not typically peg to the British 0.0 0.0
pound or the Japanese yen. In the case of the euro, 0.1 0.1
Germanys and Frances trade importance is mostly 0.2 0.2
within Europe, where most sample countries are 0.3 0.3
0 2 4 6 8 0 2 4 6 8
Figure 4.9. Dynamic Reponses of Recipient Countries Output Exchange Restrictions (IMF classification).23 More
to US Spending Shock under Various Exchange Rate Regimes details are provided in Annex4.1.
(Impact on output, percent; quarters on x-axis)
The evidence suggests that a government spending
shock in the United States generates stronger and
1.2 1. Fixed Regime: 2. Flexible Regime: 1.2
Reinhart-Rogoff Reinhart-Rogoff more persistent impacts on countries whose exchange
1.0 Classication Classication 1.0 rates are pegged to the US dollar than on those whose
0.8 0.8 exchange rates are more flexible (Figure4.9). This is
0.6 0.6 the case regardless of which exchange regime classifi-
cation is used. The difference in the output responses
0.4 0.4
between fixed and flexible regimes is statistically sig-
0.2 0.2
nificant on impact under both classifications and also
0.0 0.0 during the second year under the Reinhart-Rogoff clas-
0.2 0.2 sification. At the same time, no difference in spillovers
0 2 4 6 8 0 2 4 6 8
is observed between fixed and flexible regimes from an
overall fiscal shock or a tax shock (not shown). Taken
1.2 3. Fixed Regime: 4. Flexible Regime: 1.2
IMF Classication IMF Classication at face value, this result seems to point to relatively
1.0 1.0
weak expenditure switching effects in the transmission
0.8 0.8 of spending shocks. This weakness could reflect that,
0.6 0.6 for a significant portion of the sample, US monetary
0.4 0.4 policy was constrained by the effective lower bound,
limiting interest rate and hence exchange rate move-
0.2 0.2
ments. Another possibility is that, as discussed earlier,
0.0 0.0
trade integration may be stronger under pegsbeyond
0.2 0.2 what can be captured by the simple import ratios used
0 2 4 6 8 0 2 4 6 8
in weighting the shocks.
Source: IMF staff calculations.
Note: t = 0 is the quarter of the shock. Solid red lines denote point estimates
conditional on exchange rate regime; dashed red lines denote 90 percent The Transmission of Fiscal Shocks
condence bands; and solid blue lines represent the unconditional estimates.
Shocks are normalized to an average of 1 percent of GDP across source countries
Model-Based Analysis
(note that this represents a less than 1 percent of US GDP shock). To complement the empirical analysis, the chapter
presents model-based simulations using a multiregion
general equilibrium modelthe IMFs G20 Model.
The model simulations are intended to be illustrative
and offer further insights into the macroeconomic
either euro area members or peg to the euro, not adjustment to fiscal shocksincluding the response
allowing for enough variation in the data to identify of exchange and interest ratesand more granularity
the effect for those with flexible regimes. on the impacts of various fiscal instruments. Overall,
The empirical framework is again modified to simulations serve as theory-based cross-checks on the
allow for regime dependence of the fiscal shocknow empirical results and provide insights into how fiscal
originating only in the United Stateswhere the shocks are propagated.24
regime definition is based on the prevailing bilat- The results are generally consistent with the empir-
eral exchange rate arrangement between the United ical findings in this chapter: simulations show that
States and the recipient country in a particular spillovers from temporary fiscal shocks can differ
period. Specifically, a fixed exchange rate regime
is defined as encompassing de facto pegs or crawl- 23In 2015, for example, the Reinhart-Rogoff classification has
ing pegs, classified using two alternative methods: more recipient countries classified as having fixed exchange rates
(1)Reinhart and Rogoff (2004) updated by Ilzetzki, compared with the IMF classification. The number of fixed-rate
Reinhart, and Rogoff (2017a, 2017b)henceforth countries varies over time. In general, there tend to be more fixed
exchange rate regimes in earlier years of the sample.
called Reinhart-Rogoff classification; and (2) the 24Additional details on the G20 Model are available in Andrle and
substantially depending on the monetary policy Figure 4.10. Impact of Fiscal Shocks on Global GDP Based on
response and the fiscal instruments used. In addition, Various Instruments
(Two-year average impact, percent)
the responses of GDP components under different
assumptions on monetary accommodation closely
0.20
resemble those identified empirically.25 In all cases, fis-
cal shocks are expressed as a share (generally 1percent) 0.18
of a particular source countrys GDPthis differs from United States
0.16
how results were presented in the empirical section France and Germany
and implies that, all else equal, shocks emanating from 0.14
larger countries will have larger spillover effects. 0.12
0.10
Spillovers on Output: Fiscal Instruments and Policy 0.08
Accommodation
0.06
The model simulations confirm substantial spill-
overs from government spending shocks. Specifically, 0.04
countries fiscal shocks. holds can avoid some of the burden by postponing consumption.
Figure 4.11. Spillovers from US Fiscal Shocks with and have the broadest global reachdue to the large size of
without Monetary Accommodation the US economy and its moderately strong trade links
(Two-year average impact on rest of the world GDP, percent)
with most regions (Figure4.12).30 Spillovers from the
United States are largest on countries in Latin Amer-
0.8
Normal response of monetary
ica and Canadaall of which account for significant
policy shares of US import demand. For shocks from France
Monetary accommodation in and Germany, spillovers are largest on Europe, given
United States only
0.6 Monetary accommodation in deep trade integration, but relatively small on other
recipient countries regions. Finally, fiscal measures in China have mean-
Monetary accommodation in
all countries ingful spillovers on each region due to the size and
openness of the Chinese economy. By region, spillovers
0.4
are slightly larger on countries in Asiagiven strong
trade linksthough spillovers on Europe, Canada,
and Latin America are not trivial. Chinas economy,
0.2 given its growing global clout, is playing an important
role in driving spillovers onto neighboring countries
through the trade channel and the impact of fluctua-
tions in demand on commodity prices (IMF 2016).
0.0
Government spending Tax revenue
shocks. Spillovers from stimulus in the United States lations closely resembles those implied by the empirical analysis pre-
sented earlier. See Blagrave and others (forthcoming) for more details.
31Several studies estimating fiscal shocks in structural vector
28In the G20 Model, monetary policy in countries with flexible autoregression models find that increases in government spend-
exchange rate regimes responds to an increase in expected future ing trigger exchange rate depreciationssee, for example, Corsetti
inflation by increasing nominal interest rates to reduce demand and and Mller (2006); Kim and Roubini (2008); Monacelli and
return inflation to target. Perotti (2010); Enders, Mller, and Scholl (2011); and Ravn,
29Spillovers are even larger under the full accommodation Schmitt-Groh, and Uribe (2012). This empirical result runs counter
scenariothey should be viewed as an upper bound, as such a to the predictions of the Mundell-Fleming-Dornbusch framework,
scenario would require an exceptional coordinated accommodation although it is consistent with some new open-economy macroeco-
by monetary policy in all countries. nomic models (Obstfeld and Rogoff 1995).
Figure 4.12. Regional GDP Impact of Government Spending Figure 4.13. Dynamic Reponses to a US Government Spending
Shocks from the United States, Europe, and China Shock
(Two-year average impact, percent) (Percent deviation from baseline)
Figure 4.14. Spillovers from US Spending Shock with and conditionsand imports, due to stronger domestic
without a US Term-Premium Increase activity in recipients. The expenditure switching
(Percent deviation from baseline)
channel operates in the opposite direction under
monetary accommodation, with recipient coun-
Spending shock without Spending shock with US term-
US term-premium increase premium increase
tries real exchange rates appreciating against the
US dollar. This occurs because the negative impact
on US real interest rates is more pronounced than
0.8 1. US Real GDP
elsewhere. Recipients trade balances still improve
0.6 because of the strong increase in demand from the
United States. Overall, as shown in Figure4.11, the
0.4 cumulative effect on global GDP is amplified under
0.2
monetary accommodation.
as permanent budget-neutral shifts in the composition Figure 4.15. Spillovers from Corporate Income Tax Reduction
of the public sector budget. The scenarios considered so Financed by an Offsetting Increase in Value-Added Tax
(Percentage-point deviation from baseline, unless noted otherwise)
far in the chapter deal with temporary fiscal impulses
associated with a change in the fiscal stance in the
source country. However, budget-neutral fiscal reforms CIT/VAT reform direct impact only
CIT/VAT reform plus assumptions on investment shift
may also have spillover effects. To demonstrate these CIT/VAT reform plus assumptions on prot and investment shift
differences, the following two scenarios are considered:
(1)a budget-neutral corporate income tax reform and 5 1. Real GDP: Countries 2. Real GDP: 5
(2) a budget-neutral infrastructure spending increase. Pursuing CIT/VAT Reforms Rest of the World
4 (Percent deviation from (Percent deviation from 4
These illustrative scenarios suggest that fiscal reforms baseline) baseline)
3 3
have limited cross-border effects, though significant
changes can still generate large spillovers. 2 2
1 1
0 0
Budget-Neutral Corporate Income Tax Reform
1 1
The direct spillovers of a (simultaneous)
2 2
budget-neutral reduction in corporate income tax 2017 19 21 23 25 27 2017 19 21 23 25 27
rates in France, Germany, and the United Statesthe
source countries in this scenariowould be slightly 2 3. Government Debt to GDP: 4. Government Debt to GDP: 2
Countries Pursuing CIT/VAT Rest of the World
negative.33 The scenarios main assumptions are that 1 Reforms 1
corporate tax rates are reduced by 15percentage
points, consumption-tax rates rise to offset the revenue 0 0
competitive disadvantage with respect to their return pure fiscal revenue gain for source countries and a
to capital, and real interest rates are slightly higher corresponding loss for other countries.38
implying lower investment. This negative impact The results suggest that investment shifting and
more than offsets the small impetus to exports associ- profit shifting could trigger more significant spill-
ated with increased demand from source countries. overs on activity and affect fiscal positions. Activity
in source countries would be considerably higher
However, beyond this direct effect, fiscal reforms with GDP increasing by almost 4percent after 10
may also affect investment and profit-reporting deci- yearsalthough significantly reduced elsewhere, by
sions. As discussed in Devereux (2008) and De Mooij about 1percent. Corresponding changes in trade
and Ederveen (2008), corporate tax rates influence balances would imply a material deterioration for
both intensive and extensive (discrete or location) deci- corporate-tax-reforming countriesas import demand
sions of firms, suggesting that multinational companies rises significantlyand an improvement for the rest
may relocate operations when faced with significant of the world, due to import compression and export
changes in relative tax rates in different jurisdictions. growth. Both investment shifting and profit shifting
In addition, both studies note that it is feasible for can also have an impact on fiscal positions, improving
multinational companies to shift profits between the primary balance of source countries and under-
countries. In the scenario, the lower corporate income mining the balance of others, above and beyond the
tax rates prompt these firms to shift operations direct effects of the corporate income tax reform itself.
both investment and the jurisdiction in which profits The marginal impact of profit shifting on public debt
are reportedto source countries, to the detriment stocks can be seen by comparing the red and yellow
of recipients. lines in panels 3 and 4 of Figure4.15it is clear that
The effect of investment and profit shifting are illus- the impact of investment shifting (measured by com-
trated by the red (investment shifting only) and yellow paring the blue and the red lines) is much larger than
(investment and profit shifting together) lines of Fig- that of profit shifting.39
ure4.15. Based on estimates in the literature on profit
and investment shifting, the scenario assumes that
foreign direct investment in countries not pursuing Budget-Neutral Permanent Increase in
reforms could decline by about $400billionthis loss Public Investment
is assumed to be distributed equally across all coun- Compared with corporate income tax reforms that
tries as a share of GDP.36 By contrast, the countries trigger investment and profit shifting, a budget-neutral
pursuing reforms are assumed to benefit by a similar permanent increase in public investment would have
amount, above and beyond the immediate impact on very modest spillovers.40 The scenario assumes a per-
investment from the corporate income tax reduction cent of GDP increase in public investment in the five
discussed above.37 Profit shifting is assumed to be a large economies considered in the empirical exercise
France, Germany, Japan, the United Kingdom, and
the United Stateswhich is financed by an increase
36This is a simplifying assumption. Countries that currently bene-
fit from a significant corporate income tax gap relative to the source 38The assumed impact of profit shifting is derived by applying
countries, or those with a significant presence of multinational cor- an estimated semielasticity of profits with respect to the tax ratea
porations based in countries pursuing corporate income tax reforms, value of 2, taken from De Mooij and Ederveen (2008)to estimates
may be more adversely affected by investment shifting. of the share of multinational firms in each country, which is assumed
37The assumed impact of investment shifting is derived by apply- to be approximately 0.6 in Germany and France and 0.3 in the
ing an estimated semielasticity of the corporate tax base to tax rate United States, and to the corporate income tax rate reduction being
changes from De Mooij and Ederveen (2008)taken to be 3.2 considered (15percentage points). The same caveats mentioned for
to foreign direct investment inflows and outflow data for France, investment shifting regarding elasticities apply.
Germany, and the United States, which proxy the foreign portion of 39The impact on public debt in this scenario is only transitory,
the corporate tax base subject to relocation. Under a large corporate with all debt-to-GDP ratios returning to baseline in the long term.
income tax rate reduction, foreign direct investment inflows would The speed of adjustment back to baseline depends on assumptions
increase as foreign multinationals choose to locate more production regarding the aggressiveness of the models fiscal ruleother assump-
in the countries pursuing reforms, and outflows would decline as tions would lead to different adjustment dynamics.
domestic multinationals choose to develop more production capacity 40This result is broadly consistent with results reported in Bussire
domestically. It is important to note that semielasticities in the liter- and others (2017), who find that most budget-neutral fiscal reforms
ature vary widely and that the estimated investment-shifting impact do not have large cross-border trade spillovers, except in the case of
of corporate income tax reform is sensitive to these assumptions. coordinated reforms in periods of accommodative monetary policy.
in consumption taxes. Such a reform would boost Figure 4.16. Spillovers from Increase in Government
the capital stock in source countries, thereby increas- Investment in Five Major Economies
(Percent deviation from baseline)
ing output permanentlythe increase in investment
resulting from the higher productivity associated with
an expansion of the public capital stock outweighs the 1.2 1. Real GDP, Consumption, and Investment in Source Countries
negative impact on domestic consumption of higher 0.8 Real GDP Real consumption
consumption taxes. However, as shown in Figure4.16, Real investment
0.4
although there would be some modest cross-border
impact due to expenditure shifting, the impact would 0.0
be muted by an exchange rate depreciation in source
0.4
countries, implying that the expenditure switching
channel will eventually offset the positive effect.41 The 0.8
impact on recipient countries trade balances is small, 1.2
but negative. 2017 18 19 20 21 22 23 24 25 26 27
profit-reporting decisions of multinational firms, could rioration in the trade balance of the country where
have large spillovers. it occurs, with corresponding improvements in the
Finally, and not surprisingly, fiscal actions with positions of trading partners. This implies that a
economically meaningful cross-border effects can fiscal expansion in the United States could exacerbate
also impact trade balances. For example, the chapter global current account imbalances, while stimulus in
suggests that fiscal stimulus tends to lead to a dete- Germany would tend to reduce them.
Box 4.1. The Spillover Impact of US Government Spending Shocks on External Positions
Consensus on the effect of government spending Figure 4.1.1. Response of Recipient
shocks on a countrys exchange rate and external bal- Countries Trade Balance and Real Exchange
ance remains elusive in the empirical literature.1 This Rate vis--vis US Dollar
may stem partly from the difficulty of isolating agents (Quarters on x-axis)
anticipation of fiscal policies, given both legislative
and implementation lags, as highlighted by Ramey Full sample Before global nancial crisis
(2011), among others. This box and a related spillover
note (Popescu and Shibata, forthcoming) examine 1.0 1. Response of Trade Balance
the impact of fiscal spending shocks from the United (Percentage points)
0.8
States on the US trade balance and real exchange rate,
0.6
from both a multilateral and a bilateral perspective,
while carefully taking into consideration the issue of 0.4
fiscal foresight. 0.2
To capture anticipation effects, the approach follows
0.0
Forni and Gambetti (2016) and relies on professional
forecasters surveys to identify fiscal shocks at the 0.2
announcement rather than implementation date.2 0.4
0 2 4 6 8 10 12 14 16 18 20
Methodologically, the fiscal foresight (news) shock
is identified in a vector autoregression using US data
2. Response of Real Exchange Rate vis--vis
from the first quarter of 1981 through the fourth 8
US Dollar
quarter of 2016.3 The analysis further extends Forni 4 (Percent)
and Gambetti (2016) to a cross-country perspective
to account for recipients macroeconomic conditions, 0
which is the main unique contribution of this exercise. 4
The results suggest that news of future government
spending leads to a real appreciation of the US dollar 8
and deterioration of the US trade balancein line 12
with theory and solving the depreciation puzzle
found in most previous studies. As discussed in Forni 16
0 2 4 6 8 10 12 14 16 18 20
and Gambetti (2016), the key intuition is that the
inclusion of additional information on fiscal expec- Source: IMF staff calculations.
tations and forecasts improves the estimation of the Note: t = 0 is the quarter of the shock. Dashed lines denote
effects of fiscal spending shocks by capturing more 90 percent condence bands.
precisely the timing of the impact. The timing is likely
The authors of this box are Adina Popescu and Ippei Shibata. significant in assessing the response of fast-moving
1For example, while the theoretical literature tends to predict
Annex Table4.1.1. Data Sources for Quarterly Fiscal Data by Source Country
Country Fiscal Data Data Source Seasonal Adjustment Note
France Government spending Eurostat1 SWDA by source Sum of government final
consumption and GFCF
Tax revenue Eurostat1 SWDA by source Current taxes on income
and wealth, excluding
social contributions
Germany Government spending Deutsche Bundesbank SWDA by source Sum of government final
consumption and GFCF
Tax revenue Eurostat1 X-12-ARIMA by IMF staff
Japan Government spending Cabinet Office of Japan SAAR by source Sum of government final
consumption and GFCF
Government total revenue Ministry of Finance and X-12-ARIMA by IMF staff Extrapolated using Denton
Cabinet Office method
United Kingdom Government spending Office for National Statistics Seasonally adjusted by source Sum of government final
consumption and GFCF
Tax revenue Eurostat1 X-12-ARIMA by IMF staff
United States Government spending US Bureau of Economic Seasonally adjusted by source Sum of government final
Analysis consumption and GFCF
Tax revenue US Bureau of Economic Seasonally adjusted by source
Analysis
Source: IMF staff compilation.
Note: For government spending, nominal levels are deflated using the GDP deflator when real levels are not directly available from the source. For tax revenue
(total revenue for Japan), real levels are calculated by deflating nominal levels using each countrys GDP deflator. GFCF = gross fixed capital formation; SAAR
= seasonally adjusted and annualized data; SWDA = seasonally and working-day adjusted data; X-12-ARIMA = US Census Bureau software package for
seasonal adjustment.
1Quarterly nonfinancial accounts for general government database from Eurostat.
Real Consumption, Haver Analytics Rebased to 2010; deflated Vietnam Seasonally adjusted, annualized,
Investment, Exports, using respective deflators for in national currency; data from
Imports each country and variable national accounts
Bilateral Goods DOTS Average of values reported None in the sample Original data at monthly
Exports/Imports by the reporter and partner frequency, aggregated by sum
countries
External Demand WEO; DOTS; Haver Export-weighted sum of None in the sample Seasonally adjusted, quarter over
Analytics partner countries real GDP quarter growth, log difference,
growth percent
Short-Term Bloomberg Finance L.P.; Three-month LIBOR, Cyprus, Estonia, Policy rate, deposit rate, target
Monetary Policy Haver Analytics three-month Treasury bill Luxembourg, Slovak rate used where LIBOR and
Rate rate, where available Republic, Uruguay treasury bill rates were not
available
Output Gap WEO; Haver Analytics Gap between real output and None in the sample Denton method used to match
potential output estimated annual output gap numbers in
by HP filter WEO
Source: IMF staff compilation.
Note: DOTS = IMF, Direction of Trade Statistics; HP = Hodrick-Prescott; LIBOR = London interbank offered rate; WEO = World Economic Outlook.
Bilateral goods exports/imports: Bilateral weights are Output gap: The quarterly output gap is first calcu-
calculated using bilateral exports/imports of goods lated as the gap between real output and potential
between 55 countries in the sample and five source output, estimated by the Hodrick-Prescott filter.
countries (5 x 55 = 275pairs). For each country pair, Then, to reconcile any potential difference between
the average is that of reported values of both countries. the estimated output gap and the annual output gap
External demand: This is calculated as a weighted numbers published in the IMFs World Economic
sum of partner countries real growth based on bilat- Outlook (WEO), the Denton proportional bench-
eral export weights. marking method is used. This method both preserves
Short-term interest rate: The three-month London the seasonality observed from quarterly estimated
interbank offered rate (LIBOR) and three-month output gap series and matches the data published in
Treasury bill rate are used. For more comprehensive the WEO when converted to annual basis.
country and historical coverage, policy, deposit, and
target rates are used where three-month LIBOR and Variables with notable trends over the sample
Treasury bill data are not available. period are detrended using country-specific linear
Americas Argentina, Brazil, Canada, Chile, Colombia, Costa Rica, Mexico, Peru, United States,* Uruguay
Asia Australia, China, India, Indonesia, Japan,* Korea, Malaysia, New Zealand, Philippines, Thailand, Vietnam
Europe Austria, Belgium, Bulgaria, Croatia, Cyprus, Czech Republic, Denmark, Estonia, Finland, France,* Germany,*
Greece, Hungary, Ireland, Israel, Italy, Latvia, Lithuania, Luxembourg, Malta, Netherlands, Norway, Poland, Portugal,
Romania, Russia, Slovak Republic, Slovenia, Spain, Sweden, Switzerland, Turkey, United Kingdom*
Source: IMF staff compilation.
*Shock-emitting (source) country. Source country is excluded from the set of recipient countries when analyzing fiscal shocks from the same source.
trends. In addition, outliersobservations with accommodate nonlinearity; that is, it allows esti-
quarter-over-quarter GDP growth rates higher than mation of spillovers under different states of the
10percent or lower than 10percent in any given economy. Moreover, the method is more robust to
quarter (very few observations)are excluded. misspecification of the data-generating process than
a vector autoregression, for which the misspecifi-
Exchange Rate Regime Classification cation error is compounded at each horizon of the
impulse response.
A measure of bilateral exchange rate arrangement
The following baseline linear model at time horizon
vis--vis the US dollar is constructed to estimate spill-
h (for h = 0, ... , H) is estimated using a panel ordi-
overs for different exchange rate regimes.
nary least squares estimator:
For the Reinhart-Rogoff classification, the exchange
rate regime is expressed as a time-varying index Z Z Shock
_________
i,t+h i,t1
=
h______it+
l=1 hl Xi,tl
L
based on the annual coarse de facto classification Yi,t1
Yi,t1
from Ilzetzki, Reinhart, and Rogoff (2017a, 2017b), + hi +ht + iht, (4.1)
ranging from 1 (most rigid) to 6 (most flexible). For
each period, if a country is assigned a value of 1 (de in which Zitis the variable of interest (real GDP,
facto peg) or 2 (de facto crawling peg), it is deemed consumption, investment, and the like) in recipient
a fixed regime. The quarterly index is interpolated country i at quarter t, Yit is real GDP in recipient
from annual data, assigning the same value for all four country i at quarter t, Shockitis the foreign fiscal
quarters within a year. For example, in 2015, this clas- shock facing country i at time t (see below), and Xit
sification yields seven fixed rate countries (Argentina, is a vector of control variables including lags of the
China, Costa Rica, India, Peru, Philippines, Vietnam) fiscal shock, lags of GDP growth, and lags of external
out of the sample of 55 countries.43 demand, measured as a weighted average of trading
The IMF pre-2008 classification (coarse) consists of partner growth rates (the number of lags L = 4 was
six categories, with 1 being the most rigid and 6 the chosen). Variables hi and htcapture the country and
most flexible.44 The classification changed in 2008, and time fixed effects. Given that the foreign fiscal shock is
post-2008 data are obtained from the IMFs website. expressed in units of recipient-country GDP (Shockit
As under the Reinhart-Rogoff classification, a country is scaled by lagged GDP Yit1), the coefficient h is
is generally classified as having a fixed exchange rate analogous to a domestic multiplier of an external shock
vis--vis the US dollar if it is assigned a value of 1 (Hall 2009; Barro and Redlick 2011). The impulse
(de facto peg) or 2 (de facto crawling peg or crawling response for H periods is constructed from a sequence
band narrower than or equal to 2percent). Again, of estimates { h} h=0
H.
the quarterly index is interpolated from annual data. The baseline fiscal shock combines country-specific
For example, for 2015, this classification yields two shocks from the five source countries (France, Ger-
fixed-rate countries (China, Vietnam) out of the sam- many, Japan, United Kingdom, United States) and
ple of 55 countries, although there are more fixed-rate weights them using trade links with recipient coun-
countries in earlier periods. tries. The assumption behind the weighting system
is that fiscal policy is transmitted mainly through
tradecountries with tighter trade links to the source
Annex4.2. Empirical Strategy would be expected to receive larger shocks in the form
Baseline Specification of larger changes in export demand, and therefore
As in Auerbach and Gorodnichenko (2013), the larger spillovers. However, the estimated spillovers
response of output in the recipient country to a fiscal capture those from all transmission channels, including
shock abroad is estimated using the local projections the financial channel. The external fiscal shock facing
method. This approach is particularly well suited to recipient country i in time t is given by
Mij,t1 sjtEj,t1
43The number of countries classified as fixed can generally 5______
Shockit =j=1 ______, (4.2)
vary over time given that the exchange rate regime classification is
Mj,t1 Ei,t1
time varying.
44Data for regime classification before 2008 is from Carmen in which j denotes source country, Mijt is country
Reinharts website, http://www.carmenreinhart.com. js goods imports from country i at time t, Mjt is
total goods imports by country j, sjt is the identified the two different states can then be analyzed by com-
fiscal shock in country j expressed in real terms in paring the estimated parameters 1h and 2h.
country js currency, and Ejt is country js US dollar For the source country, only the shock is partitioned
real exchange rate. Therefore, the second term on according to the state of the economy, which can be
the right side (sjtEj,t1 /Ei,t1
)equals the real mon- again either the cyclical position or monetary policy
etary value of the fiscal shock coming from country near the effective lower bound. The states are defined
j converted into units of recipient country is cur- in the same way as in the specification for recipient
rency. This term is then scaled by the import share countries. The source-country shock therefore becomes
(Mij,t1 / Mj,t1), which captures the relative impor- Shockitj :It1 (1It1 )
jShock
j j
Shockitj,
tance of recipient country i as a supplier of the source it+ (4.4)
countrys imports.45 Finally, the weighted shocks are in which It j is a {0;1} dummy variable indicating the
added up across the five source countries.46 The com- state in the shock-emitting country. The assumption
bined shocks are relatively small: for example, spend- behind interacting only the shock with the state
ing (tax) shocks average about 0.06 (0.1) percent of dummy is that although shocks in the source country
recipient-country GDP over the sample period. and its domestic response might be regime dependent,
their propagation to recipient countries is not.
and (Yj /Yi )is the ratio of source to recipient-country Annex Figure 4.3.1. Effects of Spending and Tax Shock on
GDPboth measured in US dollars.47 Recipient Countries Output: Comparison with Panel Vector
Autoregression
(Percent; quarters on x-axis)
Annex4.3. Robustness Tests
Local projection method Panel vector autoregression
To ensure that the baseline results are not driven by
the selected shock identification scheme or economet-
0.5 1. Output Effects of Spending Shock
ric approach, this section performs several robustness
checks. The results are robust to (1) estimation of 0.4
spillovers using a panel vector autoregression, which 0.3
accounts for the endogenous response of exchange rates
0.2
and monetary policy in recipient countries; and (2) the
use of alternative fiscal shocks based on forecast error 0.1
and narrative approaches.
0.0
Annex Figure 4.3.2. Effects of Spending and Tax Shock on sample from 2000 to 2012.49 The forecast error for
Recipient Countries Output: Forecast Errors {G,T,Y}is constructed as
each variable X=
(Percent; years on x-axis)
FEtX =Xt Xtf|
t1, (4.8)
0.8 1. Output Effects of Spending Shock
in which Xtis the growth rate of the variable from the
0.6
contemporaneous data release and Xtf|
t1is the forecast
0.4 one period earlier. A positive forecast error means an
expansionary spending shock and a contractionary
0.2
tax shock. Following Auerbach and Gorodnichenko
0.0 (2013), the forecast errors of spending and taxes are
0.2 regressed on the forecast errors of output to take
into account any changes as a result of surprises in
0.4
0 1 2 3 4 5 the business cycle. They are also regressed on lagged
macroeconomic variables growth rates (GDP, deflator,
0.2 2. Output Effects of Tax Shock investment, government spending or tax revenues) to
0.1 account for the portion of the innovation that can be
predicted from past observations. The forecast error
0.0
shocks for each source country are then constructed as
0.1 residuals from this regression, converted to levels using
base year (2010) expenditures or revenue, and replaced
0.2
in equations (4.1) and (4.2).
0.3 Spillover analysis using forecast error shocks confirms
0.4 the baseline resultsthat spending shocks have larger
0 1 2 3 4 5 spillovers than tax shocks (Annex Figure4.3.2)and
provides a strong robustness check. These shocks are
Source: IMF staff calculations.
Note: t = 0 is the year of respective shocks. Solid lines denote the response to constructed using an entirely different methodology, a
respective shocks, and dashed lines denote 90 percent condence bands. Effects different database and estimated at a different frequency
are estimated based on shocks derived from forecast errors. Shocks are
normalized to an average of 1 percent of GDP across source countries.
than the shocks used in the baseline specification. The
size of spillovers is somewhat larger compared with the
baseline, which can be explained, in part, by a stronger
response of source-country spending and revenues to
between actual variable and its forecast from the forecast error shocks compared with structural shocks
previous period) in the growth rates of government (although these impulse responses are imprecisely esti-
spending or tax revenues, this way capturing only mated because of the small sample).
unanticipated fiscal changes. This differs from the
structural shocks used in the baseline analysis, which
are based on actual changes in fiscal variables and
Identification Using Narrative Approach
can be anticipated by agents if they were announced To further confirm that the baseline results are not
earlier. The presence of such anticipated shocks could driven by the shock identification scheme, a robust-
bias the estimates because the information set of the ness check using the narrative tax shocks of Romer
econometrician is different from the information set and Romer (2010) is conducted. Several studies in the
of the agents. Because forecast errors capture unex- literature present narrative fiscal shocks (for example,
pected changes, the problem with fiscal foresight DeVries and others 2011), but the data set of Romer
is reduced under this approach, as the information and Romer (2010) is the most suitable for comparison
set of the econometrician and private agents is with the baseline analysis of the chapter given that it
more aligned. covers both expansion and consolidation episodes.50
The approach uses real-time fiscal projections by
49After 2012 the forecast data are not continuous.
the Organisation for Economic Co-operation and 50Narrative shock databases for government spending are much
Development and real-time actual data to construct less common in the literature, which precludes a robustness check of
the forecast error shocks at annual frequency on the spillovers from spending shocks based on narrative shocks.
Growth? Model-Based Results. Banque de France Work- Gorodnichenko, Yuriy, Enrique G. Mendoza, and Linda L. Tesar.
ing Paper 625. 2012. The Finnish Great Depression: From Russia with
Canzoneri Matthew, Fabrice Collard, Harris Dellas, and Behzad Love. American Economic Review 102 (4): 161944.
Diba. 2016. Fiscal Multipliers in Recessions. Economic Goujard, Antoine. 2017. Cross-Country Spillovers from Fiscal
Journal 126 (590): 75108. Consolidations. Fiscal Studies 38 (2): 21967.
Christiano, Lawrence, Martin Eichenbaum, and Charles Evans. Guajardo, Jaime, Daniel Leigh, and Andrea Pescatori. 2014.
2005. Nominal Rigidities and the Dynamic Effects of a Shock Expansionary Austerity: International Evidence. Journal of
to Monetary Policy. Journal of Political Economy 113 (1): 145. the European Economic Association 12 (4): 94968.
Christiano, Lawrence, Martin Eichenbaum, and Sergio Rebelo. Hall, Robert E. 2009. By How Much Does GDP Rise If the
2011. When Is the Government Spending Multiplier Government Buys More Output? Brookings Papers on Eco-
Large? Journal of Political Economy 119 (1): 78121. nomic Activity 40 (Fall): 183249.
Cloyne, James. 2013. Discretionary Tax Changes and the Ilzetzki, Ethan, Enrique G. Mendoza, and Carlos A. Vegh. 2013.
Macroeconomy: New Narrative Evidence from the United How Big (Small?) Are Fiscal Multipliers? Journal of Mone-
Kingdom. American Economic Review 103 (4): 150728. tary Economics 60 (2): 23954.
Cole, Harold L., and Lee E. Ohanian. 2004. New Deal Policies Ilzetzki, Ethan, Carmen M. Reinhart, and Kenneth S. Rogoff.
and the Persistence of the Great Depression: A General Equilib- 2017a. Exchange Arrangements Entering the 21st Century:
rium Analysis. Journal of Political Economy 112 (4): 779816. Which Anchor Will Hold? NBER Working Paper 23134,
Corsetti, Giancarlo, Andr Meier, and Gernot J. Mller. 2012. National Bureau of Economic Research, Cambridge, MA.
What Determines Government Spending Multipliers? . 2017b. The Country Chronologies to Exchange Rate
IMF Working Paper 12/150, International Monetary Fund, Arrangements into the 21st Century: Will the Anchor Cur-
Washington, DC. rency Hold? NBER Working Paper 23135, National Bureau
Corsetti, Giancarlo, and Gernot J. Mller. 2006. Twin Deficits: of Economic Research, Cambridge, MA.
Squaring Theory, Evidence, and Common Sense. Economic International Monetary Fund (IMF). 2016. Spillovers from
Policy 21 (48): 598638. Chinas Transition and from Migration. Chapter4 in World
De Mooij, Ruud, and Sjef Ederveen. 2008. Corporate Tax Economic Outlook, October.
Elasticities: A Readers Guide to Empirical Findings. Oxford Jord, scar. 2005. Estimation and Inference of Impulse
Review of Economic Policy 24 (4): 68097. Responses by Local Projections. American Economic Review
Devereux, Michael P. 2008. Business Taxation in a Globalized 95 (1): 16182.
World. Oxford Review of Economic Policy 24 (4): 62538. Kim, Soyoung, and Nouriel Roubini. 2008. Twin Deficit or
DeVries, Pete, Jaime Guajardo, Daniel Leigh, and Andrea Pesca- Twin Divergence? Fiscal Policy, Current Account, and Real
tori. 2011. A New Action-Based Dataset of Fiscal Consoli- Exchange Rate in the U.S. Journal of International Economics
dation. IMF Working Paper 11/128, International Monetary 74 (2): 36283.
Fund, Washington, DC. Kirchner, Markus, Jacopo Cimadomo, and Sebastian Hauptmeier.
Dornbusch, Rudiger. 1976. Expectations and Exchange Rate 2010. Transmission of Government Spending Shocks in the
Dynamics. Journal of Political Economy, 84 (6): 116176. Euro Area: Time Variation and Driving Forces. ECB Working
Eggertsson, Gauti B. 2011. What Fiscal Policy Is Effective at Paper Series 1219, European Central Bank, Frankfurt.
Zero Interest Rates? in NBER Macroeconomics Annual 2010, Klein, Michael W., and Jay C. Shambaugh. 2006. Fixed
vol. 25, 59112. Cambridge, MA: MIT Press. Exchange Rates and Trade. Journal of International Economics
Enders, Zeno, Gernot J. Mller, and Almuth Scholl. 2011. 70 (2): 35983.
How Do Fiscal and Technology Shocks Affect Real Exchange Mertens, Karel, and Morten O. Ravn. 2013. The Dynamic
Rates?: New Evidence for the United States. Journal of Inter- Effects of Personal and Corporate Income Tax Changes in the
national Economics 83 (1): 5369. United States. American Economic Review 103 (4): 121247.
Erceg, Christopher J., and Jesper Lind. 2010. Is There a Free Michaillat, Pascal. 2014. A Theory of Countercyclical Govern-
Lunch in a Liquidity Trap? International Finance Discussion ment Multiplier. American Economic Journal: Macroeconomics
Paper 1003, US Federal Reserve System, Washington, DC. 6 (1): 190217.
. 2013. Fiscal Consolidation in a Currency Union: Mineshima, Aiko, Marcos Poplawski-Ribeiro, and Anke Weber.
Spending Cuts vs. Tax Hikes. Journal of Economic Dynamics 2014. Size of Fiscal Multipliers. Post-crisis Fiscal Policy Part
and Control 37 (2): 42245. III Chapter 12, 31572. Cambridge, MA: MIT Press.
Fleming, J. Marcus. 1962. Domestic Financial Policies under Monacelli, Tommaso, and Roberto Perotti. 2010. Fiscal Policy,
Fixed and under Floating Exchange Rates. IMF Staff Papers the Real Exchange Rate and Traded Goods. Economic Journal
9 (3): 36980. 120 (544): 43761.
Forni, Mario, and Luca Gambetti. 2016. Government Spend- Mundell, Robert A. 1963. Capital Mobility and Stabilization
ing Shocks in Open Economy VARs. Journal of International Policy under Fixed and Flexible Exchange Rates. Canadian
Economics 99: 6884. Journal of Economics and Political Science 29 (4): 47585.
Nakamura, Emi, and Jon Steinsson. 2014. Fiscal Stimulus in , and Sarah Zubairy. Forthcoming. Government Spend-
a Monetary Union: Evidence from US Regions. American ing Multipliers in Good Times and in Bad: Evidence from
Economic Review 104 (3): 75392. US Historical Data, Journal of Political Economy.
Nicar, Stephen B. 2015. International Spillovers from U.S. Fis- Ravn, Morten O., Stephanie Schmitt-Groh, and Martin
cal Policy Shocks. Open Economies Review 26 (5): 108197. Uribe. 2012. Consumption, Government Spending, and
Obstfeld, Maurice, and Kenneth Rogoff. 1995. Exchange Rate the Real Exchange Rate. Journal of Monetary Economics 59
Dynamics Redux. Journal of Political Economy 103 (3): 62460. (3): 21534.
Owyang, Michael, Valerie Ramey, and Sarah Zubairy. 2013. Are Reinhart, Carmen, and Kenneth Rogoff. 2004. The Modern
Government Spending Multipliers Greater during Times of History of Exchange Rate Arrangements: A Reinterpretation.
Slack? Evidence from 20th Century Historical Data. Ameri- Quarterly Journal of Economics 119 (1): 148.
can Economic Review 103 (2): 12934. Riera-Crichton, Daniel, Carlos A. Vegh, and Guillermo Vuletin.
Poghosyan, Tigran. 2017. Cross-Country Spillovers of Fiscal 2015. Procyclical and Countercyclical Fiscal Multipliers:
Consolidations in the Euro Area. IMF Working Paper Evidence from OECD Countries. Journal of International
17/140, International Monetary Fund, Washington, DC. Money and Finance 52: 1531.
Popescu, Adina, and Ippei Shibata. Forthcoming. Spillovers Romer, Christina D., and David H. Romer. 2010. The Macro-
from US Government Spending Shocks: Impact on External economic Effects of Tax Changes: Estimates Based on a New
Positions. IMF Spillover Note, International Monetary Fund, Measure of Fiscal Shocks. American Economic Review 100
Washington, DC. (3): 763801.
Qureshi, Mahvash S., and Charalambos Tsangarides. 2010. The Rose, Andrew K., and Eric van Wincoop. 2001. National
Empirics of Exchange Rate Regimes and Trade: Words vs. Money as a Barrier to International Trade: The Real Case
Deeds. IMF Working Paper 10/48, International Monetary for Currency Union. American Economic Review 91
Fund, Washington, DC. (2): 38690.
Ramey, Valerie A. 2011. Identifying Government Spending Woodford, Michael. 2011. Simple Analytics of the Government
Shocks: Its All in the Timing. Quarterly Journal of Economics Expenditure Multiplier. American Economic Journal: Macro-
126 (1): 150. economics 3 (1): 135.
T
he Statistical Appendix presents histori- assumptions underlying the projections for selected
cal data as well as projections. It comprises economies are described in Box A1.
seven sections: Assumptions, Whats New, With regard to interest rates, it is assumed that the
Data and Conventions, Country Notes, London interbank offered rate (LIBOR) on six-month
Classification of Countries, Key Data Documentation, US dollar deposits will average 1.4 percent in 2017
and Statistical Tables. and 1.9 percent in 2018, that three-month euro depos-
The assumptions underlying the estimates and pro- its will average 0.3 percent in 2017 and 2018, and
jections for 201718 and the medium-term scenario that six-month yen deposits will average 0.1 percent in
for 201922 are summarized in the first section. The 2017 and 0.2 percent in 2018.
second section presents a brief description of the As a reminder, with respect to introduction of the
changes to the database and statistical tables since the euro, on December 31, 1998, the Council of the Euro-
April 2017 World Economic Outlook (WEO). The third pean Union decided that, effective January 1, 1999,
section provides a general description of the data and the irrevocably fixed conversion rates between the euro
the conventions used for calculating country group and currencies of the member countries adopting the
composites. The fourth section summarizes selected euro are as follows:
key information for each country. The classification of
countries in the various groups presented in the WEO
1 euro = 13.7603 Austrian schillings
is summarized in the fifth section. The sixth section
= 40.3399 Belgian francs
provides information on methods and reporting stan-
= 0.585274 Cyprus pound1
dards for the member countries national account and
= 1.95583 Deutsche marks
government finance indicators included in the report.
= 15.6466 Estonian krooni2
The last, and main, section comprises the statistical
= 5.94573 Finnish markkaa
tables. (Statistical Appendix A is included here; Sta-
= 6.55957 French francs
tistical Appendix B is available online.) Data in these
= 340.750 Greek drachmas3
tables have been compiled on the basis of information
= 0.787564 Irish pound
available through September 22, 2017. The figures
= 1,936.27 Italian lire
for 2017 and beyond are shown with the same degree
= 0.702804 Latvian lat4
of precision as the historical figures solely for conve-
= 3.45280 Lithuanian litas5
nience; because they are projections, the same degree
= 40.3399 Luxembourg francs
of accuracy is not to be inferred.
= 0.42930 Maltese lira1
= 2.20371 Netherlands guilders
Assumptions
= 200.482 Portuguese escudos
Real effective exchange rates for the advanced econo-
= 30.1260 Slovak koruna6
mies are assumed to remain constant at their average
= 239.640 Slovenian tolars7
levels measured during the period July 20 to August
= 166.386 Spanish pesetas
17, 2017. For 2017 and 2018, these assumptions 1Established on January1, 2008.
imply average US dollarspecial drawing right (SDR) 2Established on January1, 2011.
conversion rates of 1.385 and 1.409, US dollareuro 3Established on January 1, 2001.
4Established on January1, 2014.
conversion rates of 1.128 and 1.176, and yenUS dollar 5Established on January 1, 2015.
conversion rates of 111.4 and 109.1, respectively. 6Established on January 1, 2009.
It is assumed that the price of oil will average $50.28 a 7Established on January 1, 2007.
Whats New The fiscal gross and net debt data reported in the
WEO are drawn from official data sources and IMF staff
Data for Somalia are included in the emerging
estimates. While attempts are made to align gross and net
market and developing economies group composites,
debt data with the definitions in the GFSM, as a result of
enlarging the database to a total of 193 countries.
data limitations or specific country circumstances, these
Somalia is classified as a member of the Middle East
data can sometimes deviate from the formal definitions.
and North Africa region.
Although every effort is made to ensure the WEO data
Starting with the October 2017 WEO, the real GDP
are relevant and internationally comparable, differences
per capita data in Statistical Tables A1, B1, and B2 are
in both sectoral and instrument coverage mean that the
shown at purchasing power parity. This differs from the
data are not universally comparable. As more informa-
treatment of these data in the April 2017 WEO and
tion becomes available, changes in either data sources or
earlier issues, in which the data were shown in local
instrument coverage can give rise to data revisions that can
national currency.
sometimes be substantial. For clarification on the devia-
tions in sectoral or instrument coverage, please refer to the
Data and Conventions
metadata for the online WEO database.
Data and projections for 193 economies form the statisti- Composite data for country groups in the WEO are
cal basis of the WEO database. The data are maintained either sums or weighted averages of data for individual
jointly by the IMFs Research Department and regional countries. Unless noted otherwise, multiyear averages of
departments, with the latter regularly updating country growth rates are expressed as compound annual rates of
projections based on consistent global assumptions. change.2 Arithmetically weighted averages are used for
Although national statistical agencies are the ultimate all data for the emerging market and developing econo-
providers of historical data and definitions, international mies group except data on inflation and money growth,
organizations are also involved in statistical issues, with the for which geometric averages are used. The following
objective of harmonizing methodologies for the compila- conventions apply:
tion of national statistics, including analytical frameworks, Country group composites for exchange rates, interest
concepts, definitions, classifications, and valuation proce- rates, and growth rates of monetary aggregates are
dures used in the production of economic statistics. The weighted by GDP converted to US dollars at market
WEO database reflects information from both national exchange rates (averaged over the preceding three
source agencies and international organizations. years) as a share of group GDP.
Most countries macroeconomic data presented in the Composites for other data relating to the domes-
WEO conform broadly to the 1993 version of the System tic economy, whether growth rates or ratios, are
of National Accounts (SNA). The IMFs sector statistical weighted by GDP valued at purchasing power parity
standardsthe sixth edition of the Balance of Payments as a share of total world or group GDP.3 Annual
and International Investment Position Manual (BPM6), the inflation rates are simple percentage changes from the
Monetary and Financial Statistics Manual and Compila- previous years, except in the case of emerging market
tion Guide (MFSMCG), and the Government Finance and developing economies, for which the rates are
Statistics Manual 2014 (GFSM 2014)have been or are based on logarithmic differences.
being aligned with the SNA 2008. These standards reflect Composites for real GDP per capita in purchasing
the IMFs special interest in countries external positions, power parity terms are sums of individual country
financial sector stability, and public sector fiscal positions.
The process of adapting country data to the new standards
versions of the SNA older than that from 1993. A similar adoption pat-
begins in earnest when the manuals are released. However, tern is expected for the BPM6 and GFSM 2014. Please refer to Table G,
full concordance with the manuals is ultimately dependent which lists the statistical standards adhered to by each country.
on the provision by national statistical compilers of revised 2Averages for real GDP and its components, employment, infla-
country data; hence, the WEO estimates are only partially tion, factor productivity, trade, and commodity prices are calculated
based on the compound annual rate of change, except in the case of
adapted to these manuals. Nonetheless, for many countries GDP per capita and the unemployment rate, which are based on the
the impact, on major balances and aggregates, of conver- simple arithmetic average.
3See Revised Purchasing Power Parity Weights in the July 2014
sion to the updated standards will be small. Many other
WEO Update for a summary of the revised purchasing-power-parity-based
countries have partially adopted the latest standards and
weights, as well as Box A2 of the April 2004 WEO and Annex IV of the
will continue implementation over a period of years.1 May 1993 WEO. See also Anne-Marie Gulde and Marianne Schulze-
Ghattas, Purchasing Power Parity Based Weights for the World Economic
1Many countries are implementing the SNA 2008 or European System Outlook, in Staff Studies for the World Economic Outlook (Washington,
of National and Regional Accounts (ESA) 2010, and a few countries use DC: International Monetary Fund, December 1993), 10623.
data after conversion to the international dollar in the line with international standards and lifted the declara-
years indicated. tion of censure issued in 2013. Given the differences in
Unless noted otherwise, composites for all sectors for geographical coverage, weights, sampling, and meth-
the euro area are corrected for reporting discrepan- odology of these series, the average CPI inflation for
cies in intra-area transactions. Annual data are not 2014, 2015, and 2016 and end-of-period inflation for
adjusted for calendar-day effects. For data prior to 2015 and 2016 are not reported in the October 2017
1999, data aggregations apply 1995 European cur- World Economic Outlook.
rency unit exchange rates. Argentinas authorities discontinued the publication
Composites for fiscal data are sums of individual of labor market data in December 2015 and released
country data after conversion to US dollars at the new series starting in the second quarter of 2016.
average market exchange rates in the years indicated. Argentinas and Venezuelas consumer prices are
Composite unemployment rates and employment excluded from all WEO group aggregates.
growth are weighted by labor force as a share of Greeces primary balance estimates for 2016 are based
group labor force. on preliminary excessive deficit procedure (EDP)
Composites relating to external sector statistics are data on an accrual basis (ESA 2010) provided by the
sums of individual country data after conversion to National Statistical Service (ELSTAT) as of April 21,
US dollars at the average market exchange rates in the 2017. Fiscal data since 2010 are adjusted in line with
years indicated for balance of payments data and at program definitions.
end-of-year market exchange rates for debt denomi- Indias growth rates of real GDP calculated from
nated in currencies other than US dollars. 1998 to 2011 are as per national accounts with base
Composites of changes in foreign trade volumes and year 2004/05, and thereafter are as per national
prices, however, are arithmetic averages of percent accounts with base year 2011/12.
changes for individual countries weighted by the US dol- Against the background of a civil war and weak
lar value of exports or imports as a share of total world capacities, the reliability of Libyas data, especially
or group exports or imports (in the preceding year). medium-term projections, is low.
Unless noted otherwise, group composites are com- Data for Syria are excluded from 2011 onward
because of the uncertain political situation.
puted if 90 percent or more of the share of group
Projecting the economic outlook in Venezuela, includ-
weights is represented.
ing assessing past and current economic developments
Data refer to calendar years, except in the case of a
as the basis for the projections, is complicated by the
few countries that use fiscal years. Please refer to Table
lack of discussions with the authorities (the last Article
F, which lists the economies with exceptional reporting
IV consultation took place in 2004), long intervals
periods for national accounts and government finance
in receiving data with information gaps, incomplete
data for each country.
provision of information, and difficulties in interpret-
For some countries, the figures for 2016 and earlier
ing certain reported economic indicators in line with
are based on estimates rather than actual outturns.
economic developments. The fiscal accounts include
Please refer to Table G, which lists the latest actual out-
the budgetary central government and Petrleos de
turns for the indicators in the national accounts, prices,
Venezuela, S.A. (PDVSA), and the fiscal accounts
government finance, and balance of payments indicators
data for 201622 are IMF staff estimates. Revenue
for each country.
includes the IMF staffs estimated foreign exchange
profits transferred from the central bank to the
Country Notes government (buying US dollars at the most appreci-
The consumer price data for Argentina before Decem- ated rate and selling at more depreciated rates in a
ber 2013 reflect the consumer price index (CPI) for multitier exchange rate system) and excludes the staffs
the Greater Buenos Aires Area (CPI-GBA), while from estimated revenue from PDVSAs sale of PetroCaribe
December 2013 to October 2015 the data reflect the assets to the central bank. Fiscal accounts for 201022
national CPI (IPCNu). The new government that took correspond to the budgetary central government and
office in December 2015 discontinued the IPCNu, PDVSA. Fiscal accounts before 2010 correspond to
stating that it was flawed, and released a new CPI for the budgetary central government, public enterprises
the Greater Buenos Aires Area on June 15, 2016 (a new (including PDVSA), Instituto Venezolano de los
national CPI has been disseminated starting in June Seguros Sociales (IVSSsocial security), and Fondo
2016). At its November 9, 2016, meeting, the IMF de Garanta de Depsitos y Proteccin Bancaria
Executive Board considered the new CPI series to be in (FOGADEdeposit insurance).
International Monetary Fund | October 2017 219
WORLD ECONOMIC OUTLOOK: SEEKING SUSTAINABLE GROWTHSHORT-TERM RECOVERY, LONG-TERM CHALLENGES
tional law and practice. Some territorial entities included here are winder Singh, Debt Relief for Low-Income Countries: The Enhanced
not states, although their statistical data are maintained on a separate HIPC Initiative, IMF Pamphlet Series 51 (Washington, DC: Interna-
and independent basis. tional Monetary Fund, November 1999).
Table A. Classification by World Economic Outlook Groups and Their Shares in Aggregate GDP, Exports of Goods and
Services, and Population, 20171
(Percent of total for group or world)
Exports of Goods
GDP and Services Population
Number of Advanced Advanced Advanced
Economies Economies World Economies World Economies World
Advanced Economies 39 100.0 41.8 100.0 64.4 100.0 14.5
United States 37.0 15.5 16.6 10.7 30.5 4.4
Euro Area 19 28.1 11.7 41.2 26.5 31.9 4.6
Germany 7.9 3.3 12.1 7.8 7.8 1.1
France 5.4 2.3 5.7 3.7 6.1 0.9
Italy 4.4 1.9 4.2 2.7 5.7 0.8
Spain 3.4 1.4 3.1 2.0 4.4 0.6
Japan 10.4 4.4 6.1 3.9 12.0 1.7
United Kingdom 5.5 2.3 5.6 3.6 6.2 0.9
Canada 3.3 1.4 3.6 2.3 3.4 0.5
Other Advanced Economies 16 15.6 6.5 26.9 17.3 16.0 2.3
Memorandum
Major Advanced Economies 7 74.1 31.0 53.8 34.7 71.7 10.4
Emerging Emerging Emerging
Market and Market and Market and
Developing Developing Developing
Economies World Economies World Economies World
Emerging Market and Developing Economies 154 100.0 58.2 100.0 35.6 100.0 85.5
Regional Groups
Commonwealth of Independent States2 12 7.8 4.5 6.9 2.5 4.6 3.9
Russia 5.5 3.2 4.5 1.6 2.3 2.0
Emerging and Developing Asia 30 54.3 31.6 50.1 17.8 56.9 48.6
China 30.5 17.7 30.0 10.7 22.2 19.0
India 12.4 7.2 6.0 2.2 20.9 17.8
Excluding China and India 28 11.4 6.6 14.1 5.0 13.8 11.8
Emerging and Developing Europe 12 6.1 3.5 9.8 3.5 2.8 2.4
Latin America and the Caribbean 32 13.5 7.8 14.2 5.1 9.9 8.4
Brazil 4.5 2.6 3.0 1.1 3.3 2.8
Mexico 3.3 1.9 5.4 1.9 2.0 1.7
Middle East, North Africa, Afghanistan, and
Pakistan 23 13.2 7.7 14.7 5.2 10.7 9.2
Middle East and North Africa 21 11.7 6.8 14.3 5.1 7.1 6.0
Sub-Saharan Africa 45 5.2 3.0 4.3 1.5 15.1 12.9
Excluding Nigeria and South Africa 43 2.6 1.5 2.5 0.9 11.3 9.7
Analytical Groups3
By Source of Export Earnings
Fuel 28 18.6 10.8 20.2 7.2 11.7 10.0
Nonfuel 125 81.4 47.3 79.8 28.4 88.3 75.5
Of Which, Primary Products 31 4.6 2.7 4.7 1.7 7.9 6.7
By External Financing Source
Net Debtor Economies 121 49.6 28.9 46.2 16.5 66.8 57.1
Net Debtor Economies by Debt-
Servicing Experience
Economies with Arrears and/or Rescheduling
during 201216 25 3.3 1.9 2.1 0.8 5.5 4.7
Other Groups
Heavily Indebted Poor Countries 39 2.4 1.4 1.9 0.7 11.3 9.7
Low-Income Developing Countries 59 7.2 4.2 6.6 2.3 22.5 19.2
1The GDP shares are based on the purchasing-power-parity valuation of economies GDP. The number of economies comprising each group reflects those
insufficient data.
Table D. Emerging Market and Developing Economies by Region and Main Source of Export Earnings
Fuel Nonfuel Primary Products
Commonwealth of Independent States
Azerbaijan Uzbekistan
Kazakhstan
Russia
Turkmenistan1
Emerging and Developing Asia
Brunei Darussalam Lao P.D.R.
Timor-Leste Marshall Islands
Mongolia
Papua New Guinea
Solomon Islands
Tuvalu
Latin America and the Caribbean
Bolivia Argentina
Ecuador Chile
Trinidad and Tobago Guyana
Venezuela Honduras
Paraguay
Suriname
Uruguay
Middle East, North Africa, Afghanistan, and Pakistan
Algeria Afghanistan
Bahrain Mauritania
Iran Sudan
Iraq
Kuwait
Libya
Oman
Qatar
Saudi Arabia
United Arab Emirates
Yemen
Sub-Saharan Africa
Angola Burkina Faso
Chad Burundi
Republic of Congo Central African Republic
Equatorial Guinea Democratic Republic of the Congo
Gabon Cte dIvoire
Nigeria Eritrea
South Sudan Guinea
Guinea-Bissau
Liberia
Malawi
Mali
Sierra Leone
South Africa
Zambia
1Turkmenistan, which is not a member of the Commonwealth of Independent States, is included in this group for reasons of geography and similarity in
economic structure.
Table E. Emerging Market and Developing Economies by Region, Net External Position, and Status as Heavily Indebted Poor Countries
and Low-Income Developing Countries
Low-Income Low-Income
Net External Heavily Indebted Developing Net External Heavily Indebted Developing
Position1 Poor Countries2 Countries Position1 Poor Countries2 Countries
Commonwealth of Independent States Emerging and Developing Europe
Armenia * Albania *
Azerbaijan Bosnia and Herzegovina *
Belarus * Bulgaria *
Georgia3 * Croatia *
Kazakhstan * Hungary *
Kyrgyz Republic * * Kosovo *
Moldova * * FYR Macedonia *
Russia Montenegro *
Tajikistan * * Poland *
Turkmenistan3 Romania *
Ukraine3 * Serbia *
Uzbekistan * Turkey *
Emerging and Developing Asia Latin America and the Caribbean
Bangladesh * * Antigua and Barbuda *
Bhutan * * Argentina
Brunei Darussalam The Bahamas *
Cambodia * * Barbados *
China Belize *
Fiji * Bolivia *
India * Brazil *
Indonesia * Chile *
Kiribati * Colombia *
Lao P.D.R. * * Costa Rica *
Malaysia Dominica *
Maldives * Dominican Republic *
Marshall Islands * Ecuador *
Micronesia El Salvador *
Mongolia * Grenada *
Myanmar * * Guatemala *
Nauru * Guyana *
Nepal * Haiti * *
Palau Honduras * *
Papua New Guinea * * Jamaica *
Philippines * Mexico *
Samoa * Nicaragua * *
Solomon Islands * * Panama *
Sri Lanka * Paraguay *
Thailand * Peru *
Timor-Leste * St. Kitts and Nevis *
Tonga * St. Lucia *
Tuvalu * St. Vincent and the
Vanuatu * Grenadines *
Vietnam * * Suriname *
Trinidad and Tobago
Uruguay *
Venezuela
224 International Monetary Fund | October 2017
STATISTICAL APPENDIX
Table E. Emerging Market and Developing Economies by Region, Net External Position, and Status as Heavily Indebted Poor Countries
and Low-Income Developing Countries (continued)
Low-Income Low-Income
Net External Heavily Indebted Developing Net External Heavily Indebted Developing
Position1 Poor Countries2 Countries Position1 Poor Countries2 Countries
Middle East, North Africa, Afghanistan, and Pakistan Democratic Republic of
the Congo * *
Afghanistan *
Republic of Congo * *
Algeria
Cte dIvoire * *
Bahrain
Equatorial Guinea *
Djibouti * *
Eritrea * * *
Egypt *
Ethiopia * *
Iran
Gabon
Iraq
The Gambia * *
Jordan *
Ghana * *
Kuwait
Guinea * *
Lebanon *
Guinea-Bissau * *
Libya
Kenya * *
Mauritania * *
Lesotho * *
Morocco *
Liberia * *
Oman
Madagascar * *
Pakistan *
Malawi * *
Qatar
Mali * *
Saudi Arabia
Mauritius
Somalia * * *
Mozambique * *
Sudan * * *
Namibia *
Syria4 ...
Niger * *
Tunisia *
Nigeria * *
United Arab Emirates
Rwanda * *
Yemen * *
So Tom and Prncipe * *
Sub-Saharan Africa
Senegal * *
Angola
Seychelles *
Benin * *
Sierra Leone * *
Botswana
South Africa
Burkina Faso * *
South Sudan4 ... *
Burundi * *
Swaziland *
Cabo Verde *
Tanzania * *
Cameroon * *
Togo * *
Central African Republic * *
Uganda * *
Chad * *
Zambia * *
Comoros * *
Zimbabwe * *
1Dot (star) indicates that the country is a net creditor (net debtor).
2Dot instead of star indicates that the country has reached the completion point, which allows it to receive the full debt relief committed to at the decision point.
3Georgia, Turkmenistan, and Ukraine, which are not members of the Commonwealth of Independent States, are included in this group for reasons of geography and similarity in
economic structure.
4South Sudan and Syria are omitted from the net external position group composite for lack of a fully developed database.
Commerce, and/or Development; MoF = Ministry of Finance and/or Treasury; NSO = National Statistics Office; PFTAC = Pacific Financial Technical Assistance Centre.
2National accounts base year is the period with which other periods are compared and the period for which prices appear in the denominators of the price relationships used to
that average volume components using weights from a year in the moderately distant past.
4For some countries, the structures of government consist of a broader coverage than specified for the general government. Coverage: BCG = budgetary central government;
CG = central government; LG = local government; MPC = monetary public corporation, including central bank; NFPC = nonfinancial public corporation; NMPC = nonmonetary
financial public corporation; SG = state government; SS = social security fund; TG = territorial governments.
5Accounting standard: A = accrual accounting; C = cash accounting; CB = commitments basis accounting; Mixed = combination of accrual and cash accounting.
6Base year is not equal to 100 because the nominal GDP is not measured in the same way as real GDP or the data are seasonally adjusted.
Box A1. Economic Policy Assumptions Underlying the Projections for Selected Economies
Fiscal Policy Assumptions Brazil: Fiscal projections for the end of 2017 take into
The short-term fiscal policy assumptions used in the account budget performance through July 31, 2017, and
World Economic Outlook (WEO) are normally based on the deficit target approved in the budget law.
officially announced budgets, adjusted for differences Canada: Projections use the baseline forecasts in
between the national authorities and the IMF staff the 2017 federal budget and 2017 provincial budget
regarding macroeconomic assumptions and projected updates as available. The IMF staff makes some
fiscal outturns. When no official budget has been adjustments to these forecasts, including for differences
announced, projections incorporate policy measures in macroeconomic projections. The IMF staff forecast
that are judged likely to be implemented. The medium- also incorporates the most recent data releases from
Statistics Canadas Canadian System of National
term fiscal projections are similarly based on a judg-
Economic Accounts, including federal, provincial,
ment about the most likely path of policies. For cases
and territorial budgetary outturns through the second
in which the IMF staff has insufficient information to
quarter of 2017.
assess the authorities budget intentions and prospects
Chile: Projections are based on the authorities
for policy implementation, an unchanged structural
budget projections, adjusted to reflect the IMF staffs
primary balance is assumed unless indicated otherwise.
projections for GDP and copper prices.
Specific assumptions used in regard to some of the
China: Projections assume that the pace of fiscal
advanced economies follow. (See also Tables B5 to B9 in
consolidation is likely to be more gradual, reflecting
the online section of the Statistical Appendix for data on
reforms to strengthen social safety nets and the social
fiscal net lending/borrowing and structural balances.)1
security system announced as part of the Third Plenum
Argentina: Fiscal projections are based on the avail-
reform agenda.
able information regarding budget outturn and budget
Denmark: Estimates for 2016 are aligned with the
plans for the federal and provincial governments, fiscal
latest official budget estimates and the underlying
measures announced by the authorities, and IMF staff
economic projections, adjusted where appropriate
macroeconomic projections.
for the IMF staffs macroeconomic assumptions. For
Australia: Fiscal projections are based on Australian
201718, the projections incorporate key features
Bureau of Statistics data, the fiscal year 2017/18
of the medium-term fiscal plan as embodied in the
budget, and IMF staff estimates.
authorities Convergence Programme 2016 submitted
Austria: Fiscal projections are based on data from
to the European Union.
Statistics Austria, the authorities projections, and IMF
France: Projections for 2017 reflect the budget law
staff estimates and projections.
and cancellation of spending taken in July 2017. For
Belgium: Projections reflect the IMF staffs
201819, they are based on the multiyear budget and
assessment of policies and measures laid out in the
the preliminary fiscal path announced by the new
2017 budget and the 201619 Stability Programme,
government in July 2017, adjusted for differences in
incorporated into the IMF staffs macroeconomic
assumptions on macro and financial variables, and
framework.
revenue projections. Historical fiscal data reflect the
1The output gap is actual minus potential output, as a
May 2017 revisions and update of the fiscal accounts,
percentage of potential output. Structural balances are expressed debt data, and national accounts for 2014 and 2015.
as a percentage of potential output. The structural balance is the Germany: The IMF staffs projections for 2017 and
actual net lending/borrowing minus the effects of cyclical output beyond are based on the 2017 Stability Programme
from potential output, corrected for one-time and other factors, Update, adjusted for the differences in the IMF staffs
such as asset and commodity prices and output composition macroeconomic framework and assumptions concerning
effects. Changes in the structural balance consequently include
effects of temporary fiscal measures, the impact of fluctuations
revenue elasticities. The estimate of gross debt includes
in interest rates and debt-service costs, and other noncyclical portfolios of impaired assets and noncore business
fluctuations in net lending/borrowing. The computations of transferred to institutions that are winding up, as well as
structural balances are based on IMF staff estimates of potential other financial sector and EU support operations.
GDP and revenue and expenditure elasticities. (See Annex I of Greece: The fiscal projections reflect the IMF staffs
the October 1993 WEO.) Net debt is calculated as gross debt
minus financial assets corresponding to debt instruments. Esti-
assessment of implementation of legislated fiscal
mates of the output gap and of the structural balance are subject measures under the IMF and European Stability
to significant margins of uncertainty. Mechanism (ESM) program.
Box A1 (continued)
Hong Kong Special Administrative Region: Projections onward assume compliance with rules established in
are based on the authorities medium-term fiscal the Fiscal Responsibility Law.
projections on expenditures. Netherlands: Fiscal projections for 201722 are
Hungary: Fiscal projections include IMF staff based on the authorities Bureau for Economic
projections of the macroeconomic framework and of Policy Analysis budget projections, after differences
the impact of recent legislative measures, as well as in macroeconomic assumptions are adjusted for.
fiscal policy plans announced in the 2017 budget. Historical data were revised following the June 2014
India: Historical data are based on budgetary Central Bureau of Statistics release of revised macro
execution data. Projections are based on available data because of the adoption of the European System
information on the authorities fiscal plans, with of National and Regional Accounts (ESA 2010) and
adjustments for IMF staff assumptions. Subnational data the revisions of data sources.
are incorporated with a lag of up to two years; general New Zealand: Fiscal projections are based on the
government data are thus finalized well after central authorities fiscal year 2017/18 budget and on IMF
government data. IMF and Indian presentations differ, staff estimates.
particularly regarding divestment and license auction Portugal: Projections for 2017 are based on the
proceeds, net versus gross recording of revenues in authorities approved budget, adjusted to reflect the IMF
certain minor categories, and some public sector lending. staffs macroeconomic forecast. Projections thereafter are
Indonesia: IMF projections are based on moderate based on the assumption of unchanged policies.
tax policy and administration reforms, fuel subsidy Puerto Rico: Fiscal projections are based on the Puerto
pricing reforms introduced in January 2015, and a Rico Fiscal and Economic Growth Plan (FEGP), which
gradual increase in social and capital spending over the was prepared on March 13, 2017, and certified by
medium term in line with fiscal space. the Oversight Board. In line with assumptions of this
Ireland: Fiscal projections are based on the countrys plan, IMF projections assume that Puerto Rico will
Budget 2017, Stability Programme Update 2017, and lose federal funding for the Affordable Care Act (ACA)
Summer Economic Statement 2017. starting in 2018. Likewise, projections assume federal tax
Israel: Historical data are based on Government incentives, which were neutralizing the effects of Puerto
Finance Statistics data prepared by the Central Bureau Ricos Act 154 on foreign companies, will no longer
of Statistics. Projections for 2017 and 2018 are based on be available, starting in 2018, leading to additional
the 201718 budget, adjusted for the fiscal impact of revenue losses. Given sizable policy uncertainty, some
new measures announced in April 2017 (the Net Family FEGP and IMF assumptions may differ, in particular,
Plan) and for one-off revenues in 2017 arising from a those relating to the effects of the corporate tax reform,
large foreign direct investment transaction (0.3 percent tax compliance and tax adjustments (fees and rates);
of GDP). The central government deficit is assumed reduction of subsidies, freezing of payroll operational
to remain at the current ceiling level of 2.9 percent of costs, improvement of mobility, and reduction of
GDP in subsequent years, rather than declining in line expenses; and increasing health care efficiency. On
with medium-term fiscal targets, consistent with long the expenditure side, measures include extension of
experience of revisions to those targets. Act 66, which freezes much government spending,
Italy: IMF staff estimates and projections are based on through 2020; reduction of operating costs; decreases in
the fiscal plans included in the governments 2017 budget government subsidies; and spending cuts in education.
and April 2017 Economic and Financial Document. Although IMF policy assumptions are similar to those
Japan: The projections include fiscal measures in the FEGP scenario with full measures, the IMFs
already announced by the government, including the projections of fiscal revenues, expenditures, and balance
fiscal stimulus package for 2017 and the consumption are different from FEGPs. This stems from two main
tax hike in October 2019. differences in methodologies: first and foremost, while
Korea: The medium-term forecast incorporates the IMF projections are on an accrual basis, FEGPs are on
governments announced medium-term consolidation a cash basis. Second, the IMF and FEGP make very
path. different macroeconomic assumptions.
Mexico: Fiscal projections for 2017 are broadly in Russia: Projections for 201719 are IMF staff estimates,
line with the approved budget; projections for 2018 based on the authorities budget. Projections for 202022
Box A1 (continued)
are based on an oil price rule to be in effect in 2022, with adjusted for differences between IMF staff forecasts
adjustments by the IMF staff. of macroeconomic variables (such as GDP growth
Saudi Arabia: IMF staff projections of oil revenues and inflation) and the forecasts of these variables
are based on WEO baseline oil prices and the assumed in the authorities fiscal projections. IMF
assumption that Saudi Arabia continues to meet its staff data exclude public sector banks and the effect of
commitments under the OPEC+ agreement. For non- transferring assets from the Royal Mail Pension Plan
oil revenues, IMF staff estimates of the revenue impact to the public sector in April 2012. Real government
of announced policies in the Fiscal Balance Program consumption and investment are part of the real GDP
are included in the baseline. On the expenditure side, path, which, according to the IMF staff, may or may
starting in 2017, following recent reforms, the wage not be the same as projected by the U.K. Office for
bill estimates no longer include the 13th-month wage Budget Responsibility.
payment that used to be awarded every three years United States: Fiscal projections are based on the
in accordance with the lunar calendar. Expenditure January 2017 Congressional Budget Office baseline
projections take the 2017 budget as a starting point adjusted for the IMF staffs policy and macroeconomic
and reflect staff estimates of the effects of the latest assumptions. The baseline incorporates the key
changes in policies and economic developments. provisions of the Bipartisan Budget Act of 2015,
Singapore: For fiscal years 2016/17 and 2017/18, including a partial rollback of the sequester spending
projections are based on budget numbers. For the cuts in fiscal year 2016. In fiscal years 2017 through
remainder of the projection period, the IMF staff 2022, the IMF staff assumes that the sequester cuts will
assumes unchanged policies. continue to be partially replaced, in proportions similar
South Africa: Fiscal projections are based on the to those already implemented in fiscal years 2014 and
authorities 2017 Budget Review. 2015, with back-loaded measures generating savings
Spain: For 2017, fiscal data are IMF staff projections, in mandatory programs and additional revenues.
reflecting the cash outturn through May and the 2017 Projections also incorporate the Protecting Americans
budget passed by Parliament. For 2018 and beyond, from Tax Hikes Act of 2015, which extended
fiscal projections are based on the measures specified in some existing tax cuts for the short term and some
the Stability Programme Update 201720 and on the permanently. Finally, fiscal projections are adjusted to
IMF staffs macroeconomic projections. reflect the IMF staffs forecasts for key macroeconomic
Sweden: Fiscal projections take into account the and financial variables and different accounting
authorities projections based on the 2017 Spring treatment of financial sector support and of defined-
Budget. The impact of cyclical developments on the benefit pension plans and are converted to a general
fiscal accounts is calculated using the Organisation government basis. Data are compiled using SNA 2008,
for Economic Co-operation and Developments and when translated into government finance statistics,
2005 elasticity to take into account output and this is in accordance with GFSM 2014. Because of data
employment gaps. limitations, most series begin in 2001.
Switzerland: The projections assume that fiscal policy
is adjusted as necessary to keep fiscal balances in line
with the requirements of Switzerlands fiscal rules. Monetary Policy Assumptions
Turkey: The fiscal projections for 2017 are based on Monetary policy assumptions are based on the estab-
the authorities Medium Term Programme 201719, lished policy framework in each country. In most
with adjustments for additionally announced fiscal cases, this implies a nonaccommodative stance over
measures and the IMF staffs higher inflation forecast. the business cycle: official interest rates will increase
For the medium term, the fiscal projections assume a when economic indicators suggest that inflation will
more gradual fiscal consolidation than envisaged in the rise above its acceptable rate or range; they will decrease
Medium Term Programme. when indicators suggest that inflation will not exceed
United Kingdom: Fiscal projections are based on the acceptable rate or range, that output growth is
the countrys Budget 2017, published in March below its potential rate, and that the margin of slack in
2017, with expenditure projections based on the the economy is significant. On this basis, the London
budgeted nominal values and with revenue projections interbank offered rate (LIBOR) on six-month US dollar
Box A1 (continued)
deposits is assumed to average 1.4 percent in 2017 and Japan: Monetary policy assumptions are in line with
1.9 percent in 2018 (see Table 1.1). The rate on three- market expectations.
month euro deposits is assumed to average 0.3 percent Korea: Monetary policy assumptions are in line with
in 2017 and 2018. The interest rate on six-month market expectations.
Japanese yen deposits is assumed to average 0.1 percent Mexico: Monetary policy assumptions are consistent
in 2017 and 0.2 percent in 2018. with attaining the inflation target.
Australia: Monetary policy assumptions are in line Russia: Monetary projections assume that policy
with market expectations. rates will be falling over the next year or two as infla-
Brazil: Monetary policy assumptions are consistent tion continues to be close to target in the context of a
with gradual convergence of inflation toward the tight monetary stance.
middle of the target range over the relevant horizon. Saudi Arabia: Monetary policy projections are based
Canada: Monetary policy assumptions are in line on the continuation of the exchange rate peg to the
with market expectations. US dollar.
China: Monetary policy is expected to tighten with Singapore: Broad money is projected to grow in line
a gradual rise in the interest rate. with the projected growth in nominal GDP.
Denmark: The monetary policy is to maintain the Sweden: Monetary projections are in line with Riks-
peg to the euro. bank projections.
Euro area: Monetary policy assumptions for euro area Switzerland: The projections assume no change in
member countries are in line with market expectations. the policy rate in 201617.
Hong Kong Special Administrative Region: The IMF Turkey: The outlook for monetary and financial
staff assumes that the currency board system remains conditions assumes no changes to the current policy
intact. stance.
India: The policy (interest) rate assumption is con- United Kingdom: The short-term interest rate path is
sistent with an inflation rate within the Reserve Bank based on market interest rate expectations.
of Indias targeted band. United States: Following the Federal Reserves 25
Indonesia: Monetary policy assumptions are in line basis point rate hike in mid-March, the IMF staff
with the maintenance of inflation within the central expects the federal funds target rate to increase by 25
banks targeted band. more basis points in 2017 and rise gradually thereafter.
List of Tables
Output
A1. Summary of World Output
A2. Advanced Economies: Real GDP and Total Domestic Demand
A3. Advanced Economies: Components of Real GDP
A4. Emerging Market and Developing Economies: Real GDP
Inflation
A5. Summary of Inflation
A6. Advanced Economies: Consumer Prices
A7. Emerging Market and Developing Economies: Consumer Prices
Financial Policies
A8. Major Advanced Economies: General Government Fiscal Balances and Debt
Foreign Trade
A9. Summary of World Trade Volumes and Prices
Flow of Funds
A14. Summary of Net Lending and Borrowing
Table A2. Advanced Economies: Real GDP and Total Domestic Demand1
(Annual percent change)
Fourth Quarter2
Average Projections Projections
19992008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2022 2016:Q4 2017:Q4 2018:Q4
Real GDP
Advanced Economies 2.5 3.4 3.1 1.7 1.2 1.3 2.1 2.2 1.7 2.2 2.0 1.7 2.0 2.2 1.9
United States 2.6 2.8 2.5 1.6 2.2 1.7 2.6 2.9 1.5 2.2 2.3 1.7 1.8 2.3 2.3
Euro Area 2.1 4.5 2.1 1.6 0.9 0.2 1.3 2.0 1.8 2.1 1.9 1.5 1.9 2.2 1.7
Germany 1.6 5.6 3.9 3.7 0.7 0.6 1.9 1.5 1.9 2.0 1.8 1.2 1.9 2.2 1.8
France 2.0 2.9 2.0 2.1 0.2 0.6 0.9 1.1 1.2 1.6 1.8 1.8 1.2 2.1 1.4
Italy 1.2 5.5 1.7 0.6 2.8 1.7 0.1 0.8 0.9 1.5 1.1 0.8 1.2 1.5 1.0
Spain 3.6 3.6 0.0 1.0 2.9 1.7 1.4 3.2 3.2 3.1 2.5 1.7 3.0 3.1 2.1
Netherlands 2.5 3.8 1.4 1.7 1.1 0.2 1.4 2.3 2.2 3.1 2.6 1.8 2.7 3.4 1.9
Belgium 2.3 2.3 2.7 1.8 0.1 0.1 1.6 1.5 1.2 1.6 1.6 1.5 1.1 1.9 1.5
Austria 2.4 3.8 1.9 2.8 0.7 0.1 0.6 1.0 1.5 2.3 1.9 1.4 2.0 1.8 2.0
Greece 3.5 4.3 5.5 9.1 7.3 3.2 0.4 0.2 0.0 1.8 2.6 1.0 1.0 3.6 1.7
Portugal 1.6 3.0 1.9 1.8 4.0 1.1 0.9 1.6 1.4 2.5 2.0 1.2 2.0 2.0 2.3
Ireland 5.4 4.7 1.8 2.9 0.0 1.6 8.3 25.5 5.1 4.1 3.4 2.8 8.9 0.4 2.4
Finland 3.3 8.3 3.0 2.6 1.4 0.8 0.6 0.0 1.9 2.8 2.3 1.5 2.4 2.5 2.6
Slovak Republic 5.1 5.4 5.0 2.8 1.7 1.5 2.6 3.8 3.3 3.3 3.7 3.4 2.9 3.6 3.7
Lithuania 6.1 14.8 1.6 6.0 3.8 3.5 3.5 1.8 2.3 3.5 3.5 3.0 3.5 2.4 5.5
Slovenia 4.3 7.8 1.2 0.6 2.7 1.1 3.0 2.3 3.1 4.0 2.5 1.8 4.6 2.2 3.7
Luxembourg 4.3 4.4 4.9 2.5 0.4 4.0 5.6 4.0 4.2 3.9 3.6 3.0 3.9 3.9 2.9
Latvia 6.6 14.3 3.8 6.4 4.0 2.6 2.1 2.7 2.0 3.8 3.9 3.0 2.3 4.3 4.0
Estonia 5.7 14.7 2.3 7.6 4.3 1.9 2.9 1.7 2.1 4.0 3.7 3.0 3.2 3.0 4.0
Cyprus 4.1 1.8 1.3 0.3 3.2 6.0 1.5 1.7 2.8 3.4 2.6 2.2 2.9 3.1 2.6
Malta 2.2 2.4 3.5 1.4 2.6 4.6 8.2 7.1 5.5 5.1 4.4 3.2 5.9 3.9 4.4
Japan 1.0 5.4 4.2 0.1 1.5 2.0 0.3 1.1 1.0 1.5 0.7 0.6 1.7 1.4 0.5
United Kingdom 2.5 4.3 1.9 1.5 1.3 1.9 3.1 2.2 1.8 1.7 1.5 1.7 1.9 1.3 1.5
Korea 5.7 0.7 6.5 3.7 2.3 2.9 3.3 2.8 2.8 3.0 3.0 2.9 2.4 3.4 2.8
Canada 2.9 2.9 3.1 3.1 1.7 2.5 2.6 0.9 1.5 3.0 2.1 1.8 2.0 3.0 2.0
Australia 3.4 1.7 2.3 2.7 3.6 2.1 2.8 2.4 2.5 2.2 2.9 2.7 2.4 2.3 3.2
Taiwan Province of China 4.6 1.6 10.6 3.8 2.1 2.2 4.0 0.7 1.5 2.0 1.9 2.2 2.7 1.6 2.4
Switzerland 2.3 2.2 2.9 1.8 1.0 1.9 2.5 1.2 1.4 1.0 1.3 1.7 0.9 2.1 0.3
Sweden 3.0 5.2 6.0 2.7 0.3 1.2 2.6 4.1 3.2 3.1 2.4 1.7 2.1 2.7 2.6
Singapore 5.9 0.6 15.2 6.2 3.9 5.0 3.6 1.9 2.0 2.5 2.6 2.6 2.9 2.0 2.8
Hong Kong SAR 4.7 2.5 6.8 4.8 1.7 3.1 2.8 2.4 2.0 3.5 2.7 3.3 3.2 2.6 3.0
Norway 2.2 1.6 0.6 1.0 2.7 1.0 1.9 1.6 1.1 1.4 1.6 1.9 2.0 1.2 1.8
Czech Republic 4.0 4.8 2.3 1.8 0.8 0.5 2.7 5.3 2.6 3.5 2.6 2.3 1.8 3.6 3.0
Israel 3.7 1.5 5.5 5.2 2.2 4.2 3.5 2.6 4.0 3.1 3.4 3.0 4.7 2.7 3.1
Denmark 1.8 4.9 1.9 1.3 0.2 0.9 1.7 1.6 1.7 1.9 1.8 1.8 2.9 0.7 1.8
New Zealand 3.4 0.4 2.0 1.9 2.5 2.1 2.8 3.2 3.6 3.5 3.0 2.4 2.8 4.6 1.9
Puerto Rico 1.7 2.0 0.4 0.4 0.0 0.3 1.2 1.1 2.6 2.8 2.5 0.5 ... ... ...
Macao SAR ... 1.3 25.3 21.7 9.2 11.2 1.2 21.5 2.1 13.4 7.0 4.3 ... ... ...
Iceland 4.6 6.9 3.6 2.0 1.2 4.4 1.9 4.1 7.2 5.5 3.3 2.7 10.7 5.5 1.8
San Marino ... 12.8 4.6 9.5 7.5 3.0 0.9 0.5 1.0 1.2 1.3 1.3 ... ... ...
Memorandum
Major Advanced Economies 2.1 3.8 2.8 1.6 1.4 1.4 1.9 2.1 1.4 2.0 1.9 1.5 1.7 2.1 1.8
Real Total Domestic Demand
Advanced Economies 2.4 3.7 2.9 1.4 0.8 1.0 2.0 2.4 1.7 2.3 2.1 1.7 2.0 2.2 2.1
United States 2.7 3.8 2.9 1.6 2.1 1.3 2.7 3.5 1.7 2.3 2.5 1.6 2.1 2.2 2.5
Euro Area 2.0 4.0 1.5 0.7 2.4 0.6 1.3 1.9 2.3 2.1 1.9 1.5 2.4 2.4 0.9
Germany 0.9 3.2 2.9 3.0 0.8 1.0 1.3 1.5 2.4 2.1 2.0 1.5 2.3 2.0 1.7
France 2.4 2.5 2.1 2.0 0.3 0.7 1.4 1.6 1.9 1.9 1.7 1.8 1.5 1.7 1.9
Italy 1.4 4.1 2.0 0.6 5.6 2.6 0.2 1.3 1.0 1.6 1.1 0.7 1.3 1.4 1.1
Spain 4.2 6.0 0.5 3.1 5.1 3.2 1.9 3.4 2.9 2.6 2.2 1.5 2.3 2.9 1.8
Japan 0.6 4.0 2.4 0.7 2.3 2.4 0.4 0.7 0.4 1.1 0.7 0.6 0.5 1.5 0.5
United Kingdom 2.8 4.9 2.5 0.6 2.2 2.1 3.4 1.9 1.5 1.6 1.2 1.7 1.6 2.0 1.1
Canada 3.5 3.0 5.1 3.4 2.0 2.1 1.5 0.0 0.8 4.4 1.9 1.7 1.7 5.0 1.6
Other Advanced Economies3 3.7 2.6 6.1 3.1 2.0 1.5 2.6 2.5 1.9 3.1 2.6 2.6 2.0 2.7 3.3
Memorandum
Major Advanced Economies 2.1 3.7 2.8 1.4 1.1 1.3 1.9 2.3 1.5 2.1 1.9 1.5 1.8 2.1 1.9
1Inthis and other tables, when countries are not listed alphabetically, they are ordered on the basis of economic size.
2From the fourth quarter of the preceding year.
3Excludes the G7 (Canada, France, Germany, Italy, Japan, United Kingdom, United States) and euro area countries.
Table A4. Emerging Market and Developing Economies: Real GDP (continued)
(Annual percent change)
Average Projections
19992008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2022
Latin America and the Caribbean 3.3 1.8 6.1 4.7 3.0 2.9 1.2 0.1 0.9 1.2 1.9 2.7
Antigua and Barbuda 4.6 12.1 7.2 2.1 3.5 0.1 5.1 4.1 5.3 2.7 3.0 2.0
Argentina 2.6 5.9 10.1 6.0 1.0 2.4 2.5 2.6 2.2 2.5 2.5 3.2
The Bahamas 2.1 4.2 1.5 0.6 3.1 0.0 0.5 1.7 0.3 1.8 2.5 1.5
Barbados 1.8 4.0 0.3 0.8 0.3 0.1 0.1 0.9 1.6 0.9 0.5 1.6
Belize 5.7 0.8 3.3 2.1 3.7 0.7 4.1 2.9 0.8 2.5 2.3 1.7
Bolivia 3.4 3.4 4.1 5.2 5.1 6.8 5.5 4.9 4.3 4.2 4.0 3.7
Brazil 3.4 0.1 7.5 4.0 1.9 3.0 0.5 3.8 3.6 0.7 1.5 2.0
Chile 4.3 1.6 5.8 6.1 5.3 4.0 1.9 2.3 1.6 1.4 2.5 3.3
Colombia 3.4 1.7 4.0 6.6 4.0 4.9 4.4 3.1 2.0 1.7 2.8 3.6
Costa Rica 4.7 1.0 5.0 4.3 4.8 2.3 3.7 4.7 4.3 3.8 3.8 3.9
Dominica 2.8 1.2 0.7 0.2 1.1 0.6 4.4 2.5 2.6 3.9 2.8 1.5
Dominican Republic 4.8 0.9 8.3 3.1 2.8 4.7 7.6 7.0 6.6 4.8 5.8 5.0
Ecuador 3.3 0.6 3.5 7.9 5.6 4.9 4.0 0.2 1.5 0.2 0.6 1.6
El Salvador 2.6 3.1 1.4 2.2 1.9 1.8 1.4 2.3 2.4 2.3 2.1 2.0
Grenada 3.7 6.6 0.5 0.8 1.2 2.4 7.3 6.4 3.7 2.5 2.3 2.7
Guatemala 3.6 0.5 2.9 4.2 3.0 3.7 4.2 4.1 3.1 3.2 3.4 4.0
Guyana 1.8 3.3 4.4 5.4 4.8 5.2 3.8 3.1 3.3 3.5 3.6 2.8
Haiti 0.7 3.1 5.5 5.5 2.9 4.2 2.8 1.2 1.4 1.0 3.0 3.0
Honduras 4.5 2.4 3.7 3.8 4.1 2.8 3.1 3.6 3.6 4.0 3.6 3.8
Jamaica 1.3 3.4 1.4 1.4 0.5 0.2 0.5 0.9 1.3 1.7 2.3 2.8
Mexico 2.6 4.7 5.1 4.0 4.0 1.4 2.3 2.6 2.3 2.1 1.9 2.7
Nicaragua 3.9 3.3 4.4 6.3 6.5 4.9 4.8 4.9 4.7 4.5 4.3 4.5
Panama 5.7 1.6 5.8 11.8 9.2 6.6 6.1 5.8 4.9 5.3 5.6 5.5
Paraguay 2.2 4.0 13.1 4.3 1.2 14.0 4.7 3.0 4.1 3.9 4.0 3.8
Peru 5.1 1.0 8.5 6.5 6.0 5.8 2.4 3.3 4.0 2.7 3.8 3.8
St. Kitts and Nevis 3.7 1.0 2.9 0.8 0.8 6.6 5.1 4.9 3.1 2.7 3.5 2.7
St. Lucia 2.1 0.8 0.1 3.4 0.7 0.2 0.9 2.0 1.0 1.6 2.8 1.5
St. Vincent and the Grenadines 3.5 2.0 2.3 0.2 1.3 2.5 0.3 0.9 0.8 2.2 2.8 3.0
Suriname 4.1 3.0 5.2 5.8 2.7 2.9 0.4 2.7 10.5 1.2 1.2 3.1
Trinidad and Tobago 7.6 4.4 3.3 0.3 1.3 2.7 0.6 0.6 5.4 3.2 1.9 1.4
Uruguay 1.5 4.2 7.8 5.2 3.5 4.6 3.2 0.4 1.5 3.5 3.1 3.0
Venezuela 3.4 3.2 1.5 4.2 5.6 1.3 3.9 6.2 16.5 12.0 6.0 1.3
Middle East, North Africa, Afghanistan,
and Pakistan 5.2 1.1 4.7 4.5 5.2 2.7 2.8 2.7 5.0 2.6 3.5 3.8
Afghanistan ... 20.6 8.4 6.5 14.0 5.7 2.7 1.3 2.4 2.5 3.0 5.0
Algeria 4.0 1.6 3.6 2.8 3.4 2.8 3.8 3.7 3.3 1.5 0.8 2.4
Bahrain 6.0 2.5 4.3 2.0 3.7 5.4 4.4 2.9 3.0 2.5 1.7 2.2
Djibouti 3.3 1.6 4.1 7.3 4.8 5.0 6.0 6.5 6.5 7.0 7.0 6.0
Egypt 5.1 4.7 5.1 1.8 2.2 3.3 2.9 4.4 4.3 4.1 4.5 6.0
Iran 4.4 0.3 5.8 3.5 7.7 0.3 3.2 1.6 12.5 3.5 3.8 4.1
Iraq 13.0 3.4 6.4 7.5 13.9 7.6 0.7 4.8 11.0 0.4 2.9 2.1
Jordan 6.3 5.5 2.3 2.6 2.7 2.8 3.1 2.4 2.0 2.3 2.5 3.0
Kuwait 5.9 7.1 2.4 10.9 7.9 0.4 0.6 2.1 2.5 2.1 4.1 3.2
Lebanon 3.8 10.1 8.0 0.9 2.8 2.6 2.0 0.8 1.0 1.5 2.0 3.0
Libya4 4.5 3.0 3.2 66.7 124.7 36.8 53.0 10.3 3.0 55.1 31.2 2.6
Mauritania 5.2 1.0 4.8 4.7 5.8 6.1 5.6 0.9 1.7 3.8 3.0 4.0
Morocco 4.4 4.2 3.8 5.2 3.0 4.5 2.7 4.5 1.2 4.8 3.0 4.6
Oman 2.9 6.1 4.8 1.1 9.3 4.4 2.5 4.2 3.0 0.0 3.7 2.2
Pakistan 5.1 0.4 2.6 3.6 3.8 3.7 4.1 4.1 4.5 5.3 5.6 5.9
Qatar 11.3 12.0 18.1 13.4 4.7 4.4 4.0 3.6 2.2 2.5 3.1 3.2
Saudi Arabia 3.2 2.1 4.8 10.3 5.4 2.7 3.7 4.1 1.7 0.1 1.1 2.0
Somalia ... ... ... ... 1.2 2.8 3.6 3.6 3.2 2.4 3.5 3.8
Sudan6 6.2 4.7 2.5 1.2 3.0 5.2 1.6 4.9 3.0 3.7 3.6 3.5
Syria7 3.4 5.9 3.4 ... ... ... ... ... ... ... ... ...
Tunisia 4.9 3.1 2.6 1.9 3.9 2.4 2.3 1.1 1.0 2.3 3.0 4.3
United Arab Emirates 5.9 5.2 1.6 6.4 5.1 5.8 3.3 3.8 3.0 1.3 3.4 3.1
Yemen 4.1 3.9 7.7 12.7 2.4 4.8 0.2 28.1 9.8 2.0 8.5 5.5
Table A4. Emerging Market and Developing Economies: Real GDP (continued)
(Annual percent change)
Average Projections
19992008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2022
Sub-Saharan Africa 5.6 3.9 7.0 5.1 4.4 5.3 5.1 3.4 1.4 2.6 3.4 3.9
Angola 11.2 2.4 3.4 3.9 5.2 6.8 4.8 3.0 0.7 1.5 1.6 1.4
Benin 4.5 2.3 2.1 3.0 4.8 7.2 6.4 2.1 4.0 5.4 6.0 6.2
Botswana 5.2 7.7 8.6 6.0 4.5 11.3 4.1 1.7 4.3 4.5 4.8 4.2
Burkina Faso 5.7 3.0 8.4 6.6 6.5 5.7 4.2 4.0 5.9 6.4 6.5 6.0
Burundi 3.1 3.8 5.1 4.0 4.4 5.9 4.5 4.0 1.0 0.0 0.1 0.5
Cabo Verde 7.4 1.3 1.5 4.0 1.1 0.8 0.6 1.0 3.8 4.0 4.1 4.1
Cameroon 3.6 1.9 3.3 4.1 4.6 5.6 5.9 5.8 4.7 4.0 4.6 5.5
Central African Republic 1.2 1.7 3.0 3.3 4.1 36.7 1.0 4.8 4.5 4.7 5.0 5.6
Chad 7.8 4.1 13.6 0.1 8.8 5.8 6.9 1.8 6.4 0.6 2.4 3.7
Comoros 2.0 1.8 2.1 2.2 3.0 3.5 2.0 1.0 2.2 3.3 4.0 4.0
Democratic Republic of the Congo 2.4 2.9 7.1 6.9 7.1 8.5 9.5 6.9 2.4 2.8 3.0 4.7
Republic of Congo 3.5 7.8 8.7 3.4 3.8 3.3 6.8 2.6 2.8 3.6 2.8 0.3
Cte dIvoire 0.5 3.3 2.0 4.2 10.1 9.3 8.8 8.9 7.7 7.6 7.3 6.5
Equatorial Guinea 28.1 1.3 8.9 6.5 8.3 4.1 0.7 9.1 9.7 7.4 7.8 1.4
Eritrea 1.1 3.9 2.2 8.7 7.0 3.1 5.0 4.8 3.7 3.3 3.6 4.0
Ethiopia 8.1 10.0 10.6 11.4 8.7 9.9 10.3 10.4 8.0 8.5 8.5 7.5
Gabon 0.1 2.3 6.3 7.1 5.3 5.5 4.4 3.9 2.1 1.0 2.7 5.1
The Gambia 3.7 6.4 6.5 4.3 5.6 4.8 0.9 4.3 2.2 3.0 3.5 4.8
Ghana 5.3 4.8 7.9 14.0 9.3 7.3 4.0 3.8 3.5 5.9 8.9 5.4
Guinea 3.5 1.5 4.2 5.6 5.9 3.9 3.7 3.5 6.6 6.7 5.8 5.2
Guinea-Bissau 2.9 3.4 4.6 8.1 1.7 3.3 1.0 5.1 5.1 5.0 5.0 5.0
Kenya 3.3 3.3 8.4 6.1 4.6 5.9 5.4 5.7 5.8 5.0 5.5 6.5
Lesotho 3.5 4.5 6.9 4.5 5.3 3.6 3.4 2.5 2.4 4.6 3.1 5.6
Liberia ... 5.1 6.1 7.4 8.2 8.7 0.7 0.0 1.6 2.6 4.0 6.8
Madagascar 4.0 4.7 0.3 1.5 3.0 2.3 3.3 3.1 4.2 4.3 5.3 5.0
Malawi 3.8 8.3 6.9 4.9 1.9 5.2 5.7 2.9 2.3 4.5 5.0 5.5
Mali 5.4 4.7 5.4 3.2 0.8 2.3 7.0 6.0 5.8 5.3 5.0 4.7
Mauritius 4.3 3.0 4.1 3.9 3.2 3.2 3.6 3.5 3.9 3.9 4.0 4.1
Mozambique 7.8 6.4 6.7 7.1 7.2 7.1 7.4 6.6 3.8 4.7 5.3 14.0
Namibia 4.1 0.3 6.0 5.1 5.1 5.6 6.4 6.0 1.1 0.8 2.5 3.6
Niger 4.4 0.7 8.4 2.2 11.8 5.3 7.5 4.0 5.0 4.2 4.7 6.2
Nigeria 7.5 8.4 11.3 4.9 4.3 5.4 6.3 2.7 1.6 0.8 1.9 1.7
Rwanda 8.0 6.3 7.3 7.8 8.8 4.7 7.6 8.9 5.9 6.2 6.8 7.5
So Tom and Prncipe 4.3 4.0 4.5 4.8 4.5 4.3 4.1 4.0 4.1 5.0 5.5 5.5
Senegal 4.4 2.4 4.3 1.9 4.5 3.6 4.1 6.5 6.7 6.8 7.0 6.4
Seychelles 2.2 1.1 5.9 5.4 3.7 6.0 4.5 5.0 4.5 4.1 3.4 4.0
Sierra Leone 7.5 3.2 5.3 6.3 15.2 20.7 4.6 20.5 6.1 6.0 6.1 7.4
South Africa 4.0 1.5 3.0 3.3 2.2 2.5 1.7 1.3 0.3 0.7 1.1 2.2
South Sudan ... ... ... ... 52.4 29.3 2.9 0.2 13.8 6.3 3.4 3.9
Swaziland 3.6 4.5 3.5 2.0 3.5 4.8 3.6 1.1 0.0 0.3 0.9 2.2
Tanzania 6.1 5.4 6.4 7.9 5.1 7.3 7.0 7.0 7.0 6.5 6.8 6.6
Togo 1.6 3.5 4.1 4.8 5.9 6.1 5.4 5.3 5.0 5.0 5.3 5.6
Uganda 7.5 8.1 7.7 6.8 2.2 4.7 4.6 5.7 2.3 4.4 5.2 7.3
Zambia 6.4 9.2 10.3 5.6 7.6 5.1 4.7 2.9 3.4 4.0 4.5 4.5
Zimbabwe8 6.8 7.4 15.4 16.3 13.6 5.3 2.8 1.4 0.7 2.8 0.8 0.9
1Data for some countries refer to real net material product (NMP) or are estimates based on NMP. The figures should be interpreted only as indicative of broad orders of magnitude because
reliable, comparable data are not generally available. In particular, the growth of output of new private enterprises of the informal economy is not fully reflected in the recent figures.
2Georgia, Turkmenistan, and Ukraine, which are not members of the Commonwealth of Independent States, are included in this group for reasons of geography and similarity in economic
structure.
3Data are based on the 2008 System of National Accounts. The revised national accounts data are available beginning in 2000 and exclude Crimea and Sevastopol from 2010 onward.
4See country-specific notes for India and Libya in the Country Notes section of the Statistical Appendix.
5In this table only, the data for Timor-Leste are based on non-oil GDP.
6Data for 2011 exclude South Sudan after July 9. Data for 2012 and onward pertain to the current Sudan.
7Data for Syria are excluded for 2011 onward owing to the uncertain political situation.
8The Zimbabwe dollar ceased circulating in early 2009. Data are based on IMF staff estimates of price and exchange rate developments in US dollars. IMF staff estimates of US dollar values
may differ from authorities estimates. Real GDP is in constant 2009 prices.
Appendix.
Table A7. Emerging Market and Developing Economies: Consumer Prices1 (continued)
(Annual percent change)
End of Period2
Average Projections Projections
19992008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2022 2016 2017 2018
Latin America and the Caribbean7 6.7 4.6 4.2 5.2 4.6 4.6 4.9 5.5 5.6 4.2 3.6 3.4 4.6 4.2 3.6
Antigua and Barbuda 2.0 0.6 3.4 3.5 3.4 1.1 1.1 1.0 0.5 2.4 1.2 2.0 1.1 2.5 2.0
Argentina8 7.6 6.3 10.5 9.8 10.0 10.6 ... ... ... 26.9 17.8 8.6 ... 22.3 16.7
The Bahamas 2.2 1.7 1.6 3.1 1.9 0.4 1.2 1.9 0.8 2.4 2.2 2.2 0.8 2.4 2.2
Barbados 3.5 3.6 5.8 9.4 4.5 1.8 1.8 1.1 1.3 5.0 5.8 2.7 3.2 6.7 2.4
Belize 2.5 1.1 0.9 1.7 1.2 0.5 1.2 0.9 0.6 1.8 2.3 2.0 1.1 2.4 2.3
Bolivia 4.7 3.3 2.5 9.9 4.5 5.7 5.8 4.1 3.6 3.2 5.1 5.0 4.0 4.3 5.0
Brazil 6.8 4.9 5.0 6.6 5.4 6.2 6.3 9.0 8.7 3.7 4.0 4.0 6.3 3.6 4.0
Chile 3.7 1.5 1.4 3.3 3.0 1.9 4.4 4.3 3.8 2.3 2.7 3.0 2.8 2.4 2.9
Colombia 6.9 4.2 2.3 3.4 3.2 2.0 2.9 5.0 7.5 4.3 3.3 3.0 5.7 4.0 3.1
Costa Rica 11.1 7.8 5.7 4.9 4.5 5.2 4.5 0.8 0.0 1.7 2.9 3.0 0.8 2.7 3.0
Dominica 2.1 0.0 2.8 1.1 1.4 0.0 0.8 0.8 0.0 0.6 1.4 2.0 0.2 1.4 1.4
Dominican Republic 12.8 1.4 6.3 8.5 3.7 4.8 3.0 0.8 1.6 3.0 3.3 4.0 1.7 2.9 4.2
Ecuador 19.6 5.2 3.6 4.5 5.1 2.7 3.6 4.0 1.7 0.7 0.7 1.6 1.1 0.8 0.7
El Salvador 3.5 0.5 1.2 5.1 1.7 0.8 1.1 0.7 0.6 0.8 2.2 2.0 0.9 2.4 2.0
Grenada 2.9 0.3 3.4 3.0 2.4 0.0 1.0 0.6 1.7 2.6 2.0 1.9 0.9 3.0 1.8
Guatemala 7.3 1.9 3.9 6.2 3.8 4.3 3.4 2.4 4.4 4.4 3.5 4.0 4.2 4.3 4.0
Guyana 6.6 3.0 4.3 4.4 2.4 1.9 0.7 0.9 0.8 2.3 2.7 3.1 1.5 2.6 2.7
Haiti 15.3 3.4 4.1 7.4 6.8 6.8 3.9 7.5 13.4 14.7 9.0 5.0 12.5 15.3 5.0
Honduras 8.8 5.5 4.7 6.8 5.2 5.2 6.1 3.2 2.7 4.0 4.0 4.0 3.3 4.5 4.0
Jamaica 10.6 9.6 12.6 7.5 6.9 9.4 8.3 3.7 2.3 3.4 5.2 5.4 1.7 5.0 5.5
Mexico 6.3 5.3 4.2 3.4 4.1 3.8 4.0 2.7 2.8 5.9 3.8 3.0 3.4 6.1 3.5
Nicaragua 9.7 3.7 5.5 8.1 7.2 7.1 6.0 4.0 3.5 4.0 7.2 7.3 3.1 4.0 7.2
Panama 2.3 2.4 3.5 5.9 5.7 4.0 2.6 0.1 0.7 1.6 2.1 2.4 1.5 2.5 2.1
Paraguay 8.6 2.6 4.7 8.3 3.7 2.7 5.0 3.1 4.1 3.5 4.0 4.0 3.9 4.0 4.0
Peru 2.6 2.9 1.5 3.4 3.7 2.8 3.2 3.5 3.6 3.2 2.3 2.0 3.2 2.7 2.5
St. Kitts and Nevis 3.6 2.1 0.9 5.8 0.8 1.1 0.2 2.3 0.4 1.2 1.8 2.0 0.9 1.5 2.0
St. Lucia 3.0 0.2 3.3 2.8 4.2 1.5 3.5 1.0 3.1 0.2 0.9 1.5 3.0 1.4 1.2
St. Vincent and the Grenadines 2.9 0.4 0.8 3.2 2.6 0.8 0.2 1.7 0.2 1.7 1.4 1.5 1.0 1.9 1.5
Suriname 21.0 0.3 6.9 17.7 5.0 1.9 3.4 6.9 55.5 22.3 9.3 4.1 52.4 9.1 12.3
Trinidad and Tobago 5.9 7.0 10.5 5.1 9.3 5.2 5.7 4.7 3.1 3.2 3.2 3.2 3.1 3.2 3.2
Uruguay 8.3 7.1 6.7 8.1 8.1 8.6 8.9 8.7 9.6 6.1 6.3 6.1 8.1 6.2 6.7
Venezuela8 20.5 27.1 28.2 26.1 21.1 43.5 57.3 111.8 254.4 652.7 2,349.3 4,684.8 302.6 1,133.0 2,529.6
Middle East, North Africa,
Afghanistan, and Pakistan 6.4 7.3 6.6 9.2 9.8 9.2 6.8 5.7 5.1 6.8 7.7 4.9 6.2 7.5 6.5
Afghanistan ... 6.8 2.2 11.8 6.4 7.4 4.7 0.7 4.4 6.0 6.0 6.0 4.6 7.2 6.0
Algeria 2.9 5.7 3.9 4.5 8.9 3.3 2.9 4.8 6.4 5.5 4.4 4.0 7.0 5.5 4.4
Bahrain 1.2 2.8 2.0 0.4 2.8 3.3 2.7 1.8 2.8 0.9 3.5 1.6 2.3 1.0 3.2
Djibouti 3.2 1.7 4.0 5.1 3.7 2.4 2.9 2.1 2.7 3.0 3.0 3.0 2.4 3.0 3.0
Egypt 5.8 16.2 11.7 11.1 8.6 6.9 10.1 11.0 10.2 23.5 21.3 7.1 14.0 29.8 11.7
Iran 15.6 10.7 12.4 21.2 30.8 34.7 15.6 11.9 9.0 10.5 10.1 8.7 11.9 10.1 9.7
Iraq ... 2.2 2.4 5.6 6.1 1.9 2.2 1.4 0.4 2.0 2.0 2.0 1.0 2.0 2.0
Jordan 3.8 0.7 4.8 4.2 4.5 4.8 2.9 0.9 0.8 3.3 1.5 2.5 0.8 2.5 2.5
Kuwait 2.8 4.6 4.5 4.9 3.2 2.7 3.1 3.7 3.5 2.5 2.7 2.7 3.5 2.5 2.7
Lebanon 2.3 1.2 4.0 5.0 6.6 4.8 1.9 3.7 0.8 3.1 2.5 2.0 3.1 3.0 2.0
Libya8 0.1 2.4 2.5 15.9 6.1 2.6 2.4 9.8 27.1 32.8 32.1 23.5 29.9 35.1 29.9
Mauritania 6.4 2.1 6.3 5.7 4.9 4.1 3.8 0.5 1.5 2.1 3.7 4.0 2.8 1.6 4.7
Morocco 1.9 1.0 1.0 0.9 1.3 1.9 0.4 1.5 1.6 0.9 1.6 2.0 1.8 1.1 1.6
Oman 2.2 3.5 3.3 4.0 2.9 1.2 1.0 0.1 1.1 3.2 3.2 3.1 1.1 3.2 3.2
Pakistan 6.2 19.6 10.1 13.7 11.0 7.4 8.6 4.5 2.9 4.1 4.8 5.0 3.2 3.9 5.0
Qatar 6.3 4.9 2.4 2.0 1.8 3.2 3.4 1.8 2.7 0.9 4.8 2.3 ... ... ...
Saudi Arabia 1.0 4.1 3.8 3.7 2.9 3.5 2.7 2.2 3.5 0.2 5.0 2.0 1.7 0.2 5.0
Somalia ... ... ... ... ... ... ... ... ... ... ... ... 2.3 2.9 2.7
Sudan9 9.1 11.3 13.0 18.3 35.4 36.5 36.9 16.9 17.8 26.9 19.0 14.0 30.5 21.0 17.0
Syria10 4.1 2.8 4.4 ... ... ... ... ... ... ... ... ... ... ... ...
Tunisia 2.7 3.7 3.3 3.5 5.1 5.8 4.9 4.9 3.7 4.5 4.4 3.5 4.2 4.5 4.1
United Arab Emirates 5.6 1.6 0.9 0.9 0.7 1.1 2.3 4.1 1.8 2.1 2.9 1.9 1.8 2.1 2.9
Yemen 11.4 3.7 11.2 19.5 9.9 11.0 8.2 39.4 5.0 20.0 29.5 9.0 22.0 23.0 24.0
Table A7. Emerging Market and Developing Economies: Consumer Prices1 (continued)
(Annual percent change)
End of Period2
Average Projections Projections
19992008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2022 2016 2017 2018
Sub-Saharan Africa 10.5 9.8 8.1 9.4 9.3 6.6 6.3 7.0 11.3 11.0 9.5 7.8 12.5 10.4 9.2
Angola 81.6 13.7 14.5 13.5 10.3 8.8 7.3 10.3 32.4 30.9 20.6 9.5 41.9 23.4 17.6
Benin 3.1 0.4 2.2 2.7 6.7 1.0 1.1 0.3 0.8 2.0 2.1 2.0 2.7 2.2 2.0
Botswana 8.7 8.1 6.9 8.5 7.5 5.9 4.4 3.1 2.8 3.7 3.7 3.9 3.0 4.4 3.0
Burkina Faso 2.6 0.9 0.6 2.8 3.8 0.5 0.3 0.9 0.2 1.5 2.0 2.0 1.6 2.0 2.0
Burundi 10.0 10.6 6.5 9.6 18.2 7.9 4.4 5.6 5.5 18.0 20.2 17.7 9.5 18.6 21.5
Cabo Verde 2.3 1.0 2.1 4.5 2.5 1.5 0.2 0.1 1.4 1.0 1.5 2.0 0.3 1.1 1.7
Cameroon 2.4 3.0 1.3 2.9 2.4 2.1 1.9 2.7 0.9 0.7 1.1 2.0 0.3 1.2 1.1
Central African Republic 2.9 3.5 1.5 1.2 5.9 6.6 11.6 4.5 4.6 3.8 3.7 3.0 4.7 3.6 3.6
Chad 1.6 10.1 2.1 1.9 7.7 0.2 1.7 6.8 1.1 0.2 1.9 3.0 4.9 0.7 2.3
Comoros 4.0 4.8 3.9 2.2 5.9 1.6 1.3 2.0 1.8 2.0 2.0 2.0 0.8 1.9 2.1
Democratic Republic of the Congo 77.9 46.1 23.5 14.9 0.9 0.9 1.2 1.0 18.2 41.7 44.0 10.0 23.6 50.0 40.0
Republic of Congo 2.7 4.3 0.4 1.8 5.0 4.6 0.9 2.7 3.6 0.4 1.1 2.4 0.8 1.3 0.9
Cte dIvoire 3.0 1.0 1.4 4.9 1.3 2.6 0.4 1.2 0.7 1.0 2.0 2.0 1.1 1.5 2.0
Equatorial Guinea 5.0 5.7 5.3 4.8 3.4 3.2 4.3 1.7 1.4 1.7 1.8 2.1 1.6 1.7 1.9
Eritrea 16.3 33.0 11.2 3.9 6.0 6.5 10.0 9.0 9.0 9.0 9.0 9.0 9.0 9.0 9.0
Ethiopia 10.2 8.5 8.1 33.2 24.1 8.1 7.4 10.1 7.3 8.1 8.0 7.5 6.7 9.2 7.5
Gabon 0.7 1.9 1.4 1.3 2.7 0.5 4.5 0.1 2.1 2.5 2.5 2.5 4.1 2.5 2.5
The Gambia 6.5 4.6 5.0 4.8 4.6 5.2 6.3 6.8 7.2 8.3 7.1 4.7 7.9 7.6 6.4
Ghana 17.7 13.1 6.7 7.7 7.1 11.7 15.5 17.2 17.5 11.8 9.0 6.0 15.4 10.0 8.0
Guinea 15.1 4.7 15.5 21.4 15.2 11.9 9.7 8.2 8.2 8.5 8.2 7.9 8.7 8.2 8.0
Guinea-Bissau 3.0 1.6 1.1 5.1 2.1 0.8 1.0 1.5 1.5 2.8 2.5 2.5 1.6 2.5 2.5
Kenya 6.8 10.6 4.3 14.0 9.4 5.7 6.9 6.6 6.3 8.0 5.2 5.0 6.3 5.1 5.2
Lesotho 7.5 5.8 3.3 6.0 5.5 5.0 4.6 4.3 6.4 6.6 6.0 5.0 4.4 6.5 6.0
Liberia ... 7.4 7.3 8.5 6.8 7.6 9.9 7.7 8.8 12.8 9.9 7.1 12.5 12.4 9.1
Madagascar 10.3 9.0 9.2 9.5 5.7 5.8 6.1 7.4 6.7 7.8 6.8 5.0 7.0 7.7 6.8
Malawi 17.4 8.4 7.4 7.6 21.3 28.3 23.8 21.9 21.7 13.0 9.6 3.9 20.0 11.1 8.3
Mali 2.2 2.2 1.3 3.1 5.3 0.6 0.9 1.4 1.8 0.2 1.2 2.2 0.8 1.0 1.4
Mauritius 6.4 2.5 2.9 6.5 3.9 3.5 3.2 1.3 1.0 4.2 5.0 3.1 2.3 5.0 4.0
Mozambique 10.5 3.3 12.7 10.4 2.1 4.2 2.3 2.4 19.2 17.5 10.5 5.5 21.1 14.0 8.0
Namibia 7.6 9.5 4.9 5.0 6.7 5.6 5.3 3.4 6.7 6.0 5.8 5.8 7.3 6.0 5.8
Niger 2.4 4.3 2.8 2.9 0.5 2.3 0.9 1.0 0.3 1.0 2.1 2.0 2.4 2.0 2.0
Nigeria 11.6 12.5 13.7 10.8 12.2 8.5 8.0 9.0 15.7 16.3 14.8 14.5 18.5 16.0 15.1
Rwanda 6.8 10.3 2.3 5.7 6.3 4.2 1.8 2.5 5.7 7.1 6.0 5.0 7.3 7.0 5.0
So Tom and Prncipe 15.3 17.0 13.3 14.3 10.6 8.1 7.0 5.3 5.4 4.5 5.2 3.2 5.1 5.5 5.0
Senegal 2.3 2.2 1.2 3.4 1.4 0.7 1.1 0.1 0.9 2.1 2.2 2.2 2.1 2.0 2.2
Seychelles 6.3 31.8 2.4 2.6 7.1 4.3 1.4 4.0 1.0 2.8 2.3 3.0 0.2 3.0 3.2
Sierra Leone 9.8 9.2 17.8 18.5 13.8 9.8 8.3 9.0 11.5 16.9 10.6 7.7 17.4 12.0 9.5
South Africa 5.8 7.1 4.3 5.0 5.6 5.8 6.1 4.6 6.3 5.4 5.3 5.5 6.7 5.2 5.4
South Sudan ... ... ... ... 45.1 0.0 1.7 52.8 379.8 182.2 45.0 7.5 479.7 111.4 25.0
Swaziland 7.4 7.4 4.5 6.1 8.9 5.6 5.7 5.0 8.0 7.0 5.4 5.5 9.0 6.5 4.4
Tanzania 6.1 12.1 7.2 12.7 16.0 7.9 6.1 5.6 5.2 5.4 5.0 5.0 5.0 5.0 5.0
Togo 2.6 3.7 1.4 3.6 2.6 1.8 0.2 1.8 0.9 0.8 1.2 2.0 0.5 ... ...
Uganda 5.7 13.0 3.7 15.0 12.7 4.9 3.1 5.4 5.5 5.8 5.6 5.0 5.7 5.9 5.3
Zambia 18.5 13.4 8.5 8.7 6.6 7.0 7.8 10.1 17.9 6.8 7.4 8.0 7.5 5.8 8.0
Zimbabwe11 7.4 6.2 3.0 3.5 3.7 1.6 0.2 2.4 1.6 2.5 9.5 4.0 0.9 7.0 10.0
1Movements in consumer prices are shown as annual averages.
2Monthly year-over-year changes and, for several countries, on a quarterly basis.
3For many countries, inflation for the earlier years is measured on the basis of a retail price index. Consumer price index (CPI) inflation data with broader and more up-to-date coverage are
structure.
5Starting in 2014 data exclude Crimea and Sevastopol.
6Based on Eurostats harmonized index of consumer prices.
7Excludes Argentina and Venezuela.
8See country-specific notes for Argentina, Libya, and Venezuela in the Country Notes section of the Statistical Appendix.
9Data for 2011 exclude South Sudan after July 9. Data for 2012 and onward pertain to the current Sudan.
10Data for Syria are excluded for 2011 onward owing to the uncertain political situation.
11The Zimbabwe dollar ceased circulating in early 2009. Data are based on IMF staff estimates of price and exchange rate developments in US dollars. IMF staff estimates of US dollar values may
Table A8. Major Advanced Economies: General Government Fiscal Balances and Debt1
(Percent of GDP unless noted otherwise)
Average Projections
19992008 2011 2012 2013 2014 2015 2016 2017 2018 2022
Major Advanced Economies
Net Lending/Borrowing 3.4 7.3 6.3 4.2 3.6 3.0 3.5 3.4 2.9 2.7
Output Gap2 0.9 2.4 2.1 1.9 1.4 0.7 0.7 0.1 0.2 0.3
Structural Balance2 3.8 6.3 5.1 3.7 3.1 2.8 3.1 3.3 2.9 2.8
United States
Net Lending/Borrowing3 3.5 9.6 7.9 4.4 4.0 3.5 4.4 4.3 3.7 4.3
Output Gap2 1.8 3.1 2.2 1.9 1.1 0.0 0.1 0.3 0.7 0.6
Structural Balance2 4.0 8.2 6.4 4.4 3.8 3.6 4.1 4.4 4.0 4.5
Net Debt 43.2 76.8 80.2 81.6 80.8 80.2 81.3 82.5 81.1 82.8
Gross Debt 62.6 100.0 103.4 105.4 105.1 105.2 107.1 108.1 107.8 109.6
Euro Area
Net Lending/Borrowing 2.0 4.2 3.6 3.0 2.6 2.1 1.5 1.3 1.0 0.1
Output Gap2 0.9 0.5 1.9 2.8 2.4 1.9 1.3 0.5 0.0 0.6
Structural Balance2 2.5 3.9 2.1 1.3 1.1 0.9 0.8 0.9 0.9 0.4
Net Debt 54.5 68.5 72.2 74.6 74.9 73.9 73.3 71.8 70.3 62.7
Gross Debt 67.9 86.1 89.5 91.4 91.9 90.0 89.0 87.4 85.6 76.3
Germany
Net Lending/Borrowing 2.1 1.0 0.0 0.1 0.3 0.6 0.8 0.7 0.8 1.1
Output Gap2 0.1 1.0 0.5 0.3 0.1 0.1 0.4 0.8 1.0 0.8
Structural Balance2 2.2 1.3 0.2 0.1 0.5 0.6 0.6 0.3 0.2 0.7
Net Debt 50.3 58.7 58.2 57.0 53.5 50.5 48.3 45.8 43.2 33.7
Gross Debt 62.6 78.7 79.9 77.5 74.7 70.9 68.1 65.0 61.8 50.1
France
Net Lending/Borrowing 2.6 5.1 4.8 4.0 3.9 3.6 3.4 3.0 3.0 0.8
Output Gap2 0.5 1.1 1.9 2.4 2.5 2.4 2.2 1.8 1.3 0.3
Structural Balance2 3.0 4.4 3.4 2.4 2.3 2.0 1.9 1.8 2.2 1.0
Net Debt 54.6 76.9 80.6 83.5 86.1 86.9 87.8 88.5 88.7 82.9
Gross Debt 63.1 85.2 89.5 92.3 94.9 95.6 96.3 96.8 97.0 91.2
Italy
Net Lending/Borrowing 2.9 3.7 2.9 2.9 3.0 2.7 2.4 2.2 1.3 0.0
Output Gap2 0.2 0.5 2.8 4.1 4.1 3.3 2.7 1.6 1.0 0.0
Structural Balance2,4 3.6 4.1 1.5 0.6 1.1 0.9 1.1 1.4 0.8 0.0
Net Debt 94.7 106.8 111.6 116.7 118.8 119.8 120.6 121.2 119.9 109.6
Gross Debt 102.9 116.5 123.4 129.0 131.8 132.1 132.6 133.0 131.4 120.1
Japan
Net Lending/Borrowing 5.5 9.1 8.3 7.6 5.4 3.5 4.2 4.1 3.3 2.1
Output Gap2 0.8 4.6 3.7 2.2 2.6 2.1 1.8 0.9 0.7 0.7
Structural Balance2 5.5 7.5 7.1 7.1 5.1 3.9 3.8 3.9 3.2 2.0
Net Debt 64.2 117.9 120.5 117.4 119.0 118.4 119.8 120.9 120.7 114.6
Gross Debt5 165.8 230.6 236.6 240.5 242.1 238.1 239.3 240.3 240.0 233.9
United Kingdom
Net Lending/Borrowing 1.9 7.5 7.7 5.5 5.6 4.3 2.9 2.9 2.3 1.2
Output Gap2 0.8 2.6 2.6 2.1 0.7 0.2 0.1 0.1 0.2 0.0
Structural Balance2 2.5 5.4 5.7 3.9 4.8 4.1 2.8 2.8 2.2 1.2
Net Debt 34.9 73.2 76.4 77.8 79.7 80.3 80.1 80.5 80.6 76.6
Gross Debt 39.5 81.6 85.1 86.2 88.1 89.0 89.3 89.5 89.7 85.6
Canada
Net Lending/Borrowing 1.1 3.3 2.5 1.5 0.0 1.1 1.9 2.2 1.8 1.1
Output Gap2 1.6 1.0 1.3 0.9 0.5 1.2 1.5 0.2 0.4 0.1
Structural Balance2 0.3 2.8 1.8 1.0 0.0 0.5 1.1 2.1 2.1 1.2
Net Debt 34.1 27.1 28.2 29.0 27.2 25.2 27.4 24.6 22.7 14.9
Gross Debt 75.6 81.5 84.8 85.8 85.4 91.6 92.4 89.6 87.7 79.9
Note: The methodology and specific assumptions for each country are discussed in Box A1. The country group composites for fiscal data are calculated as the sum of the US dollar values
for the relevant individual countries.
1Debt data refer to the end of the year and are not always comparable across countries. Gross and net debt levels reported by national statistical agencies for countries that have adopted
the System of National Accounts (SNA) 2008 (Australia, Canada, Hong Kong SAR, United States) are adjusted to exclude unfunded pension liabilities of government employees
defined-benefit pension plans. Fiscal data for the aggregated major advanced economies and the United States start in 2001, and the average for the aggregate and the United States is
therefore for the period 200107.
2Percent of potential GDP.
3Figures reported by the national statistical agency are adjusted to exclude items related to the accrual-basis accounting of government employees defined-benefit pension plans.
4Excludes one-time measures based on the authorities data and, if unavailable, on receipts from the sale of assets.
5Includes equity shares; nonconsolidated basis.
Table A12. Emerging Market and Developing Economies: Balance on Current Account
(Percent of GDP)
Projections
2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2022
Commonwealth of Independent States1 2.5 3.2 4.0 2.4 0.6 2.1 2.8 0.0 0.9 1.3 2.2
Russia 3.8 4.1 4.7 3.2 1.5 2.8 5.0 2.0 2.8 3.2 4.0
Excluding Russia 1.8 0.2 1.7 0.6 2.3 0.1 3.0 5.6 4.6 4.0 2.2
Armenia 16.5 13.6 10.4 10.0 7.3 7.6 2.6 2.3 3.6 3.2 5.1
Azerbaijan 23.0 28.0 26.5 21.9 16.1 13.3 0.4 3.6 1.9 2.5 5.4
Belarus 12.5 14.5 8.2 2.8 10.0 6.6 3.3 3.6 5.3 4.6 2.5
Georgia 10.5 10.3 12.8 11.7 5.8 10.7 12.0 13.3 11.9 10.7 9.1
Kazakhstan 3.6 0.9 5.3 0.5 0.5 2.8 2.8 6.4 5.3 3.8 0.5
Kyrgyz Republic 0.9 2.2 2.9 3.7 13.3 16.0 16.0 9.7 11.6 12.0 9.7
Moldova 8.2 7.5 11.7 7.5 5.2 5.3 5.0 3.8 4.0 4.0 5.4
Tajikistan 3.6 9.6 7.3 9.2 7.8 2.8 6.0 3.8 6.3 6.2 4.3
Turkmenistan 16.6 12.9 0.8 0.9 7.3 6.4 14.0 21.0 15.4 14.3 11.9
Ukraine 2 1.4 2.2 6.3 8.1 9.2 3.9 0.3 4.1 3.3 3.0 3.0
Uzbekistan 2.6 6.6 5.8 1.2 2.9 1.7 0.7 0.7 0.9 0.3 2.0
Emerging and Developing Asia 3.4 2.4 0.9 1.0 0.7 1.5 2.0 1.4 0.9 0.7 0.4
Bangladesh 2.4 0.4 1.0 0.7 1.2 1.2 1.6 0.6 0.7 1.3 2.4
Bhutan 6.3 22.2 29.8 21.5 25.4 26.4 28.3 29.1 29.4 16.6 10.0
Brunei Darussalam 32.3 36.6 34.7 29.8 20.9 30.7 16.0 9.6 4.8 2.1 14.1
Cambodia 9.9 9.3 5.9 8.2 13.0 9.8 9.3 8.8 8.6 8.6 8.0
China 4.7 3.9 1.8 2.5 1.5 2.2 2.7 1.7 1.4 1.2 0.2
Fiji 4.0 4.3 5.1 1.4 9.7 7.6 1.5 5.1 5.0 4.7 3.3
India 2.8 2.8 4.2 4.8 1.7 1.3 1.1 0.7 1.4 1.5 2.4
Indonesia 1.8 0.7 0.2 2.7 3.2 3.1 2.0 1.8 1.7 1.8 1.9
Kiribati 13.3 2.2 13.5 4.5 8.4 25.1 35.2 15.4 4.6 5.9 2.9
Lao P.D.R. 17.0 15.6 14.3 24.9 26.7 18.3 16.5 10.6 9.6 10.9 4.8
Malaysia 15.0 10.1 10.9 5.2 3.5 4.4 3.0 2.4 2.4 2.2 1.8
Maldives 9.6 7.3 14.8 6.6 4.3 3.2 7.3 19.6 17.2 17.0 10.3
Marshall Islands 10.3 20.9 2.0 0.1 5.3 1.9 16.5 8.5 5.9 4.5 0.7
Micronesia 19.0 15.4 18.8 13.4 10.1 1.2 4.2 3.2 3.4 3.0 2.9
Mongolia 6.0 13.0 26.5 27.4 25.4 11.5 4.0 6.3 4.9 8.7 4.1
Myanmar 1.2 1.1 1.8 4.0 4.9 3.3 5.2 5.9 6.6 6.6 6.4
Nauru 63.8 46.3 26.1 38.1 18.8 13.5 9.5 1.7 0.7 1.3 2.1
Nepal 4.2 2.4 1.0 4.8 3.3 4.5 5.0 6.3 0.4 0.7 1.9
Palau 9.9 9.0 11.8 11.2 11.6 15.0 7.7 10.3 12.1 14.4 12.3
Papua New Guinea 8.3 20.4 24.0 36.1 31.3 2.9 19.8 20.1 18.6 17.3 16.5
Philippines 5.0 3.6 2.5 2.8 4.2 3.8 2.5 0.2 0.1 0.3 1.0
Samoa 4.9 7.0 4.3 6.3 0.4 8.1 3.0 6.1 5.7 5.0 4.0
Solomon Islands 21.9 32.9 8.3 1.7 3.4 4.3 3.0 3.9 5.0 5.0 6.0
Sri Lanka 0.4 1.9 7.1 5.8 3.4 2.5 2.4 2.4 2.5 2.3 2.0
Thailand 7.9 3.4 2.5 0.4 1.2 3.7 8.1 11.5 10.1 8.1 2.9
Timor-Leste 40.4 42.0 41.4 41.0 42.3 27.0 7.7 19.3 5.6 15.9 17.0
Tonga 19.7 19.0 16.8 12.3 8.3 10.7 14.7 12.8 13.5 13.6 11.4
Tuvalu 6.9 42.2 63.6 36.4 22.3 23.1 27.0 34.0 37.7 39.0 27.8
Vanuatu 7.9 5.4 8.1 6.5 3.3 0.3 10.5 3.7 14.4 13.6 4.7
Vietnam 6.5 3.8 0.2 6.0 4.5 4.9 0.1 4.1 1.3 1.4 0.0
Emerging and Developing Europe 3.4 5.0 6.3 4.4 3.6 2.9 2.0 1.8 2.4 2.5 2.8
Albania 15.9 11.3 13.2 10.1 9.3 10.8 8.6 7.6 9.2 8.2 6.2
Bosnia and Herzegovina 6.4 6.1 9.5 8.7 5.3 7.4 5.5 4.5 4.3 4.2 4.9
Bulgaria 8.3 1.7 0.3 0.9 1.3 0.1 0.1 4.2 2.5 1.9 0.4
Croatia 5.1 1.1 0.7 0.0 1.0 2.1 4.8 2.6 3.8 3.0 0.5
Hungary 0.8 0.3 0.7 1.8 3.8 2.1 3.4 5.5 4.8 4.2 1.4
Kosovo 9.2 11.6 12.7 5.8 3.6 7.0 8.5 9.8 11.0 11.3 10.0
FYR Macedonia 6.8 2.0 2.5 3.2 1.6 0.5 2.1 3.1 2.3 2.5 3.1
Montenegro 27.9 22.7 17.6 18.5 14.5 15.2 13.3 19.0 20.2 21.2 14.0
Poland 4.1 5.4 5.2 3.7 1.3 2.1 0.6 0.2 1.0 1.2 2.2
Romania 4.8 5.1 4.9 4.8 1.1 0.7 1.2 2.3 3.0 2.9 2.9
Serbia 6.2 6.4 8.6 11.5 6.1 6.0 4.7 4.0 4.0 3.9 3.7
Turkey 1.8 5.8 8.9 5.5 6.7 4.7 3.7 3.8 4.6 4.6 3.8
Table A12. Emerging Market and Developing Economies: Balance on Current Account (continued)
(Percent of GDP)
Projections
2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2022
Latin America and the Caribbean 0.9 1.9 2.0 2.3 2.8 3.1 3.4 2.0 2.0 2.3 2.6
Antigua and Barbuda 13.8 14.5 10.3 14.8 15.1 2.0 6.8 0.2 1.4 0.8 0.1
Argentina 2.2 0.4 1.0 0.4 2.1 1.5 2.7 2.7 3.6 3.7 4.3
The Bahamas 10.3 10.1 14.1 17.1 16.9 21.9 13.6 12.9 17.8 14.0 7.1
Barbados 6.6 5.7 12.6 9.0 8.9 9.9 6.5 4.6 3.3 3.0 2.7
Belize 4.9 2.5 1.1 1.2 4.6 7.5 9.9 9.4 8.0 6.6 4.9
Bolivia 4.3 3.9 0.3 7.2 2.4 1.7 5.7 5.7 4.7 4.8 4.4
Brazil 1.6 3.4 2.9 3.0 3.0 4.2 3.3 1.3 1.4 1.8 2.0
Chile 1.9 1.4 1.7 4.0 4.1 1.7 1.9 1.4 2.3 2.8 3.5
Colombia 2.0 3.0 2.9 3.1 3.3 5.2 6.4 4.3 3.8 3.6 2.9
Costa Rica 1.8 3.2 5.3 5.1 4.8 4.9 4.3 3.2 3.9 4.0 4.1
Dominica 22.7 15.9 14.1 17.3 9.8 7.2 1.9 0.8 6.2 7.3 0.2
Dominican Republic 4.8 7.5 7.5 6.4 4.1 3.3 2.0 1.5 1.6 2.6 3.6
Ecuador 0.5 2.3 0.5 0.2 1.0 0.5 2.1 1.4 0.7 1.6 1.6
El Salvador 1.5 2.5 4.8 5.4 6.5 4.8 3.6 2.0 1.0 2.1 4.4
Grenada 24.3 23.7 23.6 21.1 23.2 4.4 3.8 3.2 7.1 6.3 5.0
Guatemala 0.7 1.4 3.4 2.6 2.5 2.1 0.3 1.0 0.5 0.2 2.5
Guyana 9.1 9.6 13.0 11.6 13.3 9.6 5.7 0.4 2.0 1.1 3.6
Haiti 1.9 1.5 4.3 5.7 6.6 8.5 3.1 0.9 1.1 0.9 0.3
Honduras 3.8 4.3 8.0 8.6 9.6 7.0 5.5 3.8 4.0 4.2 4.4
Jamaica 11.0 8.0 12.2 11.1 9.2 7.5 3.2 2.2 2.7 3.0 3.0
Mexico 1.0 0.5 1.1 1.3 2.5 1.8 2.5 2.2 1.7 2.0 2.2
Nicaragua 8.5 8.9 11.9 10.7 10.9 7.1 9.0 8.6 8.4 8.4 10.2
Panama 0.8 10.8 13.2 10.5 9.8 13.7 7.3 5.7 5.1 3.3 2.8
Paraguay 3.0 0.3 0.4 2.0 1.7 0.4 1.1 1.7 1.1 0.4 0.5
Peru 0.5 2.4 1.9 2.7 4.4 4.4 4.8 2.7 1.5 1.6 2.3
St. Kitts and Nevis 25.2 20.4 13.0 7.6 11.1 4.9 9.7 11.4 12.8 11.1 10.9
St. Lucia 10.8 14.7 16.9 12.2 9.8 3.3 6.8 1.9 0.5 3.6 0.2
St. Vincent and the Grenadines 29.2 30.6 29.4 27.6 30.9 25.7 14.9 15.8 14.7 13.6 11.3
Suriname 2.9 14.9 9.8 3.3 3.8 7.9 16.4 2.8 9.4 6.1 1.1
Trinidad and Tobago 8.6 18.8 16.8 13.1 20.5 15.1 3.9 11.3 9.0 8.4 7.0
Uruguay 1.2 1.8 2.7 5.1 5.0 4.5 2.1 0.1 0.4 0.8 1.5
Venezuela 0.2 1.9 4.9 0.8 2.0 2.3 6.6 1.6 0.4 1.3 1.6
Middle East, North Africa, Afghanistan,
and Pakistan 1.5 6.2 12.8 12.4 9.7 5.5 3.7 4.1 1.9 1.6 0.1
Afghanistan 41.5 29.2 26.7 10.8 0.3 5.7 3.0 7.1 4.7 1.6 3.8
Algeria 0.3 7.5 9.9 5.9 0.4 4.4 16.5 16.5 13.0 10.8 6.7
Bahrain 2.4 3.0 8.8 8.4 7.4 4.6 2.4 4.7 4.6 4.2 3.1
Djibouti 6.6 2.8 13.1 18.8 23.3 25.1 31.8 30.4 21.0 18.2 17.3
Egypt 3.8 1.9 2.5 3.6 2.2 0.8 3.6 6.0 5.9 3.8 2.2
Iran 2.2 4.4 10.4 6.0 6.7 3.2 2.4 4.1 5.1 5.9 6.2
Iraq 11.5 1.6 10.9 5.1 1.1 2.6 6.5 8.7 6.3 6.7 0.6
Jordan 5.2 7.1 10.3 15.2 10.4 7.3 9.1 9.3 8.4 8.3 6.2
Kuwait 26.7 31.8 42.9 45.5 39.9 33.4 3.5 4.5 0.6 1.4 1.0
Lebanon 11.9 20.3 15.4 24.0 26.7 26.4 18.7 18.6 18.0 16.8 15.6
Libya 3 18.5 21.1 9.9 29.9 0.0 78.4 52.6 22.4 1.8 9.8 0.5
Mauritania 13.4 8.2 5.0 24.1 22.0 27.3 19.7 14.9 14.2 9.6 7.9
Morocco 5.4 4.4 7.6 9.3 7.6 5.9 2.1 4.4 4.0 2.9 1.0
Oman 1.0 8.3 13.0 10.1 6.6 5.8 15.5 18.6 14.3 13.2 6.1
Pakistan 5.5 2.2 0.1 2.1 1.1 1.3 1.0 1.7 4.0 4.9 3.0
Qatar 6.5 19.1 31.1 33.2 30.4 24.0 8.4 4.9 2.3 1.0 1.0
Saudi Arabia 4.9 12.7 23.6 22.4 18.1 9.8 8.7 4.3 0.6 0.4 1.6
Somalia ... ... ... ... 4.8 6.3 7.2 10.1 11.1 10.7 10.0
Sudan4 9.6 2.1 0.4 9.3 8.7 7.1 8.0 5.6 1.9 2.0 1.5
Syria5 2.9 2.8 ... ... ... ... ... ... ... ... ...
Tunisia 2.8 4.8 7.4 8.3 8.4 9.1 8.9 9.0 8.7 8.4 6.2
United Arab Emirates 3.1 4.2 12.6 19.7 19.0 13.3 4.7 2.4 2.1 2.1 3.7
Yemen 10.1 3.4 3.0 1.7 3.1 1.7 5.5 5.6 2.3 2.4 3.2
Table A12. Emerging Market and Developing Economies: Balance on Current Account (continued)
(Percent of GDP)
Projections
2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2022
Sub-Saharan Africa 2.4 0.8 0.8 1.8 2.4 3.9 6.1 4.2 3.4 3.6 4.1
Angola 10.0 9.1 12.6 12.0 6.7 3.0 10.0 5.1 4.8 4.5 4.1
Benin 8.3 8.2 7.3 7.4 7.4 8.6 8.4 7.2 8.7 7.1 6.2
Botswana 6.3 2.6 3.1 0.3 8.9 15.4 7.8 11.7 4.5 2.8 3.9
Burkina Faso 4.7 2.2 1.5 7.0 11.3 8.0 8.0 6.8 7.2 7.1 6.7
Burundi 1.7 12.2 14.4 18.6 19.3 18.5 17.7 13.1 12.4 11.8 9.5
Cabo Verde 14.6 12.4 16.3 12.6 4.9 9.1 5.0 3.7 6.1 6.0 5.4
Cameroon 3.5 2.8 3.0 3.6 3.9 4.3 4.1 3.6 3.6 3.5 1.7
Central African Republic 9.1 10.2 7.6 4.6 3.0 5.6 9.0 9.1 9.7 6.5 5.1
Chad 8.2 8.5 5.8 7.8 9.1 8.9 12.3 9.2 2.0 2.8 3.1
Comoros 6.9 0.2 4.9 7.3 8.3 8.0 0.4 10.1 9.5 11.3 13.1
Democratic Republic of the Congo 6.1 10.5 5.2 4.6 5.2 4.8 3.9 3.4 4.6 2.1 2.0
Republic of Congo 14.1 7.8 3.2 17.7 1.7 11.6 42.9 70.1 15.9 2.5 5.0
Cte dIvoire 6.6 1.9 10.4 1.2 1.4 1.4 0.6 1.1 2.9 2.8 2.5
Equatorial Guinea 9.7 20.2 5.7 1.1 2.5 4.3 17.7 10.5 8.0 7.4 4.7
Eritrea 7.6 5.6 0.6 2.3 0.1 0.6 2.2 0.1 0.7 0.3 1.2
Ethiopia 6.7 1.4 2.5 6.9 5.9 6.4 11.6 9.9 8.3 7.4 6.4
Gabon 4.4 14.9 21.0 17.6 7.0 7.3 5.7 10.2 9.3 6.7 1.5
The Gambia 12.5 0.7 12.3 7.9 10.2 10.8 15.0 8.9 9.4 12.0 12.0
Ghana 5.5 8.6 9.0 11.7 11.9 9.5 7.7 6.7 5.8 5.4 4.3
Guinea 5.7 6.4 18.4 20.0 12.5 13.4 15.4 31.9 25.0 21.4 11.8
Guinea-Bissau 5.8 8.3 1.3 8.4 5.0 0.6 2.0 0.9 0.1 0.6 1.2
Kenya 4.4 5.9 9.2 8.4 8.8 10.4 6.8 5.2 6.1 7.0 7.3
Lesotho 1.5 8.4 12.9 8.2 5.5 5.2 4.8 7.7 8.5 9.4 13.2
Liberia 23.2 32.0 27.4 21.5 30.1 26.9 35.2 24.7 26.7 31.3 22.4
Madagascar 21.1 9.7 6.9 6.9 5.9 0.3 1.9 0.8 4.7 5.3 4.0
Malawi 10.2 8.6 8.6 9.3 8.4 8.4 9.5 13.5 9.1 8.1 7.4
Mali 10.8 10.7 5.1 2.2 2.9 4.7 5.3 7.1 7.0 5.6 5.7
Mauritius 7.4 10.3 13.8 7.3 6.3 5.7 4.9 4.4 5.8 6.2 0.3
Mozambique 10.9 16.1 25.3 44.7 42.9 38.2 40.3 38.2 25.6 45.8 114.1
Namibia 1.5 3.5 3.0 5.7 4.0 10.8 12.6 14.0 7.3 6.6 6.0
Niger 24.4 19.8 22.3 14.7 15.0 15.4 18.0 15.5 18.6 18.3 14.1
Nigeria 4.7 3.6 2.6 3.8 3.7 0.2 3.2 0.7 1.9 1.0 0.7
Rwanda 7.0 7.2 7.4 11.2 8.7 11.8 13.4 14.4 10.2 11.2 8.0
So Tom and Prncipe 24.7 22.9 27.7 21.9 13.8 21.9 13.0 6.2 10.2 9.9 7.8
Senegal 6.7 4.4 8.0 10.9 10.5 9.0 7.5 5.3 5.1 5.2 5.7
Seychelles 14.8 19.4 23.0 21.1 11.9 23.1 18.6 18.4 15.6 14.6 14.4
Sierra Leone 13.3 22.7 65.0 31.8 17.5 18.2 17.4 19.7 21.1 18.5 15.1
South Africa 2.7 1.5 2.2 5.1 5.9 5.3 4.4 3.3 2.9 3.3 3.8
South Sudan ... ... 18.2 15.9 3.9 1.6 7.2 4.7 1.7 12.7 4.5
Swaziland 11.2 8.6 6.9 3.3 5.3 3.4 10.8 0.7 1.1 0.2 1.3
Tanzania 7.6 7.7 10.8 11.6 10.6 10.1 8.5 5.6 5.6 6.5 6.7
Togo 5.6 6.3 8.0 7.5 13.2 9.9 11.1 9.7 8.3 7.3 4.3
Uganda 5.7 8.0 10.0 6.8 7.0 8.5 7.1 4.3 5.6 7.2 3.6
Zambia 6.0 7.5 4.7 5.4 0.6 2.1 3.9 4.4 3.6 2.8 0.4
Zimbabwe 6 13.7 12.5 20.1 12.9 15.6 15.1 9.3 4.1 3.6 0.8 1.9
1Georgia, Turkmenistan, and Ukraine, which are not members of the Commonwealth of Independent States, are included in this group for reasons of geography and similarity in economic
structure.
2Starting in 2014 data exclude Crimea and Sevastopol.
3See country-specific notes for Libya in the Country Notes section of the Statistical Appendix.
4Data for 2011 exclude South Sudan after July 9. Data for 2012 and onward pertain to the current Sudan.
5Data for Syria are excluded for 2011 onward owing to the uncertain political situation.
6The Zimbabwe dollar ceased circulating in early 2009. Data are based on IMF staff estimates of price and exchange rate developments in US dollars. IMF staff estimates of US dollar
structure.
structure.
United States.
Underlying Drivers of U.S. Yields Matter for Spillovers October 2014, Chapter 2,
Spillover Feature
Is It Time for an Infrastructure Push? The Macroeconomic Effects of Public Investment October 2014, Chapter 3
The Macroeconomic Effects of Scaling Up Public Investment in Developing Economies October 2014, Box 3.4
Where Are We Headed? Perspectives on Potential Output April 2015, Chapter 3
Steady As She GoesEstimating Sustainable Output April 2015, Box 3.1
Macroeconomic Developments and Outlook in Low-Income Developing Countries
The Role of External Factors April 2016, Box 1.2
Time for a Supply-Side Boost? Macroeconomic Effects of Labor and Product Market
Reforms in Advanced Economies April 2016, Chapter 3
Road Less Traveled: Growth in Emerging Market and Developing Economies in a Complicated
External Environment April 2017, Chapter 3
Growing with Flows: Evidence from Industry-Level Data April 2017, Box 2.2
Emerging Market and Developing Economy Growth: Heterogeneity and Income Convergence
Overthe Forecast Horizon October 2017, Box 1.3
Dismal Prospects for the Real Estate Sector October 2010, Box 1.2
Have Metals Become More Scarce and What Does Scarcity Mean for Prices? October 2010, Box 1.5
Commodity Market Developments and Prospects April 2011, Appendix 1.2
Oil Scarcity, Growth, and Global Imbalances April 2011, Chapter 3
Life Cycle Constraints on Global Oil Production April 2011, Box 3.1
Unconventional Natural Gas: A Game Changer? April 2011, Box 3.2
Short-Term Effects of Oil Shocks on Economic Activity April 2011, Box 3.3
Low-Frequency Filtering for Extracting Business Cycle Trends April 2011, Appendix 3.1
The Energy and Oil Empirical Models April 2011, Appendix 3.2
Commodity Market Developments and Prospects September 2011, Appendix 1.1
Financial Investment, Speculation, and Commodity Prices September 2011, Box 1.4
Target What You Can Hit: Commodity Price Swings and Monetary Policy September 2011, Chapter 3
Commodity Market Review April 2012, Chapter 1,
Special Feature
Commodity Price Swings and Commodity Exporters April 2012, Chapter 4
Macroeconomic Effects of Commodity Price Shocks on Low-Income Countries April 2012, Box 4.1
Volatile Commodity Prices and the Development Challenge in Low-Income Countries April 2012, Box 4.2
Commodity Market Review October 2012, Chapter 1,
Special Feature
Unconventional Energy in the United States October 2012, Box 1.4
Food Supply Crunch: Who Is Most Vulnerable? October 2012, Box 1.5
Commodity Market Review April 2013, Chapter 1,
Special Feature
The Dog That Didnt Bark: Has Inflation Been Muzzled or Was It Just Sleeping? April 2013, Chapter 3
Does Inflation Targeting Still Make Sense with a Flatter Phillips Curve? April 2013, Box 3.1
Commodity Market Review October 2013, Chapter 1,
Special Feature
Energy Booms and the Current Account: Cross-Country Experience October 2013, Box 1.SF.1
Oil Price Drivers and the Narrowing WTI-Brent Spread October 2013, Box 1.SF.2
Anchoring Inflation Expectations When Inflation Is Undershooting April 2014, Box 1.3
Commodity Prices and Forecasts April 2014, Chapter 1,
Special Feature
Commodity Market Developments and Forecasts, with a Focus on Natural Gas October 2014, Chapter 1,
in the World Economy Special Feature
Commodity Market Developments and Forecasts, with a Focus on Investment April 2015, Chapter 1,
in an Era of Low Oil Prices Special Feature
The Oil Price Collapse: Demand or Supply? April 2015, Box 1.1
Commodity Market Developments and Forecasts, with a Focus on Metals in the World Economy October 2015, Chapter 1,
Special Feature
The New Frontiers of Metal Extraction: The North-to-South Shift October 2015, Chapter 1,
Special Feature Box 1.SF.1
Where Are Commodity Exporters Headed? Output Growth in the Aftermath
of the Commodity Boom October 2015, Chapter 2
The Not-So-Sick Patient: Commodity Booms and the Dutch Disease Phenomenon October 2015, Box 2.1
Do Commodity Exporters Economies Overheat during Commodity Booms? October 2015, Box 2.4
Commodity Market Developments and Forecasts, with a Focus on the April 2016, Chapter 1,
Energy Transition in an Era of Low Fossil Fuel Prices Special Feature
Global Disinflation in an Era of Constrained Monetary Policy October 2016, Chapter 3
Commodity Market Developments and Forecasts, with a Focus on Food Security and
Markets in the World Economy October 2016, Chapter 1,
Special Feature
How Much Do Global Prices Matter for Food Inflation? October 2016, Box 3.3
Commodity Market Developments and Forecasts, With a Focus on the Role Technology and
Unconventional Sources in the Global Oil Market April 2017, Chapter 1,
Special Feature
Commodity Market Developments and Forecasts October 2017, Chapter 1,
Special Feature
V. Fiscal Policy
Improved Emerging Market Fiscal Performance: Cyclical or Structural? September 2006, Box 2.1
When Does Fiscal Stimulus Work? April 2008, Box 2.1
Fiscal Policy as a Countercyclical Tool October 2008, Chapter 5
Differences in the Extent of Automatic Stabilizers and Their Relationship with Discretionary Fiscal Policy October 2008, Box 5.1
Why Is It So Hard to Determine the Effects of Fiscal Stimulus? October 2008, Box 5.2
Have the U.S. Tax Cuts Been TTT [Timely, Temporary, and Targeted]? October 2008, Box 5.3
Will It Hurt? Macroeconomic Effects of Fiscal Consolidation October 2010, Chapter 3
Separated at Birth? The Twin Budget and Trade Balances September 2011, Chapter 4
Are We Underestimating Short-Term Fiscal Multipliers? October 2012, Box 1.1
The Implications of High Public Debt in Advanced Economies October 2012, Box 1.2
The Good, the Bad, and the Ugly: 100 Years of Dealing with Public Debt Overhangs October 2012, Chapter 3
The Great Divergence of Policies April 2013, Box 1.1
Public Debt Overhang and Private Sector Performance April 2013, Box 1.2
Is It Time for an Infrastructure Push? The Macroeconomic Effects of Public Investment October 2014, Chapter 3
Improving the Efficiency of Public Investment October 2014, Box 3.2
The Macroeconomic Effects of Scaling Up Public Investment in Developing Economies October 2014, Box 3.4
Fiscal Institutions, Rules, and Public Investment October 2014, Box 3.5
Commodity Booms and Public Investment October 2015, Box 2.2
Cross-Border Impacts of Fiscal Policy: Still Relevant October 2017, Chapter 4
The Spillover Impact of U.S. Government Spending Shocks on External Positions October 2017, Box 4.1
Exchange Rate Pass-Through to Trade Prices and External Adjustment April 2007, Box 3.3
Depreciation of the U.S. Dollar: Causes and Consequences April 2008, Box 1.2
Lessons from the Crisis: On the Choice of Exchange Rate Regime April 2010, Box 1.1
Exchange Rate Regimes and Crisis Susceptibility in Emerging Markets April 2014, Box 1.4
Exchange Rates and Trade Flows: Disconnected? October 2015, Chapter 3
The Relationship between Exchange Rates and
Global-Value-Chain-Related Trade October 2015, Box 3.1
Measuring Real Effective Exchange Rates and Competitiveness:
The Role of Global Value Chains October 2015, Box 3.2
Labor Force Participation Rates in Advanced Economies October 2017, Box 1.1
Recent Wage Dynamics in Advanced Economies: Drivers and Implications October 2017, Chapter 2
Labor Market Dynamics by Skill Level October 2017, Box 2.1
Worker Contracts and Nominal Wage Rigidities in Europe: Firm-level Evidence October 2017, Box 2.2
Wage and Employment Adjustment After the Global Financial Crisis: Firm-level Evidence October 2017, Box 2.3
A Tale of Two Adjustments: East Asia and the Euro Area October 2014, Box 4.2
Understanding the Role of Cyclical and Structural Factors in the Global Trade Slowdown April 2015, Box 1.2
Small Economies, Large Current Account Deficits October 2015, Box 1.2
Capital Flows and Financial Deepening in Developing Economies October 2015, Box 1.3
Dissecting the Global Trade Slowdown April 2016, Box 1.1
Understanding the Slowdown in Capital Flows to Emerging Markets April 2016, Chapter 2
Capital Flows to Low-Income Developing Countries April 2016, Box 2.1
The Potential Productivity Gains from Further Trade and Foreign Direct Investment Liberalization April 2016, Box 3.3
Global Trade: Whats Behind the Slowdown October 2016, Chapter 2
The Evolution of Emerging Market and Developing Economies Trade Integration with
Chinas Final Demand April 2017, Box 2.3
Shifts in the Global Allocation of Capital: Implications for Emerging Market and
Developing Economies April 2017, Box 2.4
Macroeconomic Adjustment in Emerging Market Commodity Exporters October 2017, Box 1.4
Remittances and Consumption Smoothing October 2017, Box 1.5
X. Regional Issues
EMU: 10 Years On October 2008, Box 2.1
Vulnerabilities in Emerging Economies April 2009, Box 2.2
East-West Linkages and Spillovers in Europe April 2012, Box 2.1
The Evolution of Current Account Deficits in the Euro Area April 2013, Box 1.3
Macroeconomic Policies for Smoother Adjustment to Abrupt Climate Shocks April 2008, Box 4.3
Catastrophe Insurance and Bonds: New Instruments to Hedge Extreme Weather Risks April 2008, Box 4.4
Recent Emission-Reduction Policy Initiatives April 2008, Box 4.5
Complexities in Designing Domestic Mitigation Policies April 2008, Box 4.6
Getting By with a Little Help from a Boom: Do Commodity
Windfalls Speed Up Human Development? October 2015, Box 2.3
Breaking the Deadlock: Identifying the Political Economy Drivers of Structural Reforms April 2016, Box 3.1
Can Reform Waves Turn the Tide? Some Case Studies Using the Synthetic Control Method April 2016, Box 3.4
A Global Rush for Land October 2016, Box 1.SF.1
Within-Country Trends in Income per Capita: The Case of the Brazil, Russia, India, China,
and South Africa April 2017, Box 2.1
Technological Progress and Labor Shares: A Historical Overview April 2017, Box 3.1
The Elasticity of Substitution Between Capital and Labor: Concept and Estimation April 2017, Box 3.2
Routine Tasks, Automation, and Economic Dislocation Around the World April 2017, Box 3.3
Adjustments to the Labor Share of Income April 2017, Box 3.4
Conflict, Growth, and Migration April 2017, Box 1.1
Tackling Measurement Challenges of Irish Economic Activity April 2017, Box 1.2
The Effects of Weather Shocks on Economic Activity: How Can Low-Income Countries Cope? October 2017, Chapter 3
The Growth Impact of Tropical Cyclones October 2017, Box 3.1
The Role of Policies in Coping with Weather Shocks: A Model-Based Analysis October 2017, Box 3.2
Strategies for Coping with Weather Shocks and Climate Change: Selected Case Studies October 2017, Box 3.3
Coping with Weather Shocks: The Role of Financial Markets October 2017, Box 3.4
Historical Climate, Economic Development, and the World Income Distribution October 2017, Box 3.5
Mitigating Climate Change October 2017, Box 3.6
The following remarks were made by the Chair at the conclusion of the Executive Boards discussion of the
Fiscal Monitor, Global Financial Stability Report, and World Economic Outlook on September 21, 2017.
E
xecutive Directors broadly shared the assess- cross-border spillovers. Common challenges include
ment of global economic prospects and maintaining the rules-based, open trading system;
risks. They observed that global activity has preserving the resilience of the global financial system;
strengthened further and is expected to rise avoiding competitive races to the bottom in taxation
steadily into next year. The pickup is broad based and financial regulation; and further strengthening the
across countries, driven by investment and trade. Nev- global financial safety net. Multilateral cooperation is
ertheless, the recovery is not complete, with medium- also essential to tackle various noneconomic challenges,
term global growth remaining modest, especially among which are refugee flows, cyberthreats and, as
in advanced economies and fuel exporters. In most most Directors highlighted, mitigating and adapting
advanced economies, inflation remains subdued amid to climate change. Concerted effort is also needed to
weak wage growth, while slow productivity growth and reduce excess global imbalances, through a recalibra-
worsening demographic profiles weigh on medium- tion of policies with a view to achieving their domestic
term prospects. Meanwhile, several emerging markets objectives as well as strengthening prospects for strong,
and developing economies continue to adjust to a sustainable, and balanced global growth. In this con-
range of factors, including lower commodity revenues. text, as a few Directors emphasized, the IMF also has a
Directors noted that, while risks are broadly bal- role to play by continuing to strengthen its multilateral
anced in the near term, medium-term risks remain analysis of external imbalances and exchange rates.
skewed to the downside, with rising financial vulnera- Directors agreed that continued accommodative
bilities. These include the possibility of a sudden tight- monetary policy is still needed in countries with low
ening of global financial conditions, a rapid increase in core inflation, consistent with central banks mandates.
private sector debt in key emerging market economies, Fiscal policy should gear toward long-term sustain-
low bank profitability and pockets of still-elevated non- ability, avoid procyclicality, and promote inclusive
performing loan ratios, and policy uncertainty about growth. At the same time, fiscal policy should be as
financial deregulation. Directors also pointed to risks growth friendly as possible, using space, where avail-
associated with inward-looking policies, rising geopo- able, to support productivity and growth-enhancing
litical tensions, and weather-related factors. structural reforms. In many cases, policymakers should
Given this landscape, Directors underscored the prioritize rebuilding buffers, improving medium-term
continued importance of employing arange of policy debt dynamics, and enhancing resilience. Efforts to
tools, in a comprehensive, consistent, and well- raise potential output should be prioritized based on
communicated manner, to secure the recovery and country-specific circumstances, including increasing
improve medium-term prospects. They recognized that the supply of labor, upgrading skills and human capi-
major central banks have made every effort to commu- tal, investing in infrastructure, and lowering product
nicate their monetary normalization policies to markets. and labor market distortions. Social safety nets remain
The cyclical upturn in economic activity provides a important to protect those adversely affected by tech-
window of opportunity to accelerate critical structural nological progress and other structural transformation.
reforms, increase resilience, and promote inclusiveness. Directors noted that income disparities among
Directors stressed that a cooperative multilateral countries have narrowed, but inequality has increased
framework remains vital for amplifying the mutual in some economies. They saw a role that well-designed
benefits of national policies and minimizing any fiscal policies can play in achieving redistributive
objectives without necessarily undermining growth and Directors observed that the global financial system
incentives to work. Directors generally concurred that continues to strengthen, and market confidence has
there may be scope for strengthening means-testing improved generally. They recognized the substan-
of transfers in many countries and for increasing the tial progress made in resolving weak banks in many
progressivity of taxation in some others. Most Direc- advanced economies, while a majority of systemic
tors noted that any consideration of auniversal basic institutions are adjusting business models and restoring
income would have to be weighed carefully against a profitability. However, a prolonged period of monetary
host of country-specific factorsincluding existing accommodation could lead to further increases in asset
social safety schemes, financing modalities, fiscal cost, valuations and a buildup of leverage in the nonfi-
and social preferences, as well as its impact on incen- nancial sector that could signal higher risks to finan-
tives to workwhich, in the view of many Directors, cial stability. These developments call for continued
raised questions about its attractiveness and practical- vigilance about household debt ratios and investors
ity. Directors emphasized that improving education exposure to market and credit risks. In this context,
and health care is key to reducing inequality and Directors stressed the need to calibrate the path of nor-
enhancing social mobility over time. malization of monetary policies carefully, implement
Directors underlined the continued need for emerg- macro- and microprudential measures as needed, and
ing market and developing economies to bolster address remaining legacy problems.
economic and financial resilience to external shocks, Directors noted a generally subdued outlook for
including through enhanced macroprudential policy commodity prices. They encouraged low-income
frameworks and exchange rate flexibility. They noted developing countries that are commodity export-
that acommon challenge across these economies is how ers to continue improving revenue mobilization and
to speed up their convergence toward living standards in strengthening debt management, while safeguarding
advanced economies. While priorities differ across coun- social outlays and capital expenditures. Countries with
tries, many need to improve governance, infrastructure, more diversified export bases should further strengthen
education, and access to health care. In several countries, fiscal positions and foreign exchange buffers. Across all
policies should also facilitate greater labor force partici- low-income developing countries, an overarching chal-
pation, reduce barriers to entry into product markets, lenge is to maintain progress toward their Sustainable
and enhance the efficiency of credit allocation. Development Goals.
CO
MI
NG
SO
ON
Unleashing Growth and From Great Depression to Great Fiscal Policies and Gender Equality
Strengthening Resilience in Recession: The Elusive Quest for $25. 2017. Paperback
the Caribbean International Policy Cooperation ISBN 978-1-51359-036-3.
$25. 2017. Paperback $27. 2017. Paperback
ISBN 978-1-48431-519-4. Approx. 400pp. ISBN 978-1-51351-427-7. 256pp.
F R AV
CO
EE AI
MI
DO L A
NG
WN BL
SO
LO E
AD
ON