TABLE OF CONTENTS

I. Introduction..........................................................2
II. The theoretical relationship between variables......3
1. The relationship between FDI and unemployment rate..........3

2. The relationship between CPI and unemployment rate .........4

3. The relationship between GDP and unemployment rate .......6

4. The relationship between saving rate and unemployment rate

..................................................................................................7

5. The relationship between export growth and unemployment

rate ...........................................................................................7

III. Data and methodology.........................................8

1. Data source............................................................................ 8

2. Data description.....................................................................8

3. Methodology..........................................................................9

IV. Empirical Results.................................................9

1. Regression model................................................................... 9

1.1. Model 1............................................................................9

1.2. Model 2..........................................................................10

2. Test assumptions for model 2..............................................11

2.1. Multicollinearity..............................................................11

2.2. Heteroskedasticity.........................................................12

2.3. Normality.......................................................................12

2.4. Autocorrelation..............................................................13
V. Interpretation.....................................................14
1. Interpretation.......................................................................14

2. Suggested policy to reduce unemployment rate..................15

VI. Conclusion.........................................................16
REFERENCES...........................................................17
13. Harun Öztürkler, Ömer Faruk Çolak, The
relationship between current account deficits and
unemployment in Turkey. .......................................18
14. Noel Gaston, Gulasekaran Rajaguru (2011), How an
export boom affects unemployment, The Institute of
Social and Economic Research, Osaka University......18
APPENDIX...............................................................19
1. Data.....................................................................................19

2. Non- linearity test (log terms)- pvalue<0.05.......................20

3. Non- linearity test (squared terms)- pvalue<0.05................21

4. Ramsey’s RESET test for model 2- p value >0.05................21

5. Plot fitted and actual value of dependent variables.............21

I. Introduction
A person is said to be "unemployed” if he or she is looking for
work, or is willing to work at the prevailing wage, but is unable to
find a job. Unemployment refers to the condition of being
unemployed, or to the number or proportion of people in the labor
force who are unemployed. The labor force of a country consists
of everyone of working age who are actively employed or seeking
employment. People not counted include students, retired people,
stay-at-home parents and people in prisons or similar institutions.
In Vietnam at present, the adult population (in-work age
population) is 48 M. Of which, the labor force accounts for
43.01/48 = 89.5%, consisting of about 42.01 M employed and
2
about 2M unemployed. According to the formula in theories we
have: u = U/L = 2.01/43.01 = 4.65%.
Vietnam has a low unemployment rate by international standards.
While this low and stable unemployment rate is an encouraging
sign, one would be too hasty to conclude that the rate of
employment creation is keeping up well with the rate of growth of
labor force. Low unemployment rate is rather typical in low
income countries where many workers are self-employed and are
too poor to afford unemployment.
In 2008, household consumption achieved an 8% growth
compared to 10.7% growth for 2007. In 2009, this will still be the
stable source for GDP growth, especially when Vietnam
population increases by almost 2% (> 3 millions) each year. On
the other hand, higher unemployment may lead to declining
income and declining consumption, which eventually result in
enterprises reducing production and laying off workers, so higher
unemployment means slower growth. According to WLO, 1% of
economy growth equals to 0.33%-0.34% labor growth. In 2008,
Vietnam economy growth dropped from 8.48% (2007) to 6.23%,
which equals to 300,000 more unemployment of 9-million working
force. We estimate further 150,000 will lose their jobs in 2009,
equals to an economy growth of 5.1% (a decrease of 1.1%
compared to 2008). Overall, we estimate that household
consumption will achieve a 6% growth for 2009.
The significance of this research is to explore the reasons that
affected current unemployment rate in Vietnam from 1990 to
2010 through some important aspects of economic, including FDI,
CPI, GDP growth rate, saving and export.

II. The theoretical relationship between variables

This section analyzes the relationship between FDI, CPI, GDP,
saving rate, export growth rate and unemployment rate.

1. The relationship between FDI and unemployment rate

Foreign direct investment (FDI) has been recognized as an
important resource for economic development. Many scholars
3
widely believe that the benefits accrued from FDI may include the
acquisition of new technology, employment creation, human
capital development, contribution to international trade
integration, enhancing domestic investment, and increasing tax
revenue generated by FDI (Jenkins and Thomas, 2002; World
Bank, 2000). All of these benefits are expected to contribute to
higher economic and employment growth. The direct impact of
FDI can be seen through the raise in employment resulting from
the increase in the demand for employment, and the
improvement of workforce and safety nets. FDI contributes to
economic growth directly by creating employment opportunities
and indirectly through the creation of employment opportunities
in other organizations.

Looking primarily at the effects of FDI on the level of employment,

it is possible to identify circumstances under which it has a
significant positive impact. Where such investment supplements
domestic investment and involves the creation of new
“Greenfield” plants, demand for labor will tend to increase. If this
FDI is concentrated in labor-intensive industries, this increase will
be substantial. FDI can also lead to increased employment
amongst local firms as a result of backward or forward linkages so
that the direct employment by foreign affiliates may
underestimate the total impact. Foreign firms that are subject to
pressures in their home countries may also bring with them
higher labor standards and wages than the norm for the host
economy. Where a firm makes a long-term commitment, it can
provide stable employment.

2. The relationship between CPI and unemployment rate

A consumer price index (CPI) measures changes in the price
level of consumer goods and services purchased by households. A
chief measure of price inflation is the inflation rate, the
4
annualized percentage change in the Consumer Price Index over
time.
Unemployment and inflation are two intricately linked economic
concepts. Over the years there have been a number of
economists trying to interpret the relationship between the
concepts of inflation and unemployment. There are two possible
explanations of this relationship – one in the short term and
another in the long term. In the short term there is an inverse
correlation between the two. As per this relation, when the
unemployment is on the higher side, inflation is on the lower side
and the inverse is true as well.
This relationship has presented the regulators with a number of
problems. The relationship between unemployment and inflation
is also known as the Phillips curve. In the short term the Phillips
curve happens to be a declining curve. The Phillips curve in the
long term is separate from the Phillips curve in the short term. It
has been observed by the economists that in the long run the
concepts of unemployment and inflation are not related.
As per the classical view of inflation, inflation is caused by the
alterations in the supply of money. When the money supply goes
up the price level of various commodities goes up as well. The
increase in the level of prices is known as inflation. According to
the classical economists there is a natural rate of unemployment,
which may also be called the equilibrium level of unemployment
in a particular economy. This is known as the long term Phillips
curve. The long term Phillips curve is basically vertical as inflation
is not meant to have any relationship with unemployment in the
long term.
It is therefore assumed that unemployment would stay at a fixed
point irrespective of the status of inflation. Generally speaking if
the rate of unemployment is lower than natural rate, then the rate
of inflation exceeds the limits of expectations and in case the
unemployment is higher than what is the permissible limit then
the rate of inflation would be lower than the expected levels.
The Keynesians have a different point of view compared to the
Classics.

5
The Keynesians regard inflation to be an aftermath of money
supply that keeps on increasing. They deal primarily with the
institutional crises that are encountered by people when they
increase their price levels. As per their argument the owners of
the companies keep on increasing the salaries of their employees
in order to appease them. They make their profit by increasing
the prices of the services that are provided by them. This means
there has to be an increase in the money supply so that the
economy may keep on functioning. In order to meet this demand
the government keeps on providing more money so that it can
keep up with the rate of inflation.

3. The relationship between GDP and unemployment rate

Gross domestic product (GDP) refers to the market value of all
final goods and services produced within a country in a given
period. It is often considered an indicator of a country's standard
of living.
We know that when there is unemployment, the economy is not
producing at full output since there are people who are not
working. But, what exactly is the relationship between
unemployment and national output or GDP? How much would we
expect the GDP to increase if unemployment fell 1%? These are
useful and important questions to ask when trying to understand
the costs of unemployment.
An economist named Arthur Okun looked at the relationship
between unemployment and national output over the past 50
years. He noticed a general pattern and stated an equation to
explain it. His equation, Okun's Law, relates the percentage
change in real GDP to changes in the unemployment rate. In
particular, the equation states: % change in real GDP = 3% - 2 x
(change in unemployment rate)
This equation basically says that real GDP grows at about 3% per
year when unemployment is normal. For every point above
normal that unemployment moves, GDP growth falls by 2%.
Similarly, for every point below normal that unemployment
6
moves, GDP growth rises by 2%. This equation, while not exact,
provides a good estimate of the effects of unemployment upon
output.
Okun's law states that a one point increase in the unemployment
rate is associated with two percentage points of negative growth
in real GDP. The relationship varies depending on the country and
time period under consideration.

4. The relationship between saving rate and

unemployment rate
Because current account deficits reflect the part of domestic
investment that cannot be financed by domestic savings,
assuming there is no decline in national savings, increases in
current account deficits theoretically imply increased domestic
investment, and therefore it should reduce unemployment.

5. The relationship between export growth and

unemployment rate
The term export is derived from the conceptual meaning as to
ship the goods and services out of the port of a country. The
seller of such goods and services is referred to as an "exporter"
who is based in the country of export whereas the overseas based
buyer is referred to as an "importer". In International Trade,
"exports" refers to selling goods and services produced in home
country to other markets.
Any good or commodity, transported from one country to another
country in a legitimate fashion, typically for use in trade. Export
goods or services are provided to foreign consumers by
domestic producers.
Higher terms of trade lower the equilibrium rate of
unemployment. Specifically, rising export prices have been an
important factor driving reductions in the rate of unemployment.
Interestingly, so too have falling import prices.

7
III. Data and methodology
1. Data source
Main data source comes from data bank online of the World Bank.
Besides, we also collect some data from the Statistical Year Book
2005 and some relevant paper available in the General Statistic
Office’s website. There are some problem arising in collecting
data since there are small number of observations and some
differences between different sources.
We have six variables: Foreign direct investment (net inflows % of
GDP) denoted as FDI, GDP growth rate (annual %) denoted as
GDP, CPI (annual %) calculated by us from data of the GSO, Gross
domestic savings (% of GDP) denoted as SR, Exports of goods and
services (annual % growth) denoted as EX and the unemployment
rate (annual %) denoted as UR. The data is from 1990 to 2010.

2. Data description
Variab Description Mea SD Max Min
les n
FDI Foreign direct investment 6.24 0.60 11.9 2.781
(%) (net inflows % of GDP) 1 0 39
GDP GDP growth rate (annual %) 7.31 0.30 9.54 4.774
(%) 5 6 0
CPI CPI (annual %) 13.4 3.94 79.9 -1.617
(%) 67 1 08
EX (%) Exports of goods and 15.9 8.64 29.8 -
services (annual % growth) 96 4 57 10.06
9
SR (%) Gross domestic savings (% 22.4 7.62 31.6 3.327
of GDP) 44 5 68
UR (%) The unemployment rate 6.34 1.92 12.3 4.2
(annual %) 1 6

Variabl Expected Reasons

es signs
FDI (%) - (negative) Part II.1

8
GDP - (negative) Part II.3
(%)
CPI (%) - (negative) Part II.2
SR (%) - (negative) Part II.4
EX (%) - (negative) Part II.5

3. Methodology
This paper aims at evaluating the influence of some factors on the
unemployment rate in Vietnam. To analyze this impact, we have
to run an econometric model by using the Ordinary Least Squares
method (OLS) with the support of Gretl software. The
unemployment rate is the dependent variable, while the other
variables are the independent.
The formulation of regression will be represented in detail as
follow
UR = b1 - b2.FDI – b3. GDP – b4. CPI – b5. SR – b6. EX +ui
We use the linear model for two reasons. First, all of the data are
percentage so it is not necessary to present the logistics model.
However, in this paper, we also test whether there is a log
relationship between variables. Second, the linear model is easier
to be observed.

IV. Empirical Results

1. Regression model

1.1. Model 1
We run regression by Gretl and receive the result as follow:
Model 1: OLS, using observations 1990-2010 (T = 21)
Dependent variable: UR

Coefficient Std. Error t-ratio p-value

const 14.4488 0.754008 19.1627 <0.00001 ***
GDP -0.186352 0.0727045 -2.5631 0.02163 **
SR -0.213629 0.0166357 -12.8416 <0.00001 ***
FDI -0.319183 0.0381273 -8.3715 <0.00001 ***
CPI 0.00179515 0.00696843 0.2576 0.80021

9
EX 0.00114691 0.0119204 0.0962 0.92462

Mean dependent var 6.341429 S.D. dependent var 1.926298

Sum squared resid 2.683466 S.E. of regression 0.422963
R-squared 0.963841 Adjusted R-squared 0.951788
F(5, 15) 79.96634 P-value(F) 2.85e-10
Log-likelihood -8.194872 Akaike criterion 28.38974
Schwarz criterion 34.65688 Hannan-Quinn 29.74987
Rho 0.091222 Durbin-Watson 1.772619

As we can see, there are four coefficients having statistical

significance because p-values are smaller than 0.05. The two
variables CPI and EX have not had statistical significance. The
sign of the significant variables is followed our expectations.
R2 = 96.38%, adjusted R2 = 95.18%
1.2. Model 2
We run regression with three significant variables.
Model 2: OLS, using observations 1990-2010 (T = 21)
Dependent variable: UR

Coefficient Std. Error t-ratio p-value

Const 14.5879 0.533775 27.3297 <0.00001 ***
GDP -0.18772 0.0679267 -2.7636 0.01328 **
SR -0.216776 0.0117521 -18.4457 <0.00001 ***
FDI -0.321726 0.0347198 -9.2664 <0.00001 ***

Mean dependent var 6.341429 S.D. dependent var 1.926298

Sum squared resid 2.698103 S.E. of regression 0.398387
R-squared 0.963644 Adjusted R-squared 0.957228
F(3, 17) 150.1974 P-value(F) 1.97e-12
Log-likelihood -8.251987 Akaike criterion 24.50397
Schwarz criterion 28.68206 Hannan-Quinn 25.41073
Rho 0.070819 Durbin-Watson 1.825246

Our model has 3 significant variables: GDP, FDI, SR with R2 =

96.36% and Adjusted R2= 95.72%. The sign of the significant
variables is followed our expectations.
We see that adjusted R2 of the model 2= 95.72% > adjusted R2
of the model 1= 95.18%. Consequently, the model 2 is better

10
than model 1. Besides, we also run the Omit variables test and
receive results as follow:
Table 4.1
Null hypothesis: the regression parameters are zero for the
variables EX, CPI

Test statistic: F(2, 15) = 0.0409077, with p-value = 0.960024

Of the 3 model selection statistics, 3 have improved.

P-value >0.05, so we do not reject null hypothesis. We can

confirm that the 2 variables, EX, CPI are not significant for the
model.
Regression function:
UR^ = 14.6 - 0.188*GDP - 0.217*SR - 0.322*FDI
(0.534) (0.0679) (0.0118) (0.0347)

2. Test assumptions for model 2

2.1. Multicollinearity
By using Gretl, we have VIF value for 3 independent variables.
The result is quoted
in table 4.2
Table 4.2
Variance Inflation Factors

Minimum possible value = 1.0

Values > 10.0 may indicate a collinearity problem

GDP 1.146
SR 1.012
FDI 1.150

VIF(j) = 1/(1 - R(j)^2), where R(j) is the multiple correlation

coefficient
between variable j and the other independent variables

11
The table shows that VIF is smaller than 10 and near 1. As a
result, we can conclude that multicollinearity does not appear in
this case.
2.2. Heteroskedasticity
In this part, we continue testing the second assumption using
White test. The software gives us the following result in table 4.3
Table 4.3
White's test for heteroskedasticity
OLS, using observations 1990-2010 (T = 21)
Dependent variable: uhat^2

coefficient std. error t-ratio p-value

--------------------------------------------------------
const -1.70497 1.41237 -1.207 0.2527
GDP 0.530743 0.382295 1.388 0.1925
SR 0.0313004 0.0362355 0.8638 0.4061
FDI -0.0587636 0.220196 -0.2669 0.7945
sq_GDP -0.0350615 0.0248150 -1.413 0.1853
X2_X3 -0.00543290 0.00694138 -0.7827 0.4503
X2_X4 0.00986216 0.0148321 0.6649 0.5198
sq_SR 0.000300666 0.00139741 0.2152 0.8336
X3_X4 -0.00173898 0.00373021 -0.4662 0.6502
sq_FDI 0.00134211 0.00846593 0.1585 0.8769

Unadjusted R-squared = 0.300239

Test statistic: TR^2 = 6.305026,

with p-value = P(Chi-square(9) > 6.305026) = 0.709044
The null hypothesis test Ho = heteroskedasticity not present
P-value is equal to 0.709 which is larger than 0.05, leading us to
conclude that we do not reject H0. In other words, there is no
heteroskedasticity.
2.3. Normality
In this part, we answer the question whether the error term ui has
normal distribution. Thanks to Gretl, we have the following result
Table 4.4
Test for normality of residual -

12
 Null hypothesis: error is normally distributed
Test statistic: Chi-square(2) = 0.850094

with p-value = 0.653739

P-value= 0.692 >0.05, we conclude that error is normally

distributed.
2.4. Autocorrelation
This part will test whether Cov(ui,uj)=0 or not.
According to table 4.1, Durbin- Watson value- d=1.825246
dL = 1.026; dU = 1.699
Since d >dU, we can conclude no autocorrelation.
Furthermore, we can use BG test (LM test) for autocorrelation up
to order 3 with the help of Gretl
Table 4.5
LM test for autocorrelation up to order 1 -
Null hypothesis: no autocorrelation
Test statistic: LMF = 0.0854691
with p-value = P(F(1,16) > 0.0854691) = 0.773775

LM test for autocorrelation up to order 2 -

Null hypothesis: no autocorrelation
Test statistic: LMF = 2.01795
with p-value = P(F(2,15) > 2.01795) = 0.16745

LM test for autocorrelation up to order 3 -

Null hypothesis: no autocorrelation
Test statistic: LMF = 1.46658
with p-value = P(F(3,14) > 1.46658) = 0.266307
According to the result of LM test with p- value > 0.05, we can
conclude that there is no autocorrelation.

13
V. Interpretation
1. Interpretation
After running regression and testing all the assumptions for
multiple regressions, we have the following regression equation
with R2 = 96.36%
UR = 14.6 - 0.188*GDP - 0.217*SR - 0.322*FDI + ui (5.1)

According to the function, we can conclude as follows:

- Equation 5.1 represents the fact that GDP growth rate has a
negative influence on the unemployment rate. One percentage
point of GDP growth rate reduces 0.188 percentage point of
unemployment rate if other factors remain unchanged. This is
similar to our expectation and the Okun’s law which stated the
negative relationship between unemployment and GDP. Increases
in the unemployment rate should be associated with decreases in
real GDP because employed workers help to produce goods and
services and unemployed workers do not (Macroeconomics-
Mankiw). More notably, the data here is real GDP growth rate.
- According to equation 5.1, we can see the obviously negative
relationship between the saving rate and the unemployment rate.
A coefficient of -0.127 means that saving rate growth of 1
percentage point reduces the unemployment rate by 0.127
percentage points. This finding is inline with our expectation. An
increase in saving rate leads to an increase in investment.
Investment goes up, creating more jobs. Because lacking of data,
we can not examine the investment variables directly.
- Last but not least, we can observe that FDI inflows growth rate
and unemployment rate are strongly negative. Regarding to the
unemployment rate, if FDI grows by 1 percent, the unemployment
rate will decrease by 0.322 percent. This result, once again,
satisfied our satisfaction and previous studies.

14
- However, our paper cannot analyze the impacts of CPI and
export growth on the unemployment rate, which may come from
some reasons.
First, we find no particular relationship between CPI and
unemployment rate, which is different from our expectations that
these 2 variables may be negatively related. Labor market
rigidities and changes in the cost of labor are believed as a major
cause of inflation in developed countries. But it is not considered
as a major cause of inflation in most developing countries.
Chhibber and Shafik (1990) argued that “wage-push inflation” is
rare in developing countries because wages constitute only a
small part of national income. In the case of Vietnam, a large pool
of underemployment in agriculture sector makes the price of
labor extremely cheap, especially for unskilled labor. The wage-
price spiral therefore can not hold true. Similarly, the trade-off
between unemployment and inflation suggested by Phillips curve
is not applicable for Vietnam case. (Bui Thi Kim Thanh, Inflation in
Vietnam over period 1990-2007, Research paper for MA course,
November 2008)
Second, we also find no relationship between the unemployment
rate and the export growth rate since there are many differences
between needs of workers and the exporting companies. Workers
require high salary; however they do not meet the companies’
requirements of ability. As a result, although the company
increases output by changing technology, the workers still can not
meet the companies’ demand. The unemployment rate remains
unchanged.

2. Suggested policy to reduce unemployment rate

- Maintain the stability of economic growth and ensure Vietnam
economy to develop gradually. Vietnam must focus on
maintaining a proper economic growth rate and controlling
inflation rate in order to manage its economy
- Promote domestic private investment and improve effectiveness
of government investment.

15
- Make great efforts to attract foreign direct investment by
providing the foreign projects with special policy.

VI. Conclusion

From our final regression model, we can conclude that among five
variables: GDP growth rate, FDI inflows growth rate, CPI, domestic
saving growth rate and export growth rate, there are 3 variables
having negative impacts on the unemployment rate over the
period 1990-2010. These three variables’ influence has followed
our first expectations and other relevant previous studies.
However, there are two insignificant variables, CPI and export
growth rate, for which we also present our explanations.

Besides, in presenting this paper, we have to face some

problems. The most challenged problem arising in the first step is
finding data source. It makes us confused since there are some
differences among domestic and foreign sources. Finally, we
decided to choose the data from World Bank, Asian Development
Bank and GSO. Moreover, we also have difficulties using software,
choosing variables and suitable model.

To do this research paper, we hope to have an outlook into the

unemployment situation in Vietnam and try explaining the
influence of some factors on it. From the result of this paper, we
also try to suggest policies to reduce the unemployment rate in
Vietnam in recent days.

16
REFERENCES
1. James H. Stock and Mark W.Watson (2008), Introduction to
Econometrics (Brief Edition), US: Pearson.
2. N. Gregory Mankiw, Macroeconomics, 6th edition
3. Nguyen Quang Dong (2008), Bài giảng kinh tế lượng, Nhà xuất
bản GTVT.
4. Bui Thi Kim Thanh (2008), Inflation in Vietnam over the period
1990-2007, part of the author’s study programme while at the
Institute of Social Studies.
5. Nguyen Thi Tue Anh, Vu Xuan Nguyet Hong, Tran Toan Thang,
Nguyen Manh Hai (2006), The impacts of foreign direct
investment on the economic growth in Vietnam, Research report,
Capacity building project for policy research.
6. P.P.A Wasantha Athukorala (2003), The Impact of Foreign
Direct Investment for Economic Growth: A Case Study in Sri
Lanka, research paper, 9th International conference on Sri Lanka
Studies.
7. Ewald Walterskirchen (1999), The relationship between growth,
employment and unemployment in the EU, research paper,
Workshop in Barcelona.
8. Nola Reinhardt (2007), Saving rates in Latin American: Why
reformers got it wrong, research paper, the Latin American
Studies Association XXVII International Congress, Montreal,
Canada.
9. Ismail AKTAR, Latif OZTURK, Nedret DEMIRCI (2007), The
impact of FDI, export, economic growth, total fixed investment on

17
unemployment in Turkey, research paper, Kırıkkale University,
Turkey.
10. Brian W. Cashell, Inflation and Unemployment: What is the
Connection?, Congressional Research Service.
11. Rhys Jenkins (2006), Globalization, FDI and employment in
Viet Nam, Transnational Corporations, Vol. 15, No. 1 (April 2006).
12. Tran Trong Hung, Impacts of Foreign Direct Investment on
Poverty Reduction in Vietnam, IDS Program, GRIPS.

13. Harun Öztürkler, Ömer Faruk Çolak, The relationship between

current account deficits and unemployment in Turkey.

14. Noel Gaston, Gulasekaran Rajaguru (2011), How an export

boom affects unemployment, The Institute of Social and
Economic Research, Osaka University.
15. The General Statistical Office, Statistical year books (brief and
online version).
16. The General Statistical Office, Special subject of analysis.
17. Wikipedia, Okun’s law, http://en.wikipedia.org/wiki/Okun
%27s_Law, accessed 26/4/2011.
18.Sparknotes,Unemployment,http://www.sparknotes.com/econo
mics/macro/measuring2/section2.rhtml, accessed 26/4/2011.
19. World Bank data sets, Vietnam page
20. Economy Watch website, Inflation and Unemployment,
accessed 26/4/2011.
21. Asian Development Bank statistical database system online,
Key Indicators for Asia and the Pacific, last updated through SDBS
on 30 July 2010

18
APPENDIX
1. Data
Yea UR GDP SR FDI CPI EX
r
199 12.3 5.101 3.327 2.781 23.3 12.925
0
199 10.4 5.961 9.953 3.903 79.9 29.857
1
199 8 8.646 13.554 4.803 38.7 24.672
2
199 7 8.073 15.489 7.028 8.5 9.127
3
199 6.2 8.839 16.029 11.939 9.3 16.000
4
199 6 9.540 18.049 8.586 17.8 20.000
5
199 5.88 9.340 17.134 9.713 5.7 24.000
6
199 6.01 8.152 20.161 8.270 3.2 16.000
7
199 6.20 5.764 21.745 6.141 7.7 19.000
8
199 6.74 4.774 24.775 4.923 4.4 23.000
9
200 6.70 6.787 27.147 4.164 -1.6 21.101
0
200 6.28 6.895 28.891 3.999 -0.3 17.180
1
200 6.01 7.080 28.041 3.991 3.9 10.368
19
 2
200 5.78 7.341 27.084 3.667 3.2 19.950
3
200 5.60 7.790 27.911 3.543 7.7 25.620
4
200 5.31 8.442 31.393 3.692 8.3 17.776
5
200 4.82 8.229 31.668 3.939 7.5 -10.069
6
200 4.20 8.456 28.195 9.422 8.3 11.288
7
200 4.65 6.311 24.506 10.611 23.0 5.049
8
200 4.66 5.324 27.778 7.821 6.9 11.080
9
201 4.43 6.780 28.500 8.12 11.5 12.000
0
2. Non- linearity test (log terms)- pvalue<0.05
Auxiliary regression for non-linearity test (log terms)
OLS, using observations 1990-2010 (T = 18)
Missing or incomplete observations dropped: 3
Dependent variable: uhat
coefficient std. error t-ratio p-value
-------------------------------------------------------
const 8.31743 4.44485 1.871 0.1035
GDP 1.25330 0.658714 1.903 0.0988 *
FDI 0.522643 0.181570 2.878 0.0237 **
SR 0.0310408 0.0520947 0.5959 0.5700
CPI 0.0326629 0.0104909 3.113 0.0170 **
EX -0.0961463 0.0595672 -1.614 0.1505
l_GDP -8.36107 4.48064 -1.866 0.1043
l_FDI -3.16064 1.27060 -2.488 0.0417 **
l_SR -0.132992 0.846095 -0.1572 0.8795
l_CPI -0.396777 0.196710 -2.017 0.0835 *
l_EX 1.06170 0.808585 1.313 0.2306
Unadjusted R-squared = 0.799361
Test statistic: TR^2 = 14.3885
with p-value = prob(Chi-square(5) > 14.3885) = 0.0133212

20
 3. Non- linearity test (squared terms)- pvalue<0.05
Auxiliary regression for non-linearity test (squared terms)
OLS, using observations 1990-2010 (T = 21)
Dependent variable: uhat
coefficient std. error t-ratio p-value
-------------------------------------------------------
const 1.15188 2.51199 0.4586 0.6536
GDP 0.0767753 0.710194 0.1081 0.9154
SR -0.0513260 0.0715108 -0.7177 0.4847
FDI -0.389215 0.223322 -1.743 0.1033
sq_GDP -0.00567162 0.0496161 -0.1143 0.9106
sq_SR 0.00146212 0.00180773 0.8088 0.4322
sq_FDI 0.0295555 0.0150712 1.961 0.0701 *
Unadjusted R-squared = 0.357630
Test statistic: TR^2 = 7.51024,
with p-value = prob(Chi-square(3) > 7.51024) = 0.057296

4. Ramsey’s RESET test for model 2- p value >0.05

RESET test for specification (squares and cubes)
Test statistic: F = 1.356261,
with p-value = P(F(2,15) > 1.35626) = 0.287
RESET test for specification (cubes only)
Test statistic: F = 2.761559,
with p-value = P(F(1,16) > 2.76156) = 0.116
RESET test for specification (squares only)
Test statistic: F = 2.857116,
with p-value = P(F(1,16) > 2.85712) = 0.11

5. Plot fitted and actual value of dependent variables

Actual and fitted UR
13
fitted
actual
12

11

10

9
UR

3
1990 1995 2000 2005 2010

21