The Economic Costs of Obesity

and the Economics of SSB Taxes

Keynote Address

Lisa M. Powell, PhD

Distinguished Professor and Director
Health Policy and Administration
School of Public Health
University of Illinois at Chicago
Obesity and Fiscal Policy Training Workshop:
Tax Policy Design and Health Impact in Asia
World Health Organization
Asian Development Bank Institute
Asian Development Bank
July 13-15, 2018
Manila, Philippines

The views expressed in this presentation are the views of the author and do not necessarily reflect the views or policies of the Asian Development Bank Institute
(ADBI), the Asian Development Bank (ADB), its Board of Directors, or the governments they represent. ADBI does not guarantee the accuracy of the data included
in this paper and accepts no responsibility for any consequences of their use. Terminology used may not necessarily be consistent with ADB official terms.
 Outline of Presentation

• Context of Obesity and Burden

 Obesity rates
 Health burden
 Disparities in obesity

• Costs of Obesity
 Direct medical costs for children and adults
 Incremental lifetime costs of childhood obesity
 Mental health
 Human capital costs in childhood
 Productivity costs for adult workers: Absenteeism; Presenteeism
 Wage penalties
 Environmental costs

• Economics of SSB Taxes

 Rationale
 Empirical Evidence
 Examples from Tobacco and Alcohol
 Tax Design and Considerations
 Obesity among Children and Adults Worldwide

• 41 million children < age 5 and over 340 million aged 5-19 were overweight or
obese in 2016.

• The prevalence of overweight and obesity among children aged 5-19 has grown
from approximately 4% in 1975 to 18% in 2016.

• 1.9 billion adults were overweight and 650 million were obese in 2016.

• 39% of adults (39% of men and 40% of women) were overweight

• 13% of the world’s adult population (11% of men and 15% of women) were
obese in 2016.

Source: WHO. Obesity and overweight: Fact sheet, 2018.

 Childhood Overweight

Source: WHO. Report of the Commission on Ending Childhood Obesity. 2016

 Health Burden of Obesity

• Obesity during childhood increases risk of premature

onset of illness (diabetes and heart disease) and
childhood obesity tracks into adulthood

• Obesity is associated with numerous health problems

including:
• cardiovascular disease
• insulin resistance / diabetes
• musculoskeletal disorders
• some cancers
• disability
• mental health illness
 Health Burden of Poor Diet and Obesity
Example: Increasing Risk Factors in the Philippines

Source: Institute of Health Metrics and Evaluation. 2018. www.healthdata.org/philippines

Overall Medical Expenditures/Costs of Obesity

• Examples from the U.S:

• Children: $14.3 billion annually ($14.1B for prescription drugs, ER and outpatient +
$237.6M from inpatient costs)
• Adults: $209.7 billion annually (20.6% of national health expenditures)
• Incremental lifetime direct medical costs of an obese vs a normal weight 10 y old child
(accounting for weight gain among normal weight youth): $12,660
→ Cost for one cohort of 10 y olds: $9.4 billion

• Direct costs health care attributable to obesity for adults in Brazil totaled
US$ 269.6 million with morbid obesity accounting for ~24% (US$64.2M)
but it is 18 times less prevalent.

Source: Trasande & Chatterjee, Obesity, 2009; Trasande et al., Health Affairs, 2009; Cawley & Meyerhoefer, JHE, 2012;
Finkelstein et al., Pediatrics, 2014; de Oliveira et al., PLoS One, 2015
Medical Expenditures/Costs of Obesity for Children

• Compared to normal weight children, overweight and obese children

have higher health care expenditures/costs, respectively, by:
 $180 and $220 in U.S.
 €66 and €266 in Germany

• Health care costs of obese versus normal weight children:

 1.2 times higher in Canada
 1.6 times higher in Australia

Sources: Finlestein and Trogdon, AJPH, 2008; Breifelder et al., Econ Human Biology, 2011; Kuhle et al., Int J. Pediatr Obesity, 2011;
Hayes et al., Obesity, 2016
 Medical Expenditures Due to Obesity for Adults

• Estimates of incremental per capita medical expenditures among

U.S. full-time employees aged ≥ 18 by obesity grade and gender:
 Grade 1 obese (30.0 ≤ BMI ≤ 34.9): Men: $475 Women: $1274

 Grade 2 obese (35.0 ≤ BMI ≤ 39.9): Men: $824 Women: $2532

 Grade 3 obese (BMI>=40.0): Men: $1269 Women: $2395

Source: Finkelstein et al., JOEM, 2010

 Obesity Mental Health and Risky Behaviors

• Reviews found self-esteem was significantly lower among obese

children and adolescents; although some mixed results

• Among school aged children aged 11-17, obesity versus normal weight
associated with:
 Increased smoking among girls
 Younger boys more likely to be victims of bullying
 Older boys more likely to carry a weapon

• Perception of being overweight significantly raises suicide ideation and

attempts for girls

• Among adults, a recent review revealed and inverse relationship

between BMI and the risk of completed suicide; although positively
associated with suicide attempts among women

Sources: Griffiths, Parsons & Hill, Int J. of Pediatr Obesity, 2010; Russel-Mayher et al., J. Obesity 2012; Farhat et al., Am J Prev Med
2010; Dave & Rashad, Soc Sci & Med, 2009; Zhang et al., J Aff Dis, 2013
Youth Obesity and Human Capital Accumulation

• Among young (2-3y) German children obesity is associated with

reduced verbal, social and motor skills and daily activity living skills for
boys and reduced verbal skills for girls

• Obesity among 4th-6th graders found to be positively associated with

school absenteeism:12.2 days versus 10.1 days for normal weight
children

• Among 14-17 y olds, higher BMI associated with lower GPA among
white females but less consistent evidence for nonwhite females and
males

Sources: Cawley and Spiess , Econ & Hum Bio, 2008; Geier et al., Obesity, 2007; Sabia, Southern Econ J, 2007
 Productivity Costs of Adult Obesity

• Obesity, as compared to normal weight status, is associated with 1.1 to 1.7

additional days missed from work annually: a financial burden on U.S. states
ranging from 6.5% to 12.6% of total absenteeism costs in the workplace

• Individuals with a BMI ≥ 30, as compared to those with BMI ≤ 27, has 69%
more total days absent from work

• Absenteeism cost estimates range from $4.3B, $8.65B, $12.8B /year

• Presenteeism cost estimated at $30B /year

• Short-term disability in workplace: overweight and obesity 1.3 and 1.8 higher
odds of short-term disability

Sources: Andreyeva et al., JOEM, 2014; Kleinman et al., JOEM, 2014; Finkelstein, et al., JOEM, 2010; Cawley et al., JOEM, 2007;
Arena et al., JOEM, 2006.
 Productivity Losses Attributable to Obesity:
Absenteeism & Presenteeism

• Estimates of incremental per capita productivity losses from “Absenteeism” by obesity

grade and gender:

 Grade 1 obese (30.0 ≤ BMI ≤ 34.9): Men: $277 & 1.6 days Women: $407 & 3.1 days

 Grade 2 obese (35.0 ≤ BMI ≤ 39.9): Men: $657 & 3.8 days Women: $67 & 0.5 days

 Grade 3 obese (BMI ≥ 40.0): Men: $1026 & 5.9 days Women: $1261 & 9.4 days

• Estimates of incremental per capita productivity losses from “Presenteeism” by obesity

grade and gender:

 Grade 1 obese (30.0 ≤ BMI ≤ 34.9): Men: $391 & 2.3 days Women: $843 & 6.3 days

 Grade 2 obese (35.0 ≤ BMI ≤ 39.9): Men: $1010 & 5.8 days Women: $1513 & 11.0 days

 Grade 3 obese (BMI ≥ 40.0): Men: $3792 & 21.9 days Women: $3037 & 22.7 days

Source: Finkelstein et al., JOEM, 2010

 Per Capita Incremental Medical Expenditures, Absenteeism
and Presenteeism Costs, by Obesity Status and Gender, U.S.

$7000

$6000

$5000

$4000

Presenteeism $
$3000
Absenteeism $
Medical $
$2000

$1000

$0

Source: Data drawn from Table 2. Finkelstein et al., JOEM, 2010

 Obesity and Wages

• Obesity is associated with lower wages

 In the U.S., obesity wage penalty of about 1-3% for men and 2-6% for women

 In Europe, 10% higher BMI associated with earnings penalty of 3% for men and 2% for
women

 Effect found more consistently for women in U.S., particularly white women (e.g., 2 sd
increase in weight associated with 9% lower wages)

 Obesity wage penalties are found in Germany for women in white-collar jobs

 Wage penalties are larger in the U.S. in occupations requiring interpersonal skills

Sources: Baum & Ford, Health Econ, 2004; Brunello and D’Hombres, Econ & Hum Bio, 2006; Cawley, JHR, 2004; Caliendo & Gehrsitz,
IZA, 2014; Han et al., Health Econ, 2009.
 Environmental Costs

• Overweight and obesity are related to one billion additional gallons of gasoline
consumed each year in the U.S. (0.8% of the annual fuel consumption) = $2.7B
 Adds 20 billion pounds of CO2 emissions (0.5% of annual CO2 transport emissions)

• Study of OECD, estimates that a population wide 5kg weight reduction would
lower CO2 emission from transportation by 10 million tons.

• Increased obesity in the U.S. between 1990 and 2000 associated with $275 M in
jet fuel costs in 2000.

Sources: Jacobson and King, Trans Res Part D, 2008. Michaelowa and Dransfield, Ecolog Econ, 2008; Dannenberg et al. AJPM, 2004.
 Summary: Costs of Obesity

Childhood Obesity Adult Obesity

Gender, race/ethnicity, SES, region

Gender, race/ethnicity, SES, region

• Poorer physical health
• Poorer physical health • Poorer mental health
• Poorer mental health • Greater direct medical costs
• Greater direct medical costs • Greater work absenteeism
• Greater school absenteeism • Greater work presenteeism
• Delayed skill acquisition • Lower wages
• Lower test results • Greater environmental costs

Lower Utility
(Direct health effects and due to increased financial constraints)

Health and Wellbeing Disparities

Source: Powell, Lisa M., Conceptual Model: Costs of Obesity, 2016

Where are we Going: Growing Economic Costs

Source: Bloom et al., The Global Economic Burden of Noncommunicable Diseases. Geneva: World Economic Forum. 2011.
 Economic Costs Diabetes

Source: Bloom et al., The Global Economic Burden of Noncommunicable Diseases. Geneva: World Economic Forum. 2011.
Comprehensive “Actions” to Reduce Obesity and Costs
 Economics of SSB Taxes:
Rationale for using Fiscal Policy
• Rationale for fiscal policy intervention: over-consumption leading to
increased medical costs, lost productivity, etc.

• Over-consumption is associated with negative externalities that are not

accounted for in the “private” market because the cost of the externalities
are not included in the market price.

• A “Pigouvian” tax (set equal to the social cost of the negative

externality) is one way to help internalize the external costs.

• Idea is that the fiscal policy instruments change relative costs which, in
turn, impact behavior choices related to consumption.

• Impact on consumption can be measured by price elasticity of demand:

% change in consumption as a result a 1% change in price.

• Fiscal policies have broad population reach. Nonetheless, they should be

considered as part of a comprehensive policy approach.
 Selected Food Price Trends in the U.S., 1980-2014
Inflation Adjusted

160

150

140

130

120

110

100

90

80

70

60

50

Fruits & Vegetables Fresh Fruits & Vegetables Carbonated Drinks

Source: Bureau of Labor Statistics, 2015. Powell’s calculations.

Evidence: Impact of SSB Prices / Taxes on the Demand for SSBs

Mean Estimates of Price Elasticity of Demand for SSBs

U.S. Studies from 2007-2012

a Overall weighted mean (based on SSB consumption shares) from the estimates from the aggregated

SSB category and estimates from the disaggregated categories within the beverage demand system.

Source: Powell et al., Obesity Reviews, 2013

 Recent Evidence on SSB Tax Impacts

• Mexico: Federal excise tax of 1 peso/L on SSBs (approx 10% price

increase based on 2013 prices); effective Jan 1, 2014
 Average volume of taxed beverages purchased was 6% lower in first year post-tax and was 12% lower by December
2014; greater impact among low-income households (-17.4% by Dec 2014)
 Recent 2017 study finds sustained impact: household purchases of SSBs was 5.5% lower one year post-tax and
9.7% lower in the second year post-tax

• Berkeley, CA: $0.01/oz SSB tax; effective March 2015

 At 4 months post-tax, SSB consumption frequency among individuals living in low-income neighborhoods fell 21%
compared to a 4% increase in comparison cities
 Water consumption increased 63% compared to 19% in comparison cities

 At 1-year post-tax, SSB sales in Berkeley stores fell by 9.6% and sales of untaxed beverages rose by 6.9%.
o Substitution was mostly to plain water (up 15.6%). Interestingly, diet soda and energy drinks were down 9.2%.
 No statistically significant changes in usual SSB intake from self-reported data

• Chile: 13%→18% high-sugar (H) SSBs (>6.25g sugar/100 mL); 13%→18% low-
sugar (L) SSBs (<6.25g sugar/100 mL); effective October 2014
 Post-tax (Oct 2014 through Dec 2015), H-SSB purchases decreased by 3.4% (greater effect among high-SES
households); L-SSB purchases increased by 10.7%; untaxed beverage purchases fell by 3.1%.

Source: Colchero et al., BMJ, 2016; Colchero et al., Health Affairs, 2017; AJPH, 2016; Silver et al., PLOS Medicine, 2017;
Caro et al., PLOS Medicine, 2018.
Evidence on SSB Price/Tax Effects on Body Weight Outcomes

Source: Powell et al. Obesity Reviews, 2013

 Evidence from Tobacco: Price Effects on Smoking

Evidence shows that taxes that raise the prices for

cigarettes significantly reduce smoking:

• 10 percent price increase would reduce smoking by:

 2-6% reduction in HICs, with an average of 4%
 2-8% reduction in LMICs, with an average of 5%
 about half of the impact is on number of smokers, other half on
consumption among continuing smokers
• Young people 2-3 times more sensitive to price
• Lower income groups generally more responsive to price
Evidence from Alcohol: Price Effects on Drinking and Consequences

Evidence shows that taxes that raise the prices for

alcoholic beverages significantly reduce drinking:

• 10 percent price increase would reduce:

 Beer consumption by 1.7 to 4.6 percent
 Wine consumption by 3.0 to 6.9 percent
 Spirits consumption by 2.9 to 8.0 percent
 Overall consumption by 4.4 percent
 Heavy drinking by 2.8 percent
 Generally larger effects for youth and young adults

• Evidence that taxes on alcohol reduce: alcohol-related diseases and

mortality, traffic-related deaths, violence, crime, work-place accidents,
teenage pregnancy and sexually transmitted diseases.

Source: Wagenaar et al., Addiction , 2009, Wagenaar et al., AJPH, 2010, Xu and Chaloupka, Alcohol Research & Health, 2012.
Example from Tobacco: Cigarette Prices & Adult Smoking Prevalence U.S.
Inflation Adjusted 1970-2013

$6.00
37
$5.50

33 $5.00

$4.50

Price per Pack (1/14 Dollars)

29
Prevalence

$4.00

$3.50
25
$3.00

$2.50
21

$2.00

17 $1.50
1970 1973 1976 1979 1982 1985 1988 1991 1994 1997 2000 2003 2006 2009 2012
Year

Prevalence Price

Figure courtesy of Dr. Frank Chaloupka. Source: Tax Burden on Tobacco, BLS, NHIS, Dr. Chaloupka’s calculations
 Example from Tobacco: Adult Prevalence & Price, Brazil

Adult Smoking Prevalence and Cigarette Price

Brazil, Inflation Adjusted, 2006-2013
16

16 5.4

15

15
Adult Smoking Prevaleence

4.9

Price per Pack, 2013 BRL

14

14
4.4
13

13

3.9
12

12

11 3.4
2006 2007 2008 2009 2010 2011 2012 2013
Sales, Million Sticks Price per Pack, 2013 BRL

Figure courtesy of Dr. Frank Chaloupka. Source: Ministry of Health, Brazil; World Bank
Example from Tobacco: Smoking, Tax and Male Lung Cancer, France
1980-2010

6.0
# cigarettes/adult/day
300
Number/adult/day and death rates

5.5

5.0

Price (% relative to 1980)

Lung cancer death rates per 100,000 (divided
by four): men age 35-44 250
4.5

4.0
200
3.5

3.0
150
2.5
Relative price
2.0 100
1.5

1.0 50
1980 1985 1990 1995 2000 2005 2010
Year

Figure courtesy of Dr. Frank Chaloupka. Source: Jha, Global Heart, 2012
 Tax Design: Base and Type

• The appropriate tax base depends on the objective of the tax

 Public health objective to reduce sugar intake suggests a tax on all forms of
SSBs (i.e., all beverages containing free sugars).
 Broader tax base helps to minimize substitution

• Tax type:
 Excise tax versus sales tax
o Incorporated at shelf price – more apparent to consumers
o Applicable regardless of where items are sold

 Specific (per unit) versus ad valorem (% of price) excise tax

o Quantity discounts are still taxed
o Reduces incentives to switch to cheaper brands
o Needs to be adjusted for inflation
 Tax Design: Tax Rate

• How large should the tax be to generate a meaningful impact?

 WHO recommends at least 20%

• Recent taxes in Mexico, Barbados and Dominica implemented at 10%

• Penny per ounce in the U.S. equates to approximately a 17% tax

• Flat per unit tax across SSB types equates to lower a % tax on higher
priced SSBs; e.g., based on 2012 US data, a 1¢/oz equates to:
 20% for soda, 17% for sports drinks, and only 5% for energy drinks

• Tiered tax rate based on sugar content

 E.g., UK tax with thresholds of 5g/100ml (tax of 18p/L) and 8g/100ml (tax of 24p/L) ~
based on current exchange rates, this equates to a tax of about 0.7¢/oz for sugar
content above 18g/12oz and 0.9¢/oz above 28g/12oz
 Tax Design: Earmarking of Tax Revenue

• A portion of the tax revenue may be earmarked for specific

government programs.
• Earmarking can be important to help garner public support.
• Earmarking for nutrition and physical activity-related programs will
complement the intended health impact of the tax.
• Earmarking toward low-income and minority populations can help
to address health disparities
 Thank you!

Lisa M. Powell, PhD

Distinguished Professor and Director
Health Policy and Administration
Director, Illinois Prevention Research Center
School of Public Health
University of Illinois at Chicago

powelll@uic.edu