Obesity and Poverty in America: time and demographic trends

Roland Sturm, Ph.D.Senior Economist

RAND

Why the Recent Obesity Epidemic?

• Many competing hypotheses

� Transportation

� Television

� More work hours

� Not enough exercise

� Fast food

� Affluence

� Poverty

• Compatible with facts?

Different Theories Need Not Be Contradictory

• Scope of theories narrower than proponents assume

• Synthesis often better than “either … or”

• Misunderstandings because no common knowledge

– facts or assumptions?

• Conventional wisdom unreliable guide

• Let’s start with looking at the bigger picture and data about societal trends

Outline of Talk

• Poverty and Obesity

• What has changed in our lives? Trends in time and money

• Time use differences by income groups

Obesity and Poverty?

• Not new idea that obesity has adverse health effects

• But isn’t obesity really a disease of affluence?

• Traditionally the privilege of the wealthy

– Still is in many countries

– Shouldn’t that reduce socioeconomic

health disparities?

The fat cats were fatter 50 years ago

Overweight by Family Income, US men 1959-1962

0

10

20

30

40

50

Overweight Obese

Percent

<$1,000

>$10,000

Based on data from National Health Examination Survey, 1959-1962

But not women

Overweight by Family Income, US women 1959-1962

0

5

10

15

20

25

30

Overweight Obese

<$1,000

>$10,000

Based on data from National Health Examination Survey, 1959-1962

The Poor Are Often Thinner in Other Countries

Prevalence of Overweight by Family Income in China,

1999-2000

05

10152025303540

Poor Middle Rich Poor Middle Rich Poor Middle Rich

Overall Rural Urban

Source: Xu F et al., Public Health Nutrition, 2005

117 106186

9733

471

188

209

107

2445

5226

309

267

389

122

252

117

129

59

106

12

9

0

200

400

600

800

1000

1200

1990 2020 1990 2020

De

ath

ra

tes

pe

r 1

00

,00

0

All others Infections and parasitic diseases

Other noncommunicable diseases Malignant neoplasms

Diabetes CVD

Even if Beyond the Health Transition:Mortality trends in China and India, 1990 and 2020

China India

Murray and Lopez. Global Burden Disease.Harvard University Press, 1996.

In the US, Obesity and Poverty now occur together

more often …

…Among Youth

Overweight among Children and Adolescents by

Federal Poverty Level, 2001-2004

0

5

10

15

20

25

<100% <200% 200%+

By Poverty

6-11 years

12-19 years

Based on Table 74, NCHS Chartbook, 2006

… Among Women …

Based on data from NHANES, 2001-2004

Overweight and Obesity by Family Income, Women,

2001-2004

05

10

15202530

354045

Overweight Obese

<$20,000

>$65,000

… But Not Among Men

Based on data from NHANES, 2001-2004

Overweight and Obesity by Family Income, Men, 2001-

2004

0

10

20

30

40

50

Overweight Obese

<$20,000

>$65,000

Subpopulations seem to have gained similar weight

in last 20 years

Trends in Average Body Mass Index

by Education

22

23

24

25

26

27

28

1986 1989 1992 1995 1998 2001

Year

Av

era

ge

BM

I

No High School

High School

Some College

College

Source: Truong and Sturm, AJPH, 2005 Sep;95(9):1602-6

• But consistent differences at every point in time

• Reversal must have occurred prior to “obesity epidemic”

Trends in Average Body Mass Index

by Relative Income

(but income categories not really comparable over

time)

23

24

25

26

27

28

1986 1989 1992 1995 1998 2001

Year

Ave

rag

e B

MI

Lowest Income

Highest Income

Source: Truong and Sturm, AJPH, 2005 Sep;95(9):1602-6

Outline of Talk

• Poverty and Obesity

• What has changed in our lives? Trends in time and money

• The role of income

Time and Money

• What do people do?

�Time use data

• Why study time use?

�Scarcest resource

�Regardless of economic growth, a day has 24 hours

� In contrast to money, everybody has the same amount each day

• Where does the money go?

�Gross output by industry

Where Does the Money Go?

• Leisure time industries outpace GDP growth

• “Active” industries at twice the rate of GDP

– Sporting goods, fitness clubs, dance studio

• “Passive” industries grow even faster

– Entertainment electronics

– Spectator sports

• Food expenses declines as share of income, but buy much more energy

• Pushes puzzle one step back: Why the increase?

• Technological progress

– Revolution in mass production of food

Retail: Sporting Goods Doubled, While Real GDP Increased 50%

0

2000

4000

6000

8000

10000

12000

1987

1988

1989

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

Millio

ns

19

96

do

lla

rs

Sporting goods and bicycle shops

Source: Bureau of Economic Activity, Gross Domestic Product By Industry.

But Dwarfed By Growth of Home Electronics (TV, DVD)

0

10000

20000

30000

40000

50000

1987

1988

1989

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

Mil

lio

ns

19

96

do

lla

rs

Radio and TV stores Sporting goods and bicycle shops

Source: Bureau of Economic Activity, Gross Domestic Product By Industry.

Industries More Closely Related to PA: Sports Clubs Doubled …

0

2000

4000

6000

8000

10000

12000

14000

16000

18000

20000

1987

1988

1989

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001In

du

str

y i

n m

illi

on

s (

19

96

do

lla

rs)

Sports Clubs

Source: Bureau of Economic Activity, Gross Domestic Product By Industry.

But Sedentary Spectator Sports Grew Faster…

0

2000

4000

6000

8000

10000

12000

14000

16000

18000

20000

1987

1988

1989

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

in m

illi

on

s (

19

96

do

lla

rs)

Sports Clubs Professional sports clubs and promoters

Source: Bureau of Economic Activity, Gross Domestic Product By Industry

Source: USDA, Economic Research Service

While Leisure Industries Are Growing Fast, Food Industries Decline in Importance …

Food Expenditure as Share of

Disposable Personal Income

0.0

5.0

10.0

15.0

1973

1976

1979

1982

1985

1988

1991

1994

1997

2000

Perc

en

t

Food At Home Food Away From Home Total

Source: USDA, Economic Research Service

… But Supply More Calories

U.S. Food Supply, Energy per capita per day

3,000

3,200

3,400

3,600

3,800

4,000

kc

al

Source: BLS

Price Indices

50.0

100.0

150.0

200.0

250.0

300.0

1978

1980

1982

1984

1986

1988

1990

1992

1994

1996

1998

2000

2002

1982-1

984=100

FreshFruits andVegetables

ConsumerPriceIndex

SugarsandSweets

Soft Drinks

Price trends in food types likely to

exacerbate income effects

• Technological progress and production efficiencies affect relative prices

– Less change for perishable/fresh goods means higher prices

• Double hit for lower income groups

– “prudent” diets become relatively more expensive everywhere

– Perishable, more expensive foods less available in small stores

– Even though generally more food stores in low income neighborhoods, smaller size and less efficient means higher prices

Where does the time go?

Source: 1965-1985: Robinson and Godbey; 1999: my calculation using FISCT 1999: 2003-2005: ATUS

0

50

100

150

200

250

300

350

400

1965 1975 1985 1999 2003-2005

Paid Work House/Dep Care Free Time

Television Eating/Drinking

Active Leisure Time Keeps Increasing

• In BRFSS, additional 20 mins/week in last decade in physical activity for the “typical” American

• Percent of inactive keeps falling

� From 30.7 in 1990 to 27.4 in 2000

� More than 3 percentage point

• But small relative to total increase in free time

Home Production

• Drop of 5 hours/week

• Biggest reduction in cleaning and meal preparation

• Meal preparation more important for diet than

physical activity

� Wider range and variety of prepared food available at lower TIME costs

� Variety of snacks always available, including while watching TV/DVD

Trends in Time Use Differ For Children

• For adults 18-64, free time increased by 8 hours per week from 1965 to 2003-2005

• increase for both men and women

• in or out of labor force

• For children 3-12, free time decreased by 7 hours per week from 1981 to 1997

Changes in Weekly Minutes Spent on Activities from 1981 to 1997, Age 3-12

Figure 2: Changes in Weekly Minutes Spent on Activities from 1981 to 1997, Age

3-12

Source: calculation based on data from Hofferth and Sandberg (2001b)

(162)(138)

(246)

6173107

366

-400

-300

-200

-100

0

100

200

300

400

Min

ute

s p

er

week

Sch

oo

l, D

aycare

, S

tud

yin

g,

Read

ing

, A

rt

Acti

vit

y

Pers

on

al care

Sp

ort

s / O

utd

oo

rs

Sh

op

pin

g

Pla

yin

g

Oth

er

passiv

e,

Co

nvers

ati

on

s,

Ch

urc

h, V

isit

ing

, L

eis

ure

Tele

vis

ion

Source: calculations based on data from Hofferth and Sandberg (2001)

Outline of Talk

• Poverty and Obesity

• What has changed in our lives? Trends in time and money

• Time use differences by income groups

Work and Leisure Time

Among Women By Family Income

050

100150200250300350400

Paid

Wor

k

Hou

sewor

k

Dep

Car

e

Tota

l Wor

kLe

isur

e,Soc

ializ

ing

trans

porta

tion

<$20k 20-40k 40-60k >60k

More Leisure Time

Among Lower Income Full Time Workers

• Lowest versus highest earnings group among full time workers

– more TV watching (22 mins/day), more games (7 mins)

– more socializing (15 mins)

– Less reading (8 mins/day)

– Less active participation in sports (7.5 mins)

Leisure Time for Full Time Workers By Weekly

Earnings

0

50

100

150

200

250

300

Total leisure TV Active Sports

Min

ute

s/d

ay

<$450 $451-700 $701-1075 >$1,075

Differences between lowest and highest income

women (adjusted for sociodemographics)

• Positive difference: more time in low income group

• No substantial or highly significant differences in:

– Grocery shopping

– travel time for groceries

– Personal care

– Housework other than food preparation

– Religious activities

– Telephone calls

• Race/ethnicity other important characteristic

MinutesActivity

34TV

-5Eating

-12Transportation

-4Volunteer

-7Sports, Exercise

34Leisure, Relax

-13Education

-26Paid Work

13Child Care

10Food Preparation

17Sleep

Differences between lowest and highest income men

(adjusted for sociodemographics)

• Positive difference: more time in low income group

• No substantial or highly significant differences in:

– Grocery shopping

– travel time for groceries

– Personal care

– Education

– Religious activities

– Telephone calls

MinutesActivity

42TV

-10Eating

-13Transportation

-14Sports, Exercise

36Leisure, Relax

-8Paid Work

3Child Care

-14Household Work

23Sleep

Summary

• Obesity and poverty currently correlated among women and youth in US

• Income marker for vulnerable groups even if causes and consequences are less clear

• Does obesity cause poverty?

– marriage market likely pathway for women

• Does poverty causes obesity?

– Some plausible pathways exist, empirical evidence less clear, one of the themes today

• Or unobserved other factors?

– Likely for some subgroups (psychotic disorders)

• “The German Women are generally so disagreableto an English Eye, that it wou’d require great Portions to induce Englishmen to marry them…..dick und starcke, that is, thick and strong, always enters into their Description…”

Benjamin Franklin, ca. 1753, about immigrants