1 wealth and-health_of_children_in_india

WEALTH AND HEALTH OF CHILDREN IN INDIA :

A STATE-LEVEL ANALYSIS

Diane Coffey, Aparajita Chattopadhyay & Rajan Gupt

Objectives: What do relationships between wealth and health look

like within India, a nation whose states are as populous as many other countries?

We also describe how these relationships seem to be changing in recent years.

We present a state level analysis of the association between state net domestic product per capita and children’s health indicators ( Infant mortality, Height for age Z score and Proportion Stunted).

The contradictions and the questions:

Preston’s describes a strong, log-linear relationship between country level mortality and income, suggesting that cross country wealth is strongly associated with health, but that this relationship is less steep at higher levels of wealth.

Yet, measures of health do not always behave in the same way. For example, Preston’s famous cross country correlation between mortality and wealth contrasts with the puzzling result from Deaton that average height across countries, another important measure of population health, does not correlate with gross domestic product per capita.

Bozzoli et al, 2009: no association between adult heights and

GDP per capita in European countries.

Fogel, 2004 and others: in resource constrained settings, there is a

strong association between income and stature

James & Syamala, 2010:strong association between rising incomes in India and

longer life expectancies.

the relationship between income and life expectancy has become less steep over time.

Coffey, 2012for state cohorts born in India between 1970 and 1983,

there is a robust relationship between state net domestic product per capita in a cohort’s year of birth and the state cohort’s adult height.

Indian Economy grew fast in the last decade. But this economic growth has not led to commensurate improvements in health (Drèze and Sen, 2011).

Deaton and Drèze, 2002: rates of decline in child mortality in India do not match the

unprecedented rates of economic growth.

Declines in child malnutrition as measured by anthropometric measures such as height and weight have also been slow (Radhakrishnan & Ravi, 2004).

Subramanyam et al., 2011: growth does not predict indicators of anthropometric faltering.

So the questions are:

What is the relationship of child health and wealth in India?

Does this relationship change over time?

What is the relationship between growth of income and health improvements?

Data and methods:

NFHS, 1998-99 & NFHS, 2004-2005

§ Child height

§ Stunting prevalence

§ Infant mortality

o EPW Research Foundation

§ State net domestic product per capita, base year 1993

o Census of India, 1991, 2001, 2011

§ Population weights

Only the heights of children under three years old were used.

Results:

Relationship between aggregate wealth and child’s health

1

2

3

4

5

6

7

8

9

11

12

13

14

15

16

18

19

21

2223

24

25

26

10 20

0.2

.4

.6

fraction stunted according to CDC-W

HO

77

0 5000 10000 15000 20000 25000 30000st at e net domest ic product in 1997 (1993 pr ices)

f r act ion of st unt ed childr en under 3 Fit t ed values

NFHS 2

23

4

5

6

7

8

910

11

12

13

14

15

16

19

20

21

22

23

25

261

0.2

.4

.6

fraction stunted according to CDC-W

HO

77


f r act ion of st unt ed childr en under 3 Fit t ed values

NFHS 3

size of circle proport ional to approximat e populat ion of children under 6

Stunting and NSDP 1 Andhra Pradesh

2 Arunchal Pradesh

3 Assam

4 Bihar & Jharkhand

5 New Delhi

6 Goa

7 Gujarat

8 Haryana

9 Himachal Pradesh

10 Jammu & Kashmir

11 Karnataka

12 Kerala

13 Madhya Pradesh & Chhattisgarh

14 Maharashtra

15 Manipur

16 Meghalaya

17 Mizoram

18 Nagaland

19 Orissa

20 Punjab

21 Rajasthan

22 Sikkim

23 Tamil Nadu

24 Tripura

25 Uttar Pradesh & Uttarakhand

26 West Bengal

1

2

34

5

6

7

8

9

1011

12

13

14

15

16

18

19

21

22 23

24

25

20

26

-2.5

-2

-1.5

-1

height for age z-scores, CDC-W

HO

77


aver age height f or age z- scor e of childr en under 3 Fit t ed values

NFHS 2

2

3

4

5

6

7

8

910

11

12

13

14

15

1619

20

21

22

23

25

261

-2.5

-2

-1.5

-1

height for age z-scores, CDC-W

HO

77


aver age height f or age z- scor e of childr en under 3 Fit t ed values

NFHS 3


Height for age Z score and NSDP1 Andhra Pradesh

2 Arunchal Pradesh

3 Assam

4 Bihar & Jharkhand

5 New Delhi

6 Goa

7 Gujarat

8 Haryana

9 Himachal Pradesh

10 Jammu & Kashmir

11 Karnataka

12 Kerala

13 Madhya Pradesh & Chhattisgarh

14 Maharashtra

15 Manipur

16 Meghalaya

17 Mizoram

18 Nagaland

19 Orissa

20 Punjab

21 Rajasthan

22 Sikkim

23 Tamil Nadu

24 Tripura

25 Uttar Pradesh & Uttarakhand

26 West Bengal

Infant Death and NSDP

1

2

3

4

5

6

7

8

1011

12

13

1415

16

18

19

20

21

2223

24

25

26

9

0.02

.04

.06

.08

fraction of infant deaths am

ong births in 3 years prior to survey


f ract ion of inf ant deat hs Fit t ed values

NFHS 2

12

3

4

5

6

7

810 11

12

13

1415

16

19

20

21

22

23

25

26

9

0.02

.04

.06

.08

fraction of infant deaths am

ong births in 3 years prior to survey


f ract ion of inf ant deat hs Fit t ed values

NFHS 3


1 Andhra Pradesh2 Arunchal Pradesh3 Assam4 Bihar & Jharkhand5 New Delhi6 Goa7 Gujarat8 Haryana9 Himachal Pradesh

10 Jammu & Kashmir11 Karnataka12 Kerala13 Madhya Pradesh & Chhattisgarh

14 Maharashtra15 Manipur16 Meghalaya17 Mizoram18 Nagaland19 Orissa20 Punjab21 Rajasthan22 Sikkim23 Tamil Nadu24 Tripura25 Uttar Pradesh & Uttarakhand26 West Bengal

Magnitude of the association between wealth and health

The relationships between wealth and children’s health are weaker in the later survey than the earlier survey

Pooled regression :

NFHS2 NFHS3

a difference of 1.5 infant deaths per thousand

a difference of 1 infant death per thousand.

a 7 percentage point difference in stunting prevalence,

a 3 percentage point difference in stunting prevalence.

0.25 standard deviation difference in the average height for age z-score of children under 3.

0.13 standard deviation difference in the average height for age z-score of children under 3.

a 5000 rupee difference in state net domestic product per capita is associated with

State level relationship between growth of income and health improvements

1

2

3

4

5

6

7

89

10

1112

13

14

15

1619

20

21

2223

25 26

-.2

-.1

0.1

change in fraction stunted

0 . 1 . 2 . 3growt h in st at e net domest ic product per capit a bet ween NFHS 2 and 3

change in st unt ing Fit t ed values

1

2

3

4

567

8910

1112

13

14

15

1619

20

21

222325

26

-.5

0.5

1

change in average height for age z-score


change in average height f or age z-score Fit t ed values

1

2

3

45

6

7

8

910

1112

13

1415

16

19

20

21

22

2325

26

-.04

-.02

0.02

change in fraction of infant deaths


change in f ract ion of inf ant s who died Fit t ed values

size of circle proport ional to average of 1998 and 2005 under 6 populat ion

1 Andhra Pradesh2 Arunchal Pradesh3 Assam4 Bihar & Jharkhand5 New Delhi6 Goa7 Gujarat8 Haryana9 Himachal Pradesh

10 Jammu & Kashmir11 Karnataka12 Kerala

13Madhya Pradesh & Chhattisgarh

14 Maharashtra15 Manipur16 Meghalaya17 Mizoram18 Nagaland19 Orissa20 Punjab21 Rajasthan22 Sikkim23 Tamil Nadu24 Tripura

25Uttar Pradesh & Uttarakhand

26 West Bengal

Why is economic growth associated with less improvement in children’s health?

This might be the case if states which were already healthier experienced more economic growth over the period under study; in states which already had healthier children, it would have been harder to make improvements in health

There is a high correlation between state averages of children’s health in 1998 and economic growth between the two rounds of the survey.

The correlation between height for age in 1998 and economic growth between the two survey rounds was 0.4; comparable figures for stunting and infant death are -0.31 and 0.39.

States with better initial health environments (1998), experienced more economic growth in the period under study.

Findings: aggregate wealth and children’s health indicators are

positive for all three chosen health indices.

Yet the association was less steep in the mid-2000s than in the late 1990s.

negative relationship between growth in state net domestic product per capita and improvement in state level children’s health indicators between these two surveys.

That means, more state level income growth between the late 1990s and mid-2000s is associated with less improvement in child health indicators.

Discussion and Concluding remarks Finding 1:

Level of wealth and health are strongly positively associated in India.

Part of the association is because wealth allows people to afford better food, medical care and better home environments. Societies that are richer can invest in public goods like sanitation, vector control and education.

Finding 2:

But the association between wealth and health outcomes is becoming weaker with time.

Plausible reasons:

○ Based on IHDS (2005) data it is observed that within India, Gini coefficients extend from Chhattisgarh (3.8) and Delhi (3.9) to Karnataka (5.2),Kerala (5.4) Gujarat and the NE.

○ This range is even greater than for the differences between Sweden (2.4) and the United States(3.7) (Nanneman & Dubey, 2011). Gini for India is 5.2.

○ Interestingly, within India, higher income states have almost the same

average levels of inequality as lower income states

(r= 0.04).

○ Development of basic factors that help improvement of health is too slow

In MH (Comprehensive Nutrition Survey)

No toilet facility, 43 % improved sanitation, 38 %, food insecure ( HFIAP), 43% had no food to eat last month at least once, 11 % cooking inside house with no separate room,

40% Exclusively breastfed, 60 % The dietary diversity of food for 6-23 months

children are appallingly low.

( report to be released in August, 2013)

Finding :3

States which improved their state net domestic products per capita did not see corresponding improvements in children’s health, and some states which did not achieve faster economic growth did improve children’s health indicators.

Why is economic growth associated with less improvement in children’s health?

This might be the case if states which were already healthier experienced more economic growth over the period under study;

in states which already had healthier children, it would have been harder to make improvements in health.

Indeed, it was those states with better initial health environments, experienced more economic growth.

Going forward

These associations suggest that policy makers must work to improve children’s health.

1. good governance and women’s education : Long term

2. Short term:

factors like women’s health and sanitation are likely to be increasingly important reasons for differences in health across states in India and between India and the rest of the world.

make a difference in the 1000 days window improvements to women’s health and nutrition in India

Sanitation and hygiene:60% of the people who openly defecate without a toilet or latrine live in India, and the country’s progress in improving latrine coverage has been worse than Pakistan, Bangladesh and much of sub-Saharan Africa.

Health = wealth

for growth in Wealth you must have a good Health

Creating demand side through Household empowerment

Improving supply side of basic needs

Health Wealth

Thank you

1 wealth and-health_of_children_in_india

Travel

gross domestic product

relationship of child

fitted values nfhs

state cohorts adult

himachal pradesh

arunchal pradesh

andhra pradesh

prices average height