The longterm consequences of malnutrition on children: New evidence from Guatemala John Hoddinott...

17
The longterm consequences of malnutrition on children: New evidence from Guatemala John Hoddinott International Food Policy Research Institute April 2009

Transcript of The longterm consequences of malnutrition on children: New evidence from Guatemala John Hoddinott...

Page 1: The longterm consequences of malnutrition on children: New evidence from Guatemala John Hoddinott International Food Policy Research Institute April 2009.

The longterm consequences of malnutrition on children: New evidence from Guatemala

John Hoddinott

International Food Policy Research Institute

April 2009

Page 2: The longterm consequences of malnutrition on children: New evidence from Guatemala John Hoddinott International Food Policy Research Institute April 2009.

INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE

Introduction

Improving the nutritional status of pre-school children – in terms better anthropometric outcomes and reduced prevalence of micro-nutrient deficiencies – is receiving increased attention from the World Bank, bilateral donors and some developing country governments:• The World Bank’s “Repositioning nutrition as central to development”

Two rationales are given for this renewed emphasis on improving nutrition:• Its intrinsic value– Being well nourished is a “good thing” • Its instrumental value - Malnutrition has high costs in terms of poor health,

cognitive development and loss of earnings

Page 2

Page 3: The longterm consequences of malnutrition on children: New evidence from Guatemala John Hoddinott International Food Policy Research Institute April 2009.

INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE

Introduction

While there is a growing body of evidence of the instrumental value of nutrition, much of it discussed in “Repositioning …”, this evidence suffers from two significant weaknesses:• Many of the benefits are measured over a relatively short time horizon. From

these studies, it is not clear if the beneficial effects of improving dimensions of nutrition are sustained or if they “fade out” over time.

• Much of the evidence on the economic benefits is indirect. For example:• We observe a causal relation between improved nutrition and some

dimension of human capital, say cognitive skills• We assume that this causal relationship will persist• We draw evidence from the extant literature on the relationship between

cognitive skills and some economic outcome, say wages.• We assume that this relationship is applicable to the setting in which the

improved nutrition has taken place• We adduce a relationship between improved nutrition and this economic

outcome

Page 3

Page 4: The longterm consequences of malnutrition on children: New evidence from Guatemala John Hoddinott International Food Policy Research Institute April 2009.

INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE

Introduction

In this presentation, I present new evidence on the instrumental value of investing in nutrition. This is taken from a long term, longitudinal study of a nutrition supplementation trial that was fielded in rural Guatemala.

Using the data collected during the trial and afterwards, I will discuss:• The impact of one dimension of malnutrition, stunting, on wages in adulthood• The pathways by which this impact occurs and• The long term effects of the intervention itself

In doing so, I draw on ongoing and published work with my collaborators, particularly Jere Behrman, John Maluccio, Reynaldo Martorell and Raphael Flores

Page 4

Page 5: The longterm consequences of malnutrition on children: New evidence from Guatemala John Hoddinott International Food Policy Research Institute April 2009.

INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE

Guatemala Study

Page 6: The longterm consequences of malnutrition on children: New evidence from Guatemala John Hoddinott International Food Policy Research Institute April 2009.

INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE

Guatemala Study

Two matched-pair sets of villages (small & large)

Randomization into: • Atole – high protein, high calorie, micronutrients• Fresco – no protein, low calorie, micronutrients

Centrally located feeding center open at convenient hours, on-demand service

Free preventative healthcare

Enrolled all 0-7 year olds in 1969, all new births included until after supplementation ended in 1977 (n=2392)

Extensive data collected on anthropometric status

Page 6

Page 7: The longterm consequences of malnutrition on children: New evidence from Guatemala John Hoddinott International Food Policy Research Institute April 2009.

INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE

Guatemala Study

In 2002-04, these individuals re-surveyed.

Of the 2392 original participants• Excluded deceased (11%), international migrants (7%) and untraceable (4%)• 1855 (78%): Targeted sample living in Guatemala in 2002• 1471 (62% original; 79% targeted) were resurveyed

Correcting for potential attrition bias has no significant impacts on any of the results discussed here

Respondents were asked about:• Occupation, Income & wages• Underwent a physical examination; and• Completed tests on reading, vocabulary and nonverbal cognitive skills

Community level data was collected on school characteristics and changes in livelihoods

Page 7

Page 8: The longterm consequences of malnutrition on children: New evidence from Guatemala John Hoddinott International Food Policy Research Institute April 2009.

INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE

Guatemala Study

The key feature of this study is that we have data on the following:

• Nutritional status as a pre-schooler• Exposure to a randomized nutrition intervention• Family background and community characteristics• Height, measures of strength and other measures of health• Earnings in adulthood• Measures of completed schooling and selected cognitive skills in

adulthood• Reading comprehension and vocabulary (Inter-American Reading

test) and• Nonverbal cognitive skills (Raven’s Progressive Matrices)

Page 8

Page 9: The longterm consequences of malnutrition on children: New evidence from Guatemala John Hoddinott International Food Policy Research Institute April 2009.

INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE

Results (1): What is the impact of stunting on adult productivity?

Modeling strategy:• Dependent variable is log of income earned per hour worked:

• Income earned from wage labor, own business activities, own-farm production

• KEY right hand side variable is dummy variable equaling one if individual was not stunted at age 72 months, zero if stunted.

• Stunting is defined as having a height, standardized for age and sex, that is two standard deviations or more below the median height of a reference population

• Being stunted is treated as endogenous using exposure to the intervention from 0-36 months, exposure from 0-36 months x atole , log mother’s height

• We have excellent IV diagnostics• Other controls include sex, potential labor market experience, labor market

conditions at age 15, school characteristics at age 7, family background

Page 9

Page 10: The longterm consequences of malnutrition on children: New evidence from Guatemala John Hoddinott International Food Policy Research Institute April 2009.

INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE

What is the impact of stunting on adult productivity?

With full set of controls, not being stunted raises wages by approximately 31 percent

Effect is larger for men (57%) than for women (22%)

Effect is larger for individuals who were severely stunted (< 3 SD), 48%

We can compare the impacts of not being stunted with the impact of improving student-teacher ratios. Based on our preliminary estimates, it would take a 40% reduction in student-teacher ratios at age 7 to get the equivalent increase in earnings that we get when we “switch” a child from being stunted to “not stunted.”

Page 10

Page 11: The longterm consequences of malnutrition on children: New evidence from Guatemala John Hoddinott International Food Policy Research Institute April 2009.

INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE

Results (2): What are the pathways by which pre-school nutrition affects wages?

The nutrition literature suggests two pathways that could generate these results:• Individuals who were stunted as pre-schoolers might have reduced cognitive

skills in adulthood• Individuals who were stunted as pre-schoolers are shorter as adults• We can think of these pathways as education or cognition (“brains”) and height,

health or strength (“brawn”)

Modeling strategy:• Dependent variable is log of income earned per hour worked:• KEY right hand side variables are measures of cognitive ability:

• Completed grades of schooling, reading/vocabulary score, Raven’s progressive matrices

• AND• Measures of strength:

• Height, fat-free mass, hand grip strength, cardio-vascular endurance

Page 11

Page 12: The longterm consequences of malnutrition on children: New evidence from Guatemala John Hoddinott International Food Policy Research Institute April 2009.

INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE

What are the pathways by which pre-school nutrition affects wages?

Cognitive skills - “brains” - have a large effect on earnings however measured.• For example, a one standard deviation increase on a reading/vocabulary test

raises hourly returns to labor by about 40%.

By contrast, none of the measures of strength have a causal effect on earnings

This suggests that it is the impact of poor nutrition on the development of cognitive skills that is the driver of the economic case for investing in nutrition

Page 12

Page 13: The longterm consequences of malnutrition on children: New evidence from Guatemala John Hoddinott International Food Policy Research Institute April 2009.

INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE

Results (3): Does malnutrition affect cognition abilities in adulthood?

Modeling strategy:• Dependent variable are two measures of cognitive abilities:

• Reading/vocabulary score and Raven’s progressive matrices

• KEY right hand side variable is dummy variable equaling one if individual was not stunted at age 72 months, zero if stunted. We also control for grades of completed schooling and number of years in skilled employment

• ALL these variables are treated as endogenous using exposure to the intervention from 0-36 months, exposure from 0-36 months x atole , sex, being a twin, labor market conditions at age 18, school characteristics at age 6. Again we have excellent IV diagnostics

Not being stunted raises• Raven’s progressive matrices score by 0.7 standard deviations and• Reading z scores by approximately 0.5 standard deviations. This is equivalent, in

terms of effect size, to increasing schooling attainment by four grades

Page 13

Page 14: The longterm consequences of malnutrition on children: New evidence from Guatemala John Hoddinott International Food Policy Research Institute April 2009.

INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE

Results (4): Did the original intervention affect cognition abilities in adulthood?

Dependent variables are measures of cognitive abilities and outcomes:

• Reading/vocabulary scores, Raven’s progressive matrices and grades of completed schooling

• KEY right hand side variable is exposure to the intervention from 0-36 months x atole (an intent-to-treat model)

• We also control for exposure to the intervention from 0-36 months, village dummy variables, family background and school characteristics

We find that exposure to atole from 0-36 months increases scores on reading/vocabulary and Raven’s progressive matrices by approximately 0.25 of a standard deviation. It increases female grade attainment by one full grade, but has no impact on male grade attainment

Page 14

Page 15: The longterm consequences of malnutrition on children: New evidence from Guatemala John Hoddinott International Food Policy Research Institute April 2009.

INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE

Results (5): Did the original intervention affect wages and incomes in adulthood?

Dependent variable is log of income earned per hour worked:

• KEY right hand side variable is exposure to the intervention from 0-36 months x atole (an intent-to-treat model)

• We also control for exposure to the intervention from 0-36 months, village dummy variables, family background, school characteristics and labor market shocks

Exposure to atole before, but not after, age 3 years was associated with higher hourly wages, but only for men. For exposure to atole from 0 to 2 years, the increase was US$0·67 per hour which meant a 46% increase in average wages. There was a non-significant tendency for hours worked to be reduced and for annual

Page 15

Page 16: The longterm consequences of malnutrition on children: New evidence from Guatemala John Hoddinott International Food Policy Research Institute April 2009.

INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE

Summary

This work, some of which is preliminary, provides evidence of the instrumental value of investments that improve the nutritional status of pre-school children by demonstrating:

• A causal link between stunting and adult wages (result #1)

• That the causal pathways by which this occurs are those that run from stunting to lowered cognitive skills to lower wages (results #2 and 3)

• That interventions aimed at improving nutritional status can have persistent effects on cognitive skills and productivity (results #4 and 5)

As such, it reinforces the value of repositioning nutrition as being central to development

Page 16

Page 17: The longterm consequences of malnutrition on children: New evidence from Guatemala John Hoddinott International Food Policy Research Institute April 2009.

INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE

References

Result (1): “The impact of pre-school malnutrition on wages in adulthood”, in preparation.

Result (2): J. Behrman, J. Hoddinott, J. Maluccio, R. Martorell, 2008, “Labor Market Returns to Intellectual and Physical Human Capital in a Poor Developing Country” in preparation.

Result (3): J. Behrman, J. Hoddinott, J. Maluccio, E. Soler-Hampejsek, E. Behrman, R. Martorell, M. Ramírez-Zea, A. Stein, 2008, “What Determines Adult Cognitive Skills? Impacts of Pre-Schooling, Schooling and Post-Schooling Experiences in Guatemala” mimeo, International Food Policy Research Institute, Washington DC

 

Result (4): J. Maluccio, J. Hoddinott, J. Behrman, R. Martorell, A. Quisumbing, and A. Stein, “The Impact of nutrition during early childhood on education among Guatemalan Adults”, Economic Journal, forthcoming.  

Result (5): J. Hoddinott, J. Maluccio, J. Behrman, R. Flores and R. Martorell, 2008, “Effect of a nutrition intervention during early childhood on economic productivity in Guatemalan adults”, The Lancet, vol. 371, pp. 411-416.

Page 17