Young Lives Early Nutrition and Cognition in Peru: A Within-Sibling Investigation Washington, DC, October 2009.

Slides:



Advertisements
Similar presentations
Child protection risks, child poverty and social transfers Paul Dornan March 2013.
Advertisements

REGRESSION, IV, MATCHING Treatment effect Boualem RABTA Center for World Food Studies (SOW-VU) Vrije Universiteit - Amsterdam.
Association Between Parental Resources and Child Development in Peru
Methods of Economic Investigation Lecture 2
The choice between fixed and random effects models: some considerations for educational research Claire Crawford with Paul Clarke, Fiona Steele & Anna.
Income and Child Development Lawrence Berger, University of Wisconsin Christina Paxson, Princeton University Jane Waldfogel, Columbia Univerity.
The First Twelve Years: Growing-Up in Low and Middle- Income Countries November 2014 Paul Dornan.
Is the Playing Field Leveling in Peru? The Evolution of Children’s Opportunities Javier Escobal, GRADE & Young Lives –Peru (and LCSPP/PREMPR – World Bank)
2.F Developing a Contextual Assessment for OOS 15 year olds.
Family-level clustering of childhood mortality risk in Kenya
 Social & Environmental Variables The effects of SES and Parenting on Cognitive Development.
Risk of Low Birth Weight Associated with Family Poverty in Korea Bong Joo Lee Se Hee Lim Department of Social Welfare, Seoul National University. A Paper.
1 James P. Smith Childhood Health and the Effects on Adult SES Outcomes.
Children’s Multidimensional Health and Medium-Run Cognitive Skills in Low- and Middle-Income Countries Elisabetta Aurino Imperial College, London Francesco.
1 Research Method Lecture 11-1 (Ch15) Instrumental Variables Estimation and Two Stage Least Square ©
Does shame and stigma undermine children’s learning? Evidence from four low- and middle- income countries July 2015 Paul Dornan and Maria Jose Ogando Portela,
Poverty: Facts, Causes and Consequences Hilary Hoynes University of California, Davis California Symposium on Poverty October 2009.
Unpaid Care and Labor Supply of Middle-aged Men and Women in Urban China Lan Liu Institute of Population Research, Peking University Xiaoyuan Dong Department.
Beyond test scores: the role of primary schools in improving multiple child outcomes Claire Crawford and Anna Vignoles Institute of Education, University.
Matthew S. Rutledge, Mashfiqur R. Khan, and April Yanyuan Wu Center for Retirement Research at Boston College and Mathematica Policy Research 16th Annual.
Holistic early childhood development indicators
THE EFFECT OF INCOME SHOCKS ON CHILD LABOR AND CCTs AS AN INSURANCE MECHANISM FOR SCHOOLING Monica Ospina Universidad EAFIT, Medellin Colombia.
Impact Evaluation of Health Insurance for Children: Evidence from Vietnam Proposal Presentation PEP-AusAid Policy Impact Evaluation Research Initiative.
Early Life Events and Health and Labor Market Outcomes in Adulthood Rucker C. Johnson Robert F. Schoeni University of California, Berkeley University of.
Monitoring and Evaluating National Responses for Children Orphaned and made Vulnerable by HIV/AIDS Mary Mahy UNICEF Meeting on Results of Pilot Surveys.
Cognitive development among young children in Cambodia: Implications for ECED programs.
Food and Nutrition Surveillance and Response in Emergencies Session 14 Data Presentation, Dissemination and Use.
The longterm consequences of malnutrition on children: New evidence from Guatemala John Hoddinott International Food Policy Research Institute April 2009.
Seasonal Migration and Early Childhood Development in Nicaragua Karen Macours (SAIS-Johns Hopkins) and Renos Vakis (World Bank)
DO MINIMUM WAGES IMPROVE EARLY LIFE HEALTH? EVIDENCE FROM DEVELOPING COUNTRIES Farhan Majid Arijit Nandi, José Mendoza, John Frank and Sam Harper.
Predictors - shock/events and response - gender - long term illness - vaccination status - breastfeeding - health service utilisation - social networks.
Social Capital and Early Childhood Development Evidence from Rural India Wendy Janssens Washington, 20 May 2004.
Dean R. Lillard 1,3, Richard V. Burkhauser 2,3,4, Markus H. Hahn 4 and Roger Wilkins 4 1 Ohio State University, 2 Cornell University, 3 DIW-Berlin, 4 Melbourne.
Measuring Equality of Opportunity in Latin America: a new agenda Washington DC January, 2009 Jaime Saavedra Poverty Reduction and Gender Group Latin America.
Racial/Ethnic Disparities in Adults Reading to Two Year Old Children: A Population-based Study Olivia Sappenfield Emory University School of Public Health.
Has Public Health Insurance for Older Children Reduced Disparities in Access to Care and Health Outcomes? Janet Currie, Sandra Decker, and Wanchuan Lin.
WHAT IS YOUNG LIVES? Young Lives is an international research project that is recording changes in child poverty over 15 years and the factors affecting.
Parents’ basic skills and children’s test scores Augustin De Coulon, Elena Meschi and Anna Vignoles.
Presentation to Education Portfolio Committee, 29 May 2001 White Paper Early Childhood Development.
Rwanda: The impact of conflict on fertility Kati Schindler & Tilman Brück Gender and Conflict Research Workshop 10/06/2010.
HAOMING LIU JINLI ZENG KENAN ERTUNC GENETIC ABILITY AND INTERGENERATIONAL EARNINGS MOBILITY 1.
Children’s Multidimensional Health and Medium-Run Cognitive Skills in Low- and Middle-Income Countries Elisabetta Aurino Young Lives, University of Oxford.
The Choice Between Fixed and Random Effects Models: Some Considerations For Educational Research Clarke, Crawford, Steele and Vignoles and funding from.
Reproducing Inequality: Family Background and Schooling in Peru Santiago Cueto, Alejandra Miranda, Juan León, and María Cristina Vásquez GRADE - Young.
Accounting for the Effect of Health on Economic Growth David N. Weil Proponent/Presenter Section.
Who supports whom? Co-residence between young adults and their parents Maria IacovouMaria Davia Funded by JRF as part of the Poverty among Youth: International.
Ifo Institute for Economic Research at the University of Munich Employment Effects of Innovation at the Firm Level Stefan Lachenmaier *, Horst Rottmann.
TRADE LIBERALIZATION AND CHILDREN Understanding and coping with children vulnerabilities Javier Escobal Group for the Analysis of Development.
Cally Ardington, Nicola Branson, Murray Leibbrandt, University of Cape Town David Lam, Vimal Ranchhod University of Michigan January, 2009 Assessing the.
The progress of the Child Well-being Module 22 nd of June, 2010 Network on Early Childhood Education and Care Dominic Richardson, SPD / DELSA.
Measuring the Impact of Young Adult Mortality on the Wellbeing of Older Persons in KwaZulu-Natal, South Africa Marjorie Opuni-Akuamoa Advisor: Dr David.
The Effect of Health on Consumption Decisions in Later Life Eleni Karagiannaki Centre for Analysis of Social Exclusion, LSE Presentation prepared for the.
Comments to “The effects of Ecuador’s economic crisis on child health and cognitive development” by Melissa Hidrobo Eduardo Rodríguez-Oreggia EGAP ITESM.
Children’s Emotional and Behavioral Problems and Their Parents’ Labor Supply Patrick Richard, Ph.D., M.A. Nicholas C. Petris Center on Health Markets and.
We take a multi-period model of childhood investment, based on Cuhna, Heckman et al (2005), which distinguishes early from late investments. In particular,
Food and Nutrition Policy Program Using Non-Income Measures of Well-Being for Policy Evaluation Prepared for the Second Meeting of the Social Policy Monitoring.
GISSELE GAJATE-GARRIDO IFPRI APRIL 2011 Excluding the poor: the impact of public expenditure on child malnutrition in Peru.
Birth Spacing and Sibling Outcomes Kasey S. Buckles, University of Notre Dame Elizabeth L. Munnich, University of Notre Dame.
Johan Mouton© February 2006 C Hart Exploratory questions What are the most important variable that have an effect on learner achievement? What happens.
Do Remittances Improve Food Consumption of Migrant’s Household? Evidence from Nigeria Babatunde Raphael Olanrewaju Department of Agricultural Economics.
BY SANDRA BLACK PAUL DEVEREUX KJELL SALVANES QUARTERLY JOURNAL OF ECONOMICS, 2005 The More the Merrier? The Effect of Family Size and Birth Order on Children’s.
An Empirical Investigation of the Quantity-Quality Model in Mexico
Instrumental Variables and Two Stage Least Squares
Instrumental Variables and Two Stage Least Squares
Swedish Institute for Social Research (SOFI)
Instrumental Variables and Two Stage Least Squares
Transitory Shocks, Permanent Effects: Impact of the Economic Crisis on the Well-Being of Households in Latin America and the Caribbean Almudena Fernandez.
Instrumental Variables Estimation and Two Stage Least Squares
Additional analysis of Nutritional Indicators and Children’s Development Santiago Cueto GRADE Lima, Peru.
Presentation transcript:

Young Lives Early Nutrition and Cognition in Peru: A Within-Sibling Investigation Washington, DC, October 2009

Overview of proposal  Investigate outcomes of 4-5 year olds by comparing with younger siblings at a similar age ~3yrs later  Nutrition (height-for-age) and Cognitive Development (PPVT-TVIP)  First strategy: reduced form for both: Role of SES (lagged), changes in community & hh shocks (contemporaneous)  Second: structural, nutrition CD

Young Lives Team  GRADE (Lima) – Santiago Cueto, Javier Escobal  IIN (Lima) – Mary Penny & fieldwork team  University of Oxford (UK) – Stefan Dercon, Ingo Outes-Leon, Catherine Porter, Alan Sanchez

Young Lives Data- overview  12,000 children in 4 countries  Long-term cohort study core-funded by UK-DFID since 2002 (and until 2017)  2000 children in each country born 2001 (Ninos del Milenio)  1000 aged 7-8 years older than them  Random sample in Peru

YL index children  Born , surveyed in 2002, 2006 and currently in field for 3 rd Round  Will be followed 2 more rounds  In R1 and R2 we have information on  Assets, Consumption, Economic shocks  Caregiver characteristics  Time use and schooling of adults/children  Anthropometrics and cognitive development of Index Child  Breastfeeding & Early health

Siblings Data  In R3 IADB funding allows us to collect anthropometric data and cognitive development data (PPVT) of next-sibling-down (same mother, usually same father)  Anthropometrics from age 3  Only do PPVT if sibling>3yrs  Making big effort to find sibling (so far around a third of sample)  NB: IADB Funding has stimulated funding for other 3 countries to study siblings

Dates and ages of children in survey rounds RoundOneTwoThree Year Age of index child6-18mths4-5 yrs7-8 years Age of sibling-0-3 years3-6 years

ECD proxy variables  PPVT- TVIP (Spanish language version) has been used in R2- continue in R3  Used by other studies in LAC (Paxson & Schady)  Designed for spanish speakers  Height-for-age z-scores proxy long-term nutrition

Descriptive statistics *Paired-sibling sample based on households with younger siblings aged between 1 and 4 years in Round 2. Some of the children aged 1 year might not be eligible.

Situating our paper in literature  Schady (2006) review notes paucity of info on ECD outcomes in LAC (esp causal links)  Inspired by Paxson/Schady (2007) but we can go further given that we have sibling data  Causal links literature- Glewwe et.al (2001), Alderman et al (2006)

Part One: Reduced form  What is the effect of socioeconomic status, household and community characteristics on child nutritional and cognitive development?  Policy relevant- especially in context of crisis, food price hikes  Not attempting structural model, but exploring correlates, whilst controlling for household fixed effects.

Empirical issues  Todd and Wolpin (03,07) review empirical estimation strategies – and their implicit assumptions  Cross-section estimation will be biased if we can’t control for unobservables  We have lagged data on SES/caregiver/community characteristics  We have data on siblings in R3  Sibling outcomes R3 - Index child outcomes R2  On SES/HH chars R2 – SES/HH chars R1

Econometric specification  Cognitive achievement of child ‘k’ from household ‘h’: where: : cognitive achievement (PPVT std. score) or : height-for-age : child and household and community observable characteristics : child unobservable characteristics : household unobservable characteristics : random error, iid (1)

Sibling differences (reduced form) ECD outcome of child k from household h where siblings k=i, j Includes the iid error and innate ability of child k. represent time-varying household investments, where t is specified at the time when child k is aged 6-18 months

Estimation issues  We are therefore assuming:  Omitted inputs are not correlated with the error  Inputs associated with each child do not respond to own or sibling’s endowment  In practice we have only household level investments for the siblings, hence we specify inputs at time t-1 (critical period investments)  Shocks in time t can proxy for investments age 3-5

Explanatory variables  Time varying characteristics of the household  Change in poverty status (income, consumption and assets)  Water and sanitation, adult nutrition, food security  Parenting attitudes not collected for the sibling (e.g. we do not have information on attitudes or practice of breastfeeding for the sibling, nor their immunizations)  But we will have information on the psycho-social status of the mother (including maternal depression, social capital, self- efficacy and self-esteem).  Spending on various goods that are inputs to child development such as food or healthcare may also have a significant impact on ECD outcomes.  Time varying inputs at the community level include prices and the availability of certain services (pre-school, school, sanitation, social protection programs eg Juntos).

Second part  Linking early childhood nutrition to cognitive development

Differences in R2 PPVT scores between R1 stunted and non-stunted children

Differences in Round 2 PPVT scores between stunted and non-stunted children

Linking Nutrition to Cognitive Development Extensive literature - linkages between nutritional deficiencies at an early stage of child development and reduced cognitive ability, educational attainment and ultimately lower market wages at a later age. (Behrman and Lavy 1994), both a child’s health and her cognitive achievement can be understood as the outcomes of a utility-maximization process whereby parents choose to invest in a child’s human capital subject to initial conditions – genetic innate abilities –, parental taste for child’s quality and budget constraints. Parental taste for child quality and a child’s genetic ability are unobserved, OLS estimations of nutrition link to cognitive development likely to be biased.

Solving endogeneity problem  Experiments (e.g. Guatamala, Pollit et al(93) Maluccio et al (09))  Sibling differences + IV  Glewwe et al (01) birthweight  Alderman et al (06) drought shock  We follow Alderman et al 2006 and Glewwe et al 2001 in combining household fixed effects and instrumental variable techniques to deal with the endogeneity of nutrition.  We also propose to include observable differences between siblings as well as time-variant household and community characteristics as controls in the estimation.

Econometric specification  Link nutrition and cognitive achievement of child ‘k’ from household ‘h’: where: : cognitive achievement (PPVT std. score) : height-for-age : child and household observable characteristics : child unobservable characteristics : household unobservable characteristics : random error, iid (1)

Empirical strategy: part I  Using specification (1) and taking the siblings-difference between children ‘i’ and ‘j’ from hh. ‘h’’: (2)  Specification (2)-(2a) is useful because all household unobservable characteristics that are common across siblings are removed.  It also controls for observable differences across siblings that might lead to differential investments within the household (differences in age, gender, relative birth order included in vector ΔX).  Due to data constraints, in practice we estimate: (2a)

Empirical strategy: part II (C1) Parental nutritional investments could be adjusted as a child’s innate cognitive abilities are revealed. (C2) The health status and the cognitive ability of a child can be correlated through a common unobserved genetic endowment.  In specification (2a), endogeneity of child’s nutrition still remains a problem due to child-specific unobservable characteristics:  To deal with these possibilities, we further instrument the within-siblings nutrition.

Empirical strategy: part II 1.Within-siblings birth weight  Deals with (C1) but not with (C2). 2.Child-specific shocks  Deals with (C1) and (C2).  The shocks affected only the index children, at 0-2 years. Younger siblings were not yet born.  The shocks affected the younger siblings when they were 0-3 years.  Preliminary analysis identifies the following shocks as affecting a substantial number of children in the paired-siblings sample: in , food shortage events; in , job loss of the head of the hh and severe illness of family members.  Two types of instrumental variables available:

Other empirical issues: time-varying characteristics  While not explicitly included in specification (1)-(2a), differences in cognitive and nutrition outcomes between siblings could also be driven by time-varying household and community characteristics that benefit the development of one sibling over the other. To take this into account we add as controls: 1.Changes in household consumption levels (excluding food) between Rounds –to control for life-cycle patterns-. 2.Changes in community characteristics –as reported in community questionnaires from Round 2 and 3-. This includes changes in local prices, community infrastructure, etc.

Other empirical issues: preschool enrolment  Final concern:  whether differences in cognitive results might be driven by differences in age of preschool enrolment  itself driven by nutritional differences between siblings. 1.Auxiliary regressions: In the full sample, we can make estimations of the determinants of age of preschool enrolment. Testing the nutrition effect in this estimation will indicate the problem we are faced with. Preliminary results show that nutrition does not predict preschool enrolment. However, this is not entirely the endogeneity we need to address. To deal with this, a number of strategies are at hand:

2. Sub-samples: (a) For those siblings aged between 5 and 7 years, age of pre- school enrolment can be directly controlled for. This sub-sample has an estimated size of paired-siblings. (b) In sites with no preschool facilities, small or zero bias resulting from not including age of preschool enrolment. Other empirical issues: preschool enrolment

3. Sources of exogenous variation in preschool enrolment The minimum age of preschool enrolment in Peru is 3 years, whereas enrolment is compulsory at 5 years. Using month of birth as a source of exogenous variation in age of preschool enrolment could partially capture differences in cognitive achievement driven by differences in preschool enrolment. Other empirical issues: preschool enrolment

Summary  Looking at socio-economic determinants of ECD outcomes in Peru  Plus more structural look at nutrition- cognitive development nexus  Currently in the field  PDA for 50% of households  PPVT on paper for all