WOMEN’S ASSET OWNERSHIP AND CHILDREN’S NUTRITIONAL STATUS Evidence from Papua New Guinea Alice Louise Kassens Roanoke College
Jan 28, 2018
WOMEN’S ASSET OWNERSHIP AND CHILDREN’S NUTRITIONAL STATUS
Evidence from Papua New Guinea
Alice Louise Kassens
Roanoke College
BACKGROUND
• Immediate, underlying, and intermediate factors determine health and nutrition
• Immediate: lack of food, low utilization/poor quality of health care
• Underlying: family income, education, cultural
• Intermediate: mother education influences impact of clean water on child health (ex. higher education more awareness, more productive use of knowledge)
• Child health determinant groups: child characteristics, socioeconomic, household, environmental
BACKGROUND
• Previous studies show impact of mother’s income on child health (+)
• Few studies trace the influence of mother’s wealth due to lack of data on mother asset ownership in developing countries
• Previous evidence in Papua New Guinea:
• Adding a boy increases household spending, adding a girl does not
• Poor access to roads and public transportation hinders health care access, especially for women and children
• Lack of improvement in child health at odds with global trends
BACKGROUND
• Papua New Guinea is one of the world’s most rural countries (86% live in rural areas)
• Over 1/3 live in the Highland Region with poor infrastructure
• 80% in rugged/coastal terrain without access to roads/public transportation
• High dependency ratio compared to other Asian countries
• Communicable diseases are the leading cause of morbidity and mortality
• Malnutrition is thought to be the leading cause of child death
HIES
• 2009-2010 Papua New Guinea Household Income and Expenditure Survey
• Government of Papua New Guinea (GoPNG), World Bank, United Nations Children’s Fund (UNICEF), Australian Aid for International Development (AusAID)
• Third HIES in PNG (prior:1975-76, 1996-1997)
• Family demography, education, health, employment and consumption data
• Use to construct Consumer Price Index
• 4,079 households
• 22,698 individuals, 2,926 children < 60 months
MODEL
Child health: stunting, underweight, wasting (WHO standards)
Child characteristics: age (< 60 months old), gender
Socioeconomic: expenditure quintile, mother’s education, mother’s
assets
Household: number of children in HH, mother smokes, improved
toilet, clean water
Geographic: region, rural
EMPIRICAL PLAN
• OLS (1) and quantile (2) regression
(1) Yi = α0 + α1 Ai + α2 Xi + εi
(2) 𝑄𝑢𝑎𝑛𝑡𝜏 𝑌𝑖 = 𝛼 𝜏 0 + 𝛼 𝜏 1𝐴𝑖 + 𝛼 𝜏 2𝑋𝑖 + 𝜀 𝜏 𝑖
• Dependent variable = z-scores for OLS and quantile regression
• = 0.05, 0.10, 0.50, 0.90, 0.95
EMPIRICAL PLAN
• Asset measures (Ai):
• Individual asset dummies
• One per regression (6 separate regressions; only for OLS)
• Asset indexes from principle components
• Two factors, food production and shelter
• Both in one regression
• Standardized asset index (one with and one without house)
• Sum of z-scores for each individual asset
• One per regression