The effects of children on households’ fixed and transitory wealth: Insights from a longitudinal study of periurban communities in Ghana and Ethiopia (Amy Tsui, Qingfeng Li, Julia Driessen, Assefa Sema and Emmanuel Naku)

Asset indices have been used to gauge household wealth and occupants’ welfare. Their construction circumvents a number of issues plaguing direct measures of income and wealth, such as expenditure data. The standard household asset index used in multi-national surveys comprised of items related to ownership of property, housing construction quality, and amenities. Principal components analysis has been applied to derive household scores.

Asset goods stand in contrast to consumption goods owned or purchased by the household, the latter often relevant for a short period of time. Because of the transitory presence of these household goods, they are not considered reflective of a household’s true long-term economic status, and instead may represent temporary fluctuations in a household’s economic resources. The resources required to support young children, e.g., school fees, health care, and food, can reduce or increase household transitory wealth in the short term. Little research evidence on this relationship is available for households in low and/or middle income countries, particularly based on longitudinal studies. This study aims to better understand the relationship between the number and ages of children under 15 years and transitory and permanent wealth. It also looks at whether recent medical expenditures (health shocks) and recent births (fertility shocks) lower transitory and permanent wealth.

The data come from two waves (rounds) of the Family Health and Wealth Study (FHWS), conducted in 2009/2010 and 2011/2012. The objective of this study is to investigate individual- and family-level health and economic consequences of childbearing patterns. Two (Ethiopia and Ghana) of the six periurban study sites have acceptable follow-up and completion rates to warrant inclusion in this panel analysis. Criteria for household eligibility include having a resident married couple with the wife of childbearing age.

Household data on ownership of home, basic furnishings, and consumer assets, housing construction materials and quality, available water and electricity, and cooking fuel comprise the fixed wealth index (FWI), while ownership of non-essential consumer items (e.g., microwave, washing machine, electronics) and expenditures (on eating out, clothing, books, newspapers, child care) as well as household financial health (spending on taxes, having no debt, making loans, and having savings) comprise the Transitory Wealth Index (TWI). The latter index is aimed at capturing expenditures that rely on disposable household income and thus are vulnerable to reallocation by significant life events, such as births and health care needs. Principal Components Analysis was used to construct each score which was then normalized.

Guided by micro-economic household theories, we construct a panel model to assess the effects of Round 1 predictors on Round 2 wealth measures. Following exploratory data analysis, we estimate two models for each of the two measures using OLS for each of the two sites. The Round 2 values for TWI and FWI are modeled separately and for each of these models, one model adds two covariates--recent births (between Round 1 and 2) and its statistical interaction with the wife’s parity (4 or more births). The covariates of interest, largely measured at Round 1, are of three types: Round 1 Fixed Household Wealth, fertility and health needs, and characteristics of the partners. Fertility is measured through the presence of children in the household under age 5 and at age 5–14, number of births since Round 1, and wife’s high parity status. A second household economic “shock” measure is the percent of total household income spent on health care in the past. The partner covariates include the husband’s and wife’s years of schooling and her age.

Amy Tsui is a Professor and Qingfeng Li is a Research Associate at the Johns Hopkins Bloomberg School of Public Health in the United States.

Tsui et al Presentation








Copyright (c) 2004-2017