Betting On Human Capital: Options For Consumption Smoothing When Health Insurance Is Introduced In Developing Countries

Introduction: Over the last two decades, there have been a growing number of health insurance programs being implemented by governments in emerging and developing economies. While much of this literature has focused on evaluating the impact of these programs on health access and spending, there is very little known on the impact of these programs on other spending decisions of households. The health economics literature has found robust evidence that individuals with health insurance can increase current consumption spending by reducing precautionary savings. However, in developing countries, income constrained households may not be able to reallocate gains from insurance towards greater inter-temporal consumption through the precautionary savings channel. In such cases, households can choose to allocate gains from having access to health insurance towards different types of household goods that allow them to smooth their consumption over time.

Methodology: In order to examine its impact on household budgetary allocations, we use a unique nationally representative data set that collected detailed expenditure information of households in Viet Nam. We exploit variation introduced by a government supported free health insurance program that was targeted to children under the ages of six in Viet Nam. Since the program began in 2005, we also exploit variation introduced by the survey timing (2002, 2004, 2006, 2008). In our model, we compare spending outcomes between households with children under six (beneficiary) and those with children who are slightly older (non-beneficiary) between a pre-program and post-program survey period.

Results: We find that having access to the free health insurance increases household spending on current consumption. We find evidence that beneficiary households are more likely to allocate the indirect income transfer from health insurance towards spending on human capital development. Expenditures on health and food increase relative to non-beneficiary households who were not eligible for the program between the pre and post survey period. Our aggregate model indicates that households do not change any other spending, precautionary savings or asset related spending compared to the non-beneficiary households. There are no distributional effects of the program for either of these outcomes across the spending distribution. There are two explanations for these results. One is that in the absence of properly functioning financial markets, households ay prefer to allocate spending towards goods that increase human capital. Two, it is also possible that the household views this program as a transitory transfer of income since the program ends as the child ages out of it. In both cases, investing in human capital provides households the necessary channel to increase both current and future consumption.