Duke 2008 | ASHE | Equity and Efficiency in Health and Healthcare

Objectives

This study evaluates the short tem impact of the policy on free health care for children under six endorsed by the 2004 Vietnamese Government's Law on Protection, Care, and Education of Children. Since this is a major financial undertaking, it is important to assess if the Law achieves its goals.

Income and Substitution Effects

The Law provides children under six with free services from government medical establishments. Therefore, the income effect is expected to be positive for every type of health service while the substitution effect is positive only for public services. Given that the substitution effect usually dominates the income effect, eligible children are likely to switch from purchasing drugs for self-medication and private services to public services. The effect on overall out-of-pocket expenditure on health is likely to be positive, particularly among poor children who previously had financial hardships in seeking medical treatment even at a public facility.

Data and Variables

Data come from two national surveys conducted in 2002 (the Vietnam National Health Survey) and 2006 (the Vietnam Household Living Standard Survey). The former covers 36,000 households and focuses exclusively on health, and the latter has an expanded health module applied to 9,189 households. Both surveys collected detailed information on households' health service utilization, financing, and the communes' socio-economic and health service situations.

The outcomes of interest include: (1) 4-week outpatient contacts at public or private facilities or 4-week purchases of drugs for self-medication; (2) 12-month inpatient days; and (3) 12-month out-of-pocket expenditures on health. Since the outcomes are count events characterized by a high frequency of zeros, a two-part model is used in the econometric analysis. Major control variables include individual, household, and commune characteristics, as well as the availability of public and private health services at the primary level.

Empirical Specification

Policy impact can be estimated as a difference-in-difference in the outcomes of interest by the eligible group (the treated) and a comparison group (the control) before and after the Law's effect date (2002 and 2006 respectively). This paper uses children ages 6 and 7 as the control group under the assumption that any time-varying trends will affect the control and treated groups equally. The analysis also controls for age trend in service utilization.

For an individual i at age a in group g (treated or control) and year t (2002, 2006), the outcomes of interest Y_iagt can be expressed as:

Y_iagt = a + b1Treat_g +? b2Year2006_t + b3(Treat_gYear2006_t) + b4Age_a + b5(Age_aYear2006_t) + b6Z_iagt? + e_iagt

where Z_iagt is a vector of control variables and e_iagt is the error terms. The parameter of interest is b3. Further analysis will include a dummy variable on whether the child is from a poor family to assess potentially differential effects of the policy on the poor versus non-poor children.

Equity and Efficiency Implications

Equity implications of the Law are reflected in its potentially disproportional benefits to poor children (vice versa). Efficiency is improved if, as a result of free health care, children reduce self-medication and use the appropriate level of medical treatment.