Taiwan's National Health Insurance and the labor force participation decisions of married women

This dissertation analyzes, theoretically and empirically, how the introduction of the universal National Health Insurance (NHI) program in Taiwan in 1995 impacted the labor force participation (LFP) decisions of married women. The core identification strategy comes from exploiting the variation in health insurance eligibility between wives of government and non-government employees before and after the introduction of NHI. The first essay investigates the impact of NHI on the LFP decisions by married women of different ages. Using the Survey of Family Income and Expenditure (SFIE) data, I apply the difference-in-difference (DID) approach to evaluate heterogeneous LFP responses. I find sharp differences in the impacts of NHI for married women in different age ranges---with larger negative impacts for older women. For example, women aged 25-44 showed no change in employment while women aged 45-54 reduced employment by between 17 and 28 percent.;The second essay investigates how the impact of the introduction of NHI on the LFP decisions of married women aged 25-44 varied with family income levels. I find that married women in the 25th-50th percentiles is the only subgroup significantly responsive to NHI. The introduction of NHI reduced LFP of married women in the 25th-50th percentiles by between 33 and 40 percent. This labor participation disincentive response resulting from the NHI, however, shows "fade-out" effects. In addition, married women in several different family structures (the presence of children under 3 or less healthy elderly household members) did not respond differently to the introduction of NHI.;In the third essay, I examine LFP response to NHI among farm wives. I employ a difference-in-difference-in-difference (DIDID) research design to take group-specific underlying trends in overall LFP into account and estimate that the introduction of NHI reduced the overall LFP (including on- and off-farm LFP) of farm wives by 10 percent. Using a DID approach in a regression framework, I also test several alternative scenarios and find that NHI provided a great disincentive for farm wives to work off farm by 23-27 percent. This reduction is the largest for wives in small farm households.