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

On March 1, 1995 Taiwan implemented its National Health Insurance (NHI) program which provides coverage to all Taiwanese citizens. Just before NHI, 40% of its population was still uninsured. The remaining 60% was previously insured by various employment-based social insurance programs and then shifted automatically to be covered by NHI after 1995. The elderly were one of the major components of the previously uninsured population. In addition, official statistics have shown that the elderly are the highest users of health care among all age groups in Taiwan. For example, the elderly aged 60 and over amounted to only 13% of the population but were responsible for about 50% of total expenditure on personal health care in 2005.

This paper aims to evaluate the impact of Taiwan's NHI on health care utilization and health outcomes of the previously uninsured elderly. I use a panel data set, Survey of Health and Living Status of the Elderly in Taiwan, which spans periods before and after NHI. This data set contains rich information regarding demographic, social, economic and health aspects of the elderly. I adopt a difference-in-differences (DID) approach to assess the NHI impact on utilization. Due to many similarities between former social insurance programs and NHI, those who were socially insured in the pre-NHI era are used as the control group. Major findings include the following: compared to the continuously insured elderly, the previously uninsured elderly have a 13 percentage-point increase in the probability of monthly western outpatient visits, a 9 percentage-point increase in yearly hospital admissions, a 4 percentage-point increase in monthly Chinese outpatient visits and a 4 percentage-point decrease in the probability of foregoing physician visits when ill. I also find increases in their average number of monthly western outpatient visits (41%), yearly hospital admissions (85%) and monthly Chinese outpatient visits (111%). However, no significant increase in either the probability or the average number of emergency room visits is found. These results are robust to several tests of the identification strategy.

Given these findings, I examine further how NHI affects health outcomes of the previously uninsured elderly with a focus on mortality. I use a variety of parametric survival models blended with a DID design. Preliminary results show that the growth rate of mortality of the previously uninsured elderly slowed dramatically and converged to almost the same rate as the continuously insured elderly in the post-NHI period. Overall, these findings suggest that Taiwan's NHI has succeeded in fulfilling its goal of improving access to health care, at least among the elderly, resulting in important improvements in health outcomes.