全民健保對生育行為的影響

Abstract

本文藉由台灣於 1995 年 3 月 1 日實施全民健保所形成的自然實驗， 運用 1994 -- 1995 、 1997 -- 1998 共 4 年的家庭收支調查資料， 以受到全民健保影響的私部門家庭為實驗組， 並以較不受政策影響的公部門家庭為對照組， 使用 differences-in-differences 法估計全民健保對於生育行為的影響。 實證結果顯示，全民健保使得家庭有子女出生的機率較以往成長了約 23.12\%，政策效果為 0.0243 。 在母親是勞保參加者的家庭中，家庭有子女出生的機率更是成長了約 34.67\%，政策效果為 0.0276 。 但在母親不是勞保參加者的家庭中，全民健保對於其生育行為並沒有影響。 雖然本文的實證結果顯示了醫療保險會提高家庭的生育意願， 為了避免因實驗組與對照組的選樣偏誤所導致的估計誤差， 本文也檢視了近 20 年來台灣公、私部門家庭的生育行為， 發現被解釋變數在公、私部門家庭間並不具有相同的時間趨勢， 這也意謂著本文的實證結果可能無法排除 differences-in-differences 法估計上的誤差，因而得到了偏誤的政策效果。 總而言之，雖然本文的實證結果暗示了全民健保使得家庭有子女出生的機率顯著地上升， 醫療保險與生育行為之間的因果關係仍有待後續研究更深入嚴謹的探討。

Taiwan implemented National Health Insurance (NHI) in March 1995. By reducing health expenditures, comprehensive health insurance can reduce household's costs of rearing children, thereby increase household's demand for children. This paper uses data from the 1994--1995 and 1997--1998 Survey of Family Income and Expenditure and conducts differences-in-differences estimation to investigate the impact of NHI on fertility. The results show that household's fertility increase by about 23.12\%, and the effect is about 0.0243. In households where mom joins Labor Insurance, fertility increases by about 34.67\%, and the effect is about 0.0276. However, in households where mom does not join Labor Insurance, NHI does not have any impacts on fertility. To avoid possible estimation biases, we investigate both treatment and control group's fertility behavior from 1987 to 2007, and find out that both groups do not have common trend of the dependent varialbe used in this paper. Therefore, though our analyses show that NHI increases fertility, the results may suffer from some serious estimation biases. The causal relationship between health insurance and fertility behavior should still be investigated carefully in the future.