勞動參與的決定因素: 以台灣中年已婚男性為例

Abstract

近年來台灣面臨了人口高齡化的現象，有關中高齡人口的議題成為了學者與社會大眾關注的焦點，其中，自1988以來中高齡已婚男性勞動參與率至2008年為止已下降了約十個百分點，這個現象值得我們去深入了解。本篇論文主要在探討中高齡已婚男性勞動參與的決定因素，研究的資料來源為1988至2008年的人力資源及人力運用調查。其中，我選擇了55至64歲的已婚男性為對象，而總樣本數為51,730，本論文先以probit與bivariate probit模型估計每一個變數對中高齡已婚男性勞動決策的邊際影響效果，再以Oaxaca與DiNardo, Fortin, and Lemieux (DFL)分解模式，試著拆解每一個變數對整體中高齡已婚男性勞動參與率的影響性。\r\n此篇論文著重在兩個主要變數對中高齡已婚男性勞動參與的影響:妻子的勞動參與以及地區性的失業率。近年來越來越多已婚婦女投入職場，我想了解婦女勞動參與率的上升，對整體丈夫勞動參與率的影響；另外地區的失業率是表現出地區勞動市場的重要指標之一，過去的文獻提到失業嚴重的地區可能使當地勞工失業後找不到工作，或使想進入職場的勞工卻步。此篇論文研究結果顯示妻子的勞動參與會顯著的影響先生對勞動市場去留的決定，妻子影響個人的勞動參與機率6~18%左右，而1%地區性失業率的上升，則是對個人的勞動參與機率下降的影響約1.5%左右。在1988年至2008年整體中高齡已婚男性勞動參與率的分解中，勞動參與率下降了3.5%（占整體變化40%），可歸咎於地區失業率的升高。而若妻子的勞動參與沒有提升，仍維持1988年的水準，整體丈夫的勞動參與率將會下降1%（占整體變化10%）左右，本論文認為若政府能維持良好的就業市場環境，將有助於提高中高齡已婚男性人口勞動參與的比率，進一步能有效提高勞動生產力及降低社會負擔。

As the proportion of the old population increases in Taiwan, issues of older individuals’ behavior attract public attention. During 1988 to 2008, labor force participation rate of older married men declined over 10 percent. What can explain this decline? This thesis tries to find out the determinants of older married men’s labor force participation in Taiwan. I use the data from Manpower Survey and Manpower Utilization Survey from 1988 to 2008, conducted by Directorate General of Budget, Accounting and Statistics (DGBAS). The sample comprise 51,730 observations of married men aged 55-64. Older married men’s labor participation decision is treated as a dependent variable and estimates are made with a probit and a bivariate probit model. Decompositions with methodology of DiNardo, Fortin, and Lemieux (1996) and Oaxaca (1973) are conducted for explaining the decline in labor participation rate of older married men between 1988 and 2008. The results indicate that the increase in wives’ labor force participation increases husband’s likelihood of participation and prevents aggregate husbands’ participation rate from declining about 1 percentage point (-8 percent of total decline). However, regional unemployment rate negatively affects husband’s likelihood of participation and can explain at least 3.5 percent (40 percent of total decline) of the decline in husband’s participation rate. This thesis suggests the labor force participation rate could be stopped from declining if the government maintains good labor market condition.