本研究使用2006年至2016年行政院主計處五月份的人力運用調查資料，利用Heckman兩階段模型考慮勞動參與選擇性後，分別估計整體、男性與女性之工資函數，再使用Oaxaca-Blinder工資差異拆解模型，將工資差異拆解為勞工特性所能解釋的部分與特性無法解釋之部分，以了解影響勞動市場性別薪資距之原因。實證結果顯示，女性在選擇是否進入勞動市場時存在有較嚴重的自我選擇問題。工資函數部分，在考慮其他控制變數後，女性薪資相較於男性低了10.3%至14.4%，男性與女性之各個控制變數對於工資之報酬率存在差異，因而使性別薪資存在差異。工資拆解模型的估計結果顯示，性別薪資差距中，能被一般文獻所使用的變數所解釋之比例逐年降低，但本研究加入所得分配之百分位群組變數，發現整體模型可以解釋之性別薪資差距的比例大幅提升，表示在各所得分配群組中性別比例分配不平均的現象是造成性別薪資存在差異的重要因素。;This study used the 2006-2016 Manpower Utilization Survey which are acquired from the Directorate-General of Budget, Accounting and Statistics, Executive Yuan in Taiwan. We estimated the wage function of the all, the male and the female separately after considered the labor participation selection problem by Heckman two-step selection model. Then, we utilized Oaxaca-Blinder decomposition model to decompose the gender wage gap into the part that can be explained by the labor characteristic and the part that cannot be explained by the labor characteristic. And analysis the reason that impact the gender wage gap in the labor market. The empirical results indicate that female have more serious self-selection problem when decide whether enter the labor market. In the part of the wage function, the wage of the female is about 10.3% to 14.4% lower than the wage of the male. The return of every explained variable is different in gender, which cause the gender wage gap. From the estimated results of the decomposition model, the ratio of the gender wage gap that can be explained by the explained variables which are used in most literatures is decreasing year by year. But this study added the centile group variables of the wage distribution and found that the explained part of the gender wage gap raised substantially. That is, the uneven gender ratio in every centile groups is the important fact that bring out the gender wage gap.
