| 摘要: | 此論文主要研究職業死亡風險對工人工資的影響，尤其過去國內文獻中，鮮少針對性別及家戶成員（戶長與配偶）進行篩選，並特別考量同家庭內夫妻是否選擇高風險工作，探討此情況下生命價值的估計是否更高。本研究透過樣本篩選，分別探討男性樣本、女性樣本及男女混合樣本間的工資風險補償特性。
本研究主要使用Aldy(2019)透過選型配對(Assortative mating)進行重構後之特徵工資模型(Hedonic Model)作為模型，並以台灣之資料進行統計生命價值(Value of a Statistical Life, VSL)之估計，分析96至105年間戶長和配偶的工資數據，總樣本數4,640筆，並以四種樣本組合進行估計：(1)所有工作者樣本；(2)僅戶長樣本；(3)僅配偶樣本；(4)配偶相減樣本。
根據職業致命風險係數進行生命價值估算，包含所有工作者的樣本估計為141,039,987元；戶長樣本為101,371,723元；配偶樣本為105,294,878元；使用選型配對理論的夫妻樣本為250,179,018元，與傳統的估計方式相比，使用夫妻內部差異法之模型，排除異質性問題後，所估計之統計生命價值更高，且發現男性在面對工作風險時，工資補償效果顯著，而女性則未顯示出同樣的補償效果，反映出職場性別不平等現象。;This study investigates the impact of occupational fatal risk on workers′ wages, particularly focusing on the rarely addressed aspects of gender and household member selection (household head and spouse) in previous domestic literature. It specifically considers whether spouses within the same family choose high-risk jobs and explores whether life value estimates are higher in such cases. Through sample selection, the study separately examines the wage-risk compensation characteristics among male samples, female samples, and mixed-gender samples.
The primary model used in this research is the hedonic wage model reconstructed through assortative mating, as proposed by Aldy (2019). Taiwanese data from 2007 to 2016 were used to estimate the value of life, analyzing wage data from household heads and spouses. The total sample size was 4,640, and the study conducted estimates using four sample combinations: (1) all workers, (2) only household heads, (3) only spouses, and (4) spousal difference samples.
According to the occupational fatal risk coefficients, the estimated values of life are as follows: for all workers, the estimate is NT$141,039,987; for household heads, it is NT$101,371,723 ; for spouses, it is NT$105,294,878 ; and for the sample using the assortative mating theory, it is NT$250,179,018. Compared to traditional estimation methods, the model using intra-household differences, which excludes heterogeneity issues, results in a higher estimated value of statistical life (VSL). The study also finds that wage compensation for job risks is significant for men, whereas women do not show the same compensation effect, reflecting workplace gender inequality. |
