本文以1997至2015年我國55個中分類項產業的追蹤資料，分析2001年第一次法定工時縮短政策以及非薪資成本逐年增加，對於受僱人數、平均工時及加班工時的影響。實證分析方法採Hausman方法檢定雇用人數與平均工時模型之設定，並以Tobit模型探討加班時數之決定。本文首先比較勞動需求者之調查資料「薪資及生產力調查」與由勞動供給者之調查資料「人力資源調查」，結果發現在統計上「薪資及生產力調查」的工時顯著較低。實證計量結果發現，不論是使用上述哪一種資料庫，2001年法定工時縮短政策與長期非薪資成本增加，使得產業雇用量與平均總工時減少，而加班工時增加。其中，「人力資源調查」工時的邊際影響效果相對較大。將樣本進一步區分為製造業四大工業，以及按廠商平均規模、受僱員工教育程度，與自動化程度高低分類，發現與整體產業分析結果不同的是，工時政策使電子資訊工業的雇用量增加約8%。另外，低學歷密集產業受負面影響程度最大，減少約21%雇用量。相對其他產業，非薪資成本增加對民生工業及小企業較多的產業雇用量有不利的影響。 整體而言，工時縮短衝擊我國產業雇用量，廠商傾向以增加已受雇勞工的工時代替減少的雇用量，造成法定工時縮短對勞工平均工時減少有限的結果，此或亦可用以解釋截至目前為止我國勞工的工時仍偏長的原因。此外，本研究發現非薪資成本增加對工時與雇用量的影響則因產業而異，有正向也有負向。 ;The research focuses on the impact of the reduced standard-hours policy in 2001 and the increasing non-wage labor costs on labor demand in Taiwan, including employment, average working hours and overtime hours. In the past, researches mainly utilize the monthly working hours from the Earnings and Productivity Statistics (EPS) to analyze the impact of standard- hours policy. The EPS is a dataset from firm’s survey. We find that working hours between EPS and Manpower Utilization Survey (MUS), provided by households and individuals, are statistically different, with EPS hours being underestimated due to the incentives faced by firms. This study uses 2-digits-industrial panel data from both EPS and MUS from 1997 to 2015 for empirical analysis. We apply panel regression with Hausman test on employment and average working hour regression. Overtime hour regression is analyzed by Tobit Model. Empirically, we find that the reduction of standard hours in 2001 has significantly reduced employment and monthly average working hours, and increased overtime hours. The marginal effect of standard-hours policy on working hours depends on data sources, where MUS’s effect is significantly larger than that of EPS. Moreover, the impact of standard-hours policy vary with industrial features, including firm size, workforce educational level, and firm’s automatic level. The effect of non-wage labor costs on employment and working hours also depends on industrial characteristics. We thus conclude that the reduction of standard hours shocks the overall employment. Taiwanese firms are more likely to increase workers’ overtime hours than to hire more workers in responding to the policy of standard-hours reduction. This might explain why working hours in Taiwan are rather long compared to its counterparts in the world.
