论文标题
确定母性就业中的潜在结构：关于德国父母福利改革的证据
Identifying Latent Structures in Maternal Employment: Evidence on the German Parental Benefit Reform
论文作者
论文摘要
本文通过采用C-Lasso（一种最近开发的，纯粹的数据驱动的分类方法）来确定潜在的群体结构对母性对就业的影响。此外，我通过利用识别策略将尖锐的回归不连续性设计和预测就业概率的假设检验结合了尖锐的回归策略，如何评估2007年德国父母福利改革的引入如何影响不同的集群群体。 C-LASSO方法可以使母亲之间的异质就业影响分类为先验不明数的集群组，每个组都有其自身特定于组的效应。 C-Lasso使用新颖的德国行政数据确定了三个不同的聚类组前后改革。我的发现表明，孕产妇就业中明显未观察到的异质性，并且改革对所识别的集群群体的就业方式有所不同。
This paper identifies latent group structures in the effect of motherhood on employment by employing the C-Lasso, a recently developed, purely data-driven classification method. Moreover, I assess how the introduction of the generous German parental benefit reform in 2007 affects the different cluster groups by taking advantage of an identification strategy that combines the sharp regression discontinuity design and hypothesis testing of predicted employment probabilities. The C-Lasso approach enables heterogeneous employment effects across mothers, which are classified into an a priori unknown number of cluster groups, each with its own group-specific effect. Using novel German administrative data, the C-Lasso identifies three different cluster groups pre- and post-reform. My findings reveal marked unobserved heterogeneity in maternal employment and that the reform affects the identified cluster groups' employment patterns differently.
