Analysis of the labor participation and wage equation model of Japanese married women

妇女劳动参与和工资公式的经济计量模型已被广泛使用。必须考虑到这种模式估计中的样本选择偏差，因为妇女的工资只有在工作时才能观察到。传统的估计方法没有得到充分的审查，因此可能导致误导性的政策影响。本文使用家庭经济研究所的“日本消费者调查”数据，通过传统的方法，首次使用新的高效的最大似然估计方法，对日本已婚女性的劳动参与率和工资方程进行了估计，并探讨了日本的信贷紧缩对家庭福利的影响。通过引入内生信贷约束，扩充了消费增长方程的理论框架。利用1993-1999年的家庭面板数据，利用Amemiya（1985）的Type 5 Tobit模型，对扩充后的欧拉方程进行了估计。出现了若干实证研究结果。首先，我们发现，一个不可忽视的一部分人面临信贷约束。因此，我们的结果拒绝了标准的消费欧拉方程，即，生命周期永久收入假说的必要条件。其次，利用1993年的数据，在信贷约束指标可用的情况下，我们证实了信贷约束的内生性是可用的，我们证实了信贷约束的内生性对消费欧拉方程的估计产生了严重的偏差。最大似然估计结果支持我们的欧拉方程的框架与内生信贷约束。最后，对1993-1999年的全部数据的分析表明，信贷紧缩变得严重，特别是在1997年之后。

Econometric models of female labor participation and wage equations have been used extensively. Sample selection biases in the estimations of such models must be taken into account since women's wages cannot be observed unless they are working. Conventional estimation methods have not been sufficiently examined and may, therefore, result in misleading policy implications. This paper estimates labor participation and wage equations of Japanese married women through conventional methods, and for the first time, through a new, efficient, simultaneous maximum likelihood estimator that uses data from the "Japanese Panel Survey of Consumers" conducted by the Institute of Household Economy.We also investigate hoe the credit crunch in Japan affected household welfare. The theoretical framework of a consumption euler equation is augmented by adding endogenous credit constraints With household panel data for 1993-1999, we estimate the augmented Euler equation by using Amemiya (1985)'s Type 5 Tobit model. Several empirical findings emerge. First, we found that a non-negligible portion of people face credit constraints. Accordingly, our results reject the standard consumption Euler equation, i.e., the necessary condition of the life-cycle permanent income hypothesis. Second, by using data in 1993 when an indicator of credit constraint is available, we confirm that endogeneity of credit constraint is available, we confirm that endogeneity of credit constraints generates a serious bias in estimating consumption Euler equation. Maximum likelihood estimation results support our framework of Euler equation with endogenous credit constraints. Finally, analyses of the full data for 1993-1999 uncover that the credit crunch became serious especially after 1997.