Semi-parametric estimation of disequilibrium models
提出一种易于计算的半参数方法估计Fair和Jaffee的非均衡模型，无需完全指定误差项分布，蒙特卡洛研究表明在非对称误差下优于极大似然估计，并用美国劳动力市场数据演示。
Abstract This paper presents an easy-to-compute semi-parametric (SP) method to estimate a simple disequilibrium model proposed by Fair and Jaffee (1972). The proposed approach is based on a non-parametric interpretation of the EM (Expectation and Maximization) principle (Dempster et al; 1977) and the least squares method. The simple disequilibrium model includes the demand equation, the supply equation, and the condition that only the minimum of quantity demanded and quantity supplied is observed. The method used here allows one to consistently estimate the disequilibrium model without fully specifying the distribution of error terms in both demand and supply equations. Our Monte Carlo study suggests that the proposedestimator is better than the normal maximum likelihood estimator under asymmetric error distributions. and comparable to the nlaximunl likelihood estimator under synirnetric error distributions in finite samples. Aggregate U.S. labor market data from Quandt and Rosen (1988) is used to illustrate the procedure. Keywords: Disequilibrium modelsEM algorithmKaplan-Meier estimatorLeast squares methodSemi-parametric estimation