Not available in German. This paper investigates the properties of Box-Cox specifications for the labor supply of married women in France, estimated in four stages involving two dichotomous models, a regression with selectivity correction, and the calculation of sample means. Previous estimates implied wage elasticities high enough to be seriously questioned. We re-estimate the model with a more flexible stochastic specification and allow for some variation in the preference parameters across individuals. Furthermore, we account for a certain kind of heteroscedasticity. We submit all steps of the estimation to the specification tests using among others LM tests against heteroscedasticity and misspecification of the distribution, information matrix tests, Andrews' x²-tests, Vuong's LR-test for non-tested alternatives. Although we do find somewhat lower elasticities the detected heteroscedasticity or parameter stability problems apear to be seious enough to caution against the use of these models for policy simulation.
Not available in German. This paper investigates the robustness of cross-section results on innovative behaviour of firms. It discusses specifications resulting from various assumptions on market structure and strategic behaviour. A number of cross-section and pooled probit results rely on restrictions which are rejected. The specification with correlated random effects is rejected and yields poor results, but the absence of correlation between random effects and regressors is also rejected.
Not available in German. The paper describes the adaptation of the estimation method proposed by Gallant and Nychka (1987) for the selectivity model to the threshold crossing model of binary choice under independence between errors and regressors. We present Monte-Carlo and asymptotic comparisons with the probit estimator and discuss appropriate maximization algorithms, suitable choice of starting values and strategies for the choice of the number of parameters used in approximating the density. Semi-nonparametric estimation is almost as efficient as probit estimation in normal samples and performs much better in non-normal samples. We also use the method for a participation model estimated on 3658 observations with 21 explanatory variables and show that it is practicable on modern personal computers. Pseudo score test results based on this methodology are presented, with special attention to heteroscedasticity as the main remaining potential cause for inconsistency.
Not available in German. When studying particular subgroups of a population the econometrician typically has few observations, and should draw upon any additional relevant information. We illustriate, for the estimation of a participation model for lone mothers, the relative benefits derived from using the panel structure of the data in a GMM framework and from including macro information in the form of extra moments.
Not available in German. When studying particular subgroups of a population the econometrician typically has few observations, and should draw upon any additional relevant information. We illustriate, for the estimation of a participation model for lone mothers, the relative benefits derived from using the panel structure of the data in a GMM framework and from including macro information in the form of extra moments.
We propose four different GMM estimators that allow almost consistent estimation of the structural parameters of panel probit models with fixed effects for the case of small T and large N. The moments used are derived for each period from a first order approximation of the mean of the dependent variable conditional on explanatory variables and on the fixed effect. The estimators differ w.r.t. the choice of instruments and whether they use trimming to reduce the bias or not. In a Monte Carlo study we compare these estimators with pooled probit and conditional logit estimators for different DGPs. The results show that the proposed estimators outperform these competitors in several situations. (doi:10.1081/ETC-120017972)
