Some estimators for dynamic panel data sample selection and switching models
نویسندگان
چکیده
We present estimators for panel data sample selection and switching models where the regression equations are dynamic and it is allowed for the existence of endogenous regressors and correlated individual unobserved heterogeneity. We consider two types of switching models under the names of observed dynamics switching and latent dynamics switching. The dynamic sample selection model implicitly assumes an underlining latent dynamics switching regime process. The type of methods presented are different Generalized Method of Moments (GMM) estimators for dynamic panel data sample selection and switching models that may combine estimation of the models both in first time differences and in level equations (with the corresponding sample selection correction terms under one or the other case). Therefore, we consider the possibility of applying System-GMM estimators for dynamic panel data to the case of sample selection and switching models. Depending both on estimation in levels or time differences and on the types of switching models considered (observed or latent) the sample selection correction terms present different degrees of complexity. Some of this complexity can be simplified if we are willing to impose stationarity assumptions, exchangeability conditions, and/or lack of individual heterogeneity in the selection equations determining the switching regimes. In the general setting neither stationarity, exchangeability, or lack of individual heterogeneity in the selection equations are imposed. To see the performance of the proposed estimators we perform a Monte Carlo study of the finite sample properties of different Generalized Method of Moments (GMM) estimators for dynamic panel data sample selection and switching models. Finally, we present an empirical example using Spanish data on wage settlements and strike outcomes. In an economic context in which workers may strike to obtain a wage concession, the strike decision is endogenous to the wage process and the wage equation is then affected by endogenous selection. We test this as well as alternative economic hypotheses in a dynamic context.
منابع مشابه
Higher Education and Labor Market Imbalances in Iran: A Dynamic Panel Data Analysis
Higher Education (HE) in Iran have been subject to a major expansion and massification in the recent years, in a way that number of students approximately tripled from 2006 to 2016. This would have possibly affected labor market or unemployment rate of the country. Considering both provincial and national level, this study investigates the relationship between HE expansion and unemployment rate...
متن کاملRoot-N Consistent Semiparametric Estimators of a Dynamic Panel-Sample-Selection Model
This paper considers the problem of identi cation and estimation in panel-data sample-selection models with a binary selection rule when the latent equations contain possibly predetermined variables, lags of the dependent variables, and unobserved individual e¤ects. The selection equation contains lags of the dependent variables from both the latent and the selection equations as well as other ...
متن کاملBandwidth Choice for Bias Estimators in Dynamic Nonlinear Panel Models
This paper considers bandwidth selection for spectral density estimators based on panel data sets. The spectral densities of greatest interest in this paper are the ones that appear in the bias expression for xed e¤ects estimators in nonlinear dynamic panel models obtained by Hahn and Kuersteiner. The bias estimation problem is di¤erent from the usual HAC estimation problem because the need fo...
متن کاملSimple Endogenous Binary Choice and Selection Panel Model Estimators
This paper provides numerically trivial estimators for short panels of either binary choices or of linear models that suffer from confounded, missing not at random sample selection. The estimators allow for Þxed effects, endogenous regressors, weakly exogenous regressors, and heterokedastic errors with unknown distribution. The estimators, which converge at rate root n, are based on variants of...
متن کاملQuasi Maximum-Likelihood Estimation of Dynamic Panel Data Models
This paper establishes the almost sure convergence and asymptotic normality of levels and differenced quasi maximum-likelihood (QML) estimators of dynamic panel data models. The QML estimators are robust with respect to initial conditions, conditional and time-series heteroskedasticity, and misspecification of the log-likelihood. The paper also provides an ECME algorithm for calculating levels ...
متن کامل