bl ic Di sc lo su re A ut ho riz ed Pu bl ic Di sc lo su re A ut ho riz ed Pu bl ic Di sc lo su re A ut ho riz ed Pu bl ic Di sc lo su re A ut ho riz ed Unconditional Demand for Health Care in C8te d'Ivoire

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 un-observed individual e¤ects. The selection equation contains lags of the dependent variables from both the latent and the selection equations as well as other possibly predetermined variables relative to the latent equations. We derive a set of condi-tional moment restrictions that are then exploited to construct a three-step sieve estimator for the parameters of the main equation including a nonparametric esti-mator of the sample-selection term. In the second step the unknown parameters of the selection equation are consistently estimated using a transformation approach in the spirit of Berksons minimum chi-square sieve method and a …rst-step kernel estimator for the selection probability. This second-step estimator is of interest in its own right. It can be used to semiparametrically estimate a panel-data binary response model with a nonparametric individual speci…c e¤ect without making any other distributional assumptions. We show that both estimators (second and third stage) are p n-consistent and asymptotically normal. 1

We consider the estimation of sample selection (type II Tobit) models that exhibit spatial dependence. Attention is focused mainly on the spatial error dependence model (or spatial autoregressive error model, SAE), but our method can also be used to estimate the spatial lag dependence model (or spatial autoregressive model, SAL). The method considered is motivated by a two-step strategy analogous to the popular heckit model. The first step of estimation is based on a spatial probit model following a methodology proposed by Pinkse and Slade (1998) that yields consistent estimates. The consistent estimates of the selection equation are used to estimate the inverse Mills ratio (IMR) to be included as a regressor in the estimation of the outcome equation (second step). Since the appropriate IMR turns out to depend on a parameter from the second step under SAE, we propose to estimate the two steps jointly within a generalized method of moments (GMM) framework. We explore the large sample properties of the proposed estimator and undertake a Monte Carlo experiment to assess its performance in finite samples. Finally, we discuss the importance of the spatial sample selection model in applied work and illustrate its application by estimating the relative spatial efficiency of the production process within a fishery, using catch-per-unit-effort(CPUE)asameasureofefficiency. Key words and phrases: sample selection, spatial dependence, generalized method of moments.

Abstract: Sample selection is inherent to a range of treatment evaluation problems as the estimation of the returns to schooling or of the effect of school vouchers on test scores of college admissions tests, when some students abstain from the test in a non-random manner. Parametric and semiparametric estimators tackling selectivity typically rely on restrictive functional form assumptions that are unlikely to hold in reality. This paper proposes nonparametric weighting and matching estimators of average and quantile treatment effects that are consistent under more general forms of sample selection and incorporate effect heterogeneity with respect to observed characteristics. These estimators control for the double selection problem (i) into the observed population (e.g., working or taking the test) and (ii) into treatment by conditioning on nested propensity scores characterizing either selection probability. Weighting estimators based on parametric propensity score models are shown to be √ n-consistent and asymptotically normal. Simulations suggest that the proposed methods yield decent results in scenarios when parametric estimators are inconsistent.

Binary choice sample selection models are widely used in applied economics with large crosssectional data where heteroscedaticity is typically a serious concern. Existing parametric and semiparametric estimators for the binary selection equation and the outcome equation in such models suffer from serious drawbacks in the presence of heteroscedasticity of unknown form in the latent errors. In this paper we propose some new estimators to overcome these drawbacks under a symmetry condition, robust to both nonnormality and general heterscedasticity. The estimators are shown to be p n-consistent and asymptotically normal. We also indicate that our approaches may be extended to other important models.

Abstract not found

A semi parametric profil ~ likelihood method is proposed for estimation of sample selection models. The method is a two step scoring semi parametric estimation procedure based on index formulation and kernel density estimation. Under some regularity conditions, the estimator is asymptotically normal. This method can be applied to estimation of general sample selection models with multiple regimes and sequential choice models with selectivity. For the binary choice sample selection model, the estimator is asymptotically efficiency in the sense that its asymptotic variance matrix attains the asymptotic bound of G. Chamberlain. JEL classification number: 211 Keywords: Sample selectivity, discrete choice, sequential choice, semiparametric model, index model. semiparametric estimation, method of scoring, profile likelihood, kernel density, efficient estimator Corresponding address:

In this paper estimation of sample selection models using experimental data is considered with some weak restriction imposed on the error distribution. Under a normality setting, the most popular approach is the two-step method proposed by Heckman (1979). But Heckman’s approach relies on the nonlinearity of the probit function (i.e. the nonlinearity of the selection correction function) unless some exclusion restriction is imposed. Furthermore, Heckman’s method is sensitive to the underlying distributional assumption. Following this two-step method, several semiparametric estimators have been proposed for sample selection models by explicitly imposing the exclusion restriction. Using experimental data, this paper proposes a simple semiparametric matching method. There are certain advantages of our estimator over Heckman’s estimator and the existing semiparametric estimators under either the parametric setting and semiparametric setting. We do not rely on the nonlinearity of the selection correction function or the exclusion restriction. In addition, unlike other semiparametric methods, we can also estimate the intercept term in the equation of interest. The estimator is shown to be consistent and asymptotically normal under some regularity conditions. A small monte carlo study illustrates the usefulness of the new estimator. c ○ 2005 Peking University Press

Independence of the error term and the covariates is a crucial assumption in virtually all sample selection models. If this assumption is not satis…ed, for instance due to heteroscedasticity, both mean and quantile regression estimators are inconsistent. If independence holds indeed, all quantile functions and the mean function are parallel, which naturally limits the usefulness of quantile estimators. However, quantile estimators can be used to build tests for the independence condition because they are consistent under the null hypothesis. Therefore, we propose powerful tests based on the whole conditional quantile regression process. If the independence assumption is violated, quantile functions are not point identi…ed, but we show that it is still possible to bound the coe ¢ cients of interest. Our identi…ed set shrinks to a single point either if independence holds or if some observations are selected and observed with probability one. Therefore, our model generalizes simultaneously the traditional sample selection models and the identi…cation at in…nity strategy.

This paper, using survey data from Bangladesh, presents empirical evidence on externalities at household level sales decisions resulting from increasing returns to marketing. The increasing returns that arise from thick market e#ects and fixed costs imply that a trader is able to o#er higher price to the producers if the marketed surplus is higher in a village. The semi-parametric estimates identify highly nonlinear own and cross commodity externality e#ects in the sales of farm households. The vegetables markets in villages with low marketable surplus seem to be trapped in segmented local market equilibrium. The analysis points to the coordination failure in farm sales decisions as a plausible explanation of the lack of development of rural markets even after the market liberalization policies are implemented.