Estimation of Spatial Sample Selection Models: A Partial Maximum Likelihood Approach

Renata Rabovic, Pavel Cizek

Research output: Working paperDiscussion paperOther research output

Abstract

To analyze data obtained by non-random sampling in the presence of cross-sectional dependence, estimation of a sample selection model with a spatial lag of a latent dependent variable or a spatial error in both the selection and outcome equations is considered. Since there is no estimation framework for the spatial lag model and the existing estimators for the spatial error model are either computationally demanding or have poor small sample properties, we suggest to estimate these models by the partial maximum likelihood estimator, following Wang et al. (2013)'s framework for a spatial error probit model. We show that the estimator is consistent and asymptotically normally distributed. To facilitate easy and precise estimation of the variance matrix without requiring the spatial stationarity of errors, we propose the parametric bootstrap method. Monte Carlo simulations demonstrate the advantages of the estimators.
LanguageEnglish
Place of PublicationTilburg
PublisherCentER, Center for Economic Research
Number of pages78
Volume2016-013
Publication statusPublished - 31 Mar 2016

Publication series

NameCentER Discussion Paper
Volume2016-013

Fingerprint

bootstrapping
matrix
sampling
simulation

Keywords

  • Asymptotic distribution
  • Maximum likelihood
  • near epoch dependence
  • sample selection model
  • Spatial Autoregressive Models

Cite this

Rabovic, R., & Cizek, P. (2016). Estimation of Spatial Sample Selection Models: A Partial Maximum Likelihood Approach. (CentER Discussion Paper; Vol. 2016-013). Tilburg: CentER, Center for Economic Research.
Rabovic, Renata ; Cizek, Pavel. / Estimation of Spatial Sample Selection Models : A Partial Maximum Likelihood Approach. Tilburg : CentER, Center for Economic Research, 2016. (CentER Discussion Paper).
@techreport{8a4b2e5d678746858b9e128d0e6d4e47,
title = "Estimation of Spatial Sample Selection Models: A Partial Maximum Likelihood Approach",
abstract = "To analyze data obtained by non-random sampling in the presence of cross-sectional dependence, estimation of a sample selection model with a spatial lag of a latent dependent variable or a spatial error in both the selection and outcome equations is considered. Since there is no estimation framework for the spatial lag model and the existing estimators for the spatial error model are either computationally demanding or have poor small sample properties, we suggest to estimate these models by the partial maximum likelihood estimator, following Wang et al. (2013)'s framework for a spatial error probit model. We show that the estimator is consistent and asymptotically normally distributed. To facilitate easy and precise estimation of the variance matrix without requiring the spatial stationarity of errors, we propose the parametric bootstrap method. Monte Carlo simulations demonstrate the advantages of the estimators.",
keywords = "Asymptotic distribution, Maximum likelihood, near epoch dependence, sample selection model, Spatial Autoregressive Models",
author = "Renata Rabovic and Pavel Cizek",
year = "2016",
month = "3",
day = "31",
language = "English",
volume = "2016-013",
series = "CentER Discussion Paper",
publisher = "CentER, Center for Economic Research",
type = "WorkingPaper",
institution = "CentER, Center for Economic Research",

}

Rabovic, R & Cizek, P 2016 'Estimation of Spatial Sample Selection Models: A Partial Maximum Likelihood Approach' CentER Discussion Paper, vol. 2016-013, CentER, Center for Economic Research, Tilburg.

Estimation of Spatial Sample Selection Models : A Partial Maximum Likelihood Approach. / Rabovic, Renata; Cizek, Pavel.

Tilburg : CentER, Center for Economic Research, 2016. (CentER Discussion Paper; Vol. 2016-013).

Research output: Working paperDiscussion paperOther research output

TY - UNPB

T1 - Estimation of Spatial Sample Selection Models

T2 - A Partial Maximum Likelihood Approach

AU - Rabovic, Renata

AU - Cizek, Pavel

PY - 2016/3/31

Y1 - 2016/3/31

N2 - To analyze data obtained by non-random sampling in the presence of cross-sectional dependence, estimation of a sample selection model with a spatial lag of a latent dependent variable or a spatial error in both the selection and outcome equations is considered. Since there is no estimation framework for the spatial lag model and the existing estimators for the spatial error model are either computationally demanding or have poor small sample properties, we suggest to estimate these models by the partial maximum likelihood estimator, following Wang et al. (2013)'s framework for a spatial error probit model. We show that the estimator is consistent and asymptotically normally distributed. To facilitate easy and precise estimation of the variance matrix without requiring the spatial stationarity of errors, we propose the parametric bootstrap method. Monte Carlo simulations demonstrate the advantages of the estimators.

AB - To analyze data obtained by non-random sampling in the presence of cross-sectional dependence, estimation of a sample selection model with a spatial lag of a latent dependent variable or a spatial error in both the selection and outcome equations is considered. Since there is no estimation framework for the spatial lag model and the existing estimators for the spatial error model are either computationally demanding or have poor small sample properties, we suggest to estimate these models by the partial maximum likelihood estimator, following Wang et al. (2013)'s framework for a spatial error probit model. We show that the estimator is consistent and asymptotically normally distributed. To facilitate easy and precise estimation of the variance matrix without requiring the spatial stationarity of errors, we propose the parametric bootstrap method. Monte Carlo simulations demonstrate the advantages of the estimators.

KW - Asymptotic distribution

KW - Maximum likelihood

KW - near epoch dependence

KW - sample selection model

KW - Spatial Autoregressive Models

M3 - Discussion paper

VL - 2016-013

T3 - CentER Discussion Paper

BT - Estimation of Spatial Sample Selection Models

PB - CentER, Center for Economic Research

CY - Tilburg

ER -

Rabovic R, Cizek P. Estimation of Spatial Sample Selection Models: A Partial Maximum Likelihood Approach. Tilburg: CentER, Center for Economic Research. 2016 Mar 31. (CentER Discussion Paper).