Title: A Bayesian spatial sample selection model with an application to credit constraints for small businesses
Authors: Raffaella Calabrese - University of Edinburgh (United Kingdom)
Michaela Kesina - ETH Zurich (Switzerland) [presenting]
Abstract: A spatial autoregressive probit model is developed that corrects for sample selection bias. We consider a two-step model with a spatial lag of a latent variable in both the selection and outcome equation. We suggest to jointly estimate both steps using a Bayesian Markov chain Monte Carlo (MCMC) simulation approach. We explore the finite sample properties of the estimator using Monte Carlo simulations. We apply the proposed model to data on UK Small and Medium-size Enterprises (SMEs) to estimate how neighbourhood effects can help to explain the external financing decisions of UK SMEs and their access to credit.