Commuting and self-selection in Northern Sweden

The book presents the results from estimation of the job-search model on earning and commuting in the Northern Sweden. The aim of this study is to examine the potential role of the earnings in the decision-making period on the choice of being a commuter and covariance between unmeasured traits in the earning and commuting equations. Understanding reasons of selectivity helps to avoid the selection bias, which leads to inconsistent estimation of results and policy misapplications. The analysis of the male and female samples of the population that start commuting, from either urban or rural areas, was carried by the joint maximum likelihood estimation of the earning equation and commuting equation in the subsequent period. The results indicate positive selection based on the observable characteristics and a negative selection based on the unobserved traits for the female sub-samples.