The Travel-to-work. Which factors matter? An analysis on regional labor market in UK

The aim of this study is focus on the role of positive and negative monetary incentives in stimulating infra and inter-regional mobility in UK (through commuting behavior). We consider a simple theoretical model and develop an empirical application using data from the Quarterly Labour Force Survey (QLFS) 2004-2011. According to the theory, workers who commute from one labor market to another should require earning grants able to compensate, at least, monetary and non-monetary moving costs and individual evaluation of the trade-off between the latter and the value of the monetary incentive. First, we estimate the predicted wage in the region of current residence through OLS and then we verify the increase in the probability of moving to other UK GORs (Government Office Regions, representing the UK regions and the six Metropolitan Areas) due to greater hourly wages w.r.t. that predicted for the region of current residence through a Multinomial Logit model (MLM). We consider four different cases: living and working in the same region (or territorial agglomeration) which represent the baseline outcome of the MLM; working in a region whose travel distance is less than 30 minutes from the residence; working in a region whose distance is between 30 and 45 minutes from the residence; working in a region at 45 minutes of travel distance. In our analysis, we find that earnings, sex differences and individual characteristics play an important role in explaining commuting behavior, as well as the length of the employment within the same workplace. We find that ? on average ? wages 20% greater than those for the region of residence lead to an increase in probability of 2.2-3.2% (according to gender differences) to move to other regions commuting more than 45 minutes. Moreover, we find support also for the ?household responsibility hypothesis, confirming that women are less willing to commute longer and to move outside their own region of residence compared to their male counterparts. Finally, for robustness check and for taking into account possible biases in the model due to the role played by the capital, we develop the two-stage model without considering the region of Greater London, concluding. We find that, excluding the region of Greater London from the sample, there is a raise from 1% to 1.6% in the gender gap in the probability of commuting more than 45 minutes, with the consequence that female workers working in London are more willing to commute than their national counterpart.