Compensation for commuting in imperfect urban markets

We develop an urban equilibrium job search model where residential mobility is restricted due to the presence of residential moving costs. We presume a simple mono-centric model (firms are located in one location), but allow for imperfect labour and housing markets. We set out to analyse an urban equilibrium model presuming imperfect labour markets (bargaining power, search behaviour) and allowing for residential moving costs aiming to explain the empirical observation that for workers at a particular workplace (in the current model, the CBD), wages depend on residence location and therefore on commuting costs (Zax, 1991). One of the main implications of the model is that due to residential moving costs, rent gradients are less steep than predicted by standard urban theories in line with a range of empirical studies (Dubin and Sung, 1987). Using a theoretical urban economic model with multiple employment centres, Crane, 1996, and Turnbull , 1998, both obtained a similar result when presuming future job site uncertainty and the presence of residential moving costs that constrain households to live in the same place. In these models, the residential location is based not only one where the current job is located, but also on the expectation of where future jobs will be located. Although our model is quite different from the models of Crane, 1996, and Turnbull, 1998, (in particular, we presume wage bargaining and only one employment centre), these models share the assumption of residential moving costs and labour market imperfections. This suggests that the introduction of residential moving costs combined with labour market imperfections may be fundamental to our understanding of the relationship between urban labour and housing markets and the implications for commuting compensation. To conclude, we call for a better understanding of residential moving costs in urban markets. Our study should be seen as a first step towards this goal. More realistic models, for example urban models that include on-the-job-search, endogenous firm location and congestion, may be needed to confirm our results.Keywords: job matching; moving costs; rent gradient. JEL: R20; R23; J64.