Persistence of commuting habits : context effects in Germany

Based on the geo-referenced data, I analyse the commuting behaviour of employees in Germany. With the help of a behavioural economic approach, which is based on the investigation of Simonsohn (2006) for the US, I can show that it is not only the wage and the individual heterogeneity that shape commuting decisions. Instead, the commuting behaviour depends on the context individuals observe in the past. In particular, I demonstrate that the commuting behaviour is influenced by past-observed commutes: Worker choose longer commuting times in a region they just moved to, the longer the average commute was in the region they moved away. This effect applies especially for older employees, but is the same for men and women. Moreover, my robustness checks indicate that individual heterogeneity, selectivity or endogeneity issues do not drive this effect. In addition, I show if individuals stay in the new region, the effect of the previous region disappears, as workers adapt the commuting behaviour of the new region and move again within the new region. This is consistent with the prediction of behavioural economic theory, but refuses the assumption of stable taste differences.