Jobs–housing imbalance, spatial correlation, and excess commuting

► We adopt Vaughan’s model to calculate unbiased excess commuting. ► The relationship between urban structure and commuting efficiency is verified. ► Spatial correlation is also important parameter to describe commuting efficiency. ► Excess commuting and capacity utilization evaluate US and Japanese/Korean cities. In this paper, we use continuous urban structure instead of zonal model, try to calculate unbiased excess commuting with joint distribution of homes and workplaces developed by Vaughan (1974), and describe the relationship between urban structure and commuting distance explicitly and theoretically for generalized home–workplace assignment pattern. We simplify the quadrivariate distribution model to a model with three important parameters: the spread of homes, the spread of workplaces, and the spatial correlation of homes and workplaces. Then we show that excess commuting and capacity utilization are expressed by the imbalance and the spatial correlation of jobs–housing structure in a theoretical context, moreover it explicitly evaluates targeting US and Japanese/Korean cities.