Are Long Commute Distances Inefficient and Disorderly?

By joining ideas from the spatial interaction and excess commuting literatures, we integrate and organize three related aspects of commuting statistics: trip length, distance decay, and entropy. We use concepts and calculations from the excess commuting literature to improve interpretation of trip length parameters in a doubly constrained trip distribution model. We derive a direct mathematical relationship between trip length and entropy, thereby providing an analytical model to calculate the ease with which trip lengths might be reduced. The new analysis not only ties these statistics together in a mathematically useful way, but also allows for a new entropy metric—relative entropy—that determines the organization (or lack thereof) of aggregate work trips. As a result, we disentangle the entropy component from the distance component and offer new perspectives on aggregate commuting in a comparative framework. Twenty-five cities are classified in an interesting and useful way that includes the average trip length and the degree of organization in each city.