Nicolas Coulombel

Variability of Travel Time, Congestion, and the Cost of Travel

The variability of travel time modifies the rush hour traffic and the cost of commuting. The bottleneck model of road congestion with fixed peak-load demand is augmented of an additive random delay. When individuals have (α-β-γ) preferences, there exists a unique Nash equilibrium. The variability of travel time leads to departure rates that are spread more evenly over the rush hour than when travel times are deterministic. This equilibrium mechanism mitigates both congestion and the cost of unreliability. This implies that “single-traveler models,” which treat congestion as an exogenous phenomenon, overestimate the value of reliability for the rush hour. The application with the uniform or with the exponential distribution helps appraise the overestimation.

Euclidean versus network distance in business location: A probabilistic mixture of hurdle-Poisson models

While the question of spatial weight matrix specification is now largely discussed in the spatial econometrics literature, the definition of distance has heretofore attracted less attention. The choice of the distance measure is often glossed over, with the ultimate use of the Euclidean distance. This paper investigates this issue in the case of establishments locating in the Paris region. High congestion, speed limits, or physical uncrossable barriers can diminish or totally eliminate the linkage between neighboring nareas challenging the choice of the Euclidean distance in representing the spatial effects. To compare the various distance measures, we develop a probabilistic mixture of hurdle-Poisson models for several activity sectors. Each model class uses a different distance definition to capture spillover effects. The following distance measures are considered: Euclidean distance, two road distances (with and without congestion), public transit distance, and the corresponding travel times. Overall, the obtained results are in line with the literature regarding the main determinants of establishments location. However, we find that for some activity sectors, such as construction, the peak period road travel time for private vehicles is the most likely to correctly capture spatial spillovers, whereas for other sectors, such as real estate, the Euclidean distance slightly prevails. This tends to show that spatial spillovers are channeled by different means, contingently on the activity sector. In addition, we find that the proposed mixture of hurdle-Poisson models that uses several latent classes performs significantly better than the pure hurdle-Poisson models based on a single distance measure, emphasizing the usefulness of our approach.