Working papers

Abstract: Many initiatives have been introduced worldwide by governments and industries to promote carpooling usage. In France, some local authorities have introduced carpooling subsidy policies since 2019 to encourage traveling by carpooling. We estimate the effect of local carpooling subsidies on the usage of short-distance carpooling organized by platforms using a difference-in-differences design that exploits variation across French “Communautés de Communes” (i.e. local authorities) in both the amount of subsidy and the timing of subsidy policy start. We find that the implementation of the carpooling subsidy increases the number of monthly short-distance carpool trips organized by platforms by around 5.5 trips per 1,000 inhabitants in the local authority area, and that this effect increases over time. Moreover, a 1-euro increase in carpooling subsidy improves the number of monthly carpool trips organized by platforms per 1,000 inhabitants by 3.9 trips and such an effect also increases with time. These average effects mask considerable heterogeneity, with subsidy increasing carpooling usage more in larger and higher-density local authority areas and the effect being negligible for local authority areas with the smallest size and density. Our results also suggest that carpooling subsidy effects do not differ with local authority area median income. 

(pdf)

Abstract: This paper investigates the impacts of several implementations of daily mobility policies and external shocks on social costs, with a particular focus on carpooling. This social cost is calculated considering consumer's cost, external costs, as well as the expenses and incomes of public authorities and private companies. Four transport modes are considered: solo driver, carpool driver, carpool passenger and public transport. A modal choice model is then applied to trips with 6,287 different origin-destination of the eastern Lyon area. Simulations of time-gain and monetary scenarios are then realized to observe impacts on demand and consequently on the other parameters affecting the social costs. Our results show that consumer's costs explain the majority of the total social cost. Values commonly used for externalities barely impact the social cost in the simulations and traffic reduction measures impact more public and private revenues than they reduce externalities, leading to higher total social costs. Moreover, results illustrate significant variations at the geographical scale, depending on the ODs where the policies are applied. These results suggest implementing daily-carpooling incentives should be decided conscientiously considering the local context. 

(pdf)

Abstract: This paper investigates pricing decisions when a monopolistic multi-sided platform is myopic, that is unable to distinguish between two agents who participate on the same side of the platform but produce different externalities. We find that the structure of prices is the same for profit- and welfare-maximizing platforms. The unique price supplied to the two undistinguishable agents is a weighted average of the perfect information prices, where the weights depend on demand elasticities and externalities produced by the other undistinguishable agent. The prices supplied to the distinguishable agents are also affected by information asymmetry through an extra term than can be positive or negative. Introducing Hotelling competition does not affect results. We apply the model to a monopolistic short-distance carpooling platform with and without HOV lane, and show that the profit-maximizing platform does not subsidize efficiently the “good” side of the market, leading to very little traffic reduction. These results call for a discussion of the regulation of myopic platforms in general, and those of carpooling in particular. 

(pdf)