Wrong Turns and Right Tails: Identifying Detours in New York City Taxi Rides

Abstract

Credence goods are economic goods whose quality is difficult for consumers to verify due to information asymmetry. An example of a credence good is a taxicab ride. This paper presents a novel methodology of evaluating the quality of taxi services offered using aggregate New York City taxi trip data. Using my methodology, I document evidence that taxi drivers offer lower quality taxi services to passengers by taking detours in response to financial incentives. I find that taxi drivers take significant detours on trips where they are paid by distance driven, while they do not take detours when there is no financial incentive to do so, or when their perceived risk of getting caught taking a detour is high. I document heterogeneity in detour behavior amongst drivers in that drivers who take detours do so often, while those who do not take detours almost never do. I explore theories of social learning and self-selection to explain why some drivers take detours while others do not.