We present a natural field experiment designed to measure other–regarding preferences in the market for taxis. We employed testers of varying ethnicity to take a
number of predetermined taxi journeys. In each case we endowed them with only
80% of the expected fare. Testers revealed the amount they could afford to pay to
the driver ...
We present a natural field experiment designed to measure other–regarding preferences in the market for taxis. We employed testers of varying ethnicity to take a
number of predetermined taxi journeys. In each case we endowed them with only
80% of the expected fare. Testers revealed the amount they could afford to pay to
the driver mid–journey and asked for a portion of the journey for free. In a 2×2
between–subject design we vary the length of the journey and whether or not a business card is elicited. We find that (1) the majority of drivers give at least part of
the journey for free, (2) giving is proportional to the length of the journey and (3)
that 27% of drivers complete the journey. Evidence of out–group negativity against
black testers is also reported. In order to link our empirical analysis to behavioural
theory we estimate the parameters of a number of utility functions. The data and the
structural analysis lend support to the quantitative predictions of experiments that
measure other–regarding preferences, and shed further light on how discrimination
can manifest itself within our preferences.
