Reference-Dependent Preferences: Evidence from Marathon Runners [report]

Eric Allen, Patricia Dechow, Devin Pope, George Wu
2014 unpublished
Models of reference-dependent preferences propose that individuals evaluate outcomes as gains or losses relative to a neutral reference point. We test for reference dependence in marathon finishing times (n = 9, 378, 546). Models of reference-dependent preferences such as prospect theory predict bunching of finishing times at reference points. We provide visual and statistical evidence that round numbers (e.g., a four-hour marathon) serve as reference points in this environment and produce
more » ... nt and produce significant bunching of performance at these round numbers. Bunching is driven by planning and adjustments in effort provision near the finish line and cannot be explained by explicit rewards (e.g., qualifying for the Boston Marathon), peer effects, or institutional features (e.g., pace setters). We calibrate a simple model of prospect theory and show that the basic qualitative shape of the empirical distribution of finishing times is consistent with parameters that have previously been estimated in the laboratory. Recent theories of economic behavior propose that preferences are reference dependent. In other words, the evaluation of an outcome does not merely reflect tastes, risk attitudes, and wealth levels, as in classical economic models, but may also be affected by comparisons of that outcome with a reference point. For example, how an economic agent views a bonus of $1,000 might depend on the level of previous bonuses, what bonuses were distributed to other members of the organization, or the agent's expectations about what bonuses are possible (Card, Mas, Moretti, and Saez, 2012; Kahneman, 1992; Kőszegi and Rabin, 2006) . A reference point divides outcomes into gains or losses, thus creating a qualitative difference in the valuation of outcomes slightly above or below that reference point. For example, a primary feature of prospect theory, the most well-known account of reference-dependent preferences, is that the first derivative of utility is discontinuous at the reference point (Kahneman and Tversky, 1979; Tversky and Kahneman, 1992) . This property, known as loss aversion, has implications for a number of economic activities, including risky decision making, choice of consumption bundles, and effort provision (DellaVigna, 2009; Tversky and Kahneman, 1991) . A second property, diminishing sensitivity, results in a discontinuous second derivative and is captured by prospect theory's characteristic S-shaped value function that is concave for gains and convex for losses. Researchers have moved beyond Kahneman and Tversky's laboratory demonstrations of reference dependence to explain behavioral anomalies across a wide variety of field settings. 1 In a recent review of prospect theory, Barberis (2013) highlighted the key challenge to researchers testing for field evidence of reference-dependent preferences: it is often difficult to know exactly what reference points are relevant for individuals in field settings. Tversky and Kahneman (1991) proposed that "although the reference state usually corresponds to the decision maker's current position, it can also be influenced by aspirations, expectations, norms, and social comparisons." The difficulty in identifying the appropriate reference point is best illustrated by a stream of work examining the possible role that reference points play in labor supply and effort provision. Camerer et al. (1997) argued that taxi drivers have a downward-sloping labor supply curve due to reference-dependent preferences defined by daily income targets (see, also Fehr and Goette, 2007,
doi:10.3386/w20343 fatcat:hx4mq6g5fnhkvaq3bulr7sdxzu