סמינר באסטרטגיה

Neil Thakral, Department of Economics, Harvard University

We document evidence of high-frequency reference-point adjustment in the field. 

Analyzing a data set of all New York City cab fares in 2013 using non-parametric methods, we find reductions in cab driver labor supply in response to higher accumulated daily earnings and stronger effects for more recent earnings. 

The income effect is inconsistent with the neoclassical model and the non-fungibility of daily income rejects models invoking daily income targets. 

To explain the evidence, we incorporate adaptive reference points into models of loss aversion and salience. 

While loss aversion tends to overstate the main quantitative features of the data, both models capture the qualitative features.