Real-Time Search in the Laboratory and the Market

Abstract

While widely accepted labor market search models imply a constant reservation wage policy, empirical evidence strongly suggests that reservation wages decline in search duration. This paper reports the results of the first real-time-search laboratory experiment. The controlled environment subjects face is stationary, and the payoff-maximizing reservation wage is constant. Nevertheless, subjects' reservation wages decline sharply over time. We investigate two hypotheses to explain this decline: 1. Searchers respond to the stock of accruing search costs. 2. Searchers experience non-stationary subjective costs of time spent searching. Our data support the latter hypothesis, and we substantiate this conclusion both experimentally and econometrically. (JEL C91, D83, J64)