Social Comparisons in Job Search: Experimental Evidence

Using a laboratory experiment we examine how social comparisons affect behavior in a sequential search task. In a control treatment, subjects search in isolation while in two other treatments subjects get feedback on the search decisions and outcomes of a partner subject. The average level and rate of decline in reservation wages are similar across treatments. Nevertheless, subjects who are able to make social comparisons search differently from those who search in isolation. Within a search task we observe a reference wage effect: When a partner exits, the subject chooses a new reservation wage which is increasing in partner income. We also observe a social learning effect: Between search tasks, subjects who have been paired with a more patient and successful partner increase their reservation wages in the next task.