Job Search Intensity, Wage Preference, and Job Preference during an Unemployment Spell:  New Evidence from Micro-Data on 8 Million Job-Seekers  (w/ Tanya Menon)


Economists have long sought to understand the relationship between job search and unemployment duration. Our clearest empirical evidence has thus far come from time-use surveys of reservation wage and search intensity (Krueger and Mueller 2011; Krueger and Mueller 2010). We approach the question of how search evolves over an unemployment spell with highly unique micro-data from a major online job search site. Specifically, we observe a panel of daily level search data for 8 million job users over a 14 month period stretching from 2009 to 2010, and an even richer set of “click-level” data for several hundred thousand users over a 3 month period. We first describe the time-path of search intensity, search expansiveness with respect to job type, and wage preferences over the course of the unemployment spell. Using registration data, user resumes, and full texts of job listings, we then document the heterogeneity of our findings across occupation, experience, prior wage, gender, and state unemployment generosity. We then situate these findings in the context of optimal unemployment insurance and models of job search.