Academic journal article Economic Inquiry

Do Economics Departments Search Optimally in Faculty Recruiting?

Academic journal article Economic Inquiry

Do Economics Departments Search Optimally in Faculty Recruiting?

Article excerpt


When an economics department decides to recruit new faculty, it must decide in which fields to conduct the search. Casual observation suggests that the scope of recruiting searches varies widely, with some departments searching primarily in a narrow subfield and others searching in several general or even all fields. It is generally recognized that the very top departments tend to engage in very wide "best athlete" searches. Among departments that engage in narrower searches, on the other hand, it is not uncommon to hear complaints ex post that the search should have been broader. Is it therefore the case that these departments are making a suboptimal choice to search narrowly? It would seem, however, that economic departments, more so than other departments, should be making economically rational decisions when choosing search scope. (1)

In this article we develop a simple model of how economics departments can optimally choose search scope in faculty recruiting. We show that the optimal search scope is increasing in the quality rank of the department. We use postings in Job Openings for Economists (JOE) to estimate the relationship between department rank and search scope and find that higher-ranked departments engage in broader searches than lower-ranked departments. The relationship is robust to the exclusion of the top-ranked departments from the sample. Because there is some debate about how well various department rankings reflect true department quality, we instrument a reputation-based ranking with a publication-based ranking to correct for measurement error. We find that a 10-place difference in department ranking is associated with 3.5-4.8 more Journal of Economic Literature (JEL) subfields in a position announcement.


We consider a simple model of a department's choice of the number of fields to search and show that the optimal number of fields searched is increasing in department quality. Other work on employer search has noted the benefits to some employers of broadening the applicant pool. Much like our finding that departments with higher quality standards will search more broadly, Barron et al. (1985) show that employers will search over more candidates and/or more intensively if the education requirements for the position are high. Barron et al. (1997) argue that there is greater variation in productivity at higher levels of human capital; therefore it is optimal for employers searching for workers with more formal education to spend more on search. Lang (1991) shows that employers for whom a job vacancy is the most costly will offer a higher wage, therefore increasing the number of prospective applicants.

Because the need to decide ex ante the fields to advertise in JOE largely precludes the use of sequential search methods, our model is essentially a fixed-sample-size search problem. A key feature of our model is that each department has a quality cut-off that reflects the department's ranking. Higher-ranked departments will have higher quality cut-offs. Applications below this cut-off are thrown out without cost, and applications above the cut-off are reviewed more extensively with some positive cost. (2) Therefore, if a high-ranked department with a high quality cutoff searches narrowly, it may not receive any applications above its cut-off. In addition, because the higher-ranked department can ignore most applications without cost, the cost involved in expanding the search to other fields is lower than those incurred by lower-ranked departments with lower cut-offs. These two effects will be important in explaining why a higher-ranked department is better off expanding its search to more fields.

To formalize the model, there are i = 1, ..., M fields in which a department can conduct its search. Each field searched by the department produces one applicant, whose quality, [q.sub.i], is a random draw from a uniform distribution on support [0,1]. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.