Research, teaching, and ‘other’: what determines job placement of economics Ph.D.s?

ABSTRACT

This article examines job placement for Economics Ph.D.s seeking junior-level positions using a data set constructed from job candidate vitas, public websites, and queries to programme directors. Based upon multinomial logit estimation, being from highly ranked graduate institutions and having high quality publications has a significantly positive effect on placement at a top 20 academic institution or Doctoral-level institution. Teaching experience – as a teaching assistant (TA) or independent instructor – has a significantly positive effect on placement, but only for institutions ranked below the top 60, Masters and Baccalaureate institutions, and non-tenure track academic positions. We find little evidence on the effect of teaching in tenure track hires for departments with Doctoral programmes or mid-tier prestige. Moreover, teaching experience has a significantly negative effect on placement in the top group of academic institutions in Economics.

KEYWORDS:

• Job placement
• candidate
• Ph.D. program
• graduate school

JEL CLASSIFICATION:

• J24
• J44
• I23

Acknowledgements

The views expressed in this document are those of the authors and do not reflect the official policy or position of the Department of Defense or the U.S. Government. Distribution for this article is unlimited. We thank without implication two anonymous referees as well as Scott Carrell, Jesse Cunha, Wayne Grove, David Henderson, Jeff Kubik, Jonathan Lipow, Bob McNab, and seminar participants at Syracuse University and the U.S. Naval Postgraduate School for helpful comments on previous drafts. Many thanks for research assistance go to Sidney Hilton, Anthony Seifort and Mallory Vachon. In addition, we thank Kathleen Bailey, Abbie Beane, and Richard Mastowski for editing assistance, as well as Christian Zimmermann for allowing us access to his data.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ The number of Ph.D. granting programmes listed on the NBER.org website has increased since our data collection efforts began in 2010. For example, 159 schools are listed as of 6 May 2015.

² The sole exception to this rule is the ranking of the American Economic Journals (AEJs). The AEJs were not ranked categorically in Combes and Linnemer (2010), so we assume they are in the high-tier category. For reference, the AAA journals are considered the top 5 in the profession, i.e. Quarterly Journal of Economics, American Economic Review, Journal of Political Economy, Econometrica, and Review of Economic Studies and the AA journals are ranked 6 through 20; see Combes and Linnemer (2010) for a complete listing of all the rankings.

³ RePEc is a collaborative effort of volunteers in 75 countries to enhance the dissemination of research in economics and related sciences. The heart of the project is a decentralized bibliographic database of working papers, journal articles, books, books chapters, and software components.

⁴ IDEAS is a RePEc service hosted by the Economic Research Division of the Federal Reserve Bank of St. Louis.

⁵ We thank Christian Zimmerman for the full set of rankings from Zimmerman (2011). We are not permitted to share the entire author rankings due to confidentiality issues.

⁶ While ranking economics departments is inherently subjective, different rankings tend to be highly correlated when they use the same general technique. Dusansky and Vernon (1998) find that the Pearson correlation coefficient between the Scott and Mitias (1996) ranking and seven other economics department rankings range from 0.71 to 0.85. The 20 top economics departments using the Kalaitzidakis, Mamuneas, and Stengos (2003) ranking overlap substantially with the National Research Council’s ranking of the top 20 departments.

⁷ This was only done for departments that listed job placements online. If a department did not list job placements online and the placement director did not return our request for information, then those observations were omitted from our analysis, due to having no job placement information.

⁸ We use the 2010 version of the Carnegie Classification. The website’s responsibility has been transferred to Indiana University as of 1 January 2015. This may include some recategorization as well as classification of new types of institutions of higher education (see http://classifications.carnegiefoundation.org/).

⁹ Although the Carnegie Classification includes a number of other categories such as Associate colleges, all tenure track placements in U.S. institutes in our sample fall into one of our three categories.

¹⁰ Complete results from all estimated models referred to in this article are available from the authors upon request.

¹¹ All significance results regarding the unreported control variables from the estimated models in this section are based upon the 10% critical level.

¹² Obtaining reliable estimates from the multinomial logit procedure relies on the property that each characteristic have at least one non-zero value in each category of job placement. In addition, to avoid collinearity with the intercept, any characteristic represented by a dummy variable must have at least one value of zero in each category of placement. The large SEs for estimates from the high- and mid-tier Graduate Institutions for the top 20 ranked universities in Table 6 likely stems from the fact that applicants from these groups represent nearly all the hires for these institutions within the subsample of applicants from U.S. universities. Correspondingly, the large SEs of the estimated coefficient for ‘Black’ reflect that the top 20 ranked schools had almost no hires of Black race in our full sample and subsample. All the remaining SEs in our estimated models do not seem to be affected in this way.

¹³ As a robustness check, we performed a multinomial logit estimation by adapting the model in Table 7 to include separate categories for positions in tenure track Masters and Baccalaureate institutions. The results indicate one significantly positive (at the 5% critical level) teaching effect for each group – semesters teaching for Masters institutions and semesters as a TA for Baccalaureate institutions. Significantly negative estimates for high-tier and mid-tier graduate institutions occur for the Masters institutions. The remaining findings are largely the same as for the estimated model in Table 7.

¹⁴ We also investigated the effects of interactions of fields and prestige of advisors, as mentors from different fields may have different importance in hiring. Apart from a couple of significant estimates (for Finance in schools ranked 1–20 and public for schools ranked 61 and above, both negative), the results indicate little evidence of this effect.

Additional information

Funding

Research funding for the project was provided by the Defense Resources Management Institute.