PostDocs in Computer Science: Where are we now – and where should we go from here?
Data from CRA’s annual Taulbee Survey document substantial growth in the cadre of U.S. and Canadian postdoctoral fellows over the past decade. Most recently, the Computing Community Consortium (CCC) and National Science Foundation (NSF) have funded one- to two-year postdoctoral positions through the Computing Innovation Fellows Project, in hopes of retaining recent PhDs in computing research and teaching during difficult economic times. We believe it is time for the community to understand the significance of this PostDoc surge, assessing whether it is the right course of action for the field in the long term.
According to Taulbee data, the numbers of new PostDocs soared from 60 in 1998 to 159 in 2009 (three-year rolling averages). Given that the average duration of a PostDoc position is about 1.8 years, this increase is consistent with published data from NSF’s Computing and Information Science and Engineering Directorate (CISE). They report that a total of 330 PostDocs participated in CISE grants in 2009. (CISE provides about 82 percent of all Federal funding for basic computing research, so 330 PostDocs is a fairly reasonable estimate of the total number of PostDocs in the field as a whole.)
By contrast, concurrent with the growth in the number of PostDocs, the absolute number of new tenure-track faculty has declined sharply (from a three-year rolling average of 224 in 2004 to an equivalent of 151 in 2009). Moreover, the number of recent PhDs who are hired into industry immediately out of graduate school is now approximately one-half of the total number of PhD graduates in any given year (755 out of 1,800 PhD graduates in 2009), up sharply from about one-quarter several years ago (243 of 919 went to industry in 2004). More and more PhD graduates are also pursuing first-time employment abroad.
We seek to ask whether the rise in PostDocs is positive or negative for the field. Is it positive or negative for the individuals? Is it a sign of a maturing discipline? Is it matched to the activities needed to conduct computing research today?
Whatever the case may be, it is important to recognize that this is a substantial change, to understand its significance and likely impact in the short and long term, and to determine the best way forward for the field.
While there are many perspectives on this change in the field, one of the most important is the impact on the PostDoc, the individual. By its very nature a PostDoc position is a temporary training position. Wages are dramatically lower than those of tenure-track, research and teaching faculty. In some universities the benefits of a PostDoc are more like those of a student than those of employees. Benefits to employees may be greater in areas such as health care, retirement, access to childcare, and access to wellness centers. At most universities and in industry, PostDocs cannot be Principal Investigators on grant proposals, reducing their independence at the time when they would be asserting it in most alternative positions.
From a personal perspective, PostDocs are typically at an age when they marry and start families. A PostDoc position is not permanent. The individual must do another job search and, typically, must move from one geographic locale to another with the career disruption, personal disruption and expense that this causes. Relocation is more difficult for women and men who are nurturing a young family.
On the positive side, the PostDoc generally has freedom to focus almost entirely on research, presumably theirs. Typically that is impossible for individuals in alternative positions such as teaching and tenure-track faculty. However, some PostDoc advisors do assign other obligations to the PostDoc. Credentials amassed during a PostDoc experience may materially increase the possibility of finding a permanent position in a more desirable organization than would otherwise be possible.
Anecdotally, some academic hiring committees are requiring publication records that are very difficult to amass during a PhD program, thus nudging the field to accept the PostDoc position as necessary in order to be hired into a tenure-track faculty position. Is this good for the field and for new entrants to the research enterprise?
We discuss the impact on the individual in this article. There are other perspectives to be considered. It is timely to have a discussion of these trends and to consider whether the CS research community should do things differently. These trends are the result of myriad individuals and organizations making independent decisions. Butultimately, those decisions should yield a collective result that the field believes to be most beneficial.
To help facilitate this discussion about the need for, and role of, PostDocs within the computing research community, the CRA commissioned a committee in November 2010 to prepare a white paper that reports the statistics associated with academic and industry hiring, articulates the relevant issues about PostDocs in the context of the many stakeholders, and specifically solicits input from the community. The white paper is available at https://turing.cra.org/postdocs.The goal is for the community to determine the paper’s conclusion by expressing their opinions on an associated blog—in particular, to articulate whether the PostDocs cadre should grow, shrink or stay the same.
We encourage you and your colleagues within your department or laboratory to review the white paper, discuss the issue, and post your views (collectively or individually) on the companion website (https://turing.cra.org/postdocs). We seek to get a sense of the community by March 15. Following review of the comments received, the committee will prepare a revised version of the white paper articulating the community’s broad view (and consensus, if any) on this issue.
Anita Jones, University of Virginia, is a Member of the CCC Council; Erwin Gianchandani is Director of the CCC.