The path to professorship by the numbers and why mentorship matters

What does it take to land a top academic job? A look at new data shows that the job market is competitive and that publications matter, but it’s not the whole story. Mentorship matters more than we think – both in getting the job and succeeding at it.
Published in Social Sciences
The path to professorship by the numbers and why mentorship matters
Like

It’s easy for aspiring students to envision the goal: a tenured professor happily researching, publishing, and inspiring young minds - getting paid to be a “finder outer”, as Harvard social psychologist Daniel Gilbert puts it. But the path to professorship is less clear. Is it really publish or perish? And why should we care about mentorship along the way?

To get a sense of expectations for the academic job market, I turned to my own field, psychology, and scraped the CVs of over 150 current assistant professors (most hired in the past six years) from top-ranked research universities and small liberal art colleges (SLACs) – positions coveted by many budding academics. I’ve woven these data in below, and although not a random sample, it hopefully provides insights into the academic job market and what it takes to land a top job.

By The Numbers

In order to become a professor, you need a PhD.1 It comes as no surprise that PhD programs can be grueling as whip smart and hard-working graduate students stave off imposter syndrome and loneliness, attempting to gulp down new knowledge (often teaching themselves and learning on the fly) while trying to publish their own research amidst other responsibilities (e.g., teaching, mentoring, etc.).2 Adding to these challenges are uncertainties about how to secure an academic job.

Image by Nino Carè from Pixabay

I think most advisors believe that when it comes to publishing papers, “quality trumps quantity,” and that’s probably true. But the days of publishing just one or two papers and landing a tenure-track professor position in the US are long gone. While your advisor or their advisor might have landed their first faculty position with a couple first-author papers and a few co-authored papers, that CV barely qualifies you for a good postdoc these days.

"When I started my career almost 30 years ago, a few peer-reviewed publications could secure an academic job at a storied institution…Today…a CV that used to get you a job now makes you competitive for a postdoctoral fellowship."

 -  Lisa Feldman Barrett, President of the Association for Psychological Science and University Distinguished Professor of Psychology at Northeastern University

This isn’t just folklore either. By sifting through stacks of CVs, and in line with findings from others (Pennycook & Thompson, 2018), there appears to be a positive correlation around .24 between the year an assistant professor was hired and the number of publications they had, suggesting that today’s early career researchers need to publish more and more.3

Moreover, I found that today’s top R1 (institutions that grant many doctoral degrees and have high research output) assistant professor of psychology hires have around 16 publications, about half of which are first author. These numbers might sound intimidating for graduate students, though it’s worth noting that the average hire was 5 years out from their PhD, with over 80% of these hires doing a postdoc and a quarter coming from a previous faculty position (see Table 1). 

Image by Jerzy Górecki from Pixabay

Take a recent example.4 A US R1 university sought to hire a social psychologist at the assistant professor level. From 218 applicants (30% graduate students), they made four cuts using a set of predetermined criteria.

The first cut, emphasizing productivity and research quality, counted the number of publications in certain “top journals”: either prestigious journals (e.g., Nature) or field-specific journals (e.g., Journal of Experimental Social Psychology). Applicants with a sufficient number of “top journal” papers (e.g., at least 1 for graduate students, 2 for postdocs/early career faculty) moved forward. Those without were chopped, unless they had a strong research and diversity statement to pull them back into the running.

This first cut left 93 applicants (only 9 were “pulled back up” despite not meeting the publication cut-off). The second cut re-reviewed applicants’ research and diversity statements for specific criteria (e.g., communication of ideas, research synergy within the area and department, importance of discovery to the field and world), leaving 36 applicants. The third cut examined all materials except teaching statements, with almost all criteria in mind (e.g., broad thinking, productivity, quality of science, ability to win grants, contributions to diversity, open science), leaving 12 finalists who received Skype interviews.

The average finalist at this stage had nearly 10 first-author papers, 3 in “top journals.”5 Although 3 finalists were postdocs and 2 were graduate students, a full 7 were already faculty.

The fourth cut relied on the Skype interviews, a full review of the applicant’s portfolio (including teaching statements) and calls to references (letter writers, former students and colleagues), to whittle down to a final 5 candidates who received on-campus interviews (i.e., job talks). Of those 5 final candidates, 3 were faculty and 2 were postdocs. 

The initial offer was made to one of the faculty. That a faculty applicant was the department’s top choice is an example of how landing one’s first faculty position may be the hardest. Once you have a faculty position, moving to new academic jobs becomes easier.

Image by Andrew Martin from Pixabay

Expectations for SLACs also include strong research and teaching records. While systematic data for non-R1 schools are harder to come by, I found that postings for psychology jobs at top-ranked SLACs appear to elicit upwards of 200 applicants. Of those hired as assistant professors, about half have done a year postdoc, and a third are coming from a previous faculty position (often a visiting assistant professorship). Hires typically have around 11 publications, about a third as first author, plus strong teaching experience including instructor on record for several courses and various TAships (see Table 1).

Profile of Assistant Professor Hires in Psychology from top-ranked R1s and SLACs

Up to and including the year they began their job… Research University (R1)
N = 112
Small Liberal Art College (SLAC)
N = 43
# of publications 16 (22) 11 (6)
# of 1st author publications 7 (8) 4 (3)
# of instructor of record 1 (5) 3 (4)
# of TAships 1 (3) 3 (4)
# of years since PhD 5 (3) 2 (4)
% who did a postdoc 83% 56%
% who had a prior faculty position (TT or NTT) 27% 35%

Table 1. This table presents the data – statistics are medians with standard deviations in parentheses – from 155 assistant professor hires in psychology from top-ranked research universities (R1s) and small liberal art colleges (SLACs). R1s were chosen based on their graduate psychology program being ranked in the top-10 by U.S. News & World Report. SLACs were chosen based on the college being ranked in the top-10 by U.S. News & World Report, though I added a few extra similar reputable colleges to increase my sample size for this group. Every assistant professor listed on the institution’s psychology department website was included (including visiting assistant professors who are more prominent at SLACs), though adjuncts or clinical positions were excluded. As I recorded all information manually, via CVs or online profiles (e.g., university website, LinkedIn, Google Scholar), it is possible there are occasional coding errors and there was missing data for some faculty. Since I did not have access to the job applications of these now-professors, all stats (e.g., papers) are up to and including the year the person began their position. While this may produce slight overestimates, I suspect that applicants would have included work that is “submitted” or “under review” or “in press” (on their CVs/research statements and in job talks) and that hiring committees would have considered such work in their decision. To be conservative, publications do not include conference proceedings (i.e., only journal articles, chapters, commentaries, or encyclopedic entries (which were rare) counted). Data and code are publicly available here: https://osf.io/z8rhg/.


Raising the bar even more is an over-saturated academic job market. In 1970, there were around 2-3 psychology PhD graduates per assistant professor of psychology vacancy6 (Rose, 1972) yet in 2017 that number is around 8.7 This oversupply of PhDs spans across disciplines: Columbia University’s English department recently came under scrutiny for admitting 19 new doctoral candidates despite only placing 1 PhD candidate into a tenure-track job this year.

So yes, competition for academic jobs is fierce. And a look at those who landed prestigious academic jobs can be intimidating. But to learn what matters for landing faculty positions, we also need to look at those who weren’t hired and see what predicts job offers. Two surveys of job market applicants in this past year’s cycle (2018-2019) did just this.

Image by Niek Verlaan from Pixabay

The first survey polled psychology job market applicants (N = 326; 64% had some academic offer) and found that, using common academic metrics, it was easier to predict landing an interview than to predict a job offer after landing an interview (Mehr & Pennycook, 2019). For example, sending out many applications and having more and well-cited papers was associated with landing an interview, but did not predict job offers. This pattern was mirrored in a second survey of mostly life science job market applicants (N = 317; 58% had some academic offer) (Fernandes & Sarabipour et al., 2019). Every academic job search has idiosyncrasies, and recent data suggest that collegiality and professionalism matter as much as anything at the interview stage (Boysen, Morton, & Nieves, 2019).

So while you of course need to be above a threshold to be considered for an interview – a bar that seems to rise each year – ultimately the fit of your research program as well as things like your interpersonal skills, teaching philosophy, and whims of a given year’s search committee, will determine whether you land the job.

Image by pasja1000 from Pixabay

The good news is that there are many types of academic jobs to consider. In just the US, there are over 4,000 institutions of higher education, including over 1,000 community colleges, over 1,400 baccalaureate and master’s colleges/universities, and over 400 research universities (in fact only 3% of US institutions are R1 universities).8 And some smaller institutions or non-tenure track positions may have better work-life balance, better departmental climate, and allow employees to feel more valued for their teaching and service.

The perfect fit may be at any one of these types of academic jobs. For example, a colleague of mine who loves teaching and who studies how minority students can overcome the race gap in academic achievement, landed a tenure track position at a community college. She told me: “Fit totally matters…working at an institution with a diverse population that focuses on teaching while still supporting scholarship was ideal for me… I think in the current market, we focus so heavily on just getting a job somewhere that fit is often overlooked – and a bad fit will only mean that you’re soon back on the market.”

As social psychologists Jay Van Bavel, Neil Lewis Jr., and Wil Cunningham wrote, to succeed on the academic job market you have to maximize the signal (i.e., have good materials tailored to each institution) and minimize the noise (i.e., apply as broadly as your situation allows, including to smaller universities or colleges as well as to neighboring disciplines or internationally). When considering where to apply, applicants should focus on job fit, weighing all of life’s factors.

Image by Evan Peck
Image by Evan Peck

For many who want an R1 job (and who may be used to a big research university setting), they should ready themselves to begin their career with a postdoc or a “starter home” faculty job (i.e., at a smaller or less reputable school) before seeking their “dream home” faculty job.

Most importantly, PhD programs and faculty should be telling this to students at the outset. My colleague who works at the community college told me she didn’t even consider community colleges until someone, outside of the department, mentioned it in the fall of her final year, which completely changed her job search. Annual career workshops or presentations, such as Van Bavel and Lewis Jr.’s presentation on “Demystifying the academic job market” should become commonplace to help set appropriate goals and expectations for aspiring academics.

Admittedly, knowing that there are various academic jobs doesn’t remove the inequity and uncertainty in the academic job market. Fortune favors those who can apply broadly and there is still a lack of secure jobs – one of academia’s biggest issues. There is a case to be made for restructuring academia to support the rising “permadoc” workforce (Milojević, Radicchi, & Walsh, 2018) and rejecting the pyramid model where everyone is the head of their own crowded lab.


Mentorship Matters

Whether you need 2 or 20 publications, the pressure to publish places a premium on finding supportive advisors at well-networked institutions who can steer your research in productive ways and navigate peer-review.

A good mentor can also help you keep your spirits lifted when you inevitably hit roadblocks or experience rejections. PhD students are more than six times as likely to experience depression and anxiety as the general population, and nearly half will leave their program before finishing.9 Among those who finish, the #2 reason given for making it through was mentorship and advising (#1 was financial support; and neither factor alone is sufficient).

In a survey of mostly life science job market applicants, many respondents indicated that poor mentorship was a hinderance in their faculty applications. Reproduced with permission from Fernandes & Sarabipour et al., 2019.

In my view, PIs have two main jobs: 1) recruit and mentor talented students who fit with the lab culture, 2) obtain funding to support the lab’s research. PIs of course need to publish regularly, and doing so helps them obtain funding and promotions, but such aims can be well aided by supporting their students and postdocs. 

PhD students and postdocs are often the real foot soldiers of science, as they are the ones designing experiments, collecting and analyzing data, and taking the lead on writing manuscripts. As such, good mentors recognize that their own success is inextricably linked to that of their students and put their students’ interests ahead of their own. They cultivate their students’ independent research and provide guidance, from conception of research questions to writing of manuscripts and grants. Good mentors are able to step into the minutia of an experiment and zoom back out to consider the broader impacts of the work within the field. Good mentors regularly check in with their students about their career goals and feelings, offering advice and support. 

But access to such advisors is a luxury only afforded to some, perpetuating inequality in the graduate school-to-professorship pipeline. A glance through any department would likely reveal a wide range of quality in advisors. These issues are compounded with broader graduate-school-to-professorship pipeline inequalities, such as the fact that just 25% of institutions produce 86% of all tenure-track faculty (Clauset, Arbesman, & Larremore,  2015) – a statistic not explained by meritocracy, and more likely reflecting factors such as social status.

And unfortunately, our system for hiring is mismatched. Data show that hiring committees for university professorships value a profusion of publications, ideally in high-impact journals (at least when deciding who to interview), but seem to forget that PIs also spend much of their life mentoring postdocs and PhD students. Being a prolific scientist may be a signal that you know how to develop a good program of research, but it does not guarantee you know how to run a lab.

Image by Felix Mittermeier from Pixabay

Worse still, graduate students don’t gain experience leading a lab during graduate school (and only acquire some mentoring experience of undergraduates or peers), which means two things: 1) it’s hard to judge an applicant on their mentorship potential during a hiring decision, and 2) many PIs begin with little to no experience of a crucial aspect of their job.

A better system might emphasize training for assistant professors so they can learn how to effectively manage a lab. Few institutions invest in managerial trainings for new assistant professor hires. Instead, junior faculty rely on colleagues or scattered advice from academic Twitter (#phdchat) about best practices for mentoring students or developing a lab manual.

The academic job market is competitive, and we should be filling the professorship ranks with good mentors, both for the sanity of their students and the future of academics.

Acknowledgements: 

Special thanks to Jasmin Sandelson, Jillian Jordan, Gordon Kraft-Todd, Philip Pärnamets, Elizabeth Harris, Claire Robertson, Steve Hall, Gordon Pennycook, Samuel Mehr, Jon Freeman, Tessa West, and Jay Van Bavel for their insightful edits and feedback.

Footnotes:

1 Gaining entrance to a PhD program is a feat itself. Applicants tend to be a highly selective group to begin with and even among such a qualified pool, acceptance rates hover around 10-20%. In fact less than 2% of the US population holds a PhD.

2 Evolutionary anthropologist Dorsa Amir offers fantastic advice for what to expect in graduate school and how to navigate the ups and downs. And physical chemist Martin Schwartz reminds us about “the importance of stupidity in scientific research”.

3 I took a weighted average between the correlation found by Pennycook & Thompson (2018), r(64) = .48, p = .001, and the correlation in my dataset, r(145) = .14, p = .091, to give an overall rough estimate.

4 I corresponded with a faculty member on the search committee at this institution to learn about this process. To respect the privacy of applicants and the department, I was asked to keep this anonymous.

5 If counting co-authored work, the average Skype finalist had 6 papers in “top journals”. As reference, one graduate student finalist had 4 first-author papers, 2 of which were in “top journals”, and the other graduate student finalist had 3 first-author papers, 1 of which was in a “top journal”.

6 Even these 1970 numbers are likely an overestimate, as the data are limited to established psychology departments with graduate training (listed by the National Council of Chairmen of Graduate Departments of Psychology), and do not count positions offered by undergraduate-only institutions or non NCCGDP institutions.

7 See NSF data on number of graduates and this helpful blog post with data on psychology job postings.

8 The remainder of US institutions are special focus institutions (e.g., technical professions) or tribal colleges.

9 See why some left their PhD and advice they would give to current or prospective doctoral students.

References:

Boysen, G. A., Morton, J., & Nieves, T. (2019). Kisses of Death in the Psychology Faculty Hiring Process. Teaching of Psychology46(3), 260–266.

Clauset, A., Arbesman, S., & Larremore, D. B. (2015). Systematic inequality and hierarchy in faculty hiring networks. Science advances1(1), e1400005.

Fernandes, J. D., Sarabipour, S., Smith, C. T., Niemi, N. M., Jadavji, N. M., Kozik, A. J., Holehouse, A. S., Pejaver, V., Symmons, O., Bisson Filho, A. W., Haage, A. (2019). Insights from a survey-based analysis of the academic job market. bioRxivhttps://www.biorxiv.org/content/10.1101/796466v1.

Milojević, S., Radicchi, F., & Walsh, J. P. (2018). Changing demographics of scientific careers: The rise of the temporary workforce. Proceedings of the National Academy of Sciences115(50), 12616-12623.

Pennycook, G., & Thompson, V. A. (2018). An analysis of the Canadian cognitive psychology job market (2006–2016). Canadian Journal of Experimental Psychology/Revue canadienne de psychologie expérimentale72(2), 71.

Mehr & Pennycook. (2019). Poll available here: tinyurl.com/psychjobmarket

Rose, R. M. (1972). Supply and demand for psychology PhDs in graduate departments of psychology: 1970 and 1971 compared. American Psychologist27(5), 415.

Schwartz, M. A. (2008). The importance of stupidity in scientific research. Journal of Cell Science121(11), 1771-1771.

Van Bavel, J.J., Lewis Jr., N. A., & Cunningham, W.A. (2019). In the tough academic job market, two principles can help you maximize your chances. Science.

Please sign in or register for FREE

If you are a registered user on Research Communities by Springer Nature, please sign in

Subscribe to the Topic

Society
Humanities and Social Sciences > Society
  • Nature Nature

    A weekly international journal publishing the finest peer-reviewed research in all fields of science and technology on the basis of its originality, importance, interdisciplinary interest, timeliness, accessibility, elegance and surprising conclusions.

  • Nature Human Behaviour Nature Human Behaviour

    Drawing from a broad spectrum of social, biological, health, and physical science disciplines, this journal publishes research of outstanding significance into any aspect of individual or collective human behaviour.