It has become relatively common for our recent recruits into the Development Research Group at the World Bank to do a post-doc somewhere else for a year before joining. I wondered if this was part of a broader trend towards post-docs becoming increasing common in economics, and especially development economics, and so thought I’d look for some data, and then ask some people their experiences.
The NSF survey of earned doctorates in the U.S. breaks down the post-graduation commitments for those PhDs going to work in the U.S. The data for 2020 (Table 51) show that of the 693 econ PhDs going into employment in the U.S., 128 were doing post-docs (18%), 234 were going into academic employment, 194 industry, and 137 government/NGO/international institution or other jobs. So there are more than half as many post-docs as academic jobs. The public reports seem to only break social sciences down by subfield from 2014 onwards – in 2014, 83 out of 576 jobs (14%) were post-docs.
Looking at development economists, I took the sample of BREAD affiliates who had received their PhDs from 2010 or after. Of these 54 young development economists, 28 (52%) had done a post-doc. The most common form of post-doc for them was a one-year position, when they had already accepted a job that they would go onto after this post-doc, but there are also some doing two or three year post-docs. Post-docs are normally at a different institution, but some arrange to do a year as a post-doc at a Center or Institute at the university they will then join as an assistant professor.
Given this high prevalence of post-docs, I thought I would see what advice those who have recently done or are doing post-docs might have to offer to others considering a post-doc, as well as for institutions hosting them and for departments like ours, who will be receiving someone after a post-doc. I was delighted to receive thoughts from:
· Melanie Morten, Associate Professor at Stanford, who did a post-doc at the Federal Reserve Bank of Minneapolis
· Pierre Bachas, Economist in the Development Research Group, who did a post-doc at Princeton.
· Patrick Behrer, Economist in the Development Research Group, who did a post-doc at Stanford.
· Megan Lang, currently a post-doc with the King Climate Action Initiative, based at UC Santa Barbara, who will be joining the Development Research Group later this year.
· Gabriel Englander, currently a post-doc in the Environmental Markets Lab (emLab) at UC Santa Barbara, who will be joining the Development Research Group later this year.
What are the pros and cons for researchers to consider when deciding whether or not to do a post-doc?
Melanie notes there are two types of postdocs — postdocs that are lined up once the person has a faculty job already, and then postdocs that allow some time to work on research and then go on the market (either again or for the first time). The reasons for doing these two are somewhat different. Patrick gives a couple of reasons for considering the second type: having a long-term project that you can’t finish up in time for the PhD but that you think will make a much better job market paper; or potentially as a springboard to positions out of academia if you do a post-doc with a non-academic institution.
In terms of the more common one-year post-doc with a deferred job offer, the main pros to consider are:
· A delay in the tenure clock: this may be particularly important in development economics, especially for those setting up fieldwork that can take a long time. However, how much this delay helps depends on what the tenure standards of your institution are. Gabriel notes that this reason may not be as important as he ex ante thought, given the World Bank has a longer tenure clock and more transparent and attainable (hopefully) promotion guidelines than he thought he might be facing ex ante at a university.
· The chance to build your network and find new co-authors and learn new skills: Megan notes her biggest motivation for doing the postdoc were the people she wanted to have in her network, particularly those who she hoped would be future coauthors. “This was particularly important for me because I started working on energy and environment topics relatively late in grad school, so I thought that being somewhere with a strong focus in those areas would help me continue to "catch up". While this didn't involve formal coursework for me, I have heard of others who have used their postdoc to strengthen certain skills that they didn't fully develop in grad school. “
· Time with less responsibilities to get your research up and running: this is particularly relevant to those moving into academic positions, where prepping to teach new classes and teaching can take a lot of time in the first years, leaving less time for research. A post-doc can allow time to get the job market paper submitted and new projects underway. Patrick notes that without the pressure of the job market, it can be a pressure-free year to help you remember why you like doing research in the first place.
The main potential cons to consider are:
· Moving costs: moving twice in a year can be demanding, especially for those with families/partners. Melanie notes that an alternative can be more people to negotiate a post-doc with the institution you’re accepting the job with. This does not yield the new networking opportunities but gives some of the other benefits.
· Potentially lower income: post-docs will often pay less than the job you are going into. This guide estimates a typical econ post-doc salary is in the $60,000-65,000 range. However, this need not always be the case –apparently some post-docs pay is similar to starting salaries as an assistant professor.
· Potential isolation: Pierre notes that since one year is short, it limits the incentives of the host institution to invest in you, and since you are in between your cohort of PhD students and your new group of fellow faculty or workmates, it can create a bit of a sensation of floating around. Patrick also noted the possibility of ending up in a post-doc without much support or a system for meeting other people.
· Potential strings attached: some post-docs come with the expectation that you spend a portion of your time working on a specific project, which may leave less time for developing your own research agenda. For example, in Gabriel’s case “My offer said I would spend 25-50% of my time on "funded projects", which may or may not lead to an academic paper”
Advice on how to make the most of your post-doc
The researchers emphasized the importance of thinking very carefully about the goals you want to achieve, and emphasized that a year is not very long and can pass very quickly. Their main advice on how to make the most of this time is:
Consider networking an important part of your job and work hard at it. Gabriel notes “I'm glad so many people told me to network during my postdoc. I did very little networking in grad school because I'm shy, but knowing that networking was "part of my job" this year has helped me force myself to actually do it.”. Patrick adds “Be aggressive in trying to meet people at your new institution. I think doing mine during COVID both helped and hurt in this respect. On one hand it reduced the natural interaction that happens when everyone is attending the same seminars. On the other I think it made people more amenable to brief zoom coffee chats. Regardless, you need to be very proactive in meeting people if you want to build your network/collaborations at your new post-doc institution. Get coffee with people, talk to them about their research, talk to them about yours, but make sure you get out of your office and talk with people.” If you want to start new collaborations, identify early who you might want to work with.
Don’t travel too much: Melanie notes that giving lots of talks is good for exposure, but can eat up a lot of time. Travelling a lot also means you miss out on being part of the community at your hosting institution and may reduce the chances for some of the attending workshops/coffees/other networking opportunities.
Don’t set too many goals: Megan said that one very useful piece of advice multiple people have given her is to set a small number of realistic goals (e.g., get your JMP submitted, start one new project).
Think longer term, and figure out where the gains are highest from being physically present: Patrick said that while there is often an expectation that people use the post-doc to finish and send out their job-market paper, this may not necessarily be the best use of your time longer term. “My strategy was to try and start a lot of collaborative papers that I knew I wouldn’t finish during the post-doc but that I would keep working on once I left. I think it’s easier to start things with people when you’re in the same place and then you can finish those at distance. But it’s harder to start something once you’ve left so working on many things in parallel (and focusing on starting new things rather than finishing things you’re coming into the post-doc with) may help you ultimately get more out of the post-doc. So far this approach has been successful for me. Of the papers I started in the post-doc only one had been submitted by the time I finished but since I left 3 more have been sent in.” Megan suggests that many of the returns to a postdoc probably will not materialize during the actual postdoc, so it's worth thinking long-term when you're thinking about what you want to get out of it.
Quickly build a support system to help get feedback and avoid isolation: Megan offers several components of this. “Part of that support system is hopefully a senior mentor (probably the person sponsoring your postdoc) who is willing to invest in you, which is probably obvious. The other piece of it is finding other postdocs or early career economists so that you don't waste time struggling alone. For me, this looks like a mix of other postdocs at my institution, some early career faculty, and people from my PhD cohort who are willing to do things like troubleshoot econometrics questions, read abstracts/intros for conference submissions, exchange papers, sanity check new ideas, etc.”.
This last point applies even for those not doing post-docs. Melanie notes “In general, I think it’s really helpful to get good mentoring when you’re starting as a junior faculty member. An organization that I found helpful is the National Center for Faculty and Diversity who run a 10 week academic bootcamp (https://www.facultydiversity.org/). Most tenure track jobs are not long — the length of a PhD — and you need to finish a substantial number of projects in time. You need to have a clear plan for which projects to work on and prioritize to get it all done in the length of time on your tenure clock. For example, if projects will need fieldwork, you only have the first few years to get these started if they’re realistically going to be published in time for tenure. Looking back, a postdoc year is a very valuable year that you may not use fully if you haven’t planned out what you need to get done.” CSWEP also run a junior faculty mentoring course aimed at mentoring women and nonbinary faculty.
What can host institutions do to help make the post-doc experience work well?
The most important advice/request researchers had was for hosting institutions to try to actively integrate them into the community as early and as much as possible. This includes making sure they are on the mailing lists for seminars and get offered opportunities to present research early on. This can work well if there are more informal lunch talks or brown bags that often go unfilled early in the semester, where post-docs can be keen to introduce themselves and their work to others. Having professors and grad students set up coffees or lunches, and inviting post-docs to lunches or dinners with seminar speakers helps a lot. It is also useful to have a central admin point person that post-docs can go to for questions like dealing with reimbursements, computing, etc.
Patrick noted that this experience of being closely incorporated into a community worked particularly well in his case, because he went to work with people who run research labs that are organized more like the hard sciences with a dedicated lab group that included a community of part-time undergraduate Ras, full-time Ras between a BA and post-grad, PhD students, post-docs, and full-time researchers. These labs provided feedback on work and a source of collaborators, as well as regular meetings to discuss work in progress and share ideas. These types of large research labs are becoming more common in economics as well, so might offer good opportunities for post-docs going forward.
What (if anything) should institutions like the World Bank do to help people who are joining after deferring their offers for a post-doc first?
The main advice is for a fairly light touch, mostly leaving those doing the post-docs alone to focus on their research there. There was some appreciation for helping with some connections – e.g. if the post-doc is in the same geographic region, including the person on the seminar list or perhaps a conference you are organizing. As the start date approaches, inviting the post-doc to a couple of team meetings virtually so they can have some sense of what is going on, and be ready to hit the ground running was also appreciated.
Revealed preference shows that more and more development economists are finding post-docs to be, on balance, something they expect to be useful to them. Of course the responses above reflect those of people who chose to do post-docs, and not those who turned them down. Moreover, the post-docs the researchers I talked with did were all at excellent institutions which perhaps accentuated these benefits. Nevertheless, hopefully the advice above is helpful for those deciding on a post-doc, and can help those about to embark on one to make it productive.
But while post-docs offer individual benefits, there are potential downsides for the profession as a whole. There is already much hand-wringing about the rise of pre-docs, which coupled with a 6-year PhD, and then a post-doc, can really stretch out the start of a career. I would not like to see economics become too much like the hard sciences, where to be competitive on the job market, people typically need to have done one, or maybe more, post-docs, and where the funding model and production function for research depends on cheap post-doc labor. Post-docs as a chance to learn new skills, experience living in a different place, and form new collaborations seems great to me - but if the main purpose of the post-doc is to simply extend the tenure clock and make people more competitive on the job market, there is a risk of this becoming a less socially optimal arms race.