Should PhD programmes require candidates to have at least one paper published in order to graduate?

Many PhD programmes require PhD candidates to have one (or more) papers accepted or published in a scientific journal in order to be able to graduate and complete their doctorate. This is of course a well-intentioned idea which might bring some benefits. First, PhD candidates would gain experience and start learning how to navigate the publication process in scientific journals. Second, having a publication out of the PhD may also help them secure their next position. Third, the publication requirement would also encourage a feasible PhD project plan, and that research results are published and not lost forever, particularly if the new PhD moves on after completing their doctorate. Finally, having a paper accepted at a scientific journal usually means that at least three people (editors and reviewers) have evaluated the research and the editor(s) have considered it good enough for publication, so it could be interpreted as an external validation and signal of quality.

Undesired consequences of requiring publication in PhD programmes

So far so good. But on second thoughts one can foresee how this publication requirement could create perverse incentives, bring many problems for early career researchers and science in general, and basically do more harm than good. Particularly considering that many PhD programmes last 3-4 years (at least in Europe) and the publication process can take well more than a year from submission to acceptance (more on this below). Thus, a few counter-arguments that come to mind:

1. A publication requirement may add significant mental and financial stress to finishing PhD students who may already be unemployed and tired but need to get an acceptance letter in order to be able to complete their doctorate. The mental health of PhD students seems very worrying in many countries (e.g. Belgium, Spain, Germany, Sweden), so we must work hard to improve it, and be very careful not to worsen it. The publication process can take well over a year from submission to acceptance ( depending on the journal), and in my humble opinion it is totally unfair to put PhD candidates on hold until journals are able to complete the review process and take an editorial decision, as that process is totally beyond the control of the PhD candidate. I can imagine how terrible it can be to receive a rejection from a journal after a 6-month (or longer) review process and having to start a new submission elsewhere while the clock keeps ticking…

2. The PhD candidate cannot control the speed of the review process at a journal, but can (generally) choose where to submit the manuscript. Thus, the publication requirement will clearly incentivise submission and publication in fast journals, sometimes with dubious review processes, thus contributing to the strain and drain of scientific publishing. Many PhD candidates with little time ahead will simply not allow themselves to risk one (or more) rejections or resubmissions in demanding journals that care for quality, and will find a faster/easier way to fulfil the publication requirement (cf. Goodhart’s law). As a result, due to the requirement, the PhD candidate may move away from rigorous, honest journals that may be more suitable for their research and their careers, and those honest journals may also suffer from unfair competition by (too) fast journals of lower quality. Some PhD programmes require that papers are published in particular journals (e.g. in the first quartile) which often will increase the publication delay (e.g. due to rejections) and introduces another set of undesired problems. Overall, the role of journals in the future of science is somewhat uncertain, and we should be very careful to avoid incentivising publication in low quality, predatory journals.

3. The publication requirement may incentivise PhD candidates (and their supervisors) to design faster, simpler, less ambitious research projects that allow having a manuscript ready within a short period of time. Having realistic, feasible, well-designed PhD projects with contingency plans is of course a nice thing! But, ambitious, bold, risky, innovative, exploratory research projects may be discouraged or penalised over more conventional, incremental, safer projects (which are totally fine, but it would be a shame to penalise the former projects too). For example, doing field or lab research, which usually takes a good portion of the first 2-3 years of the PhD journey, may be discouraged over embarking on ongoing projects and analysing readily available data. Furthermore, the publication requirement may encourage PhD candidates to publish incomplete, immature, unreliable research (e.g. lacking enough replicates, proper analysis, code review, etc). Therefore, the publication requirement could be creating undesired incentives again.

What are the benefits of requiring having a paper published anyway?

Here I revisit some of the supposed benefits of requiring publication for PhD completion, mentioned at the beginning of this post.

First, having a paper published in an indexed scientific journal may be considered a signal of quality, as the PhD research would have received the approval of at least one external, hopefully independent, editor. But, given the abundance of dubious journals and the ongoing debilitation of peer review, can we still consider that having a paper published is a valid signal of quality as a scientist? Probably not, at least not in a straightforward manner. One can get very bad papers published in any one of thousands of journals often avid to take your money for it. And requiring publication in journals with certain impact factor is not the solution.

Second, requiring publication avoids the research being lost forever if the PhD thesis is never published afterwards. Ok, avoiding such loss sounds nice, but rather than requiring publication in a journal, which introduces the unfair stress, delays, and uncertainties outlined above, why not archive publicly the PhD thesis (as many universities already do), or require publication of at least a preprint? That would prove that the PhD candidate is able to write a coherent, well-developed research piece good enough to be shared with the world forever with their name on it, but without introducing the stress and delay of waiting for the review process at a journal.

I am not saying requiring publication is necessarily a bad thing! Publishing during the PhD can bring many benefits, as outlined at the beginning of this post. But my impression is that good PhD candidates (who may also count with good supervisors) will publish good papers anyway without it being required, and other candidates facing difficulties will find the way to have a paper published somewhere somehow.

So, having a paper published might be considered a nice thing to have, but perhaps not a requirement? I don’t know. While the publication requirement has some merit, the drawbacks could outweigh the benefits, especially when considering the stress it places on candidates and the incentives it creates for low-quality publications.

What else could we do then?

If we are worried about the quality of PhD candidates and want to ensure that they are mature enough and have learnt how to develop a sound research project building upon current knowledge and looking ahead, maybe having a paper published is not the best requirement we should make? Maybe we need to think of other quality criteria (e.g. ability to synthesise the scientific literature and build from it, to think critically and come up with interesting research questions, to design experiments or research projects, to write coherently and backed on scientific knowledge, to develop rigorous and reproducible research, etc). Maybe we need to read the actual research done (i.e. the thesis), without having to wait for months to hear from an editor. In many countries there is an external committee of experts who read the thesis thoroughly and critically and can ask as many questions as they consider to the PhD candidate, and then decide if the performed research is good enough to deserve a doctorate degree. Even better if such committee can provide advice and guidance during the development of the PhD work.

In the end, I always come back to Doug Altman’s aphorism, which is the motto of the MAYBE_lab: ‘We need less research, better research, and research done for the right reasons’.