How relevant are (journal) papers for the valuation of one’s work and results?

Question

I am a PhD student at an engineering department in Europe. At my department, and at most others in my university, research papers are the main benchmark of a scientist’s productivity. It seems like this is also true in other institutions.

For instance, last year the faculty considered the number of journal papers for each department a key performance indicator, and when a PhD student is at the end of his/her studies, he/she is often judged by colleagues based on how many papers they have published.

There is no threshold of a number of papers required in order to obtain a PhD or a senior position, and conference papers are also relevant and encouraged, but counting journal papers is the key point. In the end, a PhD thesis is just the sum of published papers during the PhD studies, and itself is not that relevant.

Naturally, I feel pressured to publish and to have lots of papers to support my upcoming degree. It seems to make sense - if you have a result, why wouldn't you publish it in a journal? Why wait for the thesis, when it's not really published and nobody will read it? And after all, submitting research to journals and conferences is also beneficial for getting a review of the research from someone who is outside one's department.

However, recently I realized that a friend of mine, also an engineer in a related field, received a PhD from another university with 0 journal papers, and just 1 conference paper where she was a second author. Her university is well regarded, and on almost all world rankings in the top 100. She was accepted for a post-doc position at another top university. I have seen a few other similar cases, and other people just not caring about papers.

I cannot judge the work presented in her thesis, but I found this shocking, and it made me reconsider. How can one be well regarded with hardly any publications? Why would anyone choose not to publish (not counting the thesis as the bare minimum)?

Hence my questions: How relevant are (journal) papers in terms of valuation of a fresh PhD, and for their future career?

EDIT

I would like to thank everyone for their reply, especially @Superbest for the edit (sorry, I am not a native English speaker), I didn't expect that my question would generate such a discussion. I'd like to clarify some things:

• In my field conference papers are less relevant than journal papers, and are used to present research in progress in order to get comments and advices.
• The graduate didn't publish a revolutionary paper which is worth like 10 other average papers. We are talking about a zero. I see no reason why would anyone choose NOT to publish at least one paper about a PhD.
• Even if the graduate had very good recommendations, why would anyone in academia, which depends on publications, hire someone without any publication record? Why would a professor think that the graduate would suddenly start to be productive?

Answer 1

1. How much people rely on your publication record to judge you depends on how much they understand your research. The further they are removed from your area of study, the more they will depend on "the numbers". People who work in your subfield will base their opinion on the actual quality of your work and on your reputation in the community. Fortunately, the most important evaluations (i.e., reference letters) usually come from people in your subfield.

2. Not all publications are created equal. Let's say I'm hiring a postdoc. Candidate A has 10 publications with almost identical titles all in low-quality journals. Candidate B has one publication that describes a significant breakthrough and is published in a good journal (I don't mean Nature or Science, just a journal that is widely read by people in the field; e.g. (in my field) SINUM or SISC). Who would I hire? Almost certainly candidate B. Indeed, Candidate A's publication record may actually be a negative factor.

3. Publications matter most in academia. A mathematician friend of mine took a job at Google after graduation. They didn't really care about his publication record; they hired him because he had created the Cython language, which was purely incidental to his research. If you go the academic route, your publications will be a critical factor in your evaluation for getting a post-doc, a tenure-track job, and tenure. If you go to industry, they may not matter so much.

4. Publishing norms vary by field. In mathematics, for instance, people may sit on a result for years while they refine it or try to develop it further. Senior people may just put their papers on the arXiv. In other fields, it seems like new developments must be published within a week, and you're expected to write vast numbers of papers each year.

Answer 2

Published papers (either as journal or conference papers, depending on your field) are an important metric in measuring a candidate's output. However, one important thing to note is that many graduate students publish a substantial fraction of their papers close to the end of their degree programs, so that their publication record may look quite sparse. More importantly, the publications may not be published in time for them to actually graduate, but may be finished up in the period immediately afterward. For instance, I wrote six papers as a graduate student, but only two were published while I was a student—the other four appeared the year after my degree was awarded. This is an entirely normal state of affairs.

Another thing worth noting is that, at least in engineering (and particularly engineering in Europe), a substantial fraction of projects are financed by industrial contracts. In such cases, the primary "fruits" of the project represent proprietary information which cannot easily be published. Under such circumstances, publications become quite difficult, and the quality of the work becomes much more significant.

Finally, while publications are important, at the level of a postdoctoral position, they are equally as important as the interview and letters of recommendations from colleagues. Somebody who comes to me highly recommended from people I know well and trust are much more likely to get the position than someone with the same credentials and quality of recommendations from people I don't know as well.

Answer 3

I generally agree with the other answers so far. I will just add that as a PhD student myself my supervisor put it like this:

Its not strictly necessary to publish many papers as a student but it makes writing and defending your thesis much easier. Your thesis must show some novel work. If you have some peer reviewed papers then this work has clearly already been accepted as novel, which basically ticks that box. Secondly you base your thesis on your papers.

Indeed I know some people whose thesis is basically 3-5 papers stapled together with an introduction and conclusion and maybe a bit of editing to make it flow better.