Opinion 167: The New Criteria for Promotion and Tenure, Based on the h-Index, Recently Adopted by the Consortium of Mathematics Departments, should be Applauded, but one should severely prosecute "h-index enhancement reciprocity agreements"

By Doron Zeilberger

Written: April 1(!) , 2018

Until recently, the decision whether to grant tenure and/or promotion was very subjective, and "noisy", to say the least. One requested six recommendation letters from "experts" and based on their assessments the departmental promotion and tenure (P&T) committee made its decision.

This is a very unfair process, since most of such letters are gross overstatements (by friends of the candidate), while some of them (by the candidate's rivals) err on the opposite direction. In other words, it is a mere popularity contest, that has nothing to do with the intrinsic merit of the candidate.

Luckily, now we have the h-index, that is an objective measure of a scientist's impact to his or her field. In the new policy one no longer bother "experts" with requests for letters of reference, forcing them to lie, or at least, grossly exaggerate (if the candidate is their friend), or giving them an opportunity to trash their competitors or rivals. In addition it is also a great waste of their time, that could have been used more fruitfully to do their own research.

The consortium's recommended cut-offs of the h-index for the various promotions are as follows.

This will save so much time, not only for the letter writers, but also for the P&T committee members, with their endless fruitless discussions that often result in unfair decisions. In fact, with this new policy, there would no longer be a need for T&P committees. The department chair will go to google scholar (or one of the other databases) and in one second figure out the h-index of the candidate, and decide right away whether or not to grant the promotion.

But there is one caveat, that one has to watch for. Like all objective measures, the h-index, too, can be tampered with. If you practice IQ tests all day, even a moron can become a genius, and if your parents are wealthy enough, they can pay for SAT tutoring that would raise your scholastic "aptitude" considerably, and the test becomes a measure of the parents' wealth.

One obvious way to raise your h-index is to cite yourself! But this is easily detectable. A more subtle way is to form "h-index enhancement reciprocity agreements". "If you cite me, I'll cite you". But don't cite just any random paper of mine, and definitely not my most cited one, this will not help me at all. Please look up (or compute) my current h-index, let's call it h, and cite my (h+1)-th most cited paper. [ In many cases this would immediately raise my h-index by one, (if ch > h and ch+1=h), otherwise you will need to cite it in several of your new articles.]

Let's make it clear by a concrete example.

Suppose that Tim Gowers asked you to help him upgrade his h-index. Whatever you do, do not cite this paper that would increase its number of citations from 733 (as of 4/1/18) to 734, but will not affect the h-index, but instead cite this article, that would raise Tim's h-index from 27 to 28. Now repeat!, and keep pumping it up. Im most cases, the cited paper has nothing to do with the citing paper, but one can always make-up a "plausible" connection.

I know of quite a few cases where friends had this "mutual h-index enhancement agreement", where they cite each other, making sure to cite the "critical paper". This is already pretty bad. But what is really shocking is that there is now a start-up company (whose name I can't reveal for obvious reasons) that, for a handsome fee, "matches" people who want to raise their h-index, and if you have many "pen-pals", before you know it, your h-index will get you the desired promotion.

To conclude, the new policy, to decide on promotion solely in terms of the h-index, is a good one, but the university ethics committee should carefully check that the above-mentioned algorithm has not been used, and each candidate would be subjected to a polygraph test.

Doron Zeilberger's Opinion's Table of Content

Doron Zeilberger's Homepage