The world may or may not have been waiting for this, but there's now some theoretical support for the Peter Principle. What relevance does this have to the pharma-biotech industry, you ask? Well, actually, you probably don't ask, because you know just the sort of thing I'm talking about. If you've spent any time in any sort of large organization, you've seen what looks like empirical proof of the Peter Principle already - actually, you may already be picturing specific examples and muttering to yourself.
The classic R&D form of the phenomenon is someone who's capable of doing good research, but just terrible at managing people. You don't have to go very far up the hierarchy to see this one. Sad to say, there are quite a few scientists who reach their "level of incompetence" (to put it in Peterian terms) as soon as they get their first direct report under them. People skills are often not necessary to get through graduate school - in some research groups, they might actually be a handicap - so not every fresh PhD is equipped with managerial skills, to put it mildly. (This topic came up around here a few months ago, in a discussion of whether you want a scientist as a CEO in this business or not).
And the problem, in research as in everywhere else. is that educating a bad manager out of being bad is difficult at best, and impossible at worst. For one thing, a substantial number of poor managers have no idea, no idea at all, that anything might be amiss on their end. And the very deficiencies that keep them from realizing this also help to make them more impervious to attempts to change it. There's empirical support for this, too - often, the first thing that incompetent people are bad at is estimating their own competence.
Now that theorists are reproducing the Peter effects in model systems, that brings up the logical next question: can this help us do anything about the problem? The authors have some suggestions, but I don't see them being implemented any time soon. That's because the Peter Principle, if it's really true, necessarily implies that you should resist the temptation to always promote your best people:
We summarize in Table 1 the percentages of gain or loss obtained for the different strategies applied. These results confirm that, within a game theory-like approach, if one does not know what way of competence transmission is acting in a given organization, as usually one has in the majority of the typical situations, the best promotion strategies seem to be that of choosing a member at random or, at least, that of choosing alternatively, in a random sequence, the best or the worst members. This result is quite unexpected and counterintuitive, since the common sense tendency would be that of promoting always the best member, a choice that, if the Peter hypothesis holds, turns out to be completely wrong.
Try getting that one past the HR department!