Strategic issues when hiring professors

I have always wondered what makes a good hiring strategy for professors at a computer science department. Here are my thoughts on two dimensions:

• Core strength or best you can get?
• Are locals better than non-locals?

Core strength or best you can get?

A common strategy is to double down on your strengths and hire more of the same. For example, at my engineering faculty, there are 35 professors of computer science, 12 of which are specializing in machine learning for medical engineering. But when is enough enough?

An alternative strategy is to hire the best you can get, without requiring a particular research focus, and see where your hire will take you. These are so-called open calls for professorship applications; not very common in Germany, but on the rise. The wisdom is that capable people can trump hoped-for synergies from focusing on your core strengths.

There is probably a healthy balance to be found, but for now I’m camp “hire the best you can get” rather than “more of the same”.

Are locals better than non-locals?

Another difficult question is whether it is better to hire people with local ties or not. There is a rule in Germany according to which you can’t hire postdocs from your own institution directly into a professorship position (“Hausberufung”), but as far as I can tell this rule has no real bite as there are too many ways to work around.

People with local ties know how things are done around wherever local is in this scenario. They are also committed to the location as they are unlikely to leave. On the other hand, given that they may have arrived where they wanted to end up, you might have hired someone into early retirement. At least they have no reason to hold up a research record because they want to go elsewhere.

People without local ties will find it easier to go elsewhere and, given that their local ties are elsewhere, might just be trying to do that. For this, they need to keep up a good research record, which in turn benefits the hiring department, at least as long as they are around. And if they never manage to leave, the department locked in that performance.

Here, I’m unsure; as always, it depends on the specific people. I’d tend to favor non-locals as it prevents inbreeding and cronyism.