Patents

Some of the research labs are part of companies that patent very aggressively. Well, they are internal economies, and incentives aside, they set up an environment at their workplace which can be quite convincing that patents are important for their employer. Researchers have worked out several stories in their head that it is indeed true (the licensing fee from a few slam-dunk patents will provide revenue for supporting a lot of researchers; if my company doesn't patent it, the competitor would and then charge a licensing fee; company A meets company B with a portfolio full of patents, both avoid shootout by not opening the portfolio for scrutiny, and simply shake hands; etc.) I am a sceptic: generally, it seems USPTO relies on the courts to work out the worth of the patents in cases where there is a real need, and most patents filed by corporate lab researchers ferment in the portfolio bags.

The question that interests me is patenting in universities, where in general, resources are scarce. Some universities are aggressive about protecting their faculty research via patents; this is presumably because they have fat budgets from prior licensing successes. Others less so, and they let individual faculty find the resources to file patents and rely on the fact that the faculty will do it only for compelling cases (eg there is a startup company in the works). This provides some sort a natural barrier and foments selective patenting. Does this work out well in practice, or are their inefficiencies?