Recently the economics blogosphere has been discussing why firms are willing to pay significant amounts of money for the advice of recent college graduates at consultancy firms. The consensus seems to be that management uses consultancy firms to justify decisions they have already made.
This discussion got me thinking about the mismanagement of firms. I keep hearing, from various sources, about how poorly many firms are run in the real world. There is a whole field in economics called the theory of the firm taking on the issue of why firms exist, and why they cannot maximize profits (essentially, people working at firms don't like doing work--surprise!). Thinking about my I/O and trade research, I always take firms to be black boxes which maximize profits. In light of real world anecdotes and the theory of the firm, this assumption doesn't seem very attractive even as an approximation.
This problem runs a nice parallel with the insights of behavioral economics. Years of experiments, both in the laboratory and in the real world, have shown that people do not maximize well defined preferences. However, most of economics marches forward modeling individuals as perfectly rational utility maximizers. The defense is usually to quote Tolstoy: "Happy families are all alike, but unhappy families are each unhappy in their own way." There is a large body of work showing how to tractably model rational individuals, but it is not clear how to even begin to model irrational behavior. Sometimes it is better to have a result built on shaky foundations rather than no result at all (or is it?).
Industrial organization people sometimes defend their modeling assumptions by positing that while people may be irrational, irrational firms will be eliminated by a competitive market. The more I hear about firm management, the less I believe that argument. As a footnote, there is a neat, fairly recent literature being built which tries to measure the extent of misallocation of resources across firms. The results still depend heavily on strong assumptions about production functions and productivity, but I think this is an important direction to be moving in.