. . . And If You Can’t Teach, Teach Gym
| Peter Klein |
You know the old adage: If you can, do; if you can’t, teach. Is it true for business?
A paper in the August 2007 Academy of Management Perspectives, “Do Business School Professors Make Good Executive Managers?” by Bin Jiang and Patrick Murphy (full text; abstract; press release), identifies 217 firms with former business-school professors in management positions and finds that these firms have higher revenues-per-employee than a control group matched by industry, location, and firm size. Faculty making early exits from their academic careers appear to be the most valuable, while neither academic area nor business-school ranking seem to matter. Conclusion:
Executive managers learn from past experiences when they draw the right lessons from those experiences. But experience alone is not enough. Given the rigorous training professors receive in order to design research that objectively parses error and data, one final supposition is that they may be particularly competent at delineating patterns in complex management and organizational experiences. They may also be especially capable of continually developing innovative questions that lead to information useful for executive decision-making amidst uncertainty.
I enjoyed reading the paper. Certainly I like to think that I’d command a high salary if I chose to give up my cushy professor lifestyle for the real world. However, I don’t find the empirical analysis convincing. Here’s why:
The study does not control for two potential sources of selection bias. First, business-school professors choosing to go into industry may systematically select high-performing firms. For voluntary exits (i.e., excluding tenure denials), this is exactly what we would expect. Second, business professors who become executives are not a random sample of all business professors. Those with particular attributes (good judgment, charisma, leadership skill) are more likely to have extramural professional opportunities than the nerdier types. In other words, the results do not tell us whether these highly valued executives are highly valued because they used to be professors — as the authors imply — or in spite of it. (Or independent of it.)
How to improve the analysis? Some suggestions:
- Try to control for professor-specific characteristics in some kind of two-stage (e.g., Heckman) model.
- Use panel data with fixed effects to see how the performance of a given firm changes when an ex-professor is added to the executive team.
- Look for instruments for the firm’s decision to hire a former professor (i.e., variables uncorrelated with current or expected future firm performance). Proximity to a top business school, perhaps? A CEO with a degree from the ex-professor’s institution? There must be a good instrument out there somewhere. . . .