No Country for Old Men: The Bowden Saga
Due to travels, I missed writing about the Bobby Bowden affair during the height of the storm. The dismay among FSU fans is strong. In an Orlando Sentinel blog poll, 42% thought it time for Bowden to go — although a minority and not necessarily a good indicator of the precise number of such fans, it is still a fairly strong signal of the uneasiness with the situation.
The topic of age and coaching performance fits into the realm of an article by my former colleague, Tom Wisley and myself, on Is There a Managerial Life Cycle?. Using NFL data from 1920-2004, we found strong evidence of improvements with age and then a gradual decline in performance that mimics (with a 10 year lag) the decline seen in athletic performance as found by Ray Fair. Skip and his coauthor, Tom Goodwin, find similar effects in academic research. In Chapter 8 of From the Ballfield to the Boardroom, I find similar results look specifically at long-tenured coaches.
For NFL coaches, by their mid 60s, the gains of the earlier years are completely offset by the decline. Coaches who stay on beyond this point perform, as a group, very poorly relative to earlier years. Of course, these are averages — any given individual coach may perform considerably better or worse than the averages. The effect may be less at the college level because of less demanding strategic abilities or greater because of the need to recruit players.
I applied these methods to the careers of the leading octogenarians in college coaching, Joe Paterno and Bobby Bowden, while also taking account of their switch from independents to conference members. For Bowden, predicted performance began declining by age 62. By his early 70s, this decline put him below his predicted performance at the outset of his FSU years. Now the model predicts his performance to fall below 50 percent wins. For Paterno, the evidence is mixed. Using all Penn State years, there is no discernible age effect. Although, if the sample were truncated in 2004, there is a strong age effect detectable. Somehow, PSU has overcome the usual course of performance with an aging coach.
The turmoil created by the political economy of these kinds of situations is readily apparent in the current Bowden saga. How do you unload a legendary coach? You either appear as hearltess winning-only fools, played up by the media, or you let your program slide, maybe irreparably into mediocrity. Jerry Jones barely survived his first couple of years in Dallas after firing legendary Tom Landry.
In the era without mandatory age requirements, such a situation was bound to crop up. Maybe, mandatory age retirements were an institutionalized means of avoiding such prickly situations. Such forced retirements are a blunt tool, lumping workers with diverse skills together, but they avoid the unseemly task of asking someone who has been a productive worker but is in obvious productivity decline to step aside. (Edward Lazear offered an alternative to this kind of productivity based answer for mandatory retirements back in a 1979 article).
With mandatory retirements now largely obsolete and illegal, another mechanism to deal with the likely productivity decline is to set up contracts well in advance where a retirement age is agreed upon. If FSU had in Bowden’s mid-60s agreed that he would retire by 75, and they could have avoided this “prisoner’s dilemma” that they now face and that was inevitable if Bowden’s performance declined but he wanted to stay on anyway.