Back in graduate school, I read an excellent paper by Barry Staw and Ha Hoang titled, "Sunk Costs in the NBA: Why Draft Order Affects Playing Time and Survival in Professional Basketball." The authors found that NBA teams fell into the sunk cost trap when it came to high draft picks. If the team expended a highly valuable first round pick on a player, they tended to stick with that player far longer than they should have. The NBA teams gave those players more playing time and retained the player longer than other players of similar ability who were drafted in later rounds. At times, coaches and general managers think to themselves, "We can't put a high first round pick on the bench." It's irrational, of course. They shouldn't be worried about sunk costs. However, we know that human beings have a hard time letting bygones be bygones. In fact, we throw good money after bad. That's precisely what happened to NBA coaches and general managers in this study.
As a graduate student in the late 1990s, I decided to see if the same effect occurred in the National Football League. Since tonight is the NFL draft, I thought that I would look back what I discovered in that paper. Note that this dataset is from a very different NFL era. For a variety of important reasons, I looked back at NFL drafts in the 1980s, prior to free agency in the NFL. What I found was slightly different than the NBA study. NFL teams did fall into the sunk cost trap in the first few years after drafting running backs out of college. That was the same as the NBA. However, I found that teams disproportionately disposed of high draft picks starting around year 4. In short, there seemed to be a reversal of the sunk cost effect. Why? It turns out that historical data clearly showed that NFL running backs had a very short career, on average. They peaked in Year 4, on average. Almost no one became a star running back for the first time (defined by having their first 1,000 yard season) after Year 4. (see chart above) So what? Well, teams seemed to be quite aware of this history, and thus, they started dumping high draft picks around Year 4 who had not performed as stars in their first few years. In fact, they dumped them at a higher rate than lower round draft picks with similar performance to date. I described this as a reversal of the sunk cost trap.
In other words, the direction of the sunk cost effect may actually change over time. Traditionally, sunk cost research has demonstrated that individuals fail to ignore prior investments when deciding future courses of action. Consistent with prior research, my examination of NFL running backs demonstrated that prior investments can lead to additional investments in the future, and to an escalation of commitment to a course of action. However, the analysis also demonstrated that sunk costs may serve as inducement for divestment, rather than additional investment, as an asset matures. In short, a large prior investment may lead to the de-escalation of commitment in certain situations.
My conclusions focused on the importance of beliefs and expectations. When we throw good money after bad, we do so because we have a strong belief that a little more investment can somehow yield a better outcome. What happens, though, if that belief is strongly contradicted by well-known historical data? Perhaps the sunk cost effect wanes or even reverses itself.
I concluded that sunk costs begin to become an inducement for divestment, rather than continued utilization, when individuals can no longer sustain the belief that this additional investment and/or utilization will yield improvement in the future. My analysis demonstrated that historical data regarding asset improvement appears to play a significant role in shaping these beliefs. In the case of NFL players, the historical data strongly suggests that players are not likely to improve after Year 4 in their career. Therefore, as assets mature, it appears that sunk costs can become an inducement for divestment if people no longer believe that improvement is possible, and if they believe that the assets have underperformed the expectations typically associated with their acquisition cost.
It's the hardest decisions often have to make in a variety of fields. When do I cut my losses? Perhaps there are a few lessons to learn from the mistakes made by NFL coaches and general managers in all too many situations.
No comments:
Post a Comment