As a professor, I have had a lot of personal experience with this particular form of overmotivation. Early in my academic career, public speaking was difficult for me. During one early presentation at a professional conference in front of many of my professors, I shook so badly that every time I used the laser pointer to emphasize a particular line on a projected slide, it raced all over the large screen and created a very interesting light show. Of course, that just made the problem worse and, as a result, I learned to make do without a laser pointer. Over time and with a lot of experience, I became better at public speaking, and my performance doesn’t suffer as much these days.
Despite years of relatively problem-free public speaking, I recently had an experience where the social pressure was so high that I flubbed a talk at a large conference in front of many of my colleagues. During one session at a conference in Florida, three colleagues and I were going to present our recent work on adaptation, the process through which people become accustomed to new circumstances (you’ll read more about this phenomenon in chapter 6, “On Adaptation”). I had carried out some studies in this area, but instead of talking about my research findings, I planned to give a fifteen-minute talk about my personal experience in adapting to my physical injuries and present some of the lessons I had learned. I practiced this talk a few times, so I knew what I was going to say. Aside from the fact that the topic was more personal than is usual in an academic presentation, I did not feel that the talk was that much different from others I have given over the years. As it turned out, the plan did not match the reality in the slightest.
I started the lecture very calmly by describing my talk’s objective, but, to my horror, the moment I started describing my experience in the hospital, I teared up. Then I found myself unable to speak. Avoiding eye contact with the audience, I tried to compose myself as I walked from one side of the room to the other for a minute or so. I tried again but I could not talk. After some more pacing and another attempt to talk, I was still unable to talk without crying.
It was clear to me that the presence of the audience had amplified my emotional memory. So I decided to switch to an impersonal discussion of my research. That approach worked fine, and I finished my presentation. But it left me with a very strong impression about my own inability to predict the effects of my own emotions, when combined with stress, on my ability to perform.
WITH MY PUBLIC failure in mind, Nina, Uri, George, and I created yet another version of our experiments. This time, we wanted to see what would happen when we injected an element of social pressure into the experimental mix.
In each session of this experiment, we presented eight students at the University of Chicago with thirteen sets of three anagrams, and paid them for each of the anagrams they solved. As an example, try to rearrange the letters of the following meaningless words to form meaningful ones (do this before you look at the footnote*):
1. SUHOE
Your solution:
2. TAUDI
Your solution:
3. GANMAAR
Your solution:
In eight of the thirteen trials, participants solved their anagrams working alone in private cubicles. In the other five trials, they were instructed to stand up, walk to the front of the room, and try to solve the anagrams on a large blackboard in plain view of the other participants. In these public trials, performing well on the anagrams was more important, since the participants would not only receive the payment for their performance (as in the private trials) but would also stand to reap some social rewards in the form of the admiration of their peers (or be humiliated if they failed in front of everyone). Would they solve more anagrams in public—when their performance mattered more—or in private, when there was no social motivation to do well? As you’ve probably guessed, the participants solved about twice as many anagrams in private as in public.
THE PSYCHOANALYST AND concentration camp survivor Viktor Frankl described a related example of choking under social pressure. In Man’s Search for Meaning, Frankl wrote about a patient with a persistent stutter who, try as he might, could not rid himself of it. In fact, the only time the poor fellow had been free of his speech problem was once when he was twelve years old. In that instance, the conductor of a streetcar had caught the boy riding without a ticket. Hoping the conductor would pity him for his stutter and let him off, the boy tried to stutter—but since he did not have any incentive to speak without stuttering, he was unable to do it! In a related example, Frankl describes a patient with a fear of perspiring: “Whenever he expected an outbreak of perspiration, this anticipatory anxiety was enough to precipitate excessive sweating.” In other words, the patient’s high social motivation to be sweat-free ironically led to more perspiration or, in economic terms, to lower performance.
In case you’re wondering, choking under social pressure is not limited to humans. A variety of our animal friends have been put to similar tests, including no one’s favorite—the cockroach—who starred in one particularly interesting study. In 1969, Robert Zajonc, Alexander Heingartner, and Edward Herman wanted to compare the speed at which roaches would accomplish different tasks under two conditions. In one, they were alone and without any company. In the other, they had an audience in the form of a fellow roach. In the “social” case, the other roach watched the runner through a Plexiglas window that allowed the two creatures to see and smell each other but that did not allow any direct contact.
One task that the cockroaches performed was relatively easy: the roach had to run down a straight corridor. The other, more difficult task required the roach to navigate a somewhat complex maze. As you might expect (assuming you have expectations about roaches), the insects performed the simpler runway task much more quickly when another roach was observing them. The presence of another roach increased their motivation, and, as a consequence, they did better. However, in the more complex maze task, they struggled to navigate their way in the presence of an audience and did much worse than when they performed the same complex task alone. So much for the benefits of social pressure.
I don’t suppose that the knowledge of shared performance anxiety will endear roaches to you, but it does demonstrate the general ways in which high motivation to perform well can backfire (and it may also point to some important similarities between humans and roaches). As it turns out, overmotivation to perform well can stem from electrical shocks, from high payments, or from social pressures, and in all these cases humans and nonhumans alike seem to perform worse when it is in their best interest to truly outdo themselves.
Where Do We Go from Here?
These findings make it clear that figuring out the optimal level of rewards and incentives is not easy. I do believe that the inverse-U relationship originally suggested by Yerkes and Dodson generally holds, but obviously there are additional forces that could make a difference in performance. These include the characteristics of the task (how easy or difficult it is), the characteristics of the individual (how easily they become stressed), and characteristics related to the individual’s experience with the task (how much practice a person has had with this task and how much effort they need to put into it). Either way, we know two things: it’s difficult to create the optimal incentive structure for people, and higher incentives don’t always lead to the highest performance.
I want to be clear that these findings don’t mean that we should stop paying people for their work and contributions. But they do mean that the way we pay people can have powerful unintended consequences. When corporate HR departments design compensation plans, they usually have two goals: to attract the right people for the job and to motivate them to do the best they can. There is no question that these two objectives are important and that salaries (in addition to benefits,