This article was taken from the February 2011 issue of Wired magazine. Be the first to read Wired's articles in print before they're posted online, and get your hands on loads of additional content by subscribing online.
Nearly everyone cheats a little bit, when incentives point them in that direction
Part of my work involves making things difficult for people. By that I mean I create conflicts of interest and see how people respond. Imagine, for example, that you're in an experiment where you are given a sheet with 20 matrix problems such as the one below; the way you solve each of them is to figure out which two numbers in the set add up to ten:
<table border="0" style="text-align: left; width: 450px; height: 200px;">
<tbody>
<tr>
<td>1.69</td>
<td>1.82</td>
<td>2.91</td>
</tr>
<tr>
<td>4.87</td>
<td>4.81</td>
<td>3.05</td>
</tr>
<tr>
<td>5.82</td>
<td>5.06</td>
<td>4.28</td>
</tr>
<tr>
<td>6.36</td>
<td>5.19</td>
<td>4.57</td>
</tr>
</tbody>
</table>
You have five minutes to solve as many as possible. After which, the friendly person at the front of the room tells you to count your correct answers and then shred your sheet of paper. Now you tell them how many of the problems you solved and they pay you £1 for each correct answer.
Let's say you got four correct answers out of 20. How many would you say you got right? Well, if you're like pretty much almost everyone else we've tested, you'd exaggerate your score by about two correct answers and as a result earn £6. This way, we've found that nearly everyone cheats, by just a little bit.
We ran the same experiment again, except that instead of pound coins for each correct answer, participants received tokens which they took to another table in the room and traded for a pound each. If you were in this situation and had to report your performance knowing you'd first get a token for every correct answer, how many would you claim you got right? We were surprised by what we found. People in this experiment reported twice as many correct answers as those in the previous experiment.
What we learned from many similar experiments is that most people will cheat when incentives are pointing them in that direction, rationalising their actions to consider themselves honourable individuals.
We pay people to perform a task and give them the chance to cheat, so they're caught between the desires to make money and to remain honest about their work. Where direct payment is deferred, as in the token condition, people become more susceptible to these conflicts.
A few years ago, a Harvard Medical School student noticed that his pharmacology professor was heavily promoting the benefits of cholesterol drugs and playing down their side effects. After some Googling, he discovered that the professor was on the payroll of ten drug companies, five of which made cholesterol drugs. As New York Times reporter Duff Wilson put it, "Under the school's disclosure rules, about 1,600 of 8,900 professors and lecturers at Harvard Medical School have reported to the dean that they or a family member had a financial interest in a business related to their teaching, research or clinical care."1
The relationships that American and some British doctors have with these companies, which often include research funding and drug samples for patients, can create a feeling of obligation that changes the way they view the company and its products, which can in turn alter their prescribing patterns.
A study by Jay W Friedman, an American dentist with a masters in public health, showed that in the US, wisdom teeth are extracted at an unnecessarily high rate2. That this treatment provides a significant source of income for dentists and oral surgeons makes this a tricky situation. Friedman suggests that some myths about wisdom teeth have developed to support the financially desired outcome. For example, one myth is that these teeth almost always cause infection, when in reality only 12 per cent do (which, Friedman notes, is roughly the same rate of infection found in appendixes, and you don't see appendix removal as standard procedure for all patients).
It's a difficult situation because dentists, like other professionals, want to do their job well and ethically. However, financial incentives can cause them to consider extraction as standard care. It is not about being a rotten apple.
This bias rears its ugly head in any profession where fee-for-service is involved. We must do what we can to minimise such conflicts of interests.
*1: DUFF WILSON, "HARVARD MEDICAL SCHOOL IN ETHICS QUANDARY," NEW YORK TIMES (MARCH 2, 2009)
2: "THE PROPHYLACTIC EXTRACTION OF THIRD MOLARS: A PUBLIC HEALTH HAZARD", AMERICAN JOURNAL OF PUBLIC HEALTH, 97(9), 1554-1560, (2007)*
This article was originally published by WIRED UK
