The USS Quad Damage

This article and its content has been talked about in multiple TED talks and various articles. Since Dan Ariely is not only entertaining but usually enlightening, this is a good thing. However, with this research, I believe he’s mis-stepped. Basically, the research says:

If you offer people increased cash incentive to do things, they will be worse at it. However, if the task is menial, then incentive is directly proportional to the performance.

This is used to make the case that the argument “CEOs need high salaries to attract the best talent” is invalid. However, according to the test, the incentive is always fixed and known a-priori, and also given to a random sampling of people. There is no test showing that the talented ones cannot readily achieve these goals and thus “rise to the top”. In fact, doing worse with higher incentives might be the exact mechanism by which the best CEOs' talent is chosen. That is, CEOs with more talent will still succeed with high incentives, where lesser CEOs will falter. The test I propose is similar to this:

# Give everyone 1 unit of money, and a simple problem to solve. # If the person solves the problem successfully, double their money. Otherwise, halve it. # Repeat and look for people who are consistently the top performers. Are they consistently able to solve more problems?

You might be tempted to say “well of course, they're being paid more”, but it’s not so simple! If I get to, say $32, then get a question right, I’ll go to $64. If I get it wrong, I’ll go down to $32 again. So if I’ve got a 50% success rate, I can stay on my $32 salary. So the real question is, at what point will someone stay at a particular salary, and does this say anything about their “talent”? The test needs a fair amount of jiggling, esp. w.r.t. menial tasks, but I feel this sort of test speaks more to talent remuneration than the existing test.

Secondly, look at this from a game theory perspective: What is the maximum amount of money I can make from doing a menial job? Remember, doing it better means you get more money, so all you have to think about it execution. Is this not a more desirable (i.e. less risky) way of getting money? If I was going to do a creative job which was at risk, what sort of remuneration would I be looking for when compared to the menial job. You could test this by having a market for problems, some menial and some creative, and paying some amount for them. The problem is how to adjust them and make conclusions? However, I believe this is the sort of test that needs to be done to prove the sorts of conclusions people are drawing.