As I’ve mentioned before, I am teaching behavioral economics this semester. So far, I’ve introduced the concept of prospect theory and loss aversion- specifically, how people how people are made more unhappy by a loss than they are made happy by what they see as an equivalent gain- and we discussed how loss aversion can lead to anomalies in behavior in financial markets such as the disposition effect, i.e. the tendency fro investors to be more willing to sell securities that have gained in value since they were purchased than to sell securities that have lost in value.
The concept of loss aversion also has potential implications for the mechanics of incentive systems. Since I’m not at all weird, I gave my students the following thought exercise:
Consider the following scenarios:
- You are running to catch a really cute stray puppy so that you can give him a good home.
- You are running to catch your puppy because he wandered off without his leash.
In which scenario would you exert a higher level of effort (assuming of course that you actually like dogs)? What does this have to do with loss aversion as it relates to incentives?
My students, not being monsters, universally stated that they would work harder in the second scenario, despite the fact that I pointed out that the only actual difference between the two scenarios is the wealth effect of one puppy. (I can’t wait to get my student evaluations.) Of course, this isn’t quite a fair example, since I would hesitate to consider it irrational to have a sentimental attachment to one’s dog. I mean, look at this face:
But people tend to have similar preferences even when there is no reasonable emotional attachment to an item to be had:
Not surprisingly, given what we know about loss aversion, there is empirical evidence that people will work harder to avoid what they perceive as a loss than to get what they perceive as a gain. For example, Tanjim Hossain and John List conducted a field experiment in a Chinese consumer electronics factory and found that productivity was (modestly) higher when workers were provisionally given an amount of money and told that it would be taken away if productivity didn’t meet a certain threshold than when they were told that they would get a bonus if productivity met that same threshold. Similarly, Roland fryer, Steve Levitt, John List, and Sally Sadoff ran an experiment with teachers in the Chicago public school system and found that student performance was higher when teachers were given a lump sum payment at the beginning of the year and told that some or all of it could be clawed back if the students didn’t meet performance targets than it was when teachers were told that they would get a bonus at the end of the year if their students met performance targets, even though the ultimate amount of money that the teachers received for a given level of student performance was the same in both cases. (It’s interesting that the researchers got this result even though most of the teachers in the “loss” treatment didn’t spend the up-front money.) This latter result is particularly fascinating since there is debate on whether teaching is the sort of activity that can be effectively incentivized in the first place.
One of my students ran with this concept and asked the following: “Does this mean that those guys in restaurants that put a bunch of dollar bills on the table as a tip and then threaten to take them away when service is not up to par are actually on to something?” My response was that it depends on two factors: first, whether the server feels like the dollars on the table are his and incorporates them into his idea of a psychological neutral point, and second, what the douchebag customer’s utility function for getting his food spit in looks like. I then only barely managed to avoid going on a tangential rant about tipping.