research
Uri Gneezy delivers the 29th BSE Lecture
The 29th BSE Lecture entitled "Incentives and Behavior Change" was delivered at Banc Sabadell Auditorium on April 1, 2014 by Prof. Uri Gneezy (Rady School of Management, University of California San Diego).
In his lecture, Prof. Gneezy cited cases from Netflix and gym memberships in the United States to daycare and "lotteries with regret" in the Netherlands to illustrate the effects, both positive and negative, of incentives on human behavior. Watch the video and then read an interview with Prof. Gneezy by one of his co-authors, Prof. Pedro Rey-Biel (UAB and BSE).
Video: Highlights from the 29th BSE Lecture
Interview: Uri Gneezy on Behavioral Economics
prepared by Prof. Pedro Rey-Biel (UAB and BSE)
What's Behavioral Economics and why is it useful? How can we use Behavioral Economics to convince people to change or improve themselves?
Economics has always been about the behavior of economic agents, being consumers, producers, governments, etc. However, most of the models used since the 1950s until recently assumed that those agents were completely rational as Mr. Spock, and thus very simple with respect to the things agents cared about and with respect to how capable they were of behaving as perfect predictors and strategists. More recently, Behavioral Economics is pushing to make agents more interesting, such as Homer Simpson, and thus finding a coherent way to incorporate richer motivations, emotions and cognitive mistakes, into our models. As such, we have been greatly inspired by other social sciences such as Psychology, Sociology, Anthropology or Political Science, incorporating their insights and methods, in particular experiments, and combining them with our perhaps more structured way of researching.
Explain why incentives are not always used in the right way. Give me one example of an incentive well used and another wrongly used.
Economics is right in that "incentives matter". However, by having restricted ourselves until recently regarding what motivates agents, we have designed incentive schemes that only take into account a limited set of motivations, and thus they only provided the correct incentives in situations in which agents were mostly motivated by what we incorporate into our models. This has worked for many situations, since economic rewards, which is the main motivation incorporated into models until recently, go a long way.
However, there are cases in which agents do not only care about their earnings, but they compare themselves with others' earnings, or they care about their self-image or the image they project to others when taking certain actions which may lead to certain economic rewards. In such cases, well designed incentives will capture more aspects of the motivations of the agents and will be more effective.
Why is it so important for companies or policy makers to test decisions before putting them into practice? Give me an example of a failure that could have been avoided if a field experiment had been done.
Too many times firms implement important changes in the way they do things based on a theory, or even a hunch, that the change will work. These decisions are too important, they can cost millions of dollars or make many individuals worse off, to be taken without evidence. Field experiments can be used to test changes, such as prices or incentive schemes, at a small scale, with a fraction of the cost.
A good example is how firms decide the price of goods whose quality is difficult to evaluate, such as wine. In a field experiment we conducted we just varied the price of wine bottles to check which price would maximize profits. When experimenting with different prices for the same bottle of wine, $10, $20 and $40, we found that demand was not monotonically decreasing in price, and actually peaked with the $20 price. This is probably due to buyers interpreting the price as a signal of quality. We learned this with a minimal cost for the winery, since their only cost were the possibly reduced sales while we were conducting the experiment. Since then, they are selling their bottles at $20 and making much more money than with their old price. Field experiments may cots money, but the real cost is the amount of money you do not earn because of not experimenting!
Does Behavioral Economics go against or complements classical economic and financial theory?
Behavioral Economics has nothing against classical economic theory. The main advantage of economic theory versus the other social sciences is that the use of mathematics, the fact that "Economics is written in Greek letters," has helped us be more precise in our analysis and predictions. Let's not underestimate this. However, if we want to make our research more powerful we need to deal with richer, more interesting, economic agents, which really capture what individuals care about and what they are capable of. There should be a constant dialogue between classical and so-called behavioral economics, the same way that there should be a two-way dialogue between theory and empirical work.
Related research
- Charness, Gary and U. Gneezy. "Incentives to Exercise." Econometrica, 77: 909–931, 2009.
- Gneezy, Uri, S. Meier and P. Rey-Biel. "When and Why Incentives (Don't) Work to Modify Behavior." Journal of Economic Perspectives, 25(4): 191-210, 2011.