A portion of the interview is published here. We know you will enjoy it.
WHAT IS BEHAVIORAL ECONOMICS?
Traditional economics is based upon the assumption that economic actors are rational. The problem is that they aren’t.
Behavioral economics seeks to study this irrationality and find patterns in it. It has led me to study everything from reluctance to save for retirement to the inability to think while experiencing sexual arousal.
CAN YOU GIVE US AN EXAMPLE OF THE RESEARCH YOU CONDUCT?
In one study we tried to figure out how often people cheat. To do this, we brought a bunch of students into a room, gave them 20 simple math problems and an impossibly short period of time to complete them. The students submitted their answer sheets to the experimenter, and were rewarded with $1 per correct answer.
Then we tried to figure out how much they would cheat in this task if no one were monitoring their performance.
We conducted the same basic study again, but this time we asked the students to count how many questions they got correct, shred their papers, and only then report to the proctor how many questions were solved correctly. Now, the students answered significantly more questions “correctly” than before. We deduced from this that some of them cheated.
Surprisingly, the results weren’t that a few people cheated a lot. Instead, most students appeared to cheat just a little. I call this the personal fudge factor.
WHAT OTHER IRRATIONALITY DID THIS REVEAL?
One of the things that was most surprising to me was that we found that people react differently to money, and to money-substitutes.
We conducted the exact same study but instead of paying the students directly with money, we gave them a token for each correct answer. The students were told that they needed to walk 12 feet to the next table and turn in those tokens for $1 each. In essence the payment was the same, but when they had a chance to lie, they would do it for tokens and not for money.
Rationally, cheating rates shouldn’t have changed, but what we found was that cheating doubled.
WHAT ARE THE IMPLICATIONS OF YOUR RESEARCH?
The cheating study explains, at least in part, why so often we see certain bad behaviors in business. Ask yourself this question: Are you more likely to take a pencil home from work or take 10 cents from a petty cash box?
While these two things are identical in their cost to a business, the average person does not view them the same way. People are much better at being honest when they’re around money. It is when we create an intermediary step that problems start to occur.
There are numerous examples like that. But my goal is to find irrationalities that are actionable so that we can recognize our limitations and act accordingly. And the really good news is that we are starting to see success on this front. The United Kingdom just opened a cabinet-level office of Behavioral Economics. This demonstrates that policy-makers find our conclusions useful and practical.
WHAT ADVICE DO YOU HAVE FOR MARKET RESEARCHERS?
What people do is not what they say. In general, most of our findings suggest that we should not think very highly of focus groups.
I suspect that many market research companies use focus groups because they want a statement to put in a PowerPoint presentation for a client. But the problem is that what people say doesn’t always reflect what they think or, even more important, the reasons for their actions.
SO, THE QUESTION IS, WHAT SHOULD YOU ASK?
You can ask people what they have done, or what they think they will do in the future. They can answer those questions. But, the moment you get into reasons, into why, interpreting their answers as correct becomes much more tricky.
It is also important to ask about categories that are ones which people can accurately quantify. The more concrete your response scale is, the more likely people are to answer your questions in an informative way.
HOW WOULD YOU RECOMMEND RESEARCHERS USE SCALES?
Take the question, “How often have you done X?”
If you create a scale from “very rarely” to “very frequently”, that’s not as useful as offering, “2 times last week” or “3 times last week”.
The more concrete you get, the more people feel inclined to answer accurately and honestly.
ANY MORE ADVICE ON HOW TO USE SCALES?
Often people use a 5-point response scale, but we find most people have an aversion to the extremes. This means that when we use a 5-point scale, effectively we are using a 3-point scale. That makes our sensitivity of our measurement less useful.
I would encourage researchers to think whether people will have that extreme aversion, and if that is the case, to have more levels. A continuous scale with just 2 anchors in the extremes in such cases is ideal.
WHAT IS YOUR CURRENT RESEARCH FOCUSING ON?
We have recently done a set of studies on loss aversion. Say you are a salesperson, and you can sell either television A or television B. Selling Television B is more profitable for you, but the manufacturer of Television A pre-pays your commission, and you have to give the commission back if you don’t sell a television.
What we find is that because of loss aversion, people often sell television A. They just hate giving the money back, and because of that they do something that is financially inefficient.
WHAT ARE THE APPLICATIONS OF THIS?
We are trying to apply these benefits to 401ks. We think it can increase the percentage of people who save for retirement. Let me explain.
Many companies match for employee 401ks. How does the matching happen? The employee adds a certain amount each month, and the company matches. If the principle of this type of loss aversion for money exists, a better way to get people to save for retirement is for the company to put all the matching money for the year in their account already. Then every month that they don’t put in their full matching, the company takes back some of it. We are currently trying to get companies to test this 401k idea.
HOW DO YOU SEE BEHAVIORAL ECONOMICS INFLUENCING TRADITIONAL EC0NOMICS?
I don’t want it to. I think traditional economics is a good thing. It provides interesting models and some important insights.
I would like to see traditional economists do their research and behavioral economists do ours. Then, when a practical question comes up (in business or in public policy), we should run a test comparing two alternate ideas and see which one works.
I think both should influence how we make decisions in the real world insofar as their predictions bear out through the evidence.