Freakonomic’s Dubner: Art Of Incentives

Behind every good incentive is good data, but finding good data takes asking the right questions.

CinthyaMurphy_200x200.png
Reviewed by: Cinthia Murphy
,
Edited by: Cinthia Murphy

Stephen Dubner is a journalist, who, among many accomplishments, has the distinction of being one of the co-authors—along with University of Chicago economist Steven Levitt—of “Freakonomics: A Rogue Economist Explores the Hidden Side of Everything,” a book that has been challenging conventional economic thinking for a decade now. Since that publication in 2005, Dubner and Levitt have followed with two other books, “SuperFreakonomics” and, the most recent, “Think Like a Freak,” released last year.

Dubner has an interesting take on the world, and one he is going to share with attendees at the Inside ETFs conference on Jan. 25 in Hollywood, Florida.

ETF.com: You're going to be speaking at our conference in January. Can you give me a preview of what you'll be talking about?

Stephen Dubner: I really believe in learning—everybody trying to learn more about what they do and trying to get better, whether it's professional, personal, political, whatever.

The more work we do, the more I've come to believe that the way you get better at something is by getting really good feedback. The way you get really good feedback in the world is by collecting a bunch of data and seeing what really happens.

In any realm, whether it's investing or politics—which isn't very rational sometimes—there are a ton of conventional wisdoms and rules of behavior that were arrived at maybe through experimentation, maybe through trial and error, but often also through guesswork and hunches.

But in any institution or industry, when you do something long enough, we tend to accept that that must be the best way to do it. The message I try to bring to a group like this is to work really hard to get unpolluted data—data that reflects real behavior, not just what people say they're going to do, which is a particular problem with investing.

If you ask people 100 questions about their risk appetite and investment goals, you'll get answers. But then if you compare those to how people really behave when the markets get very volatile, you'll see that there's very little correlation between how people say they want to invest or want to behave around investment, and how they actually do.

I want to advocate for finding ways to get data that reflects reality. And then, importantly, use those data to understand the incentives that make people behave as they do. The goal is to better understand human behavior and therefore make better decisions.

ETF.com: There is so much data available to us today. How do you suggest we do better data mining? How do you separate the noise from the useful?
Dubner:
That's a really interesting question, in part because about 10 or 15 years ago, the complaint was the opposite—that we don't have enough data to answer the questions. Now, a lot of people feel we're overwhelmed with data; what do we do?

I think a big issue is that the people who tend to be really good at collecting and storing data—especially in corporations—are the “data people”; it leans toward the IT side. However, what we need to get better at is coming up with the right questions to ask of the data.

There’s this whole notion of big data fixing everything and the notion of the data scientist as the white knight. But a lot of firms—and a lot of institutions, including governments and so on—aren’t very good at asking good questions of the data. It’s not just about getting the data, but using it well.

Raw data can sometimes answer some really important questions. But often, you need to put the data under a microscope and look at it in a bunch of different contexts to see what it's really telling you about how people behave. That’s where you need creativity, and you need people who are clever, and people who are a little bit iconoclastic.

It’s really about challenging conventional wisdom. I think the only way to encourage more of that is to start at the top in firms and have the leadership really show that it is OK—in fact encouraged—to try to come up with totally different questions and totally different ideas.

It’s important to incentivize employees throughout a firm to take a truly contrarian position when it comes to challenging the data to learn more than you already know.

ETF.com: In your opinion, who uses data well today? The Fed? Public schools?

Dubner:
The Fed, hard to say—because macroeconomics is much less science than macroeconomists would have you think.

I just interviewed [former Federal Reserve Chairman] Ben Bernanke, and it was really fascinating. He's obviously an incredibly bright guy, and I would argue, a very dedicated civil servant for several years.

But it's interesting. You take someone who—in his role as an academic economist—was inherently skeptical of broad conclusions and very critical of his own theses, and you put someone like that in a policy environment, and a lot of that goes out the window.

That’s because being true to the data and what the data can and can’t tell you is a lot easier to do in academia than it is in the real world where you have people saying, “I don't want to hear your theories; I just want you to tell us what to do.” Inevitably, you get a lot of shorthand and a lot of shortcuts.

The Fed has access to a lot of macro data, but I don't know if I can ague that it uses it very well.

The public school system? I would say no. If you went to sleep 130 years ago and woke up today, very little around you would be recognizable at all. Our transportation is different. Our communication is different. Our medicine is different.

But the schools are the same. It's one teacher typically in a box with 20 or 30 kids, and some kind of chalkboard or communication model. It's unbelievable how little that model has been expanded or challenged.

Now, look at a company like Google. It’s a data company. That's what they do, and they do it well. Facebook is essentially a data company; they do it very well. Uber is an unbelievably good data company.

You know who else does it well? Academic economists. Now, I'm biased because these are the people I know, and I write about them a lot and collaborate with them. But I will say that where economics has an advantage over other social sciences is that they work with big, big data sets in a very sophisticated way.

Does that mean that they have an understanding of human nature the way psychologists do? I would say almost never. But if you want to talk about a group of people who can harness big sets of data, and get it to tell us some important truths, I would say that academic economists are pretty good at it.

ETF.com: Are all the data making for better investors?

Dubner: I admit I haven't read a lot of investing literature lately. But [over the years] I’ve gotten the clear impression—and I'm not saying this to flatter you guys; you're the ETF crowd—that some sophisticated form of passive investing is going to beat almost all forms of active investing for the median investor—without question. The data show that very, very clearly.

What does that really show? It shows that there's a lot of overconfidence among investors. There's a lot of risky behavior without realizing the risk among typical investors. There's a lot of acting in the belief that they understand the data probably better than they do.

So, there's a lot of ways to look at it. Different people have different preferences and different incentives. Some people really like to invest actively and even wildly because it’s partly entertainment.

I don't happen to think that's a great idea, because money is kind of nice to have, but it’s really surprising to me how many people spend so much energy developing their own investing prowess when most of the evidence suggests that they actually don't have any investing prowess.

ETF.com: You're focused on this idea that we live in an “incentives” economy. Where do we—and don’t we—use incentives properly?
Dubner:
The short answer is we often don't incentivize properly. It’s really hard to say where we use them correctly and incorrectly because every scenario is different.

If incentives were simple, when you want people to, say, eliminate violent crime from society, what you do is increase punishment. And so, if anybody is involved in any kind of violent behavior, and they're found guilty of it, then the minimum sentence is going to be a lifetime in prison. Or death. If you did that, it would almost certainly decrease violent crime.

On the other hand, the trade-offs are massive. Probably none of us wants to live in a society like that. So any incentive is subject to abuse and overreach.

But the other problem is that an incentive that might work very well in one kind of firm won't work well in another, because firms have different cultures. For example, if a firm offers a cash incentive for a certain sales goal, that might work really well for six months or a year or a year-and-a-half. Then you may see performance trail off, because people just get used to what they have, and they get less satisfied with what they have.

That's part of what makes us human and awesome. But it's also part of what makes it difficult to predict their future—yesterday's behavior doesn’t not necessarily predict tomorrow's behavior.

If you want to be good with incentives, you have to get a lot of data. You have to do a lot of experimentation to figure out what works. You have to constantly monitor the data to make sure that the incentives are still returning what you want them to. You have to be ready to update or change them when they start to backslide or fail.

ETF.com: Are there types of incentives that work better?

Dubner: One thing that's a fairly good general rule is that we believe very much societally in the power of moral incentives—urging people to do the right thing because it's the right thing.

But the data show that that rarely works. It doesn't mean people are immoral. It just turns out that moral incentives typically are not very successful, which doesn't mean they won't work in some cases.

Financial incentives aren’t necessarily better. But social incentives are often really strong—what you might call the “herd mentality.” If you want people to do “\‘the right thing,” whatever that is, you don’t tell them, “You should do this because you’ll be a better person for it; it's the right thing to do.”

That will typically produce a much weaker outcome than if you simply tell people, “Did you know that a lot of your neighbors or people in your community are doing this thing? You might want to think about doing it, too.” And that kind of herd mentality or social incentive is typically stronger than a moral incentive.

The other thing about incentives is that a lot of them are created by people in a position of authority or power, who happen to be so in part because they're very disciplined, motivated, rational people.

They create incentives that they would respond to, but then, they unleash them into the world where not everybody is as disciplined and rational and motivated as they are. That will often lead to a failure.

ETF.com: You're a journalist, a former rock band musician, and you've essentially created a new line of economic thinking—freakonomics—with Levitt. Has your own decision-making process changed since you started this work with Levitt 10 years ago? Do you see the world differently?

Dubner: You know, I think I do. But I'll be honest with you: It's kind of hard to tease out the effect of the influence of the work on my life. A lot of other things changed. My kids are now teenagers, so that just brings you into a whole lot of different kinds of scenarios and interactions.

It’s really hard to say. The one thing I've become is a big believer in what you'd call “heterogeneous preferences”; just the idea that different people want different things and respond to different kinds of incentives.

Even though that sounds really obvious, a lot of our social and political discussion in this country seems to pretend that that's not the case. It seems to pretend that, no, there is a right way to think about X or Y or Z.

When you realize there are these heterogeneous preferences, it leads you to be a little bit more humble about assuming that the solution you might dream up is not the best solution for everybody always.

If you're going to be in a position where you're making policy or rules or decisions for groups of people, you need to be really aware of the fact that you need to have room in there for people to exercise their preferences and not go afoul of your rule just because they see the world a little bit differently.


Contact Cinthia Murphy at [email protected].

Cinthia Murphy is head of digital experience, advocating for the user in all that etf.com does. She previously served as managing editor and writer for etf.com, specializing in ETF content and multimedia. Cinthia’s experience includes time at Dow Jones and former BridgeNews, covering commodity futures markets in Chicago and Brazil equities in Sao Paulo. She has a bachelor’s degree in journalism from the University of Missouri-Columbia.