From Push to Nudge: A Q&A With the Authors of the Latter


“‘Libertarian paternalism’ is just the sort of phrase that makes me stop paying attention,” Levitt recently blogged.

But he (and I) couldn’t stop reading about it in Richard Thaler’s and Cass Sunstein’s book, Nudge, which uses urinals, ABBA, and Homer Simpson (and cutting-edge research) to argue that by simply giving more thought to the way they present choices to people — or “nudging” — choice architects can preserve freedom of choice while dramatically influencing the choices people make.

Here’s an excerpt:

Think of Mr. Spock of Star Trek fame as someone whose Reflective System is always in control … In contrast, Homer Simpson seems to have forgotten where he put his Reflective System … One of our major goals in this book is to see how the world might be made easier, or safer for the Homers among us (and the Homers lurking somewhere in each of us).

Sunstein and Thaler have agreed to answer our questions — individually at first and then jointly — about the book.

Q: You talk about heuristics and decision-making biases which influence most people’s thinking. Which heuristic or bias has taken you for a ride? Are we as a society more or less vulnerable to heuristics than we were 20 years ago and why?

INSERT DESCRIPTIONCass Sunstein in his office.

CASS: The one that has most gotten to me is the availability heuristic, which means that people assess probabilities by asking whether examples easily come to mind.

For about two years after 9/11, I was scared to fly, even though I knew, from my own work, that the risks were really low. And after a bad incident with Chinese food (I have a severe shrimp allergy, and the vegetarian dish contained shrimp), I have been ridiculously nervous about shrimp hiding somewhere in Chinese food.

Fortunately, I am also subject to optimistic bias — with respect to just about everything — and so I now fly contentedly and eat Chinese food happily if sometimes a bit warily.

I don’t know if our society is more vulnerable to bad heuristics and errors than it was 20 years ago. Certainly there’s a lot of vulnerability to those things, but the same as been true for a very long time.


RICHARD: We both fall prey to the perils of inertia, in part because we are busy and disorganized. Perhaps that is why we spend so much time in the book talking about ways in which things can be made “automatic.”

Q: You use the term “libertarian paternalism” to describe “nudges” that agents of change (or choice architects) can use (without using force or removing options) to influence people to make better choices while still preserving their freedom to choose.

What’s the most ingenious nudge you’ve found yourself influenced by? Are you as influenced by nudges as others are? What factors can make us more immune to nudges?

CASS: My most ingenious nudge is: automatic payment of bills. I used to pay late, a lot of the time, through sheer inertia. Now I do a lot better on that count. And sure, I am influenced by nudges, especially in the form of visible chocolate things in cafeteria lines and at airports.

The best inoculation against bad nudges is to stay away from them — to find some way to tie yourself to the mast. Ulysses was a good, early behavioral economist.

RICHARD: I spent a few months visiting the new business school at the University of California in San Diego this winter. When you approach the building from the parking lot, the first thing you reach is a staircase. The elevator is another 50 feet away. I used that nudge to encourage me to walk up the stairs to my fourth floor office. I am proud to say that I never once took the elevator, even if my backpack was heavy.

Q: When does a nudge risk becoming a shove and how do you avoid crossing that line?

A: There is no bright line here. A nudge clearly becomes a shove when it is mandatory, but the harder it is to opt out, the more a nudge turns into a shove.

We have tried to devise policies where the costs of opting out are small, and in some cases we suggest strategies we call “one-click paternalism” because you can opt out with one mouse click.

If you have to fill out a long form to opt out, then that can become pretty burdensome. Both of us are horrible form filler-outers and so we are sensitive to the shove-like potential of paperwork.

Q: Is there a situation where it would be imperative to shove instead of nudge. How would you, as a libertarian paternalist, justify such a situation?

A: When children and third parties are at risk, mandates and shoves may be OK. We are not opposed to mandatory vaccination laws, in part because those who don’t get vaccinated endanger others.

Many antipollution laws are fine too. A full answer here would point to the costs of bargaining: when people can’t contract their way to a sensible outcome, because of collective action problems and a lack of information, the argument for a mandate gets stronger.

Shoves that aren’t that big an intrusion, such as mandatory seatbelt laws, are OK too, if they can be shown to save a lot of lives. But generally, we like freedom of choice.

Q: Can even the most well-intentioned nudges produce unintended consequences?

A: Yes — consider the case of organ donations, which you have blogged about. In the U.S. people have to take some action — such as filling out a form and signing the back of their driver’s license — to become organ donors.

But some European countries have adopted “presumed consent” under which people are assumed to be willing to be donors unless they have signed some form. The evidence shows that this policy produces more organ donors. However, one unintended consequence is that family members may be more willing to overrule the wishes of the (deceased) donor if those wishes were merely implicit (that is, the donor did not opt out) rather than explicit.

A policy in this domain that we find appealing is “mandated choice.”

Under this plan when you apply for a driver’s license you would be required to check one box or the other. This solves the unintended consequences problem, and is probably politically more acceptable since some people have strong feelings about this issue.

Q: One important role of choice architects is setting default options — or defining what happens when people do nothing. How can something so seemingly innocuous be so important?

Two reasons. One involves suggestion:

If the default rule favors savings, or organ donation, you might think, “Umm, someone smart decided that this rule makes sense,” — or, “well, most people are probably doing that, so I should too.”

The other reason involves inertia. We think that many people follow the “yeah, whatever” heuristic, especially when they have many things to do. One of us (Sunstein) subscribes to lots of bad magazines because of that heuristic. He hasn’t bothered to cancel them.

Q: You write that human brains function either on their more deliberate and self-conscious Reflective System or their intuitive, rapid Automatic System (Homer Simpson), and that voters seem to primarily rely on the latter. (So much so that it’s possible to predict the outcome of elections by testing automatic responses.) Can you make any predictions about the presidential election based on which candidate has more “automatic appeal?”

A: We are betting on the guy from Illinois — the one who used to teach at the University of Chicago Law School. Admittedly, we might be biased, but we have seen first hand the way people immediately relate to him, and that should help.

Also, it is important not to push this evidence too far. Although initial impressions are important, certainly many voters do use their reflective systems to consider each of the candidates with care. We think the country is lucky to have three smart candidates who can give voters plenty to think about.

Q: You tout libertarian paternalism as a promising foundation for bipartisanship. How, exactly, does it bring Democrats and Republicans together?

A: Democrats want to use government power to make people’s lives go better; Republicans respond that people know more than politicians do.

Republicans like to maintain free markets; Democrats respond that free markets can get people into trouble.

We think that both might be able to agree that nudging can maintain free markets, and liberty, while also inclining people in good directions.

Examples include the Save More Tomorrow plan (by which employees agree to devote some of their wage increases to savings), and automatic enrollment plans, (by which employees are assumed to want to enroll in standard savings programs).

Save More Tomorrow and automatic enrollment are helping people to have a lot more money for their retirements. Both Republicans and Democrats gave enthusiastic support to a provision in the 2006 Pension Protection Act that gave employers a small nudge to adopt these policies.

We also like the idea of a Greenhouse Gas Inventory (G.G.I.), by which major contributors to climate change would have to disclose their emissions. The G.G.I. would impose no regulatory mandate, but it would probably do some real good.

Q: What, in your opinion, has been the most valid critique of Nudge and how did you respond?

A: We are probably not the best judges of what the most valid critique of the book is. We think, for example, that the pictures of the authors could have been more flattering. But this is another case where our own biases might possibly be playing a role.

What we can say is what the most frequent criticism of our earlier work on the topic has been:

Many critics from the right have repeatedly reminded us that government officials can also be subject to decision making biases, and might well be captured by special interests. Needless to say, the thought that public officials, even at the very highest levels of government, might not be the brightest bulbs in the universe has occurred to us.

We have several responses to this concern.

On the point that choice architects are human, we of course agree, but first, we stress that — as Adam Smith made clear — there are gains from specialization. It is possible for the government to hire a competent expert to design a choice environment in which individuals have an easier time making good decisions.

Second, we remind people that it is often impossible to avoid choosing some specific choice architecture. For organ donations, we need some rule — be it opt in, opt out, mandated choice, or flip a coin.

It is important to stress that we do not advocate a larger role for government. We advocate a smarter role, and one that makes the lives of the citizens easier and better.

Q: In the book, you describe how nudges can solve the mortgage crisis, fix Medicare, overcome the fight for gay marriage, and address other hot-button issues. Which of these proposed nudges is most likely to catch on and why?

A: We think our mortgage and credit proposals have a good chance of catching on.

The basic idea is that people should get clear disclosure — through a kind of “simplified transparency” — which would allow them to see how much they’re paying. We believe that the trick is to require firms to disclose their prices via electronic files that could be uploaded into private web pages that would compare offerings.

This policy has become necessary because the world has become so much more complicated. It used to be that all mortgages were 30-year, fixed-rate with 20 percent down. Now mortgages vary along dozens of dimensions. Even most economists find choosing a mortgage extremely difficult, so it is not surprising that many of the homeowners now in trouble didn’t understand all the terms of the loan they were taking.

We’re also pretty enthusiastic about the prospects for a Greenhouse Gas Inventory.

Q: How did you nudge people to read your book?

A: Mostly by trying to write a book that is fun and easy to read. We had a rule when writing the book: if either of us found something boring, we took it out!

We also tried to offer occasional humor, sometimes stealing lines from Homer Simpson, who suffers from every bias we know about.

We spent more time than most authors worrying about the cover. We hope that the mother elephant nudging the baby elephant is a friendly, engaging image. Finally, having a one word title should make it easy to remember the title of the book. Hint: it is Nudge.

If this discussion has inspired a nudge idea, pitch it to Cass and Richard on their blog.

Leave A Comment

Comments are moderated and generally will be posted if they are on-topic and not abusive.



View All Comments »
  1. ykjay says:

    I’m curious whether the authors would consider the X-Prize or the DARPA Grand Challenge prizes to be a nudge, or whether that’s something on a different scale.

    Thumb up 0 Thumb down 0
  2. Charles D says:

    You’ve nudged me. I’m going to go buy the book now.

    Thumb up 1 Thumb down 0
  3. Chesapean says:

    The “nudge” paradigm is disturbing.

    Everyone should learn that their behavior can be engineered to follow predictable patterns — beneficial and not. The essence of the “nudge” paradigm, however, seems to be that any practitioner of “nudge” technology can be as good for human welfare as another: YOU TOO can invent urinal flies!

    Or pet rocks, one might add.

    To engineer choice is the opposite of developing human capacity for selection. Only persons capable of the latter are qualified to attempt the former.

    Show me a “libertarian paternalist” and I will show you a man or woman who — most likely — cannot teach.

    Thumb up 0 Thumb down 0
  4. Nikhil Punnoose says:

    Chesapean, I think you’re missing the point. True, this might not help much in “developing human capacity for selection”, but I think one key assumption is that most people don’t do that anyway. For those who wish to develop this ability, this is unlikely to provide a deterrent, and that, simply, is the essence of the “nudge paradigm”.

    Thumb up 0 Thumb down 0
  5. JP says:

    I can’t wait to get this book! The ideas make a lot of sense to me since I pay my bills and rent late every month and I have a bunch of subscriptions to magazines I never read just like the author. Whenever a company offers automatic bill payment I always do it and feel relieved.

    Thumb up 0 Thumb down 0
  6. David says:

    I have to say that “choice architects can preserve freedom of choice while dramatically influencing the choices people make” sounds like a contradiciton – and a sinister one at that!

    It seems that we are willing to contemplate using manipulation of people (exploiting emotion or ignorance/laziness) to get them to behave the way we want – in the name of saving public money?

    Add in new technology like brain scans and the world could get very controlling very fast…

    Is it not true that a more expensive government may be a less invasive one?

    We could allow people much more freedom – a greater range of choices in their behaviour – if we stopped trying to manipulate them into the most efficient, least expensive option.

    PS – I see voluntary automatic payments, for convenience sake, as a completely different category of issues to ‘presumed consent’ in organ donation…

    …I’m not sure how they fit in the same book.

    Thumb up 0 Thumb down 0
  7. GWU Econ says:

    I’m halfway done with the book and I think it’s a great read. The policy recommendations are great and it’s really surprising to see the statistics on how much change proper choice architecture can bring about in complicated areas(retirement plans, credit issues, etc.) Congrats

    Thumb up 0 Thumb down 0
  8. Rick says:

    Chesapean, it would seem to me that Nudging is an attempt to encourage outcomes when teaching either has failed or is not available — not as a substitute. It’s important to recognize that nudges already exist, designed or not. Every situation has some “natural” state which creates a set of incentives and disincentives. Nudging is simply arranging those in a more beneficial manner until a more thorough educational program can supercede.

    While it might be an ideal to have perfect information and have all individuals be perfect rational actors at all times, that simply isn’t reality. Nudging serves as a compromise; a recognition that when information is imperfect or when actors aren’t rational, we can design systems in such a way as to make things a little bit better than they’d be otherwise.

    Thumb up 0 Thumb down 0