« "Is the macroeconomy self-equilibrating?" is a stupid question | Main | Oprah Winfrey, victim of racism - or marketing? »


Feed You can follow this conversation by subscribing to the comment feed for this post.

Paternal used in the pejorative is sexist. Fathers are people too.

Patrick - fair enough. What phrase/word would you use instead?

Given that almost everyone cares about at least a few persons, it is strange how seldom that is included in economic analyses. Even at the core of micro, the first welfare theorem, everyone is completely selfish by assumption since the contrary would introduce a couple of billion public goods (where only a subset of income distributions would be Pareto-efficient).

But hey, if a theorem makes it easier to suck up to the rich, why change it.

The NPR show "Intelligence Squared" recently held a debate on the legalization of all drugs. While I am largely supportive of the idea, one opponent raised a point which--though it applies only to a tiny minority of drug users--is still valid: that the impact of drugs on human agency can be so devastating that only severe external force offers any hope of saving that person. We accept this (maybe) with severe mental illness, that some people are too ill for their own good, and there should be some protective powers that are used either temporarily (getting "formed", or picked up by the cops for a psych assessment) or in extreme cases permanently (institutionalization). Schizophrenics and crack addicts are relatively rare, but the problem is that an assumption of impaired agency underlies all paternalist concerns, even if it's just mild impairment (i.e., your time preferences are wrong). So, the persuasiveness of the paternalist argument is that impaired agency lowers the quality of the free choice that we are said to falsely idolize.

On the other hand, bad decision-making is endemic. Something like 30 percent or more of people with serious illnesses don't follow the recommended treatment, even if it's free. The practical question is, if we're going to allow libertarian paternalism, who's going to monitor if it's working and put the brakes on it when and where it fails? After all, we apparently need truckloads of it.

While I don't agree with a lot of libertarian paternalists, I am deeply sympathetic to their concerns. But I fear that underneath this is an impulse to engineer all private behavior in a direction measured by its comprehensibility to outsiders and their private norms. As a non-smoker, you can't comprehend why anyone (especially someone who's, say, had a heart attack) would continue smoking. Fair enough. But what about other decisions that turn out to be poor in retrospect? Should all those sad-sack failed entrepreneurs on Dragon's Den have been nudged away, years earlier, from blowing their entire life savings and their kids' college funds on bad business ideas? I'm not fond of slippery-slope arguments, but this is one area where the intuitive appeal of the idea, married to its weakly defined scope, is a big concern.


One thing about us libertarian paternals is that we listen to people's stated preferences. A lot of people say, "I really should save for retirrment," but never do it. They just constantly regret not saving.

But making savings opt-out instead of opt-in, people generally report much more happiness with their outcomes. It has to be either opt-in or opt-out, so if you go through people's preferences, you can find the things that people want, but don't do, and maximize utility.

I don't see examples like Dragon's Den holding weight because those people really want to try, and who knows whether their business will be successful? Humans are horrible at predicting the future. Whereas we know reasonably well what will happen with cigarettes.

That said, of course, as a libertarian paternalist I support people to do whatever drugs they want so long as they are aware of the risk and pay the externalities.

Daniel - in terms of equation 4 above, you're saying that you know what c* is.

I'm not sure it's always that easy to know. Take excessive drinking, for example. It's easy to say "people would be happier if they didn't drink". But drinking is often self-medicating some underlying mental health issue. Not all forms of depression respond to seratonin inhibitors, and drinking numbs the pain. For someone with depression that doesn't respond to conventional treatments, what is c* anyways?

Shangwen, will reply later.

The formulae from (3) onward assume that A has telepathic abilities with respect to B's utility... You need to include some model of the "intentional stance" by which A estimates B's utility. For example, Adam Smith discussed "sympathy" and assumed something like u_B(x_B) = u_A(x_B), i.e. putting yourself in somebody else's shoes.

Vlad - when I wrote down equation (3), I was explicit in saying that A cared about B. Caring preferences are most commonly used in economics of the family. Telepathy isn't the only way of discovering a person's preferences - it is possible to get to know another person by spending time with them.

I agree with you that, if we're talking beyond family/close friends, it's useful to think about how it is possible to know what another person wants. I would also say that this is something "libertarian paternalists" don't talk about enough. The Moral Sentiments Adam Smith, as opposed to the Wealth of Nations Adam Smith.

Shangwen: "if we're going to allow libertarian paternalism, who's going to monitor if it's working and put the brakes on it when and where it fails"

The problem with libertarian paternalism is that it's inherently contradictory - one can't be both libertarian and paternalistic.

The political brilliance of the term libertarian paternalism is that it's inherently contradictory, and so both social conservatives and social liberals can see something they like in it.

I agree, Frances, I don't know what's best for any individual person. People are best at knowing what they want. And your example of the self-medicating person is valid.

But the beauty is, we don't have to know what's best for people. We can allow people to tell us what they want, and then follow their expressed preferences.

Say, for example, there's a person who enjoys going out and drinking with friends. But then, when drunk, he continues drinking and then endangers herself and others. I know many people like this. So, this person goes into his Service Ontario office, and puts a limit on himself, so she can have no more than 5 drinks a night. This limit is put on his chip provincial id card, which the person taps every time she orders a drink. It records how many drinks the person has, and cuts him off after five.

This limit can, of course, be changed anytime the person is sober.

That is how government intervention can help people reach their own goals, and can intervene to make people better off without removing their agency.

Daniel: "we don't have to know what's best for people. We can allow people to tell us what they want, and then follow their expressed preferences". Any reason why people couldn't pursue their preferences themselves?

In any case, explicit expression of one's preferences is not the basis of LP. Even if Thaler didn't intend it, its appeal to many in the chattering classes is that subtle manipulation (hence the "paternalism" in the name--that's why LP isn't called "help") is justified when people either fail to see what's good for them, or fail to devote enough resources to something that they clearly should be doing (cutting their ice cream consumption, for example). What I dislike about LP so much is the way it has been taken up by certain elites as a brilliant solution to a very arbitrary menu of human frailties. If it's such a problem, use something that works--a consumption tax--instead of an elaborate manipulation scheme.

A lot of positive changes in social behavior have happened in ways inconsistent with the LP assumption that humans are gross engines of unconstrained appetite. Consider the felicitous plummet in alcohol consumption over the past 200 years. That was a combination of many things, only some of which (consumption taxes, age limits) are explicitly regulatory.


I agree with almost everything you wrote, except for your statement that "explicit expression of one's preferences is not the basis of LP." Thaler constantly brings up listening to people's preferences in Nudge. Yes, many of the chattering classes see LP as a way to reintroduce paternalism, without seeming like one of those backwards social-conservatives. They use it to try to make everyone be just like them, and punish people who live lifestyles dissimilar to their own, and see people (other than themselves, who are of course immune to such plebeian things), as you put it, as "gross engines of unconstrained appetite." I detest those people, and if you just want to argue against them, I am with you 100%.

However, I think it is unfair to lump the idea of LP in with that. For example, I call myself an environmentalist, yet I detest most enviromentalists. Most enviromentalists don't understand tax incidence, and try to make corporations (which, to them, are EVVVILLLLL) pay for everything through really bad regulation, thinking that consumers won't pay. They hate economists. They eat organic (which almost all studies show is significantly worse for the environment), and indulge in homeopathic and natruopathic "medicine."

I am a vegan, yet I detest most vegans. Most vegans are extremely judgemental people, who are more interested in feeling superior to other people than trying to make the world a better place.

I am a libertarian paternalist, yet I detest most libertarian paternalists.

I have found that the easy way to tell if a person is genuinely interested in helping people reach their own goals (or if the person thinks that their life is the ideal one and everyone should want to be like them) is to ask the person examples of their own failures at being rational in their own life. People who are genuinely interested in helping people will laugh as they tell many stories of their own stupidity; people who want everyone to be like them will get all huffy—how dare you suggest that they are not rational! This method has not failed me yet.

"Any reason why people couldn't pursue their preferences themselves?"

How many people never eat more than they should, and mean to eat healthy—but then don't? How many people save as much as they feel they should for retirement? The number is terrifyingly small.

Humans make consistent and predictable errors in judgement. Frances asks why anyone would care about another's choices; I answer, "For the same reason I want poor children to be able to eat 3 meals a day and get a good education, and as equal a chance at opportunity to live a good life: compassion and a sense of justice."

In Nudge, Thaler talks of two things: first, there always needs to be a default option, and when there is research about which option is better, make the best option the default. So, make organ donations opt-out instead of opt-in. Instead of trying to ban huge pop containers, like in New York, make smalls the default option in any meal. Second, framing matters, and we cannot avoid framing, so we should choose framing that benefits people: In school cafeterias, make it so that the healthy options are at eye-level, and the junk food is above. And we have to frame things in some way or another: do we say 1 in 10 people will die, or 10% of people will die? We know that saying 1 in 10 will push people away from the thing we're talking about. So, if we want to play up the risk, we say 1 in 10. If we want to downplay it, we say 10%. But we have to say one or the other. We have to present the information, just like we have to put the food somewhere in the display at the cafeteria.

I feel that this is a better alternative to consumption taxes, because consumption taxes are more intrusive. Yes, when there is an externality, pigouvian taxes all the way. And if junk food, for example, creates higher health care costs, putting a pigouvian tax on them for the amount of the externality can make sense. But taxing junk food in order to do social engineering? You're just punishing people for a lifestyle choice that they may really want. There are many people who intentionally get fat because that's the way they want to be (ex: http://en.wikipedia.org/wiki/Fat_fetishism ). There are people who derive more pleasure from eating junk food than they would from a slightly longer life without it. Why should they be punished by being made poorer? Because that is what a social-engineering consumption tax does. Instead, put the healthy food in the obvious areas, and the junk food in the less-obvious areas. People can still choose whatever they want, with no punishment, but people are much more likely to choose the healthy option. Thus, I avoid Frances' argument that I need to know each person's happiness function. I don't. I can generalise from research, and make my best guess. The information must be presented; there must be a default option. We choose the one the research shows is more likely to be the one people want.

I add to that that people often fall into bad habits when their own happiness would be maximized by doing the thing they express a desire to do. How many people want to lose weight, and spend a lot of money on things, only to fall back into their bad habits? That's why things like StickK are so effective (http://www.stickk.com/about.php). I think that government can step in to help in much the same way by establishing systems to help people meet their stated objectives, especially in regards to things that alter behaviour, like drugs. Hence my alcohol example, in which the government helps people stick to their expressed limit.

"There is a lot of discussion right now about libertarian paternalism. It may be helpful to think, systematically, about why one person ever has any reason to want to interfere with another person's consumption choices."

I should also add, it's not just other people's choices I want nudged. I also want my own nudged. Part of the reason I am a libertarian paternalist is that I think libertarian paternalism will make my own life better.

Shangwen: the main reason alcool consumption has plunged maybe that life is finally bearable...

Daniel I. Harris: most people are very bad at judging what's best for them. It's only that I am usually even worse at guessing what they need. Anyway, we can take comfort from the words of a famous english philosophers school:" You don't always get what you want but you sometimes get what you need"..

Daniel, thanks for a great explanation of good Libertarian Paternalism.

I especially like the opt-in / opt-out issues, and what is the default. Like with retirement, and also organ donations. Also food placement. There MUST be a default, it should more often be the "good for society" when it involves gov't. Like in a gov't cafeteria.

When the Lib Pater becomes no-choice Paternalism only, like Obamacare "tax" on the healthy, it's bad. And most proponents of Lib Pater seem more interested in more paternalism rather than more choice. (Your alcohol limit card is clever cute.)

But, in a non-gov't, profit maximizing place, expect to see the high-profit margin junk food stuff in the easiest to see & buy places. And I support such choices.

Tom Grey - "When the Lib Pater becomes no-choice Paternalism only, like Obamacare "tax" on the healthy"

Obamacare isn't about libertarian paternalism. It's about reducing risk of financial catastrophe/inability to afford card at an individual level and pooling it across the population, plus getting a grip on what is the most obscenely expensive health care system in the world. Yes, some people lose and some people gain.

But the freedom of not having to worry about being able to afford health care is pretty sweet. Despite all of the US rhetoric about freedom, Americans don't actually feel all that free - certainly no more free than Canadians do - see this old post http://worthwhile.typepad.com/worthwhile_canadian_initi/2013/04/were-free-up-here-too-eh.html

Fran: the american health care system is wonderful except for three small defects: it's not a system, it's not about health and it doesn't care.
Many insignificant posts in the last few hours. Just celebrating my return from TypePad hell.

Shouldn't the Coase Theorem handle smoking indoors? The proprietor would set smoking policies in a profit-maximizing manner at restaurants/bars. But we have laws prohibiting smoking in such places precisely to prevent proprietors from allowing smoking in cases where it's preferred.

The comments to this entry are closed.

Search this site

  • Google

Blog powered by Typepad