Skip

To Nudge or Not to Nudge
April 14, 2008 3:33 PM   Subscribe

Richard Posner, Gary Becker and Cass Sunstein debate "Libertarian Paternalism", a fancy and ugly term used to describe how some behavior economists think recent findings about human fallibility might bear on law and public policy. (previously)
posted by AceRock (30 comments total) 8 users marked this as a favorite

 
The Economist called it "the avuncular state", which is about 1000x more charming as a piece of terminology. Dan Lockton refers to "architectures of control". He hasn't posted much lately, but when he does, it's pretty good.
posted by mr_roboto at 3:42 PM on April 14, 2008 [3 favorites]


I see a big deal on the horizon here. Let's say that the future public just doesn't see the point in diverting billions to take care of all of those people in the legal limbo of a comatose state, or in any state of vegetation - and not just because of money, but because of ethical concerns regarding suffering (although limited funds makes money an ethical concern too). So, the default setting for public health care is then set to terminate, and not just unplug to a starvation mode, but terminate with a pink shot, unless a family steps forward and agrees to pay for some of it. The big deal about this is that it opens the door to a cultural acceptance of letting conscious people make that personal decision on their own in a state of chronic illness.
posted by Brian B. at 4:14 PM on April 14, 2008 [1 favorite]


Cass Sunnstein is also blogging on the Volokh Conspiracy about libertarian paternalism. I think there is only one post up now, but there should be more throughout the week.
posted by Falconetti at 4:23 PM on April 14, 2008 [1 favorite]


There's an almost grudgingly glowing review of Sunstein's and Richard Thaler's book over at Freakonomics, from last Friday.
posted by cortex at 4:34 PM on April 14, 2008


wink wink, nudge, nudge
posted by jammy at 4:35 PM on April 14, 2008


I would tend to agree with Posner and Becker on this: using one's own value judgments to affect the choices of others is just as objectionable--if not more so--if it is done subtly through "choice architecture" than straight up coercion. Such scheming always, of necessity, involves substituting the values of the bureaucrat for the consumer, in a way which is compatible with personal autonomy.

And related arguments that individualism is somehow failing because it refuses to acknowledge the dependencies of people and communities upon each other don't resolve the question such arguments raise, i.e. if each person is not permitted to choose for themselves, exactly who is going to choose for them? And how does that person gain such power? Are they elected? By a majority? What of the minority? Or are they self-selected on the basis of simply being right? I don't think "being right" is sufficient grounds for such an enterprise.

Such suggestions quickly dissolve to reveal the underlying argument: "I should get to tell you what to do." Scratch any revolutionary and you find a closet aristocrat.
posted by valkyryn at 4:38 PM on April 14, 2008 [3 favorites]


Dang it, that should be incompatible.
posted by valkyryn at 4:39 PM on April 14, 2008


Scratch any revolutionary and you find a closet aristocrat.

Not even in spitting distance of true. Try not to paint with such a broad brush.
posted by languagehat at 4:49 PM on April 14, 2008 [1 favorite]


Ok. Scratch some revolutionaries and you find closet aristocrats.
posted by Slap Factory at 5:00 PM on April 14, 2008


Such scheming always, of necessity, involves substituting the values of the bureaucrat for the consumer, in a way which is compatible with personal autonomy

What I understand you saying is this: "It's better for the government to force me to put money into the Social Security trust fund than for it to sign me up by default for some kind of retirement plan that I can choose to opt out of."

Am I correctly representing your position?
posted by Slothrup at 5:08 PM on April 14, 2008


if each person is not permitted to choose for themselves, exactly who is going to choose for them

Isn't it the idea that people are "choosing" for themselves because paternal libertarianism would be nudging people towards what they really want based on social science and that the reason they aren't making the choices that they themselves consider best is because people aren't totally rational when making decisions. If one accepts that as true, and to a certain degree it must be, then the question is more not whether we are letting bureaucrats substitute their judgment for our own, but rather how do we "scientifically" determine what people think is in their best interest and how do we make sure that the bureaucrats implement those determinations rather than their own judgment. I think there are so many problems in those premises that instituting the avuncular state would be pragmatically impossible on a grand scale, but may be a good metric to incorporate when deciding what the "ground rules" of a program are (defined contribution plans, organ donorship, etc.).
posted by Falconetti at 5:14 PM on April 14, 2008


languagehat, though I'll freely admit to the hyperbole, I think there's some truth there. I do believe that the old rallying cry that those in power should not have it was really a backhanded way of saying that the agitators are the ones that should have it.

It isn't difficult to see such a phenomenon in "libertarian paternalism." Arguing that people would make better choices if the context of their choice was changed is a thinly-veiled way of saying that the person doing the context-changing should be making those choices instead.
posted by valkyryn at 5:15 PM on April 14, 2008


Slothrup, I don't think so, because I didn't accurately state my position. See my second comment.
posted by valkyryn at 5:16 PM on April 14, 2008


languagehat, though I'll freely admit to the hyperbole, I think there's some truth there. I do believe that the old rallying cry that those in power should not have it was really a backhanded way of saying that the agitators are the ones that should have it.

languagehat was surely alluding to anarchism, and, the easy irony of Fresh Prince Kropotkin aside, there's a long intellectual and practical history of revolutionaries for whom the elimination of all state power (and all coercive power in general) is the larger part of the point.
posted by dyoneo at 5:33 PM on April 14, 2008


Becker's argument strikes me a textbook example of the libertarian tendency to ignore all the data contradicting their position, on the assumption that no one has implemented real libertarianism yet. His first objection- that preventing people from screwing up prevents them from learning- is basically tautology. If the evidence shows that people are making good decisions on their own, libertarianism is working, and if it shows them screwing up, you should give them less help so that they learn faster. It's unfalsifiable. Unless, of course, you have an example of people making poor decisions in a certain context, with significant penalties, repeatedly, for decades- like, say, cigarette smoking- in which case you define the problem away, saying that the individuals studied are by definition being rational, since they chose to smoke, and call it "rational addiction", as Becker does.

His second objection, that libertarian paternalism starts a slippery slope to straight up paternalism, is only an objection to libertarians. He finds it difficult to believe that anything might make Americans in 1950 more or less rational on any issue at all than Americans today or that anything might make Americans today more or less rational on any issue than people of other nationalities. Basically, culture doesn't exist, and all people everywhere have equal capabilities, and have since time immemorial. That's a useful simplifying assumption when creating mathematical models of human behavior, of course, but when the data doesn't match Becker's model, he throws out the data. He ridicules, without any particular reason, the idea that culture might systematically affect human behavior, and doesn't even provide an alternative explanation for the examples he chooses.

His third objection is that libertarian paternalism still requires substituting the rule-maker's judgment for an individual's, which is dangerous. Of course, this completely ignores the impetus behind libertarian paternalism, which is that we have actual fucking data showing that people are, for example, losing thousands of dollars by not taking 60 seconds to sign a form to enroll in their employer's retirement program. But Becker wonders whether we're sure that the person's judgment wasn't actually rational. The supporters of libertarian paternalism aren't suggesting that the government should start nudging people randomly on whatever issues suit the bureaucrat's fancy, they're suggesting that we nudge on the issues for which actual research shows that people are actually being irrational.

Although I read the Becker-Posner blog regularly, I have no particular familiarity with Becker's work. But his arguments here exemplify everything that advocates of heterodox economics dislike about the classical University of Chicago approach.
posted by gsteff at 5:40 PM on April 14, 2008 [10 favorites]


I would tend to agree with Posner and Becker on this: using one's own value judgments to affect the choices of others is just as objectionable--if not more so--if it is done subtly through "choice architecture" than straight up coercion.

But any government policy or architectural design must have some "default" behavior. Why not chose that behavior such that it benefits the individual, or society as a whole? What you seem to be ignoring is that a "gentle coercion" exists even in the status quo: this new movement in behavioral economics seeks to recognize it and redirect it to benefit the common good.
posted by mr_roboto at 5:42 PM on April 14, 2008 [3 favorites]


mr_roboto, changing default behaviors is one thing, but the articles discuss significantly more invasive changes than that. Signing up employees automatically for retirement contributions is one thing, but there you've got only two possible "defaults": either you're automatically signed up or you automatically aren't. But, for example, arrangement of food serving lines doesn't have a "default" position; it's all designed. One can have a discussion about what constitutes "good" design or "bad" design, and even have a discussion about what the design should be for and should encourage/discourage or if it should attempt to do either. But those are evaluative decisions which involve balancing a series of value choices against each other. Having someone other than the person making the choice and the person offering it decide the "best" way to present the choice bugs me.

Ultimately, cafeteria lines have to be arranged somehow--piling the food on the floor in the middle of the kitchen wouldn't work very well--so because some value system has to be imposed I'm not particularly concerned about which one it is. But there are other examples in the article, say, for example, disclosures before medical procedures or on financial documents, which aren't so cut-and-dried. I'd rather my doctor, someone with whom I have a personal relationship and has assumed a duty of care to me, make decisions about how to present medical options than have a bureaucrat in state capital or outside DC do it for him.

I do appreciate and approve of the idea that presenting people with an option is never a "neutral" thing; the way it is presented can affect the outcome. And for certain things, which I could identify but not enumerate, this is probably pretty unobjectionable. But I'm very leery of any attempt to use this fact to achieve policy goals.
posted by valkyryn at 6:20 PM on April 14, 2008


But, for example, arrangement of food serving lines doesn't have a "default" position; it's all designed.

What I mean by "default" is that once the design has been accomplished, the serving line, for instance, has some fixed arrangement that corresponds to a default behavior. The default is a condition of the interaction of behavior with architecture after the design has been fixed, not some ideal state that precedes design.

Ultimately, cafeteria lines have to be arranged somehow--piling the food on the floor in the middle of the kitchen wouldn't work very well--so because some value system has to be imposed I'm not particularly concerned about which one it is.

Exactly.
posted by mr_roboto at 6:26 PM on April 14, 2008 [1 favorite]


This is a philosophical position on "rulership" (whether it be dictatorial, democratic, or emergent) that I've held with little variation for decades, and can be summarized as: Choose rulers on capacity to deal with past, present, and future problems. As ruler, decide, based on meaningful data, what you want people to do. Ensure that this is within human capabilities. Model the likely outcome of people doing that. Gather data about what they currently do. Figure out how you can change what they do, to what you want them to do, with the least disruption for the greatest gain. Explain your intentions. Try the change. Gather data about what effect your changes had. Question your desired outcome. Measure the ruler's performance. Repeat.

The current system can be summarized as: Choose rulers on desire to be ruler, adjusted for capacity to loudly assert idealist fantasy and/or defeat other contenders. As ruler, decide, based on the aforementioned idealist fantasy, what you want people to do. Tell them to do it. Punish them for not doing it. Reward them for doing it, if you guess it will help. Gather data about how and whether they are obeying. If they are not, punish them more and/or reward them more. Repeat.

Several things are givens, no matter what. Firstly, people will rule and be ruled. Libertarianism and anarchy are basically fantasies. It's about time we chucked them both into the same bucket of history that manifest destiny and rigid caste are in.

Secondly, people will rule and be ruled according to self-interest first, idealism second, practicality a distant third. The appropriate response to that, in my view, is to strongly tie idealism to practicality, ie devise an "idealism of practicality", a "belief in what actually works", and tie self-interest to pursuing that idealism. One way to help with that is ruthless disparagement and mockery of positions that are clearly not based on gathered fact, and not subject to alteration as circumstances change. Relentlessly point out that not only is the Emperor naked, his endowment is small, and he is in very poor physical shape. The paramount virtue of such a society must be honesty. Dishonesty, whatever the motivation, taints the data. If the data is tainted, the conclusions are suspect.

Thirdly, people are very, very used to being ruled according to the current system. They will assume that, if the train timetable is altered, if the food import tariffs are changed, then the change is intended to be that way forevermore. Thus, if they don't like it, they had better kick up an enormous fuss, or else it will stay. The switch to leadership based on "choice architecture", or social engineering, will itself need to be engineered. Any group within a nation intending to have that nation adopt this method, will need to push the idea of "trials", and not in the criminal sense, and repeatedly demonstrate alterations in policy according to the results of those trials.

Fourthly, whatever changes, some people will gain, and some suffer. Traditionally rulers overcome this problem by (a) ensuring those who support them gain, and those who do not, suffer; (b) lying about it: "everyone will be better off ..." Anyone who suffers can be expected to resist change, to some extent; those who gain, to support change, to some extent. However, "choice architecture" as envisioned ought to have the advantage of passing the test of fairness. It's an inherent requirement, to present a policy honestly, that those who are expected to suffer, and to gain, should be told of these expectations. How many of us, how much, and why. Everything done under this view of leadership ought to, even in the minds of those who suffer, be fair. This isn't an unreasonably high bar to set: people can be, if engaged properly, a fair-minded species. It's just that for a long time, for a lot of reasons, our rulers have been committed to unfair-mindedness as a strategy to retain power.
posted by aeschenkarnos at 6:48 PM on April 14, 2008 [3 favorites]


aeschenkarnos, that sounds alright, but I'm deeply skeptical about our ability to gather data about why people do the things they do. People are unpredictable, beset by mood, and subject to an array of forces large enough in number and subtle enough in nature that controlling for them in a scientifically significant way is impossible. What you're could well be described as a totalitarian state, which didn't work very well for either Russia or China.

Still, the kind of technocratic system you describe actually manages to come reasonably close to the way many federal agencies behave, but I think we all know just how well that works. There are questions of agency capture, debates about whether technocratic or public-choice models better describe agency behavior, and anxiety about political responsiveness of unelected administrators.

The "scientific" form of government you describe is a bit too fanciful for my tastes, assuming, as it does, that people are capable of acting apart from their self-interest long enough to rule in a "scientific" fashion.
posted by valkyryn at 7:45 PM on April 14, 2008


using one's own value judgments to affect the choices of others is just as objectionable--if not more so--if it is done subtly through "choice architecture" than straight up coercion.

You know, it's really easy to say that. But I'd be hard-pressed to defend this view.

I mean, forcing people to do things is worse than letting people retain their freedom but setting things up in a way that you realize will likely result in a subconscious tendency to choose one thing over the other. The former case eliminates freedom altogether; the latter maintains freedom and just involves a subtle influence.

That doesn't mean you have to like the idea of libertarian paternalism, of course, but I don't see the point in saying it's no better than coercion, when actually, it is better than coercion because it results in more freedom.
posted by Jaltcoh at 7:57 PM on April 14, 2008


I'd rather my doctor, someone with whom I have a personal relationship and has assumed a duty of care to me, make decisions about how to present medical options than have a bureaucrat in state capital or outside DC do it for him.

I think Sunstein's use of words like "nudge" and "paternalism" may be confusing matters. In this example, the state government cannot be neutral in regard to its influence over the context in which your doctor makes his decisions; if it decides to be minimally interventionist that is an equally active form of choice-architecture as if it decides to be maximally interventionist. The Sunstein position really seems to amount mainly to acknowledging that this is true, and since it's true, we might as well build choice architectures consciously instead of haphazardly. I think this does offer a real opportunity to bridge liberal-conservative divides, as the authors argue, but it won't bring libertarians on board, mainly due to libertarianism being a fantasy politics built on fantasy premises.
posted by game warden to the events rhino at 8:14 PM on April 14, 2008 [1 favorite]


I'm suspending judgment on the theory behind all this until I do more reading, but in the meantime, why would anybody deliberately choose the term "libertarian paternalism" for their theory? Okay, I suppose one can argue whether or not this falls under libertarianism, but paternalism is such negatively loaded vocabulary, jeez.
posted by bettafish at 8:27 PM on April 14, 2008


I mean, forcing people to do things is worse than letting people retain their freedom but setting things up in a way that you realize will likely result in a subconscious tendency to choose one thing over the other.

This is tricky I think. Slavoj Zizek, whose Lacanian-Marxism is ridiculous, but who is entertaining and does make some valid observations tells the parable of the child who has to visit his grandparents. A parent might tell the child: "We are visiting your grandmother today and that is that." The parent is forcing the child to do something, but they are being honest. The more insidious tactic is to say "You know, your grandmother loves you very much and she would be so happy if you would visit her today. Wouldn't you like to see your grandmother?" In the latter case, not only must you do the thing, you must like it. I would be scared that consciously tweaking unconscious choices on a grand scale would slip easily into this creepy nether realm where the state does more than guide our choices but ends up guiding our thoughts. I am getting a little far afield when it comes to Sunnstein's "liberatarian paternalism" I realize, but I don't know if the quoted sentence is really true.
posted by Falconetti at 8:58 PM on April 14, 2008


I'm looking forward to seeing libertarians come out against Corporate advertising. Never in history has a population been subjected to a more intrusive and ubiquitous pshycological conditioning program with no consent than the current state of advertising.
posted by afu at 9:08 PM on April 14, 2008 [1 favorite]


You forgot about grade school.
posted by b1ff at 10:51 PM on April 14, 2008


All these absolutes, pshaw. The basic libertarian philosophy is great, but, like pure capitalism, it sometimes leaves society with a less than satisfactory result. Libertarianism is just a goal, not some silly moral absolute without other consequences to consider. Fuck Posner anyway; the guy is just too predictable, lacking in creative thought, like a machine into which you pour your facts and out comes his version of economic efficiency, which is so sad for so obviously brilliant of a mind.
posted by caddis at 11:04 PM on April 14, 2008


The phenomena discussed by Sunstein and Thaler in their book, in which sometimes non-obvious or inconspicuous changes to our surroundings---our environment---our situation---have large impacts on our behavior, are pervasive. And the "default" choices, or "choice architectures", or situations we face are rarely random or accidental. Rather, they are already being designed intentionally for someone's benefit. The problem is whose benefit is currently guiding those design choices, and how oblivious we normally are to the influences to which we are subject.

We speak of "freedom of choice", "free will", "individual choice", "consumer sovereignty", "maximizing utility", "choosing what is in one's own best interests", "personal responsibility", and so on... but we are continually, non-consciously influenced by situational factors designed to steer us toward a particular option. We push the button to shock the guy in the other room to death simply because the person standing behind us is wearing a white coat. We complain about bad drivers without noticing that their driving environment is designed to keep their attention anywhere but on the road. We ridicule the obese or complain that they are seated next to us without noticing that, while the population's waistlines are rapidly expanding, the airplane seats are getting smaller. We argue over smokers' rights and individual choice without ever wondering how, while banning the use of marijuana, we protect an industry that kills almost half a million Americans every year.

Professor Jon Hanson at Harvard Law School, among others, has long been examining and critiquing the ubiquitous manipulation of our situation (unfortunately, the links to his published works are mostly subscriber only). "Situationism" as he uses the term emphasizes the importance of our environment or situation---the non-conspicuous factors that influence our decisions without our conscious awareness---over the chooser's "preferences" or disposition. It's an old idea, but it is now more important than ever. Professor Hanson's situationist site is invaluable for tracking developments in the social sciences related to situationism, and also just for some good reads. I've been wanting to make an FPP of it for a long time, but just never seem to have the time to do it right.

At any rate, I strongly urge and heartily welcome an honest and open discussion about the fact and rationale of such choices, that must of necessity be made as a routine aspect of implementing almost any government program. Better that we as a body politic are aware that we are being manipulated, and that we openly delegate design of our "choice architectures" to people who may, in some small way, be held accountable to us, than to wax poetic on the comfort of the blinders we currently wear. But "accountability" is a problem, and situationist work puts the lie to the easy suggestion that lunch lines be designed to influence children toward healthy choices. The challenge is in designing a government that, by default, chooses such "default" rules, rather than taking the money to put the Coke machine at the cafeteria door. It's no easy task, and our current government is nowhere near it.
posted by dilettanti at 6:43 AM on April 15, 2008 [4 favorites]


languagehat was surely alluding to anarchism

Bingo.

the easy irony of Fresh Prince Kropotkin aside

I hope you're not suggesting that Kropotkin was one of those who wanted power, because that's the opposite of the truth. He couldn't help being a prince.
posted by languagehat at 10:19 AM on April 15, 2008


I think he's suggesting that Kropotkin got in one little fight and his mom got scared and said "you're mov—

Sorry.
posted by cortex at 10:50 AM on April 15, 2008


« Older Brain Cake.   |   The World's Smallest Cell Phone! Newer »


This thread has been archived and is closed to new comments



Post