Friday, February 22, 2013
This is an interesting review by famous nudgemeister Cass Sunstein. It has to do with a book entitled: Against Autonomy: Justifying Coercive Paternalism. Cambridge U. Press, weighing in at a hefty 95 clams. (That price ain’t exactly nudging me into considering a purchase, I’ll tell you that much. So, until that price does come down I’ll rely on the précis given by Sunstein.)
First a Sanka Freeze dried version of the most interesting kernel of the argument contained in the book, some reaction, and them some cut and paste of the Sunstein review itself.
Sanka freeze dried version of the bold aspect of the argument:
A. Human beings have various end states they desire to attain.
B. Most of these desired ends have some sort of impact on others. Because of this, states can have defensible reasons to intervene in the attainment of these ends. (Mill’s Harm Principle)
C. Some ends do not have such ramifying effects on the public welfare. The only harms, in such cases, are harms incurred by the individuals in question. (We concentrate on this latter sort of case because cases like these form the material for the boldest claim in the book, according to the reviewer. Call these Insulated Harm Cases, or IHC)
D. In IHC cases people not only fail to do things that will help them attain the end states they desire, but they do things that will harm that attainment, make it more difficult or impossible. They work in directions not only contrary to those ends, but antithetic to them. (Note: we are not talking about activities that do not help to attain the ends, but also do not hinder, something we might call ‘neutral activities.’ These are not under discussion in the review or book.)
E. In these IHC cases, the author maintains the state should intervene in the lives of such individuals in such ways as will constrain their choices, so that they will no longer work contrary to their own desired end states. The set of methods the state can use includes less controversial and more controversial elements. The author advances that, in the right circumstances, two from the more controversial subset can be aimed at IHC cases, that is: Making certain actions or items illegal (banning them), or mandating that individuals obtain certain items or take certain actions.
The state can only intervene in IHC situations if the following conditions are met:
1. A utilitarian calculation shows that the benefits generated outweigh the harms caused.
2. The people being coerced must actually desire the attainment of the ends in question.
3. The measures must actually work.
Now, if you run some of the less controversial methods used by states through this rubric, you’ll get some results that are Nanny state-ish but not glaringly counter-intuitive. That is, you’ll see some basis for regulatory constructions that do not ban items but still constrain choice in regard to them.
But, if you focus on more controversial methods used by states, and their application to more controversial practices (smoking for instance) the author believes you can make a case for “banning” the practices and associated items. (Actually, there are indications in the review that the author displays an ambiguous attitude to banning, seeming to consider it family related yet somehow different from prohibition, something she does not recommend. Yet, no work is done in the review to explain the different attitudes toward the two, nor any conceptual distinctions. I come away from the review thinking the distinction is confused, and is somehow based upon differing utilitarian estimates. If that is so, then the issue between banning and prohibition seems largely semantic. If the utility is what counts or sorts between the two, it doesn’t matter if we call it ‘banning,’ ‘prohibition’ ‘Al,’ ‘Johnson,’ or ‘Susquehanna Hat Company’ does it? But, be that as it may, I want to concentrate on the bold claim made in the review, and presumably by the author, about truly IHC cases. So, that will be the focus from here on.)
As I hinted above, it is pretty hard to find true IHC cases. Most of the things we do have some impact on others (family most obviously). But, nothing prevents us from constructing a hypothetical. So, here goes:
Fred is independently wealthy, and provides his own health care (hires his own private and exclusive doctor). Fred desires to be healthy. Fred also loves to smoke. He clearly recognizes the fact that his smoking is not good for his health. Fred has a great philanthropic streak and has invested in such a way that he generates more than enough benefits to offset any subtractions he might make from the overall good. He lives green to boot (except for the devil weed addiction). Fred is also a great respecter of the law. If a law is passed, he’ll obey. Now, seeing this, the state passes a law that bans Fred from smoking. Only Fred. It also passes a binding law that constrains itself from ever expanding the scope of this law. The public believes this, so no public angst is created at the possibility of a universal ban..er..prohibition being the next step down the proverbial slippery slope.
Looking at the 3 conditions above, his case passes each. The benefits he will accrue from quitting smoking are well attested. He will, no doubt go through the trials of withdrawal, might resent the imposition & etc. But his life will be longer, he will recognize the act is really in his own interest, and, most tellingly, he sees that it is in line with what he truly believes is in his long term interest, what he truly wants to do. Lastly, the measure will work, because Fred is a good law abiding citizen.
So, it follows, to paraphrase the language of the review, (which we must presume does reflect the language of the book): ‘Even when there is only harm to self, the government may and indeed must act paternalistically so long as the benefits justify the costs.’ (Emphasis mine)
Does this pass the intuition smell test? Do you think that the government would be acting ethically in doing this? If you do not, then this is an accurate description of your result: Despite the utilitarian support for such an action, you think it is morally wrong. It follows that there is something else going on in your judgment than utilitarian calculus. You value Fred’s autonomy perhaps. You might think, regardless of Fred’s inability to stop himself from indulging the Devil’s little imps, that outside entities simply do not have a right to take over his decision making, or remove options without his consent, especially given that he is a true IHC case. You may even think, along Rawls lines and come to the conclusion that you would not agree to such schemes regarding IHCs while you are behind some sort of pre-incarnate veil of ignorance hammering out the terms of social contract.
None of this has even begun to engage other problems with the view. The author seems to place great reliance upon the fact that modern psychological and behavioral studies suggest that individuals do not always act rationally in choosing means to their ends, while at the same time she places great reliance upon the abilities of experts in various fields being able to determine either what is in general good for individuals, or what other individuals really want for themselves. Yet, we see another ambiguity in the argument presented even in regard to this. Consider this sentence from the review: “Conly emphatically rejects “perfectionism,” understood as the view that people should be required to live lives that the government believes to be best or most worthwhile.”
One might object, on behalf of Conly, when considering the Fred case, that the government is not coming at him from the ‘perfectionist’ point of view, not passing a law intended to get him to follow every jot and tittle of some vision of the best or most worthwhile life. No, she could protest, we envision the government as passing a law that only deals with his health. That is all. If you really want to get down to brass tax, it has only to do with lung and heart health, and perhaps a judgment that a longer life is more valuable than a shorter life. What is more, Fred agrees. So, we do not fall into that perfectionist trap.
I would say, even in light of all of this, it would be overstepping moral bounds to pass this little bit of legislation. Quite aside from the Fred case, assumptions are made here that Sunstein admits are universalist in flavor, such as that all people would prefer longer healthier lives over shorter, more intensely pleasurable lives involving use of tobacco, alcohol, or tasty but unhealthy foods. While it is no doubt true that most people prefer the former, it is by no means obvious that a vast majority do. Given that is the case, then, one wonders if such intrusions are respectful of those folks autonomy. (This despite the fact that some of them may be a financial burden on the health care system, that may impact others.)
Epistemological limitations exist for third persons when it comes to determining what any person ‘really’ wants, limitations which are compounded when we consider collections of people. Aside from interviewing him or her often and in depth, how is a government to gage what a person really wants or values? This problem is obviously compounded when whole populations are considered.
If we gage by behavior, then, as Conly and Sunstein both admit, that behavior does not clearly indicate that an overwhelming majority of folks want to be healthy in their eating habits, for instance. How are we to tell if this is actually a case of most of the population working at cross-purposes with their ‘true’ intentions or values? Interview everyone, or most everyone? Carry out studies that are deemed representative?
And, even assuming we can successfully pull this feat off, what do we do about the minority, or plurality that are not in line with the general consensus? Do we ban anyway? How did that work during prohibition? Did not this very approach end up a utilitarian disaster?
But, wait, we are not talking prohibitions, just bannings. What government bans will depend on projected social unrest I suppose. It’s true that banning trans-fats probably won’t cause the populace to rise up, where banning..er..prohibiting beer did (and would). So, what we ban is going to be determined on a case by case basis utilizing such utilitarian projections of social affect, combined with state of the art scientific data as to the utility of said items for what some more state of the art data tells us are the true values or ends of some majority of citizens? And, this is supposed superior to allowing adults to make their own choices.
But that same sort of data already shows epistemic or motivational limitations exist in the first person. If we really do not know what is good for us individually either as means or ends, why believe that someone looking in from the outside can figure it out, using scientific and psychological methodologies aimed only at externally observable behavior? If we cannot motivate ourselves, why believe that a complex society of individuals can be any more effectively motivated by clever nudges or bans?
Furthermore, why be naïve about governmental wisdom as to ends or self-awareness with regard to motivations, and best means toward ends? Why believe that legislative bodies can use the information, assuming it is accurate, in a way devoid of institutional self-interest, political or financial? Why also, believe that scientists undertaking studies under the auspices of, and funded by government grants, will unerringly strive to be objective, rather than have their methods and perceptions colored by a self-interested motivation to provide the funder with what the fundee believes the funder wants to hear?
In short: It is awfully naïve to assume a sort of ability in the individuals that make up these institutions, that one spends a great deal of time demonstrating does not exist with equal perspicacity in the general populace. Great hay is made in defense of the proposition that you cannot trust corporately funded research on matters germane to corporate self-interest. By equal reasoning, you can no more trust government funded research on matters germane to governmental self-interest or researcher self-interest. What better way is there to insure the permanence of the institutions that execute legislative or regulatory paternalism than to present studies that show how effective it can be in furthering the will of the people and the common good? What is more, before we commit to believe in a body’s ability to clearly perceive and attain long term best interests, should we not examine the behaviors of such bodies, and grow suspect if they display inability to weigh long term over short term gain in other regards? The present financial situation with regard to government should give pause, for precisely this reason.
That’s it for reaction. Now to the excerpts from Sunstein’s review:
Until now, we have lacked a serious philosophical discussion of whether and how recent behavioral findings undermine Mill’s harm principle and thus open the way toward paternalism. Sarah Conly’s illuminating book Against Autonomy provides such a discussion. Her starting point is that in light of the recent findings, we should be able to agree that Mill was quite wrong about the competence of human beings as choosers. “We are too fat, we are too much in debt, and we save too little for the future.” With that claim in mind, Conly insists that coercion should not be ruled out of bounds. She wants to go far beyond nudges. In her view, the appropriate government response to human errors depends not on high-level abstractions about the value of choice, but on pragmatic judgments about the costs and benefits of paternalistic interventions. Even when there is only harm to self, she thinks that government may and indeed must act paternalistically so long as the benefits justify the costs.
Conly is quite aware that her view runs up against widespread intuitions and commitments. For many people, a benefit may consist precisely in their ability to choose freely even if the outcome is disappointing. She responds that autonomy is “not valuable enough to offset what we lose by leaving people to their own autonomous choices.” Conly is aware that people often prefer to choose freely and may be exceedingly frustrated if government overrides their choices. If a paternalistic intervention would cause frustration, it is imposing a cost, and that cost must count in the overall calculus. But Conly insists that people’s frustration is merely one consideration among many. If a paternalistic intervention can prevent long-term harm—for example, by eliminating risks of premature death—it might well be justified even if people are keenly frustrated by it.
To Mill’s claim that individuals are uniquely well situated to know what is best for them, Conly objects that Mill failed to make a critical distinction between means and ends. True, people may know what their ends are, but sometimes they go wrong when they choose how to get them. Most people want to be healthy and to live long lives. If people are gaining a lot of weight, and hence jeopardizing their health, Conly supports paternalism—for example, she favors reducing portion size for many popular foods, on the theory that large, fattening servings can undermine people’s own goals. In her words, paternalism is justified when
the person left to choose freely may choose poorly, in the sense that his choice will not get him what he wants in the long run, and is chosen solely because of errors in instrumental reasoning.
Because of her focus on the means to the ends people want, Conly’s preferred form of paternalism is far more modest than imaginable alternatives.
At the same time, Conly insists that mandates and bans can be much more effective than mere nudges. If the benefits justify the costs, she is willing to eliminate freedom of choice, not to prevent people from obtaining their own goals but to ensure that they do so. Following a long line of liberal thinking, and in a way that responds directly to potential objections, Conly emphatically rejects “perfectionism,” understood as the view that people should be required to live lives that the government believes to be best or most worthwhile.
Because hers is a paternalism of means rather than ends, she would not authorize government to stamp out sin (as, for example, by forbidding certain forms of sexual behavior) or otherwise direct people to follow official views about what a good life entails. She wants government to act to overcome cognitive errors while respecting people’s judgments about their own needs, goals, and values.
For coercive paternalism to be justified, Conly contends that four criteria must be met. First, the activity that paternalists seek to prevent must genuinely be opposed to people’s long-term ends as judged by people themselves. If people really love collecting comic books, stamps, or license plates, there is no occasion to intervene.
Second, coercive measures must be effective rather than futile. Prohibition didn’t work, and officials shouldn’t adopt strategies that fail. Third, the benefits must exceed the costs. To know whether they do, would-be paternalists must assess both material and psychological benefits and costs (including not only the frustration experienced by those who lose the power to choose but also the losses experienced by those who are coerced into something bad for them). Fourth, the measure in question must be more effective than the reasonable alternatives. If an educational campaign would have the benefits of a prohibition without the costs, then Conly favors the educational campaign.
Applying these criteria, Conly thinks that New York’s ban on trans fats is an excellent example of justifiable coercion. On the basis of the evidence as she understands it, the ban has been effective in conferring significant public health benefits, and those benefits greatly exceed its costs. Focused on the problem of obesity, Conly invokes similar points in support of regulations designed to reduce portion sizes.
She is far more ambivalent about Mayor Bloomberg’s effort to convince the US Department of Agriculture to authorize a ban on the use of food stamps to buy soda. She is not convinced that the health benefits would be significant, and she emphasizes that people really do enjoy drinking soda.
Conly’s most controversial claim is that because the health risks of smoking are so serious, the government should ban it. She is aware that many people like to smoke, that a ban could create black markets, and that both of these points count against a ban. But she concludes that education, warnings, and other nudges are insufficiently effective, and that a flat prohibition is likely to be justified by careful consideration of both benefits and costs, including the costs to the public of treating lung cancer and other consequences of smoking.
Conly is right to insist that no democratic government can or should live entirely within Mill’s strictures. But in my view, she underestimates the possibility that once all benefits and all costs are considered, we will generally be drawn to approaches that preserve freedom of choice. One reason involves the bluntness of coercive paternalism and the sheer diversity of people’s tastes and situations. Some of us care a great deal about the future, while others focus intensely on today and tomorrow. This difference may make perfect sense in light not of some bias toward the present, but of people’s different economic situations, ages, and valuations. Some people eat a lot more than others, and the reason may not be an absence of willpower or a neglect of long-term goals, but sheer enjoyment of food. Our ends are hardly limited to longevity and health; our short-term goals are a large part of what makes life worth living.
Conly favors a paternalism of means, but the line between means and ends can be fuzzy, and there is a risk that well-motivated efforts to promote people’s ends will end up mischaracterizing them. Sure, some of our decisions fail to promote our ends; if we neglect to rebalance our retirement accounts, we may end up with less money than we want. But some people who often rebalance their accounts end up doing poorly. In some cases, moreover, means-focused paternalists may be badly mistaken about people’s goals. Those who delay dieting may not be failing to promote their ends; they might simply care more about good meals than about losing weight.
Freedom of choice is an important safeguard against the potential mistakes of even the most well-motivated officials. Conly heavily depends on cost-benefit analysis, which is mandated by President Obama’s important executive order on federal regulation.14 It is also a crucial means of disciplining the regulatory process.15 But the same executive order emphasizes that government agencies must identify and consider approaches that “maintain flexibility and freedom of choice for the public.” Officials may well be subject to the same kinds of errors that concern Conly in the first place. If we embrace cost-benefit analysis, we might be inclined to favor freedom of choice as a way of promoting private learning and reflection, avoiding unjustified costs, and (perhaps more important) providing a safety valve in the event of official errors.
Conly is quite aware of the many difficulties that would be associated with efforts to prohibit the manufacture and sale of alcohol and cigarettes, but here the problems seem to me more significant than she allows. True, smoking produces extremely serious public health problems—over 400,000 deaths annually—and it is important to take further steps to reduce those problems.16 But any ban would raise exceedingly serious difficulties, not least because it would be hard to enforce. A full analysis would have to consider such difficulties, as well as the claims of free choice. Black markets in cigarettes are not exactly what the United States most needs now.
Notwithstanding these objections, Conly convincingly argues that behavioral findings raise significant questions about Mill’s harm principle. When people are imposing serious risks on themselves, it is not enough to celebrate freedom of choice and ignore the consequences. What is needed is a better understanding of the causes and magnitude of those risks, and a careful assessment of what kind of response would do more good than harm.