Against Autonomy: Justifying Coercive Paternalism by Sarah Conly (Cambridge University Press, 2013), 256 pages (ebook edition reviewed).
Bowdoin philosophy professor Sarah Conly has given us a remarkably timely book. Against Autonomy makes an important contribution to the trending discussion of what some call the “nanny state” and others might call simply “petty fascism” (or maybe just “fascism”). It is former Mayor Mike Bloomberg’s efforts to prevent New Yorkers from drinking so much soda, or the ongoing and increasingly aggressive efforts to stop people from smoking, or the various schemes to prevent us from carrying so much credit and saving so little for retirement, and so on and on, that Conly seeks to justify.
Conly’s basic argument is developed in the first chapter and is rooted in an undeniable fact: we (most all of us) are not very good at instrumental reasoning. We are prone to confirmation bias, unreasonable discounting of future consequences, undue optimism about the risks we face, and all the various errors in reasoning that make up the content of a typical course in critical thinking or elementary logic. As a result, we make bad choices: we smoke; we eat unhealthy food and too much of it; we run up our credit-card balances and don’t save for retirement. And those cognitive deficiencies generally are not remedied by more information, or by gentle nudges toward more rational choices. The solution is what Conly calls “coercive paternalism.” We should use the power of government to render some choices simply impossible, or at least very difficult, to make.
Through the middle part of the book, Conly attempts to defend this simple argument against a range of objections. She explains that hers is a paternalism of means, not ends. The problem for which coercive paternalism is a solution is not that people choose the wrong ends, but that they make choices that hinder their ability to achieve the ends they themselves have chosen. For example, we choose to smoke even though we would surely number good health among our ends. She develops a novel understanding of the value of autonomy that enables her to respond to a variety of traditional objections to paternalism or coercion. And she develops criteria for determining whether or not coercive paternalism is appropriate in particular situations.
But the penultimate chapter on applications, particularly the section on cigarettes, reveals the basic problems with her argument. (Full disclosure: I smoked for 35 years, I enjoyed every cigarette I smoked, including the first one, and I quit only because of illness.) Despite advancing an essentially utilitarian argument, she simply ignores all sorts of costs and imagines all sorts of benefits.
Conly concludes that it “seems reasonable to give a ban a try” because banning smoking meets the four criteria for taking a coercive paternalist approach. A ban on smoking would promote good health and longer life, which are important goals for virtually everyone. It is likely to be effective enough to significantly reduce the number of smokers. The benefits very likely outweigh the costs; the misery of being deprived of cigarettes is more than compensated by the health and financial benefits enjoyed by those who are made to quit or prevented from starting. And the failure of other, softer efforts to prevent smoking suggests that coercion may be necessary.
Costs and benefits
Let’s look at those claims in reverse order. That other efforts to prevent smoking have failed does not constitute a very strong reason for supposing either that a ban is necessary or that it will be effective. At various points in the book, Conly mentions the difficulties caused by Prohibition and the war on drugs. Here, she brushes those difficulties off:
Cigarettes, however, are different. For most people, alcohol and marijuana don’t appear to be as harmful as cigarettes, and we know that…. On the other hand, we do accept that smoking is dangerous. A wholesale exclusion of cigarettes from the market would not strike us as a pointless exercise, but as something genuinely protective.
The appeal to the all-knowing “we” and its cousin “most people” is a common ploy in Conly’s book. But there are plenty of people who think liquor and pot are the devil’s own weapons. And large numbers of people in virtually every culture have embraced smoking. How, then, could a ban strike us as anything but a pointless exercise?
And the earlier efforts to ban something for which there is significant demand are also instructive concerning the balance of costs and benefits. If the only cost is the misery of being deprived of cigarettes, then the benefits might outweigh the cost. But, as Prohibition and the war on drugs illustrate, the costs are much greater than that, for everyone. And concerning the effectiveness of banning cigarettes, numerous studies show that the war on drugs has had virtually no impact at all on drug use in this country. In sum, it seems “reasonable” to ban cigarettes only if one ignores the difficulties of implementation.
The first criterion presents a somewhat different problem, more conceptual than practical. It is certainly true that smoking is an obstacle to achieving good health and longevity, ends which most of us may be presumed to have. But that does not mean that no one has chosen smoking as an end, e.g., as part of a life devoted to sensuous pleasures. That may in fact be a “bad” end according to some reasonable standard. It may well interfere with other ends the person has. But smoking is by no means the only thing people choose that suffers from those problems. Maybe people who live beyond their means and who do not save for retirement have decided to live fast, die young, and leave a good-looking corpse. Maybe gluttons have decided that a full measure of life’s sensuous pleasures is more important than longevity. Maybe adrenaline junkies — mountain climbers, race-car drivers, combat photographers, kayakers — have decided that an intense but possibly short life is superior to a long one. Distinguishing means from ends is not as easy as Conly seems to think, and neither is constructing a paternalism of means rather than ends.
A substantial portion of the book’s 256 pages is devoted to rethinking the value of autonomy and individuality, and the meaning of related concepts such as respect and human dignity. Conly is at pains to show that coercive paternalism is not disrespectful or a denial of human dignity. Her arguments are too lengthy and complex to take up in detail here, but I cannot resist a brief response to her discussion of John Stuart Mill.
Mill, of course, is the great icon of anti-paternalism and the view that individual human beings have the right to choose for themselves the life they seek, the goals they will pursue, and the means they will use to reach their goals. According to Conly, Mill assumes that individual persons armed with knowledge and free from coercion will generally choose effective means to reach their ends. But because we are mostly very bad at instrumental reasoning, that assumption is false. And because we reason better when we are in a “relatively objective position,” government can help us out of that dilemma.
I think Mill assumes no such thing. He does say that each person is in a better position than anyone else to decide what will give him the best life. But Mill knows quite well that not all who choose for themselves will achieve their ends. He also says in On Liberty that the current crop of human beings are “but starved specimens of what nature can and will produce.” Mill viewed freedom as a great experiment, by means of which we would find out how to achieve human well-being.
The point here is, rather, that we can indeed expect others, and particularly those in government, to take a “relatively objective position” with regard to our well-being, while they continue to take a very subjective position with regard to theirs. Just look at the operations of your favorite federal bureaucracy. Conly seems utterly oblivious to that. She says that her account of coercive paternalism, since it justifies coercion only to help people achieve ends they themselves have chosen, does not justify incompetent or self-serving coercion. And of course, that is correct. But I am inclined to ask why we should expect there to be much coercion that isn’t incompetent or self-serving. There certainly doesn’t seem to be much now.
I could advance other criticisms of Conly’s arguments, but they all derive from the same basic flaw in paternalism. The notion that people who are not very good at making decisions for themselves will be much better at making decisions for others is problematic on its face, and it is certainly not borne out by the empirical evidence. Even so, the soldiers of the nanny state seem to be on the march these days. Insofar as Conly’s book reveals their weakness, it is timely indeed.
This article was originally published in the April 2004 edition of Future of Freedom.