vendredi 3 juin 2011

Emotion, reason and policy: Thinking liberally about feeling | The Economist

American politics

Democracy in America

Emotion, reason and policy

Thinking liberally about feeling

May 31st 2011, 14:35 by W.W. | IOWA CITY

ROSS DOUTHAT offers a partial defence of Steven Levitt's "daughter test" by way of a reply to my blog post on the subject:

[T]hinking “what if I my daughter did this/were in this position?” is a way to take an argument from the abstract to the viscerally real, and to bring moral and legal gray areas into a sharper focus. It isn’t a mathematical proof, or a system of inputs that spits out an automatic, universal answer: For instance, a more pro-life father would take a different view of the abortion question that Levitt does. (Or to take the example that kicked off Levitt’s riff, one could imagine someone else saying that Internet gambling doesn't pass his daughter test, because a gambling addiction can be as destructive as cocaine.) But that doesn't mean that it isn't a useful way of thinking about public policy. The fact that I would want to be able to involve the police if my daughter became a streetwalker, but not if she became a Hari Krishna, tells me something important about what kind of legal regime I should support. (There’s a touch of Kantianism in it: One’s (legal) preferences for one’s daughter should become a universal law…) And the fact that Wilkinson disagrees doesn’t prove that he believes in logic and reason, whereas I believe in raw emotion. It just proves that his answer to the daughter test is—for now, at least—different from my own.

Just be clear, I don't believe I argued that Mr Levitt is jerked around by his feelings, but my judgments about policy are plucked fresh from the crystal stream of cool reason. My main point was that sentiment dominates judgment so thoroughly, that the successful application of dispassionate rationality is so rare, that we cannot realistically expect even virtuosi in the arts of rational analysis, such as Mr Levitt, reliably to apply the results of their expert cogitations to policy. This doesn't mean we shouldn't try our best to apply reason to policy. It just means that our best is going to be based on the recognition that our best is sometimes pretty bad.

I would also note that the principle that "One's (legal) preferences for one's daughter should become a universal law" has a touch of Kantianism in it only because Kant liked to talk about "universal laws". Otherwise, this is pretty much the opposite of the sort of thing Kant had in mind. Categorical imperatives apply to us simply in virtue of the alleged fact that each of us has a rational will capable of motivating us to act on practical principles that otherwise might leave us cold. Our "preferences" about our children, or about anything, provide the basis for neither morality nor law. The urge to generalise one's own (contingent, historically-conditioned) paternal sentiments and codify them in law will lead to conflict between those with different sentiments and the imposition of rationally unjustified coercive restrictions by the winners. Frederick Rauscher's encyclopedia entry on Kant's political philosophy nicely captures how far anything like Mr Levitt's literally paternalistic test strays from canonical classical liberalism:

"There is only one innate right," says Kant, "Freedom (independence from being constrained by another's choice), insofar as it can coexist with the freedom of every other in accordance with a universal law". Kant rejects any other basis for the state, in particular arguing that the welfare of citizens cannot be the basis of state power. He argues that a state cannot legitimately impose any particular conception of happiness upon its citizens. To do so would be for the ruler to treat citizens as children, assuming that they are unable to understand what is truly useful or harmful to themselves.

Now, I don't think Kant's moral anthropology stands up to scrutiny. Morality is not an ahistorical product of universal pure reason. It is, as Kant's great empiricist frenemy David Hume argued, a dynamic body of conventional rules. However, as Gerald Gaus argues in his profound new work of moral and political philosophy, "The Order of Public Reason", Kant and Hume are two great tastes that taste great together. Mr Gaus argues both that

... as Kantians observe, morality is the dictate of impartial public reason and that it has a history and so is path dependent and indeed in a significant sense a society chooses its morality. Without appeal to social evolutionary processes, the Kantian ideal of common self-legislation is either hopelessly controversial or indeterminate; without appeal to the critical perspective of the Deliberative Public Justification Principle, the evolutionary view cannot distinguish authoritarian from non-authoritarian evolved positive moralities. An adequate account of morality needs both Kant and Hume.

The "Deliberative Public Justification Principle" states, more or less, that a moral rule has teeth only if each of us has good reason to accept the rule as binding—to see it as something that applies to us, something that makes sense for us to take on board and be guided by. Something in the neighbourhood of Mr Gaus' principle of public justification specifies what it means to take liberty and liberalism seriously. If some of us think it would be a good idea to limit others' choices, we owe them an argument that they have good reasons to abide by these limitations, and if they really can't see it, if they have reasonable grounds on which to reject such a limit on their liberty, we owe it to them back off.

Mr Douthat maintains that the fact that our gut judgments differ "doesn't mean that [the daughter test] isn't a useful way of thinking about policy", but I think it does mean that. That is, unless Mr Douthat simply means to say that thought experiments like the daughter test can usefully clarify the role our emotions play in our policy judgments. In that case, I agree. But I deny that this sort of thing "tells me something about what kind of legal regime I should support." It tells me something about the kind of legal regime I probably do support. But no one's moralised disgust at, say, the idea of gay marriage or righteous indignation over income inequality tells them which policies they ought to affirm.

It's true, as Mr Douthat concludes, that my answers to the daughter test are different from his. But my argument against prohibitions on drugs, gambling, abortion and prostitution is not that these restrictions do not fail my daughter test. My argument against them is that they fail the test of public justification, that many of us have perfectly reasonable grounds for rejecting and resisting the imposition of these constraints. Now, I do not claim that a commitment to liberal ideals is rationally mandatory, that my own attachment to those ideals has no emotional basis, or that liberal arguments do not require sentimental preparation to land with persuasive force. But I do claim that religious conservatives who inhabit liberal cultures like ours, and especially those like Mr Douthat, who have made a profession of public deliberation, are too liberal at heart to deny the ideal of public justification and its duties.

We are, all of us, a confusion of sentiments. The advance of liberal civilisation consists in no small part in unwinding the tensions between our liberal and illiberal impulses. It may be true, as Jonathan Haidt argues, that the feasibility of a robust liberal social order requires the ongoing cultivation of certain distinctively conservative moral sentiments and practices. If he's right, some tension between our liberal and illiberal sentiments may be essential to liberal civilisation. One can imagine an argument like this underpinning the public justification of certain limits on liberty. Everybody has good reason to affirm restrictions on freedom really required for the stability of society. But this sort of argument, about the value or necessity of certain moral emotions, is not an argument from moral emotion. It is justification for letting certain feelings rip, within reason.

Maybe it will turn out that Mr Levitt's answers to the daughter test just happen to be the right ones. Maybe Mr Douthat's are. But, obviously, our feelings can't tell us what is within reason. Sadly, sometimes reason can't tell us what's within reason. Rationality is hard and we're not very good at it. But that's no reason to give up and go with our guts. That's a reason to try harder, and better.

Readers' comments

The Economist welcomes your views. Please stay on topic and be respectful of other readers. Review our comments policy.

You must be logged in to post a comment. Log in to your account.
Don't have an account? Register

1-20 of 51
JGradus wrote:
May 31st 2011 2:48 GMT

I must say W.W., I like you a lot more when you go all "I have actually read philosophy" on people's asses.

Now, if you could only embrace the Nordic Model, we would be going somewhere :)

May 31st 2011 3:34 GMT

If I followed my gut/emotions I would probably have started serving a life sentence awhile ago.

willstewart wrote:
May 31st 2011 3:49 GMT

'But my argument against prohibitions on drugs, gambling, abortion and prostitution ... is that they fail the test of public justification..'

Do they? Do you actually have any numerically-supported arguments on public justification, or are your feelings more significant than you think?!

Anderson-2 wrote:
May 31st 2011 3:50 GMT

Douthat's and Levitts' daughter test is wrong on a number of levels. The daughter test clouds rather than clarifies judgement.

The first is what WW points out, not wanting your daughter to do something is a motivation for raising her right, not for banning the activity.

Secondly, prohibition mostly just raises the costs of an activity. Women become prostitutes (at least in the US) mostly because of abuse, poverty and demand for the service.

I don't abuse my daughter, and we are not poor, so it is unlikely my daughter will be a prostitute. But I would rather have to worry about a prostitute daughter than to worry about a prostitute daughter in prison. I would rather worry about my daughter's drug addiction than have the cost of her drug addiction drive her into prostitution and then have to worry about my daughter being in prison on drug and prostitution charges after having given a lot of money to some psychopath who raped her and got her pregnant and who's idea of competition and animal spirits involves automatic weapons fired a random into a porch party.

The test should be, if one thinks that some drugs are bad for their daughter, does one think that banning those drugs actually reduce the likelihood that their use will damage one's daughter. And if one thinks that it does, does the reduction justify the costs of prohibition and the cost of the knock-on problems that it creates?

Personally the only rational argument for banning some drugs that I have every heard is what effect might a freer market and competition have on the drugs we use? My legalized-drug nightmare is my daughter popping one of Pfizer's latest wonderfun pill at a party and waking up thinking Rush Limbaugh makes sense, and that Paul Ryan's budget ideas will cure all that ails us.

Doug Pascover wrote:
May 31st 2011 4:10 GMT

This post seems a little self-indulgent, ironically, but ok.

I agree very much with W.W. on this. The daughter test seems like a useful way to think about parenting, not policy. I suppose you can use the daughter test to imagine what you want the world to be like after all the governing is done, but you still need to consider the nature of the state, the function and disfunction of government and, as W.W. suggests, any obligation of citizens to stand down from official judgement and coercion against their neighbors before you can sensibly consider what kind of policy to support.

Anderson-2 wrote:
May 31st 2011 4:29 GMT

It is also a fundamentally illiberal concept in every sense of the word.

For instance, we very sensibly restrict minors in ways we do not restrict adults. Douthat doesn't want his 10-year-old daughter to take drugs so he wants to keep the ban on drugs. Well nobody is arguing that minors should be allowed to buy and use heroin. So obviously he also wants to control his adult daughter's behavior, too, and he wants me to pay more taxes and subsidize criminals and the rest of it so that he can continue to control his daughter's behavior all her life, by proxy.

This is crazy.

May 31st 2011 4:35 GMT

The daughter test is very dangerous when it comes to thinking about people who have different preferences from our own. I believe this is one reason why there has historically been so much prejudice against gays. Otherwise decent men cannot imagine themselves engaging in homosexual sex and find the whole idea distasteful. Therefore they want to ban it on the basis that everyone involved is either a pervert or a victim.

What is needed is the leap of imagination that says "I would never choose to engage in gay sex but there are sane, well-adjusted people who would and denying them their choice is every bit as wrong as it would be to criminalize my relations with my girlfriend." The daughter test makes this sort of leap very difficult because you are trying to ground everything in your own experience, but your experiences cannot possibly cover the total of human activity or desires.

Dylin wrote:
May 31st 2011 5:08 GMT

I was very much surprised to learn of this conception of the 'daughter test' as a tool for reflection, as I have consciously used the same manner of thinking, but not towards public policy.

I would often find myself thinking about the merits of a potential girlfriend. Whenever I was offput by a certain behavior, the best way I could explain it to myself was to suggest, 'what if I had a daughter that acted that way.' Now, this manner of evaluating a mate never would be applied to incidents involving reasonable choices that I disagreed with (choosing a college major, deciding to suddenly leave college to pursue other passions). It only applied to things that (directly or indirectly) gave me that sinking feeling which follows from a loss of respect or sense that the behavior was dishonorable (leaving college would seem to be unreasonable to some, but I defer to an individual's pursuit of happiness). I'm sure this principle of using a real or imagined daughter's well-being as a tool for reflection works very well for some people who creatively seek to define their moral stance in occasions where there is a vague sentiment of unease about potential consequences or implications of some social interaction. I had not previously considered this viewpoint so broad as to extend to public policy, but I suppose that this ‘daughter test’ is a desirable tool for any kind of stakeholder to further define or characterize their interests- in a person or government- such that it affords them a less-obstructed view of what is most honorable and reasoned from their viewpoint. I think this 'test' is a natural mental conception that is employed once someone realizes they need to protect potential long-term stake or investment but must use a sort of conceited empathy that can easily define how offspring will be implicated. However, this tool may be completely inconsiderate of the stakes of others’ daughters.

May 31st 2011 5:31 GMT

WW, you argue for banning the torture of high-value terrorists. You're in the minority. Does your preference satisfy the "Deliberative Public Justification Principle"? I would argue that "Let people torture. It's not my place to stop them," is far worse than the Daughter Test.

I'm not nearly as well-read in philosophy but what's the difference between "Deliberative Public Justification Principle" and the Daughter Test + pluralism? If you aggregate everyone's Daughter Tests don't you arrive at the same place in practice? If you internalize the principle of pluralism, you'd arrive at the same place even in your head.

@Anderson-2, I think both Levitt and Douthat would agree that the daughter test doesn't tell you whether an activity should be banned. It only tells you whether there is a moral justification for regulating it. From there, you'd have to do cost-benefit analysis to determine whether prohibition is the best regulation. Neither Levitt nor Douthat argued for banned prostitution or drugs.

May 31st 2011 5:57 GMT

@Unhappy Realist, even heterosexual premarital sex would fail the Daughter Test. It's why we have statutory rape laws. When they're older, we may not want them to engage in premarital sex but it becomes more like our preference that they pick a certain career or religion. It's not something we want imposed by force. In the case of sex, enforcement of a ban may not be something over which we're willing to violate our daughters' privacy. If she's having sex in public, by all means, arrest her.

The Daughter Test may be more useful the area of sexual activity. To rework UR's argument: "I would never choose to engage in public sex but there are sane, well-adjusted people who would and denying them their choice is every bit as wrong as it would be to criminalize my private relations."

Anderson-2 wrote:
May 31st 2011 6:11 GMT

RR, but that is what he is arguing

"If the answer is that I wouldn’t want my daughter to do it, then I don’t mind the government passing a law against it. I wouldn’t want my daughter to be a cocaine addict or a prostitute, so in spite of the fact that it would probably be more economically efficient to legalize drugs and prostitution subject to heavy regulation/taxation, I don’t mind those activities being illegal."

May 31st 2011 6:24 GMT

@Anderson-2, you're right. Although, Levitt does assume that the prohibitions would be effective, an assumption you don't share. Still, I suspect that Levitt would concede that, given limited resources, we need to do a subsequent cost-benefit analysis to prioritize prohibitions. That doesn't render the Daughter Test useless.

May 31st 2011 6:59 GMT

This debate is really showing what I think is the fundamental problem with technocracy. We identify people as experts because we hope that they'll use their acumen and training to make policy. But then they go and make many of those decisions using the same methods that the unwashed masses use, like the "daughter test" or misremembering Kant.

Commenter#38 wrote:
May 31st 2011 7:57 GMT

The bigger problem with "pure logic" is that it often fails to take into account the uncertainty of the world. Also many who believe they are following "pure logic" focus their thinking down a narrow path, constrained by the assumptions they make and by the rules they set for themselves.

I think there is some argument to be made for the idea that intuition plays a role in helping to overcome these limitation, and the "daughter test" may be part of that. I think both intuition and logic need to come into play in decision making. Otherwise, we might as well get a bunch of college students who happen to have read Kant to come up with policy for us.

LexHumana wrote:
May 31st 2011 9:19 GMT

Levitt is explaining motivations, not justifications.

If the government passes a law that dovetails with my personal biases, I am not going to complain. If the government passes a law that is dissonant with my personal biases, it will raise my hackles. The fact that a law may fly in the face of reason is irrelevant -- as long as the electorate is comfortable with the law, they are the ones that have to live with it.

For example, is it right or wrong for France to ban face coverings? Is it right or wrong for Germany to phase out nuclear power? Is it right or wrong for Saudi Arabia to cut off the hands of thieves? Is it right or wrong for Switzerland to ban minarets? Is it right or wrong for Bloomberg to ban smoking in all public areas? Is it right or wrong for Nevada to have legalized gambling and legalized prostitution? Is it right or wrong for the CIA to use waterboarding?

Ultimately, none of these questions are dependent on which side's eggheads come up with better "rational" numbers to prove their point.

colm5 wrote:
May 31st 2011 9:27 GMT


"even heterosexual premarital sex would fail the Daughter Test. It's why we have statutory rape laws. When they're older, we may not want them to engage in premarital sex but it becomes more like our preference that they pick a certain career or religion."

Statutory rape laws have nothing to do with premarital sex, they're there to protect young people before they're capable of making an informed decision. This is why when two consenting people below the statutory rape age to have sex, neither is arrested, regardless of their marital status, since the participants are presumably on equal footing in terms of emotional and mental development.

Statutory rape laws are no different than child labor laws, meant to prevent exploitation. Marital status doesn't enter into it.

bampbs wrote:
May 31st 2011 9:49 GMT

"My main point was that sentiment dominates judgment so thoroughly, that the successful application of dispassionate rationality is so rare, that we cannot realistically expect even virtuosi in the arts of rational analysis, such as Mr Levitt, reliably to apply the results of their expert cogitations to policy. This doesn't mean we shouldn't try our best to apply reason to policy. It just means that our best is going to be based on the recognition that our best is sometimes pretty bad."

Yup. Well said.

May 31st 2011 9:55 GMT

I usually agree with W.W. but I think the "daughter test" is a valid way to bring a measure of empathy to the fact that policies have consequences, often life and death consequences, for real people. As a physician (although not currently clinically active), I often tried to advise patients in a way consistent with what I would have preferred given their circumstances, but ultimately relying on their expressed wishes and my best "sound clinical estimate."

As it relates specifically to torture, I absolutely agree with W.W.'s rejection of it, and applying the "daughter test" yes I could most certainly be driven to commit acts of torture myself. But then I should later answer to a court of law for doing so, both to avoid legitimizing torture and to have an "objective" (meaning a legally trained judge and peer jurors) assessment of "how wrong" it was to torture to prevent a daughter's rape/murder or something of that nature. The same argument goes with the ticking nuclear bomb scenario. Presented with a man I was made to believe had the necessary knowledge to prevent a nuclear attack on a major city but who would not volunteer that knowledge, yes, I could come up with some pretty nasty ways to help his (or even, with admittedly slightly higher threshold, her) tongue move in useful ways. But afterwards, whether it's the daughter or the city, off to court to plead mitigating circumstances. Turning the scenario around, if I were a juror on either case, I would insist that torture is wrong but probably be quite sympathetic and lenient towards any person left with that unenviable dilemma.

Robert North wrote:
May 31st 2011 10:29 GMT

I wonder if there is an economic answer? did the emergence of liberal society coincide with technological/scientific and colonial expansion?

Anderson-2 wrote:
May 31st 2011 10:45 GMT


Levitt's assumption is that "it would probably be more economically efficient to legalize drugs and prostitution subject to heavy regulation/taxation." My assumption is similar, but I would replace probably with certainly, and that doesn't even get into a host of other mega unintended consequences, like say, the stupefyingly bloody civil war we are funding down in Mexico, at enormous cost to a country that should be much richer than it is and for no other reason than many people, like levitt, get a "queasy feeling" about the idea of legalization. I spent many months studying Spanish and working in Mexico, and it was truly a transcendent time for me, but I don't think I would want my daughters to do so now. Though they being smart girls would probably go and have a great time and mock me for my concern.

Looks like Levitt's afraid that increased availability might lead to more destructive addiction.

I lived and worked for a couple of years in Cambodia. Pot was a traditional seasoning and cost less than beef, by weight. Cocaine was easy, and heroin was easier and it was all cheap and risk free. Addicts brought their problems with them, but mostly people made rational decisions about use and abuse and I can't think of anyone that I knew or heard about who developed a debilitating habit in country. Just the opposite really.

1-20 of 51

Aucun commentaire:

Enregistrer un commentaire