Paternalist Slopes
I submit this in response to Glen Whitman's post by the same name. It's a topic that has had my interest lately.
I) There is a three step approach to policy-making:
1) Identify market failure (sometimes traditional, i.e. externality, but increasingly prevalently cognitive bias, bounded rationality, etc.)
2) Recommend and implement policy
3) Declare victory!
I first realized how badly the defenders of state action needed to be called on this approach in this exchange with Mike Huben. Mike thinks that state action in private, sexual matters is easily justifiable:
Why does the analysis of such policy go no further? We know that policy-makers (both legislators and the technical experts in the agencies directed to carry out policy) are subject to the same biases the rest of us are. Look at this history of marijuana laws presented by Charles Whitebread before the California Judges Association. At every step along the route, it was done for the benefit of the users themselves, but the evidence to support each piece of legislation was so flimsy as to defy rational explanation. Especially notable is the fact that at first, marijuana was demonized as being the drug morphine users would fall back on, and later the same people argued it was the gateway to heroin, without apparently blushing. Yet, here we are.
We also know that policies frequently have unintended consequences themselves; think of the effect of subsidizing sugar on the Everglades, or rent control on housing in NYC. We know government failure exists as surely as market failures do; rent seeking is as real as adverse selection, and bringing public good production under federal oversight does not make it less a public good. So it seems rational to me that instead of having, "3) Declare victory!", we should instead have an analysis of the policy, analysis of private alternatives to the policy, comparisons of the two, and a recognition that the least worst policy may be to fail to legislate proposals simply because we can create them without regard to their efficacy or the alternatives. As I wrote on the Becker-Posner Blog entry on Sunnstein and Thaler's Libertarian Paternalism, "The analysis rarely if ever proceeds to look for problems with the policy solution (both market and gov't failures may be present), possible secondary market solutions (e.g. Consumers Reports and Carfax in the case of lemons), or even whether the policy solution will actually address the failure rather than just intend to address it."
I do have reservations, however, about Whitman and Rizzo's vagueness argument: as I pointed out in my comment on Utilitarian Blindness (thanks to my wife for making me confront this), the simple fact that we cannot make all relevant calculations of utility does not mean we should not undertake a policy. I am here simply pointing out that perennial fans of policy-based solutions have not even identified all of the trade-offs, much less attempted to calculate and compare them.
II) There is also an implicit belief that people *should* be coldly and unrelentingly rational. Should they, really?
Despite Spock's faithful adherence to logic, Kirk could regularly beat him in contests. Who was the only successful Kobayashi Maru captain? This is entirely plausible because these were games of strategy where logic could either run you into an undesirable corner (think of the prisoner's dilemma) or fail to illuminate a potential successful but entirely non-linear direction (think Hofstatder's explanation of Godel's Theorem in Godel, Escher, Bach: an Eternal Golden Braid), and at very least fail to yield answers when they were needed because there were too many unknowns.
That last possibility is one which I think most important. When making time-dependent, strategic decisions, the information required to make a perfectly calculated, rational decision are frequently not available or not knowable. The human advantage is in having an evolved set of mechanisms for making decisions based on scant data, a process which we call, "gut feel", "instinct", or "intuitive". Think of Malcolm Gladwell's Blink. Over time, these abilities may be developed to a high degree - chess grandmasters, extraordinarily successful athletes, and polymath entrepreneurs use a combination of calculation and intuition.
That mechanism may fail us when researchers have set up a game and are watching. They know the "unknowable", and are specifically trying to trick the subject's mechanism into taking the wrong tact. I wonder, however, if anyone has ever tried analyzing what happens when they show participants the outcome and then run the experiment again? My guess is that participants get better at overcoming their biases. I would also guess that now that we humans have identified the phenomenon of cognitive bias, over time this information and strategies for avoiding its problems will disseminate to the public at large. Those who successfully incorporate this knowledge will be successful in life, possibly more successful at breeding, and thus we may breed cognitive bias out over millenia.
If people won't be cold, calculating, automatons, then it is the intent of scientific planners that government take over their decision-making for them. How can this be anything but a slippery slope? Where, in the Scientism of Comte and Croly, the Progressives and New Dealers, in the central planners, is there room for arational or even irrational behavior? If there is none, then I think we can safely assume their descent down the slippery slope even if they don't intend it; if not now, then later.
I) There is a three step approach to policy-making:
1) Identify market failure (sometimes traditional, i.e. externality, but increasingly prevalently cognitive bias, bounded rationality, etc.)
2) Recommend and implement policy
3) Declare victory!
I first realized how badly the defenders of state action needed to be called on this approach in this exchange with Mike Huben. Mike thinks that state action in private, sexual matters is easily justifiable:
"Private sexual behavior has plenty of unintended consequences, most notably unwanted pregnancy and disease transmission. Both are important reasons for state action, although such action should be as little intrusive as possible. Thus, for example, we have sexual education requirements in schools. And required VD tests before marriage."Unintended or not, these actions are not normally thought of as external. Assuming we are talking about normal adults, the actors know the risks for both, and the consequences will fall on them, not a third party. Absent an externality (I thought of at least two weak ones, but it isn't my job to provide Huben's argument), what is the real failure here? Yes, there is imperfect knowledge here, but that is true of *every* transaction, and hardly a compelling motivation since the authorities lack perfect knowledge themselves. There is also information asymmetry, but there are private means of addressing this: dating, gossip, reputation, etc. And even if we could identify a failure, do the policies address it? People have sex all the time without getting married and therefore entering the realm of VD test requirements. Further, some states don't require such tests, and others show their root in the legislative biases of bygone eras as they continue to test for syphilis but not AIDS. Despite increasingly common and frank discussions of sex in schools, teenage pregnancies remain high, much higher I believe than before the new openness (an era brought about, in part, by the pill). But no, merely identifying the problem and a policy that nominally addresses it seems to be enough. The 3-step process is a simplified but essentially correct version of what is formally known as Second Best Theory.
Why does the analysis of such policy go no further? We know that policy-makers (both legislators and the technical experts in the agencies directed to carry out policy) are subject to the same biases the rest of us are. Look at this history of marijuana laws presented by Charles Whitebread before the California Judges Association. At every step along the route, it was done for the benefit of the users themselves, but the evidence to support each piece of legislation was so flimsy as to defy rational explanation. Especially notable is the fact that at first, marijuana was demonized as being the drug morphine users would fall back on, and later the same people argued it was the gateway to heroin, without apparently blushing. Yet, here we are.
We also know that policies frequently have unintended consequences themselves; think of the effect of subsidizing sugar on the Everglades, or rent control on housing in NYC. We know government failure exists as surely as market failures do; rent seeking is as real as adverse selection, and bringing public good production under federal oversight does not make it less a public good. So it seems rational to me that instead of having, "3) Declare victory!", we should instead have an analysis of the policy, analysis of private alternatives to the policy, comparisons of the two, and a recognition that the least worst policy may be to fail to legislate proposals simply because we can create them without regard to their efficacy or the alternatives. As I wrote on the Becker-Posner Blog entry on Sunnstein and Thaler's Libertarian Paternalism, "The analysis rarely if ever proceeds to look for problems with the policy solution (both market and gov't failures may be present), possible secondary market solutions (e.g. Consumers Reports and Carfax in the case of lemons), or even whether the policy solution will actually address the failure rather than just intend to address it."
I do have reservations, however, about Whitman and Rizzo's vagueness argument: as I pointed out in my comment on Utilitarian Blindness (thanks to my wife for making me confront this), the simple fact that we cannot make all relevant calculations of utility does not mean we should not undertake a policy. I am here simply pointing out that perennial fans of policy-based solutions have not even identified all of the trade-offs, much less attempted to calculate and compare them.
II) There is also an implicit belief that people *should* be coldly and unrelentingly rational. Should they, really?
Despite Spock's faithful adherence to logic, Kirk could regularly beat him in contests. Who was the only successful Kobayashi Maru captain? This is entirely plausible because these were games of strategy where logic could either run you into an undesirable corner (think of the prisoner's dilemma) or fail to illuminate a potential successful but entirely non-linear direction (think Hofstatder's explanation of Godel's Theorem in Godel, Escher, Bach: an Eternal Golden Braid), and at very least fail to yield answers when they were needed because there were too many unknowns.
That last possibility is one which I think most important. When making time-dependent, strategic decisions, the information required to make a perfectly calculated, rational decision are frequently not available or not knowable. The human advantage is in having an evolved set of mechanisms for making decisions based on scant data, a process which we call, "gut feel", "instinct", or "intuitive". Think of Malcolm Gladwell's Blink. Over time, these abilities may be developed to a high degree - chess grandmasters, extraordinarily successful athletes, and polymath entrepreneurs use a combination of calculation and intuition.
That mechanism may fail us when researchers have set up a game and are watching. They know the "unknowable", and are specifically trying to trick the subject's mechanism into taking the wrong tact. I wonder, however, if anyone has ever tried analyzing what happens when they show participants the outcome and then run the experiment again? My guess is that participants get better at overcoming their biases. I would also guess that now that we humans have identified the phenomenon of cognitive bias, over time this information and strategies for avoiding its problems will disseminate to the public at large. Those who successfully incorporate this knowledge will be successful in life, possibly more successful at breeding, and thus we may breed cognitive bias out over millenia.
If people won't be cold, calculating, automatons, then it is the intent of scientific planners that government take over their decision-making for them. How can this be anything but a slippery slope? Where, in the Scientism of Comte and Croly, the Progressives and New Dealers, in the central planners, is there room for arational or even irrational behavior? If there is none, then I think we can safely assume their descent down the slippery slope even if they don't intend it; if not now, then later.
Labels: philosophy, politics, regulation
<< Home