-
Website
http://willwilkinson.net/flybottle -
Original page
http://www.willwilkinson.net/flybottle/2006/01/10/nim-pub-and-cognitive-paternalism/ -
Subscribe
All Comments -
Community
-
-
Top Commenters
-
Popular Threads
- A Little Mystic Nationalism
- Scott Winship on Income Inequality
- Now Let Us Praise Results-Facilitating Virtue!
- Why Are There So few Women in Philosophy?
- Hey, I Can’t Actually Quite Imagine a World in Which Things Are Exactly as Different as the Need to Be to Give Me What I Want, but It Would Be Neat if I Could!!
BTW, you've got a dangling asterisk with no associated footnote.
But perhaps then it's merely semantics. An "unstable" institution could be defined in such a way that such a thing, by definition, could not exist. But if one chooses to see stability as a continuous measure, it probably wouldn't be difficult to think of examples of fragile-yet-existent institutions which we might be inclined to label as "unstable".
You apparently assume that implementing markets and implementing coercive policies require the same kinds of competencies. Are you saying that if a government is not of sufficient quality to, say, manage the market for beef, then it probably isn't in a position to enforce contracts in the market for beef?
It seems easy enough to imagine that a goverment could effectively work to enforce contracts (by commiting to exercising coercive power against parties that violate contracts), and yet also believe that such a government would be incompetent as implementer of a cattle market.
So, yes, people who have internalized different norms may behave differently in otherwise identical external circumstances. Different possibilities will occur to them, and they will order the possibilities differently, because different norms will assign different prices to them.
Suppose we have found, through trial and error, that requiring people to put money aside for their retirement (or even requiring them to specifically exercise an option not to) helps counteract what you would concede is a nonideal cognitive tendency to discount the future excessively. If we take NIM seriously, then we should be suspicious of an ideologue who comes along and wields the coercive/non-coercive, state/market dichotomy in a utopian fashion.
You're both right. Which is why it's so nice to think aloud in front of really smart people! And which shows just how hard NIM makes to think in useful generalities.
Michael, You're right that the competencies to enforce contracts and to actively run a market are different. What I hazily had in mind by "implement a market" was something like setting in place the legal/regulatory framework for the definition of property rights and contract enforcement. My sense is that the cattle futures market, say, depends on the willingness of government to acknowledge and enforce certain kinds of contractual arrangements. Innovations in new forms of contracts won't have teeth unless the government will recognize them as valid. So maybe it's better to say "facilitate" rather than "implement" a market. One of the things that makes good governments good is that they do a good job of facilitating markets, and know that they'll do a bad job of managing them, and so don't try.
Pith, I agree that an ameliorative policy about coercing people into savings might work well. What I'm interested in is what government institutions need to look like such that policymakers accurately identify what has been sussed out by trial and error, and implement an effective policy without, at the same time, using the same power and discretion to implement excarbating or harmful policies. Are there aspects of the current policy climate that, say, exacerbate the tendency to "excessively" discount? Perhaps government assurances that people will not be allowed to fall into poverty in old age leads to present overconsumption.
I think lurking in the background here is a fuller NIM story about the institutions of expertise. You need economists, political scientists, psychologists, etc., to identify nonideal cognition, the problems that arise from it, and potential solutions. But these experts don't exist without institutions of expertise, like universities, think tanks, etc. A society with high-quality institutions of expertise will have a leg up on setting effecive ameliorative policy.
But you also need a way to integrate institutions of expertise into government institutions, while avoiding problems of confirmation bias in choosing which experts to listen to, appointing them, etc. The US Fed I think is an interesting case of a political institution that does a decent job of applying policy on the basis of expert opinion without suffering from too much political capture.
I should say that I think the coercive/non-coercive distinction IS very morally salient. Other things equal, less coercive policy is morally better. So if there is some less coercive means of raising the cost of undersaving, then we ought to prefer it. But, as Michael's comment implies, state coercion, in the form of the commitment to enforce contracts, is often the assurance at bottom of a well-functioning market. The decision to enforce some contracts, but not others, is a decision about how to bring coercion to bear. Some markets, such as, say, dynamic auction markets for landing spots at airports, are coercively forbidden (as far as I understand). "Letting the market work" in this kind of case involves removing a policy of coercion, and, at the same time, committing to enforcing new forms of contracts, which is an new policy about the deployment of coercion. So.. there's a bigger point in there somewhere ...
Thanks guys, this is really helping.
How is the commitment to enforce contracts a form of coercion? Perhaps this is just a verbal quibble, but enforcing a contract amounts to enforcing the property rights of the parties to said contract. Enforcing property rights is not coercive.
Pithlord,
Referring to a rate of future discounting as excessive seems to imply that there is a correct rate. Do you mean to imply this? Do you believe that you know what this rate is?
I don't need to know what the "correct" rate for discounting is to know that the rate revealed by behaviour is too high any more than I need to know what the correct calorie intake for a 35 year old male is to know that I overeat. I know I save too little, and I know many others are worse. That is perfectly consistent with their being a range of reasonable discount rates, and even my not knowing what that range is.
Just as pharmaceutical companies engage in double blinded clinical studies to remove the possibility of human bias in the determination of the safety and efficacy of drugs, so should blinding be used in the planning and execution of ameliorative public policies.
The basis of this is best understood by taking Brian Tracy's most recent book, Something for Nothing, as a useful description of non-ideal cognition. He says that all humans are naturally (read: genetically):lazy, greedy, ambitious, selfish, vain, ignorant, and impatient. He then explains that all humans have a heirarchy of needs. Starting from the bottom: safety, security, comfort, leisure, love, respect, and fulfillment. The issue is how humans get from their basic attributes to the fulfillment of their needs. The right systems dissuade the pernicious manifestation of our basic attributes, while the wrong systems encourage them. Free markets without an easily accessed safety net are an example of the former. Collective and paternalistic economies are an example of the latter.
So, it would be easy to just say that a well-run free market is the answer to the NIM/PUB problem. However, we really can't get there from here. So what to do? Start using something akin to blinding in the dispensation of ameliorating government policies.
The idea is that the non-ideal aspect of government is almost entirely due to the disproportionally large influence of special interests. Money from the masses (from federal taxes, for example) goes to specific entities in specific geographies, often for far-from-ideal uses. Imagine - and this is where I am at the moment - imagine a system whereby requests for federal funding are blinded. Decisions as to what organization gets what would be disconnected from the non-ideal (that is, corrupt) process that happens today. This is just means based redistribution, but with a check against inherent human expediency.
Still thinking about how to make it work, but I believe there's something there. The money to be made by getting a drug to market, even if the drug doesn't work, is so substantial that we have to expect a constant barrage of invaders looking for a chink in the armor. Blinding, thus far, has been pretty effective in stopping the cheaters. And, as it is philosophically similar to the idea of checks and balances, I can't help but think it could be a plausible solution, even it is only an incremental one.
Incidentally, you can check out my review of Tracy's book at:
http://www.enlightenedcaveman.com/2006/01/book-...
Great post!
I reckon our disagreement re: coercion is probably about how we define the term coercion, rather than anything more substantive. I mean it as the initiatory infraction against the property rights of another, so when I say contract enforcement isn't coercive, I really mean that it isn't coercive, not that it's coercive but justified.
Pithlord,
Generaly I hesitate to assign a comparative operator between two quantities when I don't know the endpoints of the range in which one of them falls. If your discount rate is D and the range of reasonable discount rates is (A,B) where A and B are unknown to me, I have no idea how I could ever be reasonable in making claims about whether or not D is on (A,B). I could be wrong.
I think you're asserting the consequent or something. If I know what D is, and I know what B is, then I know whether B
Odd. I would think it is much easier to accurately rank magnitudes than to know what they are.
More importantly, I took the point of your original comment to be that our revealed preferences can't be wrong. But they can. I can like carbs too much, or smoke too much, or read crap. I can even know that my preferences are bad, and want to change them.
This does *not* mean that someone is entitled to impose my true preferences on me. There may be a value to autonomy that trumps other-imposed perfection. But we can be liberals without saying that revealed preferences are just given, above cognitive crticism or improvement. That is too high a price to pay for liberalism.
You write "Odd. I would think it is much easier to accurately rank magnitudes than to know what they are."
I'm thinking of a number A and a range from X to Y. I won't tell you what the values of A, X and Y are. Rank them.
You continue "More importantly, I took the point of your original comment to be that our revealed preferences can’t be wrong. But they can."
Maybe so. I lack the confidence to claim that any particular instance of revealed preferences is wrong. Thre is too much temptation to confuse the right preferences for others to have with my own likes and dislikes.
I have here a pug and a Newfoundland. i won't tell you what their weight is. Which one is heavier?
I suspect you get the right answer more often than chance.
I don't disagree with you that we should be cautious about imposing what we believe to be the right preferences over those that revealed by uncoerced behaviour. One reason is that we might be wrong. Another is we might be biased. But even if we are right, and unbiased, there is still a third answer, which is that we are all entitled to make bad choices if we pay the consequences.
That third answer is a pretty good one, 99% of the time. But you wouldn't use it for four year olds wanting to cross Sixth Avenue.
I have prior empirical knowledge about what different breeds of dog weigh. For your analogy to work, I'd need prior knowledge about what fraction of their income people ought to save, or the range in which this fraction falls. I don't get my knowledge about "ought" from empirical evidence and my introspection just comes up blank on the question of how much others should save.
But suppose that I though I really knew what other people should be saving. Well, I tell my friend he saves too little. He tells me that actually I'm saving too much. How do we decide? At best, we should be able to adress the issue with some method sufficiently clear that we can tell whether or not we are even applying it. Well, we can look at the consequences of different savings rates, but how much we value those effects is subjective. Then what?
Finally, I'd agree with "...we are all entitled to make bad choices if we pay the consequences," except that it has some odd implications for a person who really wants to do something terribly bad: May I smash your foot if I'm willing to pay your hospital bill and serve jail time? I guess there are bullets to bite that go along with any view, but I'd rather bite the bullet that says a kid really has the right not to be interfered with crossing the street.
Paternalism is justified by reference to the imperfection and irrationality of human nature, but these same traits count against human beings acting through government to limit imperfection and irrationality.
This is a good point, but the level of abstraction at which you raise it makes it sound more damaging to the paternalist case than it really is. A smoker intentionally lights up under the influence of nicotine cravings. Even a legislature consisting only of chain smokers will not be led by their nicotine cravings to write a law promoting smoking. The imperfections/irrationalities of human beings in government are not necessarily the same as those of the individual. In a particular instance (e.g., smoking bans) a paternalist might argue that government is clearer-sighted and deliberative IN THAT INSTANCE than the individual. (At the same time, leaders' besetting vice is a tendency to lose touch with reality -- submitting them to elections by the people, sobers up the government in ways that IT needs sobering up.)
I can certainly see how two imperfect friends, who are imperfect in different ways might make each other more perfect by offering their advice and recognizing each other's area of superiority. Individuals and their democratically elected, constitutionally limited government might balance each other similarly.
As for your final conclusion, if I understand what you mean by ameliorative markets correctly (e.g., sulfur-tax or carbon-tax, cap and trade?), these seem best suited to getting agent A to recognize a cost to agent B as a cost to himself -- they "internalize" externalities.
When we come to correcting for an individual's failure to appreciate his own long-term good, they seem to fall afoul of a much more pernicious version of the problem you raise for paternalism: they attempt to use the individual's long term rationality to correct for his deficiency in long-term rationality. They try to correct individual long-term irrationality (with respect to health) by appeal to his long-term rationality (with respect to money). This seems silly, because very few people value their money more than their health -- the problem is an irrational discounting of future satisfactions brought by either health or money. What you need to compensate for the long-term irrationality of smoking is a short term inconvenience, like having to go out and smoke in the cold, not another long-term disincentive, like finding yourself more deeply in debt at the end of the year than you would otherwise have been (a disincentive you can avoid by never adding up how much cigarettes have cost you over the year).