Friday 11 February 2011

The Coalition for Fun

What do I need to do to be a founding member of Sorens's Coalition for Fun?
Formerly a discipline devoted to research on sanitation and epidemiology, public health is now more or less an explicitly ideological field devoted to ginning up panic over freely chosen, private behaviors and to cheerleading for paternalist government action to prohibit or discourage them. Take any fun activity enjoyed by those who are not urbanized, (generally) white, middle-aged, highly educated professionals – smoking, shooting, drinking, eating tasty food, calling a friend in the car, generally exercising “personal freedoms” – “public health advocates” are agin’ it. (Of course, you don’t see them agitating against marathon running or rock climbing or bungee-jumping or long-distance hiking or extramarital sex. Fun, risky things that urbanized, highly educated professionals like.)

...

The politics behind today’s public-healthery are sinister. They are driven by loathing of the poor, the overweight, people who shop at Wal-Mart and eat at fast-food joints, people with a Southern accent (who probably own guns and might even smoke) – the list goes on. Otherwise, why aren’t progressive-dominated governments trying to ban colorful local diners and casual-dining chains the way they are fast-food chains – even though the former often have higher-calorie food and don’t even report calorie counts to their customers? Of course, it’s the white middle classes who frequent quaint diners and TGI Friday’s.

So – I raise a glass of whisky and a cigar to the hope of a backlash against “public health,” a new Coalition for Fun.
Jason Sorens at Pileus nails it. I'll be guest blogging over at Pileus later this month; it looks like I'll be in very good company.

I've a piece in today's New Zealand Medical Journal taking another swipe at the healthists. The page proofs are at the end of the post for folks without a subscription.
Let’s begin with CHB’s [Peter Crampton, Hoek and Beaglehole] case against efficient markets. They argue that because models of perfect markets require a set of conditions not found in the real world, extensive and comprehensive government intervention in individual health choices is necessary. You could just as reasonably argue that because Earth has an atmosphere, we needn’t worry about falling off of cliffs: theories of gravitational acceleration of 9.8 metres per second squared are derived for a vacuum and so do not here apply.

The conditions under which markets can be shown to maximise efficiency—the benchmark case against which market failure is measured—are sufficient rather than necessary. We can be at an optimum even if the conditions fail.3 Under those idealised conditions, it is impossible to make any person better off without simultaneously making someone else worse off.

Where the idealised conditions fail, we have some guidance about policies that may improve outcomes, but do not necessarily do so. The market failure is necessary but not sufficient for policy to meliorate outcomes. Proving a particular failure does not give us carte blanche to implement any intervention we like; rather, it tells us where an intervention might be targeted. And it also tells us when intervention isn’t warranted.

As case in point, consider the potential for market failure caused by imperfect information about calorie counts. If consumers are mistaken about true calorie counts, they might eat more or less than they would under conditions of full information. Perhaps. Let’s leave aside for the moment the ease with which any consumer could investigate calorie counts at most fast food restaurants simply by checking their websites— if he actually cared. But experiments making calorie counts really salient at point of fast food purchase show no effect on purchases.4
The rather mixed evidence on the effects of information provision suggests to me that there was no real information market failure. If your reaction to the evidence is “well, let’s try a different intervention then and claim a different market failure as justification”, you’re no longer making the case based on market failure; you’re just being paternalistic. Public health activists have been abusing market failure theory to give a sciency flavour to what is actually just paternalism.5
I get really irritated at the claim that, if we have any kind of market failure, we can throw standard economics out the window. The claim's made a lot in the alcohol social cost literature - from the early Single et al papers through to Collins and Lapsley. If we take a straight neoclassical line, the failure tells us the type of interventions that might be warranted. If you still don't like the outcome after the correction's been made, that's too bad. If we take a more Austrian line, it's the presence of these failures that opens up market opportunities for entrepreneurs to come in and transform deadweight losses into profits.

I don't know whether the "any failure justifies any intervention" line comes from a deliberately opportunistic reading of principles-level texts or if we've screwed up as a profession in teaching principles of micro.

Cursed Rhetoric Page Proofs

8 comments:

  1. Of note is this recent submission to the Southern District Health Board
    http://www.odt.co.nz/regions/otago/146512/boards-submission-called-unrealistic

    ReplyDelete
  2. "...it looks like I'll be in very good company."

    Wow, really? That author wrote a terrible article. Unable to recognise that those against phones in cars are probably not motivated by paternalism, and apparently unaware of the distinction between first and second order desires ("If people are aware of the risks of an activity, and do it anyway, doesn’t that very fact show that they are better off being permitted to do it?")

    Of course worst of all was the assumption of malice behind the motives of his ideological opponents ("The politics behind today’s public-healthery are sinister. They are driven by loathing of the poor, the overweight...") Ridiculous, sloppy thinking, you deserve better company Eric.

    ReplyDelete
  3. @varkanut: Yes, bans on cell phones in cars are more plausibly motivated by worries about external costs. But the degree of risk prevented by such regulation is so tiny (no difference in risk of hands-free and phones), it does seem awfully bloody-minded.

    It's also reasonable to take seriously Elster-type concerns about self-constraint and higher order preferences while still sticking with revealed preference. First off, there are markets in the higher order constraint mechanism - gyms that make you pay more if you fail to show up and the like. If folks don't choose to use those, isn't it more plausible that the higher order preferences aren't all that strong and that regulations mandating their enforcement would be utility-reducing? Further, isn't it rather dangerous to assume that we know folks higher order preferences and consequently regulate against their revealed lower order preferences?

    Loathing might be too a strong term. But I'm not sure that contempt wouldn't be an accurate one. Why push for mandatory calorie counts on fast food restaurants but not on high end restaurants whose fare can also be fairly energy-dense? And how else can we assume that the regulators can step out to an Archimedean point where they're immune from behavioural anomalies while setting regulations?

    ReplyDelete
  4. Hi Eric, thanks for the reply. I'm not up to date on the scientific literature of cell phone usage and driving safety, but my understanding is that compared with the other things we regulate (DUI) the risks are still substantial. Citing the fact that hands-free phones are just as dangerous just potentially gives grounds for banning hands-free.

    I didn't intend to get into a debate about paternalism itself, but rather wanted to point out the sloppy/lazy thinking in that article. It's difficult to discuss a topic so broad; obviously some paternalism steps over the line (think proposals to ban smoking, or even the recent UC initiative to create a smoke-free campus), but then there are much more reasonable paternalistic policies.

    I don't think the issue of knowing the higher-order desires of others is as serious as you make it out to be. We're always hearing about how 4 out of 5 smokers want to quit, people are making resolutions to eat healthier, exercise more, manage their expenses better in the new year. It's just not that mysterious what most people consider the good life! So a policy to (e.g.) remove tax from healthy food would quite clearly, as far as I can see, help the majority to achieve/sustain their higher-order preferences, with just simple incentives and no overt regulation, and would surely maximise utility. I just don't get what the big deal is.

    ReplyDelete
  5. "We're always hearing about how 4 out of 5 smokers want to quit, people are making resolutions to eat healthier, exercise more, manage their expenses better in the new year. It's just not that mysterious what most people consider the good life!"

    Aside from the high likelihood of poor survey methods, people have strong incentives to "signal" certain preferences even if they don't actually hold them. For example, if you want to look good in the eyes of the pretty woman questioning you, you'll claim "Oh yeah, I'd love to quit smoking; such a nasty habit". I think signalling explains a lot of the discrepancy between proferred preferences and actual behaviour, not irrationality/lack of control/other market failures etc. Signalling behaviour is not a cause for government intervention ("That guy claims to love donating to charity but doesn't do it, we ought to force him!"). If it were you'd have just as good a case for banning misrepresenting preferences as regulating behaviour (same outcome: proferred preferences match behaviour).

    "So a policy to (e.g.) remove tax from healthy food would quite clearly, as far as I can see, help the majority to achieve/sustain their higher-order preferences, with just simple incentives and no overt regulation, and would surely maximise utility. I just don't get what the big deal is."

    You've not studied economics, have you? As the price of one good relative to another changes (as, for example, taxes or subsidies change the price) people will substitute from one good to the other, BUT their income will also have changed which affects their purchases of both goods. Check this out, for example: http://scienceblogs.com/obesitypanacea/2010/04/fat_tax_or_health_food_subsidy.php

    "However, as the cost of healthy foods was lowered, the total number of calories purchased actually increased. In other words, people were using the money they saved on healthy foods to purchase more unhealthy foods."

    You probably need to do a little more reading on some of this stuff.

    ReplyDelete
  6. "I think signalling explains a lot of the discrepancy between proferred preferences and actual behaviour, not irrationality/lack of control/other market failures etc"

    Pretty dubious. Do you have any evidence that attempts to impress the surveyor create bias in answers? The much simpler answer is that despite your claims, people aren't perfect and they know they have plenty of room for improvement. Hence the whole idea of second-order preferences; impulse, lack of will, addiction etc. results in plenty of people having first-order desires they're not happy with.

    That post you linked to doesn't support your argument the way you think it does. As noted by the blogger the study, though fascinating, cannot be used to draw conclusions on food subsidies etc. The subjects had to spend all the total (fixed) pool of money on food! Of course making some food cheaper is going to increase total calorie intake. More interesting would be if they allowed participants to do with the saved money what they want. That would give more solid answers.

    What about that recent study in NZ that removing GST on fresh fruit and veggies increases consumption of those goods by around 10 or 11%? Sounds quite promising to me, though of course it didn't take in to account any possible substitution toward buying more junk food.

    ReplyDelete
  7. "Pretty dubious. Do you have any evidence that attempts to impress the surveyor create bias in answers?"

    The signalling hypothesis is most commonly discussed with respect to education (do people use higher-education as a signal of intelligence/productivity/employability/etc?). Read: http://folders.nottingham.edu.cn/staff/zlizsg1/ME/Readings/Topic6/Weiss%281995%29.pdf for a good survey. He argues that there is a good case for signalling explaining education decisions.

    With respect to people "signalling" by lieing on surveys, I found a bunch of evidence on ye olde google. First, somewhat anecdotal, but interesting, evidence: http://blogs.msdn.com/b/oldnewthing/archive/2004/10/12/241228.aspx

    Second, a discussion about lieing on surveys (not very formal, mixed evidence here, though a strong impression that lieing on surveys is common): http://www.news.cornell.edu/stories/Jan09/surveys.lying.mw.html

    More on lieing, with some studies to back it up: http://www.truthaboutdeception.com/lying-and-deception/how-often-lovers-lie/experiments.html

    An interesting direct study, suggests that women lie more often to male than female interveiwers (though, surprisingly, men don't): http://karlan.yale.edu/p/LyingaboutBorrowing.pdf

    I don't mean to say that signalling/lying about preferences explains ALL of the discrepancy between stated preferences and observed behaviour, just that it can explain a lot of that gap, substantially weakening the case for intervention on the basis of people's "lack of self-control", or whatever.

    I can't find the study you're talking about. I found this, http://www.heartfoundation.org.nz/files/healthy%20eating/GST%20off%20food_background%20discussion%20paper_July%202010_final.pdf but they also point out the likelihood of substitution. And substitution is the important bit. It doesn't matter if people eat 10% more fruit after a tax change if they also eat, say, 30% more Mcdonalds: there'd be a worsening of "health", not an improvement. In what way does that sound good to you?

    ReplyDelete
  8. @James: Thanks for taking up the banner; I was light on internet over the weekend.

    @Varkanut: Do consider the horrible mess made of GST when you start exempting things. Read my prior posts on this.

    ReplyDelete