Soda Jerk

The one lesson I most want my students to learn is this: You can’t just say anything. It’s important to care about making sense. So I find it particularly galling when people violate this rule while presenting themselves to the public as economists. It undercuts the single most important lesson we have to teach.

THe latest culprit is the unchastened serial offender Robert Frank, writing in the Business section of the Sunday New York Times. His argument has two parts, one philosophical and one economic. In both cases he substitutes blather for analysis. I’m less concerned about the philosophical part, because it’s such obvious nonsense that I can’t imagine anyone will take it seriously. But the fact that he got the economics wrong, and more importantly, his implied message that it doesn’t matter whether you get the economics wrong, seems calculated to undermine the public’s faith in economists. That’s the part I take personally.

Frank’s subject this time is New York Mayor Bloomberg’s failed attempt to curb the sale of large sugary drinks. While acknowledging that such a ban would curb individual freedom in some dimensions, Frank argues that it would simultaneously enhance individual freedom in others — namely, it would enhance your “freedom” to prevent your child from drinking lots of soda.

Now, I do not doubt that for some parents, a ban on large sugary drinks would make it easier to prevent children from drinking lots of soda, but to call this an enhancement of freedom, you (or Robert Frank) would have to use the word “freedom” in a very unorthodox way. By Frank’s definition, a ban on Democratic campaign ads would enhance your “freedom” to prevent your children from voting for Democrats. Would Frank endorse such terminology? Or suggest that this effect, in and of itself, might suffice to consider the advertising ban a generally pro-freedom initiative?

Frank even goes so far as to warn us that “pro-freedom slogans provide no guidance about what to do when specific freedoms are in conflict, as they are here”. Well, one might as well say “as they are everywhere“. So in Frank’s world, you apparently can’t oppose slavery on general pro-freedom grounds, because your right to be free conflicts with my right to enslave you. Freedom of speech, freedom of the press, freedom of religion — as long as any of these conflicts with my freedom to restrict your freedom, they are not intrinsically freedom-promoting.

Well, maybe Frank’s got some general account of what “freedom” is, and maybe everything he says makes sense in light of that account, and maybe his column was just too short to hold it all. So I’m willing to let all this go. But when he gets to the bad economics, I draw the line:

In 2010, the mayor himself praised a proposal for a penny-per-ounce tax on soda in New York State…The case for reintroducing such a proposal is strong. We have to tax something, after all, and taxing soft drinks would let us reduce taxes now imposed on manifestly useful activities. At the federal level, for example, a tax on soda would permit a reduction in the payroll tax….

AAAAAAAAAAAGH! First of all, the payroll tax is equivalent to a broadbased consumption tax. If you earn a dollar and use it to buy a dollar’s worth of goods, it really doesn’t matter whether we tax the dollar as it enters your pocket or as it leaves. So what Frank is saying is equivalent to this:

At the federal level, for example, an increase in the sales tax on soda would permit a reduction in sales taxes on other goods.

Alright, let’s grant him that if we raise the tax rate on soda, we’ll probably increase the tax revenue from soda (though this is at least not certain). In that case, taxing soda more does indeed mean we can tax something else less. But by and large, it’s good policy to tax all goods pretty much equally (though there are multiple subtleties here). After all, if we tax sodas more and (say) orange juice less, then some people who prefer soda will choose to buy orange juice, just to avoid the tax. This does nobody any good, and leaves the consumer less satisfied. That’s what economists call a “distortion”. So let’s reword Frank yet again:

At the federal level, for example, we can always increase taxes on some goods while decreasing taxes on others, thereby increasing the distortionary effect of the tax system.

In other words, Frank’s argument for increasing the tax on soda is also a (bad) argument for increasing the tax on anything. Hey, if we raise the tax on newspapers, we can lower the tax on sodas. But we don’t want to do that, because we don’t want to create an artificial incentive to relax with a Coke instead of reading the New York Times — even when the Times sinks this low.

Share/Save

93 Responses to “Soda Jerk”


  1. 1 1 Bennett Haselton

    His argument rests on the fact that some people (smokers and consumers of 32-ounce beverages) are making irrational decisions that they will regret later, so you can increase the total good by distorting their present choices, to steer them toward choices that they’re less likely to regret. Your reasoning that distorting choices “leaves the consumer less satisfied” assumes that consumers are always making rational choices that they won’t regret later.

    I think the real problem with his argument is that on the one hand, he’s implying that a tax on large sodas will protect people by reducing purchases of large sodas — but on the other hand, he’s saying that such a tax would bring in more revenue, which depends on people continuing to buy large sodas even after the tax increase! (Or, at least, it depends on people not decreasing large soda purchases by a greater proportion than the proportion of the tax increase.)

  2. 2 2 Jonatan

    I agree that he made these errors. But behind this unclear thinking on his part is a point that I don’t think you address.

    In his view taxing the sodas has two advantages:

    1. It make less people drink soda. You say that this will leave the consumer less satisfied. He says this isn’t the case because “few people go to their graves wishing that they and their loved ones had drunk more sugary soft drinks.”

    2. It will bring in some tax revenue, which will allow us to lower a different tax, which actually does have harmful consequences (again, in his view.)

  3. 3 3 Flávio

    Isn’t Frank assuming a high negative externality on comsuming soda and arguing for Pegouvian tax on it? He might be wrong about this assumption, but I don’t see why it is bad economics.

  4. 4 4 Steve Landsburg

    Flavio: As I read Frank’s column, the externality/Pigovian angle is not at all the point he’s trying to make.

  5. 5 5 Steve Landsburg

    Bennett and Jonatan:

    His argument rests on the fact that some people (smokers and consumers of 32-ounce beverages) are making irrational decisions that they will regret later, so you can increase the total good by distorting their present choices, to steer them toward choices that they’re less likely to regret

    But this, of course, doesn’t make a shred of sense. If my brain contains two “agents”, one of whom wants me to drink more soda now and the other of whom wants me not to, in what sense do we “increase the total good” by favoring one of these agents over the other?

    It make less people drink soda. You say that this will leave the consumer less satisfied. He says this isn’t the case because “few people go to their graves wishing that they and their loved ones had drunk more sugary soft drinks.”

    Sure. The part of my brain that’s in charge right now wants me to drink more soda. The part that will be in charge on my deathbed wants me not to. Once again, why should public policy favor one of these over the other? Taxing (or banning) soda makes one part of me happier and another part sadder. In what sense is that a net gain?

  6. 6 6 Harold

    To re-phrase Bennett Hasselton, the whole point of this tax is to introduce a distortion in the market. The consumer choosing orange juice even though they would prefer soda is exactly the desired outcome.

    This is deliberately selected as an exception to the “by and large” rule that everything should be taxed equally in order to discourage a particular activity, and in this case it looks as though they have the economics right. You may disagree about whether the objective is a good one, but the economic argument supports their desired outcome.

    As bennet says, we have long done this with cigarettes. Raising the price through taxes has been shown to reduce smoking, and also raise money for the treasury.

  7. 7 7 Steve Landsburg

    Harold:

    This is deliberately selected as an exception to the “by and large” rule that everything should be taxed equally in order to discourage a particular activity, and in this case it looks as though they have the economics right.

    How can the economics be right when no economic argument is made? As far as I can see, there is, in Frank’s column, no economic argument for discouraging this particular activity.

  8. 8 8 RandomWords

    @ Jonatan: How is only the preference relevant that you have at the end of your life? Last time I checked life not only consisted of evaluating what you did in the past, it also consisted of LIVING.

  9. 9 9 EricK

    I reckon a lot of people do go to their graves wishing they had been a bit more reckless and done more things they would have enjoyed at the time. And surely this includes eating and drinking tasty, but perhaps unhealthy, things.

  10. 10 10 Steve Landsburg

    RandomWords (#8): Exactly. And more to the point: If Robert Frank disagrees with this, where is his argument?

  11. 11 11 Harold

    “But this, of course, doesn’t make a shred of sense. If my brain contains two “agents”, one of whom wants me to drink more soda now and the other of whom wants me not to, in what sense do we “increase the total good” by favoring one of these agents over the other?”

    There are two points here: is there a valid argument, and if so did Frank make it. It is quite possible that Frank failed to make such a case, and therefore SL points could be technically valid. It is also quite possible that there is such a case, and we can discuss this regardless of the Frank issue.

    I think it a bit rash to dismiss this whole argument as “not making a shred of sense”. We can study people and find out how they make choices and if they tend to make predictable errors. Thus if we ask smokers at various points in thier life if they made the right choice, we can evaluate whether the choices people make at any instant are likely to be the “right” one – that is the one that maximises utility over a long period. If we discover that people tend to make the “wrong” choice, it is entirely reasonable to encourage them to make the right one.

    Regardless of whether we can predict the exact nature of the wrongness, we do know that people are pretty useless at making the right choice. To deny this makes not a shred of sense. Since we know this it makes a great deal of sense to try to understand how these errors are made.

    There is also an argument from healthcare costs – I guess an externality.

  12. 12 12 TjD

    I thought the economic argument was implied. Everybody gets penalized ( through tax or insurance costs ) by having more health core costs by having more diabetes patients by having large soda’s. Tax on soda’s might generate some revenue and it ought to reduce health care costs, plus it improves the US image as a freebie.

  13. 13 13 suckmydictum

    @1 and @9

    If people actually do go to their graves regretting their past consumption of unhealthy things, maybe someone will make an argument suggesting that on their death beds, people also regret not going bungee jumping. Should the government subsidize bungee jumping for the young?

  14. 14 14 Dave

    Harold: “Regardless of whether we can predict the exact nature of the wrongness, we do know that people are pretty useless at making the right choice.”

    Yet we trust politicians to restrict our freedoms because they know better, even though they’re just useless people too?

  15. 15 15 Flávio

    I think the externality assumption is present, for example, in:

    “Evidence suggests that the current high volume of soft-drink consumption has generated enormous social costs.”

  16. 16 16 Steve Landsburg

    Harold:

    Thus if we ask smokers at various points in thier life if they made the right choice, we can evaluate whether the choices people make at any instant are likely to be the “right” one – that is the one that maximises utility over a long period.

    In what sense is the choice that “maximizes utility over a long period” better than the choice that maximizes utility over a short period?

  17. 17 17 Gordon / Brooks

    @ Harold — Well said (#11) re: utility over the long term.

    @ Steve,

    1. Picking up on Harold’s point (#11), would you have the same position that we must simply view “two agents” non-judgmentally (as if of equal value) if the example were more extreme. What if there were some narcotic immensely more instantly and unshakably addictive, practically destructive, and eventually physical-pain-inducing (if not from the high, then from other effects causing physical pain) than even the worst currently in existence, and if the only pleasure were brief and enjoyable but not more enjoyable or so different from other, less harmful drugs — not a benefit anyone would reasonably say was worth all the “costs”, yet some people were still choosing to take that first hit that inevitably brings all of those costs, because they want that momentary pleasure. Leaving aside the question of criminalization or any other government action (which is not my question), would you simply say that we can’t judge which option is “better” for that person — taking that drug or not — simply because there are “two agents”, one that wants to take that first hit, and another that, in retrospect (or even at the time), would say it’s clearly not worth it?

    2. Perhaps I’m missing something in your argument that the payroll tax is equivalent to a consumption tax. The former reduces the incentive to work, and presumably will induce some people to work less. The latter reduces the incentive to consume, and presumably will induce some people to consume less. How are they equivalent?

  18. 18 18 Jack

    Is there never a case to be made for trying to shape people’s (particularly children’s) preferences?

    When schools integrated, over the years many white children lost their innate opposition to blacks. Was this shaping if preferences not a good thing?

  19. 19 19 Jack

    In a sense, though, While i would be grateful to someone who prevented a sugar addiction/soda pref in me but would be irritated of course if I already had the addiction.

  20. 20 20 Daniel

    @ Flavio,

    I believe Steve thinks that even in an opinion piece for the NYT, you should explicitly state your economic know how in explicitly economic terms, instead of using words the average reader can understand.

    Steve,
    Frank knows that lowering income taxes increases your ability to purchase all other goods, maybe he didn’t make it explicit in the text of this piece but if you’ve ever read his intermediate microeconomics text book you’d know how much time he spends with indifference curves and how intuitively he understands them. I guess he made the wrong assumption that everyone would just assume that increasing your after tax income would allow you to spend more on all other goods, so he didn’t feel the need to explain it.

    And like Flavio pointed out here, he does describe a negative externality in perhaps not so graceful of words, and this is probably his driving economic reason for thinking imposing a tax on soda makes economic sense.

    Do you have a problem/disagree with the part of economics that deals with externalities because it seems when you discuss prices and distortions you don’t often mention how externalities can be accounted for by distorting the equilibrium price if we can calculate marginal externality per unit, and estimate the supply curve and demand curve. Is this your problem with externalities, that estimating those three things is really hard?

  21. 21 21 Cowboy Prof

    Frank does not realize that there are also costs (regrets) associated with living longer. He may go to his deathbed regretting that he drank too many sugary drinks. However, since I don’t drink many sugary drinks, I may be in my deathbed for a really long unpleasant duration with a slow terminal disease wishing I would have had more sugary drinks that would have resulted in a quicker death.

    There is an assumption among all the people looking out for my health that extended life is all benefit and no cost. This is not necessarily true.

  22. 22 22 Al V.

    Living in New York City, there are many things that have negative externalities: cars, dogs, smokers, people handing out fliers. I live near a Catholic School, and the school buses dropping off children impede the city bus I take to get to work every day. I don’t care about sugary drinks, but I would like a tax on school buses.

    It Frank appears to be claiming that it is acceptable to either tax or restrict sugary drinks because it transfers a cost from parents to others. But why? Parents choose to have children (ideally), so why should we subsidize their choices? Similarly, why do state and federal governments grant a tax deduction for children? All this does is transfer the cost of government from parents to non-parents.

  23. 23 23 Steve Landsburg

    2. Perhaps I’m missing something in your argument that the payroll tax is equivalent to a consumption tax. The former reduces the incentive to work, and presumably will induce some people to work less. The latter reduces the incentive to consume, and presumably will induce some people to consume less. How are they equivalent?

    Gordon/Brooks: The technical answer is that they lead to exactly the same Euler equations and hence exactly the same consumption and labor choices. The intuitive answer is that a tax on consumption makes work less desirable, because we work in order to consume, and a tax on work makes consumption less desirable, because consumption requires work. It’s exactly this that Frank appears to forget when he suggest that it’s better to tax consumption goods than to tax wages.

  24. 24 24 Bob Murphy

    “So I find it particularly galling when people violate this rule while presenting themselves to the public as economists.”

    Oh thank goodness… For a terrible moment I thought Steve was going to link to one of my blog posts.

  25. 25 25 Bob Murphy

    It’s not the same point Steve was making, but for those who want to see how Frank’s “let’s tax bads not goods” argument is far too simplistic, check out my EconLib article on the so-called tax interaction effect. Even something as “obvious” as taxing carbon rather than labor, is not as clear-cut as pro-carbon-tax people suggest.

  26. 26 26 Neil

    The government should put a tax on half-baked externality arguments.

  27. 27 27 Daniel

    @ Bob Murphy,

    Great article and I agree with you that we should account for the cost of collecting taxes and the distortion effects when deciding on an optimal tax levels.

  28. 28 28 Jonathan Campbell

    Steve – do you think we do a disservice to heroion addicts when we take away their needles?

  29. 29 29 Ken Arromdee

    The intuitive answer is that a tax on consumption makes work less desirable, because we work in order to consume, and a tax on work makes consumption less desirable, because consumption requires work. It’s exactly this that Frank appears to forget when he suggest that it’s better to tax consumption goods than to tax wages.

    A tax on consuming *everything* may have the same effect as a tax on work (subject to some qualifiers–for instance, is using the money to make an investment counted as consumption). But a tax on consuming a particular item may not necessarily have the same effect; if work becomes less desirable because it is harder to work to get more of that particular item, you could substitute other items that are nearly as good as the first one.

  30. 30 30 Harold

    #15 SL: “In what sense is the choice that “maximizes utility over a long period” better than the choice that maximizes utility over a short period?”

    Perhaps I am mis-using the term. I am imagining integrating the area under a utility / time curve. If I get a “spike” in pleasure worth say 10 “utils” (u) for 10 seconds, then I have a total utility of 100us. If I get to 8 units, but it lasts for 20 seconds, I have 160us, so greater value. Is this a reasonable way to think about this, or have I got this wrong?

    This is the sort of evaluation I make when choosing between one delicious chocolate, or a bag of mediocre confectionary. The individual chocolate is undoubtably a higher peak, but I may choose the whole bag because it lasts longer. I would then say that the bag has a higher utility, but this must incorporates some time element.

    I want “utility” to mean “value” in some reasonable sense. If it only means what I choose right now based on my current preference, then it means something like attractiveness, or desirability. This has no necessary connection to value. If this is all it means, then it has little to say about what we should or should not do, and the whole project is, if not entirely pointless, then very significantly diminished.

  31. 31 31 Steve Landsburg

    Harold: Utility in classical economics is defined to be that which is maximized. If people choose sugary drinks, those sugary drinks must contribute to maximizing the discounted sum of lifetime utility.

    In order to incorporate things like addiction, regret, limited self-control, etc., once introduces notions like “hyperbolic discounting”, which amount, in effect, to saying that your brain is inhabited by competing agents with different lifetime utility functions. Maximizing one of those functions necessarily entails failing to maximize the other.

    So why should we prefer one of these agents’ utility function to the other’s?

  32. 32 32 Ken B

    “So why should we prefer one of these agents’ utility function to the other’s?”

    Because we know better.

    Or at least, that’s the reason people always give when you boil it down. It’s true for your 8 year old kid. I’m not convinced it’s often true of your 38 year old neighbor, but that’s who it always gets applied to.

  33. 33 33 Gordon / Brooks

    Forgive me, but in response to the “Soda Jerk” title, here’s a movie scene ending with a great line (note: contains profanity)
    http://www.youtube.com/watch?v=MNUbLbRslyk

  34. 34 34 Bearce

    This is a really good post and exercise. However, there still does seem to be some confusion amongst psters, these ones particularly stand out.

    2. It will bring in some tax revenue, which will allow us to lower a different tax, which actually does have harmful consequences (again, in his view.)

    I thought the economic argument was implied. Everybody gets penalized ( through tax or insurance costs ) by having more health core costs by having more diabetes patients by having large soda’s. Tax on soda’s might generate some revenue and it ought to reduce health care costs, plus it improves the US image as a freebie.

    Individuals prefer a payroll tax to a sales tax. This is because when you tax a certain good it has both an income and substitution effect, thus reducing overall welfare significantly. A payroll tax just has an income effect, which has a less distorting effect.

    As far as reducing healthcare costs by increasing revenue through a soda tax, you could argue that. And to be fair, he does argue Pigovian-like at the very end of his article (mentions ‘social costs’ of drinking soda). But if we taxed everything that raised insurance premiums, should we not place a concussion tax on parents who allow their kids in pop-warner football or individuals who find it thrilling to snowboard down steep mountain slopes and risk breaking bones? In my opinion, this argument opens the gateway for a lot of healthcare related taxes for pretty much any risky activity. A soda externality does NOT have the same kind of externalities in magnitude as smoking does.

  35. 35 35 Floccina

    What Frank really thinks:

    Fat people are ugly, I do not like to see lots of fat people around. If we tax soda fewer people will be fat, so I am for restrictions and taxes on soda.

  36. 36 36 Harold

    #30 “So why should we prefer one of these agents’ utility function to the other’s?”
    This is a question that must be answered, I agree. Thanks for putting it like that.
    If we view a person as a collection of these agents, each with its own utility function, then the actions of one agent affect the other. So maximising the addict’s utility perhaps diminishes the utility of the father agent, the husband agent, the motorist agent etc. The sum of the ultilities of these other agents is much greater than that of the addict, so total utility is maximised by diminishing the addicts utility. These agents form a market, but exchange is not always possible between them.

  37. 37 37 Floccina

    BTW I think that there is a better argument to be made and that is:

    The majority of people would like to drink less soda but fall to temptation when the time comes and so want to remove the option from themselves ahead of time. It is like locking the refrigerator before you smoke some pot and giving the key to friend until the munchies pass. This argument has weaknesses also but it seems more defensible to me.

  38. 38 38 Gordon / Brooks

    An NYC negative externality:

    I live in Manhattan and ride the subway often. A VERY common occurrence (I’d say the norm) is that at least one person is wide enough that they, in effect, occupy 3 seats*, making two reasonably slender people stand (often very uncomfortably squeezed together) who otherwise could sit. If we could efficiently tax that directly I’d be all for it. But for now, just want to note that externality of excess caloric intake in NYC.

    * Most subways have rows of individual concave, seat-like spots to sit. Some just have essentially a long, flat bench, but I assume the former have been favored because without individualizing the benches, people probably often sit too far apart, reducing the seating capacity in practice.

  39. 39 39 iceman

    On the HC costs, I would just note that a *self-imposed* externality seems like something a bit different. “Good news – we’ve decided to pay your medical bills! Bad news — you are now prohibited from doing X,Y and Z because it’s too expensive for us to cover. What’s that?…sorry no you can’t opt out.”

    Maybe we have to tax soda once we account for the enormous psychic benefit derived by those CowboyProf calls “all the people looking out for my health”. Maybe the infamous “utility monster” is just someone trying to save your soul.

  40. 40 40 Steve Landsburg

    Bearce:

    Individuals prefer a payroll tax to a sales tax. This is because when you tax a certain good it has both an income and substitution effect, thus reducing overall welfare significantly. A payroll tax just has an income effect, which has a less distorting effect.

    This is true only if labor is supplied inelastically (actually, not even then — we also need a bunch of cross-elasticities to be zero), in which case it is equally true of a broadbased sales tax.

  41. 41 41 Ken

    Jack,

    You seem to not know that segregation was government imposed preferences. Schools were and are government run. This is yet one more argument to eliminate this gov monopoly. And no, the gov does not have the right to force racial integration in privately run schools. You have the right to spend the money you earn and you have the right to associate with those you want and ignore those don’t.

  42. 42 42 Jack

    Ken,

    Are you telling me that without government segregation would have never happened? What do you mean by government imposed preference?

  43. 43 43 Al V.

    @Floccina 34, I’m all for soda taxes. In fact, I favor 1000% soda taxes, if we can use the increased revenue to reduce cigarette taxes. I’m sure you can guess why.

  44. 44 44 Daniel

    Ken,

    We have very few examples of education systems that are not held by the government. Can you give me a few examples, so I can get a better idea of where you’re coming from.

  45. 45 45 Henri Hein

    Jonathan, @27:

    I cannot speak for Steven, but I do.

  46. 46 46 Henri Hein

    Harold @11:

    “we do know that people are pretty useless at making the right choice”

    I could not disagree more. This is double wrong: the proposition is wrong, and the implication is wrong. In general, people turn out to make pretty rational choices, when they make choices for themselves and there are not distortive effects.

    The implication, namely, that therefore we should have some other people make choices for them, makes neither philosophical or logical sense. People are decent, though not perfect, at making choices for themselves. People have turned out to be terrible, terrible at making choices for others.

  47. 47 47 Ken

    Jack,

    Yes, that’s exactly what I’m telling you. Segregation occurred due to government imposed preferences of separating the races. That these laws came into existence at all is proof that the private sector didn’t mind mingling of the races. If people voluntary separated by race Jim Crow laws would have been unnecessary and irrelevant.

    Daniel,

    What are you talking about? Examples of what? Largel non-gov owned “education” system? This was largely the case in the 18th and 19th centuries.

  48. 48 48 nobody.really

    Classical economics suggests that, all else being equal, society can optimize welfare by reducing people’s constraints to arrange their own circumstances – that is, maximizing people’s freedom to choose – assuming people have free will. But do they?

    In The Big Questions, Lansdburg answers yes – yet, if I understand correctly (and I may well be mistaken), Landsburg also believes in determinism. We are free to follow our preferences – but we are not free to choose our preferences, or to refrain from following them. In short, my behavior is (purely?) a function of physical phenomena both within and beyond my body. But public policy can alter the phenomena that will dictate my behavior.

    I make my choices within the constraints of culture – a culture I do NOT choose. In other words, culture is this big, fat externality, affecting people who have not chosen to be affected. I cannot know in the abstract whether policies that alter my culture will make society better off or worse off. True, those changes may be imposed upon me without my consent – but the existing culture was also imposed upon me without my consent, so it’s not obvious why I would think in the abstract that either culture is more intrusive on my autonomy than the other.

    Some people opposed the soda regulation on the grounds of “freedom to choose.” I understand Frank to argue that a person’s propensity to drink sugared beverages is heavily influenced by culture. I understand Frank to argue that governments should have some discretion to alter culture, just as they might seek to control other externalities. That is, the focus of Frank’s argument is that no one in the US chooses his own culture, so no one is in the position to object on the grounds of “freedom of choice” that the culture is being changed.

    Frank provides little argument to support the view that a culture that provides less support for the consumption of sugary beverages is better than our current culture. I suspect Frank opted not to expend text supporting this proposition because 1) he regards the proposition as popularly accepted and 2) defending the proposition would draw focus away from his main point, discussed above.

  49. 49 49 Daniel

    @Ken,

    So I’m still not sure what you’re suggesting. Do you suggest that we move to a system of education in which the government neither subsidizes the purchase of privately run education systems nor provides it directly through state and local government, but rather individuals choose how much they want to spend on their children’s education, or were you proposing that parents get a subsidy for education?

  50. 50 50 Ken

    Daniel,

    Do you suggest that we move to a system of education in which the government neither subsidizes the purchase of privately run education systems nor provides it directly through state and local government, but rather individuals choose how much they want to spend on their children’s education…?

    Yes.

  51. 51 51 Daniel

    Ken,

    I suppose if you could convince an individual state to test this out there would be nothing preventing it right? Why do you think we’ve ended up with 50 states adopting a mixed public/private school system?

    Separately, should a child’s level of education be based on their parents income/education at birth? That seems like what would be the de facto result of your proposal.

  52. 52 52 Ken

    Daniel,

    I suppose if you could convince an individual state to test this out there would be nothing preventing it right?

    I do what I can, but when was the last time government actually shrank? Also, teacher’s unions are some of the best institution through with to direct tax payer money into the pockets of politicians.

    Why do you think we’ve ended up with 50 states adopting a mixed public/private school system?

    Control. Specifically indoctrination.

    Separately, should a child’s level of education be based on their parents income/education at birth?

    Should a child’s house in which he lives be determined by the value his parents create? Should the type of clothes a child wears be determined by the value his parents create? Should a child’s diet be determined by the value his parents create? Should the material level at which a child lives depend on the value his parents create?

    The answer to those questions and yours is clearly yes.

    That seems like what would be the de facto result of your proposal.

    So? The poorest Americans live better than 99% of all humans that ever lived. They live this way due to the wonderous creativity of the private sector, often in spite of the roadblocks thrown up by politicians and myopic bureaucrats.

    Additionally, you falsely think that government run education is actually good for the poorest, but you’re wrong about this. A recent article in NY discussed how 80% of students entering college need remdial courses before doing actual college work. This means that 80% of the academically strongest students are functionally illiterate. Just how well educated do you think the poorest students are faring in our government run, union choked schools?

    Did you know that literacy rates were higher in the US during the late 1700′s than today? The Federalist Papers were published in the newspapers of that time. Today’s newspapers are written at a sixth grade level.

    You worry about the poor in private markets, but fail to understand that the only system in recorded history that lifted anyone without an army from absolute poverty to the astonishing material comfort that the poorest Americans enjoy are privately run markets. If you were truly worried about the poor, you’d be worried about the encroaching police state and using that police state to indoctrinate children.

  53. 53 53 Daniel

    Ken,

    “So? The poorest Americans live better than 99% of all humans that ever lived. They live this way due to the wonderous creativity of the private sector, often in spite of the roadblocks thrown up by politicians and myopic bureaucrats.”

    Some would argue that it was free markets, well regulated throughout most of the postwar 20th century that brought much of the gains in the period following the great depression. I’m not saying they’re necessarily right, but I know many economists that make this argument. Certainly life is much better today than it was at the beginning of the 20th century for the poorest, and some would argue that the social safety net has at least some to do with that.

    ” A recent article in NY discussed how 80% of students entering college need remdial courses before doing actual college work. This means that 80% of the academically strongest students are functionally illiterate. Just how well educated do you think the poorest students are faring in our government run, union choked schools?”

    Citation please. Also, this isn’t relative to anything, so how am I supposed to evaluate it? Perhaps colleges have more recently discovered that remedial classes make students more likely to graduate from college. More students in the U.S. are also going to college than ever before so this could effect an increase in the level of students that require extra makeup classes before being on par with their peers. It seems the last half of the 20th century has seen a higher percentage of the population graduate from college than any other point in history. It could be due to lax standards, but you haven’t provided me with enough information other than to point to a single data point, and say see government must be failing.

    “Did you know that literacy rates were higher in the US during the late 1700′s than today? The Federalist Papers were published in the newspapers of that time. Today’s newspapers are written at a sixth grade level.”

    Really, do you have any data to support your supposition? At least from 1870 onward it seems we’ve drastically increased literacy rates.

    http://nces.ed.gov/naal/lit_history.asp

    I know, it’s produced by the government, but if you have an independently run estimate of literacy overtime than present it, instead of just saying something based on the fact that federalist papers were published in newspapers at the time. How many people read the federalist papers? Also in the 1700′s how many academic journals were there? How many engineers, scientists, and mathematicians did we have as a percent of the population versus today?

    It’s easy to say that it was the free market that created this enormous growth in wealth, but over the same time period that we’ve seen this enormous growth of wealth, we’ve also seen an enormous growth in public involvement in education. Provide me with some links that support your position based on actual data and I can evaluate your opinion based on reality rather than your logical assumptions.

  54. 54 54 Mike H

    There’s a pwoerful principle being discussed here :

    “few people go to their graves wishing that they and their loved ones had spent more time at work” Therefore, we should ban work.

    “few people go to their graves wishing that they and their loved ones had bought more cars” Therefore, we should ban car sales.

    “few people go to their graves wishing that they and their loved ones had eaten less ravioli” Therefore, we should make ravioli compulsory.

    “few people go to their graves wishing that they and their loved ones had run for president” Therefore, we should ban presidential nominations.

    “few people go to their graves wishing that they and their loved ones had paid more consumption tax” Therefore, we should scrap all consumption taxes.

    “few people go to their graves wishing that they and their loved ones had spent more time commenting on blogs” Therefore, um..

  55. 55 55 Mike H

    “few people go to their graves wishing that they and their loved ones had spent more time pointing out the silly typo in the previous comment”

  56. 56 56 Jonathan Campbell

    Henri – do you think there is ever an instance when it is a good idea to convince someone that he should change his mind about a plan he has (e.g. shoot up heroin, drink soda, commit suicide)?

    If so, do you judge it as unfortunate to the extent you are never able to convince him? (if not, why bother trying in the first place?)

    And lastly, to the extent that you think it is ever unfortunate, can you be sure that the the fact that compelling him to change his mind, while being costly since it infringes on freedom, is worth it since it avoids the unfortunate outcome of him doing the stupid thing?

  57. 57 57 Steve Landsburg

    Mike H (#53): Bravo.

  58. 58 58 Henri Hein

    Jonathan:

    First, let us establish some semantics. There may be cases where it is a good idea to convince a close friend, family member or partner to do something different, for their own sake. I cannot think of a situation in which it makes sense for a stranger to try to convince someone to do something different. At least, not beyond some friendly nudging or helpful pointers.

    That said, I don’t take it for granted it is unfortunate if such convincing sometimes fails. Sometimes it is bad, sometimes it is just as well.

    Lastly, I am sure of only one thing. I am not sure I know what is best for a close friend, and I am not sure if my close friends know what is best for me. I *am* sure that no stranger knows what is best for any of us.

  59. 59 59 Mike H

    @Henri #57 “There may be cases where it is a good idea to convince a close friend, family member or partner to do something different, for their own sake. I cannot think of a situation in which it makes sense for a stranger to try to convince someone to do something different”

    If it’s truly for their own sake, why would it matter who is doing the convincing??

  60. 60 60 Harold

    Steve: “Utility in classical economics is defined to be that which is maximized. If people choose sugary drinks, those sugary drinks must contribute to maximizing the discounted sum of lifetime utility.”
    Then utility itself is not important to people. Maximised utility may have very little to do with anything that matters. Free markets maximise ultility, but why should that be in any way important? We must assume that utility has some relationship to something important. We must assume that there is some reason why people make choices.

    #45 Henri Hein: We are talking about a comparative here – I can argue that people are both good and bad at making these choices. Is Usain Bolt good at the 100m? Compared to every other human, then yes, but compared to a cheetah, or a motor bike, then not so good.

    We do know from research (yes, empirical data – we must not ignore it) that people are swayed by factors like the framing effect. This is just one example of how people are bad at making choices. I can come up with loads of others. Steve has done so with red balls. We know that people are bad at evaluating anything with lots of data because they make different evaluations depending on how the data is presented. We know that people ignore evidence they don’t like. I will provide references and examples if you wish, but I think this sort of stuff is well known. Given the weight of research it is difficult to maintain that we are “good” at these things if to be good means consistency, weighing evidence and avoiding mistakes.

    However, this does not mean we are necessarily “bad”. We do have a pretty good processor, and we know a lot about ourselves.

    So by some measures at least we are not very good at these things, but we do not necessarily know if there is any better way. My contention is that it is ignoring reality if we start from a supposition that people make the right choices all the time.

  61. 61 61 Jack

    Excuse me for the stilted writing as im writing on my phone, but I believe we should have a policy that nudges or at least educates (hopefully right on the bottle) parents to feed or allow their children as few soft drinks as possible…yes, a soft drink every so often, one every few days even, is not going to the end of the world for your child’s health.

    but when you start drinking more than, say, one a day, especially in an otherwise unbalanced diet, folks that is the road to diabetes, sugart Soft drinks provide empty calories. If you’re eating a lot of calories, you should at least get some nutrition out of it, give your kit ice cream not soda! at least the isulin spike shoild in theory be less. and especially if you’re a kid when you need nutrition most! sure, screw the adults age 18 and older they can do what they want. And again a coke every now and then isnt going to kill you, but parents should know it’s not good for their kids to drink all the time.. especially as most of them have lots of caffeine too.. Because they are addictive and they provide no nutritional value and daily, above-average consumption can lead to real disease (diabetes).

  62. 62 62 Ken

    Daniel,

    about 60 percent of the population was literate between 1650 and 1670, a figure that rose to 85 percent between 1758 and 1762, and to 90 percent between 1787 and 1795. In cities such as Boston, the rate had come close to 100 percent by century’s end.

    A long-awaited federal study finds that an estimated 32 million adults in the USA — about one in seven — are saddled with such low literacy skills that it would be tough for them to read anything more challenging than a children’s picture book

    One in seven puts the literacy rate today at 85%, which is less than 90%.

    Nearly 80 percent of New York City high school graduates need to relearn basic skills before they can enter the City University’s community college system.

    Provide me with some links that support your position based on actual data and I can evaluate your opinion based on reality rather than your logical assumptions.

    It’s unlikely that you will do this. I’m sure you don’t have the time much less the training to do this evaluation. That you claim in your opening statement much of the gains in the US occurred after 1945 shows that you don’t live in reality at all.

    Your second to last sentence also shows just how little understanding you have, mixing up correlation with causation. The reality is that as Americans became richer, so did politicians ability to take more wealth and intrude ever more into people’s lives. It’s no surprise at all that government grew at the same time people’s wealth did. In fact, it’s expected.

  63. 63 63 nobody.really

    I commend Ken for making a valiant effort to document changes in literacy rates over centuries. That said, I suspect we must take these reports with a grain of salt.

    about 60 percent of the population was literate between 1650 and 1670, a figure that rose to 85 percent between 1758 and 1762, and to 90 percent between 1787 and 1795. In cities such as Boston, the rate had come close to 100 percent by century’s end.

    A little more context may be appropriate:

    As University of Colorado historian Gloria L. Main says, “Measuring ‘literacy’ is problematical . . . because the term encompasses not one skill but two, each with a separate range of competency”—that is, the ability to read and the ability to write. In the eighteenth century, the two were taught as separate skills. Many people learned the rudiments of reading but never to write—in those cases, a mark in place of a name may give us a false negative about someone’s ability to read. It’s also possible that some people who could not read, or could not read well, learned the alphabet well enough to write their names. In those cases, signing a name gives a false positive.

    Despite the caveats, we can generalize about patterns of literacy. In 1974, University of Montana scholar Kenneth Lockridge’s groundbreaking book, Literacy in Colonial New England, surveyed evidence from legal records and offered provisional conclusions — “The exercise is bound to be tentative, as it uses a biased sample and an ambiguous measure” — but made the case that, among white New England men, about 60 percent of the population was literate between 1650 and 1670, a figure that rose to 85 percent between 1758 and 1762, and to 90 percent between 1787 and 1795. In cities such as Boston, the rate had come close to 100 percent by century’s end.

    Lockridge and his successors showed that literacy was higher in New England and the mid-Atlantic colonies than in the South, and higher in the cities than in the countryside. Traders and shopkeepers were more literate than farmers. They showed that American literacy was high by European standards. As the University of Delaware’s F. W. Grubb wrote in 1990:

    Of all European countries perhaps only Scotland surpassed America in literacy by 1800. Not only had the European literacy revolution been transplanted to the American periphery during the colonial period, but colonial literacy had somehow leaped past that of northwestern Europe.

    In short, the statement about a literacy rate of 90% was a statement about white men in New England. It was explicitly not a statement that could be generalized to people outside of New England. It was explicitly not a statement that could be generalized to areas that were more rural. And it was implicitly not a statement that could be generalized to other minority groups, or to women.

    Moreover, given the inherent ambiguity in the term “literacy,” it is far from clear that different authors use the term to mean the same thing. Thus it’s unclear that reports about illiteracy today really mean the same thing as reports about illiteracy in 1800. Do 1 in 7 people today sign legal documents with an X? By that measure of literacy, I suspect contemporary US society is doing pretty well.

    (That said, you can find various blog posts about how people get away with signing credit card receipts with pretty much anything at all – Xs, wavy lines, doodles of trees and bunnies – so perhaps the ability to sign your name to legal documents is not a useful measure of literacy anymore.)

  64. 64 64 iceman

    Someone here mentioned orange juice – I just checked my fridge and OJ has the same calories and about ¾ the sugar as Mountain Dew.

    nobody.really: “governments should have some discretion to alter culture…the existing culture was also imposed upon me without my consent, so it’s not obvious why…either culture is more intrusive”

    Important question, my take: “culture” reflects the net aggregate result of *individual* preferences and choices. It’s not monolithic but sets social norms that make one gathering of people different from another in interesting ways. “Imposed” in this context is too strong. To say one is helplessly influenced by the preferences of others is to say one cannot function in the real world. Do you see any difference between a govt giving people information about their choices versus directly prohibiting certain choices? To me the former will reflect an expression of culture in the degree to which a given group adjusts its behavior. (BTW I hear greater awareness has in fact been causing obesity rates in the US to level off/decline recently, before things like soda bans.) The latter does not “shape” culture by fiat, it stifles it. We can force a group to take their customs etc. underground, but you haven’t changed culture if you haven’t changed attitudes and beliefs.

    Henri Hein #45 – nice. PJ O’Rourke once described a decision box divided into quadrants: if I’m buying something for myself with my own money, I’m strongly motivated to get good quality at a good price…by the time I get to buying something for someone else with someone else’s money, “who gives a sh!t?”

  65. 65 65 Daniel

    @ Ken,

    I second what nobody.really said. You can’t just fit two very different datasets together and assume they’re measuring the same thing. I won’t get in to an insulting war with you, but it seems if you had any sort of statistical training you should know this. If you did know this and you chose to present it anyway it seems you’re manipulating data to make a point.

    “It’s unlikely that you will do this. I’m sure you don’t have the time much less the training to do this evaluation. That you claim in your opening statement much of the gains in the US occurred after 1945 shows that you don’t live in reality at all.

    Your second to last sentence also shows just how little understanding you have, mixing up correlation with causation. The reality is that as Americans became richer, so did politicians ability to take more wealth and intrude ever more into people’s lives. It’s no surprise at all that government grew at the same time people’s wealth did. In fact, it’s expected.”

    Again we go to insulting people’s intelligence instead of providing actual data. I very much understand that correlation is not causation, but it should at least make you take a pause in your assumptions. There are things that I think the government is good at and things I think it is not good at. If you claim to have some expert knowledge about why government run education is so horrible you should share your sources instead of simply pointing to single data points. It just seems obvious to me that when people don’t have very good backup for their assumptions their first reaction to criticism is to start insulting the person that’s challenging them.

    To your point about most of the gains not coming after 1945, it all depends on what you find to be an important increase in quality of life. I find, the internet, rapid decrease in death from heart disease, and the ability to have anything I want shipped to me at anytime, anywhere a pretty big chunk of the increase in quality of life, but others may disagree. That’s not my point, my point is that an increase in government at our current levels of spending hasn’t seemed to put much of a damper on innovation or increases in worker productivity. You may have evidence that explains away this lack of negative correlation, but I’ve yet to see it.

  66. 66 66 nobody.really

    nobody.really: “governments should have some discretion to alter culture…the existing culture was also imposed upon me without my consent, so it’s not obvious why…either culture is more intrusive”

    Important question, my take: “culture” reflects the net aggregate result of *individual* preferences and choices. It’s not monolithic but sets social norms that make one gathering of people different from another in interesting ways. “Imposed” in this context is too strong. To say one is helplessly influenced by the preferences of others is to say one cannot function in the real world.

    Thanks for responding to my remarks in English; that’s my native language. Some might argue that I somehow chose English as my native language, and that to say that this choice was imposed upon me by culture simply reflects my inability to function in the real world. I see things differently.

    Do you see any difference between a govt giving people information about their choices versus directly prohibiting certain choices? To me the former will reflect an expression of culture in the degree to which a given group adjusts its behavior…. The latter does not “shape” culture by fiat, it stifles it. We can force a group to take their customs etc. underground, but you haven’t changed culture if you haven’t changed attitudes and beliefs.

    Yes, I appreciate distinctions between informing, nudging, taxing, prohibiting, compelling. I acknowledge the advantages of pursuing behavior changes through less-intrusive means. Yet I believe all of these strategies can influence culture. Yes, people can resist efforts to change culture. But I believe that behavior changes (whether induced through information, nudging, taxing, etc.) can also change culture — even as people resist those changes.

    Bloomberg’s policies may have been unwise, but I don’t regard them as unreasoned. These policies may be disrespectful of people’s different preferences – but this is true of culture in general. Relative to the impositions of culture, Bloomberg’s impositions at least have the benefit of having been democratically adopted.

  67. 67 67 Jack

    Iceman: “Someone here mentioned orange juice – I just checked my fridge and OJ has the same calories and about ¾ the sugar as Mountain Dew.

    At least with orange juice you’re not getting caffeine and are normally getting a multitude of naturally-occurring vitamins. … And if it’s pulpy perhaps some fiber that will slow the insulin spike.

  68. 68 68 Jonathan Campbell

    Henri (#57) – suppose a stranger isn’t aware that he will fall off a cliff if you do not try to convince him to swerve?

    Obviously this is a little different since it involves merely conveying information to the stranger rather than purely trying to prescribe behavior. However in practice there may not be a bright line between the two types of convincing.

  69. 69 69 Henri Hein

    Jonathan:

    Granted, though I would distinguish between conveying information and trying to convince someone. If I met climber in Yosemite who were planning on climbing East Buttress on El Capitan, but were not sure of the difference between a nut and a cam, I would have a conversation with them. However, if they were still bent on going, and I didn’t think they’d be a danger to others, I would let them go and make their mistakes.

    I think there is a line between information and action. If you want to educate drug users about the dangers of that lifestyle, I don’t have a problem with it. But taking their needles (or any other paraphernilia) away is pernicious.

    Aside: In Copenhagen, they do the opposite of your original suggestion. They give addicts clean needles. There is no measurable impact on drug use, but contagious diseases, HIV in particular, dropped after they started with that.

  70. 70 70 Henri Hein

    @nobody.really:

    You are one of my favorite commenters here. I rarely agree with you, but I always enjoy your posts.

  71. 71 71 iceman

    #65 – “I acknowledge the advantages of pursuing behavior changes through less-intrusive means”

    So we agree that some types of changes are in fact “more intrusive on my autonomy”…that’s all I was really trying to establish. Of course we can force behavioral changes, and eventually they can become ingrained in social norms (at least to the extent we can control black markets)

    How are “natural” cultural influences not in general adopted at least as democratically?

    BTW clever line :) but language doesn’t seem like the most relevant component of culture as it relates to the kinds of behavioral response differences we’re discussing here. Sure the noises we make to convey the concepts of “soda” and “obesity” sound different in different localities, but the issues remain.

  72. 72 72 Henri Hein

    @Harold,

    I’m an empericist myself, so I don’t have a problem with that. Here are three problems I do have.

    You establish that people are not perfect at making choices, something I already stipulated, and then assume they are bad at it. That is an unwarranted jump. We also have a load of evidence that people respond rationally to incentives, towards their own interest, and that they understand their own preferences pretty well.

    My second problem with a lot of these studies is that there is a ‘right’ choice, which is often projected by the researcher or reporter writing about it.

    Here is my fundamental problem: even if we assume people are sub-optimal at making choices, it does not follow that a third-party can help them. Your original contention, and Frank’s, rephrased as an example, is that Ally is not good and making choices, so therefore we should have Bob making that choice for her. By the premises assumed, Bob is not good at making choices for himself either. So I don’t know why I would think Bob is any good at making choices for Ally.

  73. 73 73 nobody.really

    How are “natural” cultural influences not in general adopted at least as democratically?

    BTW clever line :) but language doesn’t seem like the most relevant component of culture as it relates to the kinds of behavioral response differences we’re discussing here.

    Consider the case of les Québecois (sp?): They want their kids to grow up speaking French. They live in a sea of English speakers. Their mechanism to preserve the French language is to have language regulations that make Bloomberg’s proposal seem like the mildest nudge in comparison. They recently tried to crack down on an Italian restaurant for listing “pasta” on the menu, instead of employing the French term which is … well, no one could say what it is. (Restaurants in France have the temerity to use the word “pasta,” but what do they know?)

    This language control policy seems intrusive as hell. Yet in the absence of some mechanism, it also seems very likely that each generation’s use of French would dwindle. Conversational French in Quebec is already riddled with English nouns; whatever its practical merits, this language definitely lacks the euphonious cadence of the continental version. So, if you really care about this stuff, what do you do?

    You could look to the example of the Europeans who first settled this land: the Pilgrims. (Forget the Vikings already.) They came in pursuit of freedom – religious freedom! Except that they already had freedom in Holland. But the problem was that there was too damn much freedom in Holland, and their kids were growing up to become freedom-loving Dutch! So the Pilgrims crossed an ocean and endured a winter than killed off half their members not for freedom, but for control. And they got what they were looking for: control of language, religion, fashion, music videos, the whole shebang.

    So if you consider the crazy shit the Pilgrims went through just to control their own damn kids, I guess the Quebecois don’t seem so bad. And Bloomberg seems downright benign.

    I rarely agree with you, but I always enjoy your posts.

    Be assured, I’m doing my best to change that.

  74. 74 74 iceman

    “They live in a sea of English speakers.”

    So what’s more democratic, allowing the language most people prefer to speak to prevail, or sending around menu police?
    (The more I learn about Quebec the less surprised I am that it lags other provinces economically.)

    “The Pilgrims crossed an ocean” — *voluntarily* of course — “to control their own damn kids” – kids are a different matter; Bloomberg is doing this to *consenting adults*.

    Still looking for a good example of a “natural” cultural factor that I would consider “intrusive” in the same sense as govt imposition.
    But I could send you an old betamax player if you still feel aggrieved about that one.

  75. 75 75 Ben Southwood

    On the less mentioned freedom point: there is one highly respectable view of freedom that is highly cogent and plausible, that holds there is a fixed amount of freedom; the question is its distribution among individuals.

    Hillel Steiner makes the case here: http://www.jstor.org/discover/10.2307/4544864?uid=3738032&uid=2&uid=4&sid=21102052860317

  76. 76 76 Harold

    #72 Henri Hein: “You establish that people are not perfect at making choices, something I already stipulated, and then assume they are bad at it. That is an unwarranted jump.”

    As the term “bad” has no strict definition, as I explained, it is a perfectly good term to apply, given the weight of evidence. Perhaps we can agree that we are bad at some aspects of decision making. Link to a Wiki list:
    http://en.wikipedia.org/wiki/List_of_biases_in_judgment_and_decision_making

    Once we have established (as we have) that the decisions people make when selecting a large soda (or anything) are less than perfect, we must accept that in principle it is possible to improve on the choices individuals make. If these choices erred in a random way, then in practice it would be impossible to apply a blanket policy that would improve the outcome. However, if people all tended to make the same error, then it would be possible in practice as well as in princiiple to improve the choices people make. The important thing is that these interventions must be based on proper research. Otherwise we may expect them to suffer from the same problems as the original choice.

    Confirmation bias is one well documented cognitive error. People consistently interpret data that supports their existing belief or desire.

    Hyperbolic discounting is where people “irrationally” under value events more distant in time. For example: “when offered the choice between $50 now and $100 a year from now, many people will choose the immediate $50. However, given the choice between $50 in five years or $100 in six years almost everyone will choose $100 in six years, even though that is the same choice seen at five years’ greater distance.” This produces choices that are inconsistent over time. I believe this explains why we should favor some agents over others in Steve’s multi agent description. If we applied exponential discounting, then this problem would disappear, whilst we would still value the present and immediate future events over the more distant future. But we do not – or at least we tend not to.

    Above are two cognitive errors that would lead to over-consumption of large sodas. They are well documented and well researched and widely accepted. From the list of “errors” there are few that would lead to under consumption. I spotted pessimism bias -which is specifically related to depression.

    Overall, given what we know about how people make choices, it seems overwhelmingly likely that large sodas will be over-consumed. I would welcome an analysis that contradicted this conclusion, as I could easily be victim of confirmation bias.

    However, even if we do establish that intervention will on balance improve the choices people make, that does not mean we should necessarily implement that policy. We must also evaluate the costs.

    I don’t know if the soda tax should be implemented. The costs of interfering in this micro way may be large, and the benefit small. However, I do argue that the tax has a sound rational and empirical basis in economics and psychology.

  77. 77 77 Tel

    Once we have established (as we have) that the decisions people make when selecting a large soda (or anything) are less than perfect, we must accept that in principle it is possible to improve on the choices individuals make.

    Presuming you are not a creationist, you might want to consider that evolution has been working on exactly that problem for umpteen million years, and devoted energy equivalent to all the sunlight shining on the Earth. How much time and energy have you devoted to your conclusions?

    when offered the choice between $50 now and $100 a year from now, many people will choose the immediate $50. However, given the choice between $50 in five years or $100 in six years almost everyone will choose $100 in six years, even though that is the same choice seen at five years’ greater distance.

    There’s nothing remotely irrational about this, and it is not at all the same choice at five years’ greater distance. A bird in the hand is worth two in the bush as the old saying goes.

  78. 78 78 Demosthenes

    I want to go on record as saying that I completely understand Harold’s (#11) utilitarian reasoning. Unfortunately, I don’t believe he has taken it far enough. Allow me to complete the chain.

    I believe that Harold, and those who argue like him, have an irrational belief that they are my fathers, or perhaps my keepers…or, in any case, that what I do with my life is of some concern to them, and that they ought to be able to protect me from the consequences of my decisions. This irrational belief has led, inevitably, to them acting irrationally by proposing that it should be the goal of public policy to distort consumer choices for what the distorters believe is the consumer’s ultimate good.

    On the theory that what is sauce for the goose is sauce for the gander, I would like to make a similar proposal: to wit, every time A proposes taxing or outlawing X because the results would be for B’s own good, A should be made to submit to the “tax” of allowing B to whack A for 30 seconds with a baseball bat.**

    It is obvious how, in the long run, the choice distortion introduced by this tax will increase overall social utility, by encouraging people to mind their own damn business.** In addition, I think that all the A’s of the world will, upon reflection, come to approve of this practice — since we have noticed that they are making the wrong choice, and are simply encouraging them to make the right one. Whaddya say, Harold?

    * Depending on locality, cricket bats are acceptable substitutes.

    ** To avoid the obvious logic loop, this proposal will be exempt.

  79. 79 79 Demosthenes

    Also, to Harold @ 76:

    “As the term ‘bad’ has no strict definition, as I explained, it is a perfectly good term to apply, given the weight of evidence.”

    If the term ‘bad’ has no strict definition, then neither does the term ‘good.’ Thus, your notion that the term ‘bad’ is a perfectly good term to apply under these circumstances…is in itself a bad notion.

    What scares me most is that you probably vote.

  80. 80 80 Harold

    Tel: Are you seriously saying that the fact we have evolved means that our choices about sodas must be optimum?
    “A bird in the hand is worth two in the bush as the old saying goes.” There is nothing wrong with discounting – as I said with exponential discounting we value the present more than the future, but it does not yield values that are inconsistent over time. Hyperbolic discounting results in dynamic inconsistency.

    Demosthenes: I am following market theory to suppose there is an optimum decision that results in an efficient outcome that is not arbitrary. Your suggestion of a baseball bat tax would not help.

    We have a wonderful theory of economics that explains how resources can be efficiently distributed based on individual choices. The power of this is awesome – as illustrated by essays such as “I Pencil”. This is somewhat like Newtons theory of gravity. Good enough to get us to the Moon and back and drop artillery shells with pinpoint accuracy 10 miles away. However, you need to refine the effect of gravity to account for air resistance etc. It seems to me that to define utility as merely what people choose is like ignoring air resistance and defining the target as whatever your shell happens to hit.

  81. 81 81 Ken B

    Harold: “Tel: Are you seriously saying that the fact we have evolved means that our choices about sodas must be optimum?”

    No, he is saying Bloomberg’s guesses won’t be better. Tel is pointing out that one’s assumptions about what is best for other people is subject to a great deal of uncertainty, and usually backed by very little. His jibe about evolution is to point out how little but prejudice backs most such attempts to meddle with the choices of others.

  82. 82 82 Demosthenes

    If you think I’m trying to define utility at all, then we are definitely speaking at cross-purposes. Allow me to correct that.

    I have no interest in maximizing utility…not, at least, as an end in itself. I know that the further one travels down that road, the more distant one gets from the protection of one man’s choice against the dictates of the mob (or whoever’s pulling their strings this decade). I know that we, as a society, cannot even agree on a full application of the concept of “utility” to our individual and collective circumstances, let alone come to decisions about how to “maximize” it in a way that would leave everybody satisfied. The best way to avoid these problems while still having a healthy utility is by decentralizing choice. Hayek…it’s all in Hayek.

    Meanwhile, I am greatly interested in protecting personal choice — especially against those would-be tyrants who, to paraphrase C.S. Lewis, are only concerned about my welfare — even though a multitude of individuals making choices will almost certainly lead to a system that is not as efficient as it could be. It’s not that I have a problem with an efficient system. It’s that people like you advocate for efficiency…and tend to see those choices of mine that they dislike as inefficiencies to be corrected…and attempt to correct them by putting their thumbs on my scale.

    I promise you, Harold, that a baseball bat tax like the one I propose would certainly help correct the inefficiencies your predilections create in the market of choice.

    “It seems to me that to define utility as merely what people choose is like ignoring air resistance and defining the target as whatever your shell happens to hit.”

    Making one’s aim to maximize utility — rather than to maximize freedom of choice and action — is akin to deciding the target, not just for yourself, but for everyone. And then appropriating their resources to hit it. The first power is the province of a god; the second is the domain of a thief. Neither is your proper function.

  83. 83 83 Harold

    Ken B: Who said anything about guesses? If we just went along with Bloomberg’s “feelings” then you have a point. That is why we do research. I have pointed out the results of a lot of research that would suggest large sodas would be over-consumed due to cognitive errors which are well studied and understood. I invited counter analysis, but all I get is “your guess is a good as mine”. Well, mine isn’t a guess. My assumptions about what is best for other people may often be “backed by very little”, but what if in some situations it is backed by a lot?

    Demosthenes. I assume you would say that even if it were backed by a lot, it makes no difference because “fredom of choice and action” is of prime importance. Well, if you are basing your policy on an arbitrary positioning of “freedom of choice” as a prime goal then there is no way to argue with you, nor any reason to agree with you. I had assumed naively that freedom of choice was a prerequisite towards the greater goal of efficiency, however defined. I fail to see how the aim of maximising utility appropriates resources of others, since this is the supposed outcome of free-market economics. Maximising utility maximises everyones personal choice of benefits, not mine. In the absence of market failures, then the free market achieves this. In the presence of marjet failures, then it fails to achieve this. I see the cognitive errrors I indicate above as market failures. If you disagree, then we can discuss this.

    A simple question: If there were such a thing as a clear and identifiable market failure, and a simple policy would correct it, should we impliment such a policy? In other words, are your objections in principle, or just in practice?

  84. 84 84 Ken B

    @Harold 83: I’d bet there’s more guess work than you think. Take a ban on large sodas. Most of the targeted people will buy more than one. What, other than costing them money, and using more paper cups, have you really achieved? In the case of a man who does drink less as a result, how do you know he values the minuscule improvement in health more than the foregone pleasure?

  85. 85 85 Harold

    Ken B: “how do you know he values the minuscule improvement in health more than the foregone pleasure?” Because of the research that shows that people are likely to make the sort of error that will over-value the soda and under-value the health costs.

    I have also said that even if there could be benefit shown, you would need to evaluate the costs, so the policy may not be worth implementing in any case.

    I do not know the details of this particular case – it may be based entirely on guesswork for all I know. Nevertheless, it seems that a good case *can* be made for such a policy producing benefits, even before we get on to healthcare costs for the rest of us.

  86. 86 86 Ken B

    @85:
    Certainly costs and benefits should be weighed. But by whom? Why should I weigh — guess — your preferences when I can let you do it. That’s what prices and markets are for.

  87. 87 87 Harold

    @86 – Prices and markets are great as long they are working. How can I stop this going round in circles? I suggest that a reasonable goal is one where everyones preferences are maximised (Pareto efficiency)
    Step 1. – in a perfect market everyones preferences are maximised.
    Step 2. where there is market failure everyones preferences are not maximised.
    Step 3. cognitive errors are a form of market failure
    Step 4. Intervention that completely corrected for the cognitive error would restore maximisation of everyoness preferences.

    So far it is theory. Next the practice:
    Step 5. We can identify cognitive errors that people are likely to make through research.
    Step 6. If we have identified such errors, there is scope to maximise everyones preferences by intervention.

    Now to the particular:
    Step 7: large sodas are an example where such errors have been identified, and intervention is possible through tax or ban on large sodas.

    Feel free to challenge any of these points, or poke holes in the way it is constructed.

  88. 88 88 Ken B

    Harold: For any individual you are guessing if he’s making the alleged cognitive error aren’t you?

    It’s one thing to restrict your choices for my good, or mine for yours. I support all kinds of stuff like that: quarantines, driver’s licenses, restrictions on storing nitroglycerine. It’s different trestricting your choices for what I think of is your benefit. Rarely am I a better judge than you are. Sometimes I might be I agree — you should listen to more Stravinsky and less Abba — but mostly I’m just guessing or imposing a prejudice. If there *is* evidence available it usually cuts against restrictionists: I already know the risks of tight rope walking (say) but enjoy it. You are conflating a reasonable prediction about certain costs and benefits — health, propriety — but are pretty much in the dark about others: pleasure, boredom.

  89. 89 89 Ken B
  90. 90 90 Eliezer

    “It’s important to care about making sense.”

    I think this sentence should say, it’s important to make sense. Who cares what anybody cares about?

  91. 91 91 Harold

    #88 Ken B. “For any individual you are guessing if he’s making the alleged cognitive error aren’t you?” Why does it matter for any individual, if I know that the policy positively affects more people than not? It is the overall good we are seeking, surely. Even markets do not promise that everyone is better off, only that overall the gains are greater than the losses.

    “If there *is* evidence available it usually cuts against restrictionists” You are just hand waving away the argument here. What if the evidence is not against the restrictionists?

    There are two (at least) possible idealogical approaches here. I say that IF government knew that by restricting certain activities then there would be more gains according to individuals preferences than by not restricting, then the Govt should implement the restriction. The other approach is to say that the Govt should not get involved, and people on average will be worse off.

    This is a “what if” thought experiment. It would be possible to agree with me that IF such a situation existed, then action should be taken, but that such a situation could not exist. Then we could talk evidence. Alternatively, you could disagree, and then we could talk about the purpose of government or something. AT the moment I am not clear about where we disagree.

  92. 92 92 Demosthenes

    @ 87:

    “Prices and markets are great as long they are working.”

    And who decides when they’re not?

    As to your argument, let’s say I accept every step. You still need a step #8 to make it go through in the way you wish…that we SHOULD do this. All the “is”-es in the world don’t imply an “ought.”

    An example: let us say that at some point in the future, after the appropriate technology has been invented, somebody makes your same argument about some “cognitive error” which I will not specify. Their proposed intervention is to financially incentivize the purchase and installation of a special brain implant that delivers a mild but painful electric shock to the implantee every time such an “error” is even contemplated. Assume no physical damage from the shock.

    So, we have here an “error” (of unspecified nature) that has been identified, and a proposed intervention enacted with help of a financial penalty. Further, I think it’s obvious that this intervention would be much more effective in solving its problem than a soda tax would be in disincentivizing soda purchases.

    My question is two-fold:

    1) Would you agree that such a step #7 could be sustained by the same steps #1-#6 that you have formulated here?
    2) If the answer is no, as I assume it probably would be, on what grounds would you be inclined to argue against my step #7?

  93. 93 93 Demosthenes

    To be clear: I recognize that there are several obvious differences between the soda tax case and the behavior modification chip case. My real question, more properly formulated, is “Which of those differences, if any, makes SUCH a difference that you would be inclined to reject the substitution of my step #7 for yours?”

  1. 1 Disposing of the Classical Liberal Tradition in Minutes
Comments are currently closed.