Mayor Bloomberg and the Granny State, or: When is a Soda Ban Not a Ban?

Not Your Typical Grandma

Yes, that’s right: granny state, not — as conservatives are wont to call it — the nanny state.

Dan’s thoughtful post the other day suggested but did not spell out an important theoretical implication of New York City’s prohibition on large servings of sugared soft drinks: it represents an almost-classic form of the “nudge,” the policy tool advocated by Cass Sunstein and Richard Thaler.

As Dan noted, Mayor Bloomberg observed that the soft drink ordinance does not ban people from drinking more soda: it only prohibits them buying it all at once, in one serving.  From a standard economic perspective, this is a distinction with virtually no difference: if people want more soda, they will just go and get seconds.  But the assumption behind the policy is that people won’t do that.  Why?  Well, perhaps it’s because they are embarrassed to ask for seconds, but somehow I doubt that anyone really believes that fast food customers are so demure.  Rather, it is because they don’t really have that preference in the first place, or perhaps more precisely, their preferences are unstable: they are influenced by circumstances, biases, irrationalities, and just plain chance.

It is thus quite similar to the sort of classic nudge outlined by Sunstein and Thaler in their book, viz. where to place desserts in the high school cafeteria line.  Researchers found that placing them in the back at the end of the line yields very substantial drops in the amount of dessert consumed, and that should not happen if people’s preferences are stable.  Instead of doing nothing or banning them, “choice architects,” Sunstein and Thaler argue, should just put them where preference stability begins to break down.  Put another way, the soda “ban” isn’t a “ban” at all: it’s a nudge.  (It differs slightly from a classic nudge because most choice architects don’t have a choice: they have to institute some default rule.  Government doesn’t really have to do that here, but that isn’t apposite for the point I am making here).

That means that the New York City soda policy is not just relevant because, as Dan observes, “the food system (from farm to table) uses a lot of energy and produces significant water pollution,” and “[m]ore food equals a bigger environmental footprint.”  This is true, but it is also relevant because the policy serves as a real life experiment into the benefits and limitations of behavioral economics on a larger scale.

Oh, and Grandma?  That’s easy.  A “nanny” is someone who takes care of a small child, often a toddler or even an infant.  A “nanny” tells someone what they can and cannot do.  They control someone’s life.  Grandma is someone that the child asks for seconds.  And as any parent who has suffered through Grandma spoiling their children rotten, grandparents are usually anything but disciplinarians.  If there are rules, they are very gently applied.  Just like a nudge.

, , ,

Reader Comments

2 Replies to “Mayor Bloomberg and the Granny State, or: When is a Soda Ban Not a Ban?”

  1. I know you say it isn’t apposite, but the face of it, I can imagine many objecting that this ‘near-classic’ example of a nudge does not deserve comparison with the classic form. If the government is imposing a default rule that would otherwise would be set by private parties in a market transaction, it would seem to be libertarian paternalism without the libertarianism.

    That’s why I think it’s worth considering some other ways in which consumer choice here is inevitably structured by government regulation vel non. (Both of these, I believe, were also discussed in ‘Nudge’).

    The first is the phenomenon of anchoring. Consumers tend to ‘anchor’ off the choices presented to them, and to irrationally assume that a ‘good’ choice likely lies somewhere within the range of presented options. Consumers tend to assume that, if sugary sodas are presented to them in a range from 42- to 128-ounce portion sizes, one of those options is probably an appropriate selection. It requires a large epistemic leap to reach the conclusion that NONE of those options are healthy choices. This is true either because people assume regulation when there isn’t any (people say of campaign ads, “well, they wouldn’t let them say that on TV if it were totally untrue!”), or simply because we can’t make choices in a vacuum (if you ask poor history students to write down the last four digits of their phone number, and then guess the year of the Battle of Hastings, they anchor off of their phone number, for no rational reason).

    The second is the way in which we evaluate risk. People tend to fear an even remote risk of an acute and catastrophic danger much more than a surer but steadier death by a thousand cuts. If the New York Post ran a series of photos of rats in pizza parlors, people would not just expect but demand the Mayor shut down every last one. Very few would say that the answer to the problem is better disclosure of risk and individual responsibility. We have a societal consensus that that no one should ever have to run any risk of eating rat-za. But the chance of dying from a pest-borne food illness is almost nil compared the chance of dying, eventually, of heart disease or obesity-related illnesses. There are some rational reasons to prefer one regulation over another, but the fact that people respond instinctively to the two situations so differently shows that we apply the ‘discount rate of risk’ in a pretty unconsidered manner.

    This isn’t to say that the soda-size regulation, or any other intervention into people’s choice architecture, is necessarily justified. But it might be worth thinking at the outset about how we should structure our decision about how to talk about choice architecture.

  2. I know you say it isn’t apposite, but the face of it, I can imagine many objecting that this ‘near-classic’ example of a nudge does not deserve comparison with the classic form. If the government is imposing a default rule that would otherwise would be set by private parties in a market transaction, it would seem to be libertarian paternalism without the libertarianism.

    That’s why I think it’s worth considering some other ways in which consumer choice here is inevitably structured by government regulation vel non. (Both of these, I believe, were also discussed in ‘Nudge’).

    The first is the phenomenon of anchoring. Consumers tend to ‘anchor’ off the choices presented to them, and to irrationally assume that a ‘good’ choice likely lies somewhere within the range of presented options. Consumers tend to assume that, if sugary sodas are presented to them in a range from 42- to 128-ounce portion sizes, one of those options is probably an appropriate selection. It requires a large epistemic leap to reach the conclusion that NONE of those options are healthy choices. This is true either because people assume regulation when there isn’t any (people say of campaign ads, “well, they wouldn’t let them say that on TV if it were totally untrue!”), or simply because we can’t make choices in a vacuum (if you ask poor history students to write down the last four digits of their phone number, and then guess the year of the Battle of Hastings, they anchor off of their phone number, for no rational reason).

    The second is the way in which we evaluate risk. People tend to fear an even remote risk of an acute and catastrophic danger much more than a surer but steadier death by a thousand cuts. If the New York Post ran a series of photos of rats in pizza parlors, people would not just expect but demand the Mayor shut down every last one. Very few would say that the answer to the problem is better disclosure of risk and individual responsibility. We have a societal consensus that that no one should ever have to run any risk of eating rat-za. But the chance of dying from a pest-borne food illness is almost nil compared the chance of dying, eventually, of heart disease or obesity-related illnesses. There are some rational reasons to prefer one regulation over another, but the fact that people respond instinctively to the two situations so differently shows that we apply the ‘discount rate of risk’ in a pretty unconsidered manner.

    This isn’t to say that the soda-size regulation, or any other intervention into people’s choice architecture, is necessarily justified. But it might be worth thinking at the outset about how we should structure our decision about how to talk about choice architecture.

Comments are closed.

About Jonathan

Jonathan Zasloff

Jonathan Zasloff teaches Torts, Land Use, Environmental Law, Comparative Urban Planning Law, Legal History, and Public Policy Clinic – Land Use, the Environment and Loc…

READ more

POSTS BY Jonathan