Thursday, August 29, 2013
NYC’s Soda Ban: What’s All the Fizz About?
Reaction to New York’s 16-oz. soda limit has mostly come in two flavors. In the right-hand tap we mostly have complaints about paternalism and the nanny state. On the left, some grousing that (legal obstacles notwithstanding) a soda or sugar tax would have been a better policy. (Then lurking under the counter are the local government nerds, but let’s leave them be for now.) I’ll confess that the paternalism argument is too much for me to swallow. As lots of folks have now shown, there is a perfectly ordinary externality case for obesity control, regardless of whether the policies help us to better control ourselves.
One could say something similar about lots of modern nudges. Many of them -- smart meters, smart buildings -- are aimed at a classic externality problem, such as climate change. There isn’t really a paternalism story there. Maybe we could debate whether the nudge-y approaches are less “coercive” than your usual command and control regulation, but since no one has an especially good definition of coercive, I don’t think we’ll get anywhere. And indeed, I think that is what you’d see if you read the various back-and-forths on coercion between Sunstein et al. and their critics.
Maybe the best argument Sunstein and others (in particular, Sam Issacharoff and some smart economists) offer is that nudges are less “coercive” in the sense that they are more efficient (though they don’t typically put it that way). Usually, the nudge disproportionately affects people who need it the most---sticky pension defaults are most effective for procrastinators, and they’re the people who aren’t saving enough. So the “deadweight loss” of the nudge is small: it doesn’t bother people who don’t need it. But it’s not so clear how we tell this story about externalities. Are procrastinators more likely to emit greenhouse gasses?
This is a pretty narrow way of thinking about the efficiency of nudges. There is no secret formula for policy evaluation; we know how to mix up a good batch of regulation. Environmental economists compare taxes to “command and control” alternatives; crim. law scholars compare fines to prison and shaming. We can infuse this same analysis into the NYC debate, or analysis of any old nudge. Or, maybe not quite the same old classic analysis--maybe more of a New Coke flavor. I’ll say more about that tomorrow.
TrackBack URL for this entry:
Listed below are links to weblogs that reference NYC’s Soda Ban: What’s All the Fizz About?:
What are activities that do not have externalities?
What evidence is required to justify the effectiveness of a specific nudge? In other words, let's grant the externalities of obesity costs - what empirical test would the proposed soda size ban have to pass to justify its adoption?
Posted by: Mark | Aug 29, 2013 4:08:52 PM
Good questions, Mark. On your first one, no serious policy analyst suggests that externalities alone are a sufficient condition for regulation. Regulation costs money and distorts people's behavior. So even externality correction is generally only justified if the harm you can prevent can exceed those costs. This, by the way, is why slippery-slope arguments against externality correction are often tenuous (I'm thinking, e.g., of that ridiculous NYT column a couple of months back fretting that we'll soon be taxing annoying humming); the bottom of the slope is a straw man.
In any event, by any measure the fiscal externalities of obesity alone are many, many times the threshold I'm suggesting, running to six figures in many estimates. In contrast, smoking is a hazier case; Jon Gruber estimates the externality cost of smoking at around $1 per pack. But the internality cost he puts at more like $32 per pack, which is to say that, unlike soda, the anti-smoking case may be largely a paternalistic one.
On the second question...funny you should ask. Read the next post.
Posted by: BDG | Aug 29, 2013 9:46:08 PM
The problem with analyzing soda bans under the rubric of externalities imposed by obesity is that it looks only at inputs (and not outputs) and then only 1 type of input. In other words, a person who exercises imposes very different externalities (or none at all) through his soda drinking than the person who doesn't. And yet, we regulate soda drinking and not the lack of exercise.
Posted by: DrGrishka | Sep 2, 2013 8:36:43 PM