Leigh Caldwell Petitions the State

Because I didn’t want to register with the UK government site on which Leigh Caldwell posted his ideas for behavioral analysis in the structure and deployment of services, I’ll comment here.

The rationale for Leigh’s wild and irresponsible proposals*:

While some behavioural interventions are being explored through the Cabinet Office’s Behavioural Insight Team (with some success) these tend to be relatively simple adjustments to framing of specific choices available to citizens. A deeper re-examination of the economic assumptions used in public service contracting and forecasting could lead to real improvements in outcomes and efficiency.

Some specific sectors that Leigh targets:

• Health care, where behavioral modelling of the consumption of health services may lead to more efficient deployment and use of resources.
• Education, where behavioral analysis could help bring incentives and signalling in line with cost savings in order to reduce spending while maintaining quality.
• Welfare and social security, where structuring incentives could help raise people out of poverty by building productivity, and encouraging formation of savings. Thus, reducing dependence on the state in the long run.

Now, I’m only a smidgen a behavior economist (having read varied works from the Santa Fe institute), but I’ve always been at least cautiously optimistic about the prospect of what Richard Thaler refers to as “libertarian paternalism“. Leigh is a crazy lefty*, but I trust that he believes in choice, and understands that choice is often not the problem in and of itself. Choice sets often are given various cognitive biases. Thus, choice architecture can preserve the ability of the individual to make a choice, but incentivize choices that are in the best interest of the decision maker.

I’m not too familiar with the first two categories Leigh lists, but I am broadly familiar with the third (which is the most “popular”). There are several ways in which the government could better structure incentives to produce superior long-run results, but I want to focus on one real-world example. The Oportunidades program, an the anti-poverty program in Mexico. The program is centered around providing cash transfers that are linked to incentive goals:

Oportunidades is the principal anti-poverty program of the Mexican government. (The original name of the program was Progresa; the name was changed in 2002.) Oportunidades focuses on helping poor families in rural and urban communities invest in human capital—improving the education, health, and nutrition of their children—leading to the long-term improvement of their economic future and the consequent reduction of poverty in Mexico. By providing cash transfers to households (linked to regular school attendance and health clinic visits), the program also fulfills the aim of alleviating current poverty.

The Oportunidades program has by many measures been very successful in reducing extreme poverty in Mexico. In the long run, these types of behavioral-influenced programs can lead to considerable long-run gain in productivity, health, and personal finance.

I think that the British government would do well to invest in research of this type. While I doubt that behavioral economics will revolutionize the field of economics as a whole (at least until highly useful simulations are commonplace, right now we have variants of sugarscape and the game of life/prisoner’s dilemma), behavioral analysis can be extremely useful at the margin.

As an aside, I am curious whether this web interaction between the government and private citizens/businesses is a Conservative thing, or just something the British government does?

---

*Just kidding. I consider Leigh a good friend.

Is behavioral economics neoclassical economics in disguise?

A new paper by Nathan Berg and Gerd Gigerenzer asks this question:

For a research program that counts improved empirical realism among its primary goals, it is surprising that behavioral economics appears indistinguishable from neoclassical economics in its reliance on “as-if” arguments. “As-if” arguments are frequently put forward in behavioral economics to justify “psychological” models that add new parameters to fit decision outcome data rather than specifying more realistic or empirically supported psychological processes that genuinely explain these data. Another striking similarity is that both behavioral and neoclassical research programs refer to a common set of axiomatic norms without subjecting them to empirical investigation. Notably missing is investigation of whether people who deviate from axiomatic rationality face economically significant losses. Despite producing prolific documentation of deviations from neoclassical norms, behavioral economics has produced almost no evidence that deviations are correlated with lower earnings, lower happiness, impaired health, inaccurate beliefs, or shorter lives.

My guess is that this critique will bring the behavioralists and neoclassical economists together in joint attack. The authors propose a new approach they call ecological economics, and summarize what the field should do improve like this:

“To make behavioral economics, or psychology and economics, a more rigorously empirical science will require less effort spent extending as-if utility theory to account for biases and deviations, and substantially more careful observation of successful decision makers in their respective domains.”

The Future of Paternalism

I’ve recently challenged paternalism supporters to tell me, if the most recent headline paternalism that bans Happy Meals in San Francisco does not constitute taking us down the slippery slope, then what would? Matt Steinglass at the Economist has responded to this challenge by accepting it’s legitimacy, and reversing it, but not answering it. Granted, speculating what people will decide the government should crack down on next is a tricky game. I mean who would have guessed that Happy Meals were on the chopping block? But from the logic of this ban, we can see a wide range of policies the future might hold.

First, lets look at the logic of this policy. Here is how Steinglass makes the case:

Now, not every parent objects to their children eating unhealthy fast food. But I do, and there are a lot of other parents like me, especially in places like San Francisco. For such parents, the Happy Meal represents an effort by some adults to profitably exploit and exacerbate the tensions in other adults’ parent-child relations over food.

So what are other policies by which some adults profitably exploit and exacerbate the tensions in other adults’ parent-child relations over food?

First note that we don’t need a one-to-one relationship between that which is attracting the children in the first place -in this case a toy- and something that is unhealthy -in this case unhealthy food. Matt can actually get a Happy Meal with chicken McNuggets, 1% milk, and slices of apple with low-far caramel dip. In fact, at 390 calories, 32% of which come from fat, McDonalds will be able to sell this food with the toy after the ban, which only prevents Happy Meals with more than 600 calories and 35% of the calories from fat. So it’s not like the only way for Matt to get this toy for his kids is to buy them the “cheap Jumbo McFattyburger”. McDonald’s has given them a relatively healthy option that allows parents to cave into their children on the toy but not on lunch. What more do you want parents? Must we hold your hand the whole way through raising your children, cordoning off everything you don’t have the will to tell them they can’t have?

Given the nature of the justifications for this policy -protecting parents from having to tell their kids no- what would be logical extensions of it?

Well without leaving McDonalds there’s some obvious ones. For one thing, the playland probably draws kids as often as the toys do. So new McDonalds could be banned from being built with playlands. Along the same lines,  fast food restaurants could be banned from being near places where children play in the first place, like parks and schools. I know this one has been called for before, and I’m not sure, but it may even be in affect in some places.

Making fast food less attractive may protect parents when they happen to be near a McDonalds with their kids, but it doesn’t protect them from having McDonalds reach out to children in the first place and getting it into their heads that their food and toys are awesome. If you’re going to stop this problem, it must be at the root. One way to do this is to ban advertising of fast food targeted at children. This would probably start with children specific magazines and TV shows, but move to a general ban.

However, these policies only target fast food, and we know that kids beg for unhealthy food all over the place, not just there. So let’s look at the grocery store.

One obvious example is banning candy from checkout lines. Sure many stores have aisles that are candy free for just this purpose, but McDonalds has a healthy lunch option as well.

Another thing that could be done is to counteract the behavioral economics and marketing used by grocery stores and food companies to target children. Unhealthy food that children like, for instance sugary cereals, could be required to be placed on top shelves where it’s harder for children to see and reach for them. More extreme than this would be to ban any cartoon characters from unhealthy food containers. I’m pretty sure many people would take a bullet for Cap’n Crunch and those elves from Rice Crispies, so the bar for unhealthy would be set pretty high… at first.

Eventually there could be a regulatory agency that has to approve all foods to ensure that no behavioral economics or marketing wizardry draws children to it whatsoever.

In addition to how stuff is sold, anything that is enumerated on a nutrition label could be regulated. We know that portion size is a problem, so maximum portion sizes could be set. Also you might also see maximum calories for single serving foods like hot pockets and candy bars.

In order to prevent an outcry from adults there may be a two-track regulation where food is categorized either as “mostly for adults” or “mostly for children”. Foods that are “mostly for children” will be regulated more strictly and on many dimensions, while “mostly for adults” remain freer … at first. Eventually all foods could fall to the more draconian regulations as two-track regulation fails because parents will just buy unhealthy “mostly for adults” food for their children.

If you think setting the maximum number of calories and other nutritional measures for every kind of food seems absurd then you should read the Institute of Medicine report on salt regulations, which pretty much suggests doing this this for sodium content.

Many of these suggestions probably seem so extreme as to be unbelievable. But they seem extreme from the vantage point of where we are on the slippery slope. The farther down we go and the more accustomed we become to these sorts of things then the less and less radical the currently extreme examples will seem. I’m not predicting all of these will happen everywhere, but some of these, or things like these, will be pushed for, and some of them will be put into place.

My question to paternalism supporters is which of these things would be ok with you? And by what basis do you reject them but approve the Happy Meal regulations? To those on the fence about paternalism, or ok with the happy meal ban but wary of more strict regulations, pay attention to the answer your paternalism allies give to these questions; as you may someday find yourself on the other end of this argument.

Behavioral Economics in the Cafeteria

There is a very interesting article at the Atlantic from David Just and Brian Wansink from the Cornell Center for Behavioral Economics in Child Nutrition Programs. They discuss their work on improving healthy eating in school cafeterias using the subtle wizardry of behavioral economics. There have been some impressive results:

One school in upstate New York was able to increase consumption of salads by close to 300 percent by simply moving their salad bar six feet from the wall and placing it near a natural bottleneck in the check-out line. Another school increased fruit sales by 105 percent by moving the apples and oranges from stainless steel bins into a well-lit and attractive basket.

It is encouraging to see behavioral economics being put into creative use like this. The authors argue that “It is difficult to teach a high school student how to make healthy choices in the real world if only escarole and tofu on are on the school lunch menu”. But is this teaching them to choose better or tricking them into choosing better? After all, if behavior is so amenable to subtle tricks like this then what hope can there be that any behavioral changes will actually last? I can’t tell if articles like this should make us more or less hopeful. Yes, the good scientists here are making a difference, but are we really so impressionable?