When Are Nudges Effective?

All hail the nudge?  This NY Times article discusses the challenge of bringing about behavioral change in the Mississippi Delta so that people eat more healthy foods and fewer fried foods.  When people ignore information that will “help them”, how do you explain that?  Do you say that they are impatient? or do you blame their region’s culture?   If we aspire to be good statisticians, how would you rigorously test your hypotheses?   

If we paid these individuals to eat healthy food, would this be effective as households would learn about tasting new foods and might feel the short run health benefits.  

Harvard’s Roland Fryer is running field experiments in which he pays kids for school achievement and for effort.  I interpret his efforts as trying to “change the culture”.  If studying becomes “cool”  in inner-city schools (because you can make some $ doing it), then kids will devote more effort and school performance will improve.

Author: Matthew E. Kahn

Professor of Economics at UCLA.

7 thoughts on “When Are Nudges Effective?”

  1. “discusses the challenge of bringing about behavioral change in the Mississippi Delta so that people eat more healthy foods and fewer fried foods.”

    Hmm. And what exactly is the proof that it is fried foods that are the problem? My understanding is that
    – the claim that fat is problematic is based on essentially nothing. It is a claim analogous to using yellow plants to treat jaundice — fat looks like artherosclerotic plaque, so must be the cause of it. The epidemiological studies that supposedly support this claim are weak and contradictory, as is usual for epidemiological studies, while there are no physiological studies that I know of.

    – there are ample reasons to assume that, in a wide variety of circumstances, in rich countries and poor, the primary cause of obesity is eating too many carbohydrates (sugar, yes, but not only sugar); in this US this particularly occurred post1980 as the “fat causes heart attacks” propaganda caused people to replace a substantial fraction of their fat calories with carbohydrates.

    – more generally, nutrition “science” has not covered itself in glory over the last 40 years, repeatedly trumpeting as vitally important some minuscule effect — which is then revealed ten years later to, yes, be unimportant. The most recent findings regarding salt are just one more example of this sort of thing.

    In other words, nudges regarding food are about the worst place to look because any rational individual knows to ignore 95% of what they are told on the subject. Far more worthwhile would be to concentrate on issues where the appropriate behavior and its payoff are actually agreed upon and rigorously understood. And these are not hard to find — vaccinate kids, issues around substance abuse, don’t get pregnant as a teenager (or as an adult with no money), use energy efficient devices, save money, both for retirement and to have a safety cushion, etc etc.

    As for Roland Fryer, I guess, as a good economist, he already knows the answer he wants, so is completely uninterested in the copious literature on the noxious effects of replacing older types of social contracts (including studying, being a good neighbor, or childcare) with purely monetary incentives…

  2. If I were a billionaire philanthropist interested in nudging kids to learn, I would open a chain of “Quiz Parlors” or “Exam Arcades” across the country, to which kids could come in their spare time and take tests for money. There would be no pass/fail criteria; just so many dollars per correct answer. I would not care whether the kids did their learning at home, in school, at the library, or on the internet. “Pay for performance” would be my motto. I would not care whether the kids went right out and spent their money on bling or saved it for college tuition. If I wanted to “nudge” kids by offering them college scholarships in the (to them) far-distant future, I’d do that instead. But I believe immediate, tangible rewards are a more effective “nudge”.

    There are obvious logistical problems with such an approach to education philanthropy. But there is one major advantage: I, the billionaire philanthropist, would get to decide what questions to ask and what the right answers to them are. For instance, religious parents who objected to their little darlings learning about evolution in biology class would not be MY problem. I can imagine somebody pointing out that kids could learn “evolution” well enough to make money answering my questions without believing in it; my response is that if a creationist kid can correctly answer properly structured biology questions, then I don’t care whether he “believes in” evolution any more than I care whether he “believes in” trigonometry.

    Would teachers start “teaching to the test”? I don’t know. Maybe their students would demand they do so. Of course, there would be no such thing as “the” test. Remember: nobody “fails”, nobody “passes”; it’s not a “test”. Would it be effective philanthropy? I don’t know that either. Philanthropy it certainly would be, especially since I would be free to pay more for correct answers in my inner-city storefront Quiz Parlors than in any suburban mall Exam Arcades I cared to open.

    Alas, I’m still working on making my first billion, and I suspect my project would cost a bit more than that.


  3. Tony, teachers wouldn’t have to *start* teaching to the test, they already do. There’s lots of evidence that this sort of teaching isn’t particularly successful, and that’s because many of the most important things to be taught are not really testable. For one small example, public speaking (aka show-and-tell in the younger years). What’s the right answer to that? Not sure either how your plan would work for special ed kids (who are usually about 10-15% of any school’s population). But your heart’s in the right place.

  4. The standard hand-flapping about eating and obesity don’t help Matt’s implicit assertion about rationality and obesity. Just saying.

  5. Ohio Mom,

    You are absolutely right that my “pay for performance” approach would not directly benefit special-needs kids. All I can say to that is, it could not do worse than a more conventional project of promising them college scholarships if they get good grades.

    And you are also right that such skills as public speaking would be awfully hard to test for. On the other hand, a talent for speaking in public without knowing what you’re talking about is not really something I want to foster in the next generation. We have too much of that in ours.

    As for “teaching to the test”, what can I say? We have to teach “to” something. About the only thing George Will ever wrote that I agree with is that education consists of putting something into children’s heads, not letting something out. Perhaps the most important thing we can put into children’s heads is the idea that knowledge is valuable. All I suggest is that a bit of cash money can go a long way toward convincing them of that.


  6. There’s actually pretty good reason for doubting that obesity is due to simply eating too many calories compared to your level of exertion. A person gaining weight will typically do so very gradually indeed, at a rate which corresponds to an excess of maybe 50 calories a day. A reliable excess of fifty calories a day, mind you. I deny anyone to consciously control their energy balance that closely, short of living in a laboratory.

    It appears, rather, to be the case that something is causing body weight set points to gradually creep up, with the body very accurately tracking the changing set point, and causing you to modulate your activity levels and food intake to conform to it. There are a number of theories as to why this is happening, but it’s even seen in laboratory animals.

Comments are closed.