Cheap Donuts, Expensive Broccoli and Soda Taxes

Josh Wright —  17 June 2011

David Leonhardt’s NY Times column laments the fact that the price of healthful foods has increased relative to unhealthful foods.   He presents this useful chart:

The fact that relative prices have changed is often invoked in support of various “fat taxes,” e.g. consider the recent salvo of proposed soda taxes (For example, here and here).   Or consider the claim of Kelly Brownell (Yale) and Thomas Frieden  in the New England Journal of Medicine that “Sugar-sweetened [beverages] … may be the single largest driver of the obesity epidemic.”  Really?  That would be a pretty impressive claim if true.  But its not.  What do the data really say?  For starters, and on a more general level, consider the research by Gelbach, Klick and Strattman (which inspired the title of this post), which concludes that “relative price changes can only explain about 1 percent of the growth in BMI and the incidence of being overweight or obese over this period” and that  ” a 100 percent tax on unhealthful foods could reduce average BMI by about 1 percent, and the same tax could reduce the incidence of being overweight and the incidence of obesity by 2 percent and 1 percent respectively.”   It should also be pointed out that there are also some significant social benefits that flow from low prices of calorie dense food.

More recently, Klick and Helland have looked at the evidence on soda taxes and obesity more directly.   One of the problems here from an empirical perspective is that one must account for substitution toward other calorie dense beverages when the price of soda increases.  There are others.  Its a nice, accessible piece in Regulation on the various studies in this area: what they find, what data they use, their methodologies, and what the evidence actually says and what it cannot.  I encourage readers interested in this policy area to read it (as well as the underlying studies).  Here is how Klick and Helland conclude:

While politicians at all levels of government in the United States have been drawn to soda taxes as a way to both raise money and fight obesity, the evidence suggests that taxes may in fact do neither. Yes, individuals do seem to be price sensitive when it comes to soda and other sugar-sweetened beverages.  That implies, however, that any increase in tax rates will be offset largely by declining demand for soda specifically, but not for calorie-rich foods overall. While many public health advocates grab on to any indication of price sensitivity to support taxes as a way to reverse the upward trend of obesity, no study finds that this effect is very large in terms of the ultimate effect on body weight, as individuals substitute to consuming other calorie-dense beverages, adjust their eating habits in ways that have little net effect on BMI, or generally undo the positive effects of reduced soda consumption.

Keep that in mind the next time you hear a claim that a soda tax is likely to cause a significant reduction in obesity.


15 responses to Cheap Donuts, Expensive Broccoli and Soda Taxes


    1) Healthy foods will have higher (and relatively inelastic) demand from people who can afford them due to greater utility, and thus prices will naturally be higher
    2) Rice is over 1000 calories per dollar, even at the inflated prices paid in the US and other countries compared to international bulk rate ($0.50 per kg bulk, compared to $2 per kg in US stores)
    3) Poor people have long been known to prefer tasty food instead of healthy food, as a way to avoid feeling poor ( quoting George Orwell in Wigan Pier: “The ordinary human being would sooner starve than live on brown bread and raw carrots. And the peculiar evil is this, that the less money you have, the less inclined you feel to spend it on wholesome food.”)
    4) Reducing income inequality by eliminating unemployment would allow more people to be able to afford food that is both tasty and healthy.


    And how confident are we that bureaucrats will accurately identify healthful and junk foods? While soda may seem obvious, what of meat or fats?

    Nutrition science is not a settled field, and until then I’d be leery of intervening too much.


    I have a real concern about your methodology here. You are attempting to rebut a prestigious study in the NEJM, a top tier medical journal with countervailing evidence from a working paper that has not yet been peer reviewed apparently inteneded to be published in a social science journal.

    At best that seems very problematic. At worst, it fails utterly.

    I know of no cites in the peer-reviewed medical literature that rebuts effectively this NEJM study. If you know of one, I look forward to reading it.

    But need it be said that the peer review process for the NEJM is far more brutal than one that has never even occurred at a social science journal.

    In law, it would be asserting that my study that I INTEND to publish in Journal of Tofu Studies is somehow of equal weight than one written by Alex Kozinski in the Harvard Law Review.


      My methodology here? What — blogging about something I’m interested in? The fantastic thing about these studies, Matt, is that one can read the studies and the critiques in the articles cited. I’ve read the NEJM study; and I’ve read most of the papers in this literature. My own view, upon review of the available evidence, is that the evidence in favor of a link between soda consumption and obesity (and in turn, soda taxes to reduce obesity) is expressed above. So I’ve said as much here. Seems like a reasonable methodology to me for a blog — but to each their own. Anyway, go read the Klick & Helland survey cited above. If the explanations therein do not satisfy you, whereas placement in a peer reviewed journal would, then by all means wait for that.


      The NEJM study correlated obesity with consumption of sugar-containing beverages but made the unfortunately common mistake of finding a correlation and assuming a cause-and-effect relationship. The fact that the article was published in a peer-reviewed journal means little: peer review is an ineffective method for ensuring quality and accuracy.

      Other studies have shown that the increased prevalence of obesity in the USA correlates with dietary changes that substitute carbohydrate calories for fat calories. Yes, some of those increased carbohydrate calories are from sugary beverages, but more are from breads, starchy vegetables (potatoes, rice, corn, etc.), cereals, crackers, cookies, and candies. It’s politically popular to demonize soda manufacturers and fast-food restaurants as the major contributors to obesity, but the facts indicate that the food pyramid propaganda put out by our federal government is the biggest contributor.


    Government meddling is largely responsible for the obesity epidemic — advising us to eat a high-carbohydrate/low-fat diet. Per investigative science journalist Gary Taubes, it is carbohydrates that cause the insulin secretion that puts on fat. Eat a low-carb diet — no flour, sugar, juice, potatotes, and lots of meat, bacon, and buttered green beans — and you will likely be effortlessly thin. I am. And I have the health stats of an elite athlete, and with very little exercise.


      And let’s not forget subsidies. The government’s subsidized corn to the point that high calorie corn syrup is so cheap it’s in nearly everything. Cut those subsidies and watch as junk food prices normalize…


    Obesity is the number one health problem among our poorest persons. People cannot use SNAP (the modern-day equivalent of food stamps) funds to buy soda, candy, or cookies. So, the poor use their SNAP debit cards to buy sugar-sweetened juice drinks, sports beverages, energy bars, granola bars, crackers, sugarcoated dried fruits, and other foods that contain as much or more sugar than the banned items. This is strong evidence that junk food taxes will not reduce obesity but will only redirect price-sensitive buyers to other equally nonnutritious choices.

    If the federal government wants to reduce obesity, the very first action it should take is to discard the belief that a single diet is good for all. The second thing it should do is reverse its demonization of fats. The number one cause of the increased prevalence of obesity and adult-onset diabetes in the USA is the government-promulgated advice to substitute carbohydrates for fats. Fats should not be shunned because they are slower to digest than carbohydrates, they are better at suppressing appetite, they suppress appetite for longer periods, and they don’t contribute to blood sugar surges. Eating fats, in and of itself, does not lead to obesity or atherosclerosis. Evidence for this has been around for decades, but our federal government still emphasizes consumption of starches (grains and vegetables) and sugars (fruits and vegetables) and recommends minimal intake of fats.


    Let me get this right. Are you saying that when I go into 7-11 to quench my thirst after jogging and find that the price of sugary Pepsi is higher, I am likely to buy a donut instead? Or, are you saying that I am likely to buy a diet Pepsi plus a donut? And if so, wouldn’t that just suggest that the right approach is to tax sugary snacks or sugar generally? Or — Are you maybe suggesting that the belief in downward-sloping demand curves gets trumped by the demand for libertarian purity?


      Steve, you are taking blog post titles too seriously! This is leading you to focus too much on the donuts. The point was merely that the healthy food (broccoli) is getting more expensive than unhealthful food (donuts, or e.g. soda). Some have claimed this is a major determinant of increasing obesity rates and have suggested taxes. Or maybe you are hungry 🙂 Nothing I’m saying or have said denies downward sloping demand curves. Its about the cross-price elasticities of other beverages. I’m am also not sure what libertarian purity has to do with this. Yes, a broad across the board tax would eliminate these substitution effects — i.e. if we taxed all food and / or beverages it would have a much better shot at reducing calorie consumption. Hey, perhaps that is coming. Good luck. But real world policy experiments and state variation in prices. For example, Fletcher (2009) finds that children did not substitute toward diet sodas or water (as is generally assumed when folks make these obesity claims) but that the calorie reduction in soda was *MORE* than offset by an increase in milk consumption and in juice and juice drink consumption.


        I guess a blog post is not the place to discuss parental responsibility. Or, the fact that “juice drinks” are mostly sugar too. Does Fletcher show that children have downward-sloping demand curves?



Trackbacks and Pingbacks:

  1. Food law roundup - July 11, 2011

    […] “High-calorie food is too cheap” argument of NYT’s Leonhardt is open to doubt [Josh Wright] “Is obesity really contagious?” [Zoë Pollock, The Dish] Knives out among scientists […]

  2. Link: Cheap Donuts, Expensive Broccoli and Soda Taxes | Kartellblog. - June 19, 2011

    […] Cheap Donuts, Expensive Broccoli and Soda Taxes […]

  3. Cheap Donuts, Expensive Broccoli and Soda Taxes « Truth on the Market | | The Money BooksThe Money Books - June 17, 2011

    […] reading here: Cheap Donuts, Expensive Broccoli and Soda Taxes « Truth on the Market This entry was posted in Uncategorized and tagged consider-the, consider-the-claim, […]