A new paper shows that food prices affect obesity – but the impact is so small that even huge taxes won’t have more than a tiny impact on BMI.
Cheap Donuts and Expensive Broccoli: The Effect of Relative Prices on Obesity
Jonah Gelbach, Jonathan Klick & Thomas Stratmann
Americans have been getting fatter since at least the mid 1980s. To better understand this public health problem, much attention has been devoted to determining the underlying cause of increasing body weights in the U.S. We examine the role of relative food prices in determining an individual’s body mass index, arguing that as healthful foods become more expensive relative to unhealthful foods, individuals substitute to a less healthful diet. Using data from the National Health Interview Survey (NHIS) for the period 1982-1996, we find that individual BMI measures, as well as the likelihood of being overweight or obese, exhibit a statistically significant positive correlation with the prices of healthful relative to unhealthful foods. These results are robust to endogenizing the relative price measure. While the magnitudes of our estimates suggest 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, they do provide some measure of how effective fat taxes would be in controlling the obesity epidemic. Our estimates imply, for example, 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.
Andrew Leigh's Archived Blog 2004-2010 
Thanks Andrew – that’s another interesting paper I wouldn’t have discovered on my own.
I’ve got a quibble with theory behind the study. The authors assume that “unhealthful food is a substitute for healthful food” (and vice versa). If the price of unhealthy food rose relative to healthy food then price sensitive consumers would shift away from unhealthy food.
But I’m not sure whether healthy food really is a substitute for unhealthy food — at least not in all circumstances. Do people really buy fries and donuts just because they’re hungry?
It may be that people consume unhealthy food — particularly fast food, sweets, snacks and treats — simply because it feels good to consume them. Even when you’re not hungry, eating fries can be pleasant. It can be a way of compensating for boredom, frustration or stress.
If this was true then the substitute for fries or donuts would not be carrot sticks or broccoli. It could be television, cigarettes, alcohol, pornography, video games — anything that occupies the mind (and or body) and feels good.
Perhaps the differences between obesity levels in more and less educated individuals have something to do with the relative levels of stress and frustration (which would affect demand) or the range of options available for dealing with stress and frustration and their relative prices.
As readers may be aware I am involved in a system called Water Rewards (which we expect to launch with Actew in Canberra in about 3 months). We have designed another system based on the same idea called Health Rewards and our initial targets are MacDonalds and one of the major super market chains.
Our modelling for Water indicates that we will get at least twice the effect for a given cost to the consumer in terms of a price increase using Rewards as opposed to a system based on price increases – such as Water Trading – to get a desired outcome. With Water Rewards it is a reduction in household use of water by some percentage for a given increase in price.
With Health Rewards our target is a reduction in BMI. Judging by the small effect of price increases on BMI reported it will not be too hard to get at least a doubling of effect of a price increase on unhealthy foods.
With Health Rewards the proposition is to give people a Reward when they purchase any MacDonalds meals or any food item in say Coles. The trick is that they will only be able to use their points on “healthy” things – like school sports, gym classes, or MacDonalds healthy meals or Coles Fresh Fruit or Vegies. The proposition to MacDonalds is that they can use it for promotional purposes and they can get tie ins with schools etc – and get sales of their “better” meals. A critical psychological point about Rewards is that they are not discounts.
We believe that when we are dealing with changing behaviour with amounts that are relatively insignificant then price alone is not enough. The BMI study indicates this.
If there is any economist or budding economist who would like to join us and look at the data we will get from our Water Rewards pilot then please contact me. One of the tasks I am going through at the moment is trying to design Water Rewards in such a way that we can measure the effect and prove one way or the other if it is as good as we expect.
Kevin – but what if people just get a free Reward meal along with the Big Mac they were going to buy in the first place?
Purely anecdotal evidence, but I have a (tall and thin) Canadian colleague who says that he is one of four brothers, two of whom live in Europe (including him) and the other two live in North America. The two who live in North America weigh 30 kg more than the two who live in Europe. So his view is that they have the same upbringing and levels of education, and very similar genetics (but they are not sets of twins), so that lifestyle is the explanation – use of cars vs walking and diet choices. I haven’t asked him whether he thinks his two North American brothers are more stressed and anxious.
backroom girl -a healthy meal and a big mac is better than 2 bigmacs?
Pingback: Tim Worstall
I’m not up to reading 53 pages now so I shouldn’t comment, but ……
Are the authors sure they have isolated all other factors – one I’m thinking of is sloth. I could have, this evening, got out vegetables (washing, peeling, cutting, cooking) and roast some potatoes or whatever, but instead I had biscuits, chocolate and licorice: a 3 course meal!
Healthy food is more work and involves more planning, and skills/knowledge than unhealthy ‘food’ – which is usually just in a packet, or the freezer. I’m not sure it’s that much cheaper – I think people are going for convenience (as well as taste, as Don said), as much as price.
Andrew here is a story called a Fat Story which you may notice has borrowed heavily from “A Milk Story”:)
Once upon a time there a country who thought that having a well fed population was a sign of prosperity and a well run economy. The Prime Minister was heard to boast on many occasions on the success of his government’s policies particularly as it related to the consumption of fat foods. The more fat consumed the greater the measure of prosperity that his economists called the GFP (Gross Fat Product). His treasurer was particularly proud because he collected taxes on the fat production and the more taxes the greater the budget surplus.
The country started to perform spectacularly as a highly populated country to its north found that its people also loved fat and could not buy enough of it. They started to import fat in such large quantities that the world wide prices increased and the budget surplus of the lucky country increased even more. This delighted the PM because he knew he would be able to spend much of the surplus advertising his achievements before the next election.
There were a few people who raised warning bells about the effect of the consumption of fat and said that if too much was consumed then the population would start to die prematurely because of overheating. However, the PM remained a fat sceptic.
Not withstanding the PM’s sceptism a world wide movement of people arose who believed that the effects of too much fat were bad and they became known as the Lean Movement or as some called them the Leanies. They invented a way of measuring the bad effect of fat and they called it the BMI for individuals. To measure the effect globally they came up with the term Global Accretions which was derived from the aggregation of all the BMI’s. They were able to convince enough countries that the global community needed to limit Global Accretions and they even managed to get most countries to agree to what become known as the Fato Protocol so named because it was signed in Fato.
The protocol required all countries to limit their BMI’s so that the total Global Accretions were limited. There were exceptions and even though an exception was made for the PM’s country he would not sign it because his best mate would not sign it either and as he said one in all in. The exception was interesting because it allowed the country to count the BMI’s of all animals – not just humans. Even though the people grew fatter and the protocol was not signed the PM was able to boast that they were meeting their Fato targets because his farmers were unable to grow sheep because of a drought and the country was able to support more kangaroos which were leaner than sheep.
All went well until suddenly the population realised that Global Accretions were starting to cause consequences that if left unchecked would cause most of the people in the world to overheat due to the layers of fat.
The PM remained a skeptic but the population become convinced that there was a problem and so the PM became a Fat Realist and finally a Fat Pragmatist.
To solve the problem he formed a task force of all the fattest people in the country and asked them to come up with a scheme to solve the problem. They devised a scheme called Fat Trading. A cap on the amount of fat to be consumed was set and people were given fat permits which permitted them to eat fat. That is, unless you had a Fat permit you were not allowed to eat fat. However people could trade Fat permits and if someone did not want eat fat they could trade their permits to someone who did want to eat fat. If you ran out of permits you could always eat other things. The PM was told by his economists that this must work because of market forces and all the PM had to do was to sit back and let the market do its magic. The Treasurer was also very pleased because he knew that if he ever needed some more money for advertising he would just sell a few more permits for “special reasons”.
The PM duly approved the scheme but he made sure that all the fattest people got free permits. He also made sure that that the total cap was set so that the price of permits would not rise too high. The economists pointed out these problems but overall they were pleased because they had a trading scheme in place. They believed that when the price of permits rose people would change their behaviour and substitute other fat free foods. They knew that it was important to have a market as markets always solved every problem.
Unfortunately it did not work very well. The price rose 100% but this only decreased the overall consumption of fat by 1%. The people with fat permits also became richer as they were able to sell some of their permits and they were able to offset their permits by grazing kangaroos in their back yards. National Accretions rose as on average people continued to get fatter because the fat people put on fat at even greater rate than the Leanies who stopped eating any fat at all. Accretions continued to go up even though the cap continued to drop and many people became both rich and fat through trading permits.
The economists were happy because they saw the market at work as predicted. Fat foods became expensive and Lean foods became cheap and the market had worked its magic. Pity about National Accretions.
You might like to read an article we have published this morning in On Line Opinion: You are what you grow by Michael Pollan http://www.onlineopinion.com.au/view.asp?article=5923. It ties in nicely with this topic.
Susan, thanks. The Gelbach paper utterly demolishes the Pollan thesis, doesn’t it?
Russell/Peter, the best analysis I’ve seen on the contribution of exercise vs eating to the rise in US obesity is Cutler, Glaeser & Shapiro (paper here, my oped summary here). They find that exercise might have explained the rise in BMI prior to 1970, but after that, it’s all changing consumption patterns.
Kevin, I enjoyed your story. The point is well taken – economists often don’t pay enough attention to getting the initial distribution of tradeable permits right.
Kevin asked whether “a healthy meal and a big mac is better than 2 bigmacs?”
Clearly that would be the case. I’m just not sure how with your reward system would ensure that people were substituting healthy for unhealthy calories rather than just adding them on to what they would have eaten anyway.
It’s the perennial problem with policy – the unintended consequence. It may not take too large a proportion of people doing the unintended thing to cancel out the benefits of those who do the intended thing.
backroom girl you are exactly right. We have little idea what will happen when we build a system. I once built a system to do an innovative invoicing system for a small business. I went back a year after it had been operating and asked the business man the most important consequence of the automation – expecting to hear about increased profits, more sales etc. He said that his golf handicap had dropped as he did not have to work on Wednesday afternoon.
It is my belief that policy makers should build their systems so that the desired policy outcomes can be measured and the system easily adjusted to handle the inevitable emergent properties – both good and undesirable.
You may be interested in knowing that the way to build and program modern information systems is to first write down how we are going to test it once it is built. Once you decide how to test it then you can build it. I suspect that if policy people did the same then they would come up with more robust policies.
With the big macs we would try to get all rewards participants to give us their BMI’s and we would monitor what they did with their Rewards and see if the BMI’s changed.