Junk Food Tax or Health Food Subsidy - Which Results in Healthier Food Purchases?

 Image by Jeff Keen.

In the past few years several prominent researchers have argued for the adoption of taxes on junk food as a means of reducing their consumption.  Often, as in a recent editorial in the New England Journal of Medicine, the argument is made that money collected through the tax could then be used to subsidize healthier foods.  This is an idea that I've found very appealing - we make the bad foods more expensive, the good foods less expensive, and people will probably shift at least some of their purchases to those healthier options.  But a very interesting new study by Leonard Epstein and colleagues suggests that things might not be so simple.

The paper starts with some very interesting background information on the cost of food over the past few decades.  For example, relative to other goods and services, current food prices are 38% lower than in 1978 (although the absolute cost of food has increased due to inflation).  And while overall the absolute cost of food has increased, this increase has been far greater in healthy foods than in unhealthy foods.  The graph below shows the increased cost for various foods since 1983.  As you can see, the cost of fresh fruit and veggies has increased by nearly 200% - 3 times greater than the increase seen in sugars and sweets, and roughly 6 times the increase seen in carbonated beverages (For a terrific exploration of why junk food can be so much cheaper than healthier alternatives, be sure to check out The Omnivore's Dilemma by Michael Pollan).

i-1f341d9eae311349e3eba1cec0717370-food prices.jpgBased on data published by Finkelstein et al., as cited by Epstein et al.

Epstein and colleagues point out that food prices have a strong influence on purchases, which makes it seem very reasonable that the relative changes in the cost of healthy and unhealthy foods over the past 25 years could be influencing food purchases, and therefore obesity rates. For this reason, changing the cost of food through the use of taxes and/or subsidies are obvious targets to curb caloric consumption at the societal level.  But they also point out that intentionally manipulating the price of food could have unintended consequences, especially with respect to subsidies for health foods.  For example, if health foods are subsidized, it is likely that people will buy more of them, which seems like a good thing.  But it is also very possible that people may use the money they save on subsidized health foods to buy even more junk food - an unintended consequence that I had never really considered.  Thus, the authors performed a small experiment to determine the effect of both fat taxes and health food subsidies on food purchasing behaviours.

Participants in the study included 42 mothers who were also the primary food shopper for their family.  The mothers were then placed in a laboratory fitted out to resemble a grocery store, and given $22.50 per family member and told to:

"imagine that she had no food in her house and that the money she was given was to be used to purchase groceries for her family for the week".

Participants were told to spend all of their money, and each participant went "shopping" 5 times - once with all foods priced accurately, twice with the cost of healthy foods lowered (by either 12.5% or 25%), and twice with the cost of unhealthy foods increased (again by 12.5% and 25%, respectively).  So, what happened?

As you might expect, as the cost of unhealthy foods was increased, the amount of total calories purchased was significantly reduced.  However, as the cost of healthy foods was lowered, the total number of calories purchased actually increased.  In other words, people were using the money they saved on healthy foods to purchase more unhealthy foods.  A health-food subsidy of 12.5% resulted in about an 800 calorie increase in total calories purchased, while a health-food subsidy of 25% resulted in an increase of about 1,500 calories.  So it seems that the health-food subsidy may not just increase the purchase of health foods - it may increase the purchase of all foods, regardless of their nutritional value.

Now there are obviously a lot of caveats to a study like this, and the authors are quite cautious in how they interpret their results.  For starters, participants were told that they had to spend all of their money during each trial, which makes it almost impossible for the health-food subsidy to result in anything but an increase in total food purchases.  So for that reason alone I'm pretty hesitant to take this study as evidence that subsidizing health foods is a bad idea.  But it is interesting, and I'm really curious to see if this finding is supported by studies looking at more "real world" settings.  

The paper is published in the journal Psychological Science and it's free to the public, so I'd really recommend you check it out.  Some of their graphs (which I couldn't re-publish here due to copyright issues) are especially worth the download.

So what do you think - are taxes and/or subsidies a good or bad idea?

Travis Saunders

ResearchBlogging.orgEpstein, L., Dearing, K., Roba, L., & Finkelstein, E. (2010). The Influence of Taxes and Subsidies on Energy Purchased in an Experimental Purchasing Study Psychological Science, 21 (3), 406-414 DOI: 10.1177/0956797610361446

Enjoyed this story? Subscribe to Obesity
and have future stories delivered regularly to your email
account or your RSS reader.

You can also follow
Travis on Twitter

More like this

why is this an either-or scenario? most of europe puts taxes on oil and subsidizes low-fuel cars

take the money you make through the JFT and support the subsidy ;)

What about the implications for programs like WIC that subsidize the purchases of healthy foods (milk, cereal, infant formula, etc.) for certain families? Perhaps the stereotype of welfare families buying chips and candy isn't that off base...

While interesting, I guess this is why economic studies always frustrate me. They have to spend all the money? Then how is this useful? So the participant buys all the healthy (and perishable) items for meals that week, then spends any left over on nonperishable (unhealthy) stuff. This is just out of my ass, but what % of healthy food lasts for a long time vs. unhealthy food?

By Rob Monkey (not verified) on 01 Apr 2010 #permalink

Rather than simply focusing subsidies on healthy foods and taxing the unhealthy foods, it would seem to make more sense to subsidize healthy lifestyles. Numerous studies have shown that incentives do get people to do things they might not otherwise do so subsidies are effective, but as this study shows, money can be directed in directions that are unintentional. If we instead subsidized people for keeping their bodies healthy within the limits of their abilities, the selection of healthy food would be part of achieving the goal as opposed to being a goal in of itself. Its hard to keep cholesterol levels and blood pressure down on a poor diet.

Unfortunately they didn't, but I imagine someone will soon, because it really is the most important scenario to consider. If anyone happens to know about such a study, please send it my way!

It's an unfortunate comment on my own cynicism that my first thought was: don't be ridiculous, the subsidy savings around here would be spent on smokes.

Practically speaking, comparing Canada and the UK (having lived in both countries), Canadian food is much more expensive, processed food doubly so. Anecdotally (I haven't seen studies, just observed friends and family), Canadians seem to be much more in the habit of cooking from scratch, much less likely to choose frozen/battered mystery fish and oven chips.

Makes me wonder if the way forward really is something to keep the price of processed stuff higher than "real food," and the subsidy scheme would appear to work as well as any other.

And the results when they sent all men to do the shopping? I would suspect the results would be a lot different.

And I'm not trying to be sexist either. When I'm doing the shopping, I notice that the carts which are loaded with junk food, frozen dinners and artificial fruit juices are being pushed by women. This may have nothing to do with saving money, but I suspect that since women tend to have more issues with food addictions, they transfer that problem to the family groceries.

'Women tend to have more issues with food addictions?' Men are more likely than woman to be obese. But, I will grant you that the scene in the movie where someone eats the whole pint of ice cream while crying over their body image usually features a woman, yes.

I imagine those women you see are shopping for families and don't have a lot of time/money. While men are more likely to be shopping for themselves or as a hobby. And yes, I AM trying to be sexist.

I'd submit that you can reproduce their graphs. It's just data. We had a bit of a test case with this on ScienceBlogs a few years ago and the net result is that you can reproduce data graphs as long as they don't have some special artistic value like Annie Liebowitz photos in their column fills.

"Men are more likely than woman to be obese." Actually, men are more likely to be OVERWEIGHT, but statistically, a larger percentage of women are OBESE. In the Black population women are more overweight and obese in all BMI stats vs. men.

People need to start eating REAL FOOD, that's not in a bag, box or plastic container. If it has a shelf-life, it doesn't belong in your body.

And that doesn't take into account that BMI isn't an accurate indicator of whether someone is healthy. It's a statistic that's been overused for ages. dAN, I agree with Tiffany. Most of the women I see shopping are obviously shopping for a family, whereas men are shopping only for themselves. Shopping for a family: more likely to be low on time, working while trying to cook, clean, raise the kids, take them to school, buy things that they will actually eat without kicking up a fuss. Shopping for only yourself: more likely to have plenty of time, and more likely to be obsessing over your body image so that you can find a mate (both genders), thus buying healthier food.

There are a lot more factors to people eating healthily than just telling people in a different class or racial group to yourself to "stop eating so much junk food". (This is aimed at dAN, not this blog as a whole, as I haven't read much on here yet).

By Katherine (not verified) on 06 Apr 2010 #permalink

Fascinating! THank you for bringing this to our collective attention. I had no idea this kind of reseach was even going on. It seems obvious, but I think its brilliant. I wonder what they will do with this information?


Very interesting study, but on a somewhat related topic, it seems to me that rather than tax unhealthy foods, it would be far better just to eliminate the existing subsidies on corn syrup and other unhealthy ingredients. The prices of junk food should then rise naturally, I would think.

Bit hard for me to understand the increased purchase of all food type calories if healthy food prices are reduced and unhealthy food prices are raised. Ingrained grocery purchasing patterns or junk food habits, or both?

Very insightful blog post.

I presume some of the cost rise in fresh produce reflects global market access? As recently as just 10 years ago, the seasonal differences in grocers produce variety and cost was marked. Now we see inflated prices for produce year-round, with the largest price increases occurring at the start of the present recession when crude and automotive fuel prices soared. Once grocers had us acclimated to the jump in fruit and vegetable costs, prices have never come back down again!

Recent study on soda pop tax suggested that minor tax increase per can would do very little to curtail consumption rates among hardcore consumers. The same would probably hold true for other snack foods.

At least some of the problem may lie in the convenience factor. Modern families tend to prepare very few meals from scratch due to perceived 'time and labor savings' massive oversupply of convenience foods. Lack of knowledge of the relative nutritional value of prepared versus freshly prepared foods is also a factor, correct?

I really like this blog.

Stairwell design in commercial and residential buildings is primarily a safety response as a nonmechanical exit path for occupants (mechanized lift failure) and entry passage for emergency response personnel.

I have worked in buildings that had stairwell access doors set to deploy fire alarms when opened. Majorly PO'ed management and firemen if you ignored the signs and mistakenly used the stairwell to speed your travels around the office.

In some government buildings, for security reasons after 9/11 each floor's access door was locked out from the inside of the stairwell, meaning one is headed on a one-way path down to the ground floor or second floor exits, which had doors that are locked out on the floor side to prevent unwanted street entry.

These stairwells are not suitable for squeezing in extra bit of exercise obviously. Thankfully, they appear to be in the minority of commercial building stairwells.

In several government agency and more than a few university research facilities I've worked in, we were thankful for ready access to stairwells because the lifts had chronic maintenance problems. These hazards frequently entrapped unlucky riders (resulting in adrenaline-rush yelling, pounding on doors and audible swearing) between floors or refused to stop at designated floors. One learned to read the symptoms by the number of times we heard emergency bells going off or prolong periods of inactivity. Eleven flights of stairs wasn't a major deterrent if you feared being marooned for hours in the lifts.

At one state university, the elevators would mysteriously respond to commands to move from one floor to another in the wee hours of the night in 5-story building, without visible occupant intervention...at least, not human. The building was chockablock with very large cockroaches not endemic to the US, most likely escapees from a long-ago research project.

My first thought is to give people the freedom to do what they want, and deal with the consequences themselves, such as not paying for any medical costs for self-imposed obesity or smoking. Maybe if people had to pay out of their own pockets they might take better care of themselves.

However, I really don't think that is practical and people will just hurt or kill themselves anyways and cause medical costs either way.

So... what to do then? I don't know the answer to the health care debate, it's pretty darn complicated.

Whoever claims that purchasing more calories is unhealthy, is not very educated. Unfortunately, this is a common presumption of the society at large, induced by the media. Purchasing calories is the point of food shopping. Calories are what fuel the body--you dont go shopping for food to NOT fuel the body. Being able to purchase more calories when health-foods are subsidized is proof that those subsidies are more efficient--families are better able to feed themselves when these foods are subsidized than when "junk" food is taxed. This country suffers from 15% food insecurity--a fact that the people all racing to fit some model of physical appearance seem to forget. Penalizing the poor and food-insecure by increasing prices does nothing to help them eat. Instead, the subsidies of organic, local, fresh foods work not only to provide them with more food, but with higher quality food.
By the way, what is called "obesity" is often a symptom of malnutrition--obesity rates are the highest among people with food insecurity: the poor, blacks, Southerners, etc. Further, there is no causational data to support the idea that high level of body fat cause the diseases often blamed on it--type II diabetes, heart disease, cancer, etc. If the presence of fat caused these diseases, liposuction would be a cure. All the data everyone is so familiar with is CORRELATIONAL (please, remember your science 101 classes). This makes perfect sense when you take into account that the people who have high rates of obesity are also the poor who have almost very limited access to health care. Further, the low-quality food that they are forced to eat contain horrible levels of persistant organic pollutants (POPs), heavy metals, and volatile organic compounds, that HAVE been shown to cause these diseases. Type II diabetes has been directly linked to POPs, for example--obese people with low blood POP levels simply do not get type II diabetes.
The 15% of people in this country that cannot afford enough food need calorie- and nutrient- rich food, not lean food, in order to have less disease. I hope this helps everyone separate out the notions of eating lean because of a desire to be thin for cultural/appearance reasons, and what truly eating healthfully is. Lean eating is NOT THE SAME as healthy eating, and being skinny isn't being healthy.

By Elle Reis (not verified) on 18 May 2010 #permalink

Just to be clear on my above post, I contend with the presumption that an increase in calories necessarily means that lower-quality food was purchased. Healthy food can be, and often is, high calorie. By your presumption, the healthiest food of all would be iceburg lettuce--which has no nutritional value whatsoever--no nutrients, calories, anything except non-digestible fiber. Your presumption would have had the mother buy almost no calories for her family, and labeled that as "healthy". No one could live off of iceburg lettuce.

By Elle Reis (not verified) on 18 May 2010 #permalink

Hi Elle,

We're going to have to agree to disagree on the role of body fat in disease. There are longitudinal studies showing that the accumulation of visceral fat is associated with increased risk of mortality (for example, this paper by our friend Dr Jen Kuk: http://www.nature.com/oby/journal/v14/n2/abs/oby200643a.html). And while liposuction doesn't improve health (for reasons we discuss here: http://scienceblogs.com/obesitypanacea/2010/03/how_do_fat_cells_protect…) the removal of visceral fat absolutely does improve health (details here: http://diabetes.diabetesjournals.org/content/48/1/94.abstract). We can't ever do a proper randomized controlled trial causing obesity in humans for obvious reasons, but we can in animals. And there are a wealth of studies showing that overfeeding rats results in both obesity and dysfunctional metabolism. None of these situations that demonstrate a strong link between body fat and disease/death can be explained by food insecurity, socio-economic status, POP's, or the like. It would be silly to discount the role of these factors in health, but it's just as silly to discount the role of obesity.


I did not say OVERFEEDING didn't cause obesity and metabolism disorders--sure it does.

What I said was not all "OBESITY" is a result of overfeeding (or for that matter, unhealthy eating of any kind, taking into account the shortcomings of the BMI chart and genetic set-point, but that's another discussion entirely).

I ABSOLUTELY agree that bad food choices lead to bad health. Obesity does not. And although obesity is VERY FREQUENTLY a SIGN of poor eating, it is not always. But when it is, it is merely a sign of poor habits, not a cause of disease.

What is called obesity can be a result of overeating, set-point, or of a host of other factors--any factor that causes a person's metabolism to shut down, often which are found in the poor and the food insecure: lack of frequent meals, lack of routine, disease, low-fiber diets, diets deficient in certain nutrients (which increases hunger), stress, eating in a rushed fashion, and the hormone disruption like that caused by many POPs.

Again, the presumption you are making is that all obesity is caused by overeating, and the answer is always to reduce calories.

I present you again with the fact that the highest rates of obesity are among the most food-insecure people in the country. To presume to restrict these people's diets out of some magnanimous arrogance, presumes they make their food choices out of ignorance or gluttony, rather than economic desperation--that presumption is insulting and colonialism. Not to mention incorrect.

You cannot presume to know a person's health by their weight. Despite recent language to the contrary, "obesity" is not a disease--at most it is an symptom, and there are times when it is not.

By Elle Reis (not verified) on 18 May 2010 #permalink

Hi Elle,

Again, we're going to have to agree to disagree. You said that "I ABSOLUTELY agree that bad food choices lead to bad health. Obesity does not." - the links I provided earlier report convincing evidence that the accumulation of body fat does, in fact, lead to bad health. I'll again refer you to the study showing that surgical removal of visceral fat results in improvements in insulin resistance. We've discussed a lot of the limitations of BMI here before, as well as the specific (and limited) situations where body fat is actually protective. But I find it difficult to read the papers that I linked to before and not conclude that the accumulation of body fat, in certain specific depots, is a very bad thing.

You jump to a lot of conclusions in this comment, and it seems pretty clear that we're not going to be able to have an amicable and evidence-based discussion. Suffice it to say that I feel that discussions of evidence-based public health policy DO NOT amount to colonialism, nor magnanimous arrogance. I think it's best we agree to disagree and move forward.


I'm perfectly able to have an amicable, and certainly fact-based, discussion.

I would again warn about interpreting correlational data for causational data.

What is deemed as obesity on the (we can all agree, on this, I think) very limited and fairly flawed BMI chart, does not necessarily mean a high ratio of visceral fat, in relation to subcutaneous fat. Much of the data on visceral fat is correlational, but from what I have seen, and I am sure I have not seen everything, conclusions have been drawn from many of these studies on the health effects of the RATIO of visceral fat to subcutaneous.

I am not proposing that I know all the ins and outs of health--rather, my point is primarily that the situation surrounding weight is FAR more complex than the media, and even educated people who seem to buy into such oversimplified stereotypes like sheep, would make it out to be.

Body types, weight, body fat, body chemistry, food, and health are far, far more complex (science has yet to understand quite a bit of it) than the mere less-calories-per-day-is-healthy paradigm that abounds in our society (a result of poor understanding of science aligning with superficial and unrealistic cultural ideals of body image).
What is healthy for one person isn't for another--there is no one-size-fits-all solution for food. We need to understand and accept diversity, and appreciate the scores of vastly complex factors that affect a person's body before we presume to know what exact weight they should be or food they should be denied the right to buy (presuming you believe that laws SHOULD deny people the right to buy certain foods, which you seem to, and I certainly don't though we need FAR better labeling).

Again, 15% of this country is food insecure--even more are nutritionally and food-safety insecure. The food insecure are poor--to presume that they merely choose bad food choices out of some inferior ability to understand or control themselves IS arrogant ethnocentrism, and in the fact that white people make these laws that would disproportionately affect black people--is racism. It is patronizing to presume to limit the ability of a certain group of people to make their own decisions.

Rather, subsidize local, organic food, and allow these people to buy more calories per dollar(the point of grocery shopping for the food insecure), but they will be buying HEALTHIER calories that are rich in protein and micronutrients.

You seem upset by the challenge I present to you, and I am sorry you feel this way. But I think you should be able to see that I am knowledgeable in this area, and this topic deserves discussion, in the larger research community, even if you are unwilling. By the way, I found your research to be interesting, aside from what I see as the grossly oversimplified presumption you based your analysis on, and I cited your work in my recent article.

By Elle Reis (not verified) on 20 May 2010 #permalink

It is indisputable that the obesity defined as an excessive of energy, represented by an increase of corporal fat, has adverse effects in health and longevity. The obesity is a pathology that has diverse presentations, depending on the country or region.
In chili one wants to reduce the obesity and to promote the healthy supply that also is pleasant.
One of the solutions is that the school education worries about this topic and insentive the physical activity.

The obesity is a chronic disease that brings sequels for the health. This blog delivers a lot of information. Important mas is that the persons worry about supporting saludablemente.
Do not eat scrap!!

To the delight of the fast food industry, quick food options have become engrained in the mind of the consumer as a choice between convenience and nutrition. For students and employed adults who don't have time to prepare meals, convenience inevitably wins out almost every time.

To the delight of the fast food industry, quick food options have become engrained in the mind of the consumer as a choice between convenience and nutrition. For students and employed adults who don't have time to prepare meals, convenience inevitably wins out almost every time.

There should be a label for PH Balance and oxidation level on all foods.

What do you all think?

I would be interested to see the results of a study where there is a smaller, less involved government that doesn't hand out any subsidies but lets businesses survive on their own.