VoxEU Column Health Economics

Obesity and price sensitivity at the supermarket

Is increasing obesity due to changes in relative food prices? High-energy density foods are less expensive per calorie than fresh fruits and vegetables. Using data from Israel, this column shows that price sensitivity has a significant impact on obesity. In fact, price sensitivity may be more crucial than income.

As I make my way down the aisles of my favourite vegetable stand (Whole Foods quality in an over-crowded market with narrow aisles), I am struck by two competing thoughts. My first reaction is “wow, the vegetables really look fantastic – indeed I can envision the great salad I plan to whip up tonight.” But, then a second, less happy thought enters my consciousness – the fruit and vegetable prices are really high.

The so-called Mediterranean diet in Israel is based in part on these wonderful vegetables. Great fresh salads are such a part of the culture here that – believe it or not – McDonalds has one of the freshest salad offerings around. They do not cut up veggies until you place your order, so their salads are fresher than some well-known salad bars. Of course, as my son confirms, McDonalds offers its more familiar fares as well.

Despite the great vegetables, Israelis are eating more processed food, and obesity is a serious and growing problem in the country. Shockingly, obesity levels are not far from the very high US levels. And it is not just the US and Israel. Over the last three decades, the number of people suffering from obesity has tripled.

The obesity epidemic

According to the World Health Organisation, there are over one billion overweight adults, 300 million of whom are obese. Additionally, childhood obesity is a very serious problem and has reached “epidemic” levels in some countries. Obesity is a serious health condition since excessive body mass is an important risk factor for cardiovascular disease, stroke, hypertension, Type 2 diabetes, and some cancers. The cost of treating obese individuals has put huge strains on large and growing healthcare budgets.

There is a great deal of work that argues obesity is in large part due to changes in technology and relative prices. Lakdawalla and Philipson (2002), for example, write that technological innovations have led to weight increases over time in the population by making production less physically demanding. They also note that engaging in physical activity has become more expensive, both in terms of the direct cost and the opportunity cost of time.

The cost per calorie

In a 2007 New York Times piece, Michael Pollan, author of The Omnivore's Dilemma, writes that the difference (in price per calorie) between fruits and vegetables on one hand and processed food on the other has increased dramatically in the US. Between 1985 and 2000, fruit and vegetable prices in the US increased by about 40%, while the price of soft drinks dropped by 23%. These seem like large changes in relative prices. According to Pollan, the change in relative prices is in large part due to the US farm bill, which provides generous subsidies for corn and soy, which are prime ingredients in high-density “processed food.” Corn syrup, for example, is the primary ingredient in most soft drinks. The farm bill provides virtually no help to farmers growing fresh produce. If that is indeed the case, US government policy truly seems schizophrenic here – bemoaning and trying to combat obesity on one hand and indirectly encouraging it through the farm bill on the other hand. We might want to tax rather than subsidise junk food (Dubois 2007).

And what about the absolute price levels? Drewnowski and Barratt-Fornell (2004) conducted a simple "experiment" in a Seattle supermarket and found that, per calorie, carrots cost virtually five times more than cookies or potato chips, and orange juice costs virtually five times as much as soft drinks. That is a huge difference!

According to Drewnowski and Barratt-Fornell (2004), dry snack food (i.e., processed or junk food) like chocolate, French fries, cookies, and candy all contain very little water and consequently may cause more weight gain than fruits and vegetables. Although the differences in absolute prices vary from country to country, the phenomenon that high-energy density foods are less expensive per calorie than fresh fruits and vegetables is fairly universal. In Israel, vegetables cost roughly three times that of candy (per calorie), while fruits cost more than three times as much as white bread.

Obesity, socioeconomic status, and prices

There is evidence that obesity is a problem for those with lower socioeconomic status (Ruhm 2007; Baum and Ruhm 2009). It makes sense that the high price of fruits and vegetable prevents low-income individuals and families from consuming these foods and that the relatively low price of high density foods has led to an increase in the consumption of the high density foods. But is the problem really just limited to low-income groups?

Anastasia Shabelansky and I examine this question in recent research (Gandal and Shabelansky 2009). The data, which comes from the very thorough 1999-2001 (MABAT) Israeli Health and Nutritional Survey, has several unique features that enable us to examine this issue.

  • First, researchers measured the weight and height of the individuals in the study to calculate the Body Mass Index (BMI).

BMI is the most widely used index for body size and is equal to the weight of the individual (in kilograms) divided by height squared (in meters). A BMI greater than or equal to 30 is a commonly used proxy for obesity. In most other studies, researchers must rely on self-reported levels, and there is evidence from Cawley and Burkhauser (2006) that overweight individuals are more likely to under-report their weight.

  • Secondly, we have data on measured "waist circumference" which is an alternative proxy for obesity.

This measure is attractive because it is not just the amount of fat that matters for health, but rather the location – and abdominal obesity is correlated with morbidity.

  • Finally, we have data on price sensitivity to food products.

Regarding this last point, consumers were asked the following question: When you buy food, how important is price. The range is from 0 to 3, where “0” means that price is “not important at all,” “1” means “not too important, “2” means “important,” and “3” means “very important.” Interestingly enough, our price sensitivity variable is only slightly (negatively) correlated with income. Since price sensitivity is the prime variable of interest and since it is still the case that women make most of the family food purchases in Israel, we conducted the analysis for women.

Price sensitivity and obesity: An opening for a policy lever?

Our main result is that, even after controlling for income and other factors, greater food “price-sensitivity” at the supermarket is associated with a higher obesity rate. We find that women who stated that price was not important at all when purchasing food products had a BMI 1.3 units below those who stated that price was “very important.” A reduction of 1.3 units in the BMI for all obese women would move approximately 25% of women who are in the "obese" category to the "overweight" category (a BMI between 25 and 30.) Our results are similar when we use waist circumference as a proxy for obesity rather than BMI.

We also included income in the study as well. Based on our estimates, the “income” effect was relatively small; in both specifications, a doubling of household income was associated with a reduction in BMI and waist circumference roughly equal to just 25% of the price sensitivity effect. Additionally, the price sensitivity effect is robust to dividing the data into high- and low-income groups. This suggests that sensitivity to food prices cuts across all income classes and is not just an issue for low-income households. Although more work needs to be done, these results suggest that high prices of fruits and vegetables affect both high- and low-income households.


Baum, C., and C. Ruhm, 2007, “Age, Socioeconomic Status and Obesity Growth,” NBER Working Paper 13289.

Cawley, J., and R. Burkhauuser, 2006, “Beyond BMI: The Value of more Accurate Measures of Fatness and Obesity in Social Science Research," NBER Working Paper 12291.

Dubois, P. (2007), "Obesity’s on the rise – let’s have the courage to tax junk food!," VoxEU.org, 17 October

Drewnowski, A., and A Barratt-Fornell, 2004, “Do Healthier Diets Cost More?Nutrition Today, 39:161-168.

Gandal, N., and A. Shabelansky (2009), "Obesity and Price Sensitivity at the Supermarket," CEPR Discussion Paper 7443.

Lakdawalla, D., and T. Philipson, 2002 “The growth of obesity and technological changes: a theoretical and empirical examination,” NBER Working Paper 8946.

MABAT, First Israeli Health and Nutrition Survey 1999-2001, Israel Center for Disease Control of the Israeli Ministry of Health.

Pollan, M., 2007, "You are what you Grow," New York Times, 22 April2007.

Ruhm, C, (2007) "Why are disadvantaged adults more likely to be obese?," VoxEU.org, 10 September.

1,575 Reads