There are countless proposals to reduce obesity in the United States, but causal links are difficult to establish. Fairly and effectively targeting low-income populations is an unresolved challenge.
Turn on the television in the United States and you will quickly learn that many people are fat, and most would like to be skinnier. Contestants on the “Biggest Loser” submit themselves to getting weighed in their skivvies, and undergo brutal weight loss programs to have a chance at a cash prize. MTV follows a younger demographic of obese people on “I Used to be Fat.” TLC follows a fat African American family trying to lose weight on “One Big Happy Family.” Most bizarrely, the inhabitants of a small town in West Virginia allow themselves to be publicly berated for eating breakfast pizza by Jamie Oliver, who intends to lead a “Food Revolution.”
TV actually has it right. One third of American adults are obese, and judging by sales to the $60 billion diet industry, many would like to lose some weight. After climbing steadily for two decades, the obesity rate for adults may in fact be stabilizing. The trend appears similar for children. The effects of rising obesity on cardiovascular health, diabetes, and hypertension threaten to reverse substantial gains made in increasing life expectancy in the second half of the twentieth century.
No demographic group has been entirely immune from the rising obesity tide, but the impact has been much greater for racial minorities and people in southern states. Interestingly, among adults there exists a negative obesity-income gradient for women, but not for men. The same holds true for education. The figure below, from the CDC, illustrates that the growth of obesity for men and women between the 1988-1994 the 2005-2008 affected all income groups.
There have been a surfeit of proposals to reduce obesity ranging from changes to the school lunch program, improved labeling of food items, elimination of the subsidies that keep unhealthy food artificially cheap, promotion of healthier neighborhoods, and better education of consumers. The problem is that there is a lack of evidence that any of these interventions would have a large and sustained impact on the dietary patterns of Americans, especially those in the lowest socioeconomic groups. To know how to intervene on obesity, we must first have a sense of what factors are driving changes in the American diet. Some compelling quasi-experiments from economists give us a sense of how the obesity epidemic has unfolded at the population level.
Food is relatively cheap in the United States, and unhealthy food is cheapest of all. One hypothesis is that obesity is a rational response to the price of food and other goods, and the utility that any of those purchases is likely to yield. Consumers decide how much food to consume based on how much disutility consuming the food will produce both in monetary terms and in health terms, against how much satisfaction they are likely to get from eating the food. One prediction from this hypothesis is that a decrease in food prices will cause a commensurate increase in food consumption and obesity. The effect is likely to be stronger for the poor for whom food is one of the few available forms of leisure available, and where individuals are not expected to live long lives (“if I can’t afford to take a vacation, and I’m not likely to live very long anyway, I might as well enjoy this junk food now”). In the aggregate, the obesity epidemic in the United States began during the 1970s when food prices declined sharply. However, obesity continued to rise despite a leveling off of food prices in the mid 1980s (see this), suggesting that changes in food price alone cannot explain rising obesity. A review of studies on food prices and obesity in the Milbank Quarterly in 2009 concludes that there exists a small, but significant relationship between food prices and obesity. To curb people’s dietary patterns, however, would require large taxes or subsidies:
“Based on the findings of the studies we reviewed, we estimated that small taxes or subsidies were not likely to produce significant changes in BMI or obesity prevalence but that nontrivial pricing interventions might have a measurable effect on Americans’ weight outcomes, particularly those of children and adolescents, low-SES populations, and those most at risk for overweight. Even though they would have only a small impact on individual behavior, such interventions could have a large impact at the population level when applied broadly.”
Availability of Food
In some neighborhoods there are burger and pizza joints on virtually every corner, but not a grocery store for miles. The disproportionate rise of fast food in urban areas, and the resulting spatial inequality in the availability of food (including the emergence of “grocery store deserts”), is one of the most widely discussed explanations for rising obesity. While the correlation between obesity and fast food in the neighborhood is well established, few studies have convincingly established causal mechanisms.
A 2010 paper in the American Economic Journal by Janet Currie and coauthors provides perhaps the most thorough and rigorous analysis of fast food availability and obesity. The authors create a spatial dataset of fast food establishments linked to the addresses of women during their pregnancies and for ninth grade students (pregnant women are observed through vital statistics data, students take a fitness exam that collects body mass index). They find a large effect of proximity of fast food for student weight gain: Among ninth graders, a fast food restaurant within 0.1 miles of a school results in a 5.2 percent increase in obesity rates. Among pregnant women, a fast-food restaurant within 0.5 miles of residence results in a 1.6 percent increase in the probability of gaining over 20 kilos. While the effect for pregnant women is substantially smaller on average (they claim that adults are much less sensitive to travel costs than children), they find that minorities and low education women are most sensitive to the opening of a fast food establishment. The authors subject their data to a number of specification checks to ensure that the observed effect is not confounded by selection bias (as communities with higher propensity to be obese could also attract fast food establishments), and find the results are robust.
By now, many people are familiar with the idea that “friends can make you fat.” Although surely a caricature, this idea is rooted in the 2007 New England Journal of Medicine study by Christakis and Fowler. The study, remarkable in its innovation and scope, constructs a social network with 12,067 subjects in Framingham, Massachusetts observed over 30 years tied together through friendship and family relationship. The headline finding of the study is that a person’s chances of becoming obese increased by 57% if he or she had a friend who became obese in a given interval. Similar effects were found among siblings and spouses. Remarkably, no significant relationship was found between geographic neighbors not connected through friendship or between people who are identified by others as friends, but do not reciprocate the friendship tie. Although there is a well-established correlation between quitting smoking and gaining weight, the authors find that the spread of smoking cessation did not account for the spread of obesity in the network.
The Framingham study provides plausible evidence that it is not merely shared exposures, or selection into social networks (homophily), that drives rising obesity, but that friends and family can also induce others to gain weight (induction). This idea is plausible, since behaviors among social groups are often mutually reinforcing or normative – the best natural experiment of this is observing how dietary and other habits change dramatically among college freshmen (and indeed, randomly assigned roommates have a significant effect on grades). Nevertheless, it is still a matter of some controversy whether the NEJM paper identifies a purely causal relationship between friends, or whether there is some remaining confounding. In a companion response to the Christakis and Fowler article, Cohen-Cole and Fletcher argue that the latter cannot be completely ruled out, but that:
“Though we advise caution in interpreting the available evidence of a social contagion in weight, we concur with CF on the use of network phenomena, broadly writ, to help to ameliorate the epidemic. Some of the encouraging evidence mentioned in CF, in particular the fact that alcohol and smoking cessation programs are more effective when coupled with peer support, can be exploited independent of the degree of induction present.”
Each of the quasi-experimental studies reviewed here ties in with a potential intervention, taxes and subsidies to change the relative prices of junk food, zoning laws and subsidies to affect the mix of urban food establishments, interventions to provide support and weight loss help within social networks. Any and all of these things may be sensible policies to pursue (and indeed many of them are already in the works).
Still we can ask, who is likely to benefit the most from programs of these types? The available evidence suggests that fresh food retailers are often reluctant to establish in poor areas because they are uncertain that even with subsidies the populations will generate enough demand for fresh fruit and vegetables to allow them to make a profit. Making fresh food available is only as good as there exists the knowledge and ability to prepare those foods – anecdotal research from my own previous research and from systematic study of nutritional knowledge suggest that many young people, particularly low-income, lack basic cooking and nutritional knowledge. And from an equity standpoint, we know that all consumption taxes, especially on basic staples like food, are borne most heavily by the poor who spend the largest share of their income on those goods.
Meanwhile, those fighting against obesity in low income communities continue their struggle to help people to change entrenched eating habits in places where the options are limited and the temptations abundant. That fight continues, without motivational trainers, cash prizes, or television cameras.