Journal Issue: Childhood Obesity Volume 16 Number 1 Spring 2006
How to Choose among the Policy Options
Given the three different economic rationales for government intervention, what policies are most appropriate? From an economic perspective, the primary goal is to repair the problem in the market. For example, the government can directly address the lack of information by requiring companies to provide it. One simple way to improve the food markets' efficiency is to expand the NLEA to require that detailed nutritional information accompany all foods and menus. Adult consumers, at least, appear to respond to such information. One study finds that a media campaign urging people to shift from whole-fat to low-fat milk changed consumer purchases.62 Another study documents a consumer shift from high-fat toward low-fat salad dressing after product labels were required to reveal fat content.63 Another finds that the NLEA decreased weight gain for white females who read labels while shopping and estimates that the NLEA's benefits totaled $101 billion a year.64
It is not clear, however, how to present nutrition and calorie information so that consumers, especially children and adolescents, can use and understand it. Further research is needed on how to make nutritional information comprehensible to children.
When the government makes information available, it must be careful to put it in a proper context; failure to do so can lead to unfortunate unintended consequences. For example, in 2003, in an effort to inform parents, Arkansas passed a law requiring schools to weigh children and to notify parents of their child's body mass index. Although the law provides parents with information they may have lacked, it has had some unforeseen negative consequences. Some muscular children have been incorrectly classified as overweight, and concerns have arisen about privacy and self-esteem.65 Failing to put the information into its proper context also raises a risk that parents may impose ineffective or harmful fad diets.
Although the government can correct the problem of incomplete information in a relatively straightforward manner, it cannot so easily fix the other two problems—societal costs of obesity and irrational consumers. The typical economic response to societal costs is to tax whatever imposes the cost or to subsidize behaviors that could decrease the societal costs. In this context, that would imply taxing obesity and subsidizing weight loss for the obese, but taxing people based on their weight or changes in weight would be difficult to implement and is politically unattractive. Subsidizing consumers who lose weight or maintain a healthy body weight would be similarly difficult to implement.
It is also hard to imagine how the government could implement a policy to enable children to become entirely rational consumers who take into account the future consequences of their actions. In the absence of an ideal policy to reduce societal costs and address children's “irrational” behavior, the question becomes whether other, “second-best” policies could both decrease obesity and do more good than harm for society overall.
Many possible second-best interventions exist. From the economic perspective, the correct way to choose among them is to analyze their cost-effectiveness. The first step in such analysis is to estimate all the costs and benefits associated with each intervention; the second is to rank the interventions according to how cheaply they achieve the policy goal, thus allowing policymakers to use a fixed budget most efficiently—to get, in other words, the most bang for the buck. Few anti-obesity interventions, however, especially those targeted to youth, have as yet been subjected to cost-effectiveness analysis. Nevertheless, I describe several possible second- best interventions.
Protecting Children from Advertising
If the government decides to protect children from advertising, one particular venue for intervention is the public school system, where the government is solely responsible for the advertising environment. Under budgetary pressure, some school districts have signed contracts with Channel One, which gives them televisions, educational materials, and cash in exchange for allowing Channel One to advertise products such as candy, food, and soda pop directly to children in the classroom for two minutes each day.66 The risk is that children who are a captive audience for such food advertising may increase both their consumption of the advertised foods and their risk of obesity. A cost-benefit analysis can determine whether the benefits of working with Channel One exceed the costs.
Some observers have advocated banning all food advertising to children in all venues.67 Many developed countries, including Canada, Great Britain, and Australia, have banned all television advertising to children.68 In the United States, however, Congress has historically tolerated little regulation of commercial speech. For example, the United States is one of only two industrialized countries (the other is New Zealand) to permit direct-to-consumer advertising of pharmaceuticals.69 In 1979, the Federal Trade Commission (FTC) sought to regulate the television advertising of sugary cereals to children because of concerns about tooth decay. Congress, however, chose to recognize broad latitude for commercial speech and blocked the FTC from pursuing the case.70
The U.S. government has shielded children from advertising in some cases. The 1992 Telephone Disclosure and Dispute Resolution Act, for example, bans advertising of 1-900 phone numbers to children younger than age twelve and requires that advertising directed to children younger than age eighteen include the warning that children need their parents' permission to use the service. And in the mid-1990s the Food and Drug Administration adopted rules against advertising cigarettes near schools or in campaigns targeted to children.
Using Taxes and Subsidies to Change Behaviors That Cause Obesity
As noted, taxing or subsidizing people based on their weight or changes in weight would be politically unattractive and difficult to implement. However, some second-best tax and subsidy policies to alter behaviors may be feasible. For example, policymakers could implement taxes and subsidies that either discourage the consumption of certain foods or encourage physical activity. To evaluate whether such policies are worthwhile, policymakers must weigh their costs and benefits.
One could, for example, tax certain foods. Even though consuming food per se does not impose costs on society, a food tax might sufficiently decrease consumption that obesity would fall, cutting the costs imposed on society. Such taxes have been shown to affect food choices. A series of recent experiments confirms that even schoolchildren's purchases are sensitive to changes in the relative prices of foods.71
But taxing food involves several problems. The first is that although proponents often call for a “junk food” tax, it is not obvious which foods most contribute to obesity.72 Any food, if consumed in sufficient quantity, can contribute to calorie surplus and weight gain. Second, food taxes would be regressive, falling more heavily on poor families who spend a larger share of their income on food than do wealthier families.
The other major option is to subsidize behavior that decreases obesity-related societal costs. In essence, local governments already do that when they subsidize public parks, pools, and athletic facilities and when they provide free physical education, nutrition education, and sports teams in public schools. Government subsidies for installing sidewalks or for full-service supermarkets that stock fresh fruits and vegetables to operate in low-income or minority neighborhoods are other possible interventions.
Before governments increase funding for these programs, though, they should subject them to cost-benefit analyses. So far, subsidies for youth physical activity appear to have little effect. Children's physical activity, for example, appears uncorrelated with the availability of local facilities or with neighborhood safety. Increased physical education requirements are associated with small changes in physical activity but have no detectable impact on weight or the probability of overweight.73
Regulating Food Markets in Schools
Again, a special venue for intervention is the public school, where the government is responsible for the food environment. For example, states could require all schools to remove vending machines for soda and candy. Because children are not generally capable of choosing foods to achieve energy balance, energy-dense foods such as sodas and candy may be the most likely to lead to energy imbalance and subsequent obesity (although, as noted, any food can cause obesity if consumed in sufficient quantity.) Schools could reconfigure meals to consist of low energy-dense foods that facilitate energy balance and serve portions that take into account the portion size effect observed in the research literature. A potential cost of removing vending machines and no longer selling energy-dense foods, however, is that schools may lose considerable revenue from “pouring rights” contracts with soft drink manufacturers and from cafeteria sales, revenue that may be used to advance the educational mission of the school.74
A 2005 Government Accountability Office report found that many schools generate considerable revenue by selling foods outside of their school lunch programs. The report estimates that about 30 percent of all high schools generated more than $125,000 per school through such sales, and that 30 percent of all elementary schools generated more than $5,000 per school. The study found that schools typically use these revenues to offset losses associated with their other food service programs and to fund student activities. Cost-benefit analyses should take into account the impact of any decrease in these revenues that would result from a ban on energy-dense foods.75
Mending or Ending Programs That May Inadvertently Contribute to Obesity
Government intervention could also take the form of modifying or canceling programs that contribute to obesity. For example, cost-benefit analyses could assess the net benefit of agricultural production subsidies and price supports. These programs clearly benefit farmers, and while some observers argue that uninsurable crop risks and weather uncertainty justify this agriculture policy, others counter that current policy is designed primarily to transfer wealth to farmers and processors.76 Farm policy contributes to obesity by lowering food prices, but its effect on weight may be small.77 A cost-benefit analysis could help determine whether society is better off with or without current agriculture policies.
Another existing policy that the government can reconsider is the ban on lawsuits against food companies by plaintiffs who allege that the company's products made them obese. In 2004 twelve states adopted laws that block consumers from filing such lawsuits; so far in 2005 seven more states have followed, and nineteen more states are considering such laws.78 But such blanket liability waivers remove the food industry's incentives to disclose information about the food's content, to exercise restraint in advertising to children, and to ensure the food's safety. Legal scholars are generally skeptical that torts against the food industry will be as successful as recent ones against tobacco companies, in part because no subset of foods can be proven to be solely responsible for causing obesity and therefore no single food or restaurant company can be shown to be liable.79 But to encourage the food industry to keep its customers' welfare in mind, consumers need to be able to pursue such legal cases, which can always be thrown out if frivolous.
Assessing Cost-Effectiveness
Although cost-effectiveness analysis of anti-obesity initiatives is in short supply, some evidence exists. Studies, for example, have calculated the cost of saving a quality-adjusted life year (QALY) associated with specific interventions. (A quality-adjusted life year attempts to take into account the quality of the extra lifespan; for example, an extra year of life in a persistent vegetative state receives a QALY score near zero whereas an extra year of life in perfect health receives a full QALY score of 1.) The decision rule for cost-effectiveness analysis is generally to implement the policy with the lowest cost per QALY and to continue implementing policies until either the initiative's budget is exhausted or the cost per QALY saved rises above some threshold. This threshold, historically $50,000 per QALY, has more recently been raised to $200,000, but other benchmarks are also used.80 For example, Richard Hirth and several colleagues estimate that under various sets of circumstances, Americans are willing to pay from $150,000 per QALY to more than $425,000 per QALY.
The Centers for Disease Control and Prevention have conducted a multiyear project to assess the cost-effectiveness of seven “exemplary” interventions to increase physical activity. (None of the interventions was targeted at children and adolescents.) The study concludes that the lowest-cost exemplary intervention was Wheeling Walks, an eight-week, intensive community-wide intervention that promoted walking among sedentary fifty- to sixty-five-year-olds using paid media and public health activities at work sites, churches, and local organizations. That intervention cost $14,286 per QALY saved.81 The other six interventions were estimated to cost between $27,373 and $68,557 per QALY saved. These estimates were for a forty-year analytic time horizon. For shorter time horizons the costs per QALY were considerably higher (more than $100,000 for a ten-year horizon), because many health benefits of weight loss are reaped only later in life while the intervention's costs are always paid up front.
In contrast, estimates show bariatric surgery for the severely obese costs between $5,400 and $16,100 per QALY for women and $10,700 to $35,600 per QALY for men.82 Providing an anti-obesity drug to overweight patients with diabetes has been estimated to cost $8,327 per QALY.83 These studies indicate that there may be available a variety of cost-effective anti-obesity interventions, some involving prevention and others involving treatment
