February 7, 2010
On Consumer Competence
On Consumer Competence-Becker
I believe that a Consumer Financial Protection Agency will hurt rather than help consumers. Despite the claim that ignorance induced many consumers with few resources to buy houses during the boom, consumers who bought a house then with almost no down payment and low interest rates were not displaying ignorance, but good sense. They put little of their own resources at risk, and annual mortgage payments were cheap, especially in an environment where housing prices were expected to continue to rise at a rapid rate. Lenders, the Fed, and others who made these loans, or helped keep interest rates low, made the mistakes and look foolish, not consumers who bought the houses.
In the vast majority of cases, consumers, even those with little education, know their own interests far better than government officials know them. Still, I have no objections to governments giving consumers information in situations where information is difficult to acquire, as perhaps in the restaurant food safety case cited by Posner. I can also accept putting information about the harmful effects of cigarettes on cigarette packages, although I doubt if that had much effect in reducing smoking.
Posner considers in most detail the issue of growing obesity that continues a discussion we have had in this blog and elsewhere. Obesity in the United States and other countries has been increasing for the past 30 years due to much lower costs of calories, as in fast foods, and also due to more sedentary activities, mainly due to the large amount of time spent watching television, using the computer and Internet, and speaking on cell phones. Giving consumers information about the adverse health consequences of obesity might be a useful government activity, but Posner wants to go much further by essentially taxing some of the sources of obesity, such as sodas and computer games. I do not agree with him.
Posner supports aggressive actions against obesity partly because he believes that consumer ignorance of the adverse consequences of being obese helps explain its high rate of incidence. He cites the fact that obesity is much more common among those with low income and low education. People with low incomes and education differ in many respects from others, such as their greater tendency to buy older used cars that are more likely to break down. No one would suggest they do this out of ignorance rather than from the constraints of low income. Similarly, fast foods, and fewer visits to health clubs, would appeal to low income persons without any need to stress ignorance.
That ignorance is not the main factor in the high obesity rates among low income and low educated persons is supported by the fact that while the rate of obesity is much higher among African-American women than among white women, it is not so much higher among African American men than among white males. It is hard to believe that African-American men are much better informed about the harmful effects of obesity than are African-American women. This racial-gender difference in obesity rates is most likely related to differences in the life situations of African-American men and women rather than to differences in their knowledge of the health cost of obesity.
Posner's second main argument for taxing obesity is that it imposes an externality on taxpayers because they pay much of the higher cost of medical care due to obesity. The most frequently cited study of the extra medical spending due to obesity is by Finkelstein, et al. in the July 27, 2009 online issue of Health Affairs. It is a careful analysis of data for 2006 that looks at the higher spending on medical care of the obese, after holding constant income category, years of schooling, age, and many other variables. Overall, the authors find that about 25% of the population is obese, and that obesity increases medical spending by about 9%, which amounts to $147 billion (in 2008 dollars).
They also find that private payers, not taxpayers under Medicare and Medicaid, bear the majority of the additional medical costs due to obesity. Since private insurance companies are not allowed to charge higher premiums to the obese because that is considered discrimination, largely under the Americans with Disabilities Act, the higher cost of obesity paid by privately insured persons can hardly be called an "externality", unless it is considered an externality from government policy.
If we take away more than half the total increase in medical spending due to obesity because it is not borne by taxpayers, we are left with a much more modest increase in health spending "externalities" due to obesity. Even that smaller percent is too large because it neglects the "savings" due to the fact that the obese die earlier than other persons. On this externality logic, that saving should be subtracted to get a net "externality". Some analysts have even claimed that this net externality is negative, so that obese people actually reduce taxpayer spending on medical care. If that were correct, should we subsidize obesity?
I have not accepted such externality arguments related to medical care ever since several economists showed that smoking cuts down medical (and retirement) spending because heavy smokers usually die early, and do not collect much in the way of social security and Medicare benefits. Surely that would be a foolish justification for subsidizing smoking. Yet the same logic applies to attempts to justify taxes on obesity because of any medical costs obesity imposes on taxpayers.
Since even ignoring their earlier deaths, the obese raise government medical spending by rather modest amounts, that "externality" provides a weak case for taxing goods like soft drinks with sugar, fast foods, and ice cream, or taxing computer games, watching television, and other sedentary activities. As Posner recognizes, taxes on these types of foods and time use activities would be a shot gun approach to reducing obesity since the great majority of these foods and activities are consumed by men and women who are of normal weight, or merely overweight (an earlier Rubenstein, et al study does not find increased medical spending for men and women who are simply overweight). Higher taxes on these items would cause a reduction in the consumption of these items, and hence an unjustified loss in welfare, by the 75% of the population who are not obese.
To the extent that obesity imposes costs on others, it would be far better to let private health insurers price coverage based on obesity and related risk factors. This would target the desired group, and not impose inefficient costs on non-obese persons by taxing the foods they eat or the activities they choose to engage in.
Providing information about the negative health consequences of obesity will do no harm, and might help prevent some persons from getting or staying obese. Doing more than that, however, has little justification from externality arguments, and hence is a weak basis for public policy.
The Question of Consumer Competence—Posner
When is it proper for government to try to protect people, in their capacity as consumers of goods and services, from themselves? And not just children or people with serious mental problems, but normal adults. Can't normal adults protect themselves? And if they can't, won't competition among sellers protect them?
These questions are acutely raised by the proposal, now before Congress, to create a Consumer Financial Protection Agency that would protect consumers of financial products such as mortgages and credit cards and payday loans not only from misrepresentations by sellers of these products, but also from their own ignorance or poor judgment. The proposal draws on behavioral economics, which teaches that cognitive and psychological limitations frequently lead consumers to make mistakes, even when there is no fraud by sellers.
The specific proposal seems to me misconceived. Its premise is that the housing bubble and ensuing financial collapse were due in significant part to reckless borrowing to finance home purchases or borrow against home equity in order to obtain cash to buy other goods and services. The argument is that people didn't realize the risk involved in buying a house with very little (sometimes zero) equity, especially if they financed it by an adjustable-rate mortgage, which might become unaffordable by them if interest rates rose.
No doubt some people didn't realize they were taking a risk, but I don't think that that is the explanation for the housing bubble. Almost no one, including sophisticated economists and financiers, realized that the steep increase in housing prices that ended in 2006 was a bubble phenomenon. If it was not, then homebuying wasn't really risky, because one could anticipate that the market value of one's house would grow, and this would create sufficient equity to be able to refinance one's mortgage on attractive terms. There was a speculative element but it did not seem extreme because so few experts believed there was a housing bubble. Among these experts notoriously was Ben Bernanke.
I want to contrast with the proposal to curtail risky borrowing by consumers two types of consumer protection that seem to me justifiable, and this regardless of the insights of behavioral economics. One is requiring cigarette labeling and advertising to carry warnings of the health hazards of smoking. This regulation is not very important today because everyone knows about these hazards, but it was important in the 1960s when the existence and gravity of the hazards were first confirmed. Obviously individual consumers were not in a position to study the health effects of smoking—which cigarette manufacturers were busy denying—but one might think that advertisers of competing products would have had an incentive to frighten consumers away from smoking. But this would not be a realistic expectation. What would consumers think if a manufacturer of chewing gum advertised that chewing gum, unlike smoking cigarettes, does not cause lung cancer? Nor would cigarette manufacturers whose cigarettes contained less tar and nicotine than the average be strongly motivated to advertise the fact, because they would be telling the world that cigarettes are hazardous, at a time when this was not generally realized. Automobile manufacturers were slow to offer seatbelts, perhaps fearing they would be advertising the dangers of driving—and charging a higher price (to cover the cost of the seatbelts) at the same time.
My second example is inspections of restaurants and food processors by government inspectors, to prevent food poisoning. One can imagine leaving food safety to the market, reinforced by tort remedies against the sale of unsafe products. But solvency limitations would make market and tort remedies ineffectual against many sellers, especially small and new ones—so the inspection regime actually facilitates new entry, which is a dominant feature of the restaurant industry. Food poisoning can cause death, indeed multiple deaths, and when the consequences of a market failure are very grave, there is an argument for preventive regulation.
Now I want to discuss an important intermediate case, where the argument for consumer protection seems to me stronger than the case for consumer financial protection (other than against fraud), but not so strong as in the cases I just gave. That is the case of obesity. According to the Weight Control Information Network, which is part of the National Institutes of Health in the Department of Health and Human Services—and I believe reputable—two-thirds of American adults are overweight and one-third—an astonishing percentage—are obese, defined as having a Body Mass Index (the ratio of a person's weight in kilograms to the square of his height in meters) of more than 30. So, for example, a woman 5 foot 6 inches tall would be deemed obese if she weighed 180 pounds, and a man 6 foot 1 inch tall would be deemed obese if he weighed 227 pounds.
Obesity is measured differently for non-adults, but 17 percent of young children and 17.5 percent of adolescents are estimated to be obese.
These are startling figures, and considerably higher than in virtually any other country in the world. My esteemed colleague Becker has argued, however, that American obesity is not excessive in an economic sense. Obese people may simply have traded off the pleasures (and economy) of eating cheap, tasty, and nutritious food against the costs in disagreeable appearance, impaired mobility, the greater danger of and longer recovery time from surgery, and the much greater incidence of Type II diabetes and joint problems; there is also a greater risk of heart disease and possibly of dementia. Becker believes that the long-run expected costs of obesity may be small if continued advances in medical technology eliminate or greatly reduce the health problems that obesity creates, and that the realization of this possibility is one of the factors that people consider in deciding whether to allow themselves to become obese.
I am skeptical. The problem of obesity is concentrated in the poorer segment of the population, among people with limited education who may be unable to assess the health risks of obesity and as a result are unwilling to incur the slight added expense (or cost in diminished eating enjoyment of a diet less rich in sugar and butter). They may also be imperfect agents of their children; and a person who becomes obese as a child will find it more difficult to avoid obesity than people who were thin children. Governmental paternalism when directed to children is less problematic than paternalism toward adults.
There is also an externality, which is a nonpaternalistic justification for government intervention. The government, meaning ultimately the taxpayer, now pays for half of total U.S. medical expenses. The average medical expenses incurred by obese people are substantially greater than those of the nonobese, even after allowance for the shorter life spans of obese people. (It has been responsibly estimated that obesity and overweight add $150 billion a year to the nation's medical bill. This is somewhat too high because it ignores the effect of obesity in reducing longevity. On the other hand it excludes the $40 billion a year that people spend on diet programs.) Much of the additional cost to Medicare, Medicaid, and other public programs is borne by taxpayers who are not obese. Private health insurers are forbidden to "discriminate" against the obese by charging higher premiums to them, which is an unsound policy that should be changed (it won't be). But probably many obese people are not insured, and their medical expenses are paid by charity or government, for example when they seek medical care in hospital emergency rooms and cannot pay the price they are billed for that care. Failure to pay the full medical costs imposed by obesity distorts the decision of a person to become or remain obese; it is a subsidy for obesity. I find it difficult to imagine a more grotesque subsidy.
But whether it would be desirable for the government to try to reduce obesity depends on the cost and efficacy of the measures it might take. Some of the common proposals are likely to have only modest effects, such as requiring restaurant menus to disclose calories. People who are motivated to avoid obesity know or can easily discover the approximate caloric content of the various foods, and people have most of their meals at home rather than in restaurants.
Somewhat more promising measures are: instruction in nutrition and the dangers of obesity in elementary and high schools; healthful school-lunch programs; expanded compulsory physical education in schools; restrictions on foods that can be purchased with food stamps; a tax on advertising fast food; a tax on video games; a ban on food advertisements aimed at children; a relaxation of regulations of health insurance that discourage charging higher premiums to the obese (and that thus subsidize obesity); a tax on soft drinks that contain sugar; and a calorie tax. All would be relatively inexpensive measures that would have a good chance of paying for themselves. The last, the calorie tax, which would probably be the most effective measure, would be a Pigouvian tax—a tax designed to internalize an externality, and, as such, defensible on standard economic grounds if I am correct that obesity creates an externality.
Still, such a tax can be criticized on two grounds. One is that it would be strongly regressive. But its regressive effect could be offset by a more generous food-stamp program. The second objection, emphasized by Becker, is that a tax on calories penalizes people who are not obese, and they are the majority. It is the same objection that can be made to alcohol taxes as a means of curtailing drunk driving: most of the people taxed are not drunk drivers. A more efficient anti-obesity tax, in principle, but utterly infeasible politically, would be a head tax measured by weight.
I am not much impressed by "fairness" objections to taxes. Taxation is inherently arbitrary, because it doesn't match the taxes paid by a person to the services he receives from government or the costs he imposes on society. A calorie tax would raise considerable revenue, because like most Pigouvian taxes it would result in only limited substitution away from the taxed good, and the government at present is in desperate need of additional revenue.
A more efficient tax would be a tax on producers of food, based on the difference between the cost of the ingredients before processing and the price for the finished product. The tax would therefore fall more heavily on highly processed foods, which tend to be higher in calories, than on lightly processed ones.
More study is necessary, however, before the costs and benefits of a well-designed program of obesity reduction can be responsibly assesssed.