Optimal weight: are government goals for reducing obesity sensible?

Marlow, Michael L. ; Shiers, Alden F.

The rising prevalence of obesity in the United States is often referred to as an epidemic (although it has apparently leveled off since 1999). Obesity is defined as a body mass index (BMI) of 30 or higher, and it has been associated with many health problems, including diabetes, hypertension, high cholesterol, heart disease, stroke, sleep apnea, some cancers, gallstones, gout, asthma, and osteoarthritis. Based on 2005 Medical Expenditure Panel Survey data, medical spending on obesity in the U.S. non-institutional adult population has been estimated to be $168.4 billion (in 2005 dollars), which was 16.5 percent of all medical spending that year.

Concern over rising health care costs has predictably encouraged a growing number of government interventions aimed at reducing the prevalence of obesity. Examples of such interventions include restrictions on soda sales at public schools, special taxes imposed on sodas, disallowing soda sales for food stamp recipients, regulations requiring restaurants to post caloric content of menu items, bans on toys offered in children's meals with high levels of calories and salt, and restrictions on locations of new restaurants.

Researchers typically assume that reduction of obesity prevalence is desirable without addressing the more fundamental issues of its optimal level, whether its optimal level has grown over time, and whether optimal levels are identical for all individuals. In this article we develop a simple demand/supply framework to model the optimum level of obesity. We examine these fundamental issues before evaluating desirability of government interventions. Our conclusions run counter to conventional wisdom that government has the necessary information to systematically reduce the prevalence of obesity in line with optimal levels that differ between individuals.

The Model

Weight gain is caused by an imbalance between calories entering the body and calories leaving the body. Obesity arises when the intake of calories sufficiently exceeds the outflow of calories in a manner that results in a BMI of 30 or higher.

We use the model that Thorkild Sorensen proposed in a 2009 paper to show the relationships between energy input, energy output, and weight. If EI is energy input and EO is energy output, then a positive energy imbalance, EI--EO > 0, results in some energy stored, ES. EC is the energy used to convert surplus energy into tissue mass. The change in energy stored is then:

(1) [DELTA]Es / [DELTA]t = EI--(EO + EC)

Changes in energy stored result in changes in body weight. EO is composed of the basal metabolic rate, BMR, and the energy spent on physical activity. If PAF is the physical activity factor, then EO can be expressed as EO = BMR X PAF. The change in energy stored, and hence weight, can then be expressed as:

(2) [DELTA]ES/[DELTA]t = EI--(BMR X PAF + EC).

Equation (2) identifies factors that affect body weight. Energy input and physical activity are determined by choices that individuals

make. BMR and EC depend on genetics as well as other factors including body weight, the amounts of lean and fat tissue, gender, and age.

[ILLUSTRATION OMITTED]

Energy (i.e., calorie) input and physical activity choices made by consumers and producers in the economy can be expressed by demand and supply schedules of weight. Choices determine whether weight gain is positive, negative, or zero. Weight gain arises from engaging in a mix of activities that results in intake of calories exceeding outflow of calories. Eating, drinking, and undertaking sedentary leisure activities are ways of demanding excess calories and, hence, weight gain. Demand also represents the marginal benefit schedule of weight as derived from satisfaction received from consuming another calorie or enjoying an additional restful moment.

Supply of weight comes from sellers of calories and providers of less physically active lifestyles. Supply represents the marginal opportunity cost of weight gain. Costs include those associated with acquiring and consuming calories, wages that may be lost due to reduced productivity caused by rising weight, health and medical costs associated with weight gain, and costs of engaging in more sedentary lifestyles.

Figure 1 displays equilibrium price and quantity of weight as determined by the intersection of demand and supply. The equilibrium quantity represents the optimum level of weight. There is also some rate of obesity prevalence for society associated with this optimum. This quite simple model suggests several important issues associated with obesity.

First, the optimum level of weight changes as demand and supply vary over time. Factors that cause demand or supply to shift rightward result in higher optimum levels of weight. Many causes of increased demand for weight gain have been suggested. These include: increased consumption of sugar-sweetened beverages, reduction in real prices of food, urban sprawl, reduced cigarette smoking, less time spent preparing healthy meals at home, eating more food from restaurants, rising numbers of food stamp recipients, and food engineering that stimulates the brain in manners that increase eating.

Factors that have been suggested as increasing supply include technological change leading to a more sedentary lifestyle, increased availability of restaurants, a growing lack of grocery stores selling healthy foods, and agricultural policies that encourage production of "excess calories."

From the standpoint of economic efficiency, rising obesity reflects shifts of demand and supply of weight over time. This is surely a contentious conclusion given that the literature on obesity focuses on prevention of obesity rather than examining whether its rise is somehow linked to changes in its efficient level. Nonetheless, marginal benefits still equal marginal costs, although optimum levels have apparently increased over time.

Second, optimal weight, and hence optimal prevalence of obesity, is likely to be different for different individuals. Simple observation indicates a wide diversity among individuals. Genetics is known to affect weight. As expressed in Equation 2, genetics can affect weight through its effects on the basal metabolic rate and energy consumption. Subgroups of the population that are genetically more predisposed to obesity experience more weight gain and higher levels of obesity prevalence than other subgroups for identical levels of energy input and physical activity factor. Genetic predispositions to obesity are believed to partially explain why obesity prevalence has risen at different rates among groups.

[FIGURE 1 OMITTED]

This effect is illustrated in Figure 2, where group B individuals are more genetically predisposed to weight gain and thus more readily turn excess calories into additional weight than do individuals in group A. Marginal costs are also lower for group B because their bodies are genetically more predisposed to turning excess calories into weight gain. Population subgroup B will have a higher optimal weight and obesity prevalence level than group A, even if the demand for weight is the same for both subgroups. Of course, demand may vary between groups as well, thus indicating that a "one size fits all" prediction for optimal weight makes little sense.

[FIGURE 2 OMITTED]

Figure 2 illustrates that setting a goal to achieve the same obesity prevalence levels for all groups in a society is misguided. If group B is at weight [q.sub.A], then the marginal benefits of weight exceed the marginal costs of weight for group B. Group B's optimum resides at [q.sub.B]. Group B would not be at its optimum level if it were somehow coerced through government intervention into becoming slimmer in order to achieve a uniform policy goal of [q.sub.A], Adopting a "one size fits all" policy goal for weight thus exerts an "excess burden" on those subgroups that exhibit optimal weight in excess of government goals.

Healthy People 2010, a federal program to promote healthy living that was started in 2000, set a goal of achieving a 15 percent obesity prevalence rate for all categories of adults and a 5 percent obesity rate for children by 2010. The goals were not achieved by any state of the United States, yet the same obesity goals are contained in Healthy People 2020, the successor program. Table 1 exhibits obesity prevalence by state using data collected by the Behavioral Risk Factor Surveillance System. Prevalence for 1995 and 2009, and the percentage change over this period, are displayed. These data are frequently cited in news reports and by obesity researchers as evidence of an obesity epidemic that requires immediate and dramatic government intervention.

There is little reason to believe that uniform prevalence goals are derived from any economic model within a demand and supply framework as developed in our paper. The fact that one state exhibits higher obesity prevalence or a larger increase over time does not necessarily or directly correlate with the degree to which it diverges from optimal weight. Differences in obesity prevalence and their rates of change clearly differ substantially by state, but these differences surely reflect variations in demand and supply across states and over time.

Data from the National Health and Nutrition Examination Survey are also frequently cited as proof of an obesity epidemic. Data indicate that about one-third of adults in the United States are obese, with woman having a slightly higher obesity rate than men. Non-Hispanic blacks have an obesity prevalence rate that is about 36 percent greater than Non-Hispanic whites. Hispanics have a prevalence rate about 19 percent greater than non-Hispanic whites. About 17 percent of children and adolescents aged 2 through 19 years are classified as obese. Again, these data reflect that different groups of individuals have experienced different variations in demand and supply over time that do not directly indicate the degree to which various groups exhibit variations from optimal weight.

Government Intervention

Presence of externalities is often used to justify government intervention to reduce obesity. It is often claimed that the obese do not pay their full health care costs because their above-average medical costs raise insurance costs for all other insured individuals and because some portion of their medical costs are publicly funded. However, obese individuals are known to have shorter life expectancies than the non-obese and thus their lifetime medical costs are lower than their slimmer counterparts. Jayanta Bhattacharya and Kate Bundorf, in a 2009Journal of Health Economics paper, also find that obese workers with employer-sponsored health insurance pay for their greater medical costs by receiving lower cash wages than are paid to non-obese workers. In addition, Bhattacharya and Mikko Packalen, in a 2008 paper, argue there is a positive innovation externality associated with the obese that roughly matches any negative Medicare-induced health insurance externality of obesity. They conclude there is no rationale for "fat taxes" because of the Medicare-induced subsidy of obesity.

The negative externality argument is thus less than persuasive. In any case, a more efficient method to account for additional medical costs of obesity would be to directly charge insurance premiums that reflect the risk of incurring greater medical costs.

Ignorant and lazy?

Proponents of government intervention also argue that consumers lack self-control and adequate information on products such as sugar-sweetened beverages. A 2009 New England Journal of Medicine article by Kelly Brownell et al. argues:

   [M]any persons do not fully appreciate the links between
   consumption of these beverages and health consequences; they make
   consumption decisions with imperfect information. These decisions
   are likely to be further distorted by the extensive marketing
   campaigns that advertise the benefits of consumption. A second
   failure results from time-inconsistent preferences (i.e., decisions
   that provide short-term gratification but long-term harm). This
   problem is exacerbated in the case of children and adolescents, who
   place a higher value on present satisfaction while more heavily
   discounting future consequences.

Such notions are widespread, as evidenced by the constant, uncritical repetition of that notion by purported experts, policymakers, social commentators, and the media. But the scientific basis for this notion is unclear. And even if "excessive" soda consumption is a product of short-term gratification syndrome, it remains doubtful that policymakers can somehow overturn this human failing without exerting unintended adverse effects on others.

Government intervention aimed at lowering tobacco use offers several examples of unintended effects. A 2004 Health Economics paper by M. C. Farrelly et al. and a 2006 American Economic Review paper by J. Adda and F. Cornaglia both indicate that tax hikes on cigarettes have led smokers to switch to higher-tar and -nicotine brands so that they can maintain chemical intake levels as they smoke less, to the detriment of their health. A 2004Journal of Health Economics paper by Shin-Yi Chou et al. found that higher cigarette prices (stemming from tax hikes), which reduce smoking, are associated with higher rates of obesity.

Interventions are also likely to impose costs on the non-obese as well as the obese. For example, taxes imposed on alcohol mostly lower consumption of light users with little to no effect on heavy drinkers. Such interventions are also often regressive in nature, with burdens on the poor higher than the non-poor.

Policymakers also suffer from an information problem themselves when attempting to levy Pigovian taxes on supposed externalities. The "correct" tax requires knowledge that certainly does not exist. A 2010 Obesity Reviews analysis by B. Rokholm et al. of the obesity epidemic notes that clear evidence on specific causes of the obesity epidemic is lacking. The above-discussed New England Journal of Medicine article provides scant hope that "correct" soda taxes are known; the authors conclude: "As with any public health intervention, the precise effect of a tax cannot be known until it is implemented and studied, but research to date suggests that a tax on sugar-sweetened beverages would have strong positive effects on reducing consumption." This is wishful thinking given recent evidence that a one percentage point increase in the tax rate on soda was associated with a decrease of just 0.003 points in body mass. In other words, large tax increases are unlikely to exert much effect on population weight. Evidence indicates that a 58 percent tax on soda, equivalent to the average federal and state tax on cigarettes, would drop the average body mass by only 0.16 points--a trivial effect given obesity is defined as a BMI of at least 30.

Finally, there is little evidence that previous government intervention has lowered obesity among the poor. A 2004 U.S. Department of Agriculture review by P. Linz et al. concludes that, despite many low-income individuals being both obese and recipients of one or more food assistance programs, the research literature does not show that programs have lowered obesity. (The review does cite two studies that find a positive correlation between food stamps and obesity in women, although neither study tested for a causal connection.) More recently, a paper by Jay Zagorskya and Patricia Smith reports that the typical female food stamp participant's BMI is significantly more than someone with the same socioeconomic characteristics who is not in the program. For the average American woman, this means an increase in weight or 5.8 pounds. Good intentions aside, we should be skeptical of the notion that the expansion of government programs would somehow lower obesity when research has yet to prove that past programs have not inadvertently encouraged obesity.

Can "Nudges" Promote Efficient Weight?

Behavioral economists Richard Thaler and Cass Sunstein argue that policymakers should "nudge" individuals toward efficient decisions. Because they "nudge" rather than strong-arm or explicitly prohibit behaviors such as obesity, nudges are labeled "libertarian paternalism." Thaler and Sunstein believe these labels allow them to escape negative connotations attached to paternalism--policies aimed at protecting individuals who are believed unable to protect themselves. For example, they write, "People often make poor choices and look back at them with bafflement!" Behavioral economists thus attempt to correct self-inflicted behaviors that cause us to exercise too little, eat too much, take on too much debt, smoke tobacco, drink too much alcohol, and save too little for retirement.

Rearranging food placements in cafeterias so that healthy foods are more prevalent and sweets are less so is one nudge favored by behavioral economists who believe diners have difficulty controlling impulses to eat unhealthy food. Grocery managers could nudge shoppers by replacing candy with healthier snacks near checkout stands, since this location is known to spark impulse buying.

But it is important to recognize differences between "nudging" by businesses versus governments. Profits motivate businesses and thus their nudges foster efficiencies, since otherwise there would be no purpose. For example, rewards for staying in good health are nudges that are in line with raising profits. The private marketplace has responded to the increase in obesity by providing various means of reducing weight gain. Diet sodas and diet foods are readily available in stores. Sales of Diet Coke overtook those of Pepsi-Cola for the first time in 2010, making it the number two carbonated soft drink in the United States. Exercise equipment can be easily obtained and there appears to be an ample supply of health spas and gyms. Some businesses now pay their employees to lose weight. Private industry undertakes much research seeking medicines that will reduce the costs of achieving weight loss. Unlike government interventions aimed at weight reduction, the costs of these private activities are not imposed on the non-obese.

The private sector is thus actively involved within its goal of maximizing profits. Government and behavioral economists operate under no such profit constraint and thus efficiency may have little to do with their motivation. Just as government cannot match supply with demand better than markets, behavioral economists are unlikely to know how to successfully nudge us toward greater efficiency even when they believe they have uncovered irrational behavior associated with weight gain.

There are other downsides to such nudging. Consider food labeling laws that require restaurants to list their fat and calorie contents. Sounds good at first, but it might also lead some diners to exercise less caution and personal judgment simply because "nudgers" have taken on the responsibility for watching what we eat. Nudges make it less important to think on our own. Intervention may also make it appear that the "eat less, exercise more" adage no longer is a surefire recipe for controlling weight. Substituting government for personal responsibility rarely works out as planned.

There is also evidence that such nudges do not work so well. A 2009 study by B. Elbel et al. of New York City's 2008 law on posting calories in restaurant chains examined how menu calorie labels influenced fast food choices, information on patrons of fast food restaurants in New York communities was compared with that on patrons in Newark, N.J., a city without labeling laws. While 28 percent of patrons in New York said the information influenced their choices, researchers could not detect a change in calories purchased after the law. A similar conclusion was reached in a 2011 study by Eric Finkelstein et al. of a mandatory menu-labeling regulation requiring all restaurant chains with iS or more locations to disclose calorie information in King County, Wash. No impact on purchasing behavior was found, as measured by trends in transactions and calories per transaction.

Finally, it is perhaps obvious, but "libertarian paternalists" place themselves in the role of fathers guiding the actions of children. This role is appropriate when exercised by parents over children, but it remains questionable to award behavioral economists this same role over adults.

Conclusion

There is no question that the prevalence of obesity has risen dramatically in recent years. Researchers typically assume its reduction is desirable without addressing the more fundamental issue of its optimal level. Our paper suggests optimal levels of obesity have increased over time and that optimal levels are not identical for all individuals or groups. Meanwhile, the federal government has set a goal of 15 percent for adult prevalence and 5 percent for child prevalence. Adopting a "one size fits all" policy goal for weight thus exerts an "excess burden" on those subgroups that exhibit optimal weight gain in excess of government goals.

There is little evidence that obesity stems from some sort of market failure. And even if a negative externality exists, government does not command the required expertise to systematically reduce its prevalence toward optimal levels. Placing identical goals for obesity rate reduction across all individuals also exerts excess burdens on those individuals who differ from government's mandated "ideal" weight. There is also no reason to believe that "ideal" weight bears any correspondence to optimal weight.

READINGS

* "A War on Obesity, Not the Obese," by Jeffrey M. Friedman. Science, Vol. 229 (2003).

* "Adult Obesity Prevalence in Canada and the United States," NCHS Data Brief, No 56, by M. Shields, M. D. Carroll, and C. L. Ogden. National Center for Health Statistics, 2011.

* "Agricultural Policy and Childhood Obesity: A Food Systems and Public Health Commentary," by David Wallinga. Health Affairs, Vol. 29, No. 3 (2010).

* "An Economic Analysis of Adult Obesity: Results from the Behavioral Risk Factor Surveillance System," by Shin-Yi Chou, Michael Grossman, and Henry Saffer. Journal of Health Economics, Vol. 23 (2004).

* "Calorie Labeling and Food Choices: A First Look at the Effects on Low-income People in New York City," by B. Elbel, R. Kersh, V. L. Brescoll, et al. Health Affairs, Vol. 28 (2009).

* "Can Soft Drink Taxes Reduce Population Weight?" by Jason Fletcher, David Frisvold, and Nathan Tefft. Contemporary Economic Policy, Vol. 28 (2010).

* "Challenges in the Study of the Causation of Obesity," by Thorkild I. A. Sorensen. Proceedings of the Nutrition Society, Vol. 68 (2009).

* "Does the U.S. Food Stamp Program Contribute to Adult Weight Gain?" by Jay L. Zagorskya and Patricia K. Smith. Economics and Human Biology, Vol. 7, No. 2 (2009).

* "Effects of Soft Drink Consumption on Nutrition and Health: A Systematic Review and Meta-Analysis," by Lenny R. Vartanian, Marlene B. Schartz, and Kelly D. Brownell. American Journal of Public Health, Vol. 97, No. 4 (2007).

* "Evidence for a Strong Genetic Influence on Childhood Adiposity Despite the Force of Obesogenic Environment," by Jane Wardell, Susan Carnell, Claire M. A. Hayworth, and Robert Plomin. American Journal of Clinical Nutrition, Vol. 87 (2008).

* "Food Insecurity Is Positively Related to Overweight in Women," by M. Townsend, J. Peerson, B. Love, et al. Journal of Nutrition, Vol. 131 (2001).

* "Food Stamp Program Participation Is Positively Related to Obesity in Low Income Women," by Diane Gibson. Journal of Nutrition, Vol. 133 (2003).

* "Increasing Consumption of Sugar-Sweetened Beverages among U.S. Adults: 1988-1994 to 1999-2004," by Sara N. Bleich, Y. Claire Wang, Youfa Wang, and Steven L. Gormaker. American Journal of Clinical Nutrition, Vol. 89 (2009).

* "Intake of Sugar-Sweetened Beverages and Weight Gain: a Systematic Review," by Vasanti S. Malik, Matthias B. Schulze, and Frank B. Hu. American Journal of Clinical Nutrition, Vol. 84 (2006).

* "Lifestyle Interventions for the Treatment of Class III Obesity: A Primary Target for Nutrition Medicine in the Obesity Epidemic," by George L. Blackburn, Samuel Wollner, and Steven B. Heymsfield. American Journal of Clinical Nutrition, Vol. 91 (2010).

* "Lifetime Medical Costs of Obesity: Prevention No Cure for Increasing Health Expenditures," by Pieter H. M. van Baal, Johan J. Polder, G. Ardine de Wit, Rudolf T. Hoogenveen, Talitha L. Feenstra, Hendriek C. Boshuizen, Peter M Engelfriet, and Werner B. F. Brouwer. PLos Medicine, Vol. 5, No. 2 (2008).

* "Mandatory Menu Labeling in One Fast-Food Chain in King County, Washington," by Eric A. Finkelstein, Kiersten L. Strombot, Nadine L. Chan, mid James Krieger. American Journal of Preventive Medicine, Vol. 40, No. 2 (2011).

* "Neighborhood Environments: Disparities in Access to Healthy Foods in the U.S.," by Nicole I. Larson, Mary Z Storey, and Melissa C. Nelson. American Journal of Preventive Medicine, Vol. 36, No. 1 (2009).

* Nudge: Improving Decisions about Health, Wealth, and Happiness, by Richard H. Thaler and Cass R. Sunstein. Yale University Press, 2008.

* "Obesity Prevalence and the Local Food Environment," by Kimberly B. Morland and Kelly R. Evenson. Health and Place, Vol. 15 (2009).

* "Obesity, Poverty, and Participation in Nutrition Assistance Programs," FSP-04-PO, by P. Linz, M. Lee, and L. Bell. U.S. Department of Agriculture; Food and Nutrition Service; Office of Analysis, Nutrition, and Evaluation; 2004.

* "Prevalence and Trends in Obesity among U.S. Adults, 1999-2008," by Katherine M. Flegal, Margaret D. Carroll, Cynthia L. Ogen, and Lester R. Curtain. JAMA, Vol. 303, No. 3 (2010).

* "Prevalence of High Body Mass Index in U.S. Children and Adolescents, 2007-2008," by Cynthia L. Ogden, Margaret D. Carroll, Lester R. Curtin, Molly M. Lamb, and Katherine M. Flegal. JAMA, Vol. 303 (2010).

* "Sin Taxes: Do Heterogeneous Responses Undercut Their Value?" by Padmaja Ayyagari, Partha Deb, Jason Fletcher, William T. Gallo, and Jody L. Sindelar. July 2009.

* "Taxes, Cigarette Consumption, and Smoking Intensity," by J. Adda and F. Cornaglia. American Economic Review, Vol. 96 (2006).

* "The Causes, Prevalence, and Treatment of Obesity Revisited in 2009: What Have We Learned So Far?" by Caroline M. Apovian. American Journal of Nutrition, Vol. 91 (2010).

* "The Economics of Childhood Obesity," by John Cawley. Health Affairs, Vol. 29, No. 3 (2010).

* "The Effect of Obesity on Health Outcomes," by John B. Dixon. Molecular and Cellular Endocrinology, Vol. 316 (2010).

* "The Effects of Cigarette Costs on BMI and Obesity," by Charles L. Baum. Health Economics, Vol. 18 (2009).

* "The Effects of Higher Cigarette Prices on Tar and Nicotine Consumption in a Cohort of Adult Smokers," by M. C. Farrelly, C. T. Nimsch, A. Hyland, et al. Health Economics, Vol. 13 (2004).

* "The Growth of Obesity and Technological Change," by Darius Lakdawalla and Tomas Philipson. Economics and Human Biology, Vol. 7 (2009).

* "The incidence of the Healthcare Costs of Obesity," by Jayanta Bhattacharya and M. Kate Bundorf. Journal of Health Economics, Vol. 28, No. 3 (2009).

* "The Leveling Off of the Obesity Epidemic Since the Year 1999--a Review of Evidence and Perspectives," by B. Rokholm, J. L. Baker, and T. I. A. Sorensen. Obesity Reviews, Vol. 11 (2010).

* "The Long-run Growth in Obesity as a Function of Technological Change," by Tomas J. Philipson and Richard A. Posner. Perspectives in Biology and Medicine, Vol. 46, No. 3 (2003).

* "The Medical Care Costs of Obesity: An Instrumental Variables Approach," by John Cawley and Chad Meyerhoefer. 2010.

* "The Other Ex-ante Moral Hazard in Health," by Jay Bhattacharya and Mikko Packalen. March 2008.

* "The Public Health and Economic Benefits of Taxing Sugar-Sweetened Beverages," by Kelly D. Brownell, Thomas Farley, Walter Willett, Barry Popkin, Frank Chaloupka, Joseph Thompson, and David S. Ludwig. New England Journal of Medicine, October 15, 2009.

* "The Skinny on Big Box Retailing: Wal-Mart, Warehouse Clubs, and Obesity," by Charles Courtamanche and Art Carden. October 31, 2008.

* "Understanding Overeating and Obesity," by Christopher Ruhm. 2010.

* "Would Soda Taxes Really Yield Health Benefits?" by Michael L. Marlow and Alden F. Shiers. Regulation, Vol. 33, No. 3 (2010).

BY MICHAEL L. MARLOW AND ALDEN go SHIERS

California Polytechnic State University

MICHAEL L. MARLOW and ALDEN F. SHIERS are professors of economics at California Polytechnic State University in San Luis Obispo.

TABLE 1
Changing Obesity Rates

By state, for years 1995 and 2009
                                   %
                 1995   2009   Change

Alabama            19     32       69
Alaska             20     25       28
Arizona            13     26       95
Arkansas           18     32       80
California         15     26       69
Colorado           10     19       88
Connecticut        13     21       68
Delaware           17     28       61
Florida            17     27       54
Georgia            13     28      108
Hawaii             11     23      112
Idaho              14     25       77
Illinois           17     27       64
Indiana            20     30       49
Iowa               18     29       63
Kansas             16     29       81
Kentucky           17     32       92
Louisiana          18     34       92
Maine              14     26       87
Maryland           16     27       64
Massachusetts      12     22       86
Michigan           18     30       66
Minnesota          15     25       66
Mississippi        20     35       82
Missouri           19     31       62
Montana            13     24       77
Nebraska           16     28       72
Nevada             13     26       98
New Hampshire      15     26       74
New Jersey         15     24       65
New Mexico         13     26       97
New York           14     25       77
North Carolina     17     30       78
North Dakota       16     28       73
Ohio               18     30       70
Oklahoma           14     32      137
Oregon             15     24       55
Pennsylvania       16     28       71
Rhode Island       13     25       89
South Carolina     17     30       80
South Dakota       14     30      118
Tennessee          18     33       79
Texas              16     30       86
Utah               15     24       58
Vermont            15     23       60
Virginia           16     26       62
Washington         14     27       94
West Virginia      18     32       73
Wisconsin          16     29       83
Wyoming            14     25      178

Source. BFRSS data Note: Utah's data begin in 1998.