Featured Articles

Stop the Sugar: Policy Considerations for an Effective Soda Tax

Posted on July 4, 2016 by Pratt, Elijah

Elijah PrattElijah Pratt is a student at the Brigham Young University J. Reuben Clark Law School.

In this viewpoint, Pratt discusses economic considerations for implementation of an optimal soda tax, including calculating externalities and minimizing adoption of substitute goods. He argues that even though it will be difficult to implement an optimal tax, policymakers should consider soda taxes as they are a step in the right direction to correct America's obesity epidemic.

This paper was selected for publication as part of Tax Analysts' student paper competition.

* * * * *

I. Obesity Overview

A. Prevalence and Effects

Over the past half-century, obesity has become a significant problem. From 1960 to 2008, obesity rose from about 13 percent to 34 percent of U.S. adults, and extreme obesity increased from less than 1 percent to 6 percent.1 Obesity is more prevalent among racial minority groups and women. In 2008, obesity rates were 36 percent for Mexican-American men, 45 percent for Mexican-American women, 37 percent for black men, and 50 percent for black women.2 Just last year, the prevalence of obesity in 21 states was more than 30 percent of the population, and no state had an obesity prevalence of less than 20 percent.3

These statistics provide only a glimpse of the real harm that obesity is causing. Obesity is not itself a disease, but an indicator of excess body fat.4 However, obesity is a serious risk factor for several diseases and health conditions, including high blood pressure, high cholesterol, Type 2 diabetes, coronary heart disease, stroke, gallbladder disease, osteoarthritis, and cancer.5 One systematic review of obesity finds a reciprocal link between obesity and depression.6 The fact that 1 in 3 adults is at an increased risk of these obesity-related diseases and conditions should not be taken lightly.

Childhood obesity poses even greater potential harm to society. In 2011 the obesity rate for children in the United States age 2 to 19 was about 17 percent, or 12.7 million children.7 Like adults, the prevalence of obesity in children is greater in racial minority groups.8 Obese children are likely to stay obese as adults and are more likely to develop obesity-related diseases like Type 2 diabetes and cardiovascular diseases at a younger age.9 Former U.S. Surgeon General Richard Carmona characterized the alarming severity of childhood obesity by noting that because of "the increasing rates of obesity . . . we may see the first generation that will be less healthy and have a shorter life expectancy than their parents."10

The rise in obesity rates results primarily from consuming more calories.11 People gain weight if there is an increase in caloric intake or a reduction in caloric expenditure, but increased caloric intake is the more likely culprit.12 This increase in caloric intake among Americans is due to technological advancements that make food cheaper and easier to prepare, including vacuum packing, improved preservatives, deep freezing, artificial flavors, and microwaves.13 Technological advances have allowed for increased consumption of carbohydrates, including soft drinks and juices, which account for much of the increased caloric intake.14 Kelly D. Brownell of Yale University said sugar-sweetened beverages "may be the single largest driver of the obesity epidemic."15 Further, consumption of sugar-sweetened beverages has been found to greatly increase childhood obesity rates.16

Obesity imposes both medical and nonmedical costs on society. A review of studies on the medical costs of obesity "provide evidence that the aggregate annual obesity-attributable medical costs in the United States are between 5 percent and 7 percent of annual health care expenditures."17 In 2013, U.S. healthcare expenditures reached $2.9 trillion.18 Hence, in that year, obesity imposed costs of up to $200 billion on the U.S. healthcare system. Obesity creates nonmedical costs as well. Obese individuals are more likely to be absent from work, work in relatively low-paying positions, report long-term unemployment, and live in poverty.19 Obese women have been found to earn lower wages than their normal-weight counterparts.20 Finally, perhaps the most significant cost of obesity is the weight-loss struggle of the 55 million Americans serviced by the $40 billion-per-year diet industry.21 Each one of those individuals lives with the daily toll of obesity.

B. Efforts to Fight Obesity


The good news is that obesity, along with its related diseases, is largely preventable.22 The primary cause of obesity is a caloric surplus, or more calories consumed than expended. Therefore, programs that succeed in getting obese individuals and at-risk obese individuals to reduce or reverse their caloric surplus will reduce obesity. Many programs and policies, both federal and local, have been proposed or implemented to achieve this aim. The following survey of anti-obesity programs and policies is non-exhaustive, but illustrates current anti-obesity efforts in the United States.


Because of the damaging consequences of childhood obesity, many of the policies and programs are directed toward schools and children. Some argue that schools need to incorporate nutrition education in their curriculum.23 This education could include not only physical education classes but also nutritional posters in the cafeteria, in-school fitness marketing, and a general attitude of healthy living in schools.24 In conjunction with nutrition education, there is movement toward school lunch menu reform. In the 2011-2012 school year, the Los Angeles Unified School District led the way in switching from processed foods to locally grown and organic foods.25 In 2003 Arkansas enacted a law prohibiting vending machines containing junk foods and beverages in public schools.26 The same year, a new California law required the replacement of carbonated beverages with healthier options such as milk, water, and juice.27 Some states have introduced body mass index reporting laws, requiring or encouraging schools to monitor students' BMI and then report it to their parents or the legislature.28 Finally, some argue for greater standards in schools involving physical education and recess.29 The standards across states for physical education vary, and a growing number of states allow children to opt out of physical education classes.30

Anti-childhood obesity policies and programs exist outside the classroom and schoolyard. There has been a battle over the marketing of unhealthy products to children, with many experts believing the practice is a major cause of childhood obesity. However, the U.S. Supreme Court's commercial speech doctrine restricts limitations on advertising.31 Consequently, marketing to children is primarily self-regulated by the food companies.32 One of the most notable recent efforts to combat childhood obesity has been first lady Michelle Obama's "Let's Move" campaign, launched in 2010. Let's Move created the White House Task Force on Childhood Obesity, assigned to review child nutrition and physical activity policies and programs.33 Let's Move takes an all-around approach to childhood health, focusing on empowering parents and increasing healthy food and physical activity in schools.34

Many government policies seek to combat obesity. One of the least intrusive policies is a disclosure requirement for the caloric content of food at restaurants and vending machines.35 In 2008 the Los Angeles City Council resorted to zoning as a means of addressing obesity by banning new fast-food restaurants in South Los Angeles.36 Other cities have banned street vendors near schools or encouraged healthy markets to move into underserved neighborhoods by exempting them from zoning requirements.37 In 2012 then-New York City Mayor Michael Bloomberg proposed a ban on all sugar-sweetened beverages larger than 16 ounces in food establishments regulated by the Board of Health.38 However, the New York Court of Appeals struck down the proposal as "exceeding the scope of [the board's] regulatory authority."39 The federal government has recently raised insurance-based policy proposals for fighting obesity. Medicare has covered bariatric surgery since 2006.40 There are current proposals for coverage of weight-loss drugs and expanded coverage for behavioral counseling.41 Finally, many criticize the government's historical subsidization of commodities such as corn, the primary ingredient in high-fructose corn syrup, found in sweetened beverages and other foods.42 These critics urge a change to the subsidization, which would reduce the price of healthy foods.43

C. Tax on Sugar-Sweetened Beverages


The above policies and programs range from those that merely inform consumers to direct government involvement in schools, advertising, and the food industry. Some argue that the government should not involve itself in the issue at all because products and prices will change as people demand healthier food.44 However, many argue that the obesity crisis in our country is the consequence of a market failure, so direct government involvement is needed.45 A tax on sugar-sweetened beverages is a relatively recent proposal. Proponents of the tax argue that it will increase the prices of beverages linked to obesity and that as the prices increase, the quantity demanded by consumers will decrease. This reduction in the consumption of sugar-sweetened beverages will, so the argument goes, translate into lower obesity rates and better health outcomes. Further, a tax on sugar-sweetened beverages could raise revenue to fund other anti-obesity programs.


A tax on sugar-sweetened beverages is most often compared to a tobacco tax. In 1921 Iowa became the first state to tax cigarettes; by 1969 all 50 states had imposed a cigarette tax.46 Federal and state governments have continued to increase the amount of cigarette excise taxes, and some municipalities have imposed their own additional taxes.47 While cigarette taxes originally began as a means of raising revenue, many countries have increased them with the explicit purpose of decreasing cigarette consumption.48 Frank J. Chaloupka concludes that cigarette taxes are "very effective" in reducing tobacco consumption.49 He attributes the effectiveness of these taxes to the reduction in both frequency and prevalence of use.50

Several countries have already imposed taxes on sugar-sweetened beverages, including Canada, Belgium, Denmark, France, Germany, Hungary, Thailand, and the Philippines.51 Mexico's one-peso-per-liter tax, implemented in 2013, is projected to reduce consumption by 10 to 12 percent and reduce obesity rates within the country.52

The United States does not have a federal tax on sugar-sweetened beverages. In 2009 the Senate Finance Committee proposed a tax to raise revenue to fund healthcare reform.53 However, the tax failed after food industry groups spent more than $24 million lobbying against it.54 As of 2009, 33 states had a sales tax on soft drinks.55 However, the mean tax rate among states of 5.2 percent is too small to affect consumption.56 In 2010 the governor of New York proposed a higher 1-cent-per-ounce tax on non-diet sodas sold within the state.57 Like the proposed federal soda tax, the proposal was swiftly attacked by the American Beverage Association, and it was quickly abandoned.58

In November 2014, voters passed a ballot measure to institute the first citywide soda tax in Berkeley, California.59 The tax placed a 1-cent-per-ounce tax on the distributors of sugar-sweetened beverages such as soda (not diet soda), energy drinks, and sugary coffee syrups.60 Despite the American Beverage Association's $2 million effort to defeat the measure, the tax passed by at least a 3-to-1 margin.61

D. Roadmap


Within the legal, health, and economic literature, numerous papers have emerged arguing for or against a soda tax.62 This paper only touches lightly on the moral arguments regarding the tax.63 Rather, this paper combines economic theory and research with health literature to explore a potential legislator's practical considerations regarding the effectiveness of a soda tax. This paper also serves as a theoretical framework for future empirical research on soda taxes.


Parts II and III provide a background for Part IV. Part II explains the economic rationales underlying the soda tax. A clear understanding of these rationales is essential in determining whether a soda tax is effective. Part III discusses the economic principles of elasticity and substitution, critical topics to understanding a tax scheme. Part IV uses economic, health, and social science research to discuss several considerations for the implementation of a soda tax.

II. Economic Rationales for a Soda Tax

The rationales underlying a soda tax are critical in determining whether the tax is effective. In order for a soda tax to be effective, it must do more than simply influence consumer behavior in some way. The tax must address the specific underlying economic costs associated with soda consumption. Economists have advanced several rationales for a soda tax, each with implications in the context of soda and related products. These rationales fall within two general arguments: an individual's soda consumption imposes costs on others, and individuals systematically make sub-optimal soda consumption decisions for themselves. The former is referred to generally as a negative externality, and is discussed first. The latter comes in several forms, and is discussed second. A discussion of each rationale sets the groundwork for a fuller soda tax analysis in Part IV.

A. Negative Externality


The primary rationale favoring a soda tax is that soda consumption creates a negative externality.64 Ideally, individuals face the costs and benefits associated with their consumption decisions. Suppose a can of soda costs $1 and the soda gives Jim $1.25 worth of benefit. In other words, Jim is willing to pay up to $1.25 for the soda. In this case, Jim will buy the soda, and he has made a net gain of 25 cents. However, suppose that in drinking the soda, Jim imposes a 50-cent cost on Sally. In this case, Jim would still drink the soda because his benefit exceeds his cost and because he does not consider the cost imposed on Sally. But there is a net loss to everyone of 25 cents because of the additional external cost imposed on Sally. This additional external cost is called a negative externality. Generally, negative externalities arise when an individual does not face all of the costs associated with a decision.


The solution to the problem is to make Jim bear (internalize) the additional cost imposed on Sally. A tax of 50 cents on Jim, the amount of the negative externality, would artificially cause Jim to internalize the 50-cent negative externality. Then Jim would not buy the soda and the world would be better off. This type of tax placed on an activity that creates negative externalities is called a Pigouvian tax.65 It is important to note that for the tax to optimally correct Jim's behavior, the tax must equal the external cost.

The key part of the hypothetical is that a 50-cent tax was effectively imposed on Sally. So how does soda consumption impose costs on others? First, many argue that soda consumption leads to obesity, and the healthcare costs of obesity are imposed on taxpayers through Medicare and Medicaid. Another argument is that an obese population within an insurance pool increases insurance premiums for the non-obese if healthcare insurers cannot charge higher premiums to obese individuals.66 Part IV will explore the soundness of these arguments.

In contrast, the following rationales argue that even if an individual's soda consumption does not create negative externalities so that costs of soda consumption are fully internalized, the costs are not properly considered, and inefficient decisions result.

B. Imperfect Information


A tax may be warranted if information about the costs of consumption are hidden to the consumer. If Jim believes that the cost of his soda consumption, from which he derives a $1.25 benefit, is $1, but in reality Jim faces an additional hidden cost of 30 cents, it would not be in Jim's best interest to consume the soda. However, if policymakers know about these hidden costs, they might impose a 30-cent tax on Jim's soda consumption as an indirect means of informing Jim of the additional hidden costs. With a tax of 30 cents, Jim would choose not to buy the soda, not because he has learned about the hidden costs specifically, but because he considers the tax. Hence, the tax causes Jim to behave as if he knew about the hidden costs. Part IV explores potentially hidden costs of soda consumption.


C. Improperly Discounted Future Costs


Economists tend to assume that consumers are fully rational decision-makers. However, behavioral economists question that assumption.67 Even with perfect information about soda and its costs, people may still irrationally choose to overconsume by improperly discounting the future costs of soda consumption. Take smoking as an example: "Although many smokers express the desire to quit smoking in the future and expect that they will quit smoking within five years, only a small percentage of those smokers actually quit smoking."68 While smokers are aware of the future costs associated with smoking, they may underestimate the present value of those future costs. They act accordingly by not quitting smoking when it would be optimal to do so.


Suppose again that Jim values a $1 soda in 2015 at $1.25, but will incur an additional 75-cent health cost in 2020 that has a present value of 50 cents. Again, it would not be in Jim's interest to buy the soda. Even if Jim is fully aware of his future costs, he may not give them as much weight as imminent costs and benefits. Hence, Jim might irrationally discount the 75-cent future cost to 5 cents instead of 50 cents. Therefore, in 2015, Jim would perceive the soda's cost at only $1.05 and would improperly choose to buy the soda. A soda tax of 45 cents would make Jim behave as if the cost of consuming a soda were $1.50, and Jim would correctly choose not to consume the soda. Part IV discusses irrationally discounted future costs in the context of soda consumption.

III. Elasticity and Substitution

The purpose of a soda tax is to motivate consumers to forgo harmful drinks in favor of healthier choices. Economists use the principle of elasticity to describe the substitution of one product for another. Hence, understanding elasticity is crucial in analyzing a soda tax.

Elasticity is a measure of how much consumers will change their consumption when the price of that good changes. In other words, how "elastic" is the change in consumption in response to a price change? Generally, elasticity equals the percent change in the quantity of a good demanded divided by the percent change in the price of that good. For example, suppose that the price of soda increases by 10 percent, causing consumers to decrease soda consumption by 20 percent. Then the elasticity of demand for soda equals -2 (-20 percent/10 percent = -2).69 But if soda were more elastic and soda consumption decreased by 30 percent, elasticity of demand for soda would equal -3 (-30 percent/10 percent = -3). Hence, more elastic goods have more negative value, and more inelastic goods have less negative value. Technically, a good with an elasticity of demand less than -1 is elastic, and a good with an elasticity of demand between -1 and 0 is inelastic.70 This becomes important in Part IV, where the elasticities of specific goods are discussed.

Elasticity can also be thought of as a measure of the substitutability of a good. The more good substitutes a product has, the more elastic it is. If the price of an elastic good increases, the consumer can easily switch to a substitute good, and the consumption of the original good will decrease sharply. This large decrease in consumption corresponds to a high elasticity of demand. For example, Coca-Cola is arguably highly elastic because many other sodas and other beverages are good substitutes. If the price of Coke increases, consumers will switch to these other substitute goods. On the other hand, an inelastic good has few good substitutes. If the price of an inelastic good increases, the consumer cannot easily substitute another good, and the consumption of the original good will decrease slightly. This decrease in consumption corresponds to a low elasticity of demand. Gasoline is arguably inelastic because there seem to be few good substitutes. Cars won't run on Pepsi, Diet Coke, or Dr Pepper (at least not for now) -- only gasoline will do. People could switch to riding public transportation, biking, or walking, but that isn't a viable substitute in many cases. So if the price of gasoline increases, people will likely only decrease their consumption of gasoline by a little and will be forced to pay the higher price.71

Also note that the elasticity of a good depends on how broadly or narrowly the good is defined. More broadly defined goods are more inelastic, and vice versa. For example, Coke specifically is likely more elastic than soda generally because Pepsi, Dr Pepper, and Sprite are close substitutes for Coke. However, milk, water, and coffee are likely poorer substitutes for soda. Hence, the elasticity of Coke is greater than that of soda, which is greater than that of all drinks, which is greater than that of all food and drinks. If the price of all food and drinks increases, the consumption of all food and drinks will decrease little, because the only alternative is death.

Finally, people will substitute one good for another not just because there is a change in price of one good, but also because there is a relative change in the prices of goods. To illustrate with a simple economic model, suppose Bob faces the daunting task of buying apples and bananas. And assume that Bob values apples and bananas equally and that apples and bananas each initially cost $1. In this case, if Bob buys one apple, he forfeits one banana. However, suppose that the price of apples increases to $2 a piece and the price of bananas remains $1. Then there is a relative increase in the price of apples compared with bananas. Now if Bob buys an apple, he forfeits two bananas. Hence, Bob will buy a greater ratio of bananas to apples. Finally, if the prices of both apples and bananas increase to $2 a piece, the relative price of apples to bananas has not changed. If Bob buys one apple, he still only forfeits one banana. Hence, Bob will not change the ratio of apples to bananas that he buys. Yet because both apples and bananas are more expensive, he may proportionally decrease the number of both apples and bananas bought.

IV. Considerations for Implementation of a Soda Tax

What questions does a legislator need to consider regarding how to create a soda tax that effectively fights obesity? This section analyzes these questions and, wherever possible, gives answers based on available health, economic, and social science research. However, many questions have no clear answers. Hence, this section also provides a framework for future soda tax researchers.

As a preliminary matter, if the purpose of a tax is to regulate behavior, how do we determine whether the tax is effective? We could impose a steep tax on factories to reduce pollution, which may result in all factories shutting down. But because we need factories to produce goods, we might not consider this tax effective, even if it causes a large reduction in pollution. Conversely, a tax on cigarettes that is too low to cause people to reduce smoking might not be effective either. Finally, we might impose a tax on bicyclists to reduce their activity, but if bike riding isn't actually a problem, the tax probably isn't considered effective. Therefore, an effective tax is one that addresses the tax's rationales. The tax on factories is ineffective because it is much too severe for the rationale. The cigarette tax is ineffective because it isn't severe enough to address its rationale. The bicyclist tax is ineffective because there is no rationale for it at all. The consideration of whether a tax is effective is like a scrutiny analysis for constitutional purposes -- there must be a rationale for the tax, and the tax must be tailored to that rationale.

A. Does Soda Consumption Create a Negative Externality?


The primary consideration regarding a soda tax is whether soda consumption actually creates a negative externality. The negative externality rationale for a soda tax deserves special consideration since it is the most commonly cited ground for a soda tax.72 Also, the negative externality rationale is distinct in that the government is protecting someone other than the soda drinker -- namely, taxpayers and those who buy health insurance -- from the consequences of soda consumption. However, when government involves the imperfect information and discounted future costs rationales, it is protecting the soda drinkers themselves. If the government is more justified in protecting third parties, the negative externality rationale is the strongest. Hence, it deserves its own analysis.


As discussed above, a negative externality occurs when individual A's decision imposes costs on individual B, but individual A does not bear these costs, and, therefore, does not consider them. For soda consumption, the negative externality argument is that soda consumption causes obesity. Obesity then imposes costs on third parties by either of two means. First, obese individuals enrolled in Medicare and Medicaid impose costs on taxpayers who pay for increased obesity-related healthcare costs. Second, obese individuals with health insurance impose costs on the non-obese through higher premiums because insurers cannot price differentiate between the obese and non-obese.73

The first premise of both arguments is that soda consumption causes obesity. But what should be considered soda for policy purposes? The remedial tax is typically referred to as either a "soda tax" or a "tax on sugar-sweetened beverages." But if the purpose of a soda tax is to reduce obesity-related diseases, the tax should include their close substitutes, which also contribute to obesity. If only soda is taxed, individuals may switch to equally harmful substitutes for soda (perhaps milkshakes?) and the tax would lose some of its ability to fight obesity. Therefore, an assessment of whether soda consumption causes obesity raises the question of how narrowly or broadly soda should be defined. This question is considered at length in the subsection below discussing the proper breadth for a soda tax. For now, a narrow definition of soda suggests that soda contributes comparatively less to obesity than if soda were broadly defined.

  • 1. Does Obesity Create a Negative Externality Through
    Medicare and Medicaid?


The question whether obesity causes a negative externality through Medicare and Medicaid raises other important questions. First, since the argument is premised on costs being transferred from obese Medicare and Medicaid enrollees, what fraction of these enrollees are obese? Second, how much of an effect does obesity have on Medicare and Medicaid costs? Third, is coverage for government-financed insurance programs expanding? Finally, how much of these costs is due to soda consumption?

On a national scale, it is clear that some fraction of Medicare and Medicaid enrollees are obese. Medicare and Medicaid pick up the tab for half of all obesity-related healthcare costs.74 However, some states have higher obesity rates than others,75 and the obesity prevalence among Medicare and Medicaid enrollees dictates the magnitude of the negative externality. A higher obesity prevalence among Medicare and Medicaid enrollees suggests a greater negative externality. If a soda tax is enacted by a city, that city should consider its local obesity prevalence among Medicare and Medicaid enrollees.

Second, if obesity does not correspond to higher healthcare costs, there can be no externality. In fact, obesity-related healthcare costs directly correspond to the magnitude of the externality. One study estimates that obesity is associated with a 36 percent increase in inpatient and outpatient spending.76 Compare this with a 21 percent increase in inpatient and outpatient spending for current smokers.77

Third, policymakers should remain aware of how expansive coverage for government-financed health insurance programs is becoming for the obese. Increased Medicare and Medicaid coverage for the obese imposes a greater negative externality on taxpayers. Obesity certainly devastates those who live with its consequences, but while increased coverage may help the obese and their families, it cannot be ignored that the costs are inevitably transferred to taxpayers.

While soda may cause obesity, the obesity costs resulting from soda consumption may differ from other causes of obesity. Soda may cause not only obesity, but a type of obesity that has more severe health consequences than that caused by a lack of exercise. More research is needed in this area.

A Pigouvian tax should equal the amount of the negative externality. Hence, research should be directed toward answering the following question: If Jim drinks one can of soda, how much do taxpayers have to pay? This question is probably impossible to answer with certainty, but it is a helpful thought experiment in determining the proper amount of the tax. Together, all of these questions form a single question, which is perhaps merely theoretical: How much of obesity-related costs covered by Medicare or Medicaid are attributable to soda consumption?

  • 2. Does Obesity Create a Negative Externality Through
    Higher Insurance Premiums?


The obese could impose costs on the non-obese through higher insurance premiums. The obese face higher medical costs than the non-obese. If both groups must pay the same insurance premium, insurers must inevitably increase insurance premiums for the non-obese in order to stay profitable. Hence, the non-obese are effectively paying for the obesity-causing decisions of the obese.78 However, this negative externality is even worse than it appears. Because the non-obese are essentially subsidizing the physical inactivity and overeating of the obese, the obese are predicted to exercise less and eat more than if they were not subsidized. In other words, charging both groups the same premium creates a moral hazard problem that further encourages obesity-causing behaviors. So the non-obese are paying an increased premium that is further amplified by the obese population's lack of incentive to avoid obesity-causing behaviors.79

Therefore, the key question for policymakers is whether insurers can charge the obese and non-obese different insurance premiums. For example, starting in 2014, the Affordable Care Act prohibited insurers from denying people coverage or charging them more for a preexisting condition.80 The ACA's prohibition on preexisting conditions does not include obesity; insurers can still charge the obese higher premiums.81 However, some say that the ACA should not exempt obesity, arguing that many low-income individuals in urban areas lack the resources to maintain a healthy weight.82 While this policy may grant some people health insurance coverage, economic theory suggests that it will hinder incentives to exercise and eat healthy, further contributing to the obesity negative externality. But for now, because obesity is exempt from the ACA's prohibition on preexisting conditions, the ACA is not the source of a negative externality via higher insurance premiums. Nonetheless, policymakers should still look for other possible mechanisms that might force insurers to charge the obese and non-obese higher premiums.

B. What Products Should Be Included in the Tax?


Perhaps the most difficult question for policymakers is what the soda tax should cover. Should it be limited to soda or be more inclusive? This section does not answer these questions. Instead, it provides an economic framework for thinking about the breadth of a soda tax.


  • 1. Breadth Issues


A policymaker confronts at least three potential breadth problems in constructing a soda tax. First, the taxed items may be too inelastic. A soda tax may increase the price consumers pay for soda, but they may only have a small reduction in soda consumption. Therefore, consumers may be paying more for soda without any improvement in health. This problem relates to the principal criticism of a soda tax: that it is regressive. Soda taxes have been called regressive because they constitute a higher proportion of a poor person's income than a wealthy person's income.83 If, however, soda is perfectly elastic (any increase in price results in zero consumption), then a soda tax cannot be regressive because consumers are not drinking any soda. But the more inelastic soda consumption becomes, the more tax revenue is collected. If the poor and wealthy pay an equal tax amount, the tax is regressive because the tax is a higher proportion of the income of the less affluent than of the wealthy. Ideally, soda would be an elastic good for the poor but an inelastic good for the wealthy. This way, the wealthy pay the taxes, not the poor. One study estimating elasticity of demand for sugar-sweetened beverages in Mexico suggests that this is the case.84 This evidence supports the view that soda taxes are not regressive.

Second, the taxed items may be elastic, but people substitute from one unhealthy option to another. It is not enough to tax an unhealthy item. A properly constructed tax will lead to a change in consumer behavior that improves health outcomes. For example, studies have found that an increase in cigarette taxes could significantly increase the use of smokeless tobacco.85 Similarly, a tax on Pepsi may simply lead to increased consumption of Coke, which may be equally harmful. A tax on all soda products may increase consumption of other harmful foods or drinks, which may be even more harmful than soda. An effective soda tax must consider not only the harm of the taxed items but also of potential substitute items.

Third, a soda tax may theoretically become too broad. In an effort to tax unhealthy products and their unhealthy substitutes, a tax on soda could quickly become a general food tax. However, a broad tax may not create a change in relative prices sufficient for substitution. A tax on Coke may cause people to switch to Pepsi because there has been a change in relative prices between the two. However, a tax on both Coke and Pepsi may cause people not to substitute to a healthy drink, such as water, but drink just a little less of both Coke and Pepsi and more of another unhealthy drink, like Dr Pepper. People may simply be unwilling to switch from soda to water. Put differently, taxing more items is like taxing a more broadly defined market -- which is more inelastic. To maintain a change in price relativity, policymakers may consider taxing the worst products more heavily than the healthier-but-still-unhealthy substitute products. Hence, policymakers should consider not only which products are unhealthy but also which are the most unhealthy.

  • 2. An Evaluation of Current Soda Tax Proposals


In considering what products a soda tax will cover, a policymaker must evaluate both soda and potential substitutes for soda. In particular, these products' health consequences and elasticity deserve careful attention. We need to know the elasticity of narrowly defined products (for example, "carbonated sugar-sweetened beverages") to determine whether people will stop drinking them if their price is sufficiently increased. We also need to know how their elasticity changes as the product is defined more broadly: People may switch from soda to high-sugar juice, but will they switch from both soda and high-sugar juice to water?

Next, a policymaker must consider whether any economic rationales for a soda tax give special reason for concern with any of these products. For example, is it likely that information regarding the sugar content of juice is hidden to consumers? Do consumers fully understand the health implications of the sugar content in juice? If any of these rationales apply, it is more likely that people are overconsuming these products, and the products become a more suitable candidate for the soda tax.

A tax designed to reduce sugar-caused obesity may be too narrow if it arbitrarily taxes only sugar-sweetened beverages, but not other sugar-sweetened foods. Current proposals for a soda tax and studies evaluating a soda tax typically focus on beverages, not foods.86 One study refers to a tax on sugar-sweetened beverages as a tax "on soft or soda drinks, and other carbonated and un-carbonated drinks, including sports and energy drinks, sweetened with sugar, corn syrup or other caloric sweeteners."87 Berkeley's tax on "sugar-sweetened beverages" explicitly includes, but is not limited to, "all drinks and beverages commonly referred to as soda, pop, cola, soft drinks, sports drinks, energy drinks, sweetened ice teas, or any other common names that are derivations thereof."88 The Berkeley tax also covers beverages with added caloric sweeteners that could be mixed with "water, ice, powder, coffee, tea, fruit juice, vegetable juice, or carbonation or other gas."89 Note that this tax does not include diet sodas or other drinks with a noncaloric sweetener. Also, milk is explicitly excluded from the tax.90

Some have argued that "sweetened beverages . . . may be the single largest driver of the obesity epidemic."91 Sugar has a particular capacity to induce caloric overconsumption. Sugar is overconsumed because it is not satiating -- it doesn't make you full -- and people do not compensate by eating less of other foods.92 A 12-ounce can of Pepsi has 150 calories from 41 grams of some form of sugar, including high-fructose corn syrup.93 Calories from multiple cans of soda add up quickly throughout the day without satiating the consumer's hunger. Some types of sugar, such as fructose and high-fructose corn syrup, may even cause more hunger.94

A tax on sugar-sweetened beverages has two qualifiers: The product must have sugar and be a beverage. However, people may substitute taxed sugar sweetened beverages for potentially equally harmful untaxed non-sugar or non-beverage products. Hence, understanding possible mechanisms for how consumers may substitute from taxed to nontaxed products is helpful in determining the proper breadth of the tax.

One possible substitution mechanism is that consumers will simply switch from taxed sugar-sweetened beverages for the relatively cheaper sugarless-sweetened beverages. Diet sodas are the most obvious example of sugarless noncaloric beverages. Cawley and Frisvold found that after the soda tax in Berkeley was implemented, Diet Coke and Diet Pepsi prices rose.95 They explain this may be due to people switching from regular soda to diet soda, driving up the price of diet soda.96 The critical inquiry, which requires more research, is whether drinking diet soda leads to negative health consequences, even without the sugar. Some scientists believe that it does. For example, as part of the San Antonio Longitudinal Study of Aging, the authors concluded that a "striking . . . relationship was observed between increasing [diet soda intake] and escalating abdominal obesity, which represents a potential pathway for future heightened cardiometabolic risk."97 Noncaloric sweeteners, such as aspartame, found in Diet Coke, or sucralose (Splenda) have not consistently shown adverse health effects.98 However, they may increase calorie consumption either by "justifying consumption of other caloric foods" or by "promoting a preference for sweet tastes."99 If diet sodas or other noncaloric sweeteners can be demonstrated to have negative health consequences, policymakers have good reason to include them in a soda tax. Even if diet soda does not necessarily contribute to obesity, but some other health consequence, it might still be taxed. There is little good in trading obesity for some other ailment.

Alternatively, consumers may substitute sugar-sweetened beverages for sugar-sweetened non-beverages. People crave sweets; some studies have found evidence that sugar may be addictive.100 If individuals aren't simply after a drink, but the sugar itself, the potential breadth for an effective soda tax increases dramatically. Take a look down the candy aisle. While a 12-ounce can of Pepsi has 150 calories and 41 grams of sugar, a regular Snickers bar has 250 calories and 21 grams of sugar. Are Pepsi and Snickers substitutes? Which is more unhealthy? If a soda tax extended to candy, Halloween may get expensive.

Sugar is hidden everywhere. In the cereal aisle, Froot Loops, despite the jovial toucan and its promotion of fiber and whole grain, has 110 calories from 12 grams of sugar in a 1-cup serving size.101 A cup of milk adds about 100 calories and 13 grams of sugar.102 Two bowls of cereal quickly becomes the equivalent in sugar intake to a can of soda. Let's hope the child doesn't have a glass of orange juice as well. This analysis could quickly spiral out of control. Even bread is criticized for its high sugar content. Although technically not listed on the ingredients label as sugar, the starches in bread are quickly broken down into glucose, causing rapid spikes in blood sugar and insulin levels.103 Is bread a substitute for soda? Perhaps the best thing policymakers could do is read the nutrition labels on everything they are eating, with an eye out for sugar. This would quickly inform them of just how prevalent sugar is in almost everything they are eating -- not just in soda.

  • 3. How Elastic Are Sugar-Sweetened Products?


A soda tax will be more effective if the taxed goods are more elastic (that is, an increase in the price of the goods causes a large decrease in consumption). On the other hand, if even the most narrowly defined markets (soda, juice, or milk, as opposed to a more broadly defined market such as beverages) are relatively inelastic, then the tax will not be as effective. The critical question, therefore, is how elastic are the goods with all the sugar?

Andreyeva, Long, and Brownell reviewed 160 studies on price elasticity of demand for different food categories.104 They found, consistent with customary belief, that demand response to food price changes is inelastic105 (that is, the absolute value of elasticity of demand is less than 1; if the price of the good increases by 10 percent, demand for the good decreases by less than 10 percent). It is not surprising that food is inelastic -- people must eat. However, there are important differences across food categories. Fortunately for advocates of a tax on sugar-sweetened beverages, both soft drinks and juice were among the most elastic categories of food, with respective elasticities of 0.79 and 0.76.106 People are not as willing to part with their Frosted Flakes -- cereals and milk have respective elasticities of 0.6 and 0.59. Finally, consumers are unfortunately quite unwilling to give up sweets and sugars, which have an elasticity of 0.34.107 These results show that a soda tax has most potential reduce consumption of their target: sugar-sweetened beverages.

These elasticity estimates show how much consumers are substituting away from some products. Unfortunately, these estimates do not show which products consumers are substituting toward when products such as soft drinks or juice are taxed. For purposes of designing an effective soda tax, this is a critical area of needed research.108 A soda tax resulting in a switch from soda to water is much different than a switch from soda to candy.

Another study estimates the elasticity of sugar-sweetened beverages in Mexico.109 This study found greater substitution away from sugar-sweetened beverages among the Mexican population, with an elasticity around 1.1.110 This result may be due to socioeconomic differences between the United States and Mexico, but studies from both countries show that sugar-sweetened beverages are relatively elastic. Further, the Mexico study found that elasticity of demand for soft drinks is greatest among the poor.111 Critics of a soda tax argue that it is regressive -- it falls disproportionately on the poor. However, as the poor's elasticity of demand for soda becomes more elastic, the burden of the tax falls relatively more on the wealthy rather than the poor. If the poor consume much less soda when its price increases, the poor pay less in tax revenue and will suffer fewer healthcare costs. If the elasticity of demand for soft drinks is greater among the poor in the United States, as it is in Mexico, the regressive concern is diminished.112

  • 4. How Do the Economic Rationales for a Soda Tax Inform
    Which Items Should Be Taxed?


If consumers of obesity-related foods and drinks have imperfect information or discount the future costs of consumption, there is economic reason to tax these products.

There are several reasons why consumers may not be fully informed concerning their sugar consumption. People may simply not recognize sugar on an ingredients label when they see it. There are at least 56 different types of sugar, including some which may not be readily identifiable, such as dextran, ethyl maltol, evaporated cane juice, and panocha.113 If someone is trying to reduce sugar intake but cannot recognize sugar on the ingredients label, that person may fail to reduce consumption. Further, even if people are checking ingredients labels, they may fail to realize that they are consuming much more than the small serving sizes on the package. Hence, they could be misled into believing that they are consuming fewer calories or grams of sugar than they really are.114 Nutritionists have argued that serving sizes should be relabeled to more accurately reflect the amount of food or drink that people actually consume. For example, a 20-ounce soft drink bottle may contain 2.5 servings, but people rarely consume just 8 ounces of soda.115 Cereal companies have been particularly criticized for their small serving sizes.116 When was the last time you ate a mere three-fourths of a cup of Honey Bunches of Oats?117 Do you know what that even looks like? Many people do not check nutrition labels, but may rather rely on their own preconceived, often erroneous, notion of how "healthy" a certain product is. For example, a 10-ounce bottle of Minute Maid apple juice, a drink marketed as "Kids' Juice and Juice Drinks," has 32 grams of sugar,118 the same as an equivalent amount of Coca-Cola.119 Finally, even if consumers are fully aware of their sugar consumption, they may not be fully aware of the health consequences of all that sugar. People may believe that sugar is bad, but they may not fully appreciate how bad it is. A soda tax on products for which there is imperfect information regarding sugar content or its consequences could compensate for this lack of information.

Next, even if consumers are fully informed concerning their sugar intake and its consequences, they may give too little weight to future health costs of immediate sugar consumption. The addictive nature of sugar120 may systematically cause people to consume more calories than they otherwise would. Similarly, aggressive food and drink advertising campaigns may cause many people to ignore the future consequences of sugar consumption. In 2014 Coca-Cola alone spent $3.5 billion in advertising.121 Todd Putnam, a Coca-Cola employee from 1997 to 2000, said the company strove to do more than beat out other brands -- it wanted to outsell everything that people drink, including milk and water.122 Some argue that persuasive advertising strips individuals of their autonomy.123 The marketing of sugary products is particularly troublesome when directed toward children, who are likely to ignore future health costs.

The products associated with imperfect consumer information and discounted future costs should be included in the soda tax. If people do not recognize the form of sugar on an ingredient label for a particular product, a soda tax should include that product. If people consistently underestimate caloric content of a product, that product should be taxed. If consumers erroneously believe a product is healthier than it really is, that product should be taxed. Harmful addictive and overly advertised products should be taxed.

C. Geographic Breadth


A tax on sugar-sweetened beverages may lead to two types of substitution. First, as discussed above, people may substitute taxed goods for alternative untaxed goods. Second, people may substitute taxed goods for identical untaxed goods outside the geographic jurisdiction of the tax. In other words, soda drinkers in Berkeley may start to buy their fizzy drinks in Oakland or San Francisco. This possibility of substitution across geographic lines diminishes the potential effectiveness of a soda tax. Berkeley consumers may reduce their soda consumption, but not as much as if they had no substitutes down the road or across the bay.


However, Berkeley residents face transaction costs when buying soda outside Berkeley. A transaction cost is any cost associated with buying or selling a good other than the price of the good itself.124 In this case, the transaction costs are the extra time and transportation costs needed to buy soda outside Berkeley. Berkeley soda drinkers have two options: pay for the taxed soda in Berkeley, or pay for the untaxed soda outside Berkeley and the transaction costs. If the latter is cheaper, consumers will buy soda outside Berkeley, and soda consumption will not decrease to the extent it would if consumers did not buy soda outside Berkeley.

As a corollary, the greater the transaction costs consumers face in buying untaxed goods, the less substitution and soda consumption there will be. The most obvious way to increase transaction costs for soda purchases would be to expand the geographic breadth of the soda tax. A state or federal soda tax would be more effective than a citywide soda tax. Berkeley soda drinkers may drive across town to get soda, but they would be much less likely to drive to Nevada or Mexico.125

A geographically broader soda tax would also reduce arbitrage. If soda is taxed in one area and not another, an arbitrageur could buy soda where it is not taxed and sell it for a bargain price where it is taxed. This would be profitable as long as the arbitrageur's profits from the resale are greater than the transaction costs of buying and reselling the soda. As the geographic coverage of the tax increases, transaction costs would also increase, eliminating potential profitability of arbitrage.

D. How Much Should the Tax Be?


A soda tax should correct the problem associated for the rationale on which it based. If the sole rationale for the soda tax is to correct a negative externality, the tax should equal the amount of the external costs associated with soda consumption. In other words, the consumption of one can of soda imposes costs on taxpayers via Medicare and Medicaid and insureds via increased premiums. A single can of soda should be taxed equal to those costs. For example, if a can of soda imposes a ten cent cost on taxpayers and a five cent cost on insureds, a 15-cent tax should be placed on that can of soda.


However, if other rationales for a soda tax are accepted, the tax should be increased to account for those rationales -- specifically, the need to account for hidden costs and discounted future costs. If drinking one can of soda imposes an obvious 15-cent cost on taxpayers and insureds but the soda drinker faces an additional 10 cents of hidden costs, the soda should be taxed 25 cents. Similarly, if the soda drinker faces additional future costs with a present value of 20 cents but only weighs those costs at 5 cents, the soda tax should be taxed at 30 cents (15 cents on taxpayers/insureds plus 15 cents to account for the mismeasurement of future costs).

This approach suggests that different products may be taxed at different amounts. If soda imposes greater costs on taxpayers or insureds than juice or candy, soda should be taxed more. If there are more hidden costs associated with juice than with soda, perhaps juice should be taxed more. If candy is more heavily advertised or more addictive, it should be taxed more. Allowing different products to be taxed at different amounts will allow the products to have a change in relative prices. The worst products -- those with the most external, hidden, or discounted costs -- will be taxed the most. Consumers will substitute away from the heavily taxed products toward lesser- or nontaxed products.

While this theoretical approach provides a helpful conceptual framework, the complex nature of obesity, health insurance, and human psychological behavior will frustrate researchers who seek to find the correct tax for each product. Determining the cost of one can of soda on taxpayers requires answering difficult empirical questions: How much of obesity is due to sugar consumption? How much do different sugary products contribute to obesity? Does sugar consumption have diminishing returns for causing obesity? How does obesity translate into healthcare costs? How much of obesity-related healthcare costs are paid through Medicare or Medicaid? Similar questions are raised regarding the hidden information and discounted future costs associated with sugary products. However, failure to answer these questions precisely should not prevent policymakers from taking action.


Obesity is taking a devastating toll on the health and economic well-being of Americans. Sugar-sweetened beverages are a significant contributor to the obesity epidemic, and current trends suggest that obesity rates will continue to grow. Fortunately, a tax on soda will likely encourage consumers to substitute away from these drinks.

However, many research questions remain regarding the effectiveness of a soda tax. Most importantly, when consumers substitute away from soda, what products are they substituting toward? Do consumers have substitution patterns that predict what products they will substitute toward? How harmful are these potential soda substitutes and how do they affect health? Will broadening a tax to cover both soda and its substitutes cause a reduction in calorie consumption? How much external cost is created by soda consumption? Do consumers lack information regarding the health costs of soda consumption? Do consumers err in evaluating the future costs of these products?

We may never know the exact answers to these questions. However, that shouldn't necessarily prevent policymakers from taking action. Chief Justice Warren E. Burger stated, "From the beginning of civilized societies, legislators and judges have acted on various unprovable assumptions. Such assumptions underlie much lawful state regulation of commercial and business affairs."126 While the assumptions underlying a soda tax may be unprovable, an imperfect policy in the right direction is better than no policy at all.


1 Cynthia L. Ogden and Margaret D. Carroll, "Prevalence of Overweight, Obesity, and Extreme Obesity Among Adults: United States, Trends 1960-1962 Through 2007-2008," Centers for Disease Control and Prevention (June 2010). Obesity is defined as having a body mass index greater than or equal to 30, where BMI is a person's weight in kilograms divided by height in meters squared. Extreme obesity is defined as having a BMI over 40. Accordingly, a higher BMI corresponds to a greater weight to height ratio, or an excess of body fat.


3 CDC, Obesity Prevalence Maps, available at http://www.cdc.gov/obesity/data/prevalence-maps.html.

4 CDC, Defining Adult Overweight and Obesity, available at http://www.cdc.gov/obesity/adult/defining.html.

5 CDC, Adult Obesity Causes & Consequences, available at http://www.cdc.gov/obesity/adult/causes.html.

6 Floriana S. Luppino et al., "Overweight, Obesity, and Depression: A Systematic Review and Meta-Analysis of Longitudinal Studies," 67 Archives General Psychiatry 220 (2010).

7 CDC, Childhood Obesity Facts, available at http://www.cdc.gov/obesity/data/childhood.html.

8 American Heart Association, Overweight in Children (updated Aug. 17, 2015).

9 World Health Organization, Childhood Overweight and Obesity, available at http://www.who.int/dietphysicalactivity/childhood/en/.

10 American Heart Association, supra note 8.

11 David Cutler, Edward Glaeser, and Jesse Shapiro, "Why Have Americans Become More Obese?" National Bureau of Economic Research, Working Paper No. 9446 (2003).


13Id. at 1-2. See also Eric A. Finkelstein, Christopher J. Ruhm, and Katherine M. Kosa, "Economic Causes and Consequences of Obesity," 26 Annual Review Public Health 239 (2005); and Darius Lakdawalla and Tomas Phillipson, "The Growth of Obesity and Technological Change: A Theoretical and Empirical Examination," National Bureau of Economic Research Working Paper No. 8946 (2002).

14 Finkelstein, Ruhm, and Kosa, supra note 13, at 242.

15 Kelly D. Brownell and Thomas R. Frieden, "Ounces of Prevention -- The Public Policy Case for Taxes on Sugared Beverages," 360 New England J. Med. 1805, 1805 (2009).

16 David S. Ludwig, Karen E. Peterson, and Steven L. Gortmaker, "Relation Between Consumption of Sugar-Sweetened Drinks and Childhood Obesity: A Prospective, Observational Analysis," 357 Lancet 505 (2001).

17 Finkelstein, Ruhm, and Kosa, supra note 13, at 247-48.

18 Centers for Medicare & Medicaid Services, National Health Expenditure Accounts, available at http://go.cms.gov/1UFHHer.

19 Finkelstein, Ruhm, and Kosa, supra note 13, at 248-49.

20Id. at 249.

21Id. at 252.

22 World Health Organization, supra note 9.

23 Deborah L. Rhode, "Obesity and Public Policy: A Roadmap for Reform," 22 Va. J. Soc. Pol'y & L. 491, 521 (2015).

24Id. at 521-22.

25 Denis Binder, "Food Fights -- Southern California Style," 17 Nexus L.J. 97, 99 (2012)

26 Laura Hoffman, "The Fight Over Fizz: Soda Taxes as a Means of Curbing Childhood Obesity?" 5 Pitt. J. Envtl. Pub. Health L. 123, 136-37 (2011).

27Id. at 137.

28 Malia Kim, "Give Me Back My Big Gulp! The Constitutionality of Obesity Regulations Under the Due Process Clause," 80 Tenn. L. Rev. 847, 867 (2013).

29 Hoffman, supra note 26, at 138.


31See Cent. Hudson Gas & Elec. Corp. v. Pub. Serv. Comm'n of New York, 447 U.S. 557 (1980).

32 Rhode, supra note 23, at 516.

33 About Let's Move, http://www.letsmove.gov/about.


35 Rhode, supra note 23, at 505.

36Id. at 514.


38Id. at 509.

39Id. at 510.

40 Hery Min, "Large-Sized Soda Ban as an Alternative to Soda Tax," 23 Cornell J. L. & Pub. Pol'y 187, 188 (2013).

41 Noelia Duchovny et al., "Estimating the Effects of Federal Policies Targeting Obesity: Challenges and Research Needs," Congressional Budget Office (Oct. 26, 2015).

42 Rhode, supra note 23, at 519.


44 Brownell and Frieden, supra note 15, at 1806.

45See Finkelstein, Ruhm, and Kosa, supra note 13, at 252.

Even with full information about the benefits of physical activity, the nutrient content of food, and the health consequences of obesity, some fraction of the population will optimally choose to engage in a lifestyle that leads to weight gain because the costs (in terms of time, money, and opportunity costs) of not doing so are just too high. Because much of the monetary costs of obesity are financed by taxpayers, the physical activity and food consumption decisions of these individuals are not optimal from a broader societal perspective, implying a role for government in attempting to reduce obesity. It should be acknowledged, however, that for many individuals, information-based interventions or other interventions that do not affect the costs or benefits of physical activity and food consumption decisions are unlikely to be effective. For some individuals, additional incentives may be needed to encourage them to lose weight.

46 History of Tobacco Regulation, DrugLibrary.Org, available at http://bit.ly/1Pgg2x1.


48 Chaloupka, "How Effective Are Taxes in Reducing Tobacco Consumption?" Preliminary Draft for the International Conference on the Cost of Smoking (1998).



51See Judy Jou and Win Techakehakij, "International Application of Sugar-Sweetened Beverage (SSB) Taxation in Obesity Reduction: Factors That May Influence Policy Effectiveness in Country-Specific Contexts," 107 Health Pol. 83, 88 (2012) (giving a complete list of countries with SSB taxes or proposed SSB taxes as of 2012).

52 World Health Organization, "Reducing Consumption of Sugar-Sweetened Beverages to Reduce the Risk of Unhealthy Weight Gain in Adults," Sept. 2014, available at http://www.who.int/elena/bbc/ssbs_adult_weight/en/.

53 Christine Spolar and Joseph Eaton, "Food Lobby Mobilizes, as Soda Tax Bubbles Up," Huffington Post, Nov. 4, 2009.


55 Kelly D. Brownell et al., "The Public Health and Economic Benefits of Taxing Sugar-Sweetened Beverages," 361 N. Engl. J. Med. 1599, 1599 (2009).

56Id. One study concludes that "current state-level tax rates are not found to be significantly associated with adolescent weight outcomes." However, the authors suggest that "taxes would need to be raised substantially to detect significant associations between taxes and adolescent weight." Lisa M. Powell, Jamie Chriqui, and Frank J. Chaloupka, "Associations Between State-level Soda Taxes and Adolescent Body Mass Index," 45 J. Adolescent Health S57, S57 (2009).

57 J. Angelo DeSantis, "Formulating a Soda Tax Fit for Consumption: A Pragmatic Approach to Implementing the Failed New York Soda Tax," 16 Mich. St. U. J. Med. & L. 363, 367 (2012).


59 John Cawley and David Frisvold, "The Incidence of Taxes on Sugar-Sweetened Beverages: The Case of Berkeley, California," National Bureau of Economic Research Working Paper No. 21465, 2 (2015). I would like to thank the authors of this study for their correspondence.


61 Tom Lochner, "Berkeley Soda Tax: First Month's Take, $116,000," MercuryNews.Com, May 19, 2015.

62See, e.g., Brownell and Frieden, supra note 15 (for soda tax); Katherine Pratt, "A Constructive Critique of Public Health Arguments for Antiobesity Soda Taxes and Food Taxes," 87 Tul. L. Rev. 73 (2012) (critiquing two arguments made for soda taxes); Hoffman, supra note 26 (generally arguing in favor of a soda tax); and Kim, supra note 28 (analyzing whether soda taxes violate the due process clause).

63 Personally, I strongly favor a soda tax.

64See Jay Bhattacharya and Neeraj Sood, "Health Insurance and the Obesity Externality," National Bureau of Economic Research Working Paper No. 11529 (2005); Brownell et al., supra note 55, at 1601-02.

65 A Pigouvian tax is named after Arthur Pigou, who developed the idea of using taxes as a means to solve the problem of negative externalities. Arthur C. Pigou, The Economics of Welfare (1932).

66 Bhattacharya and Sood, supra note 64.

67 Pratt, supra note 62, at 94. "Contrary to the central assumption of decision theory, not all behavior is volitional, and very likely most of it is not. . . . [M]uch behavior is non-volitional or only partly volitional -- even in situations characterized by substantial deliberation" (quoting George Loewenstein, Out of Control: Visceral Influences on Behavior, in Advances in Behavioral Economics 689, 717 (Colin F. Camerer et al. eds., 2004)).

68Id. at 96.

69 Because the first law of demand says that the price and quantity demanded of a good are inversely related, elasticity of demand will always be a negative number. For this reason, economists occasionally report elasticity of demand as a positive number to avoid repetitive use of the word negative.

70 Economics Online, Price Elasticity of Demand, available at http://bit.ly/1NVLp3B.

71 To illustrate, I have been using a price increase, which leads to a consumption decrease. However, an elasticity measurement is also proper for the converse: a price decrease, leading to a consumption increase. In both cases, the elasticity of demand is negative because the price and quantity demanded of a good move in opposite directions (that is, if one is negative, the other is positive).

72See, e.g., Pratt, supra note 62, at 77-78; Brownell and Frieden, supra note 15, at 1806; and Rhode, supra note 23, at 500.

73 Bhattacharya and Sood, supra note 64.

74 Brownell and Frieden, supra note 15, at 1806; and Finkelstein, Ruhm, and Kosa, supra note 13, at 248.

75 Stateofobesity.org, Adult Obesity by State, 2014, available at http://stateofobesity.org/adult-obesity/.

76 Ronald Sturm, "The Effects of Obesity, Smoking, and Drinking on Medical Problems and Costs," 21 Health Affairs 245, 245 (2002).


78 Bhattacharya and Sood, supra note 64.





83 Pratt, supra note 62, at 122-23.

84 M.S. Colchero et al., "Price Elasticity of the Demand for Sugar Sweetened Beverages and Soft Drinks in Mexico," 19 Econ. and Human Bio. 129 (2015).

85 Chaloupka, supra note 48, at 10.

86See, e.g., Brownell and Frieden, supra note 15; Finkelstein, "Impact of Targeted Taxes on Higher- and Lower-Income Households," 170 JAMA Internal Medicine 2028, (2010); and Anurag Sharma et al., "The Effects of Taxing Sugar-Sweetened Beverages Across Different Income Groups," 23 Health Econ. 1159 (2014).

87 Sharma, supra note 86, at 1160.

88 Berkeley, Cal. Municipal Code, ch. 7.72.030(O)(1).

89Id. at 7.72.030(A)(4) (internal quotation marks omitted).

90Id. at 7.72.030(O).

91 Brownell and Frieden, supra note 15, at 1805.

92 Harvard T.H. Chan School of Public Health, Sugary Drinks and Obesity Fact Sheet, available at http://bit.ly/1dlX9Xf.

93 PepsiCo, The Facts About Your Favorite Beverage, Pepsi, available at http://bit.ly/1SPE5YP.

94 Webmd.com, "Can High-Fructose Corn Syrup Make You Hungrier?" Dec. 10, 2014.

95 Cawley and Frisvold, supra note 59, at 26.


97 Sharon P.G. Fowler, Ken Williams, Helen P. Hazuda, "Diet Soda Intake Is Associated With Long-Term Increase in Waist Circumference in a Biethnic Cohort of Older Adults: The San Antonio Longitudinal Study of Aging," 63 J. Am. Geriatrics Soc. 709, 714 (2015).

98 Brownell et al., supra note 55, at 1603.

99Id. (citing Richard D. Mattes and Barry M. Popkin, "Nonnutritive Sweetener Consumption in Humans: Effects on Appetite and Food Intake and Their Putative Mechanisms," 89 Am. J. Clinical Nutrition 1 (2009)).

100See, e.g., Nicole M. Avena, Pedro Rada, Bartley G. Hoebel, "Evidence for Sugar Addiction: Behavioral and Neurochemical Effects of Intermittent, Excessive Sugar Intake," 32 Neuroscience & Biobehavioral Reviews 20 (2008); and Serge H. Ahmed, Karine Guillem, Youna Vandaele, "Sugar Addiction: Pushing the Drug-Sugar Analogy to the Limit," 16 Clinical Nutrition & Metabolic Care 434 (2013).


102 Atli Arnarson, "Milk 101: Nutrition Facts and Health Effects," authoritynutrition.com, July 2105.

103 Kris Gunnars, "Why Is Bread Bad for You? The Shocking Truth," authoritynutrition.com, Feb. 2013.

104 Tatiana Andreyeva, Michael W. Long, and Brownell, "The Impact of Food Prices on Consumption: A Systematic Review of Research on the Price Elasticity of Demand for Food," 100 Am. J. Pub. Health 216 (2010).

105Id. at 218.

106Id. at 219.


108Id. at 216.

Experimental research in both laboratory and intervention settings shows that lowering the price of healthier foods and raising the price of less healthy alternatives shift purchases toward healthier food options. Although these studies demonstrate price effects in specific, isolated settings or on one or two individual product changes, to our knowledge, the expected effects of broader food price changes have not been systematically reviewed. Such information would be helpful in designing policies that change the relative food and beverage prices paid by all or many consumers.

109 Colchero et al., supra note 84.

110Id. at 132. The study reports the elasticity of demand as -1.1, but to compare with Andreyeva's study, I have taken the absolute value.

111Id. at 136.

112 After the tax, the poor may still pay a higher proportion of income on soda than the wealthy, and hence the tax would still be regressive. The poor's more elastic soda consumption only suggests that this proportion of income is lower -- and hence the tax is less regressive -- than if the poor's soda consumption was more inelastic.

113 Institute for Responsible Nutrition, "Learn to Recognize the 56 Different Names for Sugar," listing other names for sugar, including barley malt, crystalline fructose, and maple sugar.

114 Center for Science in the Public Interest, Food Labeling Chaos: The Case for Reform (Mar. 2010):

  • Consumers may be misled when they look at the nutrition information for a product that they assume to be a single serving and do not realize that the nutrition information is based on only a fraction of the product. For example, consumers may buy an individually packaged blueberry muffin that is labeled as containing 200 calories. But that disclosure may be based on the number of calories in only half the muffin. The actual calorie content of the muffins as packaged could be 400 calories. Some consumers may not have purchased the product, or eaten only half, if the calorie content of the product were accurately stated on the label.


116 Elaine Watson, "Serving Sizes for Cereals, Canned Soup and Cooking Sprays Are Completely Unrealistic, Say Yale Researchers," http://www.foodnavigator-usa.com, Aug. 27, 2014.

117 Honey Bunches of Oats nutritional information, available at http://bit.ly/1UwwP0b.

118 Minute Maid Apple Juice nutritional information, available at http://bit.ly/1U5MUwP.

119 Coca-Cola nutritional information, available at http://bit.ly/1OiJWVb.

120See Avena, Rada, and Hoebel, supra note 100; and Ahmed, Guillem, and Vandaele, supra note 100.

121 Investopedia, "A Look at Coca-Cola's Advertising Expenses," Aug. 13, 2015.

122 Michael Moss, Salt, Sugar, Fat: How the Food Giants Hooked Us (2013). Putnam said, "We weren't trying to get a share of the market. We weren't trying to beat Pepsi or Mountain Dew. We were about trying to beat everything." Id.

123 Roger Crisp, "Persuasive Advertising, Autonomy, and the Creation of Desire," 6 J. Bus. Ethics 413, 413-14 (1987).

124 Paul M. Johnson, A Glossary of Political Economy Terms, Transaction Costs.

125 If soda drinkers burned sufficient calories walking or biking across town to get their soda, geographic substitution might make a soda tax more effective.

126Paris Adult Theatre I v. Slaton, 413 U.S. 49, 60 (1973).