Measuring weight outcomes for obesity intervention strategies: The case of a sugar-sweetened beverage tax
Highlights
► We utilize national retail purchase data to estimate U.S. beverage demand elasticities. ► We estimate calorie intake reductions from a SSB tax, which is found to be regressive. ► Static and dynamic models are used to translate energy deficits to weight loss. ► The static model is shown to overestimate weight loss from energy reductions. ► We estimate the overweight and obesity rates before and after tax among U.S. adults.
Introduction
The prevalence of childhood and adult obesity in the U.S. has more than doubled between 1976–1980 and 2007–2008 (Flegal et al., 2010, Ogden et al., 2010). Despite a recent plateau in prevalence, obesity is a major health and economic concern in the U.S. The medical costs associated with overweight and obesity were estimated to be $113.8 billion in 2008 (Tsai et al., 2011). Concurrently, children and adults have substantially increased their consumption of sugar-sweetened beverages (SSBs), as shown in Fig. 1. As researchers examine the causes of obesity and intervention strategies, its association with SSB consumption has received increasing attention.
Responding to rising childhood obesity, the Institute of Medicine (2009) has advocated local Governments to tax calorie-dense, nutrient-poor foods and beverages. Similar recommendations have been proposed as a means of reducing SSB consumption, improving diet and health, and generating revenue to address obesity (Brownell and Frieden, 2009, Jacobson and Brownell, 2000, Powell and Chaloupka, 2009).
The White House Task Force on Childhood Obesity (2010) issued a set of recommendations to fight childhood obesity, including a deeper investigation into the effects of taxing less healthy, energy-dense foods, such as soft drinks, candy, snack foods, and fast foods. Pricing-policy interventions can be justified by economic rationales based on market failures (Brownell and Frieden, 2009, Cawley, 2004, Powell and Chaloupka, 2009). The medical costs associated with overweight and obesity generate a negative externality in the form of higher health insurance premiums and higher government expenditures under Medicaid and Medicare. Obesity also results in loss of productivity and reduced tax revenues. Furthermore, pricing interventions can correct for time-inconsistent preferences by individuals and asymmetric information.
In this study, we investigate two important issues pertaining to a SSB tax. First, we estimate income-segmented beverage demands to examine the concern whether or not a SSB tax is regressive. Second, we demonstrate that estimated reductions in body weight and obesity prevalence from reduced calorie intakes can differ greatly between two prediction models. The first model is the static rule of 3500-calorie per pound of body weight, which has been widely used to estimate the rate of weight loss resulting from a given reduction of energy intake (e.g., Finkelstein et al., 2010, Smith et al., 2010). This weight loss rule neglects known physiological changes including adaptations that reduce the energy expended by the body as weight is lost. This omission results in a sizable understatement of the reduction of energy intake required for weight loss. This study contributes to the literature by examining a second model to incorporate the adaptations of energy expenditure as well as realistic changes of body composition to calculate the time-path weight loss (see Appendix A for a full description of the model).
We estimate demand elasticities and apply a hypothetical tax rate to individual consumption data in a nationally representative survey to calculate adjustments in beverage consumption and the associated reductions in calorie intake. From our literature review, we obtain beverage demand elasticities from two recent studies using the same data source and these elasticities are used in a sensitivity analysis. With estimated calorie reductions and measured body weight and height for each respondent, we apply two alternative (static and dynamic) models to translate reduced energy intake into reductions in body weight and obesity prevalence. Our beverage demand systems are stratified by income in order to improve elasticity estimates, to examine whether the effects of a SSB tax differ by income, and whether the tax is regressive. Our results show that the static model overestimates weight loss in the first year by 63 percent and exponentially increases to a 764 percent overestimation in the tenth year. Further, the tax is found to be regressive by definition, although the burden is less than 1 percent of food and beverage spending.
Section snippets
Literature review
Two recent systematic literature reviews conclude that SSB consumption is linked to an increased risk for obesity (Malik et al., 2006, Vartanian et al., 2007). As such, a fast growing literature has been devoted to examining the potential effects of taxing SSBs with mixed results. Duffey et al. (2010) examined the consumption of soda, whole milk, pizza, and hamburger using 20-year longitudinal data and found a 1-dollar increase in the price of soda (about 50 percent of the mean price) would
Data and methods
We use two nationwide datasets to evaluate dietary, health, and economic impacts of taxing SSBs. Data from the1998–2007 NCP are used to estimate two beverage demand systems stratified by income. Beverage demand price elasticities and a hypothetical 20 percent effective tax are applied to individual food intake data from the 2003–2006 National Health and Nutrition Examination Survey (CDC, NHANES, 2009) to estimate changes in beverage consumption, calorie intake, weight loss, and body weight
Beverage demand elasticities
Because a SSB tax is the policy lever under consideration, our discussion of demand elasticities will focus on SSB own- and cross-price elasticities. For brevity, parameter estimates are not reported (available upon request) and demand elasticities are reported in Table 3, Table 4. All eight own-price elasticities for both low- and high-income populations are negative and statistically significant at the 1 percent probability level, except for low-income skim milk. Consistent with the relative
Results: calorie, body weight, and obesity prevalence simulation
In addition to food intake data, NHANES reports measured body weight and height as well as demographic data for each respondent, making the data ideal for translating reductions of energy intake to weight losses. With the estimated weight reduction for each sample person, we can use the sample weights to estimate the U.S. after-tax prevalence of overweight and obesity.
Conclusions
Research on the consequences of taxing SSBs has rapidly proliferated in the economic and public health literature. A number of studies have estimated beverage demands and used price elasticities to calculate the effect of a SSB tax on beverage consumption, tax revenue, and calorie intake (Dharmasena and Capps, 2010, Finkelstein et al., 2010, Zhen et al., 2011). Because a key motivation for taxing SSBs is to tackle the obesity epidemic by reducing calorie intake, it is important to carry the
Acknowledgements
The authors thank John Cawley and three anonymous referees for their helpful comments and suggestions. The views in this paper are those of the authors and do not necessarily reflect the views or policies of the U.S. Department of Agriculture.
References (44)
- et al.
Incorporating demand shifters in the almost ideal demand system
Economic Letters
(2001) An economic framework for understanding physical activity and eating behaviors
American Journal of Preventive Medicine
(2004)- et al.
The effects of soft drink on child and adolescent consumption and weight outcomes
Journal of Public Economics
(2010) - et al.
Intake of sugar-sweetened beverages and weight gain: a systematic review
American Journal of Clinical Nutrition
(2006) - et al.
Separating what we eat from where: measuring the effect of food away from home on diet quality
Food Policy
(2009) - et al.
Associations between state-level soda taxes and adolescent body mass index
Journal of Adolescent Health
(2009) - et al.
Determining the impact of food price and income changes on body weight
Journal of Health Economics
(2008) - et al.
The impact of food prices on consumption: a systematic review of research on price elasticity of demand for food
American Journal of Public Health
(2010) - et al.
Estimation and hypothesis testing in singular equation systems with autoregressive disturbances
Econometrica
(1975) - et al.
The public health and economic benefits of taxing sugar-sweetened beverages
New England Journal of Medicine
(2009)