The Climate Cobenefits of Obesity Reduction Anthony Underwood

The Climate Co-benefits of Obesity Reduction Anthony Underwood, Dickinson College Sammy Zahran, Colorado State University AEA 2016 ANNUAL MEETING JAN 3, 2016 UNDERWOOD: AEA 2016 1

Motivation The evidence of health co-benefits resulting from greenhouse gas (GHG) mitigation is clear: • reductions in ozone and particulate matter emissions, resulting from GHG mitigation policies will lead to immediate improvements in public health and generate significant co-benefits dramatically improving their cost-effectiveness (Buonocore, 2014; Shaw et al. , 2014) • both the drivers and impacts of climate change play a significant role in population health through a variety of direct and indirect channels (Mc. Michael, 2013) What if the reverse is also true? UNDERWOOD: AEA 2016 2

Research Question Can improvements in public health, in particular reductions in obesity and overweight prevalence, generate emissions reductions and thus climate co-benefits? In other words, Can policies designed to reduce obesity generate reductions in GHG emissions? Or more simply, Is obesity prevalence associated with GHG emissions? UNDERWOOD: AEA 2016 3

The political economy of obesity The increased prevalence of overweight and obesity worldwide is largely the result of structural factors. These factors include: 1. 2. 3. 4. an increased intake of energy-dense foods that are high in fat; an increase in physical inactivity due to the increasingly sedentary nature of many forms of work; changing modes of transportation; and increasing urbanization (Ledikwe et al. , 2006; Malik et al. , 2013). Many of these changes are likely the result of economic and social changes associated with development and need to be understood in that context. UNDERWOOD: AEA 2016 4

Body Mass Index (BMI) UNDERWOOD: AEA 2016 5

BMI is often misleading… Source: New York Times, September 3, 2015. Using data from the CDC National Health and Nutrition Examination Survey. http: //www. nytimes. com/interactive/projects/cp/summer-of-science-2015/latest/how-often-is-bmi-misleading UNDERWOOD: AEA 2016 6

Theoretical Channels The hypothesized channels through which obesity leads to GHG emissions in excess of an otherwise ‘healthy’ population, as measured by BMI, are generally: 1. increased food production, especially animal-based products, and food waste generation due to higher caloric intake of obese and overweight individuals (Edwards and Roberts, 2009; Michaelowa and Dransfeld, 2008; Walpole et al. , 2012); 2. higher fuel use from motorized transport due to increased passenger weight and the assumption that heavier individuals may use motorized travel more and choose larger fuel-inefficient vehicles (Edwards and Roberts, 2009; Michaelowa and Dransfeld, 2008; Goodman et al. , 2012); and UNDERWOOD: AEA 2016 7

Increased food production Edwards and Roberts (2009) find that an overweight population (with mean BMI of 29 and 40% obese) would require 19% more food energy for its total energy expenditure compared to a ‘normal’ population (with mean BMI of 24. 5 and 3. 5% obese). Michaelowa and Dransfeld (2008) also find that emissions from food production have increased, but do not establish a causal pathway to higher prevalence of obesity. Walpole et al. (2012) find that if all countries had the BMI distribution of the United States, it would be equivalent to having an ‘extra’ 473 million adults living on earth. UNDERWOOD: AEA 2016 8

Higher fuel use Dannenberg et al. (2004) find that the increase in the average weight of U. S. citizens during the 1990 s led to an increase in fuel use of 2. 4% and annual CO 2 emissions from U. S. air traffic by 3. 8 million metric tons. Edwards and Roberts (2009) find that an overweight population would generate 12% more transport CO 2 emissions than a ‘normal’ population. Tom et al. (2014) estimate over 205 billion additional liters of fuel were consumed to support the extra weight of the U. S. population since 1970, resulting in an extra 502 million metric tons of CO 2 emissions. UNDERWOOD: AEA 2016 9

Existing Literature Only Squalli (2014) has investigated whether obesity prevalence (as measured by BMI) is associated with higher greenhouse gas emissions at the national or regional level. ◦ using data for the fifty US states in 2010, Squalli (2014) estimates that a 10% reduction in the obesity rate reduces CO 2 emissions by 0. 7% ◦ given the cross-sectional nature of the sample and the inability to account for likely unobserved heterogeneity and spatial dependence among the 50 states, these estimates are likely biased and inconsistent UNDERWOOD: AEA 2016 10

Data UNDERWOOD: AEA 2016 11

Model Specification Several estimation strategies are plausible, most of which control out between state variation in favor of estimating within state effects, here we focus on two: • fixed effects (FE) with clustered standard errors • • robust to serial correlation and heteroskedasticity but assumes cross-sectional independence a Pesaran test for cross-sectional independence confirms spatial dependence among the states. • dynamic Prais-Winsten (PW) regression using panel corrected-standard errors • • • robust to cross-sectional dependence, AR(1) serial correlation within each state, and heteroskedasticity. subject to dynamic panel bias (Nickell bias) due to inclusion of lagged dependent variable potential for unit root UNDERWOOD: AEA 2016 12

Methodology UNDERWOOD: AEA 2016 13

Results The obesity elasticity is an estimated 0. 13, suggesting a 1% reduction in the obesity rate generates a 0. 13% reduction in CO 2 emissions. A Hadri LM unit-root test confirms some panels contain a unit root so we allow the AR(1) parameter to be panel -specific. We accept some (mild) Nickell bias in order to account for (severe) spatial dependence bias. UNDERWOOD: AEA 2016 14

Results in Context In the United States, from 1995 – 2013: • prevalence of overweight and obese adults in the United States increased from 51% to 64% an increase of 25% • total annual CO 2 emissions from energy use have increased by 73 million metric tons Our results imply that 2. 7% of this increase in CO 2 emissions (around 1. 9 million metric tons per year) is attributable to the increased prevalence of obesity The average American household generates about 12 metric tons of CO 2 per year. • increased prevalence of obesity is similar to the effect of having an additional 160, 833 households, or nearly half a million additional people in the U. S. every year. UNDERWOOD: AEA 2016 15

Climate Co-benefits of Obesity Reduction We estimate that reversion to 1997 obesity rates in every state nationwide in 2014 would reduce annual US CO 2 emissions from energy use by 143 million metric tons, or 2. 7% below 2013 levels. • yields annual climate benefits of between $5. 7 and $8. 9 billion using discount rates of 3% and 2. 5%, respectively. These emissions reductions and resulting benefits are similar to those expected via implementation of the U. S. Clean Power Plan in 2025 • EPA projects annual emissions reductions of 211 million metric tons by 2025. UNDERWOOD: AEA 2016 16

Discussion Our results are suggestive of a mutually-reinforcing relationship between policies designed to improve public health (via reduced obesity) and climate change mitigation. • Policies aimed at reducing transportation related emissions via increased active transport may achieve both direct climate benefits through reduced fuel use and indirect climate benefits via reduced obesity. This supports research stressing the importance of the built environment for both public health and sustainable development (Jackson, Dannenberg, and Frumkin, 2013). UNDERWOOD: AEA 2016 17

References Buonocore, J. (2014). Climate policy not so costly. Nature Climate Change, 4, 861 -862. Malik, V. S. , Willet, W. C. , & Hu, F. B. (2013). Global obesity: trends, risk factors, and policy implications. Nature Reviews Endocrinology, 9(1), 13 -27. Dannenberg, A. L. , Burton, D. C. , & Jackson, R. J. (2004). Economic and environmental costs of obesity: the impact on airlines. American Journal of Preventive Medicine, 27(3), 264. Medicine, 368(14), 1335 -1343. Edwards, P. , & Roberts, I. (2009). Population adiposity and climate change. International Journal of Epidemiology, 38(4), 1137 -1140. Michaelowa, A. , & Dransfeld, B. (2008). Greenhouse gas benefits of fighting obesity. Ecological Economics, 66, 298 -308. Evans, B. (2012). Climate change and the politics of fatness. Environmental Politics, 21(2), 334 -336. Goodman, A. , Brand, C. , & Ogilvie, D. (2012). Associations of health, physical activity and weight status with motorised travel and transport carbon dioxide emissions: a cross-sectional, observational study. Environmental Health, 11(52). Green, M. , Strong, M. , Razak, F. , Subramanian, S. , Relton, C. , & Bissell, P. (2015). Who are the obese? A cluster analysis exploring subgroups of the obese. Journal of Public Health, doi: 10. 1093/pubmed/fdv 040. Jackson, R. J. , Dannenberg, A. L. , & Frumkin, H. (2013). Health and the built environment: 10 years after. American journal of public health, 103(9), 1542 -1544. Ledikwe, J. H. , Blanck, H. M. , Kettel Khan, L. , Serdula, M. K. , Seymour, J. D. , Tohill, B. C. , & Rolls, B. J. (2006). Dietary energy intensity is associated with energy intake and weight status in US adults. The American Journal of Clinical Nutrition, 83, 1362 -1368. Lowe, M. (2014). Obesity and climate change mitigation in Australia: overview and analysis of policies with co-benefits. Australian and New Zealand Journal of Public Health, 38(1), 19 -24. Mc. Michael, A. J. (2013). Globalization, Climate Change, and Human Health. The New England Journal of Reisch, L. A. , & Gwozdz, W. (2011). Chubby cheeks and climate change: childhood obesity as a sustainable development issue. International Journal of Consumer Studies, 35, 3 -9. Shaw, C. , Hales, S. , Howden-Chapman, P. , & Edwards, R. (2014). Health co-benefits of climate change mitigation policies in the transport sector. Nature Climate Change, 4, 427 -433. Squalli, J. (2014). Is obesity associated with global warming? Public Health, 128, 1087 -1093. Tom, M. , Fischbeck, P. , & Hendrickson, C. (2014). Excess passenger weight impacts on US transportation systems fuel use (1970 -2010). Journal of Transport and Health, 1, 153 -164. Walpole, S. C. , Prieto-Merino, D. , Edwards, P. , Cleland, J. , & Stevens, G. (2012). The weight of nations: an estimation of adult human biomass. BMC Public Health, 12(439). Wohlfahrt-Veje, C. , Tinggaard, J. , Winther, K. , Mouritsen, A. , Hagen, C. , Mieritz, M. , . . . Main, K. (2014). Body composition, energy expenditure, and physical activity. European Journal of Clinical Nutrition, 68, 664 -670. UNDERWOOD: AEA 2016 18

Additional Results UNDERWOOD: AEA 2016 19

Additional Results The interaction term is positive and significant. obesity is acting to amplify the total CO 2 emissions response to economic growth (or decline). UNDERWOOD: AEA 2016 20