Evaluation of Changes in Grams of Sugar Sold After the Implementation of the Seattle Sweetened Beverage Tax

Abstract

IMPORTANCEAdults and children routinely exceed recommended intake amounts of added sugars established by dietary guidelines. Taxes are used as a policy tool to reduce demand for sugarsweetened beverages (SSBs) given consumption-related adverse health outcomes but may induce substitution to other sources of added sugars. OBJECTIVE To examine the extent to which changes in grams of sugar sold from taxed beverages may be offset by changes in grams of sugar sold from untaxed beverages, sweets, and stand-alone sugar after the implementation of the Seattle, Washington, Sweetened Beverage Tax (SBT) on January 1, 2018. DESIGN, SETTING, AND PARTICIPANTS This study used difference-in-differences analyses to examine changes in grams of sugar sold from taxed and untaxed products in Seattle compared with Portland, Oregon, at year 1 and year 2 post tax. This study used Nielsen scanner data from supermarkets and mass merchandise as well as grocery, drug, convenience, and dollar stores on unit sales and measurements for beverage and food product universal product codes (UPCs) for each site for the pretax period (January 8-December 30, 2017) and the corresponding weeks in year 1 post tax (2018) and in year 2 post tax (2019). Nutritional analyses assessed grams of sugar for each UPC. The analytical balanced sample included 1326 taxed beverage UPCs, 239 untaxed beverage UPCs, 2054 sweets UPCs, and 81 stand-alone sugar UPCs. Statistical analysis was performed from January to August 2021. EXPOSURES Implementation of the Seattle SBT. MAIN OUTCOMES AND MEASURESChanges in grams of sugar sold from taxed beverages, untaxed beverages, sweets, and stand-alone sugar. RESULTSAt both year 1 and year 2 post tax in Seattle compared with Portland, grams of sugar sold from taxed beverages decreased 23% (year 2 posttax ratio of incidence rate ratios [RIRR] = 0.77; 95% CI, 0.73-0.80). Sugar sold from untaxed beverages increased at year 1 post tax by 4% (RIRR = 1.04; 95% CI, 1.00-1.07) with no change at year 2 post tax. Sugar sold from sweets increased by 4% at both year 1 and year 2 post tax (year 2 posttax RIRR = 1.04; 95% CI, 1.03-1.06). There were no changes in stand-alone sugar sold. CONCLUSIONS AND RELEVANCEThis study using difference-in-differences analysis found a net 19% reduction in grams of sugar sold from taxed SSBs at year 2 post tax after accounting for changes in sugar sold from untaxed beverages, sweets, and stand-alone sugar. These results suggest that SSB taxes may effectively yield permanent reductions in added sugars sold from SSBs in food stores.

Systematic review on fiscal policy interventions in nutrition

“…The evidence from twelve studies (with 15 independent effect estimates) suggests taxes on SSBs reduced consumers purchasing such beverages ( µ = −0.18 [95%CI : −0.29 to −0.07] ; p = 0.001; Figure 5). We examined the studentized residuals and found that one study (26) had a value larger than 2.94 and may be a potential outlier in the context of this model. Our results remain significant when we removed this effect from the analysis ( µ = −0.07 [95%CI : −0.12 to 0.03] ; p = 0.001).…”

“…Based on evidence from 11 studies, our metaanalysis suggests there is no effect of taxes on SSBs on the purchasing of untaxed beverages ( µ = 0.02 (95%CI : −0.06 to 0.02) ; p = 0.33; Figure 6). Studies considered calories purchased (17), grams of sugar purchased (26), volume purchased (15,20,29,30,36,(39)(40)(41)(42). There was moderate heterogeneity in results (Q (10) = 19.61, p = 0.03, τ 2 = 0.00, I 2 = 49.01%).…”

Estimating the impact of nutrition and physical activity policies with quasi-experimental methods and simulation modelling: an integrative review of methods, challenges and synergies

“…In this context, QEM can provide evidence of the tax impact on: (i) firm response (product range and possible reformulations, tax pass-through and price); 72–76 and (ii) consumer response (including substitution patterns to other beverages and foods, at-home and out-of-home) 77 and, potentially sugar and other nutrient intakes. 78 , 79 One further key input needed in SM that QEM can provide are heterogeneous policy responses across firms and population sub-groups (e.g. CATE estimates).…”

Do sugar‐sweetened beverage taxes improve public health for high school aged adolescents?

“…Another strand of this literature looks at what people consume instead of SSBs when they become more expensive (Aguilar et al, 2021;Duffey et al, 2010;Finkelstein et al, 2013;Fletcher et al, 2010), with mostly ambiguous results which suggest that people may still be making other unhealthy choices. One important exception to this is Powell et al (2021) which finds that overall grams of sugar sold went down by 19% in Seattle following the implementation of its tax, suggesting that SSB taxes could in fact lead to permanent reductions in added sugars being purchased. Noting the importance of measuring changes in self-reported consumption in addition to purchasing behavior, others have used repeated cross-sectional surveys to assess whether people are drinking fewer SSBs.…”

“…Notably, it is free of the measurement error and response-bias which can plague survey data. It can also be useful for studying tax avoidance behavior when data is available from stores on either side of a taxed border (Bollinger & Sexton, 2018;Powell & Leider, 2020a, 2020bSeiler et al, 2021), and it can be used to study what people are consuming instead of SSB's when taxes cause increases in their prices (Aguilar et al, 2021;Finkelstein et al, 2013;Powell et al, 2021). One important issue with most scanner data, however, is that it tracks purchases made by a household head, and it cannot specify which family member is consuming which items.…”