How should we evaluate sweetened beverage tax policies? A review of worldwide experience

Abstract

Over 45 jurisdictions globally have implemented sweetened beverage taxes. Researchers and policymakers need to assess whether and how these taxes change beverage demand and supply, their intended and unanticipated health, economic and equity impacts. Lessons from such evaluations can maximise the policies’ success and impact on non-communicable disease prevention globally. We discuss key theoretical, design and methodological considerations to help policymakers, funders and researchers commission and conduct rigorous evaluations of these policies and related disease prevention efforts. We encourage involving the perspectives of various stakeholders on what evaluations are needed given the specific context, what data and methods are appropriate, readily available or can be collected within time and budget constraints. A logic model /conceptual system map of anticipated implications across sectors and scales should help identify optimal study design, analytical techniques and measures. These models should be updated when synthesising findings across diverse methods and integrating findings across subpopulations using similar methods.

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

“…lost productivity and environmental impact) of policies and are often applied within health-economic modelling to compare multiple policy scenarios, generating valuable information for priority setting. 4 , 40 Finally SMs can provide policy impact corridors by simultaneously incorporating uncertainties from multiple sources. 41 , 42 …”

“…Hence, quasi-experimental methods (QEM) using observational data for policy evaluation have become increasingly popular ( table 1 ). 4 Despite the availability of this quantitative toolbox, which is successfully applied in the social sciences, especially labour economics (see the 2021 Nobel prize in Economics, Royal Swedish Academy of Sciences, 2021), its application to identify causal effects of nutrition and physical activity policies on health outcomes is complex and potentially not fully exploited. 5 …”

“… 6 Consequently, mathematical disease simulation models (SMs) projecting the long-term health and economic consequences are increasingly considered by scholars and policy makers ( table 1 ). 4 , 7 , 8 …”

Simulation models of sugary drink policies: A scoping review

“…Fifty out of 61 included articles examined at least one tax policy ( Fig 2 ). To fully characterize taxes, we extracted information on both the tax rate (e.g., ad valorem /percentage-based, such as 20%, or unit-based, such as 1 cent-per-fluid-ounce) and how that tax rate would be implemented (i.e., excise, sales, other, unclear), based on the exact language used in the article [ 3 , 5 ]. Ad valorem tax rates were the most commonly examined tax (n = 25), with most studies of tax policies examining a 20% tax on SSBs (n = 20/25).…”

“…Simulation modeling is a powerful tool for projecting likely population health outcomes under different policy scenarios. Broadly, models use existing knowledge and data to project how consumer and supply-side behaviors (e.g., SSB consumption, product reformulation) and health outcomes (e.g., obesity, diabetes) are likely to change over time in response to policy actions [ 5 ]. Modifying model parameters allows investigators to examine different ‘what if’ scenarios, such as how expected health impacts might differ if the policy was less effective, or if consumers or suppliers respond in particular ways.…”