Table 3

Natural experimental evaluation of the UK treasury soft drinks industry levy

Step in pathwayPrécis of protocol (published in 2017)24
Observational evidenceSugar-sweetened beverage (SSB) consumption is independently associated with multiple chronic disease outcomes. SSBs currently represent the single biggest source of dietary sugar for young people in the UK. Economic theory and data strongly suggest that price is an important determinant of SSB purchases.
Policy developmentTo reduce population consumption of SSBs, a range of interventions had been proposed, including fiscal measures. Globally a number of SSB taxes had been introduced, although few had been evaluated when this study was initiated. Modelling studies suggested important potential health gains, but no comprehensive evaluations had measured impacts on reformulation or consumption.
Policy actionIn 2016, the Chancellor of the Exchequer announced a tiered soft drinks industry levy (SDIL) on industries importing or selling SSBs in the UK with the explicit intention of reducing consumption of sugar from SSBs. At the time of announcement, the UK SDIL was different from other SSB taxes: it was an industry levy (paid by manufacturers) rather than an excise tax (paid by consumers).
Evaluability assessmentThe implementation of a fiscal policy is an intervention that is highly context dependent, resulting in reactions by many stakeholders including government, civil society, industry, health sector and consumers, and the potential to affect a range of diet and health outcomes. The SDIL was unique in its construction including a tiered levy directed at industry, and its 2-year lead time from date of announcement to implementation. Randomised controlled trials are recognised as the strongest method for determining causal effects. However, in the current context where the SDIL was introduced to the whole country at once, randomisation to intervention and control groups was not feasible.
Design of natural experimental evaluationThis evaluation seeks to improve our understanding of how such interventions evolve over time within complex food systems to influence products and purchasing, consumption and health outcomes. The evaluation will thus take a ‘systems’ perspective, aiming to evaluate a range of outcomes, associated processes and their dynamic interrelationships. Interrupted time series (ITS) methods offer one of the strongest quasi-experimental research designs. Using ITS designs, consideration of a range of outcomes (eg, SSB consumption declining as consumption of lower sugar alternatives increase) and mechanistic processes (eg, the relationship between price and purchases) can be explored such that a ‘pattern’ of impacts is appraised to provide the strongest possible basis on which to draw causal inference.
Evidence synthesisFindings will be integrated and synthesised to develop a coherent overarching interpretation (and) test and refine the underlying intervention theory for the SDIL. Findings generated using different methods (qualitative, quantitative) could be triangulated to explore the extent to which they provide a consistent interpretation and conclusions about the impacts of the SDIL using pattern matching and causal process observation, thus strengthening causal inference.