Evaluating the impact of individual alcohol licensing decisions on local health and crime: a natural experiment with synthetic controls

Abstract

BackgroundLocal authorities in England can influence the local alcohol environment by contributing to the licensing process and controlling the enforcement of existing licenses. However, a gap remains in the availability of quantitative evidence of effectiveness and impact of these local interventions, including a shortage of public health evidence around individual premises. Natural experiments offer the opportunity to evaluate these interventions where formal randomisation is not possible. We aimed to assess whether it is possible to quantitatively evaluate three natural experiments of alcohol licensing decisions at small spatial scale. MethodsThree situations presenting a natural experiment were identified in different English Local Authority areas by public health or licensing practitioners: (i) the closure of a nightclub in 2013, (ii) the closure of a restaurant-nightclub following reviews in 2016, and (iii) the implementation of new local licensing guidance (LLG) in 2013/14. We obtained monthly numbers of reported incidents of emergency department admissions for alcohol-related reasons, ambulance call-outs, and various crimes at lower/middle-super-output-area level (from 2010–14 for case i, 2015–17 for case ii, and 2008–14 for case iii). Bayesian structural timeseries were used to compare trends to their counterfactuals, approximated by synthetic controls based on time series of the same outcomes in other, comparable, areas. FindingsClosure of the nightclub was associated with temporary reductions in antisocial behaviour (–18%; 95% Bayesian Credible Interval [BCI] –37 to –4); equivalent to 60 averted incidents in 4 months. Closure of the restaurant-nightclub was not associated with measurable changes in outcomes. There was some evidence that the LLG introduction was associated with a reduction in drunk and disorderly behaviour (–42%, 95% BCI –109 to 23), but this reduction equated to less than one incident per month. The unplanned end of the LLG might have contributed to an increase in domestic violence (11%, 95% BCI –10 to 35), corresponding to two additional incidents per month. InterpretationIt is possible to evaluate the impact of local alcohol policies, even at the level of individual premises, using this methodology. We provide quantitative evidence that local government actions to influence the local alcohol environment can have a positive impact on health and crime in the area but could also have unintended consequences. FundingThis work was funded by the National Institute for Health Research School for Public Health Research (NIHR SPHR), a partnership between the Universities of Sheffield, Bristol, Cambridge, Exeter, and University College London); the London School for Hygiene and Tropical Medicine; the LiLaC collaboration between the Universities of Liverpool and Lancaster and Fuse; and the Centre for Translational Research in Public Health, a collaboration between Newcastle, Durham, Northumbria, Sunderland and Teesside Universities. The views expressed are those of the authors and not necessarily those of the National Health Service, the NIHR, or the UK Department of Health.