Robert Pryce

Background: Alcohol and tobacco are risk factors for non-communicable disease and deaths in England. People using both substances have increased health risks. While increasing tax on each product has been shown to reduce consumption and harm, little research has examined effects of changing both taxes together. Objectives: Work package 1: describe potential tobacco and alcohol tax policies; work package 2: analyse purchasing patterns across population subgroups; work package 3: estimate industry tax-pass-through; consumer response to price changes for 12 products; work package 4: simulate impact of 33 policy options, United Kingdom Government 2021 proposals to restructure alcohol duty and (after project end date) 2023 alcohol duty reform. Design: Work package 1 reviewed literature and interviewed stakeholders to identify policy options. Work package 2 analysed repeat cross-sectional survey data on purchasing. Work package 3 used quantile regression on market research data to estimate tax pass-through and a two-stage Tobit model to estimate consumer responses (own-price and cross-price elasticities) for 12 products (beer, cider, wine, spirits and ready-to-drinks, split off-trade/on-trade; cigarettes, roll-your-own tobacco). Work package 4 developed tax modelling for the Sheffield Tobacco and Alcohol Policy Model (v2.1.0) to analyse work package 1 policy options. Later, version v2.4.2 was used to analyse emerging policy on alcohol duty reforms. Setting: Effects of United Kingdom tax policies in England, covering 2017–50. Participants: Life course simulation of individuals aged 18–89; 2017–50. Interventions: Tax policies changes: (1) specific duty per product, (2) duty escalators for alcohol and tobacco, and Minimum Excise Tax on tobacco, (3) hypothetical strength-based alcohol taxes, (4) United Kingdom Government’s 2021 proposals for alcohol duty reform and (5) analysis of United Kingdom Government’s implemented Alcohol Duty Reform 2023. Outcome measures: Changes in participation in drinking or smoking, amounts consumed, expenditures, Government tax revenues, retailer revenues, risks of hospitalisation and death from over 60 clinical conditions, National Health Service secondary costs, quality-adjusted life-years, mortality and health inequalities. Results: Work package 1 identified policy options. Work package 2 showed variations in expenditure by subgroup. Work package 3 showed that retailers increased prices of expensive products by more than expected after tax rises, keeping cheap product prices lower. Work package 3 showed statistically significant participation and consumption price responsiveness for all 12 products. Work package 4 compared 33 policies, showing that higher tax increases, especially for cider and hand-rolled tobacco, are effective at reducing smoking rates, drinking and deaths over 20 years. Effects were largest for people living in more deprived areas. Tailored modelling (after project end date) examined United Kingdom Government’s 2023 implemented alcohol duty reform, estimating a small reduction in alcohol consumption and deaths. In contrast, a policy to increase cider taxes to be in line with beer could result in a 30 times larger reduction in deaths. Limitations: Uncertainty in estimates remains. Data used were collected pre-coronavirus disease discovered in 2019. Sales data were from legal retailers; illicit tobacco was not examined. Each sub-study has its own data limitations. Conclusions: Combined increases in tax on tobacco and alcohol could potentially be more effective at reducing disease, National Health Service costs, deaths and health inequalities than raising tax on only one product. Potential trade-offs could exist between health benefits and effects of tobacco and alcohol tax increases on financial burden for those who continue consuming both products. Future tobacco or alcohol tax analyses should factor in inter-linkages between both commodities. Future work: We are adapting modelling infrastructure to deliver responsive within-days modelling for new government proposals and budgets. Further work on price elasticities in other countries would be useful. Future analyses will incorporate post-coronavirus disease discovered in 2019 data. We are extending modelling to other United Kingdom countries, and pricing policies beyond tax (e.g. minimum pricing). We are extending analyses to no and low alcohol products and to evaluate observed impacts of the government’s alcohol duty reforms. Funding: This synopsis presents independent research funded by the National Institute for Health and Care Research (NIHR) Public Health Research programme as award number 16/105/26.

Introduction: The increasing popularity of alcohol‐free and low‐alcohol (NoLo) drinks could reduce alcohol‐related harm if those at greatest risk replace their standard alcoholic drinks with these products. This paper aimed to identify: (i) characteristics associated with occasional purchase of NoLo drinks in Great Britain in 2021; and (ii) whether these characteristics changed between 2018 and 2021. Methods: Logistic regression using data from Worldpanel by Numerator dataset detailing alcoholic and NoLo drink purchases in the off‐trade (i.e., shops) by households in Great Britain in 2018 ( n = 14,702) and 2021 ( n = 15,257). The primary outcome was occasional NoLo purchasing (≥ 4 times in the past year), and secondary outcomes were occasional NoLo beer, cider or wine purchasing. Results: In 2021, 5.7% of households purchased NoLo ≥ 4 in the past year, compared to 3.0% in 2018. In 2021, these households were more likely to be alcohol purchasers than non‐purchasers [low‐risk purchasers (≤ 112 g per week): OR = 7.11, p < 0.001; increasing risk (113–280 g per week): OR = 10.72, p < 0.001; higher risk (> 280 g per week): OR = 12.10, p < 0.001] and less likely to be from lower social grades than grade AB (Grade C2: OR = 0.72, p = 0.008; Grade D: OR = 0.58, p < 0.001; Grade E: OR = 0.46, p < 0.001). Results were similar for occasional purchasers of NoLo beer, but there was no significant association with social grade for purchasing NoLo cider or wine. The characteristics associated with occasional NoLo purchasing did not change significantly between 2018 and 2021. Discussion and Conclusions: Households from higher social grades who purchase more alcohol are more likely to regularly purchase NoLo products.

Background and aims: Health economic models can be used to assess the effectiveness and cost‐effectiveness of public health policies for gambling. To develop such a model, we must understand how gambling behaviour is associated with risk of experiencing gambling‐related harms. This study aimed to: (1) assess the strength of association between gambling frequency and the risk of gambling‐related harm and to examine how these associations differ when lottery‐only players are excluded; (2) apply the study's findings in a hypothetical policy model aimed at reducing gambling frequency. Design: Observational study using six waves of cross‐sectional data from the Health Survey for England and the Scottish Health Survey. Setting: Survey conducted in England in 2015, 2016 and 2018 and Scotland in 2015, 2016 and 2017. Participants: The sample included 16 648 adults (aged 18 and over) who reported gambling in the past year, generally representative of the populations of England and Scotland. Measurements: Gambling frequency was measured using 6 categories which indicated frequency in the past 12 months: (a) 2 or more times a week; (b) once a week; (c) less than once a week, more than once a month; (d) once a month; (e) every 2–3 months; (f) once or twice a year. Risk of gambling‐related harm was assessed using Problem Gambling Severity Index (PGSI) score (0–27) and its four categories: no‐risk (0), low‐risk (1–2), moderate‐risk (3–7) and high‐risk (≥8). Control variables included age, sex, deprivation, social grade, presence of mental disorder and frequency of drinking alcohol. Findings: Using multinomial logistic regression and zero‐inflated negative binomial models we found that gambling at least twice weekly was associated with a statistically significantly higher PGSI score than gambling once or twice a year (incidence rate ratio = 3.528, 95% confidence interval = 2.040–6.103, P value < 0.001). Reducing gambling to guideline levels for people gambling at least twice weekly moved 10% of the sub‐sample from higher PGSI categories (low, medium and high risk) to the no‐risk category and shifted the distribution of PGSI scores down. Conclusions: There appears to be a statistically significant association between gambling frequency and risk of gambling‐related harm. Data derived from this and similar analyses can be used to model gambling policies which impact gambling frequency.

Background/Aims: To conduct an exploratory analysis of the association between the mental health and emotional wellbeing of an individual and the Problem Gambling Severity Index (PGSI) scores of their spouse or cohabiting partner. Design: Observational study using two sources of secondary data: the Health Survey for England (HSE) and the Scottish Health Survey (SHeS). Data from 2012, 2015, 2016, 2017 (SHeS only) and 2018 (HSE only) were combined to create a single data set. The data contained household identifiers and a household grid, allowing cohabiting couples to be identified. Setting: England and Scotland. Participants: 20 752 individuals (in 10 376 couples) where both partners had completed the PGSI. Measurements: Outcome measures were four self‐reported measures of emotional wellbeing: a binary measure indicating a long‐term mental health condition, a scale question on life satisfaction, the twelve‐item General Health Questionnaire (GHQ‐12) and the Warwick Edinburgh Mental Health and Wellbeing Scale (WEMWBS). Gambling harms were measured using the PGSI. Controls included socio‐demographic/economic characteristics, and a binary variable indicating whether the individual had spent money in the last 12 months on gambling. Findings: Regression models showed a statistically significant association between lower emotional wellbeing, measured by WEMWBS [Coef. 0.022; 95% confidence interval (CI) = 0.004–0.040], GHQ‐12 Likert scale (Coef. 0.021; 95% CI = 0.000–0.043) and life satisfaction (Coef. 0.036; 95% CI = 0.005–0.067), among individuals who themselves had a PGSI score of zero but whose partner had elevated PGSI scores, when controlling for the individual's own gambling participation and other socio‐demographic household and individual characteristics. There was no evidence of an effect between partner's PGSI score and an increased likelihood of a long‐term mental health condition amongst the same group. Conclusions: Decrements to an individual's emotional wellbeing appear to be strongly associated with increased Problem Gambling Severity Index (PGSI) scores among their spouse/cohabiting partner, with an individual's emotional wellbeing declining as their partner's PGSI score increases.

Background and Aims: Gambling is a public health issue and widespread advertising of gambling products may contribute to gambling harms. Sports‐related gambling advertising includes advertising around sports games or for sports betting products. This review aimed to provide the most systematic and up‐to‐date review of the literature on the association between sports‐related gambling advertising and gambling behaviour.Methods: A systematic literature search of quantitative studies up to 13 February 2024 was undertaken following PRISMA guidelines. Inclusion criteria were quantitative studies published in the English language exploring the association between sports‐related gambling advertising and gambling behaviour. Traditional database searches (Medline, Scopus, PsychInfo, Web of Science, CINAHL and The Cochrane Library) were undertaken alongside citation, author and website searches. Studies were narratively synthesised, and the overall quality of the evidence was assessed using the Mixed Methods Appraisal Tool (MMAT).Results: Twenty‐two studies were included in this review covering traditional, digital, direct, embedded, inducement and aggregate advertising. The majority (n = 16) of research was undertaken in Australia on adult populations. Results suggest that sports‐related gambling advertising is associated with increases in perceived, intended and actual frequency of (n = 6 studies) and expenditure on (n = 3) gambling, unplanned or unintended gambling (n = 2), the likelihood of gambling (n = 2), the likelihood of using a sponsor's product (n = 2) and, in some cases, the complexity or riskiness of bets placed (n = 2). Studies suggest that the self‐reported effect may be more pronounced in higher‐risk gamblers (n = 7). Preliminary evidence suggests that specific inducements which reduce the riskiness or cost of gambling appear to be particularly influential (n = 3). Limitations of the evidence base include the lack of standardised measures and use of observational designs.Conclusions: Exposure to sports‐related gambling advertising appears to be associated with increased gambling behaviour for a wide range of advertising media. This association may be more pronounced in higher‐risk gamblers who are already at increased risk of harm.

Background and Aims: Buying smaller‐sized alcohol products can reduce alcohol consumption, but larger products have better value for money, which presents a barrier to switching. We tested whether proportional pricing prompts drinkers to buy smaller alcohol products and reduce alcohol purchasing.Design, Setting and Participants: This study was an online experiment set in the United Kingdom, using hypothetical shopping tasks in which participants purchased different‐sized products presented under proportional pricing (i.e. constant price per litre throughout all sizes of the same product) or standard pricing conditions. Study 1 (comprising n = 210 participants) was a mixed experiment with pricing condition (proportional pricing, standard pricing; within‐subjects) and drink type (lager, red wine, vodka; between‐subjects) as manipulated factors. Study 2 (comprising n = 90 participants) was a within‐subjects experiment with pricing condition (proportional pricing, standard pricing) and multi‐pack type (size difference‐only, quantity‐difference only, size and quantity difference) as manipulated factors. Participants were UK adult alcohol consumers.Measurements: We measured outcome variables, including alcohol purchasing (UK units) and proportion of alcohol purchased from smaller products.Findings: Proportional pricing consistently increased the proportion of alcohol purchased from smaller products [study 1: B = 10.82, 95% confidence interval (CI) = 8.72–12.92; study 2: B = 11.64, 95% CI = 3.50–19.77], indicating a switch to smaller products. However, this did not consistently reduce the total amount of alcohol purchased among drink and product types: proportional pricing reduced the total units purchased from lager multi‐packs containing more rather than fewer products (B = −2.56, 95% CI = −4.82 to −0.30), but not from other types of lager multi‐packs or single lager products. Proportional pricing also reduced vodka purchasing (B = −3.30, 95% CI = −5.21 to −1.40), but the effect of proportional pricing on wine purchasing was moderated by hazardous drinking (B = −0.11, 95% CI = −0.17 to –0.05).Conclusions: Alcohol sales policies that require proportional pricing may reduce alcohol purchasing.

Background: Alcohol consumption changed substantially during the COVID-19 pandemic for many people. This study quantified how these changes in drinking varied across the population and their potential longer-term impact on health and health inequalities. Methods: We analyzed data from the Alcohol Toolkit Study to estimate how alcohol consumption changed during the pandemic (April 2020–November 2021) and how these changes varied with age, sex, drinking level and socioeconomic position. We combined these estimates with a range of alternative scenarios of future alcohol consumption and used the Sheffield Alcohol Policy Model to assess the long-term health and health inequality impacts of these changes. Results: Alcohol consumption in 2020–21 increased in heavier drinkers but fell in moderate drinkers. If alcohol consumption returns to pre-pandemic levels in 2022, we estimate a total of 42 677 additional hospital admissions and 1830 deaths over 20 years because of these changes. If consumption remains at 2021 levels in the long-term these figures rise to 355 832 and 12 849, respectively. In all scenarios, the biggest increase in harm occurs in the most deprived 20% of the population. Conclusions: Pandemic-era changes in alcohol consumption are likely to have a significant negative impact on public health and health inequalities, even under optimistic assumptions about future drinking.

Introduction: Evidence shows that price is an important policy lever in reducing consumption of alcohol and tobacco. However, there is little evidence of the cross‐price effect between alcohol and tobacco.Methods: This paper uses an econometric model which estimates participation and consumption elasticities, on data from the UK Living Costs and Food Survey 2006–2017 and extends the literature by, for the first time, estimating joint price elasticities for disaggregated alcohol and tobacco products. This paper presents new price elasticities and compares them to the existing literature.Results: The own‐price elasticity estimates are all negative for both participation and consumption. There is no pattern to the estimates of cross‐price elasticities. The elasticity estimates, when used in the Sheffield Tobacco and Alcohol Policy Model, produce bigger changes in consumption for the same change in price compared to other elasticity estimates in the existing literature.Discussion and Conclusions: Consumption of alcohol and tobacco are affected by the prices of one another. Policymakers should bear this in mind when devising alcohol or tobacco pricing policies.

AimsEvidence exists on the potential impact of national level minimum unit price (MUP) policies for alcohol. This study investigated the potential effectiveness of implementing MUP at regional and local levels compared with national implementation. DesignEvidence synthesis and computer modelling using the Sheffield Alcohol Policy Model (Local Authority version 4.0; SAPMLA). SettingResults are produced for 23 Upper Tier Local Authorities (UTLAs) in North West England, 12 UTLAs in North East England, 15 UTLAs in Yorkshire and Humber, the nine English Government Office regions and England as a whole. CasesHealth Survey for England (HSE) data 2011-13 (n = 24 685). MeasurementsAlcohol consumption, consumer spending, retailers' revenues, hospitalizations, National Health Service costs, crimes and alcohol-attributable deaths and health inequalities. FindingsImplementing a local 0.50 pound MUP for alcohol in northern English regions is estimated to result in larger percentage reductions in harms than the national average. The reductions for England, North West, North East and Yorkshire and Humber regions, respectively, in annual alcohol-attributable deaths are 1024 (-10.4%), 205 (-11.4%), 121 (-17.4%) and 159 (-16.9%); for hospitalizations are 29 943 (-4.6%), 5956 (-5.5%), 3255 (-7.9%) and 4610 (-6.9%); and for crimes are 54 229 (-2.4%), 8528 (-2.5%), 4380 (-3.5%) and 8220 (-3.2%). Results vary among local authorities; for example, annual alcohol-attributable deaths estimated to change by between -8.0 and -24.8% throughout the 50 UTLAs examined. ConclusionsA minimum unit price local policy for alcohol is likely to be more effective in those regions, such as the three northern regions of England, which have higher levels of alcohol consumption and higher rates of alcohol harm than for the national average. In such regions, the minimum unit price policy would achieve larger reductions in alcohol consumption, alcohol-attributable mortality, hospitalization rates, NHS costs, crime rates and health inequalities.