Alcohol (Minimum Pricing) (Scotland) Act 2012 - operation and effect 2018 to 2023: report

Footnotes

1. This is often cited as the 'alcohol harm paradox', Bellis MA, Hughes K, Nicholls J, Sheron N, Gilmore I, Jones L. The alcohol harm paradox: Using a national survey to explore how alcohol may disproportionately impact health in deprived individuals. BMC Public Health. 2016 Feb 18;16(1):111.

2. A natural experiment method is a type of social research method that can be used where the division of the population into exposed and unexposed groups is outside of the researchers' control. This method typically exploits the timing and/or location of a change such that it occurs for some places or groups of people, creating an exposed group, but does not occur in similar places or groups of people, thus creating a control group. MUP being implemented in Scotland on 1 May 2018, but not in England & Wales, is a good example of this (PHS 2023: 9).

3. There were EAGs for compliance; economic impact and price; children and young people; harmful drinking; consumption and health harm; and evidence synthesis.

4. These are: premises license holders and alcohol producers as well as those with a function related to health, prevention of crime, education, social work, and children and young people. These are the same broad groups which the Scottish Ministers were required to consult while preparing this report.

5. There were originally seven separately funded studies. However, lengthy delays in providing updated approvals for access to the linked SHeS-SMR data have prevented completion of one study using the Scottish Health Survey.

6. Realist synthesis is designed to allow evaluators to understand the effects of complex interventions in complex systems. The method focuses on explaining what it is about an intervention that contributes to an outcome, why and how it does so, for whom, and in what contexts. Where typical systematic reviews aim to control all factors other than the intervention and the outcome, which is valuable in evaluating clinical effectiveness, a realist approach to synthesis acknowledges that the context-sensitive responses of individuals to a social intervention are an inextricable part of the mechanisms of that intervention, rather than something that can be controlled for when seeking to understand the adaptive and context-sensitive responses of individuals to a social intervention. The realist approach is particularly valuable in examining complex social interventions where randomised controlled trials are impractical and reviews are reliant on evidence using more diverse, less controlled research methods. (Beeston et al. 2022: 5-6)

7. Process tracing is a method for tracing the causal mechanisms by which an observed outcome was produced by an intervention, and involves the use of systematic and transparent tests to establish the extent to which a set of causal mechanisms can be validated by the evidence, and the relative validity of competing causal explanations of the outcome. Process tracing is often used within a realist framework to improve evaluators' ability to test the mechanisms they identify. It helps evaluators to produce robust high-level assessments of causal mechanisms and complements the strength of realist synthesis in producing detailed, low-level understandings of specific social groups and contexts. (Beeston et al. 2022: 6)

8. Contribution analysis is a method that 'explores attribution through assessing the contribution a programme is making to observed results. It sets out to verify the theory of change behind a programme and, at the same time, takes into consideration other influencing factors. Causality is inferred from the following evidence: 1. The programme is based on a reasoned theory of change: the assumptions behind why the program is expected to work are sound, are plausible, and are agreed upon by at least some of the key players. 2. The activities of the programme were implemented. 3. The theory of change is verified by evidence: the chain of expected results occurred. 4. Other factors influencing the programme were assessed and were either shown not to have made a significant contribution or, if they did, the relative contribution was recognised.' (Mayne 2008: 1)

9. Lengthy delays in securing updated approvals for access to the necessary linked data have delayed completion of this study. The study is still in progress, and is unable to provide a timeline for completion at this stage. The Scottish Government will consider the findings of the study when they become available.

10. Including those with current or previous experience of homelessness.

11. The social value has been calculated using the average annual number of deaths averted by MUP (estimated by Wyper et al (2023)) and the value of a prevented fatality (VPF), calculated by the Department for Transport (DfT) to be just over £1.9m at 2020 prices. The evaluation noted that a 'systematic review of values of VPF recommends that a VPF from the relevant country be used if it exists. Guidance issued by UK Government HM Treasury on how to appraise policies, The HM Treasury Green book, uses the DfT VPF, and it has been used in other health economic evaluations' (PHS, 2023: 90). The evaluation also highlights that the definition of VPF is 'how much individuals are willing to pay for a very small reduction in the probability of death, paid for by forgoing the consumption of other goods and services. It is a measure of the value of reduced risks of death in the population as a whole arising from public policy decisions. It should not be interpreted as how much a (known) life is worth' (PHS, 2023: 90).

12. These values have been calculated based on an estimate of the mean total cost of an admission to hospital following attendance at the emergency department, calculated in Parkinson et al (2016), and subsequently updated by the evaluators using the Bank of England inflation calculator. This resulted in an estimated value of £990.57 per admission at 2020 prices.

13. Kantar Alcovision data is self-reported consumption survey data. The final evaluation report notes the following in relation to this type of data: 'Self-report surveys may be subject to biases as a result of sampling, incorrect recall or social desirability, and reaching the heaviest drinkers to take part in surveys may be particularly challenging. All the surveys described below are cross sectional, which means that different people are surveyed at each wave. Sampling errors that result in systematic differences between samples are a particular issue for cross-sectional surveys. However, survey data does allow disaggregate analysis by characteristics at an individual level.' (PHS 2023: 45)

14. A heavy drinking session in which someone drinks a lot of alcohol in a short period of time, raising their risk of harm on that occasion. Typically, this is defined as those who drink at least weekly, consuming 6 or more units for women, and 8 or more units for men, on a single occasion.

15. Alcohol misuse was defined as a score exceeding 2 on the Fast Alcohol use Screening Test (FAST). (Hodgson R, Alwyn T, John B et al. The FAST Alcohol Screening Test. Alcohol. 2002 Feb; 37(1): 61-66.

16. These were: prevalence of drinking in the last seven days; number of drinking days; number of units consumed and the prevalence of exceeding the daily limit on the heaviest drinking day.

17. Alcohol misuse was defined as a score exceeding 2 on the Fast Alcohol use Screening Test (FAST). (Hodgson R, Alwyn T, John B et al. (2002) The FAST Alcohol Screening Test. Alcohol, 37(1):61–66. https://doi.org/10.1093/alcalc/37.1.61)

18. These were: prevalence of drinking in the last seven days; number of drinking days; number of units consumed and the prevalence of exceeding the daily limit on the heaviest drinking day.

19. This survey sample was heavily weighted to the younger end of the age spectrum with 65–70% of respondents below 30 years in both Scotland and England,

20. Alcohol misuse was defined as a score exceeding 2 on the Fast Alcohol use Screening Test (FAST). (Hodgson R, Alwyn T, John B et al. (2002) The FAST Alcohol Screening Test, 37(1):61–66.

21. As part of the interview, participants were asked to complete a retrospective diary recalling the alcohol they had purchased and consumed in the last typical drinking week before treatment, using a method called Time Line Follow Back (TLFB).

22. Adjusted significance threshold after sample weighting of p=0.0004630.

23. The final evaluation report states that: 'The DRS as it is currently proposed would add a deposit of £0.20 on to every single-use drinks container, including each single item within a multipack and regardless of item size. The deposit would be refunded when the container is returned for recycling through an approved channel. DRS thus has the potential to interact with the MUP pricing structure at point of purchase. Lower-strength alcohol, such as beer and cider, are more likely to be sold in multipacks while higher-strength alcohol, such as spirits and wine, tend to be sold in single containers. There is a risk that DRS incentivises a move towards larger, single containers and higher-strength alcoholic products. The extent to which this will influence consumers' purchasing decisions and industry packaging is unknown.' (PHS 2023: 60)

24. The final evaluation report notes that 'Shopping panel data are self-reported, and therefore less reliable than automatically collected EPoS data, but do not rely on retrospective recall in the same way that surveys do.' (PHS 2023: 36)

25. The final evaluation report notes that 'while the price of alcohol purchased is not strictly the same as the price of alcohol available, it provides a proxy for compliance, and these analyses can be taken as strong evidence that retailer compliance with MUP was high, with no time lag.' (PHS 2023: 37)

26. The five key performance metrics are: the number of enterprises and business units; employment; turnover; gross value added (GVA); and output value.