Minimum alcohol pricing in Scotland was linked to 13% fewer alcohol deaths

Setting a floor price on alcohol appears to save lives, and to do so most in the poorest communities: after Scotland introduced minimum unit pricing, deaths wholly attributable to alcohol fell by an estimated 13.4%, according to a controlled evaluation published in The Lancet [s1]. The effect on hospital admissions was smaller, at 4.1%, and did not reach statistical significance — a distinction that matters for what the policy can and cannot claim [s1].

What the policy is

Since 1 May 2018, every alcoholic drink sold in Scotland has carried a minimum price per unit of alcohol, initially set at £0.50, under an order made by Scottish ministers [s2]. Minimum unit pricing (MUP) works differently from a tax: it sets an absolute price floor tied to the amount of pure alcohol in a product, so it bites hardest on cheap, high-strength drink — the white ciders and value spirits associated with the heaviest, most harmful drinking — while leaving mid-priced and premium products largely untouched. The policy was designed as a targeted instrument, not a broad levy.

The floor was raised from £0.50 to £0.65 per unit from 30 September 2024, and the policy — originally passed with a sunset provision — was continued rather than allowed to lapse [s3].

How the study tested it

The Lancet evaluation used a controlled interrupted time series design, a strong quasi-experimental method for judging a policy that cannot be randomised [s1]. It tracked deaths and hospitalisations defined as wholly attributable to alcohol — conditions that do not occur without alcohol, such as alcoholic liver disease — using routinely collected records [s1]. The pre-intervention period ran from 1 January 2012 to 30 April 2018, and outcomes were followed for 32 months after implementation, to 31 December 2020 [s1]. England, where MUP was not introduced, served as the control group, helping separate the policy's effect from trends affecting the whole of Britain [s1].

That control is the design's core strength: if alcohol deaths were already falling for other reasons, England should show it too, and the Scottish "effect" is what remains after subtracting the English trend.

What it found

MUP was associated with a significant 13.4% reduction in deaths wholly attributable to alcohol (95% confidence interval −18.4% to −8.3%; p=0.0004) [s1]. Hospitalisations wholly attributable to alcohol fell by 4.1%, but the confidence interval (−8.3% to 0.3%) crossed zero and the result was not statistically significant (p=0.064) — meaning the data are consistent with a reduction of up to roughly 8%, but also with essentially no change [s1].

The benefits were concentrated in chronic conditions, particularly alcoholic liver disease, rather than acute harms [s1]. And the reductions in both deaths and hospitalisations were seen in the four most socioeconomically deprived deciles of Scotland — the groups that bear the greatest burden of alcohol harm [s1]. Because alcohol deaths are steeply patterned by deprivation, a policy that works hardest there is tackling an inequality, not just an average.

An earlier strand of research had already indicated the mechanism was plausible: previous studies found the policy reduced alcohol sales by about 3% [s1]. A price floor that cuts consumption of the cheapest, strongest drink, followed by a fall in the chronic diseases that heavy drinking causes, is a coherent causal chain.

The limits worth stating

This is strong evidence for a public-health measure, but it is not a randomised trial, and three cautions apply. First, the mortality finding is robust while the hospitalisation finding is not; reporting the 13.4% without the non-significant 4.1% alongside it would overstate the case [s1]. Second, the follow-up window included 2020, when the COVID-19 pandemic disrupted both drinking patterns and health services, and although the England control absorbs some of that shock, no design removes it entirely. Third, "wholly attributable" outcomes are the cleanest to measure but exclude the larger pool of partly alcohol-related harm, so the estimate captures one well-defined slice of the total effect [s1].

None of this undercuts the central result. It sets its size: a measurable reduction in the deaths most directly caused by alcohol, greatest where harm is greatest, from a single, specific price policy.

How it fits the wider evidence

Minimum unit pricing is one of a small set of population-level measures — alongside taxation, availability limits and marketing restrictions — that the evidence consistently ranks above education campaigns for reducing alcohol harm. It approaches the problem from the price side, where guidance bodies still disagree even on what a "safe" level of drinking is and where the link between alcohol and cancer runs through biology many drinkers never hear about. The logic parallels other fiscal public-health tools whose real-world records this site has examined, from the healthcare costs of smoking that underpin tobacco policy to a sugary-drink tax that was introduced and then repealed.

What to watch

The open questions are whether the higher 65p floor delivers further reductions or runs into diminishing returns, whether the mortality benefit holds over a longer, post-pandemic window, and whether other jurisdictions weighing MUP see the same concentration of benefit in deprived communities. The Scottish case is now one of the better-evaluated natural experiments in alcohol policy — which is precisely why its non-significant results deserve as much attention as its headline one [s1].

Sources

• [s1] Evaluating the impact of alcohol minimum unit pricing on deaths and hospitalisations in Scotland: a controlled interrupted time series study, The Lancet, 20 March 2023. https://doi.org/10.1016/S0140-6736(23)00497-X
• [s2] The Alcohol (Minimum Price per Unit) (Scotland) Order 2018 (SSI 2018/135), legislation.gov.uk. https://www.legislation.gov.uk/ssi/2018/135/made
• [s3] The Alcohol (Minimum Price per Unit) (Scotland) Amendment Order 2024 (SSI 2024/128), legislation.gov.uk. https://www.legislation.gov.uk/ssi/2024/128/made