The relationship between the price and demand of alcohol, tobacco, unhealthy food, sugar-sweetened beverages, and gambling: an umbrella review of systematic reviews

The profile of the leading causes of disease and death has changed over time, with conditions caused by commercial determinants of health progressively replacing infectious diseases, particularly in the global north [1‐3]. Alcohol consumption, smoking, and excess weight are among the top 10 risk factors for disability adjusted life years (DALYs) globally, accounting for almost a fifth of all DALYs and over a quarter of deaths in 2019 [3]. In high-income countries, this increases to over a quarter of all DALYs and almost a third of all deaths.

Although an individual might only have one risk factor of alcohol use, smoking, or excess weight, many of these risks co-occur. For example, smokers are around three times more likely to drink at risky levels and up to 1.6 times more likely to have a poor diet [4, 5]. Multiple risks also cluster among the most deprived [4], and are therefore an important consideration for health inequalities. Multiple risks have clinical implications. A recent meta-analysis identified that the risk of disease and death from multiple risks is large, and for some outcomes, synergistic [6]. For example, the combined effect of risky alcohol consumption and excess weight on liver disease/death is 1.6 times greater than the sum of each risk on its own (95% confidence interval [CI] = 1.3, 1.9). Synergistic interaction was also seen for co-occurring risky alcohol consumption and smoking and oral cancers, where the combined effect was 3.8 times greater than the sum of each risk on its own (95% CI = 2.6, 5.4). Clearly, there are behavioural and clinical synergies between alcohol, smoking, and excess weight, but importantly, there are also policy synergies.

Current public health approaches to tackle alcohol consumption, smoking, and excess weight are somewhat siloed, as reflected in the global strategies across these three risks which include actions on price, marketing, and availability for each commodity on its own, but makes no reference to joining up these approaches to tackle multiple risks together in a comprehensive approach [7]. Within governments, health departments often have separate teams and policy approaches for alcohol, tobacco, and excess weight, and in finance departments, duty on related commodities are usually decided on a case-by-case basis. Many of the civil society organisations, charities, and medical academics who perform a vital advocacy role with policymakers also operate in these silos. Nonetheless, the evidence for the most effective and cost-effective approaches to preventing and reducing harm are broadly the same for tobacco, alcohol, and excess weight: increasing product price, restricting marketing, and reducing availability [7]. It is no coincidence that these comprise three of the four P’s of the marketing mix, the fourth being product, because the problems are generated by the products of commercially mediated organisations [8]. A requirement to reduce health harm is frequently in direct opposition to the generation of shareholder value [9].

Similarities in policy approaches to prevent and reduce alcohol-, tobacco-, and obesity-related harm can be seen across those implemented in the UK. For example, tobacco is subject to an annual tax escalator [10], and a sugar-sweetened beverage (SSB) tax came into force in 2018 [11]. Alcohol had a tax escalator between the years of 2008 and 2013 during which time deaths caused entirely by alcohol consumption (the directly attributable mortality) reduced, only to increase again when the measure was repealed [12, 13]. There are potential merits in considering policy areas together. For example, in the USA, increases in cigarette tax were associated with decreases in alcohol consumption among smokers [14]. In a UK study, the price elasticity of demand (PED) for alcohol was estimated to be -0.56, which reduced to -0.26 when the price of food was included in the analysis [15]. This suggests that pricing policies for alcohol may be undermined either if retailers offset an increase in alcohol price by decreasing the price of food, or if consumers have more disposable income due to the reductions in food price which can be spent on alcohol. Although price increases in medium-sugar drinks in the UK were associated with reductions in alcohol purchasing, increases in the price of high-sugar drinks were associated with an increase in purchasing of lager [16]. Finally, Australian modelling suggests that increasing alcohol tax is the most effective and cost-effective action to reduce obesity [17].

There is coherent logic in combining policy approaches for alcohol, tobacco, unhealthy food, and SSBs. Firstly, increases in price, restrictions on marketing, and reductions in availability are the key determinants of consumption and harm across all these policy areas [7]. Secondly, success in one policy area may lead to increased opportunities for harm in another if approached in isolation. For example, if alcohol sports sponsorship is banned, this could free up opportunities for sports sponsorship by fast food brands. Finally, data and evidence clearly demonstrate that multiple risks cluster, particularly among the poor, and this clustering results in a greater risk of disease and death [4‐6]. As such, policy decisions that consider and respond to these risks together are likely to be more effective for preventing and reducing harm and narrowing health inequalities than policy decisions considering each risk in isolation.

Taxes on harmful commodities such as alcohol, tobacco, unhealthy foods, and SSBs are primarily levied to increase revenue but also play an important role in public health. Increasing the cost of the product relative to alternative spending choices and income reduces demand by decreasing affordability of the product which leads to lower levels of consumption and harm. Associated reductions in the cost of ill-health to society mean that well-designed taxes on harmful commodities can be a highly effective and cost-effective health intervention [7]. Standard economic theory predicts that a price increase will lead to a reduction in demand, which is typically measured using the PED (the percentage change in quantity demanded associated with a 1% change in price). Effectively designed taxes also provide a financial incentive for producers to reformulate their products to less harmful versions, for example, levying higher rates of tax on stronger alcohols might lead to reductions in product strength, and taxing excess fat in a product might lead to reductions in total fat content or replacement with alternative fats. Numerous reviews have examined the impact of price and taxes on consumption of alcohol, unhealthy food, SSBs, and smoking [18‐28]. Given the behavioural, clinical, and policy synergies across these commodities, it is opportune to synthesise existing evidence to understand the similarities and differences in consumer price responses and what this might mean for bringing together tax policies. To the author’s knowledge, there has been one umbrella review which has attempted to do this, however included countries in Latin America only [29]. Therefore, the aim of this umbrella review was to identify and summarise evidence from systematic reviews that report the relationship between price and demand or price and disease/death for alcohol, tobacco, unhealthy food, and SSBs in any country. Since gambling is a taxable commodity and there have been recent efforts to conceptualise gambling using a public health approach [30, 31], we also considered gambling.

The protocol for our umbrella review was pre-registered (PROSPERO CRD42023447429) [32] and our write up complies with the PRISMA 2020 reporting guidelines [33].

We undertook a comprehensive umbrella review of systematic reviews to identify the relationship between the price of alcohol, tobacco, unhealthy food, SSBs, or gambling, and demand or disease/death. Despite alcohol and tobacco being taxable commodities for centuries [81, 82], we identified fewer reviews in these areas compared to reviews on unhealthy food and SSBs which have only recently been viewed as taxable commodities [83, 84]. We did not identify any reviews for gambling. Published evidence on the effectiveness of tax as a measure to prevent and reduce gambling-related harms is limited, however its potential use as a public health tool has been debated in recent scientific papers and the results of an e-Delphi consensus study highlight them as a possibly effective public health policy [31]. Goyder and colleagues argued that gambling taxation might work in a similar way to alcohol, tobacco, and SSB tax [85], and Sulkunen and colleagues suggest that increases in gambling tax might make it less profitable to providers, which may reduce their interest in expanding [86]. Others have suggested that gambling taxes might have a harmful impact by increasing the losses of gamblers, and it is excessive losses which mediate gambling-related harm, not the gambling itself [87]. UK estimates of the PED for gambling range from -1.5 for remote gaming to -0.5 for betting pools [88].

Our review clearly demonstrates that, at the most basic level, price interacts with income to affect demand, such that price increases of alcohol, tobacco, unhealthy food, or SSBs are associated with reductions in demand. Such consistency in findings clearly highlights the fundamental role of tax as a public health policy. Our review suggests that a 10% increase in product price is associated with reductions in demand in the order of 9.% for SSBs, 6% for unhealthy food, 5% for tobacco, and 1% for alcohol. This inverse relationship was most consistently seen for alcohol, and least consistently for SSBs. The relationship between price and demand was seen among adult and child populations, with some review-level evidence to suggest that children and young people were more price responsive in the case of tobacco [67, 71], as might also be the case for SSBs [20], although most reviews did not explicitly make this comparison. Although most reviews included research from high-income countries, where studied, the inverse price-demand relationship was seen across low-, middle-, and high-income countries. There was a lack of research on heavy consumers. Although the reviews present a consistent overall picture of the impact of price/tax on demand, there was large variation in estimates across the individual studies which may relate to methodological differences across studies. Elasticity estimates also vary over time, with some suggestion that products have become less price elastic, which probably reflects increasing affluence as products become more affordable [89].

Reviews on tobacco, unhealthy food, and SSBs generally provided evidence that lower income, education, or socioeconomic status groups were more price responsive compared to their less deprived counterparts for example [27, 47, 55, 66], but no alcohol reviews included this evidence. Nonetheless, data from primary research studies suggests that more deprived drinkers are more price responsive [90, 91]. A common argument against the imposition of taxes is that they may have a proportionally greater financial impact on people with lower incomes relative to those with higher incomes. However, lower income groups as a whole reap greater health benefits [92, 93]. To the extent that lower income individuals are more price sensitive, they will be more likely to cut back on the intake of taxed commodities, often starting from a higher level of consumption, and thus experience greater health gains.

Raising taxes not only has direct public health benefits but can also generate considerable revenue for governments. If price increases do not respond proportionately to tax increases, (i.e. if the PED < 1), government revenue will increase when taxes increase because the decrease in consumption is more than offset by the extra tax paid by those who continue to purchase the taxed product. Our review suggests this is the case for alcohol, tobacco, and unhealthy food, however some of the PED estimates for SSBs exceed one, suggesting government revenue might reduce because of the greater demand response. Nonetheless, SSB taxes are not widely implemented so any revenue generated by their introduction is revenue that would not have otherwise existed. Increased revenue associated with tax increases may go some way in compensating for the societal and human costs associated with alcohol, smoking, and excess weight.

Risky alcohol use, smoking, and excess weight, commonly co-occur and cluster in the most deprived [4, 5]. Multiple risks also result in large risks of disease or death which are synergistic in the case of alcohol and excess weight for liver disease and alcohol and smoking for oral cancers [6]. From these perspectives (behavioural and clinical), there is clear rationale for joining up policy approaches to prevent and reduce harm. The results of our review further highlight the potential of joining up policy approaches, specifically taxation and price increases. While taxation shares the same mechanism of action for alcohol, tobacco, unhealthy food, and SSBs, it should be noted that the end-goal of policy implementation might differ. An appropriate goal for smoking is usually complete abstinence, whereas for alcohol, the aim is to decrease the number of people drinking at levels which increase their risk of health harm (noting abstinence might be the goal for people with alcohol dependence). Although the WHO have recently recognised that there is no safe level of alcohol consumption [94], in practice, most countries have developed low-risk drinking guidelines at levels above zero which are typically set at what is considered to be an acceptable level of risk [95‐97]. Diet however is multifaceted and may comprise aspects such as consuming fruit and vegetables, whole-grain high-fibre foods, and limiting sugar, fat, and salt. While this umbrella review has demonstrated the effectiveness of tax as a public health measure for alcohol, tobacco, unhealthy food, and SSBs in isolation, it has not been able to explore the impact of simultaneous tax increases across these commodities, namely because the evidence does not exist. Future research should aim to understand the potential public health impact of a holistic approach to tax policy spanning these commodities.

While risky alcohol consumption, smoking, and excess weight all represent a substantial public health burden in and of themselves, there are clear behavioural, clinical, and policy synergies across these risks. The evidence supports tax/price increases as effective policies for reducing demand, which suggests there might be merit in a joined-up approach to tax, with the treasury representing an arbiter of public health and the National Health Service (NHS) costs.