This section is divided into sections addressing the five questions outlined in the methods.
How (if at all) do particular health taxes change consumption behaviours and what do we know about the health-related impacts of such taxes?
Like make taxes on tobacco and alcohol products [
3,
10‐
12], the majority of taxes on healthy food and non-alcoholic beverages were intended to improve population health by reducing product consumption (see [
13]). Definitions of ‘unhealthy’ or ‘junk’ foods vary within included studies but were commonly defined to target foods high in fat, salt and/or sugar [
14]. In some cases, definitions included products high in caffeine or products that had been subjected to intensive processing, such as processed meat [
15]. For non-alcoholic beverages, the most common targets of taxes were SSBs, which can include soft drinks or soda, cordials, other sugar-added juices, and ‘isotonics’ [
16‐
18]. A small number of studies also included milk-based products (e.g. milk desserts [
19]) or full fat or high-sugar milk [
20,
21].
Taking a reduction in product consumption as the primary aim of these taxes, Table
1 summarizes the number of studies, by study design type, which found either positive health impacts or no/negative health impacts. Two modeling studies [
16,
18] have been included in counts of ‘positive’ and ‘negative’ impacts because they found both positive and negligible/negative health impacts. One mixed methods study using modeling and evaluation methods was also double-counted in Table
1 as it found both positive and negative health impacts [
17].
Table 1
Number of studies identifying positive health impacts by study design type
Modeling | 17 | 16 | 3 |
Experimental | 0 | 0 | 0 |
Evaluation | 8 | 4 | 4 |
Mixed methoda
| 1 | 1 | 1 |
Total | 26 | 21 | 9 |
Table
1 suggests that modelling studies (e.g. [
18‐
20]) were more likely to find a positive health impact than evaluations [
24‐
26], perhaps because these studies often model the impact of higher tax rates than those that have been evaluated.
Nonetheless, four evaluation studies identified positive health impacts of the (generally lower level) taxes they assessed. Evaluating the effect of the Danish fat tax (2011-2013) on risk of ischemic heart disease (IHD), Bodker and colleagues found marginal changes in population risk of IHD [
24]. Smed et al., also evaluating the Danish tax, used retail scanner data to estimate the impact of the tax on population risk of IHD, stroke and heart failures [
26]. Although the results for each disease varied, the study estimated there was a small overall reduction in mortality from non-communicable diseases (mostly in men and young women). Overall, the researchers estimated the tax averted or delayed 123 deaths per year, although given the absence of a control group, a causal link to the tax cannot be drawn [
26]. In another context, Fletcher and colleagues evaluated the impact of changes in soft drink taxes at state level (which were, on average, around 3%) in the United States on BMI, obesity, and overweight [
25]. Using nationally-representative data, the authors found that soft drink taxes had a statistically significant, albeit small effect (decrease) on BMI, obesity, and overweight. These three studies caution that low taxes on unhealthy products may influence consumption behaviour, however are unlikely to lead to substantial population health changes. In another American-focused study (although using a different national dataset from Fletcher et al.), Kim and Kawachi found that between 1991 and 1998, states without taxes on SSBs or snack foods, or states that had repealed a similar tax, were greater than four and 13 times as likely, respectively, than states with a tax to experience a relatively high increase in population obesity [
28].
The four included evaluation studies that found no, or negative, health impacts were conducted in the United States context and examined the effect of SSB taxes and weight-related measures (e.g. BMI or obesity) in young people. In contrast to their study above, which examined adult populations, Fletcher and colleagues found that current state SSB taxes in the United States had no significant effect on children’s weight, finding that in fact children consumed more calories from SSBs in states that had implemented an SSB tax than in states that had not (although this was not statistically significant) [
29]. The researchers posit that in this case, the consumers are likely not reacting to the small and possibly hidden taxes on SSBs. In a separate article [
30], Fletcher et al. again found existing SSB taxes did not significantly reduce weight in young people, which was attributed to youth substituting other high-calorie drinks such as whole milk. Using cross-sectional data on American adolescents, Powell and colleagues found no statistically significant associations between BMI and state-level SSB taxes in grocery stores and vending machines [
31]. Sturm et al. also examined existing SSB taxes in the United States and their impact on young people’s obesity. Using longitudinal data from an early childhood study, the authors found no significant relationship between current taxes (usually no higher than 4% in grocery stores) and children’s SSB consumption or obesity [
32]. In contrast to modelling studies which often model taxes at higher rates (and more often find positive health impacts), the above evaluation studies provide valuable insight into the effectiveness of existing taxes implemented at lower rates.
Table
2 summarizes the number of studies, by tax rate, distinguishing between rates of less than 20% and those of 20% or more (since this is the most commonly used threshold across the literature reviewed (e.g. [
21,
29,
30])) and product type, distinguishing between SSBs and food products. In total, 22 studies are included in Table
2. Again, certain studies are included more than once if they considered separately taxes of different rates or the health effects of tax rates on different products [
17,
20,
35]. Studies which did not make the tax rate explicit or which focused on taxes applied to both SSBs and food are excluded from the table. Studies involving taxes applied to sugar/sweeteners are classified as food product taxes.
Table 2
Number of studies identifying health impacts by tax rate and product
Tax rate of <20% SSBs | 3 | 5 |
Tax rate of 20% + SSBs | 8 | 0 |
Tax rate of <20% food products | 4 | 3 |
Tax rate of 20% + food products | 3 | 0 |
Total | 18 | 8 |
Taken collectively, the studies in Tables
1 and
2 suggest there is considerable evidence that taxes on SSBs and unhealthy food products can have positive health impacts. However, as Table
1 demonstrates, the majority of studies included in this review were based on modelling or experiments involving
potential taxes. This is despite the fact that instances of such taxes exist in many countries. For example, Finland, France, Latvia, and Hungary have implemented taxes on both foods and beverages high in added sugar [
36]; Portugal and Hungary have implemented taxes on products high in salt [
36], Hungary has implemented a tax on foods high in fat, and Denmark introduced (and later repealed) a tax on saturated fat [
36]. In addition, there have been several instances of taxes on sugar-sweetened beverages, including in Mexico, two US cities and various small island states [
37‐
39]. This suggests there are substantial opportunities for developing the available research evidence concerning the evaluation of the health impacts of taxes that have been implemented on food and beverages.
Table
2 shows that evidence in support of applying taxes to unhealthy products is strongest for taxes on SSBs set at a rate of 20% or more of the price (e.g. see [
13]). The evidence for health impact from lower taxes on SSBs is weaker, with the number of studies finding positive health impacts equal to those that found no positive impact. The evidence of taxes on food products is more mixed and difficult to assess since many of the studies involve complicated bundles of taxes (e.g. [
33]).
Of the small number of studies that commented on the relationship between the type of tax applied and health impacts, there was a consistent finding that specific taxes (i.e. a fixed value based on the quantity, size or weight of the product) are associated with stronger health benefits than
ad valorem taxes, which are proportional to the price. Applying specific taxes means that all products covered by the tax are taxed equally. In contrast,
ad valorem taxes mean that more expensive, premium products attract a higher tax, which tends to increase price differences across brands, providing more scope for consumers to respond to new or higher taxes by selecting a cheaper brand or version (e.g. [
37]). This is a finding which parallels evidence regarding tobacco taxes [
2].
Looking in more detail at the studies that involved evaluating taxes that had been implemented (rather than those modelling the effects of potential taxes), most of which focused on the US, the evidence for the impact of taxes on consumption patterns and health outcomes is mixed. As of 2014, most US states had applied some level of taxation to soft drinks, largely for revenue raising purposes [
40], and these do not appear to have had a significant impact on consumption of soft drinks. For example, in an analysis of sales data from Maine and Ohio, one study found that the rate of taxation on the price of soft drinks was, at 5.5% and 5% respectively, insufficient to create a statistically significant change in consumption [
9]. This finding was consistent with an evaluation by Sturm and colleagues [
32], which found that existing taxes on soda, at rates that are typically around 4% in grocery stores in most states, did not have a statistically significantly effect on soda consumption and obesity rates in the US [
32]. Other countries have implemented a higher rate of tax on soft drinks than in the US. For example, in September 2013, Mexico implemented a 10% tax on soft drinks and an 8% tax on unhealthy snacks [
41]. It is estimated that the tax on soft drinks contributed to a 6% average decrease in purchasing of taxed beverages by December 2014, with purchasing reductions being greatest in low income households [
42].
Twenty-three studies considered the estimated or actual health impacts of taxes applied in conjunction, or comparison, with a range of other health-related interventions. Several studies examined the impact of using subsidies – i.e. negative taxes on ‘healthy’ products - alongside taxes on ‘unhealthy’ products. Most often, the subsidies were applied to fruit and vegetables [
18,
21,
43]. Other ‘healthy’ products that were included in the analyses observed were grain-based products high in fibre, fresh fish, and bottled water [
14,
44,
45].
Several of these studies indicate that a combination of taxes and subsidies can have large behavioural and health impacts [
44,
46]. However, it is difficult to ascertain from the findings reported in these modelling studies and experiments whether taxes, subsidies or a combination of the two, are most effective in achieving such impacts. Many of the studies point out, however, that a key advantage of employing subsidies in combination with taxes is that the former can help to offset the inequitable (or regressive) burden of the latter.
A number of the studies considered (likely or actual) differential health impacts by population group. Of these, eight found that taxes on food/beverages were likely to have a greater impact on younger population groups [
22,
23,
26,
27,
47‐
50] and 15 found that public health impacts are likely to be largest for lower income groups [
22,
25,
27,
32,
33,
42,
44,
46,
48,
49,
51‐
56]. In contrast, two studies [
23,
34] found no significant differences between income groups. This suggests that taxes on unhealthy food and beverages may contribute to addressing health inequalities, but that more research is required. As we discuss in more detail later on, 27 of the included studies highlighted the regressive burden of taxes on food and beverage products, suggesting that there is a balance to be struck between the inequitable burden of ill-health and the inequitable burden of taxes.
Overall, there is considerable evidence that high tax rates (i.e. those that raise the unit price by 20% or more) on beverages are likely to have a positive impact on health behaviours and outcomes. The evidence is similar for taxes targeting unhealthy foods, though there are a smaller number of studies and the taxes in question were often more complicated. This finding is consistent with a recent review, which found that food taxes and subsidies are associated with changes in consumption behaviours [
57] and also reflects what is known about alcohol and tobacco taxes [
3,
10‐
12]. However, as noted, it is apparent that such tax rates are far higher than those that have actually been implemented. Hence, it may be that, as Fletcher and colleagues noted, “typically imposed beverage taxes aren’t large enough or transparent enough to lead to meaningful behaviour change.” ([
23], p.1064).
Can health taxes on manufacturers and retailers change behaviours?
Most studies focus on health taxes that are applied at the point of sale, and are intended to try to motivate consumers to change their consumption decisions. It should be noted, however, that a tax on manufacturers may or may not be intended to change behaviour in relation to a finished good, but rather to the use of specific raw materials (ingredients).
We identified three studies targeted at manufacturers or retailers. One such study, by Miao, Beghin, & Jensen [
19], modelled an approach to taxation that targeted the process of adding sweeteners to products, and compared this with a consumption tax on sweetened products [
19]. The rationale was that a tax on sweetener would incentivize producers of high-sugar products to reduce sweeteners in food processing by increasing the unit cost of these products to the manufacturer (while the consumption tax would change consumer-purchasing patterns). As the tax increases the cost of production, suppliers (manufacturers) may respond by increasing the price of the finished good and/or decreasing supply of the product in response to the reduction in profits they make by selling it. However, in some cases, it may not be economically advantageous for suppliers to pass on higher costs to consumers, or to reduce supply in response to higher costs. In this instance, it may be regarded as beneficial to change the formulation of product, e.g. by reducing the fat or sugar content. In this case, the authors conclude that both approaches are potentially effective, but that taxing added sweeteners is likely to have a smaller impact on consumers’ real expenditures than taxing final products.
Another study, which assessed the impacts of a set of complex unhealthy food taxes implemented in Hungary, undertaken by Hungary’s National Institute for Health Development (cited in [
56]), found that substantial changes were subsequently made to the manufacturing of certain products. A survey of manufacturers suggested that the taxing of products exceeding a minimum threshold of certain ingredients such as sugar and fat led 40% of manufacturers to modify their recipe; 30% removed the ingredient entirely, and 70% reduced the level of the ingredient [
58].
In theory, Scotland’s public health supplement on large retailers selling tobacco and alcohol had the potential to discourage retailers from selling either alcohol or tobacco (the latter was a more likely outcome, given the relative profitability of the two types of products). In practice, however, this evaluation found that the level of the tax was too low to stimulate changes in retail practice, which enhanced the predictability of the associated revenue (as discussed above) [
4].
What is the degree of support among public and policy communities for non-traditional health taxes and are there means of increasing support?
Several papers provided insights into three broad categories of factors affecting the feasibility and implementation of new health taxes. The first concerns public opinion regarding proposed, or actual, taxes. Here, studies consistently find that public support for new consumption taxes, or tax increases, is low [
60‐
63], though some suggest that there is public or ministerial support for sugared beverage taxes in some contexts [
39,
64,
65]. A four-country study in the Western Pacific region by Thow et al. suggests that, although governments are ultimately concerned with raising revenue, framing a tax around health promotion can assist in getting such a tax onto the policy agenda in the first place [
39]. For example, a tax on unhealthy food products introduced in French Polynesia in 2002 was framed as a response to concerns regarding poor nutrition and non-communicable disease [
39]. The tax enjoyed broad ministerial support, which was attributed to the tax’s earmarking for public health and other cultural, educational, and youth-focused initiatives, which benefited seven of the 17 ministers in the government [
39]. More generally, support among the general public seems to be higher when credible commitments are made to earmarking funds for specific health activities and objectives, such as subsidizing healthier foods or targeting child obesity (e.g. [
63,
64]).
We identified a smaller number of studies that considered the media coverage of proposed or actual health taxes which might be expected to both reflect and shape public opinion. In some cases, such as the Danish fat tax and the Scottish public health supplement (described above), industry interests opposed to the tax have been able to dominate media coverage, helping to secure further opposition to the tax (which, in both these cases, was eventually dropped) [
4,
24,
66]. In contrast, [
67] analysis of an SSB tax in Mexico provides an example of a supportive media, in which public health advocates successfully utilized media campaigns to raise the public and political profile of the issue and communicate with the public. Less positively, [
68] analysis of debates about potential soda taxes in three US states found that, despite public health advocates’ ability to dominate media coverage with pro-tax messages, none of the proposals were implemented. Hence, while media support for a health tax proposal may be important for it to succeed, it is not sufficient [
38].
Twelve studies considered policy design and implementation factors shaping the fate of proposed and actual health taxes. Studies considering political factors suggested that political support for, and opposition to, health taxes are likely to be key to understanding why some taxes are implemented and others are not (or why some taxes are repealed) (e.g. [
3,
20,
31,
64,
65,
67,
68]). These studies also suggest that opposition to health taxes can develop relatively quickly. For example, [
69] highlights how political opposition to a proposed soft drink tax arose in 2009 in New York State and contributed to the tax proposal being withdrawn prior to implementation, while in contrast [
67] outlines the substantive advocacy efforts in Mexico to combat multi-stakeholder opposition, leading to Mexico’s tax being successfully implemented.
In a study of taxes implemented in Pacific Island nations, [
39] identify industry lobbying in Fiji as a cause of the decision to abolish the country’s domestic excise tax on SSBs. Two studies of the short-lived Danish fat tax both argued that lobbying by food industry interests helped secure political opposition to the tax once it had been implemented, while there appears to have been only limited efforts by the government to secure broader public support. In the absence of such support, political opposition increased and a decision was taken to drop the tax after less than a year (in advance of any analysis of its health impacts) [
24,
66]. The assessment of Scotland’s levy on large retailers reached similar conclusions in relation to industry opposition and the political sustainability of a policy framed as a health tax for which the health rationale appeared to shift over time [
4].
What are the key critiques of health taxes and their implementation and what options exist to manage these challenges?
We identified three key criticisms of taxes on unhealthy products. Twenty-seven of the included studies highlighted the regressive nature of the health tax examined (e.g. [
58,
70,
71]). Poorer groups may be more price sensitive than other groups, and therefore more likely to change their behaviour in response to a tax. In this sense, taxes may play a role in addressing health inequity. In addition, it is important to acknowledge that the regressivity of existing taxes does not necessarily imply that tax
increases will be regressive since, if poorer consumers are more responsive the burden of the tax may shift more to wealthier consumers [
74]. This argument is often made in relation to tobacco taxation – see [
2]. However, if price elasticity is low (as is typical for many unhealthy products), those with lower incomes who continue to buy these products have less to spend on basic needs, such as housing, heating, and healthy food, potentially at the expense of their health and general welfare. Available research does not sufficiently address the question of whether, among low-income consumers, the overall benefits of tax-induced price increases (i.e. reducing consumption of unhealthy products) outweigh the risk of harm from financial hardship for those who do not reduce consumption. More generally, existing evidence concerning outcomes in terms of progressivity/regressivity is limited by the fact that nearly all studies addressing this issue that we identified were based on modelling or predictive experiments. If this particular combination of fiscal measures has occurred in practice, we were unable to find any evaluation studies that covered the issue of regressivity.
For policy actors concerned about the regressive potential of taxes on unhealthy products, one potential response to this would be, as noted above, to use the revenue from such taxes to subsidize other ‘healthy’ foods, such as fruit and vegetables. In this way, it may be possible to put together a package of policies in which there can be some confidence that the overall impact on poverty will be negligible [
72].
A second criticism, put forward by Fletcher et al., is that food and beverage taxes may simply lead to consumers substituting the taxed products for similar, non-taxed alternatives which are not necessarily healthier, such as sports drinks or juice [
29]. This is an issue that has also been identified in the context of differential taxes on different types of tobacco products [
2]. It implies that there is a need to carefully assess behavioural changes in response to taxes intended to achieve health goals, and that policymakers need to stay alert to the possibility that such taxes may need to be revised or expanded in response to changing behaviours.
The third key criticism is that implemented taxes are often too low to have a meaningful health impact, a criticism that is borne out by the empirical evidence, as noted above [
29]. It may be appropriate to consider the level of a tax before deciding whether or not it is appropriate to frame it as a ‘health tax’. Lower and incremental taxes are more likely to provide a stable source of revenues (which may, or may not, be spent on health-related activities) but they are less likely to achieve behaviour changes [
59,
71].
In addition to these three criticisms, it is evident that, while framing new taxes or tax increases as mechanisms for increasing health spending may increase public support, funds may not always be clearly earmarked in practice [
2,
4]. Where this occurs, this may undermine support for such taxes in the longer-term.
Looking back across our five research questions, it is apparent that the results of our review are consistent with those focusing on traditional excise taxes on alcohol and tobacco. Those reviews show that increasing taxes leads to reduced consumption among the population and can be a valuable source of revenue for government [
10,
73‐
75]. A review by Chaloupka and colleagues shows that revenue from tobacco taxes may be more reliable than those discussed in the present review, however, because there are fewer substitutes available for tobacco products, and the demand for them is therefore relatively inelastic [
74]. With regard to public support for alcohol and tobacco taxes, studies find greater public support for these approaches when the tax is earmarked for healthcare or for combating tobacco- or alcohol-related harms [
74,
76]. The concern with the regressivity of health taxes is also relevant for alcohol and tobacco. A recent study confirmed that alcohol taxes are regressive, although the authors interpreted this effect to be small [
77]. A review by Hill and colleagues found that tobacco price increases via taxes has a greater impact on low-income groups compared with those with high incomes (although, similar to the present review, this effect is argued to be positive given its potential to reduce socioeconomic inequalities) [
78].