Preventing Non-Communicable Diseases Using Pricing Policies: Lessons for the United States from Global Experiences and Local Pilots

Preventing non-communicable diseases (NCDs) in an effective and sustainable way will require forward-looking policy solutions that can address multiple objectives. This was true pre–COVID-19 and is even more true now. There are already examples from across the globe and within the United States that show how these may be possible. Although there are still many unknowns around how the design, targeting, level, sequencing, integration, and implementation of fiscal policies together can maximize their NCD prevention potential, there is already clear evidence that health taxes and particularly sugar-sweetened beverage (SSB) taxes are cost-effective. Nonetheless, policies alone may not succeed. Political will to prioritize well-being, protections against industry interference, and public buy-in are necessary. If those elements align, pricing policies that consider the context in question can be designed and implemented to achieve several goals around reducing consumption of unhealthy SSBs and foods, narrowing existing nutritional and health disparities, encouraging economic and social development. The US and its local and state jurisdictions should consider these pricing policy issues and their contexts carefully, in collaboration with community partners and researchers, to design multi-duty actions and to be prepared for future windows of opportunities to open for policy passage and implementation.

• Policies alone may not succeed; political will to prioritize wellbeing, protections against industry interference, and public buy-in are necessary.
• If these elements align, pricing policies that consider the context in question can be designed and implemented to achieve several goals around reducing consumption of unhealthy beverages and foods, narrow existing nutritional and health disparities, and encourage economic and social development.
Policies to prevent non-communicable diseases (NCDs) such as diabetes, hypertension, cardiovascular diseases, and cancers should be important national and local priorities for several reasons. First, the current prevalence of NCDs is alarming, with 73% of the world's population expected to die from an NCD. 1 NCD prevalence is higher in developed countries compared with developing countries, but age-standardized mortality rates are generally rising in developing countries, which represent a large and rapidly growing share of the world's population, 2 while declining in developed countries. 3 Second, there are clear links between NCDs and the severity and mortality of infectious diseases such as coronaviruses and tuberculosis. 4 People with NCDs have weakened immune function and more widespread inflammation, and vaccines or treatments for infectious diseases (like COVID-19) may not work as well for them. 4 Even when only accounting for healthcare costs (borne by public and private entities and individuals), economic costs of NCDs are large and are certainly worse when lost productivity of both the ill and their caregivers are also accounted for. 5 Moreover, NCDrelated costs are a larger burden for already low-resourced communities and individuals across most countries, thus widening existing income and sociodemographic disparities.
According to the World Health Organization (WHO), preventive policies are the most cost-effective actions countries can undertake to reduce the health and economic burdens of NCDs. 6 Pricing policies on foods and beverages to increase the price of unhealthy beverages and foods (e.g., taxation, tariffs) or to decrease the price of healthier beverages and foods (e.g., subsidies, cash transfers) are one of the NCD prevention approach governments have used or considered. 7 This brief summarizes the existing literature through a narrative review of papers based around two major questions: What can the United States learn from other countries' experience in designing and implementing pricing policies on food and beverages as part of their NCD prevention strategies? What knowledge gaps remain and how can US jurisdictions take the lead in informing on these policy debates and lessons for other US jurisdictions and other countries?

Taxing SSBs and Other Unhealthy Foods
Taxes are a critical pillar of NCD prevention and have been applied to tobacco, alcohol, and more recently sugar-sweetened beverages (SSBs). They are often collectively called "health taxes. " 7,8 The basic principle is that imposing or increasing taxes on these items will result in higher prices, lower consumption, and thus are a way to prevent or minimize the development of future cases of related NCDs. To date, over 45 countries, cities, or regions within countries have instituted SSB taxes, including eight US localities. 9,10 Studies assessing SSB taxes have used price changes and consumption changes (purchases, sales, or intakes) as primary outcomes. Overall, price changes from taxation are heterogeneous and highly dependent on key factors such as the levels of consumption, tax design, market shares of beverage brands given the geographical coverage of the tax, and strategic behaviors by beverage companies and retailers. 11,12 Consequently, changes in purchases and consumption are also heterogeneous, particularly across income levels, age groups, and prior beverage consumption levels. [12][13][14] Nonetheless, global metaanalysis shows that the average consumer will lower their SSB purchase by 10% if SSB prices rise 10% (implied price elasticity of demand of −1), 15 whereas a meta-analysis limited to the Americas shows that the implied price elasticity of demand is −1.36. 16 There is slower momentum around food taxes for two reasons. First, they are more difficult to justify because foods are more complex mixtures of nutrients whereas SSBs have no nutritional value. Second, there are concerns around such taxes creating a higher burden among people who are poor. However, there is mounting evidence showing that increased proportions of ultra-processed products (UPPs) in diets are linked with increased risk of obesity; many measures of cardiovascular disease, diabetes, hypertension, mortality, and cancers; and related mortality and total mortality. [17][18][19][20][21][22][23][24][25][26][27][28][29][30][31][32][33] As such, when the economic burden of having these diseases is accounted for, the poor may benefit when taxes are imposed on unhealthy foods.
Meanwhile, UPPs are gaining popularity globally, 34-40 forming a growing share of the diets of people who are very poor and now reaching infants and preschoolers. [41][42][43][44][45][46][47] In countries with available data (mostly emerging economies and high-income countries), SSBs represent about 2%-7% of food purchases and 4%-10% of kcal/day, whereas UPPs (which include SSBs) represent 17% to over 25% of purchases (15%-60% of kcal/ day) based on the age-gender group and country. Indeed, previous policy simulations show that if Chile implemented an unhealthy UPP tax aligned with their innovative and integrated food labeling and marketing regulations, there would be clinically meaningful reductions in household purchases of attributes of concern (sugar, sodium, saturated fats) linked with the most common NCDs primarily from the targeted unhealthy UPPs. 48,49 Such a broad-based tax on unhealthy UPPs builds on the two known and evaluated national taxes on some subset of UPPs in Mexico's nonessential food tax and Hungary's junk food tax. 50-52 Both showed significant reductions in the purchases of the targeted products, albeit tempered by their low tax rates (and hence small effect size). Consequently, taxes to date demonstrate strong promise for changing demand and supply of unhealthy beverages and foods. However, there is also some emerging evidence that sugar content-based taxes, although more effective in encouraging sugar reduction, are being avoided with a growing number of non-nutritive sweeteners in products, the long-term health implications of which are inconclusive at this time. 53 Indeed, it should be expected that any tax targeting current products or attributes of concern will be met with subsequent introductions of new ingredients and products and substitutions among consumers with an alternative product that may or may not be supportive of health. Therefore, researchers and regulatory agencies must be vigilant and thoughtful in establishing mechanisms (e.g., requiring products to declare the amounts of non-nutritive sweeteners and new additives) with which to periodically and quickly assess improve these regulations to ensure that they evolve with the food landscape to best protect people's health.
Health implications on a population level (e.g., flatting of diabetes prevalence rates or reductions in obesity incidence) will take years to emerge, so researchers have used consumption changes to estimate longer-term health and economic implications. These broadly show meaningful reductions in incidence and prevalence of NCDs and thus healthcare cost savings in all countries studied to date. 54-59

Decreasing Prices or Increasing Affordability of Healthier Alternatives
Lowering the prices or increasing the affordability of healthier alternatives are also viable pricing policies. One approach is to shift the existing tax structure to create larger price differentials between products of concern versus those considered healthier. For example, Chile went from a 13% ad valorem tax on beverages to 0% for plain waters and plain dairy-based drinks, 10% for all non-alcoholic beverages with sweeteners and <6.25 g sugar, and 18% for all non-alcoholic beverages with sweeteners and ≥6.25 g sugar. Because of the relatively small price increase for higher sugar drinks (+5% points) and the small decrease in prices of lower sugar drinks (−3% points) from the prior 13% tax rate, evaluations of this tax restructure have shown that the price changes were partially absorbed by suppliers, and purchase changes were consequently small. 60 Similarly, efforts to lower costs of healthy foods (e.g., agricultural subsidies or removal of past taxes on producers) are often only partially passed on as price reductions for the public 61 and so are not necessarily an effective way for governments to be spending or lowering revenue if a key purpose is to change consumption behavior.
Another approach that more directly influences choices is to increase the ability for the population to afford healthier alternatives. 62,63 This can be especially effective when targeted toward lower-income populations because it can serve two purposes by addressing equity concerns particularly when paired with tax policies to counter regressivity arguments and by providing reinforcing messages about what are healthier options vs unhealthy ones. In many countries, cash transfers have been the primary method used to bolster purchasing power. This is typically targeted toward specific subpopulations that need to meet eligibility criteria (e.g., income and asset ownership, age, health condition) and are critical mechanisms for delivering healthy food assistance.

Tax Design and Implementation Considerations for the United States
Findings to date suggest that future health taxes in the United States and elsewhere should consider the following issues. First, the baseline levels of consumption of various UPPs for the country and by subgroup (especially by income) and price elasticities of demand is important to understand. These will inform on the scope of products covered by the tax, substitutions and the level of tax to result in purchase changes.
Second, the tax structure should be aligned with the primary objectives. If the goal is sugar reduction, then a sugar density-based tax structure will achieve this more effectively, as suggested by recent findings in Portugal and the United Kingdom on their SSB taxes. 64,65 Likewise, a tax on nonessential UPPs defined based on a country's food-based dietary guidelines and the NOVA classification system 66 could be based on tiered nutrient cutpoints for UPPs with higher tax levels for products with nutrients of concern in excess of these cutpoints. If the goal is revenue generation, then a specific volume-based tax across a broader scope of unhealthy foods or beverages may result in greater tax revenue given a weaker incentive to reformulate but also less public support. Meanwhile, ad valorem taxes on SSBs have been shown to have lower pass-through onto prices compared with specific taxes, with sales taxes being particularly ineffective. 67 For food taxes, ad valorem taxes would be more feasible to implement given very large variations in prices across food categories, but again, there should be higher tax rates for UPPs with more or a higher density of the nutrients of concern. If the goal is equity enhancement, then impacted communities' needs and agency will need to be incorporated into the decisionmaking process and identification of the objectives of the tax. 68 Additionally, the geographical coverage of the tax jurisdiction has implications on the ease of cross-border shopping and highlights the need for national-or province-/state-level taxes over local taxes (although some countries are small and surrounded by other countries without similar taxes).
Third, one critical concern from a health perspective is how much the attributes of concern (e.g., sugar or calories) are reduced. Although these taxes affect high-income consumers less and lower-income consumers more, 13,52 evaluations to date suggest the reductions from SSB tax levels to date only translate to 5-22 kcals per capita per day. Even if these reductions are sustained, they are unlikely to have meaningful impacts on the broad swath of health outcomes in a timely manner though research shows the 10-20-year time horizon will produce important results. 57, 69 One way to accelerate this is to increase the current tax rates. A few Gulf states have instituted 50%-100% excise taxes on different subsets of SSBs and Bermuda has implemented a 75% import tax on sugar, SSBs and candies. While originally considered infeasible, tax levels for tobacco in some locations now have tax rates ranging from 100%-1000%, and an annual tax raise of $1 or more is often seen. 70,71 Thus, increases in tax levels for SSBs or nonessential foods over time might be possible.
In findings to date, taxes levied directly on and collected from manufacturers and distributors are easier to enforce and allow producers of taxable products to determine whether and how to pass-through the tax across their portfolio of products. These excise taxes can be built into existing collection systems. Moreover, the US Food and Drug Administration already requires and recently updated nutrition label regulations that facilitate nutrient-based tax designs. This provides a clear and consistent message to both manufacturers and consumers in terms of what items are discouraged. Taxes levied directly on suppliers also help with the framing of the tax as aiming to improve food offerings rather than simply functioning as a revenue raiser with the tax burden falling on the public. Indeed, the public's understanding and perception of the intention of such taxes will be critical in shifting social norms and thus lowering demand for SSBs and UPPs.

Tax Revenue Use
One way that local US jurisdictions have been trailblazers has been around determining revenue use. In the US localities with SSB excise taxes, many have included the creation of local commissions or committees tasked to do determine revenue use. An evaluation of community investments from SSB tax revenues in Albany, Berkeley, Oakland, and San Francisco provide useful guidance. 77 Use of those revenues include funding existing programs like the Berkeley Unified School District's garden program and capital improvement projects in Oakland's parks and recreation department as well as new programs like adding water stations in schools and educational programs for pre-diabetic adults on lifestyle changes through the YMCA. 77 This allows for trifold action via lowering demand for unhealthy beverages, generating revenue to use for lowering prices of healthy foods for individuals with low incomes, and communicating a consistent message.
Applying an equity-enhancing approach provides agency over the purpose of health taxes 68 to communities impacted by the harms of SSBs and UPPs and can help garner support around local "health" or "health-promoting" taxes because there is an explicit connection to health. Overtly revenuedriven tax policies are often seen as money-grabs, associated with fiscal mismanagement by governments, and are income regressive in nature. Therefore, to the extent that governments or advocates of such policies can communicate their revenue use in compelling ways to support communities' needs (even without strict earmarking), they can boost public support for such measures.

Supporting Healthier Eating
With regards to supporting healthier eating, there are also already examples of how local US jurisdictions are modeling approaches to use "healthy food pricing incentives" toward NCD prevention efforts while enhancing equity. These incentives can come in the form of subsidies, rebates, discounts, and matches, and they have primarily focused on fruits and vegetables to date because moderate to good quality evidence supports the use of pricing incentives to increase consumption or purchase of fruits and vegetables. 78 Organizations across the United States have been experimenting with targeted incentive programs through Medicare and Medicaid among beneficiaries of assistance programs-such as Temporary Assistance for Needy Families (TANF), Supplementary Nutrition Assistance Program (SNAP), and the Special Supplemental Nutrition Program for Women, Infants, and Children (WIC)-who have dietsensitive health conditions (e.g., hypertension, diabetes). 62 Such "healthy food prescriptions" provide increased financial access to healthy options in stores or at farmers markets via added benefits for qualifying foods through pilot programs such as SuperSNAP (currently in 42 counties in North Carolina) and Wholesome Wave, which are funded by a mixture of federal grants and private foundations. These programs are identifying and creating innovative approaches for involving retailers, local farmers, software developers, data scientists, and clinicians to manage incentive transfers and to track purchase, utilization, and health outcomes. With the current economic, social, and health threat caused by COVID-19, the expansion of these options is more pressing than ever. Investments into these innovative programs that allow at-risk populations (e.g., low-income populations of children, elderly people, and families) to access nutritious foods are critical for preventing both NCDs and the severity of infectious diseases like COVID-19.
However, such incentive efforts may not be directly transferable or scalable outside of high-income countries with existing funding or technological and implementation capacities. Indeed, most studies have occurred in the United States (19 studies) and other high-income countries (8 studies), with only one study each in Peru and South Africa. 78 Nonetheless, there are elements of incentive programs analogous to cash transfer programs in low-and middleincome countries (LMICs) in terms of targeting (e.g., Bolsa Família in Brazil), many of which have been evaluated and monitored over decades. Funds from health taxes could be directed toward supporting such social support programs, improving school feeding programs or being reinvested into communities based on their needs.

Remaining Knowledge Gaps
As more countries and localities implement health taxes, we will continue to learn from them and find ways to improve their design and implementation to maximize their potential for NCD prevention, particularly among individuals with low resources. However, we are still learning how these designs can best improve health because changes in health outcomes will take many years to manifest. This makes establishing causality difficult. In addition, it is well accepted that any single policy (unless extremely dramatic and covers a significant share of unhealthy consumption) will unlikely result in fast improvements, thus requiring multi-prong policies (e.g., integrating labeling, marketing restrictions, and tax policies) and accompanying implementation, outcome, and impact evaluations designed to assess the additive or multiplicative associations of these policies together. Finally, our understanding of how the various combinations of foods, ingredients, and chemicals we are exposed to affects our health is still evolving. Therefore, policy designs based on these attributes also need to evolve.
Meanwhile, healthy incentives are understandably popular but can be expensive and thus should be carefully designed and implemented to improve NCD prevention cost-effectiveness. Critical design decisions include the target population; how to determine eligibility or whether to make it conditional on certain behaviors (vs. making it too difficult to enroll and use); length of eligibility; frequency/cycle of disbursement; the amount of incentive; and food selections covered, their baseline levels of consumption, and substitutes. 78 Important implementation decisions are needed around the modes of enrollment, disbursement, and redemption; frequency of benefits; and incorporating reminders and nutrition education. 78 Evaluations of pilot programs should incorporate cost-benefit or cost-effectiveness analyses to help determine how to improve these incentive programs most strategically to maximize their impact and narrow health disparities. Finally, although our focus here is in terms of NCD prevention, given the costs of such programs, the political feasibility of such efforts could be bolstered when they also serve to improve educational outcomes, support the local economy, enhance agricultural practices, and link with educational or other social outcomes.

Summary
Preventing NCDs in effective and sustainable ways will require forward-looking policy solutions that can address multiple objectives. This was true before COVID-19 and is even more true now. There are already examples from across the globe and within the United States that show how these solutions may be possible. Although there are still many unknowns around how the design, targeting, level, sequencing, integration, and implementation of fiscal policies can together maximize their NCD prevention potential, there is already clear evidence that health taxes-particularly SSB taxes-are cost-effective. 8,79 Future expansions of the WHO OneHealth tool to incorporate such pricing policies in reducing the burden of NCDs are needed. 80 Nonetheless, policies alone may not succeed. Political will to prioritize well-being, protections against industry interference, and public buy-in are necessary. If those elements align, pricing policies that consider the context in question can be designed and implemented to achieve several goals around reducing consumption of unhealthy SSBs and foods, narrow existing nutritional and health disparities, and encourage economic and social development. The United States and its local and state jurisdictions should consider these pricing policy issues and their contexts carefully, in collaboration with community partners and researchers, to design multi-duty actions and to be prepared for future windows of opportunities to open for policy passage and implementation.