Sugar-based beverage taxes and beverage prices: Evidence from South Africa's Health Promotion Levy

https://doi.org/10.1016/j.socscimed.2019.112465Get rights and content

Highlights

  • In April of 2018, South Africa implemented a tax on sugar-sweetened beverages.

  • Prices of carbonates increased by 1.006 ZAR/litre post the introduction of the tax.

  • Prices of untaxed beverages did not increase.

  • Among carbonates, price increases were not greater for higher sugar beverages.

Abstract

A growing number of jurisdictions are introducing taxes on sugar-sweetened beverages (SSBs) in efforts to reduce sugar intake, obesity, and associated metabolic conditions. A key dimension of the impact of such taxes is how they induce changes in the prices of the taxed beverages and their un-taxed substitutes. At present these taxes have typically been based solely on volume. More recently, however, due to the potential to target the source of SSBs' health harms and to incentivize product reformulation, SSB taxes are being levied based on sugar content. In April of 2018 South Africa implemented such a tax, the Health Promotion Levy (HPL), at a rate of 0.021 ZAR (approximately 0.15 US cents) for each gram of sugar over an initial threshold of 4 g/100 ml. Drawing on a dataset of price observations (N = 71, 677) collected in South Africa between January 2013 and March 2019, we study changes in beverage prices following the introduction of the HPL. We find null price increases among un-taxed beverages and find significant price increases for carbonates, the largest taxed product category. However, within carbonates we find similar increases in price for low- and high-sugar brands, despite the underlying difference in tax liability. In addition, while we find evidence of product reformulation, we find significant price increases among the brands that reduced their sugar content. While the findings are broadly consistent with the price changes of volume-based SSB taxes, future considerations of price effects of sugar-based SSB taxes need to account for the opportunity for intra-firm heterogeneity in price response among large multi-product firms.

Introduction

In response to rising prevalence of obesity and its comorbidities, a number of jurisdictions have introduced or are in the process of introducing taxes on obesogenic foods and beverages and in particular on sugar-sweetened beverages (SSBs). SSBs are non-alcoholic beverages containing added sugar, with common examples including carbonated sodas, juice drinks, and sports and energy drinks (Hu, 2013). Excessive consumption of SSBs is strongly associated with weight-gain, type 2 diabetes mellitus, and other metabolic conditions (Feeley et al., 2013; Malik et al., 2013; Malik et al., 2006; Te Morenga et al., 2013; Te Morenga et al., 2014; Vorster et al., 2014).

One key rationale for such taxes is Pigouvian in nature. Pooled or publicly financed healthcare provision results in consumers externalising the costs of the treatment of the diseases associated with their consumption of SSBs (Brownell et al., 2009). However, is the source of this externality cost the SSB product in its whole, or is it its constituent ingredients? SSBs' association with obesity is driven by their high sugar content and its liquid form, often not compensated for via equivalent reduction in calories from other foods, which is rapidly absorbed by the liver (DellaValle et al., 2005; DiMeglio and Mattes, 2000; Mourao et al., 2007).

As recommended by the World Health Organization, some countries have introduced taxes that differentially tax soft drinks based on how much sugar they contain (WHO, 2016). Chile and the United Kingdom tax beverages at different rates relative to several discrete sugar content thresholds (Caro et al., 2018; gov.uk, 2016; Nakamura et al., 2018). By taxing ingredients rather than whole products, one introduces an incentive for producers to reformulate products to reduce the concentration of the taxed ingredient (Blecher, 2015). Such a mechanism does not exist for uniform per volume taxes such as the one peso per litre tax on drinks containing added sugar implemented by Mexico in 2014, where SSBs of differing sugar contents are taxed at equivalent rates (Colchero et al., 2016). The notion of ingredient-based taxation is common in the treatment of alcoholic beverages where taxes are often levied relative to absolute alcohol content, while cigarette products are more appropriate to uniform taxation due to the uniformity of the harms (Blecher, 2015).

Price is a critical tool that governs the ultimate behavioural and public health impacts of excise taxes. From a public health standpoint, the extent to which a tax might induce reductions in consumption of the unhealthy taxed products is determined, in conjunction with how price elastic demand is, by the extent to which prices respond to the tax. There is a significant literature examining the impact of volume-based taxes. These studies find significant or entire pass through of taxes particularly in low- and middle-income settings. For instance, Colchero et al. (2015) and Gogger (2015) find Mexico's one peso per litre tax was on average entirely passed through to consumer prices, with some heterogeneity across product size and geography. Evidence from local soda taxes implemented by cities in the United States suggests some variation in pass through across cities (Cawley et al., 2018a, 2018b; Cawley and Frisvold, 2017; Falbe et al., 2015; Silver et al., 2017). However, due to their heretofore limited implementation there is at present no published evidence on the price effects of sugar-based SSB taxes.

Conventional economic theory suggests that profit maximizing firms will increase their products' prices with the magnitude of this price change being mediated by the price elasticity of consumers' demand (Hines, 2008). However, in the face of an ingredient-based tax, producers face an additional decision which is whether or not to reformulate their products to reduce the levels of the taxed-ingredient and the associated tax liability. Further, reformulation involves costs, some fixed but others variable and determined by the extent that firms are price-takers and do not hold significant monopsony power in the market for sugar (or sugar substitutes). Firms may also respond by re-focusing advertising efforts (Blecher, 2015). All of these mechanisms interact with firms being multi-product firms. There is thus not much known ex-ante about how firms (and particularly their products' pricing) respond to sugar-based SSB taxes.

South Africa presents an opportunity to study the effects of such sugar-based SSB tax policies. Facing an increasingly severe burden of disease attributable to excess sugar and SSB consumption, South Africa implemented a tax on SSBs on April 1, 2018. This new tax instrument titled the Health Promotion Levy (HPL) was introduced through the passage of the 2017/18 Rates and Monetary Amounts Bill (Stacey et al., 2017; National Treasury, 2016, National Treasury, 2018). The tax is levied at 0.021 ZAR (approximately 0.15 US cents) for each gram of sugar over a threshold of 4 g per 100 ml on non-alcoholic drinks subject to the tax (National Treasury, 2018).

Do sugar-based taxes result in price increases? If so, is the pass-through complete? Does the incentive for product reformulation interact with firms' pricing responses? We seek to address these gaps by providing evidence on South Africa's implementation of the Health Promotion Levy. Drawing on price data collected for compilation of Statistics South Africa's Consumer Price Index and exploiting the discrete introduction of South Africa's HPL on the sugar content of SSBs, we estimate the change in prices of taxed and untaxed products following the introduction of this sugar-based tax. We proceed with a description of the HPL, our data and econometric approach, a presentation of our results, and close with a discussion and conclusion.

A generic call for a tax on SSBs was first made in 2012 by South Africa's National Department of Health in their National Strategy for Prevention and Control of NCDs, 2013–2017, and then again in 2015 in their National Strategy for Prevention and Control of Obesity, 2015–2020. In February 2016, the National Treasury formally announced its intentions to implement a tax on SSBs as of the next fiscal year, April 2017, and subsequently released a policy paper outlining the nature of proposed tax (National Treasury, 2016). Following what was a protracted legislative process, including extensive public consultations, the policy was only signed into law in December 2017. This process saw implementation delayed a year, with the HPL going into effect in April 2018.

The intention, outlined in the National Treasury policy paper, was to tax SSBs to reduce harms arising from excessive sugar content and to levy the tax in such a way so as to create an explicit incentive for producers to reduce the sugar-content of their taxable products (National Treasury, 2016). While other settings, such as the United Kingdom, opted for a tiered tax with rates increasing in discrete steps with increasing sugar content, the National Treasury proposed a tax linear in sugar content, with the rate set at 0.0228 ZAR/gram of sugar. This original proposal, which would have produced a burden of 20% of the price of the most popular soft drink brand was opposed by the beverage industry. A revised proposal was adopted which exempted the first 4 g of sugar per 100 ml from taxation, and taxed each gram over the 4 g threshold at 0.021 ZAR (depicted in Fig. 1). This compromise significantly reduced the burden of the HPL to 10–11% of the price of the most popular soft drink brand.

The formal delineation of which products are subject to the HPL is done via the World Customs Organization's Harmonized System designations (See Supplementary Table 1). Practically, beverages subject to the HPL include carbonates (sugar-sweetened and artificially-sweetened), concentrates, fruit nectars, sports and energy drinks, and ready-to-drink teas with their respective tax liabilities being determined by their sugar content. Beverages not subject to the HPL include non-flavoured bottled waters and 100% fruit juices. In the context of the South African drinks market, as displayed in Fig. 2, sales of carbonates dwarf sales of other beverage types, and consequently the impact of the HPL on these products is of particular consequence. Beverage manufacturers, and importers of beverages, are legally responsible for payment of the HPL to the South African Revenue Service, with tax liability determined by tests of sugar content undertaken by accredited laboratories. Small manufacturers, defined to be those using less than 500 kg of sugar in a year, are exempt from paying the HPL.

While not formally earmarked, some of the revenue raised from the HPL will be “soft-earmarked” for health promotion activities across government. As of December 2018, revenue raised had exceeded forecasts and reached approximately 2 billion ZAR. This is about 0.15% of South Africa's total tax revenue for the 2018/19 fiscal year.

Section snippets

Overview

We take three econometric approaches to studying the effect of the introduction of South Africa's HPL on prices. The first is a simple pre-post analysis, which identifies changes in average price across various taxed and un-taxed product categories following the introduction of the HPL. A second approach seeks to estimate the pass-through, price change relative to tax liability, among taxed products. And the third and final approach studies differential price change among brands that were

Data

The primary data utilized in this study are retail prices for non-alcoholic beverages collected by Statistics South Africa's Consumer Price Index (CPI) unit. These prices are collected by in-store observation in urban areas across South Africa on a monthly basis (StatsSA, 2017). Products sampled are intended to be the most popular item for each product type and unit size in each store. This is operationalized by enumerators recording the prices of the products occupying the most significant

Results

Table 1 presents summary statistics across beverage categories for the full analytical sample. The mean retail price per litre is lowest among bottled water and highest among juice beverages. Mean sugar content is 9.441 g/100 mL among high sugar carbonates (Sugarb4 g/100 mL) and only 2.990 g/100 mL among low sugar carbonates (Sugarb<4 g/100 ml), which include both entirely artificially sweetened beverages as well as lower sugar content beverages.

Table 2 presents estimates of the pre-post

Discussion

The ultimate reductions in disease risk which excise taxes are intended to achieve are determined by how market actors respond to the incentive structure that these policies impose. In the case of SSB taxes, many of the existing policies currently implemented, have been taxes levied per volume. There has, however, been some implementation of sugar-based taxes, with the rationale that these target the source of SSB's health harms. The dimensions of how industry respond to an excise tax, increase

Conclusion

With growing interest in the use of tax and fiscal policy to reduce the harms associated with SSB consumption, there has been some general guidance issued to levy taxes on SSBs with higher tax rates on higher sugar content products relative to those of lower sugar products (WHO, 2016). However, due to the novelty of such sugar-based SSB tax policies, there is limited actual experience from which best practices on how to design and implement such taxes could be based. Our study provides evidence

Declaration of interests

None.

Acknowledgements

Financial support for this work was provided by the International Development Research Centre (Grant number: 108424-001) and Bloomberg Philanthropies. The funders had no role in the undertaking of this research. The authors would like to thank Patrick Kelly and Marietjie Bennett of Statistics South Africa.

References (33)

  • J. Cawley et al.

    The Pass-Through of the Largest Tax on Sugar-Sweetened Beverages: the Case of Boulder, Colorado

    (2018)
  • J. Cawley et al.

    The Impact of the Philadelphia Beverage Tax on Prices and Product Availability

    (2018)
  • M.A. Colchero et al.

    Changes in prices after an excise tax to sweetened sugar beverages was implemented in Mexico: evidence from urban areas

    PLoS One

    (2015)
  • M.A. Colchero et al.

    Beverage purchases from stores in Mexico under the excise tax on sugar sweetened beverages: observational study

    BMJ

    (2016)
  • D.P. DiMeglio et al.

    Liquid versus solid carbohydrate: effects on food intake and body weight

    Int. J. Obes. Relat. Metab. Disord.

    (2000)
  • Euromonitor

    Passport Database

    (2018)
  • Cited by (59)

    • Decomposing consumer and producer effects on sugar from beverage purchases after a sugar-based tax on beverages in South Africa

      2022, Economics and Human Biology
      Citation Excerpt :

      In comparison, that figure is 43% (0.8 g/capita/day) for high-LSM households. Given the documented increases in taxable carbonates’ prices relative to pre-HPL trends compared to other beverages (Stacey et al., 2019), this difference is consistent with greater price sensitivity among middle-LSM households in South Africa (Burger et al., 2017). On the producer side, we found that reformulation led to very similar decreases in the sugar concentration of beverage purchases (taxable and nontaxable) by both middle- and high-LSM households (−0.3 g/100 ml on a cumulative basis; Supplemental Table S8), indicating that despite clear differences in beverage purchases by LSM level (Supplemental Table S6), beverages purchased by each group were not reformulated differentially overall.

    • The economics of malnutrition: Dietary transition and food system transformation

      2022, Handbook of Agricultural Economics
      Citation Excerpt :

      This results in less political pressure for prevention. Conditions like in Mexico have arisen also in South Africa, where the Health Promotion Levy introduced in 2018 could help limit SSB consumption (Stacey et al., 2019) and achieve significant health improvements with relatively little loss of consumer welfare for the poorest (Saxena et al., 2019). A final category of price instruments includes subsidies for healthy items.

    View all citing articles on Scopus
    View full text