Introduction

South Africa (SA) has one of the highest rates of obesity in sub-Saharan Africa, with levels continuing to increase [1–3]. One third of men and two thirds of women are overweight or obese. This compares globally to 43 percent of men and 44 percent of women who are overweight and 16 percent of all adults who are obese [4,5]. In SA, diabetes and heart disease are among the top five leading causes of death behind COVID-19, HIV and stroke [6]. At the time of this study the rates of overweight and obesity in children and adolescents in SA were comparable to those seen in high income, industrialized countries 10–15 years prior suggesting a continued rise in prevalence across the population [7].

Alongside its overweight and obesity, and wider noncommunicable disease (NCD) burden, SA has a declining but continued high burden of HIV. Approximately 18% of 15–49 year olds were living with HIV at the time of the study; of which 5.5% were young people aged 15–24 [8–10]. The intersection of HIV-NCD coexisting in developing economies is an evolving area of public health research, policy and practice, and the double burden of NCDs and HIV has already and will continue to impact the SA health system, economy, and communities [11,12].

South Africa’s rates of obesity and NCDs have been attributed to a nutrition transition linked to globalisation and greater freedom of movement for black populations, particularly following the end of Apartheid, and is similar to other transitioning economies in Africa [13–17]. The nutrition transition included increased consumption of food away from the home as well as increased consumption of convenience foods (e.g., pre-packaged meals), which are high in saturated and trans fats, salt, sugar, animal sourced and processed food, and sugar sweetened beverages (SSBs) [15]. Evidence suggests that consumption of sugary drinks is associated with weight gain across populations globally, including in SA [2,18,19].

Further to the nutrition transition, SA has many economic and structural similarities to other post-colonial, sub-Saharan countries, however the documented racialised income inequalities; food insecurity; brand loyalties; and particular geography, health, and cultural life in townships is a unique legacy of Apartheid policies. In 2018, 11% of the SA population was living in hunger with many more being food insecure [20]. Black South Africans are consistently in the lowest income groups [21]. There is evidence of strong brand association with regards to food choices and personal identity, particularly among lower income communities [22]. Like the global market, Coca-Cola Co., PepsiCo Inc. and the Danone Groupe have come to hold the majority of the total soft drinks sales market in SA [21–24].

When the South African Health Promotion Levy (HPL) came into effect in 2018 it had the stated goal of lowering the consumption of SSBs across the South African population. The HPL was implemented as an excise fixed rate tax on sugar content, taxing beverage products that exceed 4 grams of added sugar per 100ml [25]. At the time, this was the equivalent of an 11% levy on sugary beverages or 2.1 cents per gram; payable by the manufacturer on products to be sold domestically in SA. For the price of products to impact population diets, there is the rationale to use measures that will influence food supplies dominated by cheap, processed, energy-dense products. A primary purpose for taxing SSBs is to decrease consumption across a population, and thereby reduce diet-related risk factors for NCDs, in addition to raising awareness about SSB over consumption and health [25–27]. There are a range of SSB taxes and levies in place around the world, including a growing number in sub-Saharan Africa following on from the HPL success; evidenced by reductions in purchasing and consumption of SSBs across the population [26,28–31].

Globally, there has been a move to implement policies to address the drivers of NCDs. The last two decades has seen increased political support for the global NCD agenda evidenced by the United Nations’ high-level meetings on NCDs, in particular those in 2018 and 2025 and by the World Health Organization’s (WHO) on-going promotion of these policies. The implementation of the HPL is an example of the efforts in SA to address NCDs [25].

In the years leading up to the introduction of the HPL, 15–24 and 25–34 year olds had the two highest levels of SSB consumption in SA [32]. As well, 18 is the age at which individuals are considered to be adults in SA and may be starting to make their own decisions regarding purchasing and budgeting for food. This paper explores young peoples’ experience and perceptions of food choice, food access, the HPL and health in Khayelitsha, a township characterised by persistent inequalities, poverty and health burdens, during the period soon after implementation. It draws on a qualitative study that used focus group discussions through a critical theory approach to explore participants’ knowledge of the HPL in the context of their access to food and their everyday lives. We examine the knowledge participants had of the HPL 18 months after it came into effect and what, if any, bearing the HPL had on their day-to-day lives then. We examine the role of on-going income inequalities and high rates of unemployment on the lives of young, low-income South Africans after the implementation of the HPL. Using the HPL and taxes on SSBs as a critical lens through which to understand the interplay between culture, policy and lived experiences. In considering the impacts of policies such as the HPL, it is important to pay adequate attention to the social and historical context, and the broader realities of people’s lives. Examining the recent past to appreciate the path dependencies that extend into present and future policy debates, to avoid limiting their success. Many evaluations and examinations of fiscal policies are quantitative and this study adds to the collection of qualitative studies examining the lived experience of SSB consumption and the HPL [1,30,33–35].

Methods

Findings

Our findings are presented here in two primary thematic categories as follows: first our discussions on food, including the HPL as experienced at the time; and second our discussions on health. The narrative presentation of the findings brings together the key sub-themes identified, allowing for the nuanced connections between all of them Table 1. The HPL discussions are expanded on in greater detail given its role as the contextual basis for the study. The findings are expanded on in the discussion through an analysis of culture, policy and the lived experiences of the young people we spoke to as part of the wider context of the HPL.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Overview of primary themes.

https://doi.org/10.1371/journal.pgph.0005901.t001

Participants were split approximately in half between women (n = 37) and men (n = 34) and ranged in age from 18-35. The largest sub-age groups represented were those 31–35 (n = 24) and 21–25 (n = 22). Two of the focus groups were women only and one was men only, with the other four groups mixed. All of the participants were Black South Africans. Two thirds (n = 44) of the young people in the discussions classified themselves as unemployed; thirteen said they were students; and fourteen stated they were employed (either part- or fulltime).

Participants described living in a stressful environment with low incomes and high levels of crime. This could be attributed in part to high levels of unemployment; communal xenophobia at the time towards non-South African black immigrants; and the physical location of townships such as Khayelitsha [37,44]. In addition to the stress of living in Khayelitsha, family structure was identified to be a contextual determinant of what participants ate and drank. The participants lived in one of three types of households: 1) larger, multi-generational households with one or more of income earners, where those bringing in income were the primary decision makers regarding food and budgeting, 2) single, mother-led homes, where often the sole monthly income comes from child benefit grants or 3) a handful of our participants lived independently or with one or two friends or similar aged family members in informal housing on the edges of Khayelitsha, with their income coming from employment.

Discussion

This study used SSB taxes as a lens to explore young people’s immediate, lived experiences; examining how price changes and policy influenced food choices and health knowledge during the period soon after the implementation of the HPL. It offers a critical analysis of culture, policy and lived experiences within the expanded context of the HPL today and the assumed design of fiscal policies targeting commodities, as outlined in Methods. This analysis is placed here within our discussions with young people – consumers in their setting – providing an example of qualitative fiscal policy assessment and insight of a global health intervention in a local SA setting.

The primary argument against the implementation of SSB taxes is that they are by design economically regressive; in that lower income households pay a larger proportion of their income with any additional tax added to a product as compared to higher income households [26,45,46]. The global health community advocates for these policies, with real-world and modelled evidence supporting their use, to shift whole population consumption levels; with lower income groups, that usually have the highest health burdens most significantly impacted overtime [26]. An argument that the poorest will face the largest financial burden of such policies is often used to fight their implementation [26]. Since HPL implementation, several studies have found a decrease in purchasing and consumption in low-income and high-consuming sub-populations in SA.

Reduced purchasing is corroborated by our findings and many of the young people noted the rise in price of Coke and other brand name sugary beverages. Our participants spoke of a shift in purchasing based on the change in price in sugary beverages – even if many of the young people did not know why the price had changed. However, our results show that the shift was most often to less expensive, non-name brand sugary drinks or to purchasing at different times in their pay cycle or budgeting to afford Coke. This is as opposed to making a shift to lower or no sugar, healthier drink options altogether. This is consistent with evaluations of the HPL conducted after our study that found shifts to no-tax sugary beverages [47,48]. This illustrates an initial, acute economic impact of the HPL on this population that had a noticeable and immediate effect on low-income purchasing behaviour at the time.

While the HPL impacted behaviour amongst the young people in Khayelitsha, at the time of this study a large proportion of the participants indicated they had not heard about the HPL until our discussions. This is consistent with the literature and discussions with our local gatekeepers [1,49]. For example, Bosire et al. (2020) investigated perceptions and attitudes among urban South Africans who were living in Soweto on factors that contributed to their SSB intake as well as South Africa’s use of a tax to reduce SSB consumption in the three months prior to the HPL implementation in 2018 [1]. Participants in that study reported frequent SSB consumption and attributed this to habit, addiction, advertising and the wide accessibility of SSBs. Regarding the levy itself, most participants were not aware of the proposed tax and indicated cynicism with regards to the government’s stated motivation in introducing the tax [1]. Studies since our data were collected suggest that even with changes in purchasing and consumption, general knowledge of the HPL has remained low; or even decreased over time [33].

Alongside the literature on knowledge and awareness of the HPL, there are mathematical simulations and modelling studies which were conducted to estimate the potential changes in consumption and related health outcomes prior to the HPL implementation [2,32,50,51]. For example, Manyema et al. (2014) estimated the impact a 20% tax on sugary drinks might have on obesity rates in SA, constructed on estimated consumption changes based on price elasticities and thus calories consumed based on those estimated changes in consumption [32]. Following the HPL implementation, further quantitative evaluations indicate actual decreases in SSB intake [31,52].

As part of critically understanding the context and immediate experience of the HPL, in our discussions both on the health context of Khayelitsha and on the HPL, a noticeable majority of participants indicated they already knew that drinking sugary drinks was not good for their long-term health in some way. This is consistent with the findings of other studies published since HPL implementation [33,47]. This knowledge regarding the un-healthiness of sugary drinks was not linked to any prior knowledge of the HPL. In thinking critically about the role of policies on health, there is historical evidence to support the argument that structural violence – in this case high levels of unemployment and post-Apartheid structural and geographic legacies specific to SA – is a determinant of health [53]. This study, which has identified these wider determinants, adds to the small collection of qualitative studies examining the lived experience of SSB consumption and the HPL, and of SSBs more broadly within a local setting.

As highlighted earlier, the strong brand association, loyalty, and conspicuous consumption seen in our discussions is evidence of the financial and physical legacies of Apartheid, particularly in relation to Coke versus other local fizzy drink brands. We know Coke and the other leading global fizzy and sugary drinks brands have held the majority of the market for many years [22]. This market control – and the social aspiration and status associated with being able to drink Coke – were evident in our discussions. Participants told us that Coke was their preferred beverage and it was also evident in the sheer presence Coke had within the physical environment; Coca-Cola’s bright red signage and distinctive white font was everywhere. The pricing structure of Coke products in SA may have shifted because of the HPL, but at the time of our study the influence of Coke within Khayelitsha had not [54].

Post-Apartheid structural legacies, as previously mentioned, in particular SA’s racialised income inequality and life in townships influence diet, obesity rates and responses to policy. Our participants – who were all Black South Africans, and largely unemployed – reflected the high levels of poverty in SA and is consistent with Black South Africans being consistently in the lowest income groups [21]. Much of this perpetuation of poverty is driven by the geographic and physical legacy of townships. Note for example the distance Khayelitsha sits from the Cape Town city centre. As well as education and employment opportunities, there is evidence globally to support the idea that individuals’ food choice and diet is associated with levels of unemployment and poverty [55].

Furthermore, food insecurity is well-documented in townships, with evidence aligning with our results that people lacked access to food at specific times of the month depending on when they were paid or received benefits, and were often unable to consume the foods they would prefer or knew to be healthy [24,44,56–58]. Further evidence suggests that additional factors influence purchasing and consumption beyond explicit cost impacts from the HPL, such as the prices and availability of products at small retailers (i.e., spazas and tuck shops) in a low-income township [48]. These shops will have been managing changes in price too, passed on from manufacturers. In addition, consumers lack clarity in the use of tax revenue raised by the HPL [35].

The idea that macro-level social, political and economic forces shape individuals’ health, in particular the impact of policy, social structures and individual and community experiences of health, facilitate a critical assessment of lived policy [43,59]. The HPL is by design meant to shape one element of people’s food environments in an effort to reduce their levels of sugar consumption. However, other large socio-cultural structures identified through our discussions were shaping these young people’s lives at the time. Including, but not limited to, the post-Apartheid structural legacies of continued economic and geographic barriers. All paired with an interplay of old (e.g., HIV) and new (e.g., NCD) health realities.

Further to the critical assessment of consumers in their lived realities – and old and new health realities – is the impact of COVID-19 [60,61]. Evidence suggests that, since the start of the pandemic in 2020, responses to COVID-19 have impacted the control, prevention and diagnosis capacity of both HIV and NCDs [61]. Hypertension, diabetes and HIV were the leading co-morbidities of COVID-19 mortality in SA; and the pandemic exacerbated malnutrition, particularly in low-income settings such as townships, because of lockdowns and the broader global response [62–67].

The lived realities in parallel to government responses to old, new and emerging health realities were apparent at multiple points throughout our focus group discussions. Primarily when speaking about the health context of Khayelitsha and scope for policy change, the idea of “giving” or “producing” health was discussed; as a right and expectation. It came up in talking about the legacy of HIV and the SA government’s programs to promote prevention by “giving” out condoms and therefore “producing” better health. It came up when talking about the HPL itself and how the government should not tax those already living on low-incomes, but rather “give” them healthy foods through subsidisation. It also came up when discussing what the government could do to help their health beyond the HPL; they could “give” them more jobs. The “right to health” is advocated by the WHO and enshrined by the United Nations, and our findings reveal that the young people in Khayelitsha had an overwhelmingly collectivist understanding of their government’s responsibility to uphold this right [44,68]. SA developed a comprehensive response, over decades, to the HIV pandemic, however broader structural barriers (e.g., political, economic, socio-cultural) were limiting the HPL in this setting from “giving” young people a healthier food environment for which to make individual decisions.

By focusing on the interplay of culture, policy and lived experiences we examined the role of policy in shaping lives, and health, beyond the global public health discourse and accepted policy mechanisms. This is as opposed to a systematic “policy analysis” which might undertake a stakeholder analysis; or use a specific theoretical structure for policy development and agenda setting; or critique the political dynamics of a specific public policy within a set framework [69–71]. This was a point in time study and a direct or indirect response from the SA government was not included in the study design. However, it can be used to inform future policy refinement and development.

Elements of the policy mechanism that emerged from our analysis – primarily price shifting consumption, public awareness, and revenue for public use – did not fully align in Khayelitsha in the first years after the HPL implementation. While some evaluations to-date do point to shifting consumption and the population impact from the HPL, including among low income consumers, and the overall success of the HPL [27–29,31,34,47]. Our study supports others’ findings that lived realities and expectations of the HPL are not as simple as reductions in purchasing.

No population level policy is entirely equitable across the population, however using qualitative data such as ours from early in the policy’s tenure alongside evaluations overtime should be used to enhance the effectiveness of the HPL and equity issues by addressing, for example, public awareness and wider structural determinants.

Conclusions

This article utilised data from focus group discussions to reflect critically on the HPL’s impact on the day-to-day lives of young people during the period soon after implementation, within the larger context with which it sits. It highlights the role of on-going income inequalities and high rates of unemployment on the lives of young, low-income South Africans at the time. In Khayelitsha, the HPL had yet to impact levels of consumption but had impacted young people financially. This article adds to the literature on structural barriers to health (i.e., racialised unemployment); the broad impact of inequalities on health (i.e., a socio-economic environment limiting healthy food and drink consumption); and the right to health (i.e., government subsidised prevention interventions) [42,43,72]. This study also moves the literature forward by examining the effects of a policy intervention which is popular amongst many public health advocates and policymakers, and is designed to shift some of the health burdens seen across populations. Understanding the recent past of a policy’s life span is important for appreciating the path dependencies that extend into present and future policy debates. Our findings highlight the historical and contextual picture of the HPL success, as well as broader global health interventions in local settings, and provides additional contextual information for future evaluations and policy developments.

The SA government stated a desire to address widespread inequalities and the rising rates of obesity and other NCD risk factors in the country [73]. NCDs sit within a health and social context that has been dominated by the response to HIV and post-Apartheid legacy for over 30 years; and now the aftermath of COVID-19. The SA government’s implementation of the HPL was a population level policy decision supported by public health evidence and seen as a successful model for similar policies across the continent [29]. It is important for future research to examine with attention the social and historical context in which policies like the HPL are implemented and evaluated.

Supporting information

Acknowledgments

We are incredibly grateful to the young people of Khayelitsha who took the time to speak with us. The authors would also like to thank colleagues, including administrative staff, at the School of Public Health and Family Medicine at the University of Cape Town for facilitating the study, including ethics approval; for making introductions to the local community; and providing the referral for the third party Xhosa translator. We would like to also thank the local leader in Khayelitsha who managed the recruitment of our study participants and allowed us the use their church space to conduct our focus groups. The first author would like to thank her colleagues at HealthLumen for their understanding and flexibility during the completion of this study.