The economic impact of a sugar-sweetened beverages tax in South Africa.

Masters Degree. University of KwaZulu-Natal, Pietermaritzburg.

The newly implemented Health Promotion Levy (HPL) in South Africa (SA) imposes a tax on Sugar-Sweetened Beverages (SSBs) containing more than four grams of sugar per 100ml. The HPL is legislated in terms of Chapter VB of the Custom and Excise Act No. 91 of 1964, on SSBs manufactured in SA or imported into South Africa. There is a disagreement between the proponents of the tax (government) and opponents of the tax (beverage and sugar industry). The proponents of the tax contend that the SSBs tax will reduce obesity and boost SA’s economic growth by reducing government spending on prevention and treatment of obesity-related non communicable diseases (NCDs). The opponents of the tax, on the other hand, contend that the tax will be detrimental to the growth of the SA’s economy by reducing the beverage and sugar industry returns, which will, in turn, lead to labour disemployment. There are relatively fewer published empirical studies to date in South Africa that have examined the economic impact of the SSB tax. Most of the published empirical literature to date has focused more on the health impact of the SSB tax. Furthermore, there’s no empirical study that has examined the impact of the SSBs tax on the South African sugar industry. Given this background, this study aims to investigate the economic impact of the SSBs on the South African sugar industry. This is achieved by (1) assessing the impact of the SSB tax on the quantity of sugar demanded in the domestic sugar market, and (2) determining the impact of the SSBs tax on the revenue earned by the South African sugar industry from local sugar sales and Recoverable Value (RV) price received by sugarcane growers’. The analysis used time-series data for the period 1977-2019 and involved the use of the simultaneous equations regression method (Three Stage Least Square regression (3SLS)) and the Vector Error Correction Model (VECM). To deal with challenges associated with time-series data, several diagnostic tests were employed. In addition, the direction of causality between domestic sugar demand and supply variables was examined using a Wald test for Granger’s causality. Furthermore, a Johansen’s Cointegration test was used to check the presence of long-run relationships between domestic sugar demand and supply variables. The static demand and supply model (3SLS regression) results revealed that variables such as the real domestic price of sugar, previous period sugar consumption, and consumers’ real disposable income influence the domestic sugar demand in South Africa. Whereas variables such as the real domestic price of sugar, technological changes, and the real price of sugarcane in the previous period influence the domestic sugar supply in South Africa. From the 3SLS regression model, the own-price elasticity estimates of domestic demand and supply for sugar in South Africa were estimated to be -2.652, and 0.838, respectively. Implying that the domestic demand for sugar in South Africa is relatively more responsive to own-price changes, whereas the domestic supply of sugar is relatively less responsive to own-price changes. From the VECM, the short-run own-price elasticity estimates of domestic sugar demand and supply were estimated to be -0.301 and 0.762, respectively. Implying that in the short run both domestic sugar demand and supply are relatively less responsive to domestic sugar price changes. In addition, the long-run own-price elasticity estimates were estimated to be -2.243 and 1.809, respectively for the domestic sugar demand and supply equations. This implies that in the long run both domestic demand and supply for sugar in South Africa are relatively more responsive to domestic sugar price changes. This is because over time the South African sugar producers are likely to discover more alternative products to produce or develop close substitute in consumption that contains less sucrose sugar. Similarly, domestic sugar consumers are likely to discover more sugar substitutes in the long run. Based on the 3SLS model's own-price elasticity estimate of domestic sugar demand (-2.265), the quantity of sugar demanded by domestic consumers of sugar was estimated to have decreased by 184 169 tons, causing a revenue loss of about R1.68 billion in the domestic sugar market. Based on the VECM short-run own-price elasticity estimate of domestic sugar demand (-0.301), the quantity of sugar demanded in the domestic sugar market was estimated to have decreased by 209 01 tons, leading to a domestic sugar revenue loss of about R190 million. Lastly, using the longrun own-price elasticity estimates of the domestic sugar demand (-2.243), the quantity of sugar demanded in the domestic sugar market was estimated to have decreased by 154 917 tons, leading to a domestic sugar revenue loss of about R1.41 billion. Furthermore, by applying the sugar industry division of proceeds formula, the RV price received by sugarcane growers’ during the 2019/20 season was estimated to be lower by 9.167% when compared to the RV price that growers could have received if there was no SSBs tax. Given the negative impact of the SSBs tax on the South African sugar industry’s financial position, the study recommended the South African sugar industry to invest more on other alternative income streams, i.e., expand the production of biofuels using sugarcane as a feedstock. In this regard the study recommended the South African government to support the South African sugar industry stakeholders with legislation that will allow them to generate revenue from expanding the production of biofuels, electricity co-generation, and biochemical feedstock, amongst others. For instance, provides legislations that are necessary to promote the use of ethanol-blended petroleum and legislations necessary to enable sugar mills to sell electricity from co-generation. Keywords: Health Promotion Levy (HPL), South African Sugar Industry, Three-Stage Least Square (3SLS), Two-Stage Least Squares (2SLS), Vector Error Correction Model (VECM), Sugar tax.