Table 1

Data inputs and corresponding sources used in modelling of the equity impact of the minimum unit pricing policy for alcohol in South Africa

InputWealth quintiles (QI=poorest)*Source
Alcohol consumption, prices and elasticities
Prevalence of drinking27%30%33%35%38%SA DHS 2016
Prevalence of heavy drinking (more than 15 standard drinks per week)14%14%16%17%20%SA DHS 2016
Mean individual baseline consumption (standard drinks per week)20.621.420.921.720.7SA DHS 2016
calibrated to Euromonitor
Mean price per standard drinkInternational Alcohol Control Study (2014) adjusted for inflation to 2018 prices
Gibbs et al14
Occasional bingeR8.0R10.0R10.1R13.4R11.1
Price elasticity by drinker groups†Van Walbeek and Chelwa41 authors’ calculations (webappendix section 3)
Gibbs et al14
Occasional binge−0.29−0.29−0.17−0.17−0.17
Heavy drinkers−0.24−0.24−0.14−0.14−0.14
Share of disease at baseline‡
HIV20%36%32%9%3%Authors’ calculations using
GHS 2018
Intentional injury road injury
Liver cirrhosis
9%29%26%26%10%Authors’ calculations using
GHS 2018
Breast cancer7%7%22%18%47%Authors’ calculations’ using
GHS 2018
Disease-related expenditure and utilisation
Proportion of disease-related expenditures paid as OOP21%18%41%56%82%Saxena et al.22
HIV utilisation rates63%71%69%60%89%Authors’ calculations using GHS 2019 (webappendix section 5)
Trauma care utilisation rates—intentional injury39%40%40%40%47%Authors’ calculations using GHS 2019 data plus Matzopoulos et al.42 (webappendix section 5)
Trauma care utilisation rates—road injury18%19%18%18%22%Authors’ calculations using GHS 2019 data; Matzopoulos et al.42 (webappendix section 5)
Healthcare utilisation rates—liver cirrhosis52%55%54%53%63%Authors’ calculations using GHS 2019 (webappendix section 5)
Healthcare utilisation rates—breast cancer52%56%50%68%89%Authors’ calculations using GHS 2019 (webappendix section 5)
Labour and productivity
Labour force participation62%50%55%64%74%Authors’ calculations using
GHS 2019 data
Annual income per capita
610027 40049 30095 600408 900Authors’ calculations using GHS 2019 data deflated to 2018
Absenteeism (days per year)
 HIV1414141414Maffessanti and Lee-Angell43
Intentional injury1010101010Bola et al44
Road injury1818181818Parkinson et al45
Liver cirrhosis63333Matzopoulos et al46
Breast cancer66666Tangka et al47 (webappendix section 6)
  • *Wealth quintiles defined as the asset index measure provided in the SA DHS data; authors used an ordered choice regression model to predict wealth quintiles for the International Alcohol Control (IAC) data set; income quintiles used as a proxy for wealth quintiles in GHS data.

  • †Drinker groups: moderate=less than 15 standard drinks per week; occasional binge=less than 15 drinks per week but drinks more than five on at least one occasion; heavy=15 or more standard drinks per week. Standard drink=12 g or 15 mL of pure ethanol.

  • ‡Share of disease at baseline indicates how the cases of the disease/injury conditions are distributed among the quintiles.

  • DHS, Demographic and Health Survey; GHS, General Household Survey; OOP, out-of-pocket; SA, South Africa.