Methodology Overview

The TCDI South Africa website was developed by Development Gateway in partnership with the University of Cape Town. The website is intended to respond to the needs of government, civil society and academic stakeholders by providing access to demand-driven, high-quality data on national tobacco prevalence, products, and policies. The design, functionality and content for the website are led by the South African tobacco control community’s needs. This document outlines our overall website methodology.

1. Identifying Stakeholder Needs

In January and February 2020, the TCDI team selected stakeholders from 12 tobacco control organisations (see full list here). We conducted short interviews with each one to explore how they make decisions, the data used for these decisions, potential data gaps, and their preferred format for receiving data (e.g. graphs, short paragraphs, infographics). Stakeholders included representatives of government, civil-society-organisations and academia.

2. Reviewing Stakeholder Feedback and Prioritising Needs

The stakeholder feedback was documented during the interviews and later analysed to identify cross-cutting data gaps and the stakeholders’ preferred formats for receiving information. The feedback was distilled according to the frequency and urgency with which each was mentioned by the stakeholders. This process produced 6 priority data gaps and 7 data format guidelines:

Priority data gaps

1. Tobacco Prevalence
2. Illicit Tobacco
3. Impact of Tobacco Policies
4. Health Burden of Tobacco
5. E-cigarettes
6. Tobacco Agriculture

Preferred data format 

1. Short, concise and attention grabbing 
2. Simple and easy to understand 
3. Clear visuals 
4. Factsheets and individual stories 
5. Trustworthy data with clearly stated sources
6. A data repository with data from multiple themes in one place
7. South Africa or Africa-specific information

The preferred data gaps and format were confirmed with the stakeholders in an interactive workshop in June 2020, during which each data gap was discussed further to identify the sub-thematic components stakeholders wanted to learn more about (for instance, for the tobacco agriculture theme, various sub-themes, such as employment and health risks etc., were identified).

3. Research and Analysis

The TCDI team conducted research and data analysis for each theme, including systematic research of reliable primary data (e.g. data collected by academic institutions, global foundations or governments) and secondary data (e.g. peer-reviewed academic articles). Where required, the primary data were analysed by the team’s economists using statistical software such as Stata and R. The stakeholder needs were reviewed against the primary and secondary data available, relevant statistics and graphs were produced, and explanatory text was written.

4. Website Design

Using the 7 data format guidelines as a starting point, the TCDI team designed the visual elements and functionality of the website. This process included designing the visual elements (infographics, chart formats and colours) and the user functionality (menus, data exports and search functionality). 

5. Expert Review

At least two expert tobacco-control academics were selected per theme to review the quality and reliability of the content. Their feedback was integrated into the text and the design of the website.

6. Stakeholder Review

The 19 tobacco-control stakeholders were invited to small online group discussions during which they opened the website link for the first time, shared their screens and narrated their experience as they navigated through the website. The TCDI team documented the process, noting down any areas for improvement, and evaluating whether the information addresses users’ data needs and the visuals are clear and easy to understand. All feedback was collated to produce a stakeholder feedback matrix.

7. Finalise Design

Using the stakeholder feedback matrix as a guide, the TCDI team conducted additional desk research (where required) to find and analyse data to fill any remaining data gaps. In addition, visual elements and functionality were redesigned wherever necessary. The changes were incorporated into the website themes and the website was published online.

8. Update Periodically

Periodically, the TCDI team will consult with experts, review newly available data, and update the website to reflect any changes, where necessary. Please fill out the Showcase Your Work form if you have newly available data or publications you would like to share.

Smoking estimates for Kenya and Nigeria were calculated using Demographic and Health Survey (DHS) data. 2014 DHS data are used for Kenya

and 2018 DHS data are used for Nigeria. Data are only collected for women between the ages of 15 and 49, whilst for men the maximum age ranges from 55 to 59 depending on the survey year. For the purposes of comparability, we restricted the analysis to those aged 15-49 for both men and women.
DHS calculates survey weights data for men and women separately and reports smoking prevalence separately. The TCDI team calculated the combined prevalence rates by reweighting the data based on population estimates from the United Nations Department of Economic and Social Affairs Population Division. This is done by multiplying the weights in the DHS data by the proportion of males and females in the overall population in the 15 – 49 age group. The prevalence estimates are calculated by dividing the number of smokers in the sample by the total sample and multiplying by the adjusted weights.

The Policies page discusses the various tobacco-control measures used by governments and presents estimates of their effectiveness in reducing tobacco use. In addition to drawing on secondary sources, data analysis was conducted by the TCDI team to illustrate the historical relationship between excise taxes, prices, and the consumption of cigarettes.

Historical Analysis of Excise Taxes, Prices and Consumption

These graphs depict the relationship between the price of cigarettes, the excise tax levied on cigarettes, and consumption levels of cigarettes between 1961 and 2018. 

The Economics of Tobacco: Loss in Productivity due to Tobacco Use

In the Economics of Tobacco section we explore how the economic cost of tobacco (estimated to be R42 billion) could be spent on alternatives such as in low-cost housing, salaries for teachers, and child support grants. Below we present the sources and calculations.

Reconstruction and Development Programme (RDP) houses – According to South African government records, the maximum cost of building an RDP house was R183 856 in 2018 (equivalent to R166 939 in 2016 prices).

If the R42 billion lost to tobacco in 2016 was put to use building houses, it could have built over 250 000 RDP houses in that year.

Teacher salaries – The median secondary school teacher’s annual  salary was R232 000 in February 2021 (equivalent to R197 110 in 2016 prices).

If the R42 billion lost to tobacco in 2016 was put to this use, it could have paid about 215,000 teachers for a year.

Child Support Grant (CSG) – The CSG was R440 per month or R5280 per year in 2020 (equivalent to R4794 per year in 2016 prices).

If the R42 billion lost to tobacco in 2016 was used to pay Child Support grants, approximately 8.8 million children could be given the CSG for a year.

The Tobacco Excise Tax Simulation Model (TETSiM)

Assumptions in the Model

1. Conservative outcomes: Tobacco industry undershifts the tax (reduce net-of-tax price). The net-of-tax price is reduced by between 15 and 30 percent of the change in the excise tax.
2. Extreme outcomes: Tobacco industry overshifts the tax (increases net-of-tax price). The net-of-tax price is increased by between 10 and 30 percent of the change in the excise tax.
3. Perfect shifting: When we assume that the industry perfectly shifts the tax, there is a 0% increase in the net-of-tax price (the industry price stays the same).
4. Product segments: Five product segments, including imports, premium, popular, discount and illicit.
5. Illicit cigarette market: 34% of the total cigarette market. This is based on various illicit trade studies in South Africa, and relevant for roughly the 2019-2021 period.
   6
   7
   8
6. Price elasticity of demand: Range -0.4 to -0.8, with absolute elasticity being greater for lower income groups, and smaller for higher income groups.
   7
7. Cross-price elasticity of demand: Range 0 to 0.3, with 0 for imports and the most expensive brand category, and 0.3 for substitution into illicit brands.

Note:

The South African TCDI team conducted the South African E-Cigarette Survey 2022 (SAES 2022) between January and September in 2022. The survey aimed to provide data related to the use of electronic nicotine and non-nicotine delivery systems (ENDS/ENNDS) and heated tobacco products (HTPs) (collectively termed “novel products” in the study’s reports) amongst adults in urban South Africa. The Survey was implemented by Development Gateway in partnership with the Research Unit on the Economics of Excisable Products, based at the University of Cape Town, and funded by a grant from the Bill & Melinda Gates Foundation. The analysis was conducted using the statistical software, Stata.

The data collection process was designed to be representative of adults (18 years and older) living in urban metro and non-metro areas in South Africa. South Africa is split into 103 576 enumeration areas (EAs). This EA frame was overlaid with information from credit bureaus. The data were stratified by province, metro/non-metro areas, population group and neighbourhood lifestyle index (NLI), which is a proxy for income. Each EA in the sampling frame is allocated an average overall NLI. Read more on how the NLI was calculated here: https://geoterraimage.com/nli-summary

The data were collected between January and August 2022 in a telephonic survey conducted by the survey company, AskAfrika. In total, 399,810 call attempts were made; of these, 10% were reached, and 90% were unreachable (e.g. no one answered or the number was busy) or unusable (e.g. the user was deceased or their number was blacklisted). Of the calls answered, 51% provided usable data, 10% refused to do the survey, while 38% did not complete the interview. Therefore 5% (21,263) of the total phone numbers drawn in the sample resulted in complete surveys which were used in the analysis. Ultimately, there were 1,255 surveys completed by respondents that had a history of novel product use, 740 respondents had a history of smoking combustible cigarettes regularly (and no history of novel product use) and completed detailed questionnaires, and the remaining 19,268 respondents were a mixture of smokers and non-smokers and they completed short questionnaires.

The novel product prevalence rates reported on the website were calculated by dividing the number of novel product users in the sample by the total sample and multiplying by the weights. 

where i is an individual in a survey, 

x_i denotes each individual i in the population group of interest (x), 

n_i denotes novel product user i in the population group of interest, and 

w_i denotes the survey weight for individual i.

We tested whether novel product prevalence rates were statistically different between population groups of interest, by comparing confidence intervals for each estimate. The TCDI team calculated prevalence rates for the following groups:

We also explored the sequence of use for dual users. Respondents reported which product they had started first, and in the event of quitting, they reported the time since they had quit. These variables were used to assess which product was used first, and the sequence in which quitting happened. Finally, the data were analysed to understand perceptions of the addictiveness, health risks, and expense of novel products relative to combustible cigarettes.

In addition to the analysis conducted using the South African E-Cigarette Survey 2022, the TCDI team compiled research from secondary sources to discuss themes such as novel products and health and novel product policy, among others.

Download the South African E-Cigarette Survey 2022 dataset, user manual, and questionnaire below. The data and accompanying documents are also available for download from DataFirst.

Note:

For the Illicit Trade, Health Burden of Tobacco, and Agriculture themes, the TCDI team did not conduct data analysis but rather drew on secondary sources. The team undertook their initial data search and literature review between November and December 2020; and subsequently between November and December 2022. Secondary data searches occurred across various platforms, including data repositories, academic journals, and online libraries. The team also reviewed resources from trusted organisations such as the World Health Organisation and the International Agency for Research on Cancer. The TCDI team consulted with the Research Unit on the Economics of Excisable Products (REEP) and subject matter experts, who advised on data sources and reviewed each page before publication.