Panel Dataset on Energy Transition, Digitalisation and Greenhouse Gas Emissions in Sub-Saharan Africa (2000-2023)

Description

This dataset contains a country-year panel for Sub-Saharan Africa covering the period 2000-2023, assembled for research on energy transition, digitalisation, and greenhouse gas emissions. It combines indicators of energy intensity, electricity access, renewable energy use, digital connectivity, economic development, and population dynamics in a format suitable for panel-data analysis and replication. The main file, SSA_energy_digital_emissions_panel_2000_2023.csv, provides the processed analytical dataset. The companion Excel file documents the dataset structure and includes variable descriptions, country code mapping, and the full missingness and interpolation report. A key feature of the dataset is its alignment with the study’s empirical design. Greenhouse gas emissions are retained through 2023, whereas all non-GHG variables are restricted to end in 2021. Missing observations were treated within country using linear interpolation and extrapolation in Stata’s ipolate, epolate, after which the 2021 cutoff for non-GHG variables was re-imposed. The Excel documentation file reports missingness before and after processing, both overall and by year and country, together with an imputation audit for transparency. Files included: SSA_energy_digital_emissions_panel_2000_2023.csv Companion Excel file: dataset summary, variable notes, country codes, and missingness report Source: World Bank, World Development Indicators (WDI)

Steps to reproduce

Step 1 - Data acquisition: Download annual country-level data from the World Bank’s World Development Indicators (WDI) for Sub-Saharan Africa covering 2000-2023. The study uses the following indicators: ein = energy intensity (energy used per unit of output), elec = access to electricity (share of population with electricity access), ren = renewable energy consumption (share of total final energy use from renewables), ghg = total greenhouse gas emissions excluding LULUCF (overall emissions measure), mob = mobile cellular subscriptions (mobile connectivity), int = individuals using the Internet (internet penetration), gdppc = GDP per capita (income level), and pop = population growth (annual population change). Step 2 - Country coverage: Restrict the sample to 46 Sub-Saharan African countries. Eritrea and South Sudan are excluded because of extensive missing values that prevent consistent panel construction and reliable interpolation over the study horizon. The remaining countries form the final analytical sample. Step 3 - Data import and harmonisation: Import the downloaded WDI series into Stata and merge them by country code and year. Retain the study variables only and rename them to the analysis labels: ein, elec, ren, ghg, mob, int, gdppc, pop. Then create a numeric panel identifier using egen cid = group(iso3c) and sort the data by cid year. The final structure should be: cid, iso3c, country, year, ein, elec, ren, ghg, mob, int, gdppc, pop. Step 4 - Study design rule: Keep the panel for 2000-2023, but apply the study timing structure so that ghg is retained through 2023, while all other variables are restricted to end in 2021. In Stata, this is done by setting non-GHG variables to missing after 2021, for example: foreach v in ein elec ren mob int gdppc pop { replace \v' = . if year > 2021 }`. This ensures that the explanatory variables end strictly in 2021 while the emissions outcome continues through 2023. Step 5 - Missing-data treatment: Within each country, fill missing observations using linear interpolation and extrapolation in the style of Stata’s ipolate, epolate. Run by cid: ipolate variable year, gen(variable_filled) epolate for each series. After filling, re-impose the 2021 cutoff for all non-GHG variables, so that any filled values after 2021 are set back to missing. This preserves the intended design: GHG available through 2023, all explanatory variables ending in 2021. Step 6 - Final output: Save the processed file as the final panel dataset and prepare a companion documentation file reporting variable definitions, country codes, and missingness before and after interpolation. This produces the replication-ready panel used in the study.

Institutions

• Asia Pacific University of Technology & Innovation

  Kuala Lumpur, Kuala Lumpur
• Universiti Putra Malaysia

  Selangor, Seri Kembangan

Categories

Licence