Shiny web app (in development)
This website is:
A guide to Aidan Horn's TERS Dropbox folder.
Publications are still in process, so the folder will be linked here after the last publication (I could link a copy of the folder here with files omitted while research is still underway). In the meantime, direct links to important files are provided on the pages here.
A front page for our <Shiny social insurance simulation web app>
Workshop material: An overview of Andrew Donaldson and Aidan Horn's research on earnings levels and social insurance in South Africa, which began when the lockdown started in April 2020. We work from the Southern Africa Labour and Development Research Unit (SALDRU), School of Economics, University of Cape Town.
The aim of this website is to make our research transparent, reproducible and able to be utilized to undertake further research on simulating hypothetical social insurance schemes or macroeconomic modelling. It provides a link between the expertise held by the researchers and any interested researcher who wants to use the work.
Our simulation models are in Stata code at "Common Scripts/simulation.do" in our project folder, although the web app also reproduces the models in an interactive format, and we provide the simulation models as an Excel file at "LMDSA 2018/DataOUT/Mean earnings of tenths (Excel)/...in progress". The models simulate the total cost to government of various social insurance schemes, allowing for adjustments to be made to coverage rates by industry and the values of the parameters of the formulae. The social insurance formuale that we analyse so far include:
Temporary Employee/Employer Relief Scheme (TERS)
Unemployment Insurance Fund (UIF)
A new, improved "tangential linear" UIF formula, as outlined in Horn's (2021) SALDRU working paper
Employment Tax Incentive (ETI)
A modified ETI formula
A flat rate benefit
Further schemes can be incorporated in future, and we intend on using this project to advise on policy changes to pensions in South Africa.
The web app shows results other than the total cost, such as: aligning the simulated cost with actual, known expenditure; the mathematical formula of the benefit; a graph showing the benefit level at different earnings levels; the density graph of the income replacement rate (IRR); back-end code showing the algorithm of the simulation; and a graph showing where benefits are concentrated on the earnings distribution.
The first paper simulates the total cost of different hypothetical Unemployment Insurance Fund (UIF) formulae.
Horn, A.J. 2021. South Africa's Unemployment Insurance Fund Benefit Function: A Mathematical Critique. (SALDRU Working Paper Number 276). Cape Town: SALDRU, School of Economics, University of Cape Town. Available: http://opensaldru.uct.ac.za/handle/11090/1004
This paper highlights the unnecessary complexity of South Africa's Unemployment Insurance Fund (UIF) benefit function, known as the Income Replacement Rate (IRR), and the disadvantageous manner in which the IRR is low for most earners. Possible alternative formulae are described, along with the implications for total expenditure on the UIF. The paper recommends simpler (and more optimal) formulae.
In the following paper, we used the Temporary Employer/Employee Relief Scheme (TERS) simulation model to ascertain how TERS benefits were distributed across the earnings distribution, by industry. This paper builds on the results of our autoregressive modelling of aggregrate earnings levels.
Horn, A.J. & Donaldson, A.R. 2021. Temporary Employee/Employer Relief Scheme Outcomes. (SALDRU Working Paper ). (Forthcoming).
The Temporary Employee/Employer Relief Scheme (TERS) distributed payments to furloughed workers in South Africa, during the covid-19 lockdown. This paper investigates, across industries, what proportion of Unemployment Insurance Fund (UIF) contributors were covered by the scheme, and how TERS payments were distributed across the earnings distribution. We find that, because the UIF is a contributory social insurance scheme, mainly middle- and upper-income earners benefited from TERS.
Aidan Horn (hrnaid001@myuct.ac.za)
Andrew Donaldson (andrew.donaldson@uct.ac.za)