Research output from the South African Reserve Bank (SARB) and Economic Research South Africa (ERSA) have declined over recent years. ERSA’s decline in working papers published is particularly pronounced if one expresses output relative to the organisation’s budget. After a post-2010 peak in output in 2016-17 at about 15 papers per R1 million rand of budget, its output of published papers remained fairly stable relative to total expenditure until 2020-21. Putting inflation aside, output has declined to below 1 paper per million of budget since then.
It is worth comparing ERSA’s output to that of the SARB. In 2010, for example, ERSA produced roughly 20 times the number of working papers SARB produced (although some of these were also published as part of other working paper series as well). In 2023, on the other hand, ERSA and SARB produced similar numbers of papers.
Unfortunately, one cannot compare the budgets of these institutions as SARB does not publish information on how its budget is apportioned across policy departments. Both institutions also have several functions beyond publication of research so one cannot directly compare the cost differences across the institutions. Nevertheless, the chart below expresses the number of SARB working papers produced over time relative to the SARB’s total budget to give a sense of the prioritisation of research at the central bank. The profile for output to budget follows a similar profile to working paper publications. SARB’s working paper publications have declined from a peak in 2021, with output relative to the Bank’s overall budget declining to near its average since 2010.
It is worth pointing out that there may be inaccuracies in the data presented. In ERSA’s case, there are discrepancies between the number of papers listed in its annual reports and those available on the website, reflecting paper availability as well as date labelling mistakes (see charts below). It is also worth noting that working paper numbers are inflated slightly by inclusion of journal articles in ERSA’s paper count early in the sample and later in the sample by cross-publication of some Reserve Bank working papers. I present an alternative paper count where papers by SARB authors or published SARB working papers are excluded. We estimate that excluding SARB-related papers would reduce the total ERSA paper count between 2010 and 2023 by 10-20%.
In the case of SARB, papers are not always loaded on their website in a way that allows Google and Research Papers in Economics (RePEc, a repository of Economics publications tracking citations and paper reads) to pick these up. As a result, there are several duplicate papers listed, and several papers that are not picked up at all by RePEc.
Footnote
We will compare the quality of research across university economics departments and research organisations producing economic research in a future post. We showed in these earlier posts (here and here) that the South African Journal of Economics and SARB have low global rankings in terms of how influential their publications are.
For a summary of research by the CSIR, see this post.
