Finance minister Tito Mboweni’s medium-term budget policy statement (MTBPS) has in effect cancelled President Cyril Ramaphosa’s reconstruction and recovery plan. If implemented, it will dash hopes of a recovery and prolong SA’s deepest recession in a century. However, there are reasons to believe that the planned austerity measures of R300bn may not be implemented and that the MTBPS might have to be changed.
According to the Treasury’s rosy forecasts, “the economy will only recover to 2019 levels in 2024”. Given that the Treasury’s forecasts have been wrong (and too optimistic) for the past decade and the population will continue to grow, it will take longer than five years for GDP per capita to recover to 2019 levels. With current policies, SA will have a second “lost decade” until 2030. During the first “lost decade”, between 2009 and 2019, GDP per capita did not grow.
The MTBPS says: “Compared with the 2020 budget, main budget non-interest expenditure excluding technical adjustments will be reduced by R60bn in 2021/2022, R90bn in 2022/2023 and R150bn in 2023/2024.” These cuts totalling R300bn will result in three years of declining real non-interest spending — 6.9% in 2021/2022, 2.6% in 2022/2023 and 3.4% in 2023/2024. The cuts are 1.1% of the Treasury’s projected GDP in 2020/2021, 1.6% in 2022/2023 and 2.5% in 2023/2024.
According to the Treasury forecasts, government consumption spending will be the only drag on GDP growth over the next three years. It will decline by 2.5% in 2021, 2.4% in 2022 and 3.6% in 2023. There will be no infrastructure-led recovery. The MTBPS says: “SA is on course for its fourth consecutive year of contracting investment.” During 2021, there will be a fifth consecutive year of declining investment. This means the Treasury believes the president’s plans to increase investment will not succeed in the short term.
A major reason for the Treasury missing its own revenue targets every year during the past five years — by R250bn between 2014/2015 and 2019/2020 — is that it does not take into account the negative effect of its austerity measures on GDP growth. In 2012 the IMF found that its forecasts were too optimistic for countries that were implementing austerity programmes because such policies were causing more economic damage than had previously been expected. Ann Pettifor, a UK economist, explained the IMF findings: “This means a reduction of 1% in public expenditure will result in a reduction in national income of 1.5%.”
However, the Treasury has developed a fabrication that the fiscal multiplier — the additional GDP generated by an increase in government expenditure — could be less than zero in SA. Therefore, there is a free lunch that can allow the government to cut spending without any effect on GDP growth. There can even be an expansionary contraction. But that is a zombie idea, as Paul Krugman, a winner of the Nobel prize in economics, says.
The MTBPS identified R310.6bn of cuts to public sector wages — R36.5bn in 2020/2021, R83.2bn in 2021/2022, R116.1bn in 2022/2023 and R74.8bn in 2023/2024. But the savings from the proposed wage freeze will be only R79.3bn during the next two years — R28.7bn in 2021/2022 and R50.9bn in 2022/2023 — using the February 2020 budget baseline. The only way to make up the R194.8bn shortfall over the next three years would be to retrench thousands of public servants.
Already there is talk that the government has agreed to honour the last year of the 2018 wage agreement and that some ministers have told union leaders they know nothing about a wage freeze. If this is true, there will be no savings of R36.5bn this year. It could also mean there is little support in the government, except for a few people in the Treasury, for a wage freeze, let alone mass retrenchments. Under this scenario, the whole basis of the MTBPS is not credible.
• Gqubule is founding director at the Centre for Economic Development and Transformation.




Would you like to comment on this article?
Sign up (it's quick and free) or sign in now.
Please read our Comment Policy before commenting.