Public Economy Project 

Michael Sachs, Rashaad Amra, Thokozile Madonko and Owen Willcox 

 

June 2023 

  

AUSTERITY WITHOUT 

CONSOLIDATION 

Fiscal policy and spending choices 

in Budget 2023 

SCIS WORKING PAPER SERIES 

 
SCIS Working Paper | Number 60 



 

 
 

About the Authors 
 
Michael Sachs, Rashaad Amra, Thokozile Madonko and Owen Willcox are the core project team members of the 
Public Economy Project of the Southern Centre for Inequality Studies at the University of the Witwatersrand. 
 
  



 

 
 

 

CONTENTS 
 
INTRODUCTION          1 
 
PART I: MACRO-FISCAL CONTEXT – A WORSENING FISCAL IMPASSE  3 

Consolidating spending          
Containing government’s wage bill        
A changed macro-fiscal outlook         
 
PART II: MEDIUM TERM ALLOCATION AND POLICY CHOICES   11 

Introduction 
Education 
Healthcare 
Social protection 
Criminal justice and Defence 
Public infrastructure and government buildings 
Economic development and public employment 
 
PART III: IMPLICATIONS AND CONCLUSIONS      24 

The budget as an opening gambit amidst policy disarray 
Is permanent austerity feasible?  
Escaping the fiscal impasse? 
 
REFERENCES           28 

 

 

  



 

 
 

Abbreviations 
 

AIDS Acquired Immuno-Deficiency Syndrome 

ART Anti-Retroviral Therapy 

BR Budget Review  

CHW Community Health Workers 

CIT Corporate Income Tax 

Covid-SRD Covid19 Social Relief of Distress Grant 

CPI Consumer Price Index 

CWP Community Works Programme 

DBE Department of Basic Education 

DOD Department of Defence 

DPSA Department of Public Service and Administration 

ECD Early Childhood Development 

ENE Estimates of National Expenditure 

EPWP Expanded Public Works Programme 

FATF Financial Action Task Force 

GDP Gross Domestic Product 

FFC Financial and Fiscal Commission 

GEAR Growth, Employment and Redistribution Strategy 

HIV Human Immunodeficiency Virus 

IMF International Monetary Fund 

LTSM Learner and Teacher Support Material 

MTBPS Medium Term Budget Policy Statement  

MTEF Medium Term Expenditure Framework 

NDOH National Department of Health 

NDP National Development Plan 

NGO Non-Governmental Organisation 

NHI National Health Insurance 

NPO Non-Profit Organisation 

NSF National Skills Fund 

NSFAS National Student Financial Aid Scheme 

ODI Overseas Development Institute 

OECD Organisation For Economic Cooperation and Development 

PEP Public Economy Project 

PRASA Passenger Rail Agency of South Africa 

PSET Post-School Education and Training 

RAF Road Accident Fund 

SANDF South African National Defence Force 

SANRAL South African National Roads Agency Limited 

SARB South African Reserve Bank 

SASSA South African Social Security Agency 

SCIS Southern Centre for Inequality Studies 

SETA Sector Education and Training Authority 

SOC State-Owned Company 

TB Tuberculosis 

TCTA Trans-Caledon Tunnel Authority 

TVET Technical And Vocational Education and Training Centre 

UIF Unemployment Insurance Fund 

VAT Value Added Tax 

VIP Very Important Person 

WEO IMF World Economic Outlook 

  



 

 
 

Figures 
 

Figure 1: Faltering growth and deepening fiscal imbalances ................................................................................................ 3 

Figure 2: Budget plans and budget outcomes ......................................................................................................................... 4 

Figure 3: Spending growth and inflation ................................................................................................................................. 5 

Figure 4: Government wage bill trends ................................................................................................................................... 6 

Figure 5: Fiscal outlook: Budget 2023 and PEP baseline ..................................................................................................... 9 

Figure 6: Annual spending by policy function ..................................................................................................................... 11 

Figure 7: Medium-term spending growth (2019-2024) ...................................................................................................... 12 

Figure 8: Basic education spending growth.......................................................................................................................... 13 

Figure 9: Post-school education spending growth (2019-2024) ....................................................................................... 14 

Figure 10: Index of Higher education spending growth .................................................................................................... 15 

Figure 11: Students enrolled in universities .......................................................................................................................... 15 

Figure 12: Real spending per learner ..................................................................................................................................... 15 

Figure 13: Healthcare spending growth (2019−2024) ........................................................................................................ 16 

Figure 14: Real spending per user on anti-retroviral therapy ............................................................................................ 17 

Figure 15: Social protection spending growth (2019-2024) ............................................................................................... 18 

Figure 16: Real monthly value of old-age grant and child support grant ........................................................................ 19 

Figure 17: Police employment ................................................................................................................................................ 20 

Figure 18: Criminal justice and defence spending growth (2019-2024)........................................................................... 21 

Figure 19: Public infrastructure spending growth (2019-2024) ........................................................................................ 22 

Figure 20: Economic servcies spending growth (2019-2024) ........................................................................................... 23 

 

 

 

Tables 

Table 1: Revised GDP growth outlook ................................................................................................................................... 7 

Table 2: Temporary Covid-related expenditures ................................................................................................................. 16 

 

 

  



 

Page | 1  
 

FISCAL POLICY AND SPENDING CHOICES 

 IN BUDGET 2023 
Public Economy Project | June 2023 

 
 

INTRODUCTION  
 
Government’s budget is a critical instrument for building a common society and overcoming the segregation and 
inequality that thwarts growth and development in South Africa. It is also a vital element of the social contract, 
transferring income and providing services that hold our fragile society together. South Africa, however, finds itself 
in a worsening fiscal impasse. Continuous austerity over the last decade has eroded the quality and value of public 
services on which the majority of South Africans rely. Spending cutbacks have been strongly intensified since the 
Covid-19 pandemic, with pay increases for government employees held well below the rate of inflation, and across-
the-board spending reductions cutting deeply into healthcare, basic education, criminal justice, and social services.  
 
Funding for social provision depends on rising national income, but South Africa has been stagnating for a decade. 
Recently, the growth outlook has deteriorated further, as the dire implications of the failure of electricity supply and 
commodity export infrastructure have become clearer. Real economic growth is now even lower than the miserable 
estimates anticipated in the budget. Meanwhile, financial conditions have tightened as capital outflows have 
stabilised at higher interest rates, with South Africa facing an elevated country risk premium on top of global rate 
hikes.  
 
After several years of painful efforts to stabilise the public finances, this weakening of macro-fiscal conditions has 
made the goal of debt stabilisation even more distant. Despite intensifying austerity, South Africa appears no closer 
to consolidating its fiscal position. Slower growth and higher interest rates require further cuts in spending or tax 
hikes to counterbalance rising debt service costs. A sharp slowdown in spending and buoyant revenues since the 
pandemic resulted in a tentative primary surplus. But the economic “recovery” from Covid has consisted of a 
nominal boom that boosted key fiscal metrics, coinciding with a significant weakening of real growth prospects 
which have undermined fiscal sustainability.  
 
Efforts to stabilize the increase in debt take place in a context of policy disarray at the centre of government. This 
has resulted in a growing divergence between the budget plans and their execution. Annual spending far exceeds 
the targets set out in the February budget, and the spending ceiling is no longer an effective limit on cash allocations 
over the medium term. Treasury has begun to see the MTEF as a bargaining position rather than an accurate costing 
of government’s programme (which, if it were to be costed, is wholly inconsistent with the path of fiscal policy). 
Spending targets are set at unreasonably low levels in the annual budget. Although these targets are impossible to 
achieve (and are indeed exceeded each year) the overall path of spending growth has been substantially reduced. 
The budget’s expectation of a salary freeze was never going to be realised, but pay increases for government 
employees have been held below inflation. Attempts to roll back pandemic-related expenditures – such as the Covid-
19 Social Relief of Distress grant or the Presidential Employment Stimulus – are supported by an approach that 
denies that funding is available for these interventions in future budgets.  
 
This response to policy incoherence has eroded the credibility, performance, and quality of the medium-term 
fiscal framework. Budget credibility is a core element of the ability to achieve social change by mobilising the state 
apparatus behind government’s programme. Over time, a weakening of budget credibility can impact on a broader 
set of fiscal institutions, including centralised fiscal authority, effective public administration, and clear oversight 
and control of budgets. These institutions are critical underpinnings of South Africa’s constitutional democracy. If 
they are weakened, the ability of the state to redistribute resources and access long-term capital at low interest 
rates will be undermined. 
 
Three years after the introduction of the Covid-SRD grant, government has not clarified the policy framework 
governing the grant, or decided whether it will be the basis of a more extensive system of income support. In the 
interim, reliance on ad-hoc regulation and annual budget extensions creates uncertainty and confusion. Access to 
the grant has been tenuous, while its value is eroded by inflation. Funding is made available for teacher assistants in 



 

Page | 2  
 

schools, even while funding for teachers and textbooks is cut, without any policy statement about what this implies 
for the future of the public education system. The budget implies a significant contraction in the public healthcare 
system, with hospitals facing the largest cuts. NHI funding is focussed on the construction of new hospitals even 
while spending on healthcare workers, medical equipment, and operational budgets are cut. Universities and colleges 
are having core funding and infrastructure finance cut, yet the Minister of Higher Education is committed to a 
programme of “massification”, which includes the construction of two new universities. A more prominent role in 
regional security is sought, while the defence force is effectively defunded. The President announces a large increase 
in police numbers even while his government’s budget implies a substantial reduction in headcounts.  
 
The budget does contain some important spending choices, and makes valiant attempts to signal changed policy 
direction in certain areas. Funding for water infrastructure and roads is substantial and grows rapidly, as do the 
resources available for new schools and hospitals. Funding for Eskom, Transnet and Sanral should help unlock 
further rounds of infrastructure investment. The tentative extension of the social grant system to working-age 
citizens is maintained, and the budget creates a back-pocket to fund it in future. Significant additional funding is 
allocated to early learning programmes, and the school feeding programme is protected from cuts. The National 
Prosecuting Authority’s budget is spared from the deep cuts imposed on the rest of the criminal justice system. The 
budget for science and innovation also grows faster than inflation.  
 
But, now that the macro-fiscal context has worsened, further cuts to spending appear to be on the cards. South 
Africa faces large and sustained macroeconomic headwinds from the combination of fiscal, financial and monetary 
tightening. Buoyant nominal conditions and surprise inflation supported the goal of fiscal consolidation over the 
last four years. Consolidation in the years ahead will have to operate in conditions of disinflation, as the central bank 
battles supply shocks and balances capital flows, while moving towards a lower de facto inflation target. Lower 
inflation will make the task of implementing spending reductions much harder. Over the longer term, it is not clear 
how much further and how much longer spending growth can be curtailed without aggravating South Africa’s deep 
political fissures, leading to social unrest.  
 
For now, South Africa faces a cul-de-sac of permanent austerity without consolidation. This position is steadily 
eroding the capabilities necessary for the state to realise the potential that fiscal action promises. The chronic fiscal 
squeeze, coupled with failing public services and growing debt, casts a gloomy shadow over the country's 
development prospects.  
 
There is no easy road to getting beyond the fiscal crisis. In respect of the expenditure consolidation, a clearer plan 
backed by explicit policy choices is required. The very wide gap between government’s fiscal programme and its 
policy agenda needs to be addressed. This means grappling with questions of programme design in the public sector, 
and implementing institutional reforms that ease rather than exacerbate fiscal constraints. Any reform programme 
requires an effective executive authority working through a capable policy machinery. Therefore, serious public 
service reforms are urgently needed. In a context where clearer choices are made in respect of spending and 
programme design, the question of a more appropriate balance between taxes and spending in the consolidation 
process also needs reconsideration. 
 
To achieve fiscal consolidation, however, South Africa must see a sustained improvement in economic growth and 
an easing of financial conditions. While GDP per capita continues to decline there is little hope of slowing the 
retrogression in access to social and economic rights which is currently taking place. Renewed growth requires action 
to fix the electricity supply constraint, restore critical export infrastructure, and most importantly, articulate and 
implement a clear and consistent programme of structural reform and transformation to create the conditions for a 
growing economy in a common society.  
 
 
 
 
  



 

Page | 3  
 

PART 1: MACRO-FISCAL CONTEXT – A WORSENING FISCAL IMPASSE 

 
Fiscal policy can play a crucial role in fostering growth and development by ensuring the sustainable delivery of 
essential public services. Managed effectively, government’s budget can help us build a common society and 
overcome the segregation and inequality of the past, without which sustained economic growth seems unlikely.  It 
can also respond to temporary macroeconomic shocks to aggregate demand and livelihoods. South Africa however 
finds itself in a worsening fiscal impasse (see Figure 1). There is continuous austerity, dating back more than a 
decade, but South Africa appears no closer to achieving fiscal consolidation. This position is steadily eroding the 
fiscal capabilities necessary for the state to realise the potential that fiscal action offers. Instead, the budget is 
increasingly becoming a drag on development.  
 
GDP per capita has stagnated and fallen for at least a decade. Slowing growth has been accompanied by rising 
interest rates on government debt. In these conditions, rising debt service costs must be offset by higher taxes or 
reduced non-interest spending. Austerity is weakening the capabilities of the state and society that are necessary for 
sustained growth and inclusive development. It has eroded the public services on which the poor depend yet has 
failed to stabilise debt. Instead, doubts about the solvency of the sovereign have continued to rise and, in this 
context, the fiscal position itself has become a critical constraint on faster growth.  
 
 

Figure 1: Faltering growth and deepening fiscal imbalances 

(a) GDP per capita 

Constant R’000 

 

(b) Nominal GDP growth and the 10-year bond yield 

 

 

(c) Main budget deficit and primary balance 

Percent of GDP 

 

(d) Core spending vs debt costs 

Real per capita growth 

 
 

Projections reflect the release of preliminary figures for main budget revenue and expenditure, as well as GDP for the 2022 budget year 

“Core spending” is main budget non-interest expenditure excluding captial trasnfers to state-owned companies (i.e. “payments for financial assets”), skills development levy and fuel levy 
sharing with metros. 

Source: National Treasury budget data; Bond Exchange South Africa, Stats SA, Public Economy Project calculations 

60

65

70

75

80

1
9

9
4

1
9

9
6

1
9

9
8

2
0

0
0

2
0

0
2

2
0

0
4

2
0

0
6

2
0

0
8

2
0

1
0

2
0

1
2

2
0

1
4

2
0

1
6

2
0

1
8

2
0

2
0

2
0

2
2

2
0

2
4

2
0

2
6

2
0

2
8

C
o

n
st

a
n

t 
2
0
1
5
 R

'0
0
0

IMF 
Projection 
(April 2023)

-2%

0%

2%

4%

6%

8%

10%

12%

14%

16%

18%

2
0

0
0

2
0

0
2

2
0

0
4

2
0

0
6

2
0

0
8

2
0

1
0

2
0

1
2

2
0

1
4

2
0

1
6

2
0

1
8

2
0

2
0

2
0

2
2

2
0

2
4

  Nominal GDP growth

  10-year bond yield

Projection

-10%

-8%

-6%

-4%

-2%

0%

2%

4%

6%

1
9

9
6

1
9

9
8

2
0

0
0

2
0

0
2

2
0

0
4

2
0

0
6

2
0

0
8

2
0

1
0

2
0

1
2

2
0

1
4

2
0

1
6

2
0

1
8

2
0

2
0

2
0

2
2

2
0

2
4

P
e
rc

e
n

t 
o

f 
G

D
P

budget year

Main budget balance

Primary balance

PEP 

projection

-8%

-6%

-4%

-2%

0%

2%

4%

6%

8%

10%

2
0
0
0

2
0
0
2

2
0
0
4

2
0
0
6

2
0
0
8

2
0
1
0

2
0
1
2

2
0
1
4

2
0
1
6

2
0
1
8

2
0
2
0

2
0
2
2

A
n

n
u

a
l 

%
 c

h
a
n

g
e

Budget year

Core spending

Debt costs (smoothed trend)



 

Page | 4  
 

 

Consolidating spending  
 
Attempts at fiscal consolidation have focussed largely on 
containing spending. A nominal (i.e., cash) expenditure ceiling was 
introduced to limit spending growth.1 The aim was to discipline 
the annual budget planning cycle, which is largely concerned with 
cash allocations over the three-year medium term expenditure 
framework (MTEF)2. 
 
Spending grew on average at around 14% per annum between 
2000 and 2010, far outpacing GDP growth and leading to a 
doubling of real spending per citizen (see Figure 3a on the next 
page). Since the introduction of the spending ceiling in 2012, 
however, there has been a significant slowdown in the nominal 
growth of spending growth.  
 
The fiscal strategy since the Covid-19 pandemic sought to achieve 
a sharp nominal contraction of expenditure allocations in a single 
year. This aimed to reverse spending introduced as temporary 
support during the pandemic and restoring the sustainability of 
the public finances in one-fell swoop. A sharp contraction 
frontloaded in the first year of the MTEF would, it was believed, 
restore sustainability, and allow the consolidation to come to an 
end. 
 
This approach has delivered contradictory outcomes. On the one 
hand, government has failed to remain within the spending limits 
set in annual budgets. This has opened a wide gap between budget 
plans and spending outcomes,  which has been problematic in 
respect of compensation of employees – the most critical target in 
the consolidation plan. On the other hand, nominal spending 
growth has been reduced to its lowest rate ever, and compensation 
growth brought far below inflation for the first time since 1994.  
 
Figure 2a shows that the nominal contraction has not been 
delivered as planned. Successive budgets postponed the 
consolidation or revised-up expenditure estimates by billions of 
rand relative to the previous iteration, or both. The spending 
ceiling is no longer an effective limit on cash allocations over the 
medium term (Figure 2b). Even the annual budget has become an 
unreliable guide to spending in the current year, as there are large 
deviations between spending estimates tabled each February, and 
the outcomes for the year (Figure 2c). A large gap between budget 
plans and budget execution has opened. We discuss the reasons 
for this outcome, and the implications for budgets going forward, 
in greater detail in Part II.  
 
 

                                                      
1 The spending ceiling is a nominal total set for each of the three forward planning years of the MTEF. It is defined as main budget non-interest expenditure 
excluding payments directly financed by dedicated revenue sources and others not subject to policy oversight. These exclusions include payments for financial 
assets financed by asset sales in the same financial year, payment transactions linked to the management of debt and direct charges related to specific payments 
made in terms of legislation that provides for the collection and transfer of such receipts outside of the main budget, including the skills development levy 
contributions. 

2 Through the budget process departments and agencies negotiate with Treasury on cash limits to their spending commitments. Once agreed and tabled by the 
Minister of Finance, Parliament votes to enshrine the rand value of allocations into law for the current year and endorses nominal spending limits over the 
next three years in a medium-term expenditure framework (MTEF). 

Figure 2: Budget plans and budget outcomes 

a) Budget projections of core spending growth 

 

b) Spending ceiling in successive budgets 

 

c) Annual spending deviation from budget 

 

Source: National Treasury budget data; Public Economy Project calculations 

-2%

0%

2%

4%

6%

8%

10%

2
0

1
6

2
0

1
7

2
0

1
8

2
0

1
9

2
0

2
0

2
0

2
1

2
0

2
2

2
0

2
3

2
0

2
4

2
0

2
5

P
e
rc

e
n

t 
a
n

n
u

a
l 
g

ro
w

th

budget year

Budget Review 2021

Budget Review 2022

Budget Review 2023

Audited outcome

1 300

1 400

1 500

1 600

1 700

1 800

1 900

2019 2020 2021 2022 2023 2024 2025

R
 m

il
li

o
n

Budget year

Actual

2021 Budget Review

2022 Budget Review

2023 Budget Review

-12.5

-6.2

-10.9

-4.4

7.9
10.9

29.7

60.1

22.6

-20

-10

0

10

20

30

40

50

60

2014 2015 2016 2017 2018 2019 2020 2021 2022

R
 b

il
li

o
n

Budget year



 

Page | 5  
 

But while medium term and annual targets have been exceeded, 
the post-pandemic period has witnessed a further and 
substantial slowdown in nominal spending growth. Figure 3a 
shows that since the pandemic the growth in core spending has 
been brought below the expansion of nominal GDP for the first 
time in decades. So, while government has been largely unable 
to meet its own budget targets, the strategy of setting stretch-
targets for reduced cash allocations has created a context in 
which spending growth has been substantially curtailed, even 
below its post 2012 trajectory. 
 
Added to this, unanticipated inflation has helped achieve 
consolidation by raising the nominal growth of GDP, increasing 
the tax take and cutting back on the real growth of spending. 
This buoyancy partly reflected the commodity price boom that 
followed the pandemic. Global inflationary conditions 
combined with domestic supply limits (arising largely from 
loadshedding) created a surprise inflationary shock.  
 
Cash spending targets are always set with reference to treasury 
CPI forecasts in mind. These have been significantly below 
realised inflation for the last three years (see Figure 3b). With 
the help of unanticipated inflation, execution risks have been 
mitigated, and expenditure consolidation targets have been 
partially achieved.  Surprise inflation has been a key factor 
behind revenue buoyancy and acted to ease the financing 
conditions faced by government, lowering the real return on 
nominal bonds, and slowing the rise of debt service costs.  
 
In combination, these factors have “closed the jaws of the 
hippopotamus” (See Figure 3c). Spending growth has been 
brought below the rate of nominal economic expansion, 
lowering the ratio of spending to GDP. Nominal GDP 
expansion has buoyed revenues. If we exclude capital transfers 
to state-owned companies, a small primary surplus was achieved 
in 2022.  
 
But while the price shock has helped government achieve its 
fiscal consolidation targets in the short run, nominal buoyancy 
has gone together with continued real stagnation of the 
economy. This suggests a worsening fiscal position in the long 
run, a point to which we return below.  
 

Containing government’s wage bill 
 
The government wage bill is one area where the contrast 
between the failure to meet nominal targets on the one hand, 
and the achievement of a large slowdown in spending growth 
on the other hand, is most notable.  
 
Compensation ceilings are presented as fixed nominal limits in each budget. However, subsequent budgets have 
added large amounts to the salary bill, and also to the equitable share of provinces - a significant share of which is 
allocated to salaries (see Figure 3a). These additions were prompted by the realisation that deep cuts to cash 
allocations would not be possible to achieve without substantial reductions in service delivery and employee 
headcounts, particularly in labour intensive services such as basic education, healthcare, social development, and 
criminal justice.  
 
 

Figure 3: Spending growth and inflation 

a) Growth of nominal GDP and core spending* 

 

b) CPI inflation estimates in successive budgets 

 

c) Core spending* and revenue as % of GDP 

 

* “Core spending” is main budget non-interest expenditure excluding capital transfers 
to state-owned companies (i.e. “payments for financial assets”), skills development levy 
and fuel levy sharing with metros. 

Source: National Treasury budget data; Public Economy Project calculations 

-2%

0%

2%

4%

6%

8%

10%

12%

14%

16%

18%

2
0

0
0

2
0

0
1

2
0

0
2

2
0

0
3

2
0

0
4

2
0

0
5

2
0

0
6

2
0

0
7

2
0

0
8

2
0

0
9

2
0

1
0

2
0

1
1

2
0

1
2

2
0

1
3

2
0

1
4

2
0

1
5

2
0

1
6

2
0

1
7

2
0

1
8

2
0

1
9

2
0

2
0

2
0

2
1

2
0

2
2

2
0

2
3

2
0

2
4

2
0

2
5

P
e
rc

e
n

t 
a
n

n
u

a
l 
ch

a
n

g
e

Nominal GDP

Core spending
Budget 2023 

projection

2%

3%

4%

5%

6%

7%

2
0

1
6

2
0

1
7

2
0

1
8

2
0

1
9

2
0

2
0

2
0

2
1

2
0

2
2

2
0

2
3

2
0

2
4

2
0

2
5

P
e
rc

e
n

t 
a
n

n
u

a
l 
ch

a
n

g
e

budget year

Budget Review 2021

Budget Review 2022

Budget Review 2023

Actual

20%

21%

22%

23%

24%

25%

26%

2010 2012 2014 2016 2018 2020 2022

P
e
rc

e
n

t 
o

f 
G

D
P

Budget year

Core spending

Revenue



 

Page | 6  
 

Despite large nominal deviations from budget targets, 
spending growth on compensation of employees has been 
strongly curtailed, and indeed kept below the rate of inflation. 
This has led to an erosion of the real incomes of government 
employees. Salaries were frozen in nominal terms in 2020. In 
2021 government conceded a cash gratuity, costed as an 
equivalent of a 4.2% increase3, far below the rate of inflation 
of that year. In 2022, government unilaterally implemented a 
3% increase, and this year’s wage agreement implies another 
3.3%. Since the Covid-19 pandemic government employees 
have (on average) experienced a 9% fall in their basic pay 
relative to consumer price inflation.  Figure 3b shows that 
cost-of-living adjustments for government employees are 
below inflation and far below median pay settlements across 
the economy. Looking to the spending figures in the budget, 
compensation contracted in real terms in 2020 and 2022 (see 
Figure 3c). There was no wage settlement concluded before 
Budget 2023, so the Budget assumed there would be no pay 
increase, and pencilled-in an effective 4% contraction in the 
salary bill. However, the 2023 wage agreement resulted in a 
shortfall of R21 billion in 2023 and R77 billion over the 
MTEF4. 
 
Following the budget the Minister of Finance declared that the 
settlement is a substantial blow to the fiscus and warned that 
government will be “executing difficult trade-offs [that] may 
entail a rationalisation of staffing levels and deployment of 
headcount management strategies as a means to curb the wage 
bill … a higher than-budget wage agreement means less space 
for the recruitment of staff”. The National Treasury guidelines 
for the 2024 MTEF indicate that the carry-through effects of 
the 2023 wage agreement are to be funded within the existing 
baselines (National Treasury 2023b). 
 
In real terms then, the incomes of government employees 
have been substantially reduced. Salaries have not been cut as 
aggressively as treasury had planned, and wage bill spending 
continues to exceed targets. But cost-of-living adjustments 
have been held far below inflation over the last four years, 
with an adjustment aligned with inflation agreed for next year. 
Moreover, wage improvements for government employees 
have strongly diverged from private sector settlements.  
 
This has been a key element in reducing the overall growth of 
expenditure to historically low levels. Unfortunately, despite 
having achieved a historic cut in government pay, and an 
unprecedented slowdown in spending growth, the goal of 
fiscal consolidation now seems more distant than ever. This is 
because macroeconomic conditions unrelated to the budget 
have substantially worsened.  

                                                      
3 See DPSA circular 15 of 2021. The cash gratuity was extended without increase in 2022 and was integrated into the pensionable salary in 2023. National 
Treasury continues to report this reallocation from “cash gratuity” to “pensionable salary” as amounting to another salary increase in 2023. This is highly 
misleading and has created considerable confusion in public deliberation about this issue, which is at the centre of the budget programme. The reallocation 
between budget definitions in 2023 may have resulted in additional costs, for instance because it implies additional resources to the employer’s pension 
contribution. From the point of view of government employees, however, the reallocation from one accounting treatment to another adds nothing to their 
take-home pay.  

4 Treasury has continued to report the settlement as an “above inflation” shock amounting to 7.5% and reflected the materialisation of a significant fiscal risks. 
This has the effect of creating the impression that public servants pay has escalated much faster than is in fact the case. It has also confused the public debate 
about the fiscal issues and served to focus renewed attention on the salary bill.  

Figure 4: Government wage bill trends 

a) Wage bill budget estimates and outcomes 

 

b) Govt employees’ cost-of-living adjustments. 

 

* Adjustment was on a sliding scale, we take the average adjustment 

** There was no cost-of-living adjustment in 2021, but a non-pensionable cash gratuity was awarded 

equivalent to a 4.2% increase which was then extended into 2022 and integrated with base pay in 2023. 

c) Real growth of compensation spending  

 

Source: National Treasury budget data; Labour Research Service (median settlement 
derived from the percentage increases recorded in collective agreements), DPSA circulars 
(annual cost-of-living adjustments), StatsSA, Public Economy Project calculations 

550

570

590

610

630

650

670

690

710

2019 2020 2021 2022 2023 2024 2025

R
 b

il
li
o

n
s

Budget year

Actual

Budget 2021

Budget 2022

Budget 2023

Settlement (PEP calculation)

0%

1%

2%

3%

4%

5%

6%

7%

8%

9%

2
0

1
5

2
0

1
6

2
0

1
7

2
0

1
8

*

2
0

1
9

*

2
0

2
0

2
0

2
1

**

2
0

2
2

2
0

2
3

2
0

2
4

A
n

n
u

a
l 

%
 c

h
a
n

g
e

Budget year

Govt cost-of-living adjustment

Median wage settlement

CPI

-4%

-2%

0%

2%

4%

6%

8%

2
0

1
0

2
0

1
1

2
0

1
2

2
0

1
3

2
0

1
4

2
0

1
5

2
0

1
6

2
0

1
7

2
0

1
8

2
0

1
9

2
0

2
0

2
0

2
1

2
0

2
2

2
0

2
3

2
0

2
4

2
0

2
5

A
n

n
u

a
l 
%

 g
ro

w
th

Budget year

Outcomes

Budget 2023 assumption

Implied by April 2023 Setttlement

Projections



 

Page | 7  
 

A changed macro-fiscal outlook  
 
A sharp slowdown in spending growth since the pandemic, combined with buoyant revenue on the back of a 
nominal boom, has supported the achievement of a tentative primary surplus (excluding capital transfers). 
Unfortunately, while inflation has been surprisingly high, real economic growth has been even lower than the 
miserable estimates anticipated in the budget. A nominal boom has coincided with a significant weakening of growth 
prospects, undermining longer term fiscal sustainability. In recent months financial conditions have tightened, 
pointing to a worsening of the sovereign risk premium and an increase in the financing costs faced by government. 
After several years of painful efforts to stabilise the public finances, therefore, this slowing growth and tightening 
financial conditions have rendered the goal of debt stabilisation even more distant.  
 
Since the presentation of the 2022 Medium Term Budget Policy Statement (MTBPS) and 2023 Budget, the medium-
term outlook for real growth in the country has deteriorated.  Expectations of the intensity and duration of 
loadshedding have worsened. The extent of infrastructure constraints to increased commodity-export volumes has 
become apparent. There is now a significant discrepancy between the outlook published in the budget and updated 
growth forecasts for the current year, which reflects the uncertain economic impact of load shedding. The National 
Treasury's GDP growth forecasts for the medium term (2023: 0.9%) are considerably higher than those of other 
forecasters and are expected to be revised downwards in October.5 
 

Table 1: Revised GDP growth outlook 

 
 
Not only has growth slowed, but financing conditions have worsened since the budget was presented in February. 
While the alarmed market sentiment that took the rand close to R20 to the US dollar in the wake of the “Lady R” 
imbroglio appear to have eased, it has done so only with interest rates that are significantly above previous 
expectations. The tightening of financial conditions probably reflects underlying fundamental factors. The most 
significant of these is the slow-down in growth itself. In addition to generating pessimism about longer term 
sustainability and future returns on fixed investment, the combination of electricity constraints and the deterioration 
of South Africa's export logistics platform has raised concerns about the current account deficit, adding to the 
pressure on external financing. This comes as commodity prices have slowed and the recent terms of trade gains 
look to be reversed, raising the cost of external financing for commodity producers, including government bond 
yields.6 
 
Added to this, South Africa’s “greylisting” by the Financial Action Task Force (FATF) shortly after the budget 
created new concerns about the financial and economic implications of weakened state capacity and the growing 
threat posed by criminal networks. It has since become increasingly apparent that meeting the requirements to be 
removed from this greylist may be challenging, and the financial greylisting is likely to remain in place for an 
extended period7. Added to all this, doubts about South Africa's non-alignment in global affairs became elevated in 

                                                      
5 The National Treasury provide two public updates to the forecasts informing the fiscal framework – in February and October.  

6 See, for instance, Aslam et al. 2016; Drechsel and Tenreyro 2018; Shousha 2016 for consideration of the negative correlation between commodity prices and 
external financing conditions in emerging markets.  

7 Several of the factors that led to greylisting concern the performance of the criminal justice system. Strategic deficiencies identified by the Financial Action 
Task Force (FATF) include increasing investigations and prosecutions of serious and complex financial crimes, improving identification, seizure and 
confiscation of proceeds from a wider range of crimes and increasing outbound Mutual Legal Assistance requests. Other deficiencies might be address through 
reforms to legislation or financial governance and effectively implementing targeted financial sanctions while improving identification of individuals and entities 
for domestic designation. These include enhancing risk-based supervision of Non-Financial Businesses and Professions; ensuring access to accurate Beneficial 
Ownership information; boosting requests for financial intelligence for money laundering/terrorism financing investigations; updating its counter-terrorism 
financing strategy based on risk assessments. 

Real GDP grow th outlook - calender year* 2023 2024 2025

National Treasury - MTBPS 2022 1.4% 1.7% 1.8%

National Treasury - Budget 2023 0.9% 1.5% 1.8%

South African Reserve Bank - November 2022 1.1% 1.4% 1.5%

South African Reserve Bank - May 2023 0.3% 1.0% 1.1%

IMF - World Economic Outlook - January 2023 1.2% 1.3% -

IMF - World Economic Outlook - April 2023 0.1% 1.8% 1.6%

Bureau for Economic Research - 4Q 2022 0.5% 1.4% 1.8%

Bureau for Economic Research - 1Q 2023 0.2% 1.4% 2.1%

*Growth projections correspond to publication date and not forecast date

Data: National Treasury, South African Reserve Bank, International Monetary Fund, Bureau for Economic 

Research



 

Page | 8  
 

a context of growing international concerns about the implications of geo-economic fragmentation.8 The previously 
unthinkable possibility of financial or trade measures being taken against the country by its most important trading 
partners became suddenly salient. Against this backdrop, the South African Reserve Bank added a new element to 
its risk matrix: that of ‘capital outflows and declining market depth. and liquidity’. Noting that the proportion of 
government bonds held by local institutions had increased from 58% in 2018 to 75% in 2023, the SARB reported 
in its Financial Stability Review that: 

 
“This is a significant structural shift, especially considering the significant increase in government bonds issued during 
this period. It raises concerns about the capacity of South African investors to continue absorbing new issuances of 
SAGBs in future. As local participants increasingly step in to absorb the declining appetite for new SAGB issuances 
by non-residents, this raises financial stability concerns regarding market liquidity, increased volatility and higher 
domestic government bond yields. South Africa’s non-investment grade status and recent greylisting by the FATF 
could contribute to non-resident sell-offs from local markets, with non-resident investor behaviour becoming 
increasingly speculative in relation to South African assets.” (SARB 2023:6) 

 
We expect these underlying factors to persist over the medium term. Rather than constituting a temporary shocks, 
therefore, global capital markets have effectively downgraded South Africa by an additional notch. This downgrade 
will increase the government's financing costs until the several conditions that led to the downgrade are addressed.  
 
Given risk of higher bond yields and slow economic growth, achieving the primary balance required to stabilize the 
debt appears to be an even more challenging task. It would necessitate significant spending cuts or tax increases, 
over and above the consolidation achieved so far. A slower growth path over the current year and the medium term, 
with the spending and revenue proposals from the 2023 budget intact, will result in a primary balance that does not 
improve as suggested by the 2023 Budget. Instead, the primary balance could worsen over the medium term, 
primarily due to lower revenue collection, rendering the consolidation planned in the 2023 budget inadequate. If 
government is to achieve its objective of stabilising debt as a share of GDP, a continuation of the policy of fiscal 
consolidation would be necessary beyond the current MTEF. 
 
The February 2023 Budget presents a picture of frontloaded fiscal consolidation by reducing government’s 
consumption spending over the next three years, with a large cut in spending in the first year. Treasury projects 
government consumption to contract by 2.2% in 2023, its largest contraction in 25 years. The budget also presented 
an improved fiscal outlook for the medium-term expenditure framework (MTEF)9 compared with the 2022 Budget 
and MTBPS.  
 
  

Figure 3: Tightening financial conditions 

(a) Yield curve  

 

(b) Average clearing yield of fixed-coupon bonds 

 
Source: National Treasury; Bond Exchange South Africa 

                                                      
8 See IMF 2023b for a discussion of the later 

9 The three-year spending plans of national and provincial governments, published at the time of the Budget. 

6%

7%

8%

9%

10%

11%

12%

13%

< 1 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30

y
ie

ld
 

January 2023 June 2023

years to maturity

8.5

9.0

9.5

10.0

10.5

11.0

11.5

Ja
n

-2
1

M
a
r-2

1

M
a
y
-2

1

Ju
l-2

1

S
e
p

-2
1

N
o

v
-2

1

Ja
n

-2
2

M
a
r-2

2

M
a
y
-2

2

Ju
l-2

2

S
e
p

-2
2

N
o

v
-2

2

Ja
n

-2
3

M
a
r-2

3

M
a
y
-2

3

c
le

a
ri

n
g

 y
ie

ld
 (

%
)

  average clearing yield



 

Page | 9  
 

This was partly due to strong monthly revenue collections 
recorded in 2022, prompting upward revisions to revenue 
estimates over the MTEF. The Budget projects a main 
budget primary surplus for 2023/24 (0.9% of GDP), 
improving over the MTEF to 1.7% of GDP in 2025/26, 
suggesting that government’s fiscal objective of stabilising 
debt as a share of GDP will be realised, and consolidation 
will indeed “soon come to a close”. 
 
But as already mentioned, updated projections indicate 
lower economic growth in the years ahead, leading to 
further stagnation of per capita income. With slowing 
growth comes a need to revise the revenue outlook. Also, 
considering that the fiscal balances outlined in the budget 
do not present the complete picture of the government's 
expenditure path, PEP presents a fiscal trajectory which 
assumes that:  
 

 Compensation budgets will grow by the March 2023 
wage settlement agreement, for 2023 and 2024, and by 
CPI inflation in 2025. 

  There will be an extension of the Covid-SRD grant 

beyond 2023 across the MTEF, and that the value of the 

grant will grow by CPI. 

 The unallocated reserve is allocated in full to fund the 

Covid SRD grant across the MTEF. However, this is 

insufficient, resulting in a shortfall of – R4.1 billion in 

2024; R318 million in 2023 – which is funded through 

additional borrowing. 

  Public employment programmes are extended and 

implemented at the minimum cost estimate from the 

Presidency presentation “Putting South Africa to Work”. 

Our numbers take also into account the latest releases of 

macroeconomic and public finance outcomes and assume 

nominal GDP growth over the medium-term in-line with 

updated consensus forecasts, with the attendant effects 

on tax-revenue. Our revenue estimates assume Budget 

2023 GDP inflation and revenue buoyancy. This 

assessment also assumes Budget 2023 estimates of 

government’s interest rates – an assumption over which 

there is considerable uncertainty. We also treat financial 

support to Eskom announced in the budget as an ‘above-

the-line’ capital transfer, and therefore are part of the 

non-interest expenditure and the primary deficit. 

 
Although budget 2023 projected successively larger 
primary surpluses over the MTEF, by applying these 
more realistic assumptions PEP estimates primary deficits 
for every year of the MTEF, and a worsening main budget 
balance over the medium term. The debt ratio does not 
appear to stabilise over the medium term. Gross loan debt 

as a share of GDP increases at a faster rate over the medium term when compared to Budget Review 2023 
projections, without stabilisation or even a moderation over the medium term. 

Figure 5: Fiscal outlook: Budget 2023 and PEP baseline 

a) Primary Balance 

 

 

Source: National Treasury budget data; Public Economy Project calculations Both the PEP 
and BR 2023 projections reflect the release of preliminary figures for main budget revenue and 
expenditure, as well as GDP for the 2022 budget year 

b) Main budget balance as a share of GDP (%) 

 

Source: National Treasury budget data; Public Economy Project calculations 

c) Gross loan debt as a share of GDP

 

Source: National Treasury budget data; Public Economy Project calculations 

0.6%

0.9%

1.4%

0.0%

-0.8%

-0.6%
-0.6%

-1.0%

-0.5%

0.0%

0.5%

1.0%

1.5%

2.0%

2
0

2
2

2
0

2
3

2
0

2
4

2
0

2
5

sh
a
re

 o
f 

G
D

P

  Budget Review 2023*

  PEP Scenario

Budget year
*updated to reflect recent data releases

-4.6%

-3.9%

-3.6%

-3.3%

-5.7%
-5.5%

-5.7%

-6.0%

-5.5%

-5.0%

-4.5%

-4.0%

-3.5%

-3.0%

2
0

2
2

2
0

2
3

2
0

2
4

2
0

2
5

sh
a
re

 o
r 

G
D

P

  Budget Review 2023*

  PEP Scenario

Budget year *updated to reflect recent data releases

72.2%
72.8%

73.6%73.8%

76.0%

79.4%

68%

70%

72%

74%

76%

78%

80%

2
0

2
3

2
0

2
4

2
0

2
5

sh
a
re

 o
f 

G
D

P

  Budget Review 2023*

  PEP Scenario

Budget year *updated to reflect recent data releases



 

Page | 10  
 

PEPs baseline fiscal trajectory implies that despite the intention of government to frontload the previously postponed 
fiscal consolidation in the current fiscal year, the large real cuts to expenditure fall short of accomplishing the 
government's medium-term fiscal objective of stabilizing debt as a share of GDP. Indeed, achieving this goal would 
necessitate even larger and more harmful spending reductions, which would likely prove socially and politically 
implausible to enact. This underscores the importance of fostering economic growth to complement expenditure 
measures which are directed at fiscal consolidation. 
  



 

Page | 11  
 

PART II: MEDIUM TERM ALLOCATION AND POLICY CHOICES 
 

Introduction 
 
Every year, the budget gives concrete numerical form to policy goals It shows citizens the choices government is 
making about investments and the provision of services and investments. It enables government departments to 
plan to use these resources effectively. These spending choices, and how they are financed, have direct consequences 
for the welfare of citizens and the distribution of income. The quantitative nature of budget choices means trade-
offs between policy goals are made clear and explicit. Despite our concern about the credibility of spending estimates 
(see Part I), in this section we take these estimates largely at face value and ask what they can tell us about policy 
intentions. How have allocative choices been reflected within the constraints imposed by the budget process.  
 
Figure 6 shows planned spending over the medium-term expenditure framework (MTEF). Government will pay 
R333 billion each year on average to services its debts in the next three years, the consequence of previous deficits. 
The largest policy function is basic education, which will spend about R320 billion each year. Social protection, 
community development and healthcare are the next largest priorities showing that (at this broadest level of resource 
allocation) the budget is aligned to constitutional principles with the lion’s share going to mandated social spending. 
Compared to last year, community development has overtaken healthcare as the third largest allocation10 This 
reflects government’s intention to contain consumption spending and accelerate infrastructure spending. It also 
reflets the strong growth of funding for the municipal sphere of government, which continues to face operational 
and fiscal crisis.  
 

Figure 6: Annual spending by policy function 

Average over MTEF (2023−2025) | consolidated budget 

 
Source: National Treasury budget data; Public Economy Project calculations 

Note: Social protection includes “unallocated reserve” on the assumption it will fund extensions of Covid-SRD grant. “Policy functions” differs from National Treasury’s 
“budget groups”. Here, home affairs is included in the General Public Services function (security cluster in Treasury documents); Skills levy contributions are separated from 
universities and TVET colleges; Capital trasnfers to state owned companies reflects amounts defined by Treasury as “payments for financial assets” as well as the Eskom debt 
relief package.  

 
 

                                                      
10 The numbers here assume that the “unallocated reserve” is used to fund the extension of the Covid-SRD grant.  

366.8

319.1
305.2

277.0 269.8
258.9

172.6

126.0

87.2 86.0 85.8

53.5 19.9
11.9

0

50

100

150

200

250

300

350

400

R
 b

illio
n

Constitutionally-mandated social spending

Economic development, security and governance

Other



 

Page | 12  
 

 
Budget allocations are enormous, and a large share of 
spending is attached to long-standing obligations that cannot 
be easily changed in a single year. This means that shifting 
spending allocations towards new policy priorities requires 
consistent choices over several budget cycles. Expenditure 
trends over the medium-term give a sense of how the budget 
is changing over time, reflecting the underlying policy choices. 
which are difficult to discern from the static allocations shown 
in Figure 6 .  
 
Figure 7a shows average annual growth of spending by 
policy function between 2019 (the year before the 
coronavirus pandemic) to 2024 (the middle year of the 
MTEF). It omits debt-service costs, already the largest share 
of spending, and growing each year by more than 12% on 
average. In terms of policy functions, growth rates reveal the 
following: 

 Community development, economic development and 
higher education absorb a rising share of spending, growing 
faster than the average and faster than inflation. 

 Social protection is also growing faster than inflation. 
This reflects the extension of the social protection system that 
has resulted from the introduction of the Covid-SRD grant 
(and we are assuming the grant will continue to be funded).  

 Healthcare and basic education will see significant 
declines in real resource allocation and will be reduced as a 
share of the budget.  

 The criminal justice system, defence and general public 
services will see very large declines in real resource allocation.  

In the sections that follow we dig deeper into the 
consequences of these choices in respect of individual policy 
functions.  
 
Choices about the growth of policy functions are bound 
together with choices about the type of inputs which must 
be purchased to deliver government services. Transfers to 
households grow strongly on the back of the extension of the 
Covid-SRD grant and support for students in higher 
education. Capital spending is budgeted to grow robustly, and 
transfers to government agencies include allocations for water 
and road infrastructure, which drive spending in the economic 
development function. Transfers to local government also 
include large infrastructure funds. The budget shows a 
reduction in the share of consumption spending, including 
employee compensation and goods and services. These 
elements of the budget are used intensively to provide basic 
education, healthcare, social services and criminal justice.  
 
Lastly, choices are made about allocating budgets to different 
spheres of government, shown in Figure 7c. Again, this 

intersects with choices about policy functions. Education and healthcare are funded from provincial budgets; 
criminal justice, defence and social protection are national functions, while community development is largely 
delivered by local government. On average, therefore, spending by provincial governments grows at the slowest 
rate.  The provincial equitable share, which mainly funds education and healthcare, grows by only 3.3% per year 

Figure 7: Medium-term spending growth (2019-2024) 

Annual average compound growth 

a) Policy function 

Consolidated government  

 

* Assumes the “unallocated reserve” funds an extension of the Covid-SRD grant 

b) Type of purchase (economic function) 

Consolidated national, provincial and social security funds 

 

* Transfers to households includes thee “unallocated reserve” funds an extension of the 
Covid-SRD grant 
Note: Interest payments and capital transfers to SOCs are not shown 

c) Sphere of government 

Division of revenue 

 

* Excludes interest payments and bailouts to state-owned companies 

Source: National Treasury budget data; Public Economy Project calculations 

5.0%

1.0%

2.0%

2.6%

2.7%

3.8%

3.9%

5.0%

6.4%

6.7%

7.1%

7.2%

0% 1% 2% 3% 4% 5% 6% 7%

CPI

Defence

Arts and sports

Criminal justice

General public services

Healthcare

Basic education

Average

Social protection*

Higher education

Economic development

Community development

5.0%

3.0%

3.7%

4.1%

4.8%

5.7%

6.6%

6.8%

7.0%

0% 1% 2% 3% 4% 5% 6% 7% 8%

CPI

Compensation of employees

Other transfers

Goods and services

Primary spending

Govt agencies

Capital

Local government

Households*

Transfers and subsidiesPurchases

5.0%

4.1%

4.3%

9.6%

7.3%

4.3%

3.1%

3.3%

4.0%

0% 1% 2% 3% 4% 5% 6% 7% 8% 9% 10%

CPI

General fuel levy

sharing with metros

Conditional grants

Equitable share

Local government

Conditional grants

Equitable share

Provinces

National departments*



 

Page | 13  
 

between 2019-2024, while conditional grants grow slightly faster, reflecting the fact that many of these grants 
support capital and investment activities. 
 
The fastest-growing element is the local equitable share, which is an unconditional grant aimed at subsidising free 
basic services for poor households, although used to fund operational spending in many municipalities. In a context 
where many municipalities appear unable to pay for bulk purchases of electricity, and where poor households face 
large economic shocks as a result of loadshedding. Funds were added to the local equitable share to offset rising 
electricity tariffs, and this is welcome support. However, funding for local government has been growing rapidly for 
more than a decade, and yet its fiscal, operational and governance crisis has continued to intensify. The recent move 
to forgive municipal debt to Eskom, attached to ostensibly stringent conditions, may also relieve financial pressure 
on the most insolvent local authorities. However, without a clear agenda to reform the municipal sphere of 
government and its fiscal arrangements, it is likely that these funding pressures arising will deepen requiring 
increased funding out of the national budget. 
 

Education 
 
In basic education, reductions in real spending per learner are expected to intensify in the period ahead. Real 
consolidated basic education spending for each enrolled learner was more than R24 000 in 2020 but will fall to 
R21 635 by 2025. This comes against the backdrop of significant learning losses associated with schooling lockdown 
during the Covid-19 pandemic. It is notable that no clear plan or budget has been tabled to reverse the impact of 
the school shutdown, or catchup on learning losses. A recent review concluded that “There has been no attempt to 
recoup time to remediate learning losses, apart from very recent attempts in one province. The insistence on a 
largely business-as-usual approach to curriculum implementation fails to recognise and address the severe 
educational impact of the pandemic, especially on learners in the poorest communities” (Hoadley, 2023: p1). 
 
Budget estimates for basic education illustrate how lack of clarity about key budget issues make it very difficult to 
gauge government’s policy intentions from the budget numbers (see Figure 8). On the one hand the budget proposes 
to cut spending on workbooks and other learning materials from R6.4 billion last year to R5.9 billion in 2023. Cash 
transfers to schools (which cover a sizable share of their operational costs) are also lower in nominal terms. 
Estimates for school property payments decline too, despite rising utility bills. All these items, however, have grown 
robustly in recent years and – if medium-term projections are to be believed – the cuts in 2023 will be fully reversed 
over the following years. Funding for new school buildings is also set to increase over the medium term. The school 
feeding schemes is protected in the budget, retaining its value both after inflation and taking account of the rise in 
learner numbers. Allocations by the Department of Basic Education to fund teacher training through NSFAS 
bursaries, however, have been falling for some time (in constant price terms) and this trend looks set to continue.  
 
 
 
 

Figure 8: Basic education spending growth 

(a) Annual Budget (2022-2023) 

 
 

(b) Medium-term (2019-2024) 

 

* Excludes provision for teacher assistant programme 

Source: National Treasury budget data; Public Economy Project calculations 

4.9%

-8.9%

-3.2%

-2.4%

-0.6%

0.4%

2.5%

3.5%

7.4%

9.1%

11.0%

-10% -5% 0% 5% 10%

CPI

Workbooks and LTSM

Goods and services

Property payments

Transfers to schools

Teacher training (NSFAS)

Consolidated basic education

Compensation of employees

New buildings

School nutrition grant

School infrastructure grant

5.0%

2.6%

3.4%

3.9%

4.9%

5.7%

6.2%

6.3%

6.4%

7.7%

9.8%

0% 2% 4% 6% 8% 10%

CPI

Teacher training (NSFAS)

Compensation of employees

Consolidated basic education

Goods and services

School infrastructure grant

Workbooks and LTSM

New buildings

School nutrition grant

Transfers to schools*

Property payments



 

Page | 14  
 

These estimates are placed into doubt, however, by the conclusion of the 2023 wage agreement which, by our 
estimate, creates a basic education shortfall of R7.4 in the 2023 and R30 billion over the MTEF. This shortfall might 
be funded by reducing employment in the sector, which would have a huge and negative impact on learning 
outcomes (as is acknowledged very clearly in (National Treasury 2021:59, 2022b:42, 2022a:58). Alternatively, the 
shortfall might be funded by reduced funding for new school buildings, school nutrition, learning materials, or 
transfers to schools over the medium term. A third possibility is that resources will be added to the budget to make 
up for the shortfall. Indeed, past practice suggests that this will be the case: budget 2023 allocated an additional R20 
billion to the provincial equitable share “mainly to cover shortfalls in basic education compensation 
budgets”(National Treasury 2023a:60). This follows similar action in the previous year’s budget, which added 
R24.6 billion for the same purpose (National Treasury 2022a:53).  
 
It is likely that a balance between these three options – reducing headcounts, defunding other line items to support 
compensation spending, or adding new resources into the education budget - will be sought during the year, resulting 
in a combination of these measures.  
 
Added to all of this is uncertainty about the funding for teaching assistants, which makes up the lion’s share of the 
President’s flagship employment stimulus. A table in an online annex to the budget review indicates that R6.5 billion 
will be added to the equitable share in 2023 to fund the programme, but no resources are envisaged in 2024 and 
beyond. While evidence points to the potential benefits of teacher assistants when carefully selected and properly 
supported (see Ardington and Henry 2021), the significant resources allocated to this programme needs to be set 
against the concurrent fall in educator numbers and their real incomes, and the failure of government to allocate a 
budget to catch up from learning losses arising from the Covid-19 pandemic. 
 
 The subsidies going to early learning centres are significantly extended. The Early Childhood Development grant 
was allocated R520 million in 2019, increasing to a planned R1.7 billion in 2024 (a 30% increase). Relative to need 
however, the amounts remain small. The national grant is sufficient for 238 000 learners in 2022 rising to 486 000 
by 202511. However, it is estimated that there are more than 1.6 million eligible children (World Bank and DBE 
2022). Moreover, the value of the grant is insufficient, and needs to be roughly doubled to be adequate programme 
operating costs at a minimum level of service (Ibid). Instead, the grant has been fixed at R17 per child per day since 
2019, implying an 18% fall in its real value since then. There are also serious question marks about the shape of the 
ECD sector going forward. While the 
subsidy is currently channelled to 
independent ECD centres, the Minister 
of Basic Education said in her budget 
speech that “the current model of 
funding ECD programmes must be 
reviewed… it is important that a new 
holistic and inclusive model includes all 
communities of trust in the ECD space” 
(Motshekga 2023). It is unclear what this 
implies, but the Minister is keen to stress 
the role of state departments over 
independent service providers and 
appears to favour an downward 
extension of the formal schooling system, 
building on the success of grade R 
enrolments over the last decade. An ECD 
system based in government schools 
would be in stark contrast with the 
current approach and, if implemented, 
would likely involve a significant 
escalation of the programme’s costs.  
 
 

                                                      
11 This reflects the number of children that the national ECD conditional grant can finance. These funds are, however, complemented by provincial allocations 
out of the equitable share, which raise the number of learners benefitting from government programmes to the region of 630 000.  Nevertheless “provinces 
report insufficient budget to reach all children eligible for the subsidy” according to the recent expenditure review (World Bank and DBE 2022:ix).  

Figure 9: Post-school education spending growth (2019-2024) 

Selected items | Average annual nominal growth rate  

 
Source: National Treasury budget data; Public Economy Project calculations  

5.0%

-8.8%

0.5%

3.2%

3.3%

3.5%

3.9%

3.9%

3.9%

6.3%

6.3%

6.3%

6.3%

6.5%

8.4%

11.0%

11.1%

-10% -5% 0% 5% 10%

CPI

TVET infrastructure grants

TVET current grants

TVET college funding

University infrastructure grants

NSFAS grant admin

University core subsidy

University funding

TVET compensation of employees

Community adult education

SETAs

Skills development levy

National skills fund

PSET total

New university infrastructure

NSFAS

NSFAS student support



 

Page | 15  
 

Funding for the increase in the ECD grant was partly 
found by reducing the budget for higher education. 
The ENE reports that cabinet-approved decrease of 
R1.8 billion to the 2023 baseline to fund early 
childhood development. This has mainly been drawn 
from the capital budgets, with funding for university 
infrastructure slashed in 2023 to just R673 million, 
down from R3.5 billion last year. Although the 
medium-term plan is to return to normal funding 
levels in 2024, the budget pencils in an amount for 
university infrastructure grants in 2025 that is below 
the amount spent in 2017. Capital allocations for 
construction of the new Sol Plaatjies and Mpumalanga 
universities have also been cut in all three years of the 
MTEF, as has the budget for technical college 
infrastructure. Funding from these infrastructure 
budgets has (in addition to supporting ECD grant) 
also been shifted to finance the construction of nine 
Community Colleges, which aim to provide learning 
opportunities for adults.  
 
Against this backdrop, it is surprising that the Minister 
of Higher Education said in his budget speech that the 
construction of two new Universities would 
commence “in the next year or two” (Nzimande 
2023). The wisdom and feasibility of establishing a 
new “University of Science and Innovation” in 
Ekurhuleni and Crime Detection University in 
Hammanskraal, seems flawed and based on populist 
impulses rather than clear policy choices. At the very 
least, it is wholly inconsistent with the numbers 
presented in the department’s own budget.  
 
The cuts to capital budgets for universities and 
technical colleges come as funding for new student 
enrolments through NSFAS massively outpaces the 
rest of the budget (see Figure 10). The implication 
could be that a rising number of university 
undergraduates face a dwindling resource base as 
spending for lecturers, infrastructure and operating 
payments is cut. Such imbalances may lead to a 
deterioration in the quality of learning and teaching at 
higher education institutions, while the ability of 
universities to sustain research and post-graduate 
learning is also likely to come under continued 
pressure.  
 
Despite this, expanded enrolment is clearly top of 

mind for the minister, who also declared that “the time has arrived for us to drastically increase intake in TVET and 
Community Colleges” (ibid). The first goal of the national PSET plan is the “massification” of the sector, with 
targets of 1 million enrolled in Community Colleges and 2.5 million learners in public and private technical colleges 
by 2030. These claims, however, are not backed by the performance targets reported in the ENE, which see robust 
growth in university enrolment contrasting with recent falls in technical college students. Although college 
enrolments are planned to increase, by 2025 they would not have reached the level of enrolment recorded in 2020 
(see Figure 11).  
 
Even assuming these modest enrolment figures, the budget for higher education implies a falling real spending per 
learner Figure 12). NSFAS looks to be stretching each rand further, allocating less resources per learner in the period 

Figure 10: Index of Higher education spending growth 

 
Source: National Treasury budget data; Public Economy Project calculations 

 

Figure 11: Students enrolled in universities 

and technical colleges 

 
Source: Estimates of National Expenditure 
 

Figure 12: Real spending per learner 

 

Source: National Treasury budget data; StatsSA, DBE, Public Economy Project calculations 

50

100

150

200

250

300

350

400

450

500

2
0

1
7

2
0

1
8

2
0

1
9

2
0

2
0

2
0

2
1

2
0

2
2

2
0

2
3

2
0

2
4

2
0

2
5

In
d

e
x:

 2
0
1
7
=

1
0
0

NSFAS

Technical college core funding

Universities core funding

Adult education

Higher education infrastructure

Projection

budget year

0.4

0.5

0.6

0.7

0.8

0.9

1.0

1.1

1.2

2
0

1
7

2
0

1
8

2
0

1
9

2
0

2
0

2
0

2
1

2
0

2
2

2
0

2
3

2
0

2
4

2
0

2
5

M
il

li
o

n
s

University enrolment

College enrolments

Projection

0

10

20

30

40

50

60

2
0

1
7

2
0

1
8

2
0

1
9

2
0

2
0

2
0

2
1

2
0

2
2

2
0

2
3

2
0

2
4

2
0

2
5

C
o

n
st

a
n

t 
2
0
2
2
 R

'0
0
0  NSFAS  University core funding

 TVET core funding Schools

ECD conditional grant Projection

budget year



 

Page | 16  
 

ahead. Core funding for universities and technical colleges will fall faster, with real funding per learner dropping by 
9% and 3% respectively between 2019 and 2024. Schools, on the other hand, face a much larger real cut over this 
period, amounting to 11%. The value of the national ECD grant per early learner will have fallen by 22% by 2025.  
 
These are powerful indicators of inverted spending priorities in the education sector. The recent finding that 81% 
of learners in grade 4 cannot read for meaning in any language underscores the strong argument that failures at pre-
school and foundation phase reverberate powerfully throughout the whole education system, and largely explain 
the high costs and poor learning outcomes of senior and post school education. Yet government continues to 
prioritise funding to the latter, in a perhaps futile attempt to reverse the consequences of failure to allocate greater 
resources to the former.  
 

Healthcare 
 
The Covid-19 pandemic left large backlogs in treatments and procedures in its wake. This adds to the pressure the 
health system faces in a context of rising poverty and hunger, and health crises triggered by failing social 
infrastructure. If Budget 2023 is executed as tabled in February, the public healthcare system faces a large contraction 
in 2023. This comes after a decade of falling real spending per capita on healthcare, with a modest reversal to fight 
Covid-19 (see Public Economy Project 2022).  
 
The healthcare sector will lose R4 billion in funding in 2023 as temporary Covid-19 programmes come to an end. 
Table 2 shows these direct allocations since 2020. This includes spending on vaccines and grants to district health 

system to manage the pandemic.  
 
The path of compensation spending is 
highly uncertain in the wake of the 
pandemic. Provincial health 
departments employed additional staff 

on a temporary basis and it is not clear how many of these will be absorbed into permanent employment. On a 
consolidated basis, the budget estimated that compensation would fall from R164.4 billion in 2022 to R163.6 billion 
in 2023. This nominal drop reflects two assumptions. First, no cost-of-living adjustment included. Second, 
temporarily contracted staff – or at least a large proportion of them – will be let go. The first assumption has proven 
incorrect as an effective 3.3% cost-of living adjustment has been agreed. By our calculations, this gives rise to an 
R5.4 billion spending pressure in the sector in 2023, and R23.6 billion over the MTEF. In this context, there are 
serious concerns about the need to avoid paying for salary increases by defunding other components of health 
consumption (such as medicines, 
equipment machines, and services). 
 
The second assumption will be 
tested as provinces struggle with 
critical unfilled vacancies, and 
health workers struggle to find 
employment in the public sector. 
The last decade has seen falling 
employment in the sector, with 
many critical posts left vacant to 
meet budget pressures. The 
combination of a shortage of 
clinical staff and an unemployed 
surplus of trained personnel is 
difficult to sustain. National health 
policy supports a large expansion 
of employment to meet sufficient 
levels of clinical, professional and 
nursing staff, with the 2030 Human 
Resources for Health Strategy 
concluding that the shortfall in 
essential health workers will worsen by 2025 if health workforce expenditure only increases in line with 
inflation.”(NDOH 2020:19). In this context it is likely that some proportion of contract staff will see their contracts 

Table 2: Temporary Covid-related expenditures 

 

Figure 13: Healthcare spending growth (2019−2024) 

Selected items | Average annual nominal growth rate    

 
Source: National Treasury budget data; Public Economy Project calculations 

R billion 2020 2021 2022

Vaccines and other medical inventory 0.5              7.6              2.1              

COVID-19 component of the district health grant 3.4              1.5              1.9              

Total 3.9              9.1              4.0              

5.0%

-1.7%

1.4%

1.5%

2.1%

2.5%

2.7%

3.0%

3.1%

3.6%

3.7%

3.8%

3.8%

4.1%

4.5%

6.9%

8.1%

11.6%

11.9%

-2% 0% 2% 4% 6% 8% 10% 12%

CPI

Transfers and subsidies

Regulators and agencies

Machinery and equipment

Tertiary services grant

Central hospital services

Provincial hospital services

Health facilities grant

Emergency medical services

District health services

Compensation of employees

Transfers to non-profits (NDoH)

Consolidated health spending

HIV/AIDS and TB grant

Goods and services

Human resources grants

Buildings

NHI programme

District health grants



 

Page | 17  
 

renewed or be absorbed into permanent establishments. Another wildcard in the government health employment 
picture is the absorption of 50 000 community health workers onto governments payroll. The National Department 
of Health is working to permanently engage these contract workers, including them in collective bargaining 
arrangements. This approach is strongly supported by unions, and provincial departments are already acting to bring 
CHWs onto governments payroll.  
 
Figure 13 shows how the health budget is changing over the medium term. It abstracts from the impact of the Covid 
pandemic by showing annual average growth between 2019 and 2024. Only four items grow in real terms (i.e., faster 
than CPI). Two of these are conditional grants from national government. The district health grant is relatively 
small: R3bn in 2023 in a district health services budget of R113 billion. Its rapid growth signals government’s 
continued commitment to primary health care.  
 
The second are human resources conditional grants, which have been increased to absorb medical interns and fill 
of community service posts, which are critical for student-doctors to complete their training, but also contribute to 
service delivery. Ideally, absorbing more medical interns and community services doctors should be balanced with 
employment of medical practitioners and specialists to train and supervise them.12  In a context where the budget 
for compensation implies significant headcount reductions, the direction of travel may well point rising imbalances. 
More broadly, the sharp constraints on compensation and the intended resort to attrition to reduce headcounts and 
pay can adversely affect the balance between clinicians, nurses and auxiliary staff, all of whom are vital to an effective 
hospital services.  
 
Overall, hospital services bear the brunt of the cuts to operational budgets, with central hospitals facing particularly 
intense financial pressure. While much of this reflects (assumed) cuts to employee compensation in the next two 
years, the budget also implies real cuts to spending on machinery, equipment, goods and services. This amounts to 
an across-the-board reduction in the resources available for hospital services at district, provincial and tertiary level 
 
Funding for the NHI programme expands rapidly and this largely directed at capital expansion plans overseen by 
the National Department, including the construction of the Limpopo Central Hospital in Polokwane (funded from 
NHI indirect grant). The national department’s budget for capital spending has doubled from around R750 million 
each year prior to the pandemic to more than R1.5bn each year over the MTEF. Financing for provincial 
infrastructure improvements grows at a far slower pace.  
 
Expanded resources to revitalise existing facilities and build new hospitals may be a case of building in advance of 
demand. Allocations for healthcare workers, medical equipment and operating budgets are contracting in real terms 
and it is not yet clear when this decline will come to an end. Government will need to consider how expanded 
infrastructure will be brought into operation, 
the additional calls on the resource base of the 
system that are likely to be generated, and the 
reallocation of people and finances this may 
entail. When considering ambitious capital 
expansion programmes, it is important to 
remember that that capital budgets are often 
underspent. Until now, most facilities spending 
has mainly been provincial there is perhaps a 
chance that the national department will make 
a better showing, but this remains to be seen.  
 
The HIV/AIDs and TB grant constitutes a 
large share of funding for district services, 
allocated for very specific purposes. Once we 
take account of increase in the number of 
people on antiretroviral treatment, the value of 
the grant per beneficiary looks set to fall. This 
might reflect savings in the procurement of 
drugs, but in the context of a depreciating exchange rate, the estimates of need look conservative.  
 

                                                      
12 Lencoasa, Mapipa, and Chaskalson (2022) 

Figure 14: Real spending per user on anti-retroviral therapy 

 

Source: Estimates of National Expenditure Note: District health programmes grant: 
Comprehensive HIV/AIDS component, deflated with CPI and divided by “number 
of clients remaining on antiretroviral treatment” 

5 420 
5 232 

4 742 
4 613 

4 749 

4 388 

3 802 

3 395 
3 244 

0

1000

2000

3000

4000

5000

2017 2018 2019 2020 2021 2022 2023 2024 2025

C
o

n
st

a
n

t 
2
0
2
2
 r

a
n

d

PEP projections



 

Page | 18  
 

Finally, it is worth noting the fall in transfers and subsidies over the last five years. This reflects three factors. First, 
pay-outs for medico-legal claims appear to have stabilized, and new legal approaches to settlement count ensure 
that this is a permanent shift. Second, provincial allocations to NGO’s providing health services has also declined 
significantly, even while national funding for NGOs (reflected separately in the graph) have broadly kept pace with 
the budget. Third, provincial allocations to support municipal health services have also declined.  
 

Social protection 
 
Government aims to influence the course and pace of development by directly intervening in society to address the 
challenges of poverty, underdevelopment and the marginalisation of vulnerable individuals and communities. 
National and provincial departments of social development play a critical role in addressing inequality, poverty, 
gender-based violence, and femicide in the country.  To do this government allocates resources for social protection 
in the form of cash transfers, social services and social security funds. While national government manages the 
provision of cash transfers, social services are provided directly by provincial departments, and indirectly through a 
system of transfers and subsidies to nonprofit organisations. Social security funds are national entities governed by 
their own legislation.   
 
The consolidated budget for social protection increases in real terms over the medium term (see Figure 15). This is 
largely driven by the introduction of the Covid-19 Social Relief of Distress grant (Covid-SRD) grant, which we 
assume will be extended beyond 2025. This grant required government to extend income support to working-age 
adults. More than 7.5 million people received the grant in 2022 (down from more than 10 million in 2021), and 
research shows that the Covid-SRD has provided much-needed help to the unemployed.13 Three years since the 
pandemic, however, and government is yet to clarify the status and future of the grant. Not only have the number 
of beneficiaries fallen as more exclusionary regulations are enforced, but the monetary value of the grant has not 
been adjusted for inflation since its introduction at R350 per month in 2020.  
 
The budget has not made any allocation for the grant beyond 2023 but the “unallocated reserve” of R35.6 billion 
in 2024 and R44.5 billion in 2025 almost exactly covers the shortfall in the event (which we deem highly likely) that 
the grant continues to be extended over the medium-term. Factoring in moderate beneficiary growth and assuming 
the grant will be adjusted for inflation in 2024, we estimate that there could still be a shortfall of R4.5 billion in 2024 
adding to spending pressures facing the budget. 
The policy and legislative framework governing 
the current grant remains unclear, and government 
is yet to agree on whether it will be terminated, 
extended, or replaced by a more extensive and 
comprehensive system of basic income support 
for people of working age.  
 
Figure 15 shows that total spending on the old age 
and child support grants (which account for more 
than 70% of grant spending (even include the 
Covid-SRD) broadly tracks inflation over the 
period 2019-2024. Figure 16 shows value of 
monthly cash payments per recipient in terms of 
real purchasing power over a longer period. We 
have deflated monthly grant values with a weighted 
CPI index to reflect the price pressures facing 
grant recipients. As can seen the real value of both 
grants will be substantially below their 2019 level 
by 2024. The size of the real cut is particularly large 
in the case of the old age grant. For both grants, 
the erosion of the value of the grants will continue over the medium term with just in-line inflation increases, which 
will put pressure on poor households, particularly children, especially if food inflation continues to increase by 

                                                      
13 (Patel, Dikoko, and Archer 2023) 

Figure 15: Social protection spending growth (2019-2024) 

Selected items | Average annual nominal growth rate  

 
* Transfers to household assumes that the “unallocated reserve” is used to fund the extension of the Covid-SRD grant 
Source: National Treasury budget data; Public Economy Project calculations.  

5.0%

-2.5%

0.9%

2.4%

2.6%

3.4%

3.6%

4.3%

5.0%

6.4%

7.2%

-5% -3% -1% 2% 4% 6% 8%

CPI

Transfers to non-profits

SASSA

Provincial social development

Compensation of employees

Goods and services

Disability grant

Child support grant

Old age grant

Consolidated social protection

Transfers to households*



 

Page | 19  
 

double digits.14 It can be seen, however, that grant values rapidly increased in the years prior to Covid-19, and the 
current course appears to be unwinding those gains.  
 
The allocation for the South African Social Security Agency (SASSA), which administers and pays social grants on 
behalf of the department, falls in real terms over the medium term. This raises questions about the capacity of 
SASSA to deliver on its expanding mandate as the grant system is extended. Provision of social grants has also been 
affected by load-shedding, leading to interrupted services, and contributing to backlogs and longer waiting times for 
beneficiaries.  
 
Unlike cash transfers, care services that respond to the social impacts of poverty, unemployment, drug addiction, 
abuse and violence against women and children depend on the employment of social workers, psychologists, and 
other social service professionals. As is the case for health, basic education and criminal justice, the 2023 budget 
implies an erosion of services with negative consequences for social development outcomes. Provincial social 
development departments, which provide social services, are seeing real cuts in spending while transfers to non-
profit service organisations are being cut back severely.  
 
The National Development Plan (NDP) envisaged a doubling of the social protection workforce nationwide by 
2030 to 55 000 social workers. Currently, government employs approximately 18 948 social workers, with a small 
number contracted for a short-term period through the Presidential stimulus package. The department of social 
development estimates that the sector’s workforce needs are underfunded by R9.6 billion and believes that new 
employment will be phased into over the next six years15. But the compensation budget is falling in real terms, as is 
the allocation to provincial departments who are the main employers of social care workers. This is another situation 
in which policy intentions and budget realities are walking in opposite directions. On current trends, the budget 
allocation could lead to a deterioration in the quality of services and, in some cases, a lack of service provision. 
 
Transfers to non-profit organisations (which are an alternative to government-provided social services) face a large 
nominal cut in spending.  In addition to its own social workers, government uses a pool of social workers in the 
non-profit sector. These NPOs provide critical restorative and social welfare services to children and families, HIV 
services, the elderly, child protection, and people with disabilities at the provincial level. Government's transfers to 
non-profits have been cut from about R7.3 billion in 2019 to just R76 billion in 2023.  
 
The combination of falling employment of social workers and cuts in subsidies to nonprofit care services will reduce 
the availability of social welfare services to poor communities in South Africa. These shortfalls will increase the 
burden of care faced by household or family members, shifting responsibly from government to underpaid and 
unpaid women. 
 

Figure 16: Real monthly value of old-age grant and child support grant 

(a) Old-Age Grant 

 

(b) Child support grant 

 

Source: National Treasury budget data; StatsSA (CPI inflation indices by decile). Public Economy Project calculations 

 

                                                      
14 https://www.statssa.gov.za/?p=16254 

15 See Parliament of South Africa (2023) 

1900

1950

2000

2050

2100

2150

2
0

0
6

2
0

0
7

2
0

0
8

2
0

0
9

2
0

1
0

2
0

1
1

2
0

1
2

2
0

1
3

2
0

1
4

2
0

1
5

2
0

1
6

2
0

1
7

2
0

1
8

2
0

1
9

2
0

2
0

2
0

2
1

2
0

2
2

2
0

2
3

2
0

2
4

C
o

n
st

a
n

t 
2
0
2
2
 R

a
n

d

budget year

PEP

projection

450

460

470

480

490

500

510

2
0

0
6

2
0

0
7

2
0

0
8

2
0

0
9

2
0

1
0

2
0

1
1

2
0

1
2

2
0

1
3

2
0

1
4

2
0

1
5

2
0

1
6

2
0

1
7

2
0

1
8

2
0

1
9

2
0

2
0

2
0

2
1

2
0

2
2

2
0

2
3

2
0

2
4

C
o

n
st

a
n

t 
2
0
2
2
 R

a
n

d

budget year

PEP 

Projection



 

Page | 20  
 

Criminal justice and Defence 
 
Real spending on criminal justice and defence will decline in the 2023 Budget. Nominal spending increases by an 
average of 2% between 2019 and 2024, far short of CPI and implying a real decline in spending of 2.9% per year 
over the period. The decline is not evenly distributed, as can be seen in Figure 18 below. Real spending on the 
National Prosecuting Authority and the Special Investigative Unit will grow by an average of 6.7% per year, which 
hopefully should support more effective prosecutions. All other components of spending will fall on a real basis, 
and spending on the air force and prisoner care will decline on a nominal basis. These nominal cuts will have a 
direct impact on service delivery. 
 
Criminal justice, policing and defence are inherently labour-intensive activities, and spending on compensation of 
employees accounts of 70% of its budget. Goods and services account for a further 22% of spending. This means 
that these services will be severely impacted by the unbudgeted wage settlement, and government’s overall 
commitment to reducing consumption spending. 
 
It is important to note that the sector has already faced significant headcount reductions. Police headcounts have 
declined about 20 thousand personnel since the world cup in 2010 (see Figure 17). In contrast to previous budgets, 
which projected a continuation of falling headcounts, the 2023 ENE projects a reversal of this trend. The headcount 
projections show particularly strong growth in visible policing, detectives and crime intelligence, while slightly 
reducing the numbers of bureaucrats and VIP protectors.  In his budget vote speech, the Minister of Police said 
that “a total of 10 000 police trainees have been enlisted in the 2023/24 financial year with the additional funds 
allocated in the previous and current budget allocation to SAPS. A total of 10 000 police trainees will be recruited 
annually for the next two financial years which will bring a total of 30 000 new police recruits in the MTSF period” 
(Cele 2023). 
 
However, the headcount numbers in the ENE are not supported by the allocations for compensation of employees 
in the same document. Once we take account of the impact of the 2023 wage settlement, the budget implies a force 
reduction of about 8 000 personnel in 2023. This is unlikely to happen but gives an indication of the spending 
pressure faced by the department in the current year. Indeed, it is difficult to see how the police or the defence force 
will be able to accommodate the wage increase within the current baseline, and it is probable over the next year 
there will be cuts to budgets for capital, equipment and training, reduced employment through attrition and budget 
adjustments to prevent overspending. 

 
South Africa’s crime epidemic has many causes, and 
there are strong and justifiable concerns about 
inefficiency, maladministration and corruption within 
the police service. Reform to police operations, 
particularly those that separate political from operational 
authority, and stronger efforts to root out corrupt and 
criminal networks are clearly required. However, any 
process of significant reform and change management is 
likely to be lengthy and difficult. In the meantime, 
declining budget allocations will directly impact service 
provision and add additional strains to the reform effort.  
 
The South African National Defence Force faces similar 
challenges, but to an even larger degree, with the 
Minister of Defence indicating in her budget speech that 
the compensation budget was “underfunded by 
approximately R2,6 billion based on the actual feet on 

the ground”. Last year the department of defence spent R2.5 billion more than its budget allocation, mainly on 
salaries. This follows expenditure overruns amounting to R3.2 billion in 2021 and R2.2 billion in 2020.16 Yet its 
budget will be reduced further in nominal terms in each of the next two years. It is not clear how the defence force 
can continue to sustain operations in the face of such large declines, including critical peacekeeping operations in 
the Democratic Republic of the Congo and Mozambique that are vital to stability and security in the region. Again, 

                                                      
16 Spending more the allocation provided by Parliament results in unauthorised expenditure. Compensation of employees is a special case because this allocation 
is ringfenced by Parliament, so overspending on compensation of employees results in irregular expenditure.  

Figure 17: Police employment 

 

Source: Estimates of National Expenditure, Public Economy Project calculations 

160

165

170

175

180

185

190

195

200

205

2
0

1
0

2
0

1
1

2
0

1
2

2
0

1
3

2
0

1
4

2
0

1
5

2
0

1
6

2
0

1
7

2
0

1
8

2
0

1
9

2
0

2
0

2
0

2
1

2
0

2
2

2
0

2
3

2
0

2
4

2
0

2
5

T
h

o
u

sa
n

d
s

Budget year

Outcome

ENE 2023 estimate

Implied by 2023 wage agreement (PEP)

Projections



 

Page | 21  
 

unlike other departments covered in this report, the Minister of Defence is unusually forthright the consequences 
of funding shortfalls for the capacity of the state, saying that: “A scientific evaluation of the SANDF has painted a 
bleak picture of our diminishing capabilities, largely because of persistent budget cuts. The ravages of underfunding 
and unserviceable capabilities against escalating tasks has had a devastating effect” (Modise 2023) 
 
 The solution to this acute fiscal 
squeeze involves a clear agenda of 
reform to force levels and operations. 
The increasing age of the SANDF 
means that fewer full-time soldiers are 
available for deployment, resulting in 
the use of reserve forces. The 
retirement of significant numbers of 
personnel, backed by a plan to 
permanently reduce the size of the 
force, and the nature of its operations 
is needed if the defence force is to 
operate within budget. As articulated 
by the Minister: “Every appreciation 
process of the DOD points to the 
inevitable need for more boots on the 
ground to execute the many tasks 
given to the Department. To the 
contrary, Government directs the 
downward manage of the human 
resource baseline. It is our duty to 
reconcile these mutually exclusive 
positions. Sustained funding of the 
SANDF remains an unavoidable requirement.” (Ibid) 
 
The prison system continues to house 33% more inmates than it has capacity for. This average hides the variation 
between facilities and in some cases, prison facilities must accommodate inmates up to 245% of capacity. 
Tuberculosis is endemic and understaffing results in prisoners spending much of the day locked up, sometimes with 
only a few minutes of exercise outside of their cells. This results in higher levels of self-harm amongst prisoners and 
injuries to others.  There are higher levels of aggression, violence, drug use and other health issues. There is also 
evidence that overcrowding has resulted in an increase in gangsterism and smuggling of contraband, including drugs, 
negatively impacting the safety of prisoners who are not gang members (Sibisi and Olofinbiyi 2021). Prison officials 
are also sometimes harmed, and the Department has targeted keeping injuries to prisoners below 4.55% of the 
prison population.  
 
The allocation for incarceration will grow with inflation over the medium term, but the allocations for other 
components of the department of corrections, for instance, the allocations for prisoner care and rehabilitation, face 
declining real spending. Any substantial improvement in prisoners’ living conditions is unlikely. Access to 
rehabilitation programmes or skills programmes that could be used for making a living outside of prison will be 
reduced. Increasingly, the corrections system could be forced into focusing only on detaining prisoners but doing 
less and less to help those prisoners become reintegrated into society once they have served their sentences. To 
reduce overcrowding, it is imperative that more prisons are built and staffed appropriately, but the Department of 
Corrections has a poor track record of completing infrastructure projects, underspending 30% of its capital budget 
over the last three years.  
 

Public infrastructure and government buildings 
 
Government’s commitment to expanding infrastructure delivery is evident in the relatively strong growth of 
spending on several facets of public infrastructure (see Figure 19). Additions to the capital stock funded directly 
from the budget includes social infrastructure (such as school buildings, health facilities and housing), municipal 
infrastructure (funded by conditional grants) and the capital spending programmes of key public agencies such as 
the water boards, SANRAL (national roads), PRASA (commuter rail). Figure 19 also shows the growth in spending 
for public transport systems.  

Figure 18: Criminal justice and defence spending growth (2019-2024) 

Selected items | Average annual nominal growth rate  

 
Source: National Treasury budget data; Public Economy Project calculations  

5.0%

-1.2%

-0.8%

0.8%

0.9%

1.0%

1.1%

1.9%

2.0%

2.0%

2.5%

2.5%

3.0%

3.1%

3.1%

3.2%

4.6%

6.7%

-2% -1% 0% 1% 2% 3% 4% 5% 6% 7%

CPI

Airforce

Army

Military health

Military logistics

Incarceration

Navy

Court Services

Average

State legal services

Visible policing

Crime intelligence

Legal Aid

Prisoner care

Protection services

Detective services

Offender rehab

NPA+SIU



 

Page | 22  
 

 
During the Covid-19 pandemic in 2020, most infrastructure budgets were cut but allocations were restored in 2021. 
Among the directly funded infrastructure programmes reflected in Figure 19, national roads account for the largest 
share, with spending averaging R48 billion per annum over the MTEF. It is also one of the fastest growing 
programmes across the whole budget. In additional to this strong commitment to road building, government 
allocated R24 billion to SANRAL as a capital transfer in 2022 to help repair its balance sheet, which has been 
damaged by the e-tolls fiasco. Allocations for the maintenance of provincial roads amount to another R12 billion 
annually, but this budget grows far below the rate of inflation.  
 
 Water infrastructure funded by direct allocations from the budget also increases.17 The budget for local and regional 
water infrastructure grows rapidly, increasing from R8 billion in 2019 to a planned R17.1 billion in 2025. Spending 
on national water infrastructure is far smaller (around R3.5 billion per annum on average over the MTEF) by also 
increases faster than inflation. The budget for commuter rail has been increased since 2022, and this is retained over 

the medium term (although the increase 
in 2023 is meagre). Subsidies for regional 
and local bus services grow far below the 
rate of inflation. Given the sharp increase 
in petrol prices over the same period this 
must surely imply a significant reduction 
in service levels unless provinces and 
municipalities are offsetting the cut from 
their own resources. Expenditure on 
human settlements infrastructure is 
stagnant. Provinces spent R20 billion on 
housing in 2017, yet by 2025 the budget 
plans only R20.5 billion. Grants to local 
governments hardly grow faster and 
support for community libraries is 
similarly lacklustre.  
 
There are several risks to these ambitious 
infrastructure targets. The first is the high 
likelihood of underspending capital 
budgets. This is especially true when 
financing takes the form of conditional 

grants. The second is that the estimates in the outer year of the MTEF – which are the main driver of rapid medium 
term growth rates – may be moderated in future years to accommodate above-budget consumption spending. Lastly, 
executing infrastructure programmes requires effective state institutions. This is a serious constraint as instability in 
the water sector, governance failures in commuter rail and funding uncertainty for national roads have all led to 
poor outcomes in the past, despite substantial allocations.  
 

Economic development and public employment  
 
Spending on economic services and development policy is undergoing a major contraction. Only a few items will 
grow faster than inflation over the medium term – the department of science and innovation, the department of 
agriculture’s food security programme, and competition policy. Otherwise, all elements of this group face substantial 
budget pressures. Many of the economic services are employee-intensive and the contraction is no doubt part of 
the broader planned reduction in government consumption spending and reflects the unrealistic salary assumptions 
that beset the core public services. On the other hand, this does