South Africa grew for six consecutive quarters, then contracted by 0.2% in the second quarter of 2026. The disruption at the Strait of Hormuz is the convenient explanation. It is not, however, a sufficient one.
Manufacturing was already shrinking, fixed investment was already falling and construction was already under strain. Higher fuel prices found a weak economy rather than creating one. What the quarter exposed is that six quarters of growth had restored output without adding enough capacity to sustain it.
That is not simply an economic problem. It is a political-economy problem: South Africa has stabilised parts of the state without yet creating the political and institutional conditions under which businesses can confidently commit capital to expanding the productive base.
A country can restore electricity supply, improve its fiscal credibility and repair damaged institutions without necessarily creating the conditions for a new investment cycle. Recovery restores what has been lost. Expansion requires decisions about what comes next — and those decisions depend as much on the credibility of the political settlement as they do on the cost of capital.
Recovery is not expansion
A recovery returns output towards a level already reached. Expansion carries it beyond that level because capacity has been added.
Six quarters of growth from a low base therefore tells us less than the headline suggests. Electricity, gas and water, and construction remain below their 2018 levels. Manufacturing has contracted in five of the past seven quarters, was 3% lower year-on-year in Q2 and saw seven of its ten divisions shrink. Mining contracted by 3%.
These are not sectors pushing through a previous ceiling. They are sectors that have not yet rebuilt the capacity from which sustained expansion can occur.
The distinction matters politically. The state can claim progress when a failing system begins to function again. Citizens and businesses experience that improvement differently. The political system receives credit for restoring what was broken; the private sector still has to decide whether the restored system is reliable enough to justify irreversible investment.
That is the gap South Africa has not yet closed.
The national accounts rebasing due in October could make the distinction harder to see. A larger nominal denominator can improve debt-to-GDP and deficit ratios without creating additional productive capacity. The revision is routine statistical practice. What matters is how the improvement is interpreted.
Better ratios are not the same as greater capacity.
Growth is not reaching households
Even taking the growth numbers at face value, expansion remains too weak relative to the population it must support.
First-half growth of 1.4% came below Treasury’s 1.6% forecast, while the Reserve Bank sees 1.4% for the year. Population growth is close to 1%, leaving output per person flat at best after more than a decade of declining real income per capita.
This creates a political problem that cannot be solved by better aggregate numbers alone.
Households do not experience GDP. They experience employment, prices, electricity, water, transport and public services. A reform that improves electricity availability or fiscal credibility can be economically meaningful while producing little immediate improvement in a household budget.
The labour market illustrates the problem. In Q2, the labour force grew by 329,000, or 1.3%, while employment fell by 16,000. Unemployment increased by 345,000 to 8.5 million, taking the official unemployment rate to 33.6%, with broad labour underutilisation at 46.3%.
The economy is adding people to the labour market several times faster than it is adding work.
That changes the political economy of reform. A state can improve its balance sheet and restore infrastructure while large numbers of citizens see little change in their economic position. This weakens the political constituency for reforms whose benefits are delayed while increasing pressure for measures that produce more immediate distributional relief.
The reforms required to increase future productive capacity are not necessarily those that produce the fastest political returns.
Restored to function, not built for growth
Much of South Africa’s recent reform effort has restored functions rather than created surplus capacity.
Load-shedding has ended. South Africa is off the Financial Action Task Force grey list. Ratings have improved. Each represents the restoration of something that had been lost. None, by itself, creates the additional capacity required for a larger economy.
For business, the question is no longer simply whether the lights stay on. It is whether there will be sufficient reserve margin over the life of an asset; whether freight will move when production reaches volume; whether ports can absorb additional exports; whether water infrastructure can support a new facility; and whether the regulatory and political environment will remain sufficiently durable for the investment to pay back.
Stabilisation allows the existing economy to function.
Surplus capacity allows a larger economy to be built.
This helps explain private investment behaviour. Fixed investment grew by only 0.8% in the first half against an expected 2.4% and fell for a second consecutive quarter. Construction works declined by 4.0% and transport equipment by 3.4%.
The composition matters as much as the level. Businesses have been committing capital to self-generation, water storage, backup logistics and private security. This is defensive capital: investment designed to insure existing output against weaknesses in the public system.
It is economically rational. It is also politically revealing.
Capital deployed to compensate for state incapacity raises the cost of doing business without necessarily increasing productive capacity. The private sector effectively pays twice: once for infrastructure it expects the state to provide and again for private alternatives when that infrastructure cannot be relied upon.
The political settlement is part of the investment equation
This is where South Africa’s current political settlement matters.
The Government of National Unity has introduced political stability after the 2024 election, but stability and coordination are not the same thing. For investors, the question is not simply whether the government will survive. It is whether policy commitments will survive the bargaining required to keep the governing coalition together.
The 2025 budget illustrated the problem. It was tabled three times before being passed, with the dispute centred on half a percentage point of VAT rather than a fundamental change in economic direction.
The episode demonstrated that the GNU can absorb disagreement, but also exposed the difficulty of resolving competing political interests without creating uncertainty for economic actors.
That matters because fixed investment is inherently political in the broadest sense. A company building a factory, mine, data centre or logistics facility is committing capital for years. It therefore prices not only today’s policy but the probability that policy will remain sufficiently durable over the life of the asset.
With municipal elections approaching on November 4 and the ANC’s elective conference in December 2027, political incentives will not always favour coordination. Coalition partners have incentives to differentiate themselves even when government requires collective action.
This does not make policy paralysis inevitable. It does mean that policy announcements carry less value to investors when the mechanism for sustaining them across political bargaining is uncertain.
Policy uncertainty is therefore no longer principally about content, but about durability.
From policy announcements to state coordination
South Africa does not lack policy announcements. It lacks sufficient conversion of policy into coordinated, investable and functioning capacity.
Treasury can allocate money for infrastructure. Departments can announce reforms. Regulators can change rules. State-owned companies can publish turnaround plans. But productive capacity emerges only when these decisions align.
A new factory needs electricity, water, transport, digital connectivity, municipal services, regulatory approvals and predictable demand. If each element is managed through a separate institutional and political process, the investment decision becomes a calculation of cumulative uncertainty.
This is why the political economy of growth cannot be reduced to whether government is “pro-business” or “anti-business”. The more important question is whether the state can coordinate multiple institutions around outcomes that require decisions to hold over time.
South Africa’s challenge is increasingly one of state capacity and political coordination rather than policy scarcity.
The MTBPS as a political signal
This is why the Medium-Term Budget Policy Statement matters. The MTBPS cannot restore freight capacity or resolve coalition disputes. Its importance, however, lies elsewhere. It is one of the few regular instruments through which government can demonstrate whether its political settlement is capable of translating into a coherent economic direction over a medium-term horizon.
It should therefore be read not simply as a statement of revenue, expenditure and debt, but as a statement of political priorities. Three things will matter.
First is the balance between consumption and capital formation. Treasury budgeted infrastructure spending to rise by 9.7% this year while fixed investment declined in both quarters. The critical question is therefore not the headline allocation but whether government can convert that allocation into a credible pipeline of projects, with execution capacity and terms capable of attracting private capital.
Second is what the wage bill leaves behind. Compensation rises from R808.6 billion to R852.6 billion this year and is largely committed. That narrows the room for government to substitute public spending for weak private investment and makes the composition of expenditure increasingly important.
Third is what government does with an improved fiscal and statistical position. Main budget revenue between April and July was roughly R615 billion and last year’s tax revenue exceeded the estimate by R21.3 billion. If October provides additional fiscal room, how that room is allocated will reveal what government believes the binding constraint on growth actually is.
The MTBPS will therefore be a signal of priorities as much as a fiscal document.
Its timing makes the signal more politically significant. It is scheduled for 21 October, two weeks before the municipal elections.
The political return on measures that provide immediate relief can arrive within weeks. The economic return on infrastructure, institutional reform and productive capacity may take years.
That is the central political-economy test.
From stabilisation to conversion
South Africa’s second-quarter contraction does not show that the recovery is over. It shows that the recovery has not yet been converted into expansion.
The reforms of recent years have restored functions that had broken down. That was necessary. It is not sufficient.
The next phase requires a different political bargain: one capable of giving businesses enough confidence to move from protecting existing output to creating new capacity; aligning public institutions around infrastructure delivery; and sustaining policy beyond individual political cycles.
That will require more than fiscal discipline or cheaper fuel. It will require the state to demonstrate that political stability can produce institutional coordination, and that institutional coordination can produce investable capacity.
This is ultimately the test of the GNU’s economic significance.
Its significance will not be measured only by whether it avoids political breakdown, but by whether it can convert political coexistence into policy durability, policy durability into investment, and investment into employment and productive capacity.
Until that conversion occurs, South Africa risks remaining trapped between recovery and expansion: restoring the capacity it once had without building enough of the capacity it now needs.
The question for October is therefore not simply how government characterises a 0.2% contraction. It is whether, across the next three years, it can demonstrate what it intends to build, and whether businesses believe the political system can sustain the decisions required to build it
