US Ambassador Leo Brent Bozell has put a hard number on Pretoria’s land-reform gamble: billions of dollars in American public money now sit in limbo, routed through agencies that answer to Congress and to taxpayers who do not forgive losses easily. The Export-Import Bank and the Development Finance Corporation, Washington’s two largest vehicles for steering capital into frontier markets, have both pulled back. Their hesitation is mechanical, not ideological. These institutions carry explicit mandates to protect US government funds. The Expropriation Act’s “nil” compensation provisions have triggered a risk assessment that South Africa has not yet been able to satisfy.
Bozell delivered the message plainly in a recent interview with 702. American firms face no statutory barrier to investing here. The deterrent is predictive: capital goes where it can price risk, and nil compensation is, by definition, unpriceable. A factory, a data centre, a logistics hub: these assets have value because they can be collateralised, insured, resold. The Act introduces a contingency where that value may be judicially erased. For a development finance institution, that contingency is a fiduciary breach.
What the Act actually permits
The Expropriation Act of 2020 does not authorise blanket confiscation. Its framers were careful, perhaps too careful, to nest the legislation within constitutional language. Section 25 of the Constitution permits expropriation for public purpose or public interest, provided compensation is “just and equitable.” The Act attempts to specify when nil compensation meets that standard.
The conditions are narrow but consequential. Unused land held for speculation qualifies. Abandoned property qualifies. Certain parcels owned by state enterprises but not deployed for public use qualify. Most critically for investors, land whose market value equals or falls below direct state investment or subsidies received by the owner qualifies. The logic is redistributive arithmetic: where the public has already paid for the asset, further compensation would constitute double payment.
Pretoria’s defence rests on judicial process. The government argues that “just and equitable” is not synonymous with market value, that the Constitution explicitly contemplates court-determined compensation, and that nil awards would occur only after judicial scrutiny. Officials insist this is not a general confiscation policy; it is a framework for accelerating land reform within constitutional bounds.
The distinction matters legally. Whether it matters practically to a credit committee in Washington is a different question.
The Washington waiting game
Bozell was explicit about the sequencing. The US will not reconsider its investment posture until South African courts rule on the Act’s constitutional consistency. This is institutional design. The EXIM Bank and DFC both operate under statutes that require prudent stewardship of public funds. Their boards cannot approve exposure to jurisdictions where sovereign legislation may extinguish collateral value without compensating the lender.
The EXIM Bank, which insures and finances the purchase of American exports by foreign buyers, has historically supported South African infrastructure and energy projects. The DFC, created from the merger of the Overseas Private Investment Corporation and other development finance tools, provides equity, debt, and political risk insurance for private ventures in emerging markets. Together they represent the formal channel through which US government capital reaches frontier economies. Their withdrawal reduces project finance availability and signals to private American capital that the sovereign risk premium for South Africa has shifted.
Bozell’s “billions” figure lacks granular breakdown, but the order of magnitude is consistent with DFC’s announced Africa portfolio ambitions and EXIM’s historical exposure to the region. What is being withheld is not charity. It is purchase orders, equipment leases, and power purchase agreements that would have employed American suppliers and South African operators alike.
The five demands and the broader rupture
The expropriation dispute does not exist in isolation. The Trump administration has framed it as one of five specific demands in a bilateral relationship that has deteriorated across multiple vectors. The other four reveal the breadth of the friction: violence against rural communities, which Washington has characterised in terms that echo its human rights reporting; political rhetoric from South African officials that US diplomats have flagged as inflammatory; Black Economic Empowerment requirements that American firms argue impose discriminatory conditions on foreign investors; and Pretoria’s posture toward a proposed US programme for Afrikaner migration.
This is not a trade negotiation with a discrete tariff schedule. It is a comprehensive reassessment of alignment. The demands span property rights, physical security, speech, corporate governance, and migration policy. Addressing one does not unlock the others. The architecture suggests that Washington views the relationship as requiring systemic correction, not transactional settlement.
For South African policymakers, the challenge is structural. Land reform enjoys broad domestic legitimacy as a response to apartheid-era dispossession. The ANC’s electoral position, while diminished, still depends on delivering visible redistribution. Yet the instruments chosen to advance that goal have triggered a capital strike from the country’s second-largest bilateral trading partner. The government is being asked to reconcile two constituencies with incompatible timelines: voters who have waited decades for land restitution, and investors who price assets on quarterly risk assessments.
How to read the legal uncertainty
The constitutional question at the heart of the standoff has not yet produced a definitive judicial ruling. This is the vacuum Bozell identified. South African courts have extensive experience with property rights cases under Section 25, but the Act’s specific nil-compensation provisions have not been tested against the Constitution’s full text. The government maintains that the legislation merely operationalises existing constitutional authority. Critics, including American legal analysts advising the embassy, argue that the Act’s conditions for nil compensation may exceed what the Constitution’s framers contemplated.
The litigation timeline is not favourable to investment planning. South African constitutional litigation can extend across multiple years, through High Court, Supreme Court of Appeal, and potentially Constitutional Court review. Each stage introduces further uncertainty. A development finance institution cannot commit to a five-year infrastructure project while the legal status of its security interest remains subject to judicial revision.
Pretoria’s argument that courts will ultimately safeguard property rights depends on investors trusting that process. For institutional capital with alternative destinations such as Vietnam, India, Kenya, and Egypt, that trust is a discretionary premium, not a requirement.
What is actually known, and how to check
The Expropriation Act became law in 2020, replacing the 1975 Expropriation Act inherited from the apartheid era. Its full text is published in Government Gazette No. 43056, accessible through the Government Printing Works. The constitutional provisions it implements are Sections 25 and 33, which address property rights and just administrative action respectively.
The EXIM Bank’s South Africa exposure is reported in its annual competitiveness reports to Congress, available at exim.gov. The DFC’s active projects are listed in its disclosure database at dfc.gov. Neither agency has issued a formal suspension notice specific to South Africa. The withdrawal Bozell describes appears to operate through pipeline freezes and non-approval of new transactions rather than announced policy change.
No court case challenging the Act’s constitutionality has yet reached judgment. The government’s defence of the legislation is summarised in Department of Public Works briefings to Parliament, recorded in the National Assembly’s portfolio committee minutes. For investors, the absence of a ruling is itself the active variable: risk models must assign probability to outcomes that remain legally indeterminate.
The capital map redraws
South Africa’s land reform imperative is not negotiable in domestic political terms. The country’s Gini coefficient remains among the world’s highest, and land ownership concentration tracks racial lines established under colonial and apartheid policy. The Expropriation Act was crafted to accelerate redistribution without abandoning constitutional order.
The cost of that craft is now measurable in withheld dollars. Not private portfolio flows, which can reverse overnight, but long-dated institutional commitments that build physical capacity. The EXIM Bank and DFC do not speculate. They finance transformers, transmission lines, water treatment, and logistics corridors that private capital will not fund alone. Their absence reshapes the country’s infrastructure frontier.
Bozell’s warning is not a threat. It is a description of how sovereign risk is priced when legal frameworks introduce unquantifiable contingencies. South Africa can await its judicial process. Washington can too. The capital, meanwhile, goes elsewhere.
