AGOA Extended to 2028 — But SA's Car Exporters Get No Relief

On Wednesday, 2 September, President Donald Trump signed the Continuing Appropriations and Extensions Act, 2027, and buried inside a routine spending bill was a clause that quietly carried the African Growth and Opportunity Act from its December 2026 cliff-edge to 31 December 2028. Pretoria had asked for a fifteen-year renewal to anchor factory investment; it got two. And for the one sector that made AGOA matter most to South Africa — the assembly lines of the Eastern Cape and KwaZulu-Natal — even those two years are worth almost nothing.

Two years, not fifteen, and only one line changed

The extension is narrow by design. It moves AGOA's expiry date and does nothing else: no new products, no reset of eligibility, no shelter from the tariff walls the United States has built since 2025. South Africa is a heavy user of the programme — it accounted for roughly half of the $8.23 billion in goods shipped duty-free under AGOA in 2024 — so keeping that door open is not trivial for citrus, wine, macadamias and ferro-alloys, which still cross into the US market without a customs bill.

But the headline "billions in trade retained" masks where the value actually sat. For a quarter of a century AGOA's most bankable benefit for South Africa was the 2.5% duty saving on passenger cars, which underwrote a vehicle-export corridor to the US that grew from 853 units in 2000 to 24,682 in 2024. That corridor is now effectively shut — and nothing signed this week reopens it.

Section 232 overrides the saving that mattered

The reason is a separate instrument that AGOA cannot touch. Under Section 232 of the US Trade Expansion Act — a national-security provision — Washington imposed a 25% tariff on imported vehicles from 3 April 2025 and on automotive components from 3 May 2025. That 25% is levied regardless of AGOA status, so a South African-built car now lands in the US carrying a duty ten times larger than the 2.5% AGOA once waived. The preference still exists on paper; the tariff simply swamps it.

The figures translate the policy into an invoice. South African vehicle exports to the US collapsed by 83.2% in a single year, from 24,682 units in 2024 to 4,136 in 2025, and the first half of 2026 was worse still — 1,840 units, a further 36% down. In rand terms, vehicle-export earnings to the US fell to about R8 billion in 2025 from R17.7 billion the year before. This is not a market softening; it is a channel closing. The Automotive Business Council, naamsa, whose members account for 23.8% of South Africa's manufacturing output, put it plainly through interim chief executive Shinny Gobiyeza: "Section 232 continues to constrain that opportunity. Our priority must therefore be to secure a durable and mutually beneficial trade arrangement with the United States."

What it means for the exporter's landed cost

For any South African exporter shipping to the US, the practical lesson is to stop pricing off AGOA and start pricing off the tariff schedule that actually applies to the product. An agricultural or minerals exporter whose goods fall outside Section 232 keeps genuine duty-free access — the two-year extension is real money in that account, and it is worth confirming your product line and rules of origin qualify before assuming it. An automotive, steel or aluminium exporter should model the 25% Section 232 duty into every landed-cost quote to a US buyer, because that is the number the importer of record will pay at the border, and it does not expire in 2028 with AGOA.

There is a second tariff layer to watch. A 12.5% duty imposed under Section 301 in July 2026 sits over a range of South African goods, though products already caught by Section 232 are exempt from it — so a car pays the 25%, not both. The broader "reciprocal" tariff regime that began at 30% on South African exports in August 2025 is still working through the US courts and trade negotiations, which is precisely why a fixed landed-cost assumption is dangerous this quarter. Exporters carrying ad valorem duty exposure on US-bound cargo should stress-test their margins against more than one rate.

The predictability problem the headline hides

The optimists reading "AGOA to 2028" as stability are missing what the number is for. AGOA's value was never only the duty saved on any single shipment; it was the multi-year certainty that justified building an assembly plant in Gqeberha to serve the US rather than in Mexico or Morocco. A fifteen-year renewal buys that kind of capital decision. A two-year window does not — it is shorter than the lead time to commission a new production line, so it cannot underwrite one.

Worse, the clock is already ticking against even those two years. The US launched its 2027 AGOA eligibility review in June 2026, with determinations that take effect on 1 January 2027 — meaning a country can be dropped from the programme a full year before the 2028 expiry the extension just granted. South Africa's continued eligibility has been openly questioned in Washington on both political and labour grounds. So the honest description of this week's signing is not "two more years of access" but "up to two more years, subject to a review that reports in weeks."

Our Take

This extension is a reprieve worth having for agriculture and minerals and a mirage for the auto sector, and the two should not be conflated in a single cheer. The R8 billion vehicle corridor that AGOA built is not coming back on the strength of a date change while Section 232 stands, and no amount of duty-free citrus offsets losing the country's single most valuable manufactured export line to the US. Three concrete moves follow. First, exporters should segment their US business by tariff instrument, not by AGOA status — Section 232 goods (autos, steel, aluminium) need a 25%-loaded quote today, while genuinely AGOA-covered goods should be shipped hard through the 2028 window while it is open. Second, automotive suppliers should treat the diversification naamsa is signalling as operational, not aspirational: the AfCFTA continent and the EU market are where the lost US volume has to be recovered, and that repositioning takes the same two years the extension bought. Third, watch the January 2027 eligibility determination more closely than the 2028 expiry — it is the earlier, sharper risk, and the one that could turn this modest win into no win at all. Predictability was the whole point of asking for fifteen years. Two years, reviewed in weeks, is not it.