There is something almost perfectly South African about the latest attempt to improve the country’s Special Economic Zones.
The government has identified a rule that makes very little economic sense, acknowledged that it has discouraged investment, and proposed a sensible solution. Great.
To wit: National Treasury and SARS are proposing to change one of the more peculiar conditions attached to the headline attraction of the zones being the preferential 15% corporate tax rate, compared with the normal company rate of 27%, under Section 12R.
This condition was meant to stop tax avoidance. A company could lose the special SEZ tax rate if more than 20% of its deductible spending or income came from transactions with a related South African company or permanent establishment. In other words, if one part of a company is in the SEZ and another is outside, then the company could lose the entire tax benefit. The aim was to stop companies from shifting profits into the SEZ simply to pay less tax.
But the rule was very blunt. It did not matter whether the transactions were genuine or priced fairly. Once a company crossed the 20% threshold, it could lose the tax benefit.
That does not fit the way many modern businesses operate. A manufacturer might have its factory inside an SEZ, while its marketing, purchasing, IT or distribution businesses sit elsewhere. Large multinational companies often operate through networks of related companies that buy and sell goods and services from one another.
Too rigid by far
Treasury now accepts that the rule is too rigid. Its 2026 Budget Review says it can discourage both existing SEZ companies and new investors that want to make an SEZ part of a wider supply chain. The proposed change is therefore much simpler: companies will be allowed to trade freely with related businesses, as long as the prices they charge are similar to what companies not in the SEZ would charge each other.
That is clearly an improvement. It also raises the obvious question: why wasn’t it like this in the first place?
In a way, the episode nicely illustrates the broader problem with South Africa’s SEZ programme: almost every individual component is defensible, but the combination is cumbersome enough to undermine the purpose of having a special zone at all.
South Africa now has 13 designated SEZs where 224 operational investments worth R31.7bn support 28,821 direct jobs. Those numbers are not negligible. Coega, East London, Dube TradePort and particularly the Tshwane Automotive SEZ demonstrate that zones can work when they have a strong underlying economic rationale.
But the numbers are hardly spectacular for a programme that has existed in one form or another since the late 1990s. Between 2016 and 2024, national and provincial government funding and incentives associated with the programme amounted to around R24.2bn, while the zones generated an estimated R14.8bn in corporate and personal income taxes and municipal rates, according to a recent World Bank review.
That’s about 61c of measurable public revenue for every rand spent over the period. Revenues technically exceeded expenditure only in 2024, producing a surplus of R510m.
Not a disaster then. But after decades of policy effort, it is hardly Shenzhen – the astoundingly successful special economic zone that was once a sleepy Chinese fishing village. Thanks to its designation as an SEZ in 1979, it is now China’s number 1 mainland port, home to more than 17 million people, with an economy worth $560bn a year.
An obstacle course dressed up as an incentive scheme
Back in South Africa, of our 13 SEZs, only six have been approved by the finance minister for the preferential company-tax dispensation. The dtic told Parliament this year that only about 30 of the 224 operating companies have benefited from the tax incentives because the qualification requirements are so restrictive.
Pause on that for a moment. The most advertised tax benefit of South Africa’s Special Economic Zone programme applies in fewer than half the zones and, in practice, to a tiny fraction of the companies operating in them.
That is not an incentive programme. It’s more like an obstacle course with a prize at the end. And even after Treasury fixes the 20% rule, the rest of the obstacle course remains.
What’s more, under current legislation the principal section 12R incentive disappears for years of assessment beginning on or after January 1 2031.
Imagine trying to persuade an international manufacturer to build a factory whose economic life might be 20 or 30 years by offering it an attractive tax rate which, under current law, expires in little more than four years?
No wonder the World Bank has recommended extending the 15% rate to all designated SEZs. Trade & industry minister Parks Tau has said the department is considering the recommendation, along with proposals for more private-sector industrial parks, better municipal service agreements and formal intervention in chronically underperforming zones.
All sensible.
But notice the pattern. One department designates the zone. Another must approve it for tax purposes. SARS administers the tax rules. Provinces frequently own or operate the zone. Municipalities supply water, electricity and other services. The dtic supplies incentives and infrastructure money. Development-finance institutions such as the IDC and DBSA may become involved in particular projects. Investors remain subject to ordinary South African labour, environmental and company law and to the broader B-BBEE environment surrounding government incentives and procurement.
A timid approach
This marks an important difference between our SEZs and some of the famous examples internationally.
A Dubai free zone has historically said, in effect: here is a place where foreign investors may enjoy 100% ownership, easy profit repatriation, simplified licensing, customs advantages and, in some circumstances, a substantially different legal and regulatory environment.
China used Shenzhen and later free-trade zones partly as laboratories: places where rules governing foreign investment, ownership, land, markets and trade could deliberately be made different from the rest of the country.
South Africa’s proposition is much more modest.
Our version is: here is a serviced industrial park in which most ordinary South African rules continue to apply, but we’ll try to make some things cheaper and some approvals easier.
There is nothing inherently wrong with that model. But the benefits, surely, have to be exceptionally simple and predictable.
Government has spent years building SEZs, years discovering why investors don’t use particular incentives, and then years adjusting individual regulations, while continuing to designate more zones.
The successful exceptions tell us something important. Tshwane Automotive worked because Ford and its supplier network provided an economic reason for the zone to exist before government started worrying about the theoretical virtues of an SEZ. Coega has a port and an established industrial ecosystem. Dube has an airport and logistics infrastructure. East London has a deep automotive base.
Their lesson is almost embarrassingly simple: an SEZ works best when Government identifies an existing commercial opportunity and removes obstacles around it. It works much less well when government first creates a zone and subsequently goes looking for an economic reason for it to exist.
If the government really wants SEZs to compete internationally, the next stage shouldn’t be another tweak to another subsection of another tax act. It should ask the more fundamental question: what precisely is supposed to be special about doing business there?
Until that answer becomes clearer, South Africa’s SEZs are likely to remain what they have been so far: pockets of success embedded in an industrial policy that remains altogether too lumbering to fulfil much promise.