South Africa Doesn't Need New Small Business Laws. It Needs to Connect the Ones It Has

Registering a business in South Africa is not the problem. Log onto CIPC's BizPortal, reserve a name for R50, file the registration for R125, and a private company is typically issued a registration number within one to five working days. Total cost to get a legal entity into existence: under R230. That single fact should end the conversation about cutting red tape so entrepreneurs can register a business, because there's almost nothing left to cut at that step. Anyone still repeating the line that South Africans can't start companies because the paperwork is too hard hasn't looked at the paperwork in a decade.

I've registered entities through a food service franchise and a ten-location bakery network, and the CIPC step was never where the time went. It went into everything that happens in the weeks after: the separate SARS employer registration, the Compensation Fund return, the municipal process, the bank's own onboarding, each running on its own clock and answerable to nobody but itself.

That's the actual shape of the problem, and it's worth being precise about it before proposing anything, because the wrong diagnosis produces the wrong fix.

The numbers

• CIPC registers a private company for between R125 and R225 online, usually within one to five working days (CIPC BizPortal fee schedule, reviewed 2026).
• Employer registration for PAYE, UIF and SDL must be completed within 21 business days of the first hire, on a process entirely separate from CIPC and SARS income tax registration (SARS employer registration guidance, 2026).
• A 2005 Small Business Project study, still the most comprehensive regulatory cost survey South Africa has produced, put total compliance cost to South African firms at R79 billion in 2004, about 6.5% of GDP.
• The IMF's March 2026 South Africa report found that a one percentage point increase in the time a firm's managers spend on regulatory compliance is associated with roughly a one percent fall in that firm's job growth.
• The compulsory VAT registration threshold and the turnover tax threshold both rose from R1 million to R2.3 million, effective 1 April 2026 (SARS, Budget 2026).
• A business with annual turnover under R10 million can self-certify its B-BBEE status with a sworn affidavit instead of a formal audit (B-BBEE Codes of Good Practice, Exempted Micro Enterprise provisions).

The bit nobody budgets for

A new employer in South Africa faces at least four separate post-registration processes before the business is compliant in the eyes of every institution that governs it, and none of them talk to each other. CIPC issues the income tax number automatically at incorporation, which is the one part that's genuinely joined up. Everything after that is manual: a beneficial ownership filing within 10 business days, a COIDA estimated-earnings return to the Compensation Fund within 7 days of starting to trade, and employer registration for PAYE, UIF and SDL within 21 business days of the first hire, on a Department of Employment and Labour system that has nothing to do with SARS eFiling. VAT joins the queue once turnover crosses the new R2.3 million threshold.

Each of those runs on a different portal and produces a different annual return to track. None of it is complicated in isolation. All of it, together, is a part-time administrative job that a one or two-person business has to either do itself or pay someone else to do, and the fee for outsourcing it usually exceeds what CIPC charged to bring the company into existence in the first place.

Local government adds a second, more arbitrary layer. During a parliamentary committee briefing on government's own red tape reduction programme, MPs heard that the City of Cape Town could charge R5 000 for a municipal business process that cost R2 000 in Swellendam, because each municipality runs its own tariff for what is functionally the same administrative act. This isn't a rounding error. It's a structural reason the same business faces a different bill depending on which side of a municipal boundary it happens to trade on.

The burden also isn't evenly spread. Research tracked through the SME Growth Index found the smallest South African firms, those with fewer than 21 employees, spending around 5% of turnover on regulatory compliance, against 3% for firms with more than 40 employees on the payroll. Compliance cost doesn't scale down with the business. A one-person consultancy files largely the same set of returns as a fifty-person firm, just with far less revenue to absorb the cost of filing them.

A bill that promises to fix this, and won't

South Africa's regulatory complexity has grown steadily and now sits near the top of the range for similar emerging economies, according to the IMF's March 2026 country report, which drew on firm-level data alongside international comparisons. The Fund's economists found that time spent on compliance is crowding out time spent running the business, and put a number on what that costs: firms whose managers spend more time on compliance grow more slowly and are less productive.

Government's own answer to this, currently in front of Parliament, is the Business Licensing Bill. It would require every business in the country, from a spaza shop to a consulting practice, to register on a national database and obtain a municipal licence valid for five years, with municipalities given 30 days to process an application, plus one possible 14-day extension. President Ramaphosa confirmed at SONA that government intends to proceed with it despite sustained opposition.

That opposition isn't coming from one side of the political spectrum. NEASA's submission warned that handing licensing power to national government before municipalities have the capacity to implement it sets the system up to fail from day one. Free SA argued the smallest, least resourced businesses are the ones least able to absorb a new compliance requirement, and warned the Bill could push informal traders out of business rather than into the formal economy. The Helen Suzman Foundation raised constitutional concerns about how much discretion the Bill concentrates in the hands of inspectors, and what that means for property rights and fair administrative process. The Centre for Development and Enterprise went further, pointing to powers that would let officials enter business premises, and in some cases private homes, without a warrant, and called the Bill "the most anti-business bill since apartheid ended." The Western Cape government pointed out that it duplicates processes that already exist while working against government's own Red Tape Reduction Strategy.

The substantive problem is that the Bill adds a new compulsory layer instead of connecting the layers that already exist. It doesn't touch the SARS, CIPC and Compensation Fund fragmentation described above. It adds a municipal licensing requirement on top of that, in the same fragmented municipal system where MPs have already heard that fees for functionally the same registration can differ by more than double from one council to the next. The IMF's own suggestion, in the same report, is far more modest: put every permit and licence that actually applies to a business on one public list that's kept current, and let businesses apply for and track them through a single digital system that works the same way in every province. That's a plumbing fix. What's in front of Parliament is a new pipe with no connection to the old ones.

None of this means administrative and tax burden is the single reason South Africa struggles to formalise small business. The same IMF report ranks it alongside the cost and reliability of electricity, freight logistics and municipal service delivery, not above them. It's a real, comparatively cheap lever to pull. It isn't the only one, and a government that fixed every item in this article tomorrow would still have a small business formation problem, just a considerably smaller one.

South Africa already built the parts

The country doesn't lack the institutional imagination to design something simple. It has already done it more than once. It has just never connected the pieces to each other, and some of them have taken far too long to work properly even on their own.

Turnover tax has existed since 2009 and does exactly what a good simplification should: it replaces income tax, VAT, provisional tax, capital gains tax and dividends tax with a single calculation on gross turnover, for any business under the qualifying threshold, now R2.3 million. For most of its history, registering for it required an in-person SARS appointment, because eFiling didn't support it. That only changed this year, when SARS finally opened registration through its Online Query System. A genuinely good idea sat half-finished for the better part of fifteen years because nobody prioritised the last step of building it properly.

The B-BBEE Exempted Micro Enterprise provision is the cleaner example. Any business under R10 million in annual turnover, or in its first year of trading regardless of turnover, can confirm its B-BBEE status with a sworn affidavit rather than a formal verification audit. No auditor, no scorecard. It works well enough that almost nobody argues about it, which is itself worth noticing given how contested the rest of the B-BBEE framework is.

There's also a quieter cost sitting in plain sight. A qualifying Small Business Corporation pays 0% on its first R99 000 of taxable income, rising in stages toward the standard 27% rate. It's an opt-in system, which means a business that would qualify for materially lower tax has to know to ask for it. Awareness gaps in an opt-in system function the same way friction does: some money that was meant to reach small businesses simply never gets claimed.

The pattern across all three is the same. South Africa builds decent simplification tools, then leaves them stranded. Some sit half-finished for years. Others get finished but never connected to the rest of the compliance stack, or work fine on paper and simply never get promoted enough for the businesses that qualify to use them. The fix in most cases isn't a new law. It's finishing what's already legislated.

What Rwanda, Brazil and India did differently

Estonia and Singapore get cited in almost every article like this one, and I'd leave them out on purpose. They're small, wealthy states starting from a completely different administrative base. Rwanda, Brazil and India are genuinely comparable: developing economies fragmented across municipal or state lines, working from a weaker administrative starting point than a rich country's civil service.

Rwanda went from nine separate procedures and 16 days to register a business in 2008 to a single point of contact, the Rwanda Development Board, capable of issuing a certificate of incorporation in as little as six hours, at no cost. It's now consolidating additional services, building permits and agro-processing licences among them, onto the same digital platform, specifically to reduce the institutional fragmentation that used to force a founder to deal with a different office for every permit.

Brazil's MEI programme replaced a formalisation process that used to cost roughly four times the minimum wage and take about 120 days with free, same-day online registration through a single portal, bundled into one fixed monthly payment covering tax and social security. A study of the programme found it increased the number of active formal firms by 86% in the industries it covered. Brazil also offers the clearest cautionary tale on thresholds: the MEI revenue ceiling was frozen at R$81 000 from 2018, and government is only now legislating a staged increase to R$110 000 in 2027 and R$140 000 in 2028 to correct for inflation it let quietly erode the programme's reach for eight years.

India's Udyam registration is the most instructive design for the specific question this article set out to answer. It's free and entirely online, built around an Aadhaar identity check and a self-declared turnover figure, and where the business is GST-registered, that turnover figure gets pulled automatically from its existing tax data rather than checked through uploaded documents reviewed by an official. The applicant declares a number. The system checks it against a number the tax authority already has. There's no discretion in that loop for anyone to lean on, and almost no paperwork for the business to produce.

The proposal

A useful reform here doesn't require inventing anything. It requires finishing and connecting what already exists, in ways that constrain government's own behaviour rather than creating a new door for anyone to walk through with a false claim. That second part matters as much as the first. Every one of the five ideas below was chosen because there's nothing in it for a business to fake.

1. Connect the databases that already exist

BizPortal already bundles some registrations at incorporation. Extend it so PAYE, UIF and COIDA trigger automatically off a company's existing CIPC and SARS identifiers the moment a first employee is added on eFiling, instead of opening a fresh 21-day manual process on a system that has nothing to do with either. This isn't a new benefit to qualify for. It's plumbing between databases the state already owns, which means there's no new claim anyone could falsify to get in.

2. Index the thresholds instead of legislating them

Turnover tax, VAT and Small Business Corporation bands should move by a published CPI-linked formula every year, not sit fixed until a Budget Speech intervenes. South Africa's own R1 million threshold sat still long enough to quietly shrink the pool of businesses it was meant to serve, before this year's jump to R2.3 million corrected it in one move. Brazil made the identical mistake and froze its MEI ceiling from 2018 to 2026. A formula removes the multi-year lag and the political discretion, and there's nothing to game in it. Turnover is turnover, whatever the threshold happens to be set to that year.

3. Cap municipal fees and force a decision, don't add a licence

Rather than the Business Licensing Bill's compulsory new municipal layer, publish one national ceiling on what a municipality may charge for existing registration categories, and legislate that silence past a fixed number of days counts as approval. This binds the state's own behaviour instead of creating a new permission a business could misrepresent to obtain, and a rule that constrains an official's discretion is a considerably harder thing to corrupt than a rule that creates a new category of thing to apply for.

4. Let the EME model carry more weight

The self-certifying B-BBEE affidavit already works and isn't controversial. Apply the same logic, self-declared status checked against data SARS already holds, to qualification for turnover tax and Small Business Corporation rates, so a business doesn't lose out on a concession it already qualifies for simply because nobody flagged it. It's no more exploitable than the tax return the declared number is drawn from, and SARS's existing anti-avoidance rules against income-splitting between connected parties would carry across unchanged.

5. Sunset every new small-business compliance rule automatically

Three years, then it lapses, unless the department that introduced it can show, publicly, that it delivered a demonstrable net benefit. This is the direct answer to the IMF's finding that regulatory complexity in South Africa only ever grows. It constrains government departments, not entrepreneurs, so there's nothing here a business owner could manipulate in either direction.

What this means if you're the one starting the business

None of the above changes what you need to budget for right now, so plan on the system as it actually is, not as it should be. Set aside the outsourced-compliance fee alongside the CIPC fee from day one, because the first is where the real cost sits. If you're planning to hire in year one, build the 21-day PAYE, UIF and SDL registration window into your hiring timeline rather than assuming it happens automatically alongside incorporation. Check whether you qualify for Small Business Corporation rates or turnover tax before you default to the standard 27%, because that check is entirely on you under the current system. And if a municipal licence or trading permit applies to what you're doing, get the fee and the timeline in writing before you commit to a lease, because as things stand it can differ significantly depending on which municipality you're registering in.

Of the five ideas above, the ones likeliest to move first are the ones that don't need new legislation. Indexing the tax thresholds and extending the EME logic to turnover tax qualification are both administrative decisions National Treasury and SARS could make without a bill going anywhere near Parliament. Connecting the registration databases and capping municipal fees need cross-department coordination, which moves slower in South Africa than almost anything else. The sunset clause needs actual legislative appetite for government to bind its own future decisions, which is the hardest sell of the five precisely because it's the one that would work best.

Preparation has never depended on government fixing this first. It depends on knowing exactly what you're walking into before you do it, which is true whether the Business Licensing Bill passes next year or dies in committee.

Frequently asked questions

How much does it cost to register a company in South Africa in 2026?

CIPC charges R125 for a standard private company registration through BizPortal, or R175 if you reserve a name first (an additional R50 for up to six months). Total cost is typically under R230, with a certificate issued within one to five working days, according to CIPC's published 2026 fee schedule.

What is the compulsory VAT registration threshold in South Africa?

The compulsory VAT registration threshold rose from R1 million to R2.3 million, effective 1 April 2026, following the 2026 Budget Speech. The voluntary registration threshold rose at the same time, from R50 000 to R120 000, according to SARS's Budget 2026 update.

What is the Business Licensing Bill?

The Business Licensing Bill is draft legislation, published for comment in September 2025, that would require every South African business to register on a national database and hold a municipal licence, processed within 30 days plus a possible 14-day extension. It has drawn opposition from NEASA, Free SA, the Helen Suzman Foundation, the Centre for Development and Enterprise, and the Western Cape government, largely over expanded inspector powers and uneven municipal capacity.

What turnover qualifies a South African business as a B-BBEE Exempted Micro Enterprise?

A business with annual turnover of R10 million or less, or in its first year of trading regardless of turnover, qualifies as an Exempted Micro Enterprise under the B-BBEE Codes of Good Practice. It can confirm this status with a sworn affidavit rather than a formal verification audit.