Last week, the National Treasury proposed an additional 20% levy on gross gambling revenue (GGR), in addition to the provincial taxes already in place.

Treasury’s discussion paper sets out eight reasons why this is justified and overdue, including ensuring those who cause the harm contribute towards the social costs, curbing excessive gambling, aligning with international norms, preventing harmful provincial tax competition, and bringing coherence to an industry that has expanded far beyond its current regulatory framework.

Treasury is right on one point: provincial oversight designed in the apartheid era for destination casinos like Sun City cannot cope with a borderless digital industry that has exploded over the past five years (36% annualised growth).

Many countries, including the UK, Australia and parts of the EU, treat online gambling as a national regulatory and tax function. South Africa is an outlier. Aligning with global practice makes sense, as does levelling the playing field between tightly regulated, heavily taxed casinos and online operators that have enjoyed looser conditions while capturing an ever-larger share of the market (60% of GGR).

Where Treasury’s logic collapses is in its claim that the tax will discourage problem gambling. It misunderstands both how gambling addiction works and how sin taxes function. Unlike excise duties on cigarettes, tobacco and sugary drinks, which visibly increase prices and measurably depress consumption, a gambling tax will be imperceptible to punters.

The flawed analogy with sin taxes

Treasury’s central rationale is to treat gambling like other harmful products: tax it heavily so that the price rises and the harmful behaviour declines. In theory, the levy helps internalise negative externalities such as financial distress, mental health problems, family breakdown and lost productivity (in other words: making gambling companies pay for the real harm addiction causes – rather than shifting those costs onto families, communities and the state).

The premise is simple: raise the price of cigarettes, and smoking falls; raise the price of sugary drinks, and consumption declines. The evidence is clear.

After South Africa introduced a sugary-drinks levy, retail prices increased by roughly 10% and purchases dropped 6%-10% over one to two years, with the steepest reduction among lower-income consumers (about 10%). The intervention was visible and immediate at the point of purchase, and behaviour changed accordingly.

Gambling does not work like this. Increasing the tax on GGR does not raise the upfront cost of a bet. Punters will not see an extra charge added to their wagers. Instead, the tax lifts the “house edge” – the margin the operator earns on bets.

But the house edge is invisible. Behavioural biases such as social proof, present bias and intermittent reinforcement mean players do not perceive small changes in odds. People do not stop gambling because the implied margin worsens slightly; in most cases, they do not even notice it.

In contrast to sin taxes on cigarettes or sugary drinks, a tax on GGR will not change gamblers’ behaviour. Punters will continue to bet, operators will continue to operate, and the only immediate effect will be a tax windfall for the state.

Without reform, the tax becomes a revenue tool – not a harm-reduction tool.

A regressive tax on addiction

Treasury argues that the tax internalises social costs. But without a functioning national regulatory regime for online gambling, that argument collapses. South Africa is stuck with an outdated structure in which provinces license interactive gambling, each with different enforcement and oversight standards.

Taxing a system we do not adequately regulate and expecting the tax to compensate for regulatory failures is not policy – it is wishful thinking.

There is another uncomfortable dynamic. The tax is likely to raise more than R10bn a year. That revenue does not come from recreational gamblers who wager modestly and sporadically. It comes disproportionately from problem gamblers – roughly 31% of gamblers generate the majority of turnover.

If the tax fails to reduce problem gambling but efficiently captures revenue from it, we are left with something profoundly regressive: the state becomes financially dependent on harmful gambling behaviour continuing at scale.

Other jurisdictions avoid this trap by pairing national taxation with robust regulation: affordability checks, advertising limits, deposit caps, identity verification, exclusion registers and mandatory harm-reduction contributions. South Africa has none of these in a unified, enforceable form.

Provincial fragmentation remains the real problem

Treasury is correct that provincial gambling taxes create competitive distortions. Provinces can set lower rates to attract operators and collect revenue, even as those operators serve punters nationwide. Gambling conducted nationally through digital channels should be regulated at the national level.

Replacing fragmented provincial taxes with an additional national levy, while leaving fragmentation untouched, risks worsening the problem. Provinces may adjust or extend levies to protect revenue. Operators could be squeezed between conflicting regimes. Consumers will continue gambling in an environment where nobody is accountable for their protection.

To borrow a proverb: when elephants fight, it’s the grass that suffers. Here, the elephants are the state, provincial gambling boards and operators, locked in a contest over jurisdiction, revenue and control. South Africans – especially vulnerable gamblers – will bear the consequences.

A poorly designed tax, in the absence of effective national regulation, risks:

Pushing more gamblers towards illegal offshore platforms;
Increasing the financial burden on problem gamblers;
Encouraging operators to find loopholes or relocate servers abroad; and
Worsening an already-broken enforcement landscape.

Rather than reducing harm, the tax may increase it. South Africa needs national intervention in online gambling. But sequencing matters.

A coherent national regulatory framework must come first, one that legalises, licences, monitors and enforces online gambling with clear consumer protections. Once such a framework exists, a national tax can be layered on top, aligned with international best practice and calibrated to discourage harmful play.

Until then, the proposed levy risks becoming a politically attractive but substantively weak intervention – one that raises revenue but does little else. If we are serious about protecting citizens from the risks of online gambling, we cannot start with a tax. We must start with the law.

Thomas Brennan is a co-founder of Franc, a South African fintech that helps people invest simply and affordably.