A national online gambling tax without a national online gambling law

On 25 November, National Treasury’s discussion paper on a national gambling tax dropped without warning. It reads, on the surface, like a neat solution.

Take a sector that is growing fast, especially online betting, say that it creates social harm, and announce a new 20 percent national tax on gross gambling revenue from online betting and interactive gambling on top of what provinces already collect.

The headline number, more than R10 billion a year in new revenue, is designed to land well with a fiscally stressed public.

Scratch the surface and something else appears. South Africa is being asked to support a national online gambling tax at a moment when we still do not have a coherent, implemented national online gambling law.

At the same time, the paper quietly proposes a central national tax in a space where the Constitution deliberately put provinces at the centre and where provincial boards have carried the regulatory burden for more than two decades.

I have spent the better part of twenty-four years working in this sector and, as I argued during the Supabets roulette case debates, our starting point should be respect for the constitutional and institutional regime we actually have, not the one some in Pretoria wish we had.

We built a provincial model for a reason. If we want to change it, we should do that openly in legislation via constitutional amendment, not through the back door of a tax proposal.

Taxing what is still illegal

Interactive gambling remains unlawful in South Africa. Parliament passed the 2008 Amendment Act to create a national framework for interactive gambling, but that Act has never been brought into force. On the books today, online casinos are still prohibited. Only online betting, through provincially licensed bookmakers, is expressly accommodated, including on approved contingencies which include casino-style games.

Treasury’s own paper acknowledges that interactive gambling is illegal, then, in the next breath, proposes that the new national tax base will include revenue from it. The message to the illegal offshore market is effectively this: you may carry on offering products that the statute still treats as unlawful, provided that you calculate your gross gambling revenue carefully and pay a new national tax.

Worse yet, and absurdly, it intimates that, while they do not recognise the legitimacy of online betting on contingency games, licensed bookmakers in the provinces should also pay this new tax on top of their provincial obligations.

That is an inversion of basic rule of law logic.

You do not fix fifteen years of failure to implement interactive gambling legislation by quietly taxing the very conduct that Parliament still defines as prohibited. You fix it by deciding, in public, whether South Africa wants a regulated interactive market at all and, if so, on what terms.

The constitutional place of provinces

After the Supabets judgment dropped from the Supreme Court of Appeals, I argued, and still believe, that you cannot wish away the constitutional structure of gambling regulation.

Gambling, apart from the National Lottery, was allocated as a concurrent competence. Provinces were given real teeth, not symbolic roles.

Provincial boards license casinos, bookmakers, bingo halls and limited payout machine operators, they audit compliance, they carry the political heat when communities complain about harm, and they rely heavily on gambling taxes to fund local priorities.

Those provincial levies sit mainly on gross gambling revenue. Betting is already taxed by provinces, typically around six percent of gross gambling revenue, with more complex structures for casinos and other forms of gambling.

Treasury now proposes a national tax, also on gross gambling revenue, that would sit above these provincial rates. In practical terms the combined burden for online betting would move into the high twenties, even before income tax is taken into account.

This is presented as harmless alignment and as a way to tidy up online activity that does not sit neatly inside provincial borders. In substance it is an unacknowledged rebalancing of fiscal power away from the provinces and toward the centre, in a field where the provinces created and maintain most of the regulatory infrastructure.

If national government wants to redesign that settlement, the honest route is to revisit the National Gambling Act and the 2008 Amendment or, more sensibly, start again and finish the job, not to squeeze the same tax base from above.

My own view is clear.

The provincial model has served South Africa reasonably well, despite its flaws. Provincial boards are closer to the ground, closer to affected communities and closer to municipal planning. A tax proposal that sidelines their role and dilutes their fiscal autonomy should worry anyone who cares about cooperative government.

Externalities or just revenue

The paper leans heavily on the language of negative externalities. Problem gambling, mental health issues, suicidality, over indebtedness, family conflict.

Those harms are real and they deserve to be taken seriously.

Treasury then treats a 20 percent national tax on gross gambling revenue as an instrument to internalise these externalities.

There are three obvious difficulties.

First, the paper does not attempt to quantify the social cost of problem gambling in South Africa. There is no modelling of how many people are affected, what the fiscal cost of treatment and lost productivity might be, or how these costs compare to current gambling tax receipts.

The 20 percent figure does not emerge from a careful costing exercise. It is drawn from international comparisons, and only of tax rates and not necessarily feasibility, and then layered onto the existing provincial structure.

Second, there is no ring fencing at all. The more than R10 billion in new revenue is destined for the general pot.

There is no commitment that any portion of the national gambling tax will fund research, treatment, counselling, public education or enforcement capacity.

The very communities that bear the cost of gambling harm are not guaranteed any targeted benefit from the tax that is supposedly levied on their behalf.

Third, the economic incidence of the tax will not fall only on operators. In practice, higher taxes on gross gambling revenue are bound to be pushed through the value chain to players, through lower odds, fewer promotions and, in some cases, reduced investment in tools that actually mitigate harm.

A measure that is sold as principled externality pricing starts to look much more like a politically convenient consumption tax that leans on a sector which is easy to moralise about.

If we are going to invoke externalities, then we should do the work properly. That means a serious socio economic impact assessment, an honest attempt to cost gambling related harm, and a statutory mechanism that directs at least part of any new national tax directly into harm reduction.

Turning illegal operators into taxpayers

One sentence in the paper deserves to be quoted often.

Treasury states that local suppliers of online betting will be required to register with SARS and provide data similar to what they already provide to provincial gambling boards.

It then adds that, where these local industry players are involved in interactive gambling, even though that is still purportedly illegal, ignoring the provincial competence, they will be taxed on the gross gambling revenue from those activities as well.

In other words, we are officially inviting illegal offshore and unlicensed interactive operators into the tax net without first deciding whether they should be operating at all. Once you cross that line, the distinction between legal and illegal supply is badly blurred.

SARS will have a direct revenue interest in the continuation of activity that the National Gambling Act still treats as unlawful. Regulators will have to explain why actual offshore illegal operators who are under investigation on the regulatory side are carrying national tax clearance certificates on the fiscal side.

There are also real administrative questions. How does SARS audit interactive gambling revenue from an operator whose servers, or payment rails, or wallet infrastructure might sit partly offshore. How is the national tax to be separated from the provincial base when single wallets service multiple products across multiple provinces.

Treasury waves these questions away by pointing to provincial data flows. Anyone who has actually worked with those data flows, and tried to reconcile them across boards, knows it will not be that simple.

The offshore risk and channelisation

International experience matters here. Where governments have set very high headline tax rates on gambling without a realistic view of enforcement, players have simply migrated to offshore sites. Kenya is the most commonly cited African example, but the pattern repeats elsewhere. High rates imposed on a partially legal, poorly enforced market lead to a shrinking domestic tax base and a thriving parallel market that is almost impossible to police.

In South Africa, interactive casinos are already technically unlawful and enforcement against offshore operators is limited.

Provincial boards and the National Gambling Board rely on a mix of blocking, public warnings and occasional seizures of winnings.

In that setting, loading a near thirty percent combined tax burden onto licensed operators that are actually in the system is more likely to accelerate channel shift than to discourage gambling.

If the priority is player protection, then the policy sequence should be clear. First, create a legal, nationally coherent framework for interactive gambling that includes and does not sideline the provinces.

Second, make sure that locally licensed products are competitive enough, and safe enough, that the majority of South African players choose them instead of offshore options. Only once that channelisation objective is firmly in sight does it make sense to increase fiscal pressure.

What a better path could look like

It is not hard to sketch an alternative that respects the provincial model, honours the Constitution and still allows for a coherent national tax.

Start by resolving, in law, the status of interactive gambling. Government needs to decide whether to implement and modernise the 2008 Amendment framework or to replace it. Either way, the country needs a clear national set of rules for interactive products, a sensible division of licensing responsibilities between the National Gambling Board and provincial boards, and harmonised technical and responsible gambling standards.

Then, on that foundation, design a fiscal framework. If there is to be a national gambling tax on gross gambling revenue, its relationship with provincial levies should be transparent and agreed, not imposed from above. Rates should be set with explicit reference to both revenue and harm reduction, and should be reviewed periodically in light of evidence, not political pressure.

Finally, any national tax that is justified by reference to social harm should be partially earmarked for addressing that harm.

A properly allocated national responsible gambling fund, with clear governance and reporting, would do far more for credibility than a generic promise that extra money in the pot helps everyone.

Until we do that work, a national online gambling tax without a national online gambling law is not sound policy. It is a shortcut around the hard questions, and it risks undermining a provincial system that, with all its imperfections, has held the line for almost thirty years.

Wayne Lurie, Director, Lurie Inc Attorneys and head of SAROGA (South African Responsible Online Gambling Association)

[Wayne Lurie has practised gambling regulatory law exclusively for almost 24 years. He has been involved in the licensing of gambling activities, the interpretation of gambling regulations and landmark litigation both in South Africa and abroad, and has advised at every level of the gambling spectrum from multinational corporations to regulators.]

BUSINESS REPORT