Joildo Santos' article gets the data right, but skips who will pay for treatment

The data in Joildo Santos' UOL column largely checks out, but the key omission is fiscal: after MP 1.394, the betting revenue that was supposed to fund health was removed, and the treatment bill ends up entirely with the Treasury. That is the part high-risk operators, PSPs, and banking partners should care about, because it changes where the cost of problem gambling lands.

1. LENAD III: the numbers are correct, but the context matters. The column says that among bettors with income up to one minimum wage, 52.8% had some level of risky or problem gambling, versus 21.1% among those earning more than two minimum wages. That is consistent with the Unifesp survey. The missing detail is that the study was carried out between 2023 and 2024, not from 2022, even though the reference minimum wage in the research was R$ 1,212.00, the amount in force in 2022.
2. The survey reflects an earlier market, before regulation took effect. The data were collected before 1 January 2025, when the rules of Law 14.790/2023 entered into force, including identity verification, centralized self-exclusion, and limits. The underlying point is important: the vulnerability measured by LENAD III comes from an unregulated market, which is exactly the environment recreated by the illegal market after MP 1.394.
3. "Bettors" in the study are not just sports bettors. The sample covers all gambling formats. Among respondents who gamble, 71.3% play Mega-Sena, Quina, Lotomania and similar products, while 32.1% prefer bets and 28.9% play jogo do bicho. So the income split applies to the wider gambling universe, including Caixa lotteries, which MP 1.394 kept.
4. "Risk" includes low risk, not only addiction. In the population over 14, the survey found 3.4% of players with low risk, 2.6% with moderate risk, and 0.8% of problem gamblers. The column does note that the results include different degrees of risk, but readers can still easily read 52.8% as addiction prevalence. It is not measuring that.
5. IEPS: the cost estimate is not a SUS budget line. The column says at least R$ 30.6 billion a year in social costs linked to health damage, based on British references. That figure is correct, and the caveat is also correct: it is not an expense already booked in the SUS budget. The missing step is political, not statistical — after MP 1.394 removed the betting revenue earmarked for health, the treatment cost is left with the Treasury.

For high-risk payment businesses, the operational takeaway is simple: once earmarked betting revenue for health disappears, the fiscal burden does not disappear with it. It shifts to the public balance sheet, while the underlying exposure from gambling harm remains where the survey found it.