BNLData fact-check finds UOL betting column data accurate but incomplete

A BNLData fact-check of a UOL column by Joildo Santos, published on 19 October 2026, found that most of the figures cited in the column checked out, but that it omitted a central point: MP 1.394 removed the share of betting revenue that had been earmarked for health, leaving the cost of treatment with the Treasury. The column was titled “Bets: quem nunca apostou também paga a conta”.

On the LENAD III survey, BNLData said the column's figures were accurate but lacked context. The Unifesp survey recorded a prevalence of 52.8% for risky or problem gambling among those earning up to one minimum wage, against 21.1% among those earning two minimum wages or more. BNLData said the survey was carried out between 2023 and 2024, not from 2022, and that the thematic report is titled “Resultados 2023”; 2022 appears only because the reference minimum wage was R$1,212.00 at the time. The data were collected before 1 January 2025, when the rules of Law 14.790/2023 took effect, including identity verification, centralised self-exclusion and limits.

It was reported that the sample covers all forms of gambling, not only betting: 71.3% of gamblers play Mega-Sena, Quina, Lotomania and similar lotteries, while 32.1% prefer bets and 28.9% the jogo do bicho. The income breakdown applies to the whole universe of gambling, including Caixa lotteries, which the MP maintained. The survey found 3.4% of players at low risk, 2.6% at moderate risk and 0.8% problem gamblers among the population over 14. BNLData said the columnist made this caveat by stating that the results “include different degrees of risk”, but that readers tend to read 52.8% as dependence, which the figure does not measure.

The column's estimate of at least R$30.6bn a year in social costs linked to health harms, calculated from British references, was confirmed, and BNLData said the caveat that the figure is not spending recorded in the SUS budget was correct. The number comes from the IEPS dossier, which estimated total social cost at R$38.8bn, of which R$30.6bn is the health share: R$17bn from suicide deaths, R$10.4bn from loss of quality of life related to depression and R$3bn in medical treatment for the disease. BNLData said almost 90% of the value is the monetisation of deaths and quality of life, not money spent, and that the costs were calculated using a methodology used by the UK government, adjusted for purchasing power parity between the two countries. BNLData said it had already disputed that methodology, which was cited by minister Alexandre Padilha in the announcement of the MP.

BNLData said one of the study's authors stated that the individual cost to families that lose money on bets was not included in the calculation, and that the IEPS's main proposal was to increase the share of online betting taxation going to health. With the MP, that option no longer exists. BNLData said the column's description of the MP was correct: the text prohibits the exploitation, offering, intermediation and advertising of fixed-odds betting lotteries, in physical or virtual form, covering bets on real sporting events and on virtual online game events.

On financing, BNLData said the columnist was right that a debt does not disappear when a site is blocked and that care must continue, but did not say with what money. Under Law 14.790/2023, part of operators' revenue went by law to the Ministry of Health: 1% for measures to prevent, control and mitigate social harms from gambling in health areas. Between February 2025 and February 2026, about R$57m went to the National Health Fund for those actions. BNLData said this was little and that the sector itself had always argued that the allocation to health could be larger, but that it was money paid by operators and earmarked for treatment. With the MP, revenue from the regulated market went to zero, and BNLData said funding now comes from the general budget, meaning taxes also paid by people who never gambled.

BNLData said illegal platforms, which already held more than 40% of the market before the prohibition, pay no tax, do not verify age, offer no self-exclusion and do not fund any treatment. It said LENAD III found that users of digital platforms recorded a prevalence of 66.8% for risky or problem gambling, and that those platforms are now all clandestine. BNLData said the LENAD authors, the column's main source, recommended the opposite of a prohibition without counterpart, including mandatory contributions from online betting platform operators to a public fund to finance research, campaigns and treatment. It said the WHO, also cited in the column, states that unlicensed providers regularly offer unregulated products and that there is an urgent need to regulate online gambling.

In its conclusion, BNLData said Joildo Santos wrote an honest column with verifiable sources and methodological caveats, and that the numbers check out, but that the problem is what was left out: the column treats the prohibition as a starting point for demanding action from the SUS without recording that the same MP removed the only revenue linked to the problem and pushed bettors towards platforms without protection. BNLData's check table classified the LENAD III period as imprecise, the financing of treatment after the prohibition as an omission, and the other claims as confirmed, with the IEPS figure confirmed with a correct caveat.