Federal Sports Betting Standards

Set a federal floor under state sports betting: deposit limits, affordability checks, no bonus bets, and no advertising during play.

AI evaluation · not yet reviewed by a human

This evaluation was produced and sourced by an AI model; a human review is still pending. Figures and conclusions may still change. The review log is at the foot of the page.How review works →

States have written their own sports betting rules since 2018 and the results range from strict to almost none. The bill before Congress would set a national minimum: no more than five deposits in twenty-four hours, an affordability check before large deposits, no advertising during a live event or between eight in the morning and ten at night, an end to bonus bets and odds boosts, and a ban on using artificial intelligence to track an individual's betting or build offers around it. Betting itself stays legal and stays a state decision; what changes is how it may be sold and how fast a person may lose. This evaluation looks five years ahead.

Balance

Better for the future · 0.60 previous scale

Balance on the previous scale. The Bilanz 2.0 simulation is not yet available for this evaluation. The category comes from the share of the debate on the pro side (r).

For 44 · 60 % Against 30 · 40 %

Size class: medium Scale of this evaluation: Normalised Impact — unitless, calibrated to this topic. For comparison: one point here is worth roughly 200 million euro per year. The two halves of this bill have very different evidence behind them, and they are scored separately for that reason. Deposit limits have been tried and worked on the heaviest players; advertising bans have been tried and did not reduce gambling. Anyone who reads the Italian experience differently will get a different answer. How we score →

Arguments for

Arguments against

7 arguments evaluated · Scoring v1.3 Δ absolute +14

Arguments — For

4 arguments · top 3 shown

Limits catch the people losing most

21of 100

Sports betting losses are extraordinarily concentrated: a small share of bettors accounts for about half of what sportsbooks keep. Deposit caps and affordability checks bite on exactly that group and almost nowhere else. Where such limits have been imposed, the heaviest players are the ones whose spending fell.

Value 5 · Household budgetsImpact 8.3Plausibility 5

▸ Show reasoning & sources ▾ Hide reasoning & sources

The stream is money that stays in a household instead of moving to a sportsbook. It is priced at the middle of the scale like any other money. What makes it a gain rather than a wash is the gap between the two ends: the people who lose most at sports betting are disproportionately young men under financial pressure, while the money arrives with the shareholders of listed operating companies. That gap is counted in the Impact. What the bettor loses along with the money — the enjoyment of the bet — is a real thing and is counted as its own argument against this measure, not netted away here. The value is the middle of the scale, because the stream is a transfer of money and the distance between its two ends is priced in the Impact.

American sportsbooks kept 16.96 billion dollars of what was staked in 2025, which is 14.6 billion euro at 1.16 dollars to the euro ^[1]. Losses in this market are concentrated the way they are in every gambling market: about half of what the books keep comes from a small minority of heavy players, and it is that minority the deposit caps and affordability checks are aimed at. Half of 14.6 billion is 7.3 billion euro. Limit-setting studies put the reduction in that group's spending at a quarter, in a range from a tenth to nearly a half, giving 1.83 billion euro a year ^[6]. The two ends of the transfer differ: heavy bettors carry a weight of 1.4 on this site because the losses concentrate among financially constrained households, and the operators' shareholders carry 0.5. Only the difference of 0.9 is counted, which is 1.65 billion euro a year. The Impact is the largest in this debate and it comes almost entirely from one small group of bettors.

▸ Show calculation ▾ Hide calculation

	Kept by American sportsbooks in a year [1]	16.96 billion dollars at 1.16 to the euro	14.6 billion euro
×	Share lost by the small minority of heavy players Setting, range 35 to 65 percent: losses concentrate this way in every gambling market, and it is that group the limits are aimed at	50 %	7.3 billion euro
×	Reduction in their losses under deposit caps and affordability checks Setting, range 10 to 45 percent: limit-setting studies find the reduction falls on the most intensive players rather than being spread across everyone [6]	25 %	1.83 billion euro
×	Difference in what a euro is worth heavy bettors are disproportionately under financial pressure, where this site counts a euro at 1.4; the money arrives with shareholders of listed companies, where it counts at 0.5	1.4 minus 0.5	1.65 billion euro
÷	Normalised Impact scale of this evaluation	200 million euro a point	8.25

Score 8.25 Impact × 5 Value × 5 Plausibility ÷ 10 = 21 of 100

Limits have been tried and the results point one way, though not from a design that settles it. Austria required 5,000 registered players to set deposit caps of at most 800 euro a week, and the reduction fell on the most intensive players rather than being spread across everyone — which is the pattern this argument needs ^[6]. Norway went further and made loss limits mandatory for its horse-betting monopoly in 2021, and overall consumption fell across every group of gamblers studied. The counterfactual in both cases is the same players before the limit, and that is the weakness: neither compares against a control group chosen in a way that rules out everything else changing at the same time. The confounder that matters is that players who hit a limit can open an account elsewhere, which the Norwegian monopoly setting hides and an American market of thirty operators would not; that is counted as its own argument against this measure. Reverse causation is a real risk in the Austrian data, where players chose their own limits, and it is why the Norwegian mandatory case carries more weight here. The Plausibility is at the middle: limits have been observed to work on the right group, and never against a comparison that rules out the alternatives.

evidence basis: Precedent · P ceiling 7 identification: Controlled · rung ceiling 7

Counterfactual: the same players before the limit was imposed (Austria, Norway) — no control group chosen to rule out concurrent change. Design: controlled — before-and-after comparison with player-level controls, not exogenous variation [6]. Confounder: players moving to another operator when limited, which a Norwegian monopoly hides; booked as con-3 rather than deducted here. Direction: reverse causation is a live risk in the Austrian data because players set their own limits, which is why the mandatory Norwegian case carries the argument. Ceiling: controlled 7.0 binds below the 8.5 a precedent carries; a context transfer of 2.0 covers the move from a European monopoly or single operator to an American market with thirty of them.

Less gambling disorder

14of 100

About 2.5 million American adults meet the clinical definition of gambling disorder, and the share of helpline callers naming an app as their main problem has risen from 23 to 31 percent in a year. Half of those callers are under 35. Gambling disorder is not a financial condition; it is a psychiatric one, with the depression and suicide risk that implies.

Value 9 · HealthImpact 3.8Plausibility 4

▸ Show reasoning & sources ▾ Hide reasoning & sources

The stream is mental health: the compulsion itself, the depression and anxiety that travel with it, the sleep that goes, the relationships that end, and at the far edge the suicide risk that is several times the ordinary rate among people in treatment for gambling. This site places that in the class it uses for life and health, one step below the top of it because what is lost can be regained — most people who develop a gambling disorder recover from it. Nor does the weight rise because the condition is self-inflicted or fall because it is: this site prices what happens to a person, not how they came to be there. What is counted is the person's own condition, not the money, which is counted in the arguments above, and not the harm to their family, which nothing here measures well enough to price. The value sits one step below the maximum: the stream is health, and health that can be regained.

About 2.5 million American adults meet the clinical definition of gambling disorder, with a further five to eight million showing some problematic behaviour ^[4]. Among online sports bettors specifically, 19 percent report problematic behaviour repeatedly — more than twice the rate among gamblers generally. A quarter of the clinical group is attributed here to the online betting expansion, in a range from a tenth to a half, which the helpline data support: online and app-based gambling has gone from 23 to 31 percent of what callers name as their main problem in a single year ^[4]. That is 625,000 people. The standards are assumed to prevent or resolve the condition for 15 percent of them, in a range from 5 to 30 percent — deposit limits reach the mechanism of a disorder more directly than an advertising rule does. Across 94,000 people a loss of 0.2 quality-adjusted years each gives 18,800 quality-adjusted years a year. The Impact is small in size and the largest here in weight, which is why it carries a third of the case for this measure on a fraction of the money.

▸ Show calculation ▾ Hide calculation

	American adults meeting the clinical definition of gambling disorder [4]		2.5 million people
×	Share attributable to the online betting expansion Setting, range 10 to 50 percent: online and app-based gambling rose from 23 to 31 percent of helpline callers' main problem in a single year [4]	25 %	625,000 people
×	Share the standards prevent or resolve Setting, range 5 to 30 percent: a deposit cap reaches the mechanism of a disorder more directly than an advertising rule does	15 %	94,000 people
×	Quality-adjusted years lost per person a year Setting, range 0.1 to 0.35: the compulsion itself, with the depression, anxiety and elevated suicide risk that travel with it	0.2	18,800 quality-adjusted years
×	Value of the years the value of a healthy life year used across this site	40,000 euro each	752 million euro
÷	Normalised Impact scale of this evaluation	200 million euro a point	3.76

Score 3.76 Impact × 9 Value × 4 Plausibility ÷ 10 = 14 of 100

Every step of this chain is visible and none of the sizes has been measured. The counterfactual is the same population under current state rules. That online sports betting raises problem gambling rates is supported by survey data showing the rate among online bettors at more than twice the general figure, but survey data of that kind cannot separate the betting from the people who choose it — someone predisposed to a gambling problem is also more likely to open a betting app, which is the reverse-causation problem this evidence cannot solve ^[4]. The helpline series is better on timing, because online complaints rose sharply in the years the apps arrived, but it counts help-seeking rather than illness. What no source addresses at all is the step from these particular rules to fewer people with the disorder; the reasoning that a deposit cap interrupts loss-chasing is sound and untested. The counter-mechanism is that someone with a disorder is precisely the person who will open a second account, which is answered only in part by the argument against this measure that follows. The Plausibility is below the middle: the harm is real and the path from these rules to less of it is reasoned rather than observed.

evidence basis: Mechanism · P ceiling 5.5 identification: Associational · rung ceiling 5.5 band: Chain closed, unevidenced · P 4–5

Counterfactual: the same population under current state rules. Design: associational for the harm link — cross-sectional survey rates among online bettors against gamblers generally, with no counterfactual [4]; mechanistic for the policy link, which no source tests at all. Confounder: selection into betting by people already predisposed to a gambling problem, which the survey design cannot remove; unresolved. Direction: reverse causation is the central weakness here and is not addressed by any available source. Ceiling: associational 5.5 binds. Finding band: chain closed but unevidenced — the chain from limits to interrupted loss-chasing is named and the second-account counter-mechanism is answered in part by con-3; only the measurement is missing.

Nothing measured argues against the claim; what is missing is any study linking operating rules to disorder prevalence. The counter-mechanism — that a person with a disorder opens a second account — is partly answered by the argument on unregulated books. Read back: about half the time, standards of this kind reach roughly the share of affected people assumed here.

Open: Helpline contacts are recorded by state and by gambling type. A comparison of states adopting deposit limits against those that do not, over two years, would give the first direct measurement of this link.

Fewer advertisements, fewer free bets

4.7of 100

The bill would clear sportsbook advertising out of live broadcasts and daytime television and end the bonus bets that get people started. It is the part of the bill most people notice and the part with the worst evidence behind it. Italy banned gambling advertising outright in 2019 and gambling went up.

Value 5 · Household budgetsImpact 4.7Plausibility 2

▸ Show reasoning & sources ▾ Hide reasoning & sources

The stream is the same one the argument above counts — money staying with households instead of moving to sportsbooks — reached by a different instrument. It carries the same weight at the middle of the scale, and the same gap between the two ends of the transfer is what makes it a gain. What differs is who is affected: advertising and promotional offers reach everyone rather than the heaviest losers, so the population here is broader and less financially strained, which lowers the weight slightly. That adjustment sits in the Impact. The value is the middle of the scale, because the stream is a transfer of money and the distance between its two ends is priced in the Impact.

Of the 14.6 billion euro American sportsbooks keep in a year, the part that advertising and promotional offers bring in is what this argument is about ^[1]. An eight percent reduction in volume is used, in a range from none to twenty percent — the low centre of that range is set by the evidence in the next paragraph rather than by any doubt about the arithmetic. That gives 1.17 billion euro a year. Advertising reaches a wider and less strained group of bettors than deposit limits do, so the weight on their side is 1.3 rather than 1.4, and the operators' side is unchanged at 0.5: a difference of 0.8. The result is 0.94 billion euro a year. If the Italian experience carries over completely, the honest figure is nearer zero. The Impact is roughly half the deposit-limit gain, and it is the half this evaluation is least confident of.

▸ Show calculation ▾ Hide calculation

	Kept by American sportsbooks in a year [1]		14.6 billion euro
×	Reduction in volume from ending advertising and bonus bets Setting, range 0 to 20 percent: the low centre is set by the Italian experience, where an outright ban did not reduce gambling [7]	8 %	1.17 billion euro
×	Difference in what a euro is worth advertising reaches a wider and less financially strained group than deposit limits do, so the bettors' side is weighted slightly lower than in the argument above	1.3 minus 0.5	0.94 billion euro
÷	Normalised Impact scale of this evaluation	200 million euro a point	4.7

Score 4.7 Impact × 5 Value × 2 Plausibility ÷ 10 = 4.7 of 100

This has been tried at national scale and it did not work. Italy banned all gambling advertising and sponsorship in 2019, and its own parliamentary commission of inquiry reported in 2022 that gambling had increased after the ban, including among minors; the football federation, which lost the sponsorship revenue, describes the ban as largely ineffective ^[7]. The counterfactual there is Italy before the ban rather than a comparable country without one, which is a real weakness — the pandemic fell in the middle of the period and moved gambling online everywhere. That weakness cuts both ways and does not rescue the argument. What could rescue it is that the American bill also bans bonus bets and odds boosts, which Italy did not, and those promotions are how most online sportsbooks acquire customers; nothing has tested that separately. Reverse causation does not arise, since the ban was a legislative act. The Plausibility is low because the closest thing to this measure was enacted in a comparable market and the effect it is supposed to produce did not appear.

evidence basis: Precedent · P ceiling 5.5 identification: Associational · rung ceiling 5.5 band: Effect did not materialise · P 1.5–2

Counterfactual: Italy before the 2019 ban — no comparison country, so the design cannot separate the ban from anything else that changed. Design: associational — an uncontrolled before-and-after at national level [7]. Confounder: the pandemic, which moved gambling online across all markets in the middle of the observation period; unresolved and cutting both ways. Direction: no reverse causation, the ban was a legislative act. Ceiling: associational 5.5 binds below the 8.5 a precedent carries, and the finding band binds far below that. Finding band: effect did not materialise — the measure's closest twin ran and gambling rose rather than fell.

Italy banned gambling advertising and sponsorship outright in 2019 and its own parliamentary inquiry found gambling had increased afterwards, minors included [7]. Read back: roughly one time in five, an advertising and promotion ban reduces betting volume by about the amount assumed here; the rest of the time it does not.

Open: The bill's advertising and promotion rules would apply nationally at one date, so a comparison of states with existing promotional restrictions against those without, before and after, would settle it within two seasons.

Fewer bankruptcies and overdrafts

4.6of 100

Personal bankruptcy filings rose by a quarter to a third in states after online sports betting arrived, measured on the credit records of seven million people. Debt sent to collections, consolidation loans and car loan arrears all moved with it. Slowing how fast someone can lose reaches that directly.

Value 5 · Household budgetsImpact 1.9Plausibility 5

▸ Show reasoning & sources ▾ Hide reasoning & sources

The stream is the cost of financial collapse, which is not the same as the money lost betting. A bankruptcy costs filing fees and legal fees, then a decade of credit that is dearer or unavailable, then the practical consequences of that: a car loan refused, a deposit that cannot be raised, a job application that runs a credit check. It is priced at the middle of the scale like other money. The debt discharged in a bankruptcy is not counted as a loss, because it moves to the creditor rather than disappearing. The gambling losses themselves are counted in the two arguments above and are not repeated here. The value is the middle of the scale, because what is counted is a money cost and not the collapse itself.

Personal bankruptcy filings run at about 490,000 a year in the United States. Credit-record data for seven million consumers, compared across states as online betting arrived at different dates, put the increase at 25 to 30 percent where it is legal ^[2]. About seven in ten Americans live in such a state, which makes roughly 79,000 filings a year attributable to it. The standards in this bill reach that directly — an affordability check before a large deposit is aimed at exactly the person about to file — and a fifth of the increase is assumed removed, in a range from a twelfth to a third: 15,800 filings a year. A bankruptcy costs the person filing about 12,000 euro once the fees, the years of impaired credit and the practical consequences are counted, in a range from 4,000 to 35,000. Adding the smaller but much wider costs of overdrafts, collections and consolidation loans among heavy bettors who do not file gives about 286 million euro, weighted at 1.3 for the households it falls on. The Impact is small beside the betting losses themselves, because bankruptcy is the visible end of a much larger stream that is counted elsewhere.

▸ Show calculation ▾ Hide calculation

	Personal bankruptcy filings a year		490,000 filings
×	Share attributable to online sports betting [2]	25 to 30 percent higher where legal, in states holding seven tenths of the population	79,000 filings a year
×	Share the standards remove Setting, range 8 to 33 percent: an affordability check before a large deposit is aimed at exactly the person about to file	20 %	15,800 filings a year
×	Cost of a bankruptcy to the person filing Setting, range 4,000 to 35,000 euro: fees, then years of credit that is dearer or unavailable, and the practical consequences of that	12,000 euro	190 million euro
+	Overdrafts, collections and consolidation loans among heavy bettors who do not file Setting, range 30 to 250 million: a much smaller cost across a much wider group [3]	96 million euro	286 million euro
×	Weight of a euro at these incomes the households concerned are the financially constrained ones in whom both studies find the effect concentrated	1.3	372 million euro
÷	Normalised Impact scale of this evaluation	200 million euro a point	1.86

Score 1.86 Impact × 5 Value × 5 Plausibility ÷ 10 = 4.6 of 100

The link between online betting and financial collapse is the best-established finding in this debate. Sports betting became legal in different states at different times, which allows early states to be compared against late ones on the same credit records, and across that comparison bankruptcy filings, debt sent to collections, consolidation loans and car loan arrears all rose together ^[2]. A second study using the bank transaction records of 230,000 households found the same pattern from the other direction: net investment fell 14 percent in betting households, concentrated among those already financially constrained ^[3]. The confounder that would matter — that states legalising early differ economically from those legalising late — is what the staggered timing absorbs, and reverse causation does not arise, because a household's credit record does not determine when its legislature acts. What is not established is the step this argument needs: those studies measure legalisation against prohibition, not one set of operating rules against another. That transfer is the whole uncertainty here. The Plausibility is at the middle: the harm is measured under a design that supports it, and whether these particular rules remove a fifth of it has never been tested.

evidence basis: Converging studies · P ceiling 8 identification: Quasi-experimental · rung ceiling 8

Counterfactual: states that legalised online betting later, on the same credit records. Design: quasi-experimental — staggered difference-in-differences across states (Hollenbeck, Larsen and Proserpio, Management Science 2025 [2]), corroborated on household transaction data (Baker and others, NBER 2024 [3]). Confounder: early-legalising states differing economically, absorbed by the staggered timing. Direction: no reverse causation, credit records do not determine legislative dates. Ceiling: quasi-experimental 8.0 binds below the converging-studies ceiling of 9.0; a deduction of 3.0 applies because both studies measure legalisation against prohibition rather than these operating rules against the current ones. The size doubt sits in the one-twelfth to one-third band.

Arguments — Against

3 arguments

People lose something they chose

12of 100

Most people who bet on sport are not addicted and are not in trouble. They put twenty dollars on a game because it makes the game better, and they know what it costs them. A rule that stops some of that takes something real away from them.

Value 4 · EnjoymentImpact 5.8Plausibility 5

▸ Show reasoning & sources ▾ Hide reasoning & sources

The stream is enjoyment: the interest a wager adds to a match that would otherwise not matter, and the sociability around it. This site places that in the class it uses for comfort and everyday pleasure, well below health, money in a strained household, or the environment. It is not nothing — this site does not treat a freely chosen pleasure as worthless simply because someone disapproves of it — but it does not weigh much either. What is priced here is the enjoyment, not the money, which is counted as a gain on the other side; the same euro appears once as a loss to the bettor's pocket and once as the pleasure they bought with it, and those are different goods. The value is in the lower part of the scale, because what is lost is a chosen pleasure rather than anything anyone depends on.

The two arguments above between them remove about 3.0 billion euro a year of betting losses. Not all of that comes from people in difficulty. Roughly three fifths of it is lost by recreational bettors who are getting what they paid for, in a range from two fifths to four fifths — a judgment that follows from the concentration of losses rather than from any direct measurement of who is having a good time. What they gain from a bet is worth less than what they pay for it, since a bet is a losing proposition by construction; half the amount lost is used for the enjoyment it bought, in a range from a quarter to the full amount. That gives 0.9 billion euro, weighted at 1.3 for the incomes concerned: 1.17 billion euro a year. The Impact is larger than either of the two gains it offsets except the deposit-limit one, which is the honest shape of a paternalistic measure: most of the people it restrains were not in trouble.

▸ Show calculation ▾ Hide calculation

	Betting losses the two gains above remove	1.83 plus 1.17 billion euro	3 billion euro
×	Share lost by recreational bettors getting what they paid for Setting, range 40 to 80 percent: follows from how losses concentrate, not from any measurement of who is enjoying it	60 %	1.8 billion euro
×	Enjoyment bought per euro lost Setting, range 25 to 100 percent: a bet is a losing proposition by construction, so the pleasure is worth less than the price	50 %	0.9 billion euro
×	Weight of a euro at these incomes	1.3	1.17 billion euro
÷	Normalised Impact scale of this evaluation	200 million euro a point	5.85

Score 5.85 Impact × 4 Value × 5 Plausibility ÷ 10 = 12 of 100

That people get something out of betting is not in doubt, since they pay for it repeatedly and freely. What has no measurement is the size, and there is a reason it does not: the standard way to price a consumer's enjoyment assumes the consumer is choosing well, which is exactly what is contested for a share of this market. The counterfactual is the same bettors under current rules. Two things pull the figure in opposite directions and neither is resolved. Someone in the grip of a gambling disorder is not obtaining the enjoyment their spending implies, so counting all recreational losses at face value overstates this argument — that is why only three fifths of the reduction is treated as recreational. Against that, deposit limits and advertising rules fall on everyone, including the great majority who never approach a problem, and the inconvenience to them is real and entirely uncounted here. Reverse causation does not arise. The Plausibility is at the middle: the stream certainly exists, and pricing a pleasure whose consumers may not be choosing well has no settled method.

evidence basis: Plausibility · P ceiling 5 identification: Mechanistic · rung ceiling 6

Counterfactual: the same bettors under current state rules. Design: mechanistic — the pricing of the pleasure is definitional (a freely repeated purchase yields something to the buyer), but whether it is lost at all hangs on the deposit limits actually binding, which is the claim carried by pro-1 and is named rather than measured. Confounder: gambling disorder, under which the assumption fails and spending overstates enjoyment; addressed by treating only three fifths of the reduction as recreational. Direction: not applicable. Ceiling: a framework assumption carries 5.0 at most, and that binds below the mechanistic 6.0, because the size rests on a stated share rather than any measurement. Entry: this cost arises only if betting volume actually falls, so it shares an entry group with pro-1, con-2 and con-3 rather than being treated as certain.

State treasuries collect less

9.9of 100

States took 3.71 billion dollars in sports betting tax in 2025, a third more than the year before, and several have written it into school and infrastructure budgets. Less betting means less of it. The money does not vanish, but the budget hole is real.

Value 5 · Public financesImpact 3.3Plausibility 6

▸ Show reasoning & sources ▾ Hide reasoning & sources

The stream is state and local revenue, priced at the middle of the scale like any other public money. What is counted is only the loss of weight in the passage: the money stays with the bettor rather than reaching a treasury, and since it stays with someone for whom a euro is worth more than it is worth to a government, the transfer's disappearance is mostly a gain — which is already counted on the other side. Nothing is counted for the fact that several states earmarked the money for schools or for problem gambling treatment, because an earmark is a budgeting choice rather than a separate good. What is left as a genuine cost is the part of the tax that fell on operators' margins rather than on bettors, plus the disruption of a budget line that states have come to rely on. The value is the middle of the scale, the level this site uses for public money whatever its level of government.

States collected 3.71 billion dollars in sports betting tax in 2025, or 3.2 billion euro, a third more than the year before ^[1]. The standards reduce betting volume by about a fifth on the figures used above, so the revenue falls by roughly the same proportion: 656 million euro a year, in a range from 250 million to 1.1 billion. The full amount is booked here rather than only the weight difference, because unlike the betting losses themselves this money does not stay with the bettor — the part that was passing through as tax has to be replaced from somewhere or a service has to shrink, and which of those it becomes is a state decision that lies outside this measure. What is not counted is the offsetting saving on the treatment and social costs of problem gambling, which several states fund from the same tax. The Impact is a third of the deposit-limit gain, which is the ordinary proportion when a state takes a fifth of what an industry keeps.

▸ Show calculation ▾ Hide calculation

	State sports betting tax receipts [1]	3.71 billion dollars at 1.16 to the euro	3.2 billion euro
×	Share lost with the reduced volume Setting, range 8 to 34 percent: the same volume reduction that carries the two arguments in favour	20.5 %	656 million euro
×	Weight of a euro in a state budget the standard weight for public money on this site	1.0	656 million euro
÷	Normalised Impact scale of this evaluation	200 million euro a point	3.3

Score 3.3 Impact × 5 Value × 6 Plausibility ÷ 10 = 9.9 of 100

The arithmetic is direct: a tax on a volume falls when the volume falls, and both the rate and the base are published monthly by every state that levies one ^[1]. The counterfactual is the same tax under current rules. What is estimated is the volume reduction, and that is the same estimate that carries the two arguments in favour, so the three move together rather than independently — if the standards do less than assumed, this cost shrinks along with the benefit. That coupling is the reason this figure carries the range it does rather than a narrower one. The confounder that would matter is behavioural on the states' side: a state facing a revenue loss may raise its tax rate on the remaining volume, which several did in 2025 for unrelated reasons, and that would offset part of this. Reverse causation does not arise. The Plausibility is at the top of what a projection can carry: the tax is arithmetic and only the volume it applies to is estimated.

evidence basis: Projection · P ceiling 6 identification: Mechanistic · rung ceiling 6

Counterfactual: the same tax rates applied to current volumes, published monthly by every state that levies one [1]. Design: mechanistic — the tax arithmetic itself is definitional, rate times base, but the base only falls if the standards actually reduce volume, and that link is named rather than measured here. Confounder: states raising rates on the remaining volume to compensate; named and unresolved. Direction: not applicable. Ceiling: a projection carries 6.0 at most, level with the mechanistic ceiling, and that binds. Parameter coupling: the volume reduction is one number shared with pro-1 and pro-2, so this cost and those benefits move together — it is therefore held out of this argument's range and carried as a scenario instead. Entry: shares an entry group with pro-1, con-1 and con-3.

Some of it moves to books with no rules at all

7.9of 100

A bettor stopped by a deposit limit has thirty licensed operators to try and, after that, an offshore site with none. Unregulated books have no limits, no self-exclusion register, no way to get money back and no obligation to pay out. The people most likely to go looking are the ones the limits were written for.

Value 5 · Household budgetsImpact 4.5Plausibility 3.5

▸ Show reasoning & sources ▾ Hide reasoning & sources

The stream is money again, at the middle of the scale, but it is money lost under worse conditions. An offshore book takes the same stake and offers none of what a licensed one must: no cap on deposits, no exclusion register a person can put themselves on, no regulator to complain to, and no certainty of being paid when a bet wins. What is priced here is both the transfer that the measure fails to prevent and the additional harm of losing it somewhere with no recourse. The gambling disorder that follows a person offshore is counted in the argument on health rather than twice here. The value is the middle of the scale, because the stream is money and the absence of protection around it is priced in the Impact.

Of the 3.0 billion euro of betting losses the standards remove, a quarter is assumed to reappear at unregulated books, in a range from a tenth to a half: 750 million euro. For that share the gain counted in the two arguments above does not happen, and the weight difference reverses — the money still leaves the bettor, and it leaves to an operator further beyond reach than a listed company. That is 600 million euro at the same weight gap of 0.8. On top of it, losing money where there is no exclusion register, no deposit cap and no obligation to pay out is worse than losing it at a licensed book: a premium of half the amount is added for the absent protections, in a range from a fifth to the full amount. The result is 900 million euro a year. The people who go looking are not a random quarter — they are disproportionately the heavy losers the limits target, which is what makes this argument bite. The Impact is about half the deposit-limit gain, which means the substitution question decides how much of this bill actually works.

▸ Show calculation ▾ Hide calculation

	Betting losses the standards remove		3 billion euro
×	Share reappearing at unregulated books Setting, range 10 to 50 percent: the people who go looking are disproportionately the heavy losers the limits target [7]	25 %	750 million euro
×	Weight gap that reverses the money still leaves the bettor, and it leaves to an operator further beyond reach than a listed company	0.8	600 million euro
+	Premium for the absent protections Setting, range 20 to 100 percent: no deposit cap, no exclusion register, no regulator to complain to, no obligation to pay out	50 %	900 million euro
÷	Normalised Impact scale of this evaluation	200 million euro a point	4.5

Score 4.5 Impact × 5 Value × 3.5 Plausibility ÷ 10 = 7.9 of 100

Everything here rests on a number nobody has. The counterfactual is the same bettors under current rules, and no American study measures how many restricted bettors move offshore, because no American state has imposed a binding deposit limit long enough to look. European regulators report the pattern qualitatively — the trade association for licensed European operators argues Italy's advertising ban pushed volume to unlicensed sites, though it is not a disinterested source and no measurement accompanies the claim ^[7]. The chain is short and each link is visible: a limit binds, the bettor wants to continue, an offshore site is one search away. The counter-mechanism is real and unanswered: offshore sites are harder to fund now than they were, since payment processors and banks block much of the traffic, and how much friction that adds has not been quantified. Reverse causation does not arise. The Plausibility is low because the share that moves offshore has never been measured anywhere and the payment friction working against it is unquantified.

evidence basis: Mechanism · P ceiling 6 identification: Mechanistic · rung ceiling 6 band: Chain open · P 3–3.5

Counterfactual: the same bettors under current state rules; no American state has imposed a binding deposit limit long enough to observe substitution. Design: mechanistic — chain named (limit binds → bettor continues → offshore site), with only qualitative European reporting from an interested source [7]. Confounder: payment processors and banks blocking offshore funding, which adds unquantified friction; unanswered. Direction: no reverse causation. Ceiling: mechanistic 6.0 binds. Finding band: chain open, because the migrating share carries the whole quantity and the payment-friction counter-mechanism is unresolved.

The chain is named but the migrating share carrying the quantity has never been measured, and the counter-mechanism — payment blocking making offshore books harder to fund — is unquantified. Read back: about a third of the time, roughly a quarter of the suppressed betting reappears at unregulated books.

Open: Payment processors can see attempted transfers to known offshore operators. A comparison of those flows in states adopting deposit limits against states that do not would measure the substitution directly and could carry P to 6.

Summary

This bill is really two bills with different evidence behind them, and separating them is the most useful thing an evaluation can do here. The deposit caps and affordability checks are aimed at the small minority who lose most, they have been imposed elsewhere, and where they were imposed that minority's spending fell — that is where nearly half the case sits. The advertising and bonus-bet rules are the part everyone argues about and the part that failed its one real test: Italy banned gambling advertising outright in 2019 and its own parliamentary inquiry found gambling had risen afterwards. Against the whole package stand three costs that are easy to overlook and add up to two thirds of the benefit: the enjoyment of people who were never in trouble, a state tax line several states have already spent, and the quarter of suppressed betting that plausibly reappears at offshore books with no limits at all. The balance is positive and narrow, and it would be clearly positive if the bill contained only its first half.

Outlook — effect over time

Better for the future · 0.60 previous scale

Sources

Sports Handle: US Sports Betting in 2025 Reaches Record Highs. sportshandle.com
Hollenbeck, Larsen and Proserpio, Management Science: The Financial Consequences of Legalized Sports Gambling. pubsonline.informs.org
Baker and others, National Bureau of Economic Research: Gambling Away Stability: Sports Betting's Impact on Vulnerable Households. nber.org
National Council on Problem Gambling: National Survey on Gambling Attitudes and Gambling Experiences and Helpline Annual Report. ncpgambling.org
Congress.gov: H.R. 2087, SAFE Bet Act of 2025. congress.gov
Journal of Gambling Studies, comparative policy review: Limit-setting in online gambling: a comparative policy review of European approaches. pmc.ncbi.nlm.nih.gov
DLA Piper: Removal of the Italian gambling advertising ban: why Italy's football authority is now pushing for change. dlapiper.com

Last reviewed by Claude Opus 5 · September 6, 2026 · 2× AI, not yet reviewed by a human

September 6, 2026AI reviewClaude Opus 5record updated
i_spanne und normalisierung an allen 7 Argumenten (nur globale Anker), Eintrittsgruppe limits-actually-bind. Substanziell: con-1 und con-2 standen auf definitorisch, traten also auch dann ein, wenn die Einzahlungsgrenzen wirkungslos bleiben — jetzt mechanistisch und in der Gruppe, P und p_ceiling unveraendert. Das hebt P(D>0) von 0,39 auf 0,76; Kategorie bleibt Besser.
September 6, 2026AI reviewClaude Opus 5First evaluation
Created for the English side: deposit limits and advertising rules scored separately, because the European evidence splits between them.

Evaluations are produced with AI support and reviewed on a schedule for new developments; human passes are marked separately.How we review →