The number that justifies the bill is two billion dollars over ten years. Two hundred million a year. The federal deficit has not printed below one trillion dollars in any fiscal year since 2019. As a fraction of annual borrowing, the figure rounds to roughly one one-hundredth of one percent.
The number also has no source. No Joint Committee on Taxation score. No Congressional Budget Office estimate. No static-versus-dynamic qualifier. It appears in the reporting as a bare integer, carrying the same evidentiary weight as a whitepaper's roadmap slide — a plausible shape, unattached to a model. A fiscal claim whose provenance is a single unattributed integer is not a fiscal claim. It is a narrative device.
That is the first red flag. The second is the mechanism being touched, which the coverage filed under "gambling" and which is not a gambling story at all.
What Actually Moved
Congress is advancing legislation to restore full deductibility of gambling losses, reversing a cap installed by the 2017 tax overhaul. The operative provision is §165(d) of the Internal Revenue Code: gambling losses are deductible only to the extent of gambling winnings, and only for taxpayers who itemize. I will note here that the source material does not specify the cap mechanism it attributes to the earlier overhaul, which means my legislative baseline is itself unverified. Flag it, then proceed.
The distinction that matters: §165(d) is a netting rule, not a subsidy. It says a participant is taxed on net position rather than gross handle. Netting rules exist because taxing gross churn is economically incoherent. It converts a zero-sum game into a revenue machine that destroys its own participants.
The backdrop is a legalization wave. After Murphy v. NCAA in 2018, states opened sports betting jurisdiction by jurisdiction. Gross gaming revenue compounded. Lobbying capacity compounded alongside it. Bills like this one do not advance because they raise revenue or because they are popular. They advance because the industry that benefits has learned how to schedule a committee hearing.
The reason this belongs in a crypto publication is venue migration, not the casino. Binary event contracts — Kalshi under CFTC registration, Polymarket offshore and now partially repatriated, plus a long tail of on-chain betting markets — are the fastest-growing segment of the category. They settle on infrastructure the tax code was never written to read.
The Score Is a Fiction, and the Error Term Is Unknowable
Two hundred million a year against a deficit that has not closed below a trillion since 2019. Order of magnitude: the bill's ten-year cost is roughly the annual revenue of one mid-size regional casino. Debt sustainability impact: zero. Anyone trading this as a macro event is trading noise.
The interesting property is not the score. It is the score's error term. A static estimate holds wagering volume fixed and asks what the deduction costs. Remove the cap, and the after-tax cost of churn falls for the highest-volume participants. Churn rises. Handle rises. The netted base rises. Whether the score overstates or understates revenue depends on the elasticity of high-volume wagering with respect to tax friction — a parameter nobody has estimated cleanly, because the category has only existed at scale for eight years. Assumptions are just risks wearing disguises.
The Classification Problem
Here is where the bill collides with crypto, and where nearly all commentary has stopped short.
Is a binary event contract a wager, a capital asset, or a §1256 contract? Three answers, three tax regimes, and no settled authority for the instruments that are actually growing.
- Wager, §165(d): losses deductible only against winnings, no carryforward, itemized.
- Capital asset, §1211: ordinary offset capped at $3,000, losses carry forward, preferential rates on gains.
- §1256 contract: mark-to-market, 60/40 treatment, and a distinct loss regime under §1256(a).
For a one-hour BTC-up/down market settling inside a smart contract, every one of those classifications is arguable. None is settled. Which means Congress is uncapping a deduction whose scope is undefined for precisely the class of instruments expanding fastest. Value is consensus; truth is optional.
The Verification Asymmetry
Now the operational part, and the part that should worry anyone with exposure.
Winning bets are third-party reported. Casinos issue W-2G. Regulated platforms issue 1099s. Kalshi issues forms. The information return exists, it is machine-readable, and the IRS can match it against the return.
Losing bets are self-reported. There is no counterparty confirmation for a "gambling losses" line, because for an on-chain wager the counterparty is a liquidity pool, and a pool does not file a return.
Uncapping the deduction therefore increases the nominal value of the least-confirmed line item on the form. That is the entire systemic shape of the bill: it raises the payout on the claim with the thinnest audit surface. The math holds, but the humans did not verify it.
I have run this reconstruction exercise. Based on my audit experience, building a defensible netting schedule for a wallet that wagered across four chains, bridged through three routes, and cleared through two centralized exchanges is possible in principle and impractical at scale. The chain records the transfer. It does not record the taxpayer. Provenance is a story we agree to believe in.
This is not hypothetical. It is the same class of problem I modeled in 2025 when I built a formal verification framework for AI-agent contract interfaces, titled "Semantic Drift in Autonomous Transactions." If an autonomous agent places forty thousand event-contract trades in a year, the record exists. The classification does not. The taxpayer-level netting does not. You have delegated a deterministic ledger to a non-deterministic actor and then asked the tax code to reconcile both.
What the Dismissers Got Right, and Where They Stop
The dismissive read is arithmetically correct and analytically useless. Two billion dollars is noise. Everyone saying so is right.
They are right about the wrong thing. The bill is not a fiscal instrument. It is a classification instrument, and the revenue score is a decoy designed to make the classification fight look small.
Consider what the gambling industry actually buys here. Not two hundred million a year in aggregate taxpayer relief — that number is distributed across millions of itemizers and is invisible on any single balance sheet. What it buys is a precedent: that a binary event with an uncertain outcome, settled between strangers, sits inside the gambling code rather than the securities code. That precedent is worth orders of magnitude more than the score, because it determines which regulator writes the rules for the next decade of event contracts.
If a one-hour BTC market is legally a wager, it inherits gambling's tax furniture, its licensing posture, and its state-level patchwork. If it is a swap, it does not. The cohort pushing for a single consolidated regulated venue is not solving a technical problem — there is no technical problem. It is solving a jurisdictional one, and calling it liquidity fragmentation. Correlation is the comfort of the unprepared.
Who Signs the Return
The tax code was written for a world where the counterparty was a casino with a cage and a Form W-2G. The wager is migrating to a smart contract with no counterparty at all — deterministic settlement, non-deterministic attribution.
Congress is about to expand a netting rule whose verification layer does not exist for the fastest-growing segment of the category it governs. When the settlement is machine-verified and the taxpayer is self-declared, who signs the return?