Orange County, California Correspondent

I analyze business models, and the business model of a prediction market is the first thing to understand, because it answers most of the question before the lawyers arrive. A sportsbook takes the other side of your bet. It sets a line, builds a margin into it, and profits when you lose; its interest is adverse to yours by design, and every state that has legalized it taxes that margin and regulates the book as a counterparty. A prediction market does not take the other side. It matches you with another trader who disagrees, charges a fee on the transaction, and profits from volume regardless of who wins; its interest is in activity, not in your loss. That is a real structural difference, and it is the one the platforms lean on when they say they are exchanges rather than bookmakers, and they are right about the mechanism. The question is whether the mechanism changes anything that matters to a person with a phone and a problem, and I think the honest answer, from someone who prices risk for a living, is that it changes the economics for the operator and almost nothing for the user, and the law has spent the year arguing about the operator.

Let me take the three parts of the question in order. On structure, the exchange model is genuine and the contracts are genuine derivatives in form: a binary contract that pays a dollar if an event occurs, priced between zero and one, tradable until settlement, listed on a venue the federal commodities regulator designated. That form has existed in regulated finance for decades. What changed in 2025 and 2026 is the underlying: the contracts moved from inflation prints and election outcomes to whether a college quarterback throws for three hundred yards, and sports now drive most of the roughly fifty-three billion dollars the industry traded in July. A derivative on a football game is a derivative, structurally, and it is also a bet on a football game, functionally, and the circuit courts have split this year on which description controls. The Third Circuit said in April that the form controls and federal law preempts New Jersey’s. The Ninth in August and the Sixth in September said the function controls and Nevada and Ohio and Tennessee may regulate. That is not a disagreement about facts. It is a disagreement about whether the law should look at the mechanism or the user, and the Supreme Court will have to pick.

On risk, the difference runs the other way from what the platforms claim, and this is the point I most want to make. A sportsbook’s adverse interest is, perversely, a consumer protection: the book knows exactly how much you have lost to it, it is licensed by a state that requires it to track that, and in most states it is required to report irregular patterns to a regulator who can act. An exchange has no adverse interest, which means it has no natural incentive to notice you, and the federal regulator that oversees it is a commodities agency built to supervise wheat futures and interest-rate swaps, not to run a problem-gambling program. The platforms are open at eighteen rather than twenty-one, because derivatives are. They carry no excise tax, because derivatives do not. They did not, until this year, have any mechanism for reporting suspicious trading to anyone, which is how a soldier with classified knowledge of the Maduro operation turned thirty thousand dollars into four hundred thousand on an offshore exchange in January and how traders appear to have positioned before the Iran strikes in February. The sportsbooks would have flagged those bets in an afternoon, because the sportsbooks lose money on them. The exchange earned a fee.

On social impact, I will say what the capital markets say, because they are the honest party here. Kalshi raised a billion dollars at an eleven-billion-dollar valuation and signed a cable news network as a partner. Robinhood, Coinbase, and Crypto.com list the contracts next to stocks and tokens. That is not the valuation of a market-information service, which is what prediction markets were when they priced elections and inflation and were useful. It is the valuation of a gambling product with a regulatory discount, and the investors who paid it know exactly what they are buying, which is access to the sports-betting market without the sports-betting rules. The social impact follows from the valuation: a product that reaches eighteen-year-olds on a brokerage app, with no state oversight, with the same dopamine architecture as every other app, and with a federal regulator that has spent the year suing states rather than writing consumer rules. The regulator proposed rules in June to ban contracts on war and assassination, which is the minimum, and has proposed nothing on advertising, limits, age, or self-exclusion, which are the things a gambling regulator does first.

So how should policymakers treat them? Here is the analyst’s answer, and it is not the one either side wants. The exchange structure is real and should be preserved for the contracts where it adds information: elections, economic data, policy outcomes, the things the markets were built to price and price well. Those contracts should stay federal, because fifty state regulators cannot supervise a national market in the probability of a rate cut. The sports contracts are a different product wearing the same form, they add no information the sportsbooks do not already produce, and they should be regulated as what they functionally are, which means the states, the twenty-one-year age floor, the tax, the reporting, and the self-exclusion lists that thirty-some states have spent eight years building. That is a line between the underlying rather than between the venues, and it is the line the Ninth and Sixth Circuits drew, and it is the one Congress should write before the Supreme Court writes a worse one.

My assessment is that prediction markets are meaningfully different from sportsbooks in structure, that the difference favors the operator and not the user, and that the legal distinction has become a workaround precisely at the point where the product became sports: the federal form was designed for information markets and is being used to carry a gambling product across state lines without the gambling rules. Policymakers should split the product by what it prices rather than where it trades. They will not, because the platforms have an eleven-billion-dollar incentive to keep the form and the states have a tax base to defend, and the Supreme Court will decide a question of statutory interpretation that is really a question of whether the law protects the mechanism or the person. I price things. The mechanism is priced at eleven billion. The person is priced at nothing, and that is the finding.

Jacksonville, Florida Correspondent

Florida does not have legal online sports betting in the ordinary sense; the state’s compact with the Seminole Tribe gives the tribe the market, and the tribe’s app is the only lawful book, and for a few years that arrangement held. Then the prediction markets arrived, federally licensed, open to anyone with a brokerage account, offering contracts on every game the tribe takes bets on, and the compact that the state and the tribe spent a decade and a Supreme Court appeal to secure was, overnight, a monopoly on a product that anyone could buy somewhere else without the tribe’s permission or the state’s cut. I was a claims adjuster and a state staffer before I moved freight, and I want to begin there because the question asks whether the legal distinctions are workarounds, and the view from a state that built an entire gaming structure around a tribal compact is that the distinction is not a workaround. It is a demolition, and the people doing it did not have to pick up a permit.

Let me describe how the thing actually arrived at the terminal, because that is where I see it. The drivers I dispatch have had sportsbooks on their phones for years in the states that allow it and have driven across the Georgia line to use them in the ones that do not; this year they have Kalshi and Polymarket and Robinhood instead, and they do not drive anywhere, because the contracts are federal and the app does not care where the truck is parked. The product is a bet on a game. The screen calls it a contract and prices it in cents on the dollar rather than in odds, and the drivers have learned to read it the way they learned to read a moneyline, which took about a week. The thing they notice, when I ask, is that it is cheaper, because there is no state tax built into the price, and that it is easier to get on, because the age is eighteen and the sign-up is a brokerage form rather than a gaming one. Nobody in the yard has asked whether it is a swap. They have asked whether it pays, and it does, and that is the whole consumer analysis.

Now the structural question, which I will take seriously because the platforms are serious about it and the Third Circuit agreed with them. A sportsbook sets a line and takes the other side; a prediction market matches buyers and sellers and takes a fee. That is a real difference in who holds the risk, and it has one consequence I respect, which is that the exchange cannot shade the line against you, because it does not hold a position. But I adjusted claims for a living and I know that the party who holds the risk is the party who watches, and a book that holds the risk watches every account because it loses money to the ones that know something. An exchange that holds no risk has no reason to watch, and the proof arrived in January, when a soldier with classified knowledge of the operation to capture Maduro bought contracts on his ouster hours before the raid and cleared four hundred thousand dollars on an offshore exchange that had no rule against it, and again in February, when trades appeared before the Iran strikes. The books in this state would have frozen those accounts before the payout. The exchange paid, and the Justice Department had to do the work the operator should have done, and it took until April to charge anyone. That is the structural difference the question asks about, and it cuts against the platforms: the thing that makes them not a bookmaker is the thing that makes them not a watchman.

On social impact, the state-office view. Florida spent years on the problem-gambling side of the compact: the hotline, the self-exclusion list, the age verification, the limits on advertising to minors, all of it negotiated with a tribe that had an interest in keeping its license. None of that applies to a federal contract market, because the regulator that designates the markets is a commodities agency whose rules were written for grain and swaps and whose consumer protections consist of a requirement that the exchange have a rulebook. The platforms are open at eighteen. They are on the same app as the stocks. They are advertised on the news network one of them has partnered with. The federal regulator has spent the year suing nine states to protect the platforms’ right to operate and has proposed exactly one consumer rule, in June, which would bar contracts on war and assassination, which is the floor and not the ceiling. The state structure that was supposed to manage the social cost of sports betting has been bypassed by a federal structure that does not acknowledge the cost exists.

So how should the policymakers treat it? From the dock, the answer is to decide what the product is by what it does rather than by what it is called, which is what the Ninth and Sixth Circuits did this fall and what the Third did not, and to split the market accordingly. Contracts on elections and economic data are information markets with a long record of being useful, and the federal regulator is the right home for them, because a national probability of a Fed cut cannot be supervised by fifty gaming boards. Contracts on whether the Jaguars cover are a sports bet in a derivative’s clothing, and they should answer to the states and the tribes that the states have made their deals with, at twenty-one, taxed, with the watchman the state requires. That is a line through the product, not through the venue, and the platforms will fight it because the sports contracts are where the fifty-three billion dollars came from. The Supreme Court will probably get the case. It will decide a question about the word swap. The question should be about the kid at the truck stop with the app, and it will not be, because he is not a party.

My answer is that prediction markets are meaningfully different from sportsbooks in exactly one respect that matters, which is that nobody on the other side of the trade has a reason to watch the trader, and that every other difference, the age, the tax, the state oversight, the tribal compact, is a regulatory workaround that the federal form made possible and the platforms monetized. The policymakers should treat the information contracts as finance and the sports contracts as gambling and stop pretending the form settles which is which. I am not a mascot for the tribe or for the state or for the apps. I move freight. The freight used to go through a gate the state built. The gate is still there. The traffic found a road around it, federally paved, and the toll booth is empty, and somebody should ask who decided to pave the road before the Supreme Court decides whether it was legal to build.

Johannesburg, South Africa Correspondent

I am going to defend the prediction markets, which I expect will put me in a minority, and I am going to do it as a risk consultant who has spent a career trying to find out what is actually going to happen before the people in charge announce it. The single most useful instrument I had during the years of load-shedding here was not a government statement or a utility forecast; it was the price of a thing. When the state power company said the grid was stable and the diesel futures curve said otherwise, the curve was right, and I learned to read prices as the only honest witness in a room full of officials. Prediction markets are that witness, applied to events. In January a contract on Nicolás Maduro’s ouster spiked six and a half hours before American forces took him, and the whole world learned, from a price, that someone knew. In February the contracts on Iran moved before the strikes. The American response has been to prosecute the trader, which is correct, and to treat the price movement as a scandal, which is backwards. The price did what prices are for. It told the truth before the government did, and a society that wants to know things should want more of that, not less.

Let me take the question’s structural claim seriously, because it is where the defence is strongest and the critics least honest. A sportsbook is a dealer: it quotes a line, holds the risk, and profits from the customer’s losses, which is why every jurisdiction that licenses one treats it as an adversary of its own customers and regulates accordingly. An exchange is a venue: it matches two people who disagree, charges a fee, and holds nothing, which means it cannot shade the price against anyone and has no interest in who wins. That is not a legal fiction. It is the difference between a bank’s trading desk and a stock exchange, and every financial regulator on earth treats the two differently for good reason. The critics say the difference is irrelevant to the user because the user experiences a bet either way. I would say the difference is the whole point for the user, because a price set by thousands of disagreeing traders is a better estimate of the truth than a line set by a bookmaker with a margin to protect, and a person who wants to know whether a rate cut is coming or a government will fall is better served by the exchange, and the proof is that the exchanges have been more accurate than the polls and the pundits in every election they have priced.

Now the sports contracts, where the defence is weaker and I will concede what must be conceded. A contract on whether a quarterback throws for three hundred yards produces no information anyone needs; the sportsbooks already price the same event, the price is already public, and the social function is nil. The platforms moved into sports because sports is where the volume is, roughly fifty-three billion dollars in July by the industry’s count, and the volume is what justified an eleven-billion-dollar valuation and a cable news partnership, and I will not pretend that the information argument covers it. But I would make two points the critics skip. The first is that the sports contracts are also priced by an exchange rather than a dealer, which means the user gets a fairer price than the sportsbook offers, and in a country with thirty-odd state monopolies and tribal compacts the sportsbook prices are not fair, and the fifty-three billion dollars is partly a measure of how much the state-licensed books have been overcharging. The second is that the question of whether the sports contracts should be federal or state is a question about American federalism, and the three appellate courts that split on it this year, the Third one way and the Ninth and Sixth the other, were deciding a statutory question about the word swap, and I read Washington the way I read Pretoria, which is institutions first, and the institution that should decide is the legislature, which has instead produced a dozen bills and passed one Senate rule banning its own members from trading, which tells you what the members think the product is.

On risk and social impact, I will say what a consultant from a country with a serious gambling problem says. South Africa has legal sports betting, heavily advertised, taxed, and licensed, and it has not prevented a generation of young men from losing their wages on their phones; the licensing regime has mostly produced revenue for the provinces and a lobby for the operators. The American states that are suing to protect their regimes are protecting a tax base and a set of licensed incumbents, and they are dressing the protection in the language of consumer welfare, and I have seen that dress before. The genuine social risks of the prediction markets are two, and neither is addressed by state gambling law. The first is insider trading on events the trader controls or knows about, which the Maduro and Iran trades demonstrated and which the American regulator has begun to address with an advisory in February, the first criminal case in April, and proposed rules in June on war and assassination contracts, which is the right sequence and should go further. The second is the eighteen-year-old on a brokerage app, which is a problem of age and interface and advertising, and which a federal regulator could fix in an afternoon by rule, and has not.

So how should policymakers treat them? As markets, which they are, and as markets that need the rules every market needs: a ban on trading by people who control or have privileged knowledge of the outcome, enforced criminally, which has begun; an exclusion of contracts whose settlement could be caused by the trader, war and assassination and the like, which is proposed; an age floor and advertising rules appropriate to a product that is, for the sports contracts, functionally a wager; and a surveillance regime that reports suspicious trading to the authorities, which the sportsbooks have and the exchanges did not and the Maduro trade proved they must. That is a federal regime, because the markets are national and the information they produce is national, and the American states’ claim to regulate them is a claim to protect a monopoly, and I have no talent for monopolies that fail in public and then lecture the rest of us about norms.

My assessment is that prediction markets are meaningfully different from sportsbooks in structure, that the difference is real and favours the user on price and the public on information, that the sports contracts are a commercial extension that the information argument does not reach, and that the legal distinction is a regulatory workaround only in the sense that American federalism makes everything a workaround. The policymakers should keep the markets federal, write the rules a market needs, and stop pretending that fifty gaming boards defending fifty tax bases are the guardians of anyone’s welfare. The price on the Maduro contract moved six hours before the raid. A soldier is being prosecuted, as he should be. The price was right. In a country where the state power company lied to me for a decade, I have learned to trust the instrument that cannot.

Clermont, Florida Correspondent

I advise small professional firms, and I want to begin with the thing I say to every client who has discovered that a competitor is operating under a different set of rules, which is that the first question is not whether the competitor is cheating but whether the rules were ever about what they claimed. The states that are suing the prediction markets this year say they are protecting their citizens from unlicensed gambling. The states also run lotteries that pay out about sixty cents on the dollar, advertise them on billboards next to schools, and depend on the proceeds for budgets they could not otherwise balance; Florida’s lottery funds a scholarship program that half the families in this county are counting on. The states license sportsbooks and tax the margin and take a share of a product they describe, in every other context, as a social harm. So when a federally designated exchange arrives offering the same bets at a better price with no state cut, and the states respond with lawsuits in the language of consumer protection, a man who advises small businesses recognizes the shape of the complaint. It is the complaint of an incumbent whose moat has been crossed, and the moat was never the welfare of the people on the other side of it. It was the toll.

That is not a defense of the platforms. It is a clearing of the ground, because the question asks whether the legal distinctions are workarounds and the honest answer is that everyone in this fight is working around something. The platforms are working around state gambling law by calling a bet a swap and listing it on a venue the federal commodities regulator designated, which the Third Circuit accepted in April and the Ninth and Sixth rejected in August and September. The states are working around their own hypocrisy by treating a product they sell as a harm when someone else sells it. The federal regulator is working around its own mandate, which was written for commodity derivatives and never contemplated a market in whether a college quarterback throws for three hundred yards, by suing nine states to defend a jurisdiction it has not yet figured out how to exercise. And Congress, which could settle all of this with a statute, has passed a Senate rule barring its own members from trading, which is the clearest statement of what the members think the product is, and nothing else.

Now the structural question, which I think the small-business lens sees clearly. A sportsbook is a dealer that holds the risk and profits from losses; an exchange is a venue that holds nothing and profits from volume. My clients understand that distinction immediately, because it is the distinction between a shop that marks up inventory and a marketplace that charges a listing fee, and they know that the marketplace is usually the better deal for the buyer and the worse deal for the regulator, because the marketplace has no inventory to inspect and no margin to tax. The prediction markets are the marketplace model applied to wagers, and the consequence for the user is a fairer price and no one watching, which the Maduro trade demonstrated in January when a soldier with classified knowledge bought contracts on the ouster hours before the raid and the exchange paid him four hundred thousand dollars because nothing in its model gave it a reason to notice. A dealer would have noticed. A dealer loses money to the man who knows. The marketplace earned a fee and the Justice Department did the work in April. That is a real difference in structure and it is a real difference in risk, and it runs in the direction the critics say.

On social impact, I will offer the migration analyst’s observation, because this state receives the consequences of other states’ choices and has for decades. Florida gets a thousand new residents a day, a great many of them retirees from the Northeast and the Midwest who arrive with their phones and their habits, and the gambling habit is one of the habits that arrives. The state built its structure around a tribal compact that gives the Seminoles the sports-betting market, and the compact was the state’s way of saying it would permit the thing while keeping it in one hand it could negotiate with. The prediction markets have made the compact a monopoly on a product anyone can buy elsewhere, at eighteen, on a brokerage app, with no tribal or state oversight, and the retirees and their grandchildren have found the other door. The social cost of that is not a new kind of gambling. It is the old kind, delivered by a new channel that the people who were supposed to manage the cost do not control, and the cost will show up in the places it always shows up, which are the household budgets of people who cannot afford the habit, and the state will have no revenue from the channel to offset it.

So how should policymakers treat them? Here is what I would tell the client. Decide what the business is and regulate it as that business, and stop letting the name decide. The contracts on elections and economic data are information markets with a genuine public function and a long record; they belong under federal supervision, with rules against trading on privileged knowledge, which have begun with the February advisory and the April case and the June proposal on war and assassination contracts, and which should be made criminal and permanent. The contracts on sports are wagers in a derivative’s form, they produce no information the sportsbooks do not already produce, and they belong under the regime the states built for wagers, at twenty-one, taxed, with the reporting and exclusion lists, because the alternative is a national sports-betting market with no gambling rules and an eleven-billion-dollar incentive to keep it that way. That split is what the Ninth and Sixth Circuits reached by reading the statute, and it is what Congress should write before the Supreme Court decides a word. And the states should stop pretending their interest is the citizen when the citizen can see the lottery billboard from the courthouse steps.

My assessment is that prediction markets are structurally different from sportsbooks in a way that lowers the price and removes the watchman, that the legal distinctions are workarounds on every side of the fight, and that the policymakers should regulate the product by what it prices rather than by the venue that lists it. The platforms built a better marketplace and attached a gambling product to it. The states built a tolling system and called it protection. The federal regulator defended a jurisdiction it has not learned to use. And the retiree in Clermont with the brokerage app is betting on the Gators at eighteen-year-old rules through a federal door that nobody designed for him, and the Supreme Court will decide, probably next year, whether that door is a swap. I would tell the client that the door is a door, and that the people arguing about the word have all, in their own way, left it open.

Phoenix, Arizona Correspondent

I run operations for a hospital system, which means I see the end of every argument the policy people are having, in the waiting room, at two in the morning, when the argument has become a person. The gambling argument arrives there more than it used to. Arizona legalized sports betting in 2021, with the tribes and the professional teams sharing the market, and the behavioral-health intake forms in our emergency departments started showing a line that had not been there before, which was a young man, usually, with a debt he could not explain to the family that brought him in. This year the line has a new entry. The man has a brokerage app with a contract on it, and the contract is on a game, and he is eighteen or nineteen, which he could not have been at the sportsbook, and the app did not ask him the questions the sportsbook is required to ask, and he did not know he was gambling because the screen said he was trading. That is the social impact the question asks about, and I would like it on the record before the lawyers get to the word swap, because the word swap does not come to the emergency department. He does.

So let me take the structural question from the intake desk, where structure is the thing we are trained to look through. The platforms say an exchange is not a bookmaker because it matches traders and holds no position, and that is true, and in the waiting room it is irrelevant. The young man did not lose his money to the exchange; he lost it to another trader, through the exchange, which from his point of view is a distinction without a difference and from the exchange’s point of view is the entire defense. The one thing the structure does change is who was watching him, and the answer is nobody. A licensed sportsbook in this state is required to run self-exclusion lists, to verify age at twenty-one, to report irregular activity to the Department of Gaming, and to fund the problem-gambling programs that the state uses to pay for the treatment I then schedule. A federally designated contract market is required to have a rulebook, and the rulebook was written by a commodities regulator for traders in wheat and interest rates who are presumed to know what they are doing. The platforms are open at eighteen because derivatives are. The young man in my waiting room is a derivative trader under federal law. He is a kid with a problem under every other description, and the law chose the description that helped the business.

Now the legal fight, which reached Arizona directly. The state treated the sports contracts as unlicensed gambling, the federal regulator and the Justice Department sued Arizona on April 2 to stop the state from enforcing its own gaming laws, and a federal judge gave them a restraining order eight days later blocking a criminal case the state wanted to bring. I work in a system that lives on Medicaid billing and I know what it looks like when Washington tells a state how its own programs will run, and I also know that the people in Washington do not come to the waiting room afterward. The Ninth Circuit, which covers us, ruled in August that the sports contracts are not swaps and that the states may regulate them, which is the right answer and which will be appealed, and in the meantime the contracts are available on every phone in Maricopa County to anyone who has turned eighteen, because a restraining order protects the platform and nobody has a restraining order for the kid.

On the information argument, which the platforms’ defenders make and which I want to answer honestly. The markets on elections and economic data have a function; I read them myself, and they were better than the polls, and I would not take them away. The markets on whether the Suns cover have no function except the one they share with every sportsbook, which is to move money from people who do not know to people who do, with a fee for the house, and the people who do know this year have included a soldier with classified information about the Maduro operation who cleared four hundred thousand dollars before the government noticed, which tells you how much watching was being done. The hospital analogy is exact: a clinical trial produces knowledge and a casino produces revenue, and a regulator that cannot tell the difference between them has not understood either. The federal regulator proposed rules in June banning contracts on war and assassination, which is the minimum a person could ask, and has proposed nothing on the age or the interface or the advertising, which are the things that decide who ends up in my waiting room.

So how should the policymakers treat them? From the intake desk the answer is simple and I will say it plainly because that is what the heat and the drive on I-10 have taught me. Decide the age first. A product that is functionally a wager should not be available to anyone the state does not permit to wager, whatever the product is called, and the federal regulator could impose that tomorrow by rule and has not, which tells you whose comfort the regulator is protecting. Then decide who watches: the sportsbooks report irregular trading to a state regulator because they are required to, and the exchanges should be required to report it to someone, because the Maduro trade showed what happens when no one is. Then decide who pays for the harm, because Arizona’s problem-gambling fund is financed by a share of licensed revenue and the federal contracts contribute nothing to it, and the young man in the waiting room is treated on Medicaid, which is to say by the taxpayers of a state the platforms are suing. The question of whether the contracts are swaps can be argued in Washington for a decade. The questions of age, surveillance, and who pays can be answered by anyone with the authority and the will, and nobody with the authority has shown the will.

My assessment is that prediction markets are structurally different from sportsbooks in a way that removes the watchman and lowers the age, that the structural difference is real and runs entirely against the person using the product, and that the legal distinction has become a workaround in the precise sense that it delivers a wager to an eighteen-year-old on a brokerage app under rules written for grain. The policymakers should treat the sports contracts as what the Ninth Circuit said they are and regulate them at the state level with the state’s rules, keep the information markets federal because they are useful, and fix the age and the surveillance by federal rule in the meantime because those do not require anyone to win the swap argument first. The waiting room does not care about the Commodity Exchange Act. It cares who was watching the kid, and the answer, this year, was an algorithm that charged him a fee.

Dayton, Ohio Correspondent

I am going to take the engineering view, which means ignoring what the thing is called and asking what the thing does, and I want to begin by saying that the question of whether a prediction market is a derivative or a wager is a question about a label on a machine, and a process engineer learns early that the label tells you what the designer intended and the machine tells you what it does. A binary event contract is a simple machine. It pays one dollar if a stated event occurs and nothing if it does not, it trades between zero and one, and its price at any moment is the market’s estimate of the probability. That machine has been used for decades to price elections, interest-rate decisions, and inflation releases, and in those uses it does something a sportsbook does not, which is aggregate dispersed knowledge into a public number that is often better than the experts. The same machine, pointed at whether a running back gains a hundred yards, does nothing the sportsbook’s line does not already do, and the only reason to build it is that the machine is federally licensed and the sportsbook is not. The question asks whether the two are meaningfully different. The machine is the same. The input is different, and the input is the whole fight.

Let me say why the input matters, as a man who builds controls for a living. A control system is only as good as the thing it measures. When a prediction market measures a presidential election, the inputs are millions of people’s dispersed judgments about a thing none of them controls, and the output is useful because no single participant can move the outcome. When it measures the Maduro operation, the inputs include a soldier with classified knowledge who bought contracts hours before the raid and cleared four hundred thousand dollars, and the output is still useful, in a grim way, because it told the world something was happening, but the participant could not have moved the outcome either; he merely knew it. When it measures a football game, the inputs include people who can move the outcome, which is the players and the officials, and the history of sports gambling is the history of what happens when the people who can move the outcome are permitted to trade on it. The platforms have rules against that. So did every sportsbook that ever got burned, and the sportsbooks have a thing the exchanges did not until this year, which is a direct line to the leagues and the state regulators to report a line that moved for no reason. The exchange model, which holds no position and profits from volume, has no engineering reason to care why the line moved. That is the structural difference, and it is a sensor that was left off the design.

Now the Ohio side, because the state was one of the two that won in the Sixth Circuit two weeks ago, and I have watched it from a plant that sits in a county where sports betting has been legal since 2023. Ohio’s casino regulator told Kalshi in March of last year that its sports contracts were gambling under Ohio law and needed a license; Kalshi said the contracts were federal swaps and Ohio could not touch them; a federal judge in Ohio agreed with the state, a judge in Tennessee agreed with Kalshi, and the Sixth Circuit, on September 25, agreed with the state in both cases, holding that Kalshi had not shown its sports contracts met the statutory definition of a swap. The Third Circuit held the opposite in April. The Ninth agreed with the Sixth in August. That is three circuits reading the same statute and splitting on what a swap is, and an engineer reads a split like that as a specification failure: the people who wrote the statute in 2010 did not contemplate this input and the definition does not resolve it, and the Supreme Court will have to write the missing line, and a court is a poor place to write a specification.

The technical education angle, because it is mine and because the robotics kids I coach are now old enough to have brokerage accounts. The contracts are listed on the same apps as stocks and index funds, by Robinhood and Coinbase and others, at eighteen, with a sign-up flow built for a derivatives trader, and the kids experience them as the natural extension of a thing they were told was investing. That is the design decision that matters most and that nobody litigating the swap question is talking about. A sportsbook app is designed to look like what it is. A brokerage app with event contracts is designed to make a bet on a game feel like a position in a market, and the interface is the thing a kid actually interacts with, and the interface was built by the people who profit from the volume. I spent years in quality control. A product whose interface is designed to obscure its function from the user would not pass inspection in my plant regardless of how it was labeled on the regulatory form.

So how should policymakers treat them? The engineer’s answer is to regulate the input, not the label. Contracts on events that no participant can control or know in advance, elections, economic data, policy decisions, are a useful machine and should stay under a federal regulator that can supervise a national market, with the rules against trading on privileged knowledge that began with the February advisory and the April indictment and should be made permanent and criminal. Contracts on events that participants can control, which is every sporting event, should be regulated as the sportsbooks are, with the age, the surveillance, the league reporting, and the state oversight, because the engineering risk is the same and the sportsbooks’ controls exist for a reason. Contracts whose settlement a trader could cause, war and assassination and the like, should not exist, which is what the regulator proposed in June and should have proposed in 2025. And the interface should be required to say what the product is, which is a requirement so basic that the only reason it has not been imposed is that the regulator is a commodities agency that has never had to think about a user.

My assessment is that prediction markets and sportsbooks are the same machine with different inputs, that the structural difference the platforms claim is real for the operator and removes a safety sensor for the user, and that the legal distinction became a workaround at the exact moment the input changed from elections to touchdowns. The policymakers should write a specification by input and stop waiting for a court to define a word that a 2010 statute never meant to cover. The announced policy is that these are regulated derivatives. What reaches the shop floor, which in this case is a kid’s phone, is a bet on a game with a brokerage logo on it, at eighteen, with nobody watching the line. I have seen a lot of machines run outside their design envelope. They work, for a while, and then they do not, and the failure is never in the label.

Savannah, Georgia Correspondent

Georgia has no legal sports betting. The legislature has debated it every session for eight years, the professional teams have lobbied for it, the polling has favored it, and every year the bill has died somewhere between the Senate and the House over the question of whether the proceeds go to the HOPE scholarship or somewhere else, and whether it needs a constitutional amendment, and whether the Baptists in the rural districts will tolerate it, and so the largest state in the Southeast has spent nearly a decade watching its residents drive to Tennessee and Florida and the Carolinas to place bets that Georgia could have taxed. I read city-council packets the way other people read box scores, and I have read the legislature’s packets on this subject for years, and I want to begin there because the question of whether prediction markets are a regulatory workaround has, in Georgia, an answer that the states suing the platforms would rather not hear: the platforms worked around a regulation that did not exist, because the state could not agree to write one, and the thing the state is now losing is a tax base it never collected.

Let me describe what has happened here, because it is different from the states with legal books and it is the case the national coverage skips. The prediction markets arrived in Georgia in 2025 as federally designated contract markets, available on brokerage apps, offering contracts on every Bulldogs game and every Falcons game and every Braves game, at eighteen, with no state license because the state has no license to issue. A Georgian who wanted to bet on football legally has, since then, had exactly one lawful option, and it is Kalshi or Polymarket or Robinhood, and the state’s position is that this is illegal gambling, and the state’s position is also that it will not legalize gambling, and the two positions together mean that the state is asserting a right to prohibit a thing it has declined to regulate. In a berth schedule that is called holding a slot you do not intend to use, and the port charges demurrage for it. Georgia is paying the demurrage in the form of a sports-betting market that exists, is used, is taxed by nobody, and is supervised by a federal commodities agency that has never been to Atlanta.

Now the structural question, as a port planner reads it. A sportsbook holds the risk and profits from losses; an exchange matches traders and profits from volume. The difference is real for the operator and I will not dismiss it. But a planner looks at flows rather than at who owns the terminal, and the flow in both cases is money from people who do not know to people who do, with a cut to the house, and the flow on the exchange is unwatched in a way the flow at a licensed book is not, which is how a soldier with classified knowledge of the Maduro raid turned thirty thousand dollars into four hundred thousand on an offshore exchange in January without anyone noticing until the newspapers did. A licensed book in Tennessee would have frozen that account before settlement, because the book’s regulator requires it and the book’s own money was at stake. The exchange paid, because nothing in its design cared. That is the structural difference the question asks about, and it is the one that matters to a state deciding whether to let the thing operate: not who holds the risk, but who watches the water.

On social impact, I will say what the tourism side of my work tells me, because Savannah lives on visitors and visitors bring their habits. The city fills with football fans every fall, from the Georgia game in Jacksonville to the bowl season, and the fans have the apps, and the apps do not care whether Georgia has a law, and the bars on River Street have learned that a table of men watching a game on their phones is a table of men trading contracts, and the bartenders have learned the vocabulary. The social cost of that is the same cost every state with legal betting has absorbed, the young men and the debts and the families, and Georgia is absorbing it with none of the structure the legal states built: no problem-gambling fund, no self-exclusion list, no age verification above the federal eighteen, no advertising rules, no revenue. The state has the harm and not the mitigation, which is the worst position on the board, and it arrived there by refusing to choose.

So how should policymakers treat them? Here is the answer from a state that has not: decide. The contracts on elections and economic data are a federal product with a real function and should stay federal, with the insider-trading rules that began this year made permanent. The contracts on sports are a wager, as the Ninth and Sixth Circuits held this fall and the Third did not, and a state that wants to regulate them as wagers has to have a regulation to apply, which means Georgia has to pass the bill it has been failing to pass since 2018, with the age at twenty-one, the tax, the fund, and the reporting, and then it can tell the platforms to comply or leave. A state that prohibits without regulating is not protecting anyone. It is holding a slot, and the federal ships are docking in it, and the Supreme Court will decide next year whether the slot was ever the state’s, and if the Court sides with the Third Circuit, Georgia will have spent eight years failing to legalize a product that will then be legal everywhere, federally, with no Georgia rules at all.

My assessment is that prediction markets are structurally different from sportsbooks in a way that lowers the price and removes the watchman, that the legal distinction is a workaround only where there was a regulation to work around, and that in Georgia there was not, which makes the state’s lawsuit a complaint about traffic through a gate it never built. The policymakers in Washington should split the product by what it prices. The policymakers in Atlanta should pass the bill and stop pretending that a refusal to regulate is a form of regulation. I walk the riverfront before the heat and I watch the ships come in under schedules negotiated by people who understood that an unassigned berth is not a protected one. Georgia’s berth has been unassigned for eight years. Something has docked. The question now is whether the state will write the terms or keep insisting it did not give permission, which is a sentence the port stopped accepting from shippers a long time ago.

Bismarck, North Dakota Correspondent

I am going to disagree with the answer I expect most people will give to this question, which is that the sports contracts should go to the states and the rest should stay federal, and I am going to do it from compliance, because the answer sounds right and does not work and I have spent a career watching rules that sound right fail on the floor. North Dakota has a legal gambling industry that most of the country has never heard of: charitable gaming, pull tabs and electronic pull tabs in bars and clubs, which generated something over two billion dollars in wagers last year in a state of eight hundred thousand people and which the legislature spent the last session fighting over because the electronic machines have turned every bar into a small casino. We have no sportsbook. The tribes run the casinos. The attorney general has been asked what the state’s position on the prediction markets is and has, so far as I can tell, not taken one, because the state’s gaming structure was not built with any of this in mind and the people who run it would rather not find out what a federal court thinks of pull tabs. I begin there because the question asks how policymakers should treat these markets, and the honest answer from a state with a gambling regime that fits no template is that the template is the problem.

Straight talk about the structure first. An exchange that matches traders and takes a fee is not a bookmaker that holds the risk, and the platforms are right about that and the people who say it does not matter are wrong, because it matters enormously to who is supervising what. A bookmaker is a counterparty, and you regulate a counterparty by licensing it, capitalizing it, and auditing its book. An exchange is a venue, and you regulate a venue by setting the rules of the contracts it may list and the conduct of the people who trade on it. Those are different regulatory jobs, and the state gaming commissions know how to do the first and have never done the second, and the federal commodities regulator knows how to do the second and has never thought about the people doing the first, and the whole fight this year, three circuit courts splitting on whether a sports contract is a swap, is a fight about which regulator gets a job neither is built for. In my work, when two agencies both claim jurisdiction over a thing and neither has the staff or the rules for it, the thing goes unregulated while they litigate, and that is exactly what has happened, and the proof is that a soldier bought contracts on the Maduro operation hours before the raid in January and nobody with authority noticed until the press did.

Now why the split answer fails. The proposal is that elections and economic data stay federal and sports go to the states. It is clean on paper. On the floor it requires a single app on a single phone to apply fifty different regimes to one product line, to verify a user’s location and age against fifty different standards, to report suspicious activity to fifty different gaming boards for the sports contracts and to one federal agency for everything else, and to do all of it while the same user trades both kinds of contract in the same session with the same money. I have implemented compliance programs across state lines. That design does not get implemented. It gets lawyered, and the lawyering produces exactly what we have now, which is platforms that comply fully in the states that won their cases, operate under injunction in the states that are still fighting, and treat the whole thing as a cost of doing business while the volume, fifty-three billion dollars in July, goes wherever the rules are thinnest. The states that want their sports-betting regimes protected are asking for a world in which a national product is governed by local law, and that world has never existed for any national product, and it will not exist for this one.

So here is the compliance officer’s answer, and it will please neither side. The product is national, the regulator should be federal, and the federal regulator should be made to adopt the rules a gambling regulator uses, by statute, because it will not adopt them on its own. That means an age floor of twenty-one for any contract on a sporting event, imposed by rule, which the agency could have done in 2025 and has not. It means mandatory surveillance and reporting of irregular trading to a named federal body with the authority to freeze settlement, which the sportsbooks do for their state boards and the exchanges did not do for anyone until the Maduro case embarrassed them into cooperating. It means a federal excise on sports contracts at a rate comparable to what the states collect, shared with the states, so that North Dakota is not subsidizing the treatment of a habit the federal government licensed. It means the June proposal banning contracts on war and assassination made final, and the February advisory on insider trading made criminal and permanent, which the April indictment began. And it means the interface says what the product is, because a kid who thinks he is investing is not making an informed decision, and informed decisions are what the whole federal theory of derivatives regulation assumes.

I want to say the plain thing about the states, because I worked in state government and I know what their lawsuits are about. The states with legal books are protecting a tax base and a set of licensed incumbents. The states without legal books, like Georgia and like us, are protecting a prohibition they have not been able to turn into a regulation. Both are legitimate interests and neither is the consumer’s. The federal regulator, for its part, has spent the year suing nine states to protect a jurisdiction it has not learned to exercise, and the Senate, which could settle this, has passed a rule banning its own members from trading, which is the most honest thing Congress has said about the product: the people who know the most about it do not want to be caught using it. That is a tell. A compliance officer reads tells.

My assessment is that prediction markets are structurally different from sportsbooks in a way that changes which kind of regulator they need and does not change what the user is doing, that the legal distinction is a workaround only because the law was written in 2010 for a product that did not exist, and that the policymakers should stop fighting over which agency owns the product and start writing the rules the product needs, at the federal level, with the gambling rules imported wholesale for the contracts that are functionally wagers. The split by underlying is the fashionable answer and it does not survive contact with a phone. The federal answer with teeth is the unfashionable one and it is the only one that can be implemented, and the fact that it requires the federal regulator to admit it is running a gambling product is the reason it has not been proposed. Somebody should propose it. The pull-tab machines in Bismarck’s bars are better supervised than the contracts on the phones of the people playing them, and that is a finding, and I will put it on the record.