In American antitrust, clearing the federal gate increasingly means arriving at the state turnstiles.
State attorneys general play a valuable role when harms are local or federal investigators miss key facts. But serial challenges to nationally integrated conduct turn that safeguard into a standing invitation to relitigate.
The result is a system in which no settlement fully settles and no clearance fully clears. That weakens federal enforcement, rewards the most aggressive jurisdiction, and lets local politics reshape national markets.
State antitrust should supply evidence, police local misconduct, and check genuine federal failure. It should not operate as an endless appeals process with 51 possible vetoes.
When Federalism Becomes a Veto Stack
American antitrust federalism has real virtues. State attorneys general can uncover local cartels, represent residents whose individual claims may be too small to pursue, and alert federal officials to facts a Washington-centered investigation might miss. States can also provide a useful check when federal enforcers overlook a genuine competitive harm.
Those benefits do not require subjecting every national transaction or business practice to 51 independent and potentially conflicting antitrust policies.
The growing problem is sequential contradiction. The U.S. Department of Justice (DOJ) or Federal Trade Commission (FTC) investigates a national matter, settles it or closes the investigation, and a coalition of states then tries to keep the same transaction or conduct in court. The firm has not received a federal answer. It has merely cleared the first gate.
This “veto stack” turns federalism from a source of evidence into a mechanism for repeated litigation.
In an earlier Truth on the Market commentary, I argued for a sounder division of labor (see also related work by Babette Boliek). State enforcement should complement well-grounded federal initiatives and focus scarce resources on competitive harms particular to a state.
That account also recognizes that public-choice pressures and rent seeking can distort decisions at both levels of government. The point is institutional, not partisan. The proper allocation of authority is the one that best reduces the combined costs of anticompetitive conduct and mistaken intervention.
Federal Clearance, State-Sized Asterisk
Recent litigation reveals several forms of state-federal divergence. They should not be forced into a single narrative because each raises a different institutional concern. Together, though, they suggest that federal clearance increasingly comes with an asterisk.
Live Nation offers the strongest warning against claiming that continued state litigation is always wasteful. The DOJ announced a proposed settlement shortly after trial began, but most state plaintiffs stayed in the case and won a jury verdict.
One could still view that outcome as a policy error. Live Nation argues that combining promotion, venue operations, and Ticketmaster’s ticketing services creates an efficient vertical integration that lowers costs. But even assuming the jury reached the right result, one verdict cannot resolve the broader institutional question.
Firms bear the ex ante costs of overlapping authority in every case, including those in which a second prosecution produces no offsetting consumer benefit. Conflicting remedies can also persist long after a court determines liability.
HPE/Juniper raises a different concern. The DOJ first challenged the $14 billion acquisition, then settled for the divestiture of HPE’s Instant On business and licensing commitments covering Juniper’s artificial-intelligence (AI) operations source code.
The states did not file a separate Clayton Act suit. Instead, 12 states and the District of Columbia intervened in the Tunney Act proceeding, which allows courts to review federal antitrust settlements, and urged the court to reject or closely examine the decree.
John Yun’s analysis focused on a key aspect of competition to which the original complaint gave too little weight. Combining Juniper’s AI-powered networking capabilities with HPE could create a stronger rival to Cisco and increase competition. Greater international competitiveness could also advance national security interests.
The state intervention extended uncertainty over a completed integration whose competitive rationale depended on AI networking, scale, and the prospect of a more credible third platform.
Nexstar/Tegna and Paramount Skydance/WBD go further. In Nexstar, state plaintiffs sued after federal regulators approved the deal and the transaction closed.
In Paramount, the DOJ publicly ended an eight-month investigation after concluding that the asserted theories appeared to confuse harm to competitors with harm to consumers. A 12-state coalition filed a federal antitrust suit the next month.
The merits of those challenges remain distinct. The institutional message is harder to miss. A federal resolution may merely open the next round.
The High Cost of One More Veto
Antitrust enforcement is an exercise in managing error costs. False negatives allow market power to persist. False positives deter conduct that lowers costs, improves products, or creates new competitive constraints. The goal should be to minimize the expected social costs of both errors, along with the administrative and compliance costs of enforcement—not to maximize cases, remedies, or the number of officials with a vote.
Sequential state prosecution increases those costs in at least four ways.
First, it erodes the settlement value of federal enforcement. A firm cannot reliably exchange concessions for closure when non-settling states remain free to pursue conflicting relief. That uncertainty raises the price of settlement, weakens the government’s ability to secure efficient remedies, and encourages firms to litigate defensively.
Second, it delays integration. Merger efficiencies often depend on timing. Engineers leave, product road maps diverge, capital budgets freeze, and customers postpone adoption while a transaction’s legal status remains uncertain.
Third, a veto stack magnifies uncertainty. The relevant business question becomes whether any politically prominent jurisdiction might attack the conduct under a more elastic standard, rather than whether the conduct satisfies a coherent consumer-welfare rule. The most aggressive enforcer then determines expected liability, regardless of which agency has the best information.
National companies respond predictably. They design their conduct around the strictest state rule, abandon marginal investments, or avoid business models whose legality requires a fact-intensive assessment of efficiencies.
Fourth, fragmented enforcement creates remedy externalities. A state may seek divestiture, access mandates, data sharing, compulsory licensing, or conduct restrictions whose costs fall largely beyond its borders. State officials receive political credit for a visible intervention, while consumers, workers, and shareholders nationwide absorb the lost scale or innovation. The jurisdiction making the decision does not bear the full national cost.
These risks grow sharper in high technology. Dynamic competition depends on uncertain investments, complementary assets, learning effects, and rapid quality improvements that static market shares often measure poorly. My dynamic-consumer-welfare approach therefore focuses on innovation and long-run consumer gains.
The International Center for Law & Economics’ (ICLE) work on state laws with national spillovers makes a related point. State experimentation becomes costly when one jurisdiction exports the consequences of its policies to the interstate economy. Requiring every state to grow comfortable with an unfamiliar technology before firms may proceed will predictably protect incumbents and punish experimentation.
When ‘AG’ Means ‘Aspiring Governor’
Why do national technology firms make such tempting targets? Public-choice analysis offers a simple answer without accusing anyone of corruption or bad faith.
State attorneys general are elected officials or politically prominent generalists. Large technology cases generate headlines, coalition leadership, fundraising material, and a platform for higher office. Competitors, organized interest groups, and plaintiffs’ lawyers also have strong incentives to provide complaints, studies, and political support.
The beneficiaries of innovation are harder to organize. Future customers, small businesses, and users who value free or improved services rarely form a ready-made constituency.
Empirical research cannot prove that any particular case rests on political motives. It can, however, identify the incentives built into the system. Colin Provost examined when “AG” becomes shorthand for “aspiring governor” and found a relationship between multistate litigation and later campaigns for higher office. John Dove’s work on state antitrust enforcement likewise considers electoral cycles and institutional political pressures.
Those findings counsel restraint. Broader discretionary standards do not produce neutral public-interest enforcement by magic. They increase the rewards for political entrepreneurship.
Technology markets sharpen the problem. Their products are complicated, their alleged harms translate easily into populist rhetoric, and their benefits often appear as better quality, lower latency, stronger security, faster model development, or future competition rather than lower prices today.
A prosecutor can point to size, data, vertical integration, or an unhappy rival. Showing that integration improved performance or created a stronger competitor takes more work. The political reward arrives when the complaint is filed. The innovation lost to delay rarely makes the evening news.
Fifty-One Rulebooks for One National Market
Conflicting enforcement becomes more dangerous when states also apply different substantive rules.
For roughly four decades, the consumer-welfare framework has disciplined antitrust by asking whether challenged conduct is likely to reduce output, raise quality-adjusted prices, suppress innovation, or otherwise harm competition. The framework does not grant firms laissez-faire immunity. It separates competition law from a broader power to redistribute economic advantage among firms and political constituencies.
Several state initiatives point toward wider and less predictable mandates. New York’s Twenty-First Century Antitrust Act, which passed the state Senate in May and remains pending in the Assembly, would create an abuse-of-dominance regime, restrict certain vertical conduct, expand premerger requirements, and authorize private lawsuits. My critique of an earlier version warned that such a standard could protect competitors at consumers’ expense and interfere with national policy.
California’s pending AB 1776 would expand single-firm liability under the Cartwright Act. Babette Boliek’s recent ICLE critique argues that the bill would replace objective consumer-welfare analysis with protection for favored market participants. New Jersey’s S451 would make certain uses of algorithmic rent-coordination systems violations of the New Jersey Antitrust Act.
These proposals differ in scope and may change or fail. Together, they show growing support for state-specific theories that could govern conduct far beyond state borders.
The expertise concern is comparative, not absolute. Many state antitrust lawyers are excellent, and federal agencies make mistakes. But the DOJ and FTC employ larger teams of economists, technologists, merger specialists, and lawyers focused on national markets. They also retain more institutional knowledge from repeated investigations.
A state coalition formed for a prominent case may rely on federal evidence and outside counsel while pursuing a different political objective. When a dispute concerns AI infrastructure, cloud architecture, enterprise networking, algorithmic pricing, or multisided platforms, analytical mistakes become especially costly.
Broad fairness or dominance standards also invite rent seeking. Rivals can repackage requests for protection as antitrust complaints. Regulated intermediaries can seek rules that preserve their position. Local interests can impose costs on consumers nationwide.
Recent Truth on the Market commentary has described this tendency as the “nanny state” substituting regulators’ preferences for demonstrated consumer harm. The law & economics objection does not dismiss every nonprice concern. It simply observes that a standard detached from measurable competitive effects gives too much discretion to officials who face uneven political incentives.
Congress Holds the Preemption Pen
Reform cannot assume that Washington merely lends states their antitrust authority. States may sue under their own statutes, and Congress has expressly authorized state attorneys general to bring parens patriae actions for damages under federal antitrust law. These actions allow a state to seek recovery on behalf of its residents.
In California v. ARC America, the Supreme Court held in 1989 that federal limits on indirect-purchaser recovery did not preempt state laws permitting such claims. The Court reasoned that federal antitrust law did not occupy the entire field and that federal policy alone did not dictate what remedies states could authorize.
ARC America therefore poses a serious obstacle to casual claims of implied preemption. But the decision does not require Congress to preserve overlapping state remedies forever. It interpreted the statutes Congress had enacted.
Congress may regulate interstate commerce and expressly preempt conflicting state rules, so long as it speaks clearly and respects constitutional limits against commandeering state governments. The cleaner approach would establish a federal rule for private conduct and transactions in national markets, backed by an express preemption clause and a carefully drawn exception for genuinely local enforcement. Congress would govern private conduct rather than tell states how to govern.
Courts have less room to act without Congress. Modern dormant Commerce Clause doctrine generally resists invalidating nondiscriminatory state laws merely because they affect conduct beyond the state. The Supreme Court reinforced that caution in National Pork Producers Council v. Ross.
Conflict preemption remains possible when compliance with both state and federal law is impossible or when state law obstructs a clearly stated federal objective. Yet ARC America makes that objective difficult to infer from antitrust policy alone. Judicial reform would therefore require a clearer signal from Congress or a modest doctrinal shift toward giving nationally uniform federal resolutions greater finality.
How to Defuse the Veto Stack
The answer is not to push states out of antitrust enforcement. States can uncover local harms, contribute evidence, and challenge a plainly deficient federal process. But nationally integrated conduct needs a federal resolution that carries more weight than an opening bid.
A workable reform agenda should preserve state participation while limiting contradictory remedies and serial litigation. Congress can clarify when federal decisions preempt state action, courts can give greater weight to reasoned federal assessments, and federal agencies can improve coordination and transparency. AI markets make those reforms especially urgent because even a nominally local injunction can alter a national product.
Make Federal Resolution Mean Actual Resolution
Congress should make federal antitrust resolutions final for conduct or transactions with substantial, nonincidental effects across state lines. Once the DOJ or FTC completes a defined review and issues a reasoned disposition, federal substantive standards should govern exclusively.
A qualifying consent decree, litigated judgment, or published closing statement could preempt later state actions seeking inconsistent prospective relief. States should remain free to participate in the federal investigation, submit evidence, seek compensation for residents, and challenge conduct whose competitive effects are predominantly intrastate.
The dividing line should turn on function, not size alone. Relevant factors could include nationwide pricing or product design, a multistate network, integrated data or infrastructure, or a remedy that would necessarily alter operations nationwide.
A savings clause should preserve state cases involving local bid rigging, regional cartels, occupational restraints, hospital or funeral-home conspiracies confined to one state, and small mergers whose assets and customers are overwhelmingly local. The governing principle should be subsidiarity, which assigns authority to the lowest level of government capable of accounting for the relevant costs and applying adequate expertise.
Preclusion should apply only after a transparent federal process. The DOJ or FTC should explain the theories investigated, the competitive evidence considered, and why the disposition protects consumers.
States could receive a brief consultation period and limited judicial review of whether the statutory conditions were satisfied. They should not receive a second trial on the merits. Congress could also require periodic reports on matters in which a federal resolution displaced a proposed state action.
Police Conflicting Remedies
Even without full field preemption, Congress could direct courts to displace state remedies that materially frustrate a federal resolution of national conduct.
A state damages claim based on a distinct local injury may coexist with a federal decree. A state injunction requiring divestiture after the DOJ accepted a licensing remedy presents a different problem. The question should be whether the state remedy conflicts with the federal resolution, not whether state officials merely disagree with it.
Courts should ask whether the requested relief would undo the settlement bargain, destroy efficiencies the federal disposition preserved, or force changes to a nationwide operating plan that federal enforcers approved.
Courts should remain cautious without becoming passive when Congress has not spoken. If a state brings a federal antitrust claim after the DOJ or FTC issues a detailed competitive assessment, courts should give that assessment substantial weight on market definition, efficiencies, and remedy design.
That approach would preserve state standing while reducing the anomaly of one sovereign obtaining federal relief that defeats an expert federal agency’s stated national policy.
Put the Federal Case on the Record
The most immediate reform requires no new statute. The DOJ already files statements of interest in private and state litigation. Both the DOJ and FTC should adopt a published protocol for intervening when a state case threatens a federal settlement, relies on a theory the federal agency rejected, or seeks a remedy with substantial interstate spillovers.
The filing should set out the federal consumer-welfare analysis, explain the national costs of conflicting relief, and address technical questions that a generalist court might otherwise hear only through dueling experts.
Such filings would be especially useful in Tunney Act proceedings. States should have an opportunity to be heard, but the public-interest review should not become a chance to retry the government’s entire case or replace a federal settlement with a coalition’s preferred remedy.
A clear federal account of uncertainty, litigation risk, innovation benefits, and remedial tradeoffs would help courts distinguish genuine defects from political disagreement.
Make States Explain the Encore
Federal officials should say plainly that cooperation is welcome, but serial prosecution of nationally integrated conduct carries real economic costs.
Congressional hearings could require federal and state enforcers to explain their competing theories, staffing, outside-counsel arrangements, and estimates of interstate spillovers. Greater transparency would force state officials to defend both the alleged harm and the added value of a separate action after federal review.
The DOJ and FTC could also condition joint investigations on ex ante coordination agreements governing information sharing, settlement consultation, and remedy design. A state would remain sovereign and free to withdraw. It should not be able to use the federal investigation, then surprise the parties with a conflicting national remedy.
Published best practices could encourage state offices to conduct cost-benefit analysis and reserve independent action for demonstrable local harm or a clearly deficient federal process.
Give AI One National Rulebook
AI-related conduct will test this system first. Training, cloud computing, chips, networking, data centers, model distribution, and enterprise deployment cross state lines by design. A state injunction governing model access, data use, interoperability, or vertical integration could reshape a national product even when issued under a nominally local statute.
Congress could establish expedited federal review and presumptive federal primacy for AI cases with substantial interstate effects. States would retain their traditional authority over fraud, privacy, contracts, and local cartels.
Federal primacy should preserve vigorous enforcement against exclusion, collusion, and acquisitions likely to harm consumers. The error-cost analysis must also account for dynamic global competition.
As Mario Zúñiga recently argued, antitrust accounts that omit China ignore a major rival in AI development. HPE/Juniper likewise shows how integrating AI-enabled networking can create a stronger challenger. The United States will struggle to out-innovate strategic competitors if every national technology investment depends on satisfying the most economically untethered state theory.
Federal Primacy, With Guardrails
The strongest objection is that federal enforcers can be captured, politicized, or simply wrong. Live Nation’s jury verdict gives that concern real force. A system that turned any cursory federal closing letter into blanket immunity would merely trade one institutional failure for another.
Federal primacy should therefore depend on a credible process. Federal enforcers should conduct an adequate investigation, provide a reasoned public explanation, disclose the logic behind any remedy, and submit consent decrees to judicial review where required.
Preemption also should not erase state damages claims on behalf of residents where federal law expressly preserves them, unless Congress clearly chooses otherwise. Prospective injunctions and efforts to block transactions create the largest national spillovers and present the strongest case for uniform federal treatment.
States should retain broad authority to compensate localized injuries, prosecute local conspiracies, and participate in federal cases. The governing principle is matching jurisdiction to the geographic scope of the market and the remedy, rather than centralizing authority for its own sake.
A more disciplined federalism could also improve state enforcement. State offices would have more resources for procurement cartels, local licensing restraints, unlawful noncompete agreements, health-care consolidation with genuinely regional effects, and anticompetitive state regulation. In those matters, local knowledge offers a genuine advantage rather than a vehicle for imposing one state’s policy nationwide.
Federalism Without the Endless Encore
Antitrust federalism should improve information and accountability. It should not ensure that a national firm can never obtain repose.
When states continue litigating after a reasoned federal settlement or clearance, the costs extend far beyond legal fees. Integration stalls, investment falls, remedies conflict, settlements lose value, and unfamiliar innovations face a built-in handicap. Those costs hit hardest in AI and other technologies where speed, scale, and experimentation shape long-run consumer welfare and national competitiveness.
The answer is neither abolishing state antitrust nor treating federal expertise as infallible. Congress should preserve state authority over genuinely local restraints while restoring federal primacy over conduct with substantial interstate effects. Courts should police conflicting remedies when federal policy is clear. The DOJ and FTC should coordinate earlier, file statements of interest, and explain when state actions threaten consumer welfare. Congressional oversight should expose the incentives behind serial prosecution.
A national market cannot function efficiently with 51 final decision makers. States should lead where the harm and expertise are local. Federal enforcers should answer for markets and remedies that are national.
Federalism should divide the work—not multiply the vetoes.
