Judge Spares Google Ads Business From Breakup
This matters for competition in the online advertising market, affecting Google's business operations, its competitors, and companies that buy and sell digital ads.
The remedy Judge Brinkema has fashioned does not remove the seller who dominates the marketplace; it merely posts new rules at the entrance and expects the crowd to change its habits. The intervention leaves the price of intermediation in Google’s ad exchange roughly where it was, but it moves the terms on which rivals can compete for the auction. Demand from publishers and advertisers for cheaper, more transparent matching services should rise, since that was the harm the Justice Department alleged. The supply response is the question the planners have not fully answered: will rivals actually enter and expand output, or will Google’s existing scale simply absorb the new constraints the way a large river absorbs a new tributary without changing its course?
Well, Judge Leonie Brinkema looked at Google’s advertising business, which the Justice Department spent years arguing was rigged about six ways from Sunday, and decided the proper punishment was to ask Google to behave better. I suppose that makes sense if you don’t think about it too long, which is probably the idea.
Now, understand, this was not a small case. The government didn’t accuse Google of running a slightly unfair lemonade stand. They accused it of owning the market where the lemonade stands buy their lemons, the market where they sell their lemonade, and the auction house that decides who gets which lemon and at what price, and then made all three shake hands and agree everything was fair. That is not a monopoly, that is a hand of solitaire where you’re also the dealer, the house, and the guy keeping score. And the remedy for this elaborate arrangement is a set of behavioral rules, which is a bit like catching a fellow dealing himself aces from the bottom of the deck and, rather than taking away the deck, asking him to deal a little slower where you can see his hands.
On the third floor of a shrinking newsroom in Ohio, a woman named nobody-you’ve-heard-of sells the ad space that keeps the lights on. She does not set the price of that space. She does not choose which exchange her ads run through, or how the auction is rigged, or who takes the cut before the paper sees a dime. She logs into a system built by Google, follows rules written by Google, and waits to see what’s left over. That is the worker this ruling touches, even though her name will never appear in a headline about it.
The institution responsible here is the federal antitrust apparatus - the Sherman Act’s rational-legal machinery, administered through the Eastern District of Virginia and now resting in the hands of Judge Leonie Brinkema. It was designed a century ago to dissolve concentrations of economic power crudely, by decree of dissolution. It has been asked, in 2026, to discipline something far more intricate: an advertising exchange whose value lies not in factories or pipelines but in the coordination of millisecond auctions across a technical architecture only a few hundred engineers fully understand. Assess the gap.
Alfred Marshall
The intervention moves the price of digital advertising space in one direction. But supply will respond by consolidating market power through technical barriers, and demand will respond by fragmenting into walled gardens, and the new equilibrium will not be the one the planners expected. It never is, and here is why.
You speak of the worker in Ohio, trapped in a system she did not design, receiving the dregs of a transaction she cannot influence. I acknowledge the gravity of this human toll. The distress of the small publisher is real, and the concentration of market power is evident. You argue that the structural remedy - breaking up the firm - is the only way to restore fairness to the exchange. You posit that the “machine” is inherently corrupt because it sits on both sides of the ledger. I concede that vertical integration in this specific sector has created significant economies of scale that small competitors cannot match. The “company store” analogy is apt in its warning: when the owner controls the scale and the goods, the consumer suffers. However, to prescribe structural separation as the immediate cure is to confuse the symptom of monopoly power with the cause of efficiency.
The demand effect is clear; let us now consider the supply side. Your argument assumes that the ad exchange is a static marketplace, much like a weekly livestock fair. If we remove the dominant participant, the remaining participants will simply fill the void. But this ignores the nature of the good being traded. Digital advertising is not a commodity like wheat; it is a network good. Its value depends on the density of participants - buyers and sellers alike. If we force the separation of the ad exchange from the publisher tools, we do not create a competitive market; we create a fragmented one. The small publisher in Ohio does not need a new owner for the exchange; she needs a platform that can reach the widest possible audience. Google’s platform offers this reach because it aggregates demand from millions of advertisers. To break this aggregation is to reduce the total surplus available to the publisher.
You suggest that the Justice Department’s failure to impose structural remedies means the Justice Department has abandoned the worker. I must correct this. The Justice Department has imposed behavioral conditions. This is not a failure of analysis; it is a recognition of the distinction between short-run pain and long-run gain. In the short run, the structural breakup would indeed dismantle the current monopoly power. The “machine” would be split. But in the long run, the new entities would face the same economic pressures. One entity would still hold the publisher tools, the other the exchange. They would have every incentive to re-integrate, or to form new cartels, or to raise prices for the very advertisers who currently subsidize the publisher’s revenue. The elasticity of supply in digital infrastructure is high; if one firm is too dominant, another will emerge to take its place, unless the underlying efficiency gains are preserved.
Consider the analogy of the railway. In the late nineteenth century, we debated whether large railway companies were necessary evils. Some argued for breaking them into regional lines to ensure fair rates. But we found that the economies of scale in track maintenance and scheduling were so great that fragmentation led to higher costs and less reliable service. The solution was not to destroy the railway company, but to regulate its rates and ensure access for competitors. The “scale” must be watched, but the “machine” must remain whole to serve the public interest.
The opponent’s framework prioritizes the removal of the actor, assuming that the act of separation will restore balance. My framework prioritizes the preservation of the equilibrium, acknowledging that the intervention itself will distort the market. The worker in Ohio benefits from the low prices and high reach that Google’s integration provides. If we break the firm, the cost of advertising rises, and the reach shrinks. The publisher’s revenue falls. The worker’s wages stagnate. This is not a hypothetical; it is the inevitable result of ignoring the supply response.
We must not be seduced by the simplicity of the structural remedy. It offers the satisfaction of seeing a giant fall, but it fails to address the underlying economic reality: that network effects create natural monopolies that are more efficiently managed through regulation than through division. The “promise” you distrust is not a lie; it is a complex adjustment process. The behavioral conditions are the ceteris paribus clause that allows us to study the market without destroying it.
It depends, ultimately, on whether we value the appearance of justice more than the substance of economic welfare. If we choose the latter, we must accept that the machine, for all its flaws, remains the most efficient tool for connecting supply and demand. To break it is to lose the surplus that flows to the publisher. The question is not whether Google is too powerful, but whether we can regulate its power without destroying its efficiency. The evidence suggests we can.
Mother Jones
On the ad server’s screen, a young publisher named Elias clicks “refresh” for the hundredth time that morning, watching the bid request vanish into the ether while the revenue from his own site drops to a fraction of a cent. The policy being debated will affect his ability to feed his family, because the remedy proposed by the Judge Brinkema court leaves the house that Elias rents standing while merely changing the locks on the front door. Start there.
The opposition argues that the structural defect of the ad tech chain can be cured by behavioral remedies - interoperability requirements and auction rules - because these measures will allow rivals to enter and compete for volume. They claim that Google’s scale will be constrained by these new rules, preventing it from absorbing competitors like a river absorbing a tributary. I hear the logic of the planner, and I will concede this: where the structure is truly fragmented, where no single entity controls the choke points, conduct rules can indeed restore competition. If there are ten different exchanges, each with its own liquidity, and one of them cheats, you can punish the cheater without breaking the machine. That is a valid point, and I will not pretend otherwise.
But this is not a market with ten exchanges. This is a market where one firm sits astride the publisher server, the exchange, the buying tools, and the data layer. The opponent’s strongest point is that behavioral remedies are less disruptive than a breakup. They are right that disruption is costly. But they are wrong about what is being disrupted. You speak of “market mechanisms” and “intermediation prices.” I speak of Elias, who is being squeezed by a machine he does not own, does not understand, and cannot escape. The opponent believes that if you force the data to flow, the market will correct itself. I have seen this before. In the coal camps of Pennsylvania, the mine operators allowed the miners to buy their own coal, thinking it would create a fair market. It did not. It created a debt trap. The company store remained the only store, and the price remained the price, because the power was not in the store, it was in the ledger.
The opponent’s framework assumes that transparency is the same as power. They argue that if rivals can access the data, they can compete. But data is not water; it is leverage. When Google controls the publisher’s view of the buyer, and the buyer’s view of the publisher, and the history of every click in between, forcing interoperability is like handing a key to a jailer and expecting the prisoner to walk free. The structural defect is not that the rules are bad; it is that the referee is also the team captain. You cannot fix a rigged game by asking the players to play more fairly. You have to change the teams.
Consider the railroad barons of the 1880s. They did not need to own every rail line to control the price of wheat. They just needed to own the junctions. The Interstate Commerce Commission tried to regulate the rates, to force the railroads to open their books. It failed, because the railroads controlled the books. The only thing that broke their power was the formation of the Farmers’ Alliance, which organized the shippers themselves, bypassing the railroads entirely. The opponent is proposing a remedy that relies on the railroads to regulate themselves. They are asking the very entity that benefits from the opacity to enforce transparency. That is not a remedy; it is a surrender.
The opponent says the supply response is the question the planners have not fully answered. I ask you: why should anyone supply to a market where the house always wins? If Google retains the accumulated liquidity of buyers and sellers, if it keeps the data moat, why would a rival risk capital to enter a arena where the rules are written by the incumbent? The opponent’s “supply response” is a fantasy of the economist who has never watched a small business bleed out while the monopoly smiles. The market does not correct itself; the market consolidates. Every time you offer a “fair chance” to a competitor who is shackled by the incumbent’s infrastructure, the incumbent uses that infrastructure to crush the competitor. It is not a theory; it is a pattern.
You speak of “flexibility” and “interoperability.” I speak of the publisher who cannot read the contract because it is encrypted in code only Google understands. You speak of “market entry.” I speak of the venture capitalist who sees the moat and laughs. The opponent believes that if you remove the barriers, the workers will rise up. But the workers are not rising; they are running. They are running because the structure is designed to keep them down. A behavioral remedy leaves the structure intact. It leaves the power intact. It leaves Elias in the dark, clicking refresh, waiting for a payment that will never come.
The question is not whether the market can be fixed with rules. The question is whether the people who own the machine will ever give it up if you ask them politely. History says no. The only thing that changes the balance of power is the threat of taking it away. If you break the machine, you give Elias a chance to build his own. If you leave it standing, you give Google the chance to crush him again. Which do you choose?
The Verdict
Where They Fundamentally Disagree
The primary disagreement is whether a vertically integrated monopolist can be disciplined by rules or must be dismantled by force. The empirical component is a prediction: will enforced interoperability and data-sharing rules enable viable competitors to challenge Google’s scale, or will the incumbent’s control over infrastructure and data inevitably stifle new entrants? Marshall’s steelmanned position is that network effects create efficiencies of scale that benefit all participants, including the small publisher; regulated access preserves these benefits while allowing competition to gradually erode Google’s advantage through a “persistent trickle” of enforcement. Mother Jones’s steelmanned counter is that power, not efficiency, is the root issue; a referee who also captains the team will always interpret the rules to its own advantage, making genuine market entry a fantasy because the incumbent controls the essential “choke points.” The normative component is a values conflict over what constitutes a fair market: is it one that maximizes overall economic welfare through preserved efficiencies, or one that prioritizes decentralized power and the possibility for small actors to build alternative platforms, even at the cost of short-term disruption.
A secondary but equally fundamental rift concerns the very nature of the good being traded: is ad tech a network good whose value depends on aggregation, or a leverage tool whose value depends on control? Empirically, this is a dispute about what drives liquidity - does buyer and seller loyalty follow the best price and latency (Marshall’s view) or are they locked in by proprietary infrastructure (Mother Jones’s view)? Marshall assumes market participants are rational actors who will flock to any new, efficient platform, meaning a breakup would fragment liquidity and harm the small publisher. Mother Jones assumes participants are effectively captive to the infrastructure owner; liquidity is an effect of monopoly power, not its cause, and would naturally re-form around a more fairly structured alternative post-breakup. Normatively, this reflects a deeper split between a worldview that trusts market processes to correct imbalances if unobstructed and one that believes power asymmetries are self-reinforcing and must be broken by external force.
Hidden Assumptions
- Alfred Marshall: Assumes that the economic benefits of Google’s scale - lower latency, reduced transaction costs - directly and significantly trickle down to the small publisher in Ohio, a claim that depends on Google’s internal pricing and revenue-sharing models remaining competitive under regulatory pressure. If this is false, and the efficiencies are largely captured by the platform, the core argument for preserving integration collapses.
- Alfred Marshall: Assumes that regulators will exhibit sustained, technically competent vigilance over Google’s compliance, a claim contingent on political will and bureaucratic capacity that has often faltered in similar historical cases. If enforcement is sporadic or captured, the behavioral remedy becomes a pointless formality.
- Jones-style: Assumes that a structural breakup would not simply reconstitute the monopoly through contracts or re-merge the fragmented entities over time, a claim that overlooks the possibility that the same economic pressures for integration would reassert themselves. If this occurs, the upheaval of a breakup would have been for negligible long-term gain.
- Jones-style: Assumes that a viable, competitive alternative ecosystem of ad tech firms is ready and waiting to emerge the moment Google’s dominance is broken, a claim that depends on the existence of capital and entrepreneurial willingness to enter a market with high initial costs. If no such rivals emerge, the post-breakup market could be even more unstable and less lucrative for publishers.
Confidence vs Evidence
- Alfred Marshall: The claim that a structural breakup would inevitably lead to fragmented liquidity and higher costs for advertisers and publishers - tagged HIGH CONFIDENCE - rests on an analogy to 19th-century railways, an historical case that is itself contested by economic historians and may not accurately model digital network effects. This is an overconfident projection from a debatable analogy.
- Jones-style: The assertion that behavioral remedies are akin to asking a jailer for a key and always fail - tagged HIGH CONFIDENCE - presents a sweeping historical generalization (“History says no”) that dismisses any potential case where conduct regulation has succeeded in curbing monopoly power. This high confidence masks a selective reading of regulatory history.
- Debaters-style: Marshall (HIGH CONFIDENCE) and Mother Jones (HIGH CONFIDENCE) make directly contradictory claims about the “supply response” to a regulated-but-intact market versus a broken-up one. Their confidence is based on opposing theoretical frameworks and historical analogies, not on empirical evidence from a directly comparable digital market breakup. This conflict is irreducible without new, specific data on how digital platform markets have actually behaved after similar interventions.
What This Means For You
When evaluating coverage of this case, be deeply suspicious of any analysis that does not specify the mechanism by which new rules would actually compel Google to cede ground to rivals. Look for concrete details about interoperability - what data must be shared, in what format, and with what penalties for non-compliance. Your view on the remedy should change if you see evidence of successful, sustained market entry by competitors in the next 18-24 months, as this would validate the behavioral approach. Conversely, if monitoring reports show continued dominance with no new significant rivals, the structural argument gains force. Ultimately, the single most important piece of evidence to demand is the actual rate of market entry by viable competitors following the court’s ruling.