Judge Spares Google Ads Business From Breakup

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?

Consider the market mechanism at work. Ad tech is not a single market but a chain of markets - the publisher ad server, the exchange, the buying tools - and Google sits astride more links in that chain than any competitor. A structural breakup would have severed ownership across those links, forcing new firms to compete for volume at each stage. Behavioral remedies, by contrast, restrict conduct - interoperability requirements, auction rules, data-sharing obligations - while leaving the ownership structure, and with it the accumulated liquidity of buyers and sellers, intact. This is the crux of what is genuinely contested here: whether conduct rules can restore competition in a market whose defect is structural, namely that both sides of the auction gravitate to wherever the other side already is.

The strongest case for Judge Brinkema’s caution is not to be dismissed lightly. Vertical integration in ad tech does lower certain transaction costs - a single technical stack reduces latency and reconciliation errors in a business measured in milliseconds. A forced divestiture risks throwing away those efficiencies for the sake of a structural remedy whose benefits are uncertain and whose costs, in the short run, are certain. This is a fair point, and I do not think it foolish. But it answers the wrong question. The issue is not whether integration lowers cost today; it is whether that same integration, sustained year after year, forecloses the possibility of a rival exchange ever reaching the scale at which its own costs would fall to a comparable level. A young exchange needs both publishers and advertisers simultaneously, in the same way a new marketplace needs both stallholders and shoppers on its opening morning - and shoppers will not come to an empty market, nor stallholders to one with no shoppers. Behavioral rules can mandate that Google let others through the gate. They cannot, by themselves, populate the stalls.

Over the years the Justice Department will presumably continue to monitor compliance, and here the long-run adjustment becomes decisive. If enforcement is vigorous and sustained, the behavioral constraints may gradually erode Google’s structural advantage, much as a persistent trickle wears a channel through rock. If enforcement lapses, as such undertakings so often do once the newspapers move on, the equilibrium will revert to the one the intervention was meant to disturb, only with additional compliance paperwork attached.

My own reading, offered with the appropriate qualification, is that the short-run effect favors Google and the long-run effect favors whichever party - regulator or rival - proves more patient. Milder remedies buy time. Whether they buy competition depends entirely on what is built with it, and on that the court has, wisely or not, left the answer to the years still to come.