(Patrick T. Fallon/Bloomberg)

An economist retained by AT&T took aim at key government arguments Thursday in the landmark antitrust trial involving Time Warner that's now in its fourth week.

With the Justice Department's top antitrust attorney, Makan Delrahim, looking on from the government's table, AT&T's witness claimed that regulators' economic analysis of the Time Warner deal is "theoretically unsound" and riddled with inaccurate assumptions.

"The evidence doesn't support the government's claim that this transaction will harm consumers," said Dennis Carlton, an economist from the University of Chicago's Booth School of Business. His own competing study, he said, found that consumers could end up facing lower prices, not higher.

Carlton's testimony threatens to undercut that of the government's own star witness, Carl Shapiro, economist at the University of California, Berkeley.

As Carlton criticized Shapiro's economic model as "very complicated," U.S. District Court Judge Richard Leon interrupted to agree with that description. "It's like a Rube Goldberg contraption," he said, reflecting skepticism about a core piece of the Justice Department's case.

Leon welcomed Delrahim into his Washington courtroom. Although the assistant attorney general has occasionally shown up to view the proceedings, Thursday marked the first time that he participated in the trial by sitting at the litigators' table and announcing himself to the court.

"My goodness gracious!" Leon said as Delrahim approached the podium. "Honored to have you."

"Honored to be here," Delrahim replied.

Carlton took aim at a number of what he called "simplifying assumptions" by Shapiro that invalidate the government's analysis. Because Shapiro tried to simulate what would happen if a cable company were unable to air Time Warner's content for an extended period rather than for only one or two months — as most "blackouts" have been known to last — the model fails to predict anything useful, Carlton said.

"The central element of his model doesn't apply … to this transaction," said Carlton.

Carlton said Shapiro's analysis also ignored the programming contracts that lock in and help stabilize content prices. Carlton also highlighted significant declines in industry profit margins and subscriber numbers that he said Shapiro overlooked. Those dynamics, Carlton claimed, mean that AT&T would have less to gain (and rivals would have less to lose) in the event of a programming dispute with Time Warner and another cable company.

What's more, Carlton said, Shapiro failed to account for the price drops — and thus, benefits to consumers — that would occur in an extended blackout. Carlton used the example of a dispute between Time Warner and Comcast. If Time Warner "blacked out" its Turner Broadcasting channels, Comcast would not have to pay for that content, said Carlton. As a result, Comcast would likely pass those savings onto TV subscribers.

Shapiro also did not examine what happened to the price of TV content after similar mergers or acquisitions, Carlton charged. For example, Comcast did not begin charging substantially more for NBC's content after it acquired the network in 2011, he said.

Shapiro had said that the reason for Comcast didn't raise its prices was because the government's consent decree with Comcast approving the merger constrained prices.

But Carlton said Thursday that that only helps AT&T's case, because it proves that the conditions placed on the Comcast deal were effective. As a result, he said, the government ought to consider the arbitration offer AT&T has sent to 1,000 cable companies that those firms can take advantage of if they feel mistreated by AT&T-Time Warner in negotiations.

Even if you accept Shapiro's assumptions, Carlton said, the resulting price increases for consumers would be so small — as a percentage of their existing bills — as to be virtually imperceptible.

"It'd be a mistake to stop a merger just because the government witness… says there'll be a tiny price increase," said Carlton.

Justice attorneys sought to pin down Carlton on a number of fronts, critiquing his use of profit margin data that was more favorable to AT&T's case. Justice attorney Craig Conrath implied that Carlton may have deliberately cherrypicked numbers that weren't accessible to Shapiro at the time he submitted his own report.

Carlton also admitted that for some consumers, losing access to Time Warner's programming would be a major issue that can't be solved simply by watching other video content offered by, say, Apple or Facebook.

And more broadly, Carlton conceded that should the AT&T merger be approved, it was plausible — though not certain — that other companies might seek further, similar consolidation in response. The economic fruits of this type of merger, said Carlton, gives firms the incentive to seek them out. Unless Leon issues "an improper decision to bar this merger," said Carlton, "we're going to see a restructuring of the industry."

But Carlton, speaking in his native New England accent, pushed back hard against a number of Conrath's assertions, such as that the price of NBC content rose faster than that of other networks after it was purchased by Comcast in 2011. In a tense exchange, Conrath sought to use a chart from Carlton's own expert report to show the trend, but Carlton insisted loudly that the graph was merely suggestive and that his in-depth calculations did not show what Conrath was trying to prove.

"Read paragraph 190," Carlton said.