On Thursday, the Justice Department will ask a federal appeals court to unwind at least a portion of AT&T’s blockbuster purchase of Time Warner—a deal valued at $100.3 billion.

The government claims the merger violates Section 7 of the Clayton Act, an antitrust law that bars acquisitions whose effects “may be substantially to lessen competition.”

Yet as gargantuan as the deal is, Thursday’s argument will have almost nothing to do with its “bigness” or the frightening concentration of power it represents. Instead, as Columbia Law School professor Tim Wu has recently bemoaned, it will revolve around how many more or fewer pennies consumers will spend on their monthly TV subscriptions if the combination is left undisturbed.

The transaction was completed on June 15, three days after U.S. District Judge Richard Leon blessed it in a 172-page ruling. But AT&T agreed to manage Time Warner’s Turner Broadcasting System unit—which includes TNT, TBS, CNN, and the Cartoon Network—as a “separate business entity” pending the outcome of the government’s appeal to the U.S. Court of Appeals for the DC Circuit. The appeal is being heard Thursday, and will likely be decided by the end of February.

AT&T and the government both declined to comment on the appeal. “We’re going to bring a fresh approach to how the media and entertainment industry works for consumers, content creators, distributors and advertisers,” AT&T CEO Randall Stephenson said when the deal closed.

At this point a betting person would have to predict that Judge Leon’s ruling—reached after listening to 31 witnesses at a six-week trial last summer—will prevail. But there could still be more court rounds before the dust settles. The case could be sent back down to Judge Leon for further proceedings or, conceivably, continue up the ladder to the U.S. Supreme Court. Barring the latter, the appeals court’s reasoning will set important precedents determining how easy or difficult it will be to pull off similar giant mergers in the future.

The government’s case has been unusual in two ways. First, it represents the first time in four decades that the Justice Department’s antitrust division has gone to trial to block what’s known as a “vertical” merger. It more commonly challenges “horizontal” mergers, which involve companies that compete with one another. In such instances, the competition concerns are obvious, since one competitor is swallowing up another, leaving consumers with fewer options.

A vertical merger is one between businesses that do not compete but, instead, have a supplier-customer relationship. AT&T and Time Warner have that relationship because Time Warner, through its Turner unit, among others, supplies programming to AT&T’s pay-TV distribution subsidiaries, including satellite operator DirecTV, fiber-optic cable company U-Verse, and DirecTV Now, which streams television programming over the internet.

While the antitrust division’s guidelines advise that most vertical mergers are “procompetitive or competitively neutral,” the government decided that this was one of the exceptions that required intervention. Given the mammoth size of the combination, and the fact that a programming distributor was acquiring a supplier of programming that rival distributors all craved access to, the deal unquestionably merited close scrutiny.

The second unusual aspect of this case was the cloud under which it was launched. On the very day the proposed merger was unveiled, October 22, 2016, then-Presidential candidate Donald Trump vowed to stop it if given the chance.

“As an example of the power structure I’m fighting,” he said then, “AT&T is buying Time-Warner, and thus CNN, a deal we will not approve in my administration.” A few days later, a top campaign advisor reiterated Trump’s vow, stressing that Time Warner included “the wildly anti-Trump CNN.”

The campaign’s reflexive opposition—prior to any conventional antitrust analysis—raised the specter that Trump might misuse Justice Department power to punish CNN for First Amendment-protected news coverage. (Trump’s hostility to CNN has only escalated since then, of course. As president, he has regularly denounced the network as “fake news” and an “enemy of the people,” and earlier this month he stripped its White House correspondent of his security credentials.)

In November 2017, Trump’s antitrust chief, Makan Delrahim, sued to block the deal, raising more eyebrows. Eleven months earlier, when Delrahim was still a law professor, he had said of the transaction: “I don’t see this as a major antitrust problem.”

Shortly after the case was filed, AT&T asked Judge Leon to permit discovery on whether political animus had tainted the government’s decision to bring it. In legal terms, the company wanted to explore whether the suit reflected unconstitutional “selective enforcement.” This merger was being treated differently, AT&T claimed, from Comcast’s vertical merger with NBC Universal in 2011, which the antitrust division greenlighted after Comcast agreed to institute certain safeguards. AT&T had voluntarily committed to imposing those same safeguards. (If a vindictive president was, in fact, using the Justice Department to retaliate against a media critic, that might also constitute an impeachable abuse of power. But that would be up to Congress to decide, not Judge Leon.)

In February, Judge Leon denied AT&T’s request, finding the preliminary evidence of selective enforcement was too flimsy, and commenting that it was “difficult to even conceptualize how a selective enforcement claim applies in the antitrust context.” (Leon, an appointee of President George W. Bush, has won praise from the likes of The New Yorker for his maverick opinions condemning a National Security Agency surveillance program and upholding the rights of Guantanamo Bay detainees, though some of those rulings were later reversed. He’s also a character, known for peppering his opinions with exclamation points, which are seldom used in judicial decrees. His order approving the AT&T-Time Warner merger contained 19 of them.)

It’s unclear whether the DC Circuit will address Judge Leon’s blocking of the selective enforcement line-of-inquiry. AT&T’s brief only mentions the matter in passing. In an amicus filing, however, the Reporters Committee for Freedom of the Press does ask the court to declare that Judge Leon erred in this respect. “[AT&T] sought discovery to determine if evidence existed that the president translated his extraordinary rhetoric around CNN into improper pressure on the Justice Department,” committee lawyer Bruce Brown writes, “pressure that would also serve as a signal to the press as a whole that it exercised its constitutional right to free and independent coverage of this administration on pain of regulatory harassment, financial cost, and ongoing intimidation.”

Notwithstanding this malodorous elephant in the room, the case is being treated by the parties like any other antitrust case. With one exception, the same is true of the familiar cast of characters who have submitted 10 amicus briefs submitted in the case. They have lined up on predictable sides of the field, with business groups, like the Chamber of Commerce, siding with AT&T, while Consumers Union, Public Knowledge, and the pro-enforcement American Antitrust Institute are supporting the government. Similarly, a group of 27 liberal-leaning scholars are weighing in on the government’s behalf (including Harvard’s Einer R. Elhauge; New York University’s Eleanor M. Fox; and Nobel Prize winning economist Joseph E. Stiglitz of Columbia) while 37 conservative-leaning ones are backing the merger (e.g., NYU’s Richard Epstein; the University of Chicago’s William M. Landes; and Stanford’s Alan O. Sykes).

The one exception is that not a single state attorney general is supporting the government’s position—a striking void in an antitrust case, especially with respect to the Blue States. Instead, 10 state AGs have submitted a lone, bipartisan brief (seven Republicans, three Democrats) that comes down on the side of AT&T. One has to suspect that Trump’s conduct and the whiff of selective enforcement have repelled the other Democratic AGs. The bipartisan brief attributes the conspicuous absence of support for the government to other factors, however, including the dynamism of the industry, expected cost-savings for AT&T’s customers, and AT&T’s voluntary adoption of the safeguards that the government considered sufficient in the Comcast-NBCU merger.

The case is not about “bigness”

For a merger between two corporate behemoths, the case is being fought on remarkably narrow grounds. For many critics of the deal, the obvious and overriding concern it presents is the vast concentration of power—especially media power—into so few hands. Yet that’s not remotely what the appeal is about.

Though a gut-level distrust for “bigness” animated many antitrust laws when they were written, the largely conservative, “law-and-economics” school of analysis has come to dominate the field since the 1980s, redirecting judicial analysis. As a consequence, judges now focus narrowly on a deal’s impact on consumers and the prices they’ll pay.

Columbia Law School professor Tim Wu has bemoaned this situation, protesting in a New York Times op-ed last summer that the antitrust statute at issue in this case, if interpreted the way its drafters intended, “would surely have allowed the Justice Department to block the recent AT&T-Time Warner merger.” He continued: “No one can deny that the new AT&T will have more economic power and also more political power than before, even as it now carries more debt ($181 billion) than many industrialized nations.” (Wu is the author of “The Curse of Bigness: Antitrust in the New Gilded Age,” which came out this month.)

Because of this shrunken playing field, the government’s case may have been doomed before it ever began. Its suit is not about agglomeration of power, but about the prices the average consumer is likely to pay, post-merger, for his or her monthly cable subscription—assuming he or she hasn’t already “cut the cord,” as about 20% of them have.

Although AT&T is one of the nation’s two dominant wireless telephone carriers, the government’s appeal revolves solely around its role as a distributor of television programming, mainly through DirecTV and U-Verse.

Similarly, only a fraction of Time Warner’s assets are relevant to the appeal. Though the company (which now, post-merger, does business as WarnerMedia) owns Warner Bros., which makes movies and television programming, and Home Box Office, which owns the HBO premium cable network, the government’s case focuses solely on the Turner assets.

“A Beautiful Mind” and blackouts

The central theory of the suit stems from a model of bargaining behavior developed in the 1950s by Nobel Prize-winning economist John Nash. (Nash became a household name after Russell Crowe portrayed him in the movie “A Beautiful Mind.”) The negotiations at issue in this case are those that lead to “affiliate agreements.” Those are the contracts signed between programmers, like Turner or NBC, and distributors on the other, like Comcast, Verizon FIOS, DISH, or AT&T’s DirecTV.

According to the theory, each negotiating party weighs the harmful consequences of not reaching a deal. With affiliate agreements, a breakdown in negotiation results in a “blackout,” where the distributor can’t run a network’s programming in its bundle of offerings. That hurts the distributor, because it loses subscribers (both existing and potential), and also hurts the programmer, because it loses licensing fees and advertising revenue. The price and terms the parties finally accept are based on each side’s calculation of how disastrous a blackout would be for them.

The AT&T-Time Warner merger will necessarily alter those negotiations, the government argues. Post-merger, it says, Turner will demand higher prices from AT&T’s distributor rivals. And Turner will win those higher prices, it maintains, for reasons Nash’s theory explains.

Here’s why. Post-merger, if Comcast endures a blackout, some of its subscribers will defect to AT&T’s distributors—DirecTV, U-Verse, or DirecTV Now—in order to keep receiving the Turner programming. Though Turner will also suffer pain from the blackout (loss of licensing fees and advertising revenue), it will now be willing to endure more of that pain because its new parent (AT&T) will actually be benefiting from the blackout, because of the new subscribers switching over from Comcast to DirecTV. (This step of the analysis relies on another maxim of economic analysis known as “the principle of corporate-wide profit-maximization.” It holds, as the government explains in its brief, “that parent corporations and their wholly-owned subsidiaries act to maximize corporate-wide profits, not just the profits of individual divisions.”)

In the face of Turner/AT&T’s strengthened bargaining leverage, Comcast—and Verizon FIOS, DISH, and others—will agree to pay a higher price for Turner programming than they would have in the past, Nash’s bargaining theory predicts. They will then pass their increased costs on to consumers, jacking up monthly subscription bills for all consumers except AT&T’s. (AT&T’s customers, the government concedes, will benefit from the merger, due to internal cost savings.)

Though the theory might seem convoluted, AT&T itself advanced a version of it when it urged regulators to block the Comcast – NBCU merger in 2011.

Is 27 cents per month ‘substantial’?

To concretize all this abstract theory, the government hired economist Carl Shapiro of the University of California at Berkeley to compute precisely what the merger would cost American consumers, using economic modeling programs. He concluded that, by 2021, American consumers would collectively be paying $571 million a year more for their pay television subscriptions, even after subtracting the cost-savings the deal would achieve for AT&T customers.

Alas, Judge Leon was unpersuaded by Shapiro’s labors. To begin with, $571 million in net added costs is not as big a number as it might sound at first. On a per-customer basis, it works out to a measly 27-cent hike on a monthly bill of more than $100—a 0.2 percent increase. The Clayton Act only bars mergers that might “substantially” lessen competition, Leon wrote, and he “harbor[ed] serious doubts” that such a “miniscule” increase was substantial.

More important, he rejected Shapiro’s calculations. He was more impressed by AT&T’s expert, Dennis Carlton of the University of Chicago School of Business, who argued that Shapiro based his models on mistaken assumptions. Once these were corrected, Carlton testified, the modeling showed a net consumer benefit from the merger.

Judge Leon also accepted the testimony of certain industry participants, who claimed that the change in negotiating leverage that Nash’s theory predicted would be negligible in the real world.

The merger’s impact on blackout calculations would be trivial, Time Warner chief Jeff Bewkes insisted at trial. It was like the difference in being hit “by a 950-pound weight instead of a thousand pounds,” he said. “Are you going to think about it differently, feel differently? Are you going to take more risk . . . ? Absolutely not.”

Finally, AT&T expert Carlton looked at previous vertical mergers in the industry, including Comcast’s acquisition of NBCU, to see if they had, in fact, significantly altered the terms of affiliate agreements. They hadn’t, he found. (The government says the deals he looked at were not comparable.)

Clear error?

Why would the appellate court judges, reading a cold trial transcript, substitute their own factual findings for those of Judge Leon, who sat through weeks of live testimony observing the demeanor of the witnesses first-hand?

Generally, they wouldn’t. The law imposes a high burden on parties bringing appeals, requiring that they demonstrate “clear error” by the district judge before the appellate court can reverse.

The government claims it can vault that hurdle because Judge Leon rejected Nash’s well-accepted theory of economic bargaining. Similarly, it claims that Leon also refused to heed the “foundational principle of corporate-wide profit maximization”—i.e., the prediction that Turner, during blackout negotiations, would subordinate its parochial economic interests to those of its parent, AT&T.

AT&T parries that Judge Leon did no such thing. He accepted and applied both theories, the company argues. He simply concluded that, on the facts of the case, neither theory results in harm to consumers. (AT&T’s appeal will be argued by Peter D. Keisler of Sidley Austin.)

A good panel

The three-judge panel hearing the case might be considered slightly favorable to the government, in that two are appointed by Democratic presidents and one by a Republican. Circuit Judge Judith W. Rogers was appointed by Clinton; Circuit Judge Robert L. Wilkins, by Obama; and Senior Circuit Judge David Sentelle, by Reagan.

“It’s a good panel in the sense that they’re all smart,” says David Vladeck, a professor at the Georgetown University Law Center. Rogers and Wilkins are “reasonably liberal,” he continues, while Sentelle is “fairly conservative, but open-minded.”

Under the circumstances, it would not be surprising to see at least a couple judges chastise Judge Leon for giving short shrift to the selective enforcement issue.

On the merits, though, AT&T’s chances look good. Given that the government must show “clear error”; that AT&T has committed to adopt the same safeguards that were considered sufficient for the Comcast-NBCU merger; and that the government only established, at best, a 27-cent hike in monthly subscription costs, the Turner networks will probably stay where they are.

Roger Parloff is a former editor-at-large at Fortune Magazine, and has been published in Yahoo Finance, Yahoo News, The New York Times, ProPublica, New York Magazine, and NewYorker.com, among others.