In 2001, the two biggest sources of the milk that you buy at Stop-n-Shop, Whole Foods, Von's, and the like agree to merge.

A sluggish Antitrust Division of the U.S. Department of Justice okays the deal but tacks on a proviso that the dairy dukes — Dean Foods and Suiza — must divest a few of the plants that process the raw milk.

The milk mavens do spin off the plants. Yet they agree to pay the buyer, Dairy Farmers of America, millions and millions of dollars to do a poor job of running the plants and selling the milk. The payola takes the form of contracts giving DFA the right to sell tons of its raw milk to the new combo.

Do you see a per se violation of the Sherman Act?

The Sixth Circuit didn't. It held that the pact between Dean/Suiza and DFA had too much of a "vertical" aspect to fit within one of the per se realms — price-fixing, allocating customers or areas, limiting output. Food Lion, LLC v. Dean Foods Co., No. 12-5457 (6th Cir. Jan. 3, 2014). 

The ruling means that the plaintiffs, Food Lion and Fidel Breto, will have to proceed under the "rule of reason", which (unlike the per se rule) demands proof that the restraint of trade restrained trade to an "unreasonable" extent. But that represents something of a win. The district court had thrown the case out on summary judgment.

Blawgletter agrees with the Sixth Circuit that the district court erred in granting summary judgment against the plaintiffs. The trial court shouldn't have ignored expert evidence that tended to show that the agreement not to compete pushed milk prices up within a discrete area, as the panel rightly ruled. But we disagree that the plaintiffs in the case don't have a proper per se case. We don't see how the fact that the payola went to DFA instead of straight to the DFA affiliate that carried out the agreement not to compete matters.

Nor did the court help clarify a legal area that many, including lots of lawyers, find daunting if not impenetrable. For example:

• The panel wrote that "[t]he district court's decision to use the rule of reason is a question of law" but didn't explain how a "decision" could constitute a "question". The decision answers a question of law. The question of law consists in the legal rule that informs the decision. The decision and question aren't the same thing.
• The court said that the per se rule counts as a "less common method" for judging the lawfulness of restraints on trade and "should be [applied] reluctantly and infrequently". Yet the rule applies 100 percent of the time to price-fixing, for instance, and other kinds of patent efforts to conspire against the common good. In what way is it "less common"?
  
• The opinion observed that "[v]ertical restraints . . . have more redeeming qualities" than horizontal ones but doesn't say why a supplier's forcing a buyer to, for example, charge higher prices has any "redeeming qualities". Vertical restraints may be more likely to have redeeming qualities than horizontal ones, but they don't always.
• The panel relied on the Antitrust Division's and Federal Trade Commission's Horizontal Merger Guidelines from 1997 rather than the current ones from 2010. The newer Guidelines downplay the use of the "hypothetical monopolist" and other mechanical tests for judging the likely effects of a merger. The shift matters – and in fact appears to have prompted a less sluggish Antitrust Division to sue Dean Foods in 2009. (The AD and Dean Foods settled that case in 2011.)
• The court opined that "Plaintiffs should not be able to change their characterization of the conspiracy midstream in order to gain a more favorable outcome" as if courts shouldn't permit a party to adjust its allegations in light of the evidence it discovers and in response to the rulings of the courts. Why not end the case at the pleadings stage if that's the test, eh?
• The panel sought support for its per se v. rule-of-reason ruling by saying that, "especially at the summary judgment stage, this is not a 'clear cut' case of an obviously anticompetitive restraint", suggesting that the summary judgment rule imposes a standard higher than the test for upholding a jury verdict. If anything, a court must resolve doubts at the summary judgment phase of case in favor of the non-moving party — not against it.
• The opinion took the fact that the Antitrust Division did not block the deal as proof that it "sanctioned the agreement at issue here, which presumably would not have occurred if the agreement was a per se unreasonable restraint on trade." The AD did no such thing. It's decision not to sue in no way "sanctioned" the deal. And even the low-key Antitrust Division of that era felt enough worry to try to prevent post-merger abuse of market power.

We don't mean to make light of the difficulties of judging antitrust cases. It isn't at all easy. But we do feel strongly that it's really important to get the details right. And, in this case, the panel didn't do that.