Beyond Compare

$SAVE, $ACI, and the issues with consumer classes in "Relevant Markets"

Much ink has been spilled over the JetBlue Spirit trial—including by me—and yet more will be spilled once we get the parties’ post-trial briefing on Dec. 13. But I wanted to take this brief lull to talk through antitrust a little more broadly. The JetBlue / Spirit deal is unique in many ways, but it’s worth addressing how it might impact antitrust as a whole in the long run, and deals like Kroger / Albertsons in the short run.

By now, most people following these cases are familiar with the Baker Hughes framework for analyzing horizontal mergers, which defines the parties’ burdens in an antitrust merger case:

1. The DOJ or FTC must define a “relevant market” comprised of a product or group of products, and the geographic area in which these products are produced and/or traded. Then, within the relevant market . . .

   1. First, the DOJ or FTC must show the merger will create a likelihood of “substantially lessening competition” in the relevant market. This is typically shown by way of demonstrating the transaction will create undue concentration under the “HHI Index” in a particular market for a particular product, creating a presumption of harm.

   2. Second, if the government has met the first hurdle, the defendants must produce evidence of pro-competitive effects to rebut the presumption established by the government, which can be by way of

      1. demonstrating the government has failed to produce sufficient evidence of harm,

      2. showing the merger will have pro-competitive effects which will outweigh the anticompetitive harms, and/or

      3. showing the harm will be temporary, and entry by competitors will be timely, likely, and sufficient.

   3. Third, if the defendants have produced sufficient evidence, the government must provide additional evidence of anticompetitive effect.

Notably, points 2 and 3 describe burdens of production. The final step three actually “merges” with the burden of proof for the government, but this burden of proof (establishing the merger is anticompetitive by a preponderance of the evidence) is always with the government.

The burden-shifting analysis is not unique to the Clayton Act. It applies in nearly identical fashion to anticompetitive conduct (the Sherman Act) through Ohio v. American Express Co., with the ultimate question being whether or not the pro-competitive effects of the conduct can be achieved through less restrictive means.

As I discussed with Andrew Walker on our original SAVE 0.00%↑ podcast, the government always wants a narrow relevant market. It’s obviously easier to show undue concentration if the definition of competitors is very limited. But should product and geography be the only two inputs? Should there be more? Should there be less?

Brian Albrecht (who was just featured on Yet Another Value Podcast) recently published a paper about the Kroger / Albertsons merger, in which he—in part—argues for the inclusion of wholesale club stores (e.g., Costco or Sam’s Club), as well as accessibility factors like delivery platforms (e.g., Instacart) in the “relevant market” analysis. In other words, he wants to be able to broaden the scope and have fewer restrictions on how to define a market. And I don’t disagree with him. No doubt our society is more nuanced today when it comes to grocery shopping than it was in 1950.

The only problem with including these wholesale clubs and delivery services into the analysis is that these stores are more often frequented by the affluent end of the consumer market. As financial journalists have discovered, rich people f*cking love Costco. Even if wholesale clubs might be a great deal, it’s usually more educated, more well-off consumers who are willing to travel a bit further for a deal, and who can afford to front-load several months’ worth of cereal, frozen vegetable, and toilet paper expenses.

That brings us back to JetBlue and Spirit.

A bit of testimony came out during the trial through Dr. Hill’s testimony about consumer preference between the two merging brands: it turns out consumers overwhelmingly prefer JetBlue. Dr. Hill pointed to JetBlue’s outsized revenues next to Spirit on routes where the two brands fly, and he noted the bang for the buck that Blue Basic offers when compared to Spirit’s spartan service. But Dr. Hill did not volunteer which consumers he was talking about when talking about this overwhelming preference. He was, of course, largely talking about a wealthier customer; perhaps modestly so, in the grand scheme of things, but wealthier nonetheless. After all, JetBlue might be more popular in the aggregate, but I think we all know Ivanka Trump was never going to fly Spirit.

The difference in consumer class was not lost on Judge Young, who told counsel during closings:

“To me, that’s one of the key issues here. What is my benchmark for balancing the loss to Spirit consumers against the downstream to consumers—perhaps a more affluent or business class—that all airlines compete for? How will I balance that?”

It’s worth taking a pause to remember that, regardless of the JetBlue trial outcome, the relevant market in the JetBlue deal is not consumer-specific, except to the extent certain routes might be flown by less affluent customers. The relevant market is “scheduled air passenger travel” at either the local or national level.

So if we go back to our Baker Hughes framework for a moment and we focus on (iii) under Step 2 (defendants showing pro-competitive effects to rebut a presumption of harm) we are forced to wonder if we can include downstream effects to a different consumer—a likely higher end consumer—than the one harmed in Step 1 of the analysis. But how does one compare that benefit in a meaningful way to the harm that other consumers might face? Is that even possible? How do you quantify convenience to the well-off with inconvenience to those more in need?

Still further, we’re left wondering whether we can consider downstream effect at the national level, or whether we’re constrained to consider only the downstream effect within a specific o/d pair. JetBlue would, of course, prefer Judge Young to consider it all, but the case law is unlikely to support considering the national effects if Judge Young finds the o/d pairs to be a compelling market definition. At the very least though, there are some customers on even the most “vacation-y” routes (e.g. Spirits’ San Salvador to Ft. Lauderdale) who might see the benefits of a more comfortable JetBlue offering when analyzing under an o/d-specific market.

Similarly, the Kroger deal faces a somewhat bifurcated consumer market class analysis. We already know the FTC will always push for the most narrow market, but is it actually reasonable to include wholesale clubs, when historically the only “market” that mattered was the mom & pop grocery store down the street? We are now forced to wonder if we can expand the relevant market itself to include consumer demand from more affluent consumers.

It all raises another interesting question, along with a precedent-setting opportunity.

You might have asked in the months leading up to the trial, “Why didn’t the DOJ in the JetBlue case assert that the ULCC market was the relevant market?” In fact, Judge Young specifically asked DOJ counsel whether or not the ULCC market was asserted as the relevant market in closings.

Well first, we have the very obvious issue that JetBlue acquiring Spirit would not be “presumptively” anticompetitive when viewed solely through the lens of the ULCC market because JetBlue isn’t a ULCC. If the analysis is “does the pro forma entity have too much concentration in the market?” then the answer would be “the pro forma entity has no presence at all in the relevant market.” In fact, all JetBlue is doing to the ULCC market is creating a void for other ULCC’s to fill. It would almost require a reverse-HHI analysis to establish the market concentrations of the remaining players were presumptively problematic. But antitrust law doesn’t focus on the void, it focuses on the pro forma entity.

Second of all, it allows the DOJ to make an argument that would expand antitrust case law in favor of the Neo-Brandeis movement if they obtain a wholesale favorable verdict. Despite some arguments in favor of the principle, current antitrust law is not about maximizing choice. After all, as Robert Lande has said, “every contract restricts competition to some degree.” But the DOJ has slipped in an argument that seeks to expand antitrust’s scope. The JetBlue / Spirit deal, the DOJ argues, would present a distinct harm of reduced consumer choice. To the DOJ, the loss is not evidence of a harm, the loss is the harm.

That then brings us to Judge Young.

If Judge Young wanted to take the easy way out and rule for JetBlue, he would rule that the DOJ has not met step 1, having presented largely nationwide data to support harms in specific markets. If Judge Young wanted to take the easy way out and rule for the DOJ, he would rule that o/d pairs are the relevant market and JetBlue + Spirit market share is so high in at least one o/d market that JetBlue simply never was going to meet its burden of production in step 2.

Of course, Judge Young almost certainly won’t take the easy way out here. He’s a seasoned veteran with an opportunity to bolster his judicial legacy.

Judge Young—and whichever judge presides over the at-this-point-probably-inevitable Kroger trial—has the chance to craft an opinion that realigns antitrust law with, what most economists would probably agree, is the original intent: preventing mergers that lead to abuse. It does not require a PhD in economics to understand “big is bad” is a gross oversimplification, but as Brian Albrecht (an actual PhD economist) noted in his podcast appearance, firm size can be a terrible proxy for competition, and defining “relevant markets” is often arbitrary at best.

The ultimate question in JetBlue and Kroger should not be whether or not the resulting firm is “too big,” nor should it be whether or not there are a fewer options in the market. Rather, the pertinent question is whether or not the combined firm is able to wield its new weight in an abusive way and is likely to do so. As so many have noted, combining airlines 6 and 7 (with the resulting firm having ~10% market share) is hardly undue concentration when faced with 15-20% concentrations for the major players. It would therefore seem to make sense for Judge Young to craft an opinion that specifically adopts the ability to consider more affluent consumers, and specifically rejects the notion that loss of “consumer choice” is a standalone harm, provided at least that lower-end consumers are not left in the dust. Alternatively, Judge Young can reject the class comparison altogether by focusing on—and perhaps elevating the importance of—the replacement opportunity of which other ULCC’s can now take advantage. If he does either, Kroger may see an improved path to victory in the event of an FTC challenge.

In the end, that JetBlue gains market share is immaterial if higher end customers don’t see prices skyrocketing and lower end customers don’t go years without a replacement offering. Similarly, Kroger increasing its footprint is irrelevant if higher end consumers simply augment their Costco shopping list, and lower end consumers see new storefronts pop up in their communities.

It’s really just a question of whether or not the already-overloaded G-wagon suspensions can handle those extra rolls of paper towels.