This article is cross-posted in both English and French at Thibault Schrepel’s outstanding competition blog Le Concurrentialiste. Like most antitrust issues today, questions about loyalty discounts are relevant across the globe as competition regimes and courts grapple with the best way to address them.
Companies like to reward their best customers with discounts. It happens everywhere from the local sandwich shop to markets for medical devices, pharmaceutical products, airline tickets, computers, consumer products, and many other products and services.
Customers like loyalty-discount programs (or rebates) because they get more for less. And the reason so many companies offer them is because they are successful.
Everyone wins, right?
Usually. But the program could very well violate antitrust and competition laws in the United States, the European Commission, or other jurisdictions.
Complicating matters for companies that want to offer these beneficial programs is the fact that the present state of the law in most jurisdictions is, quite frankly, a jumbled mess. It isn’t at all clear whether certain programs will pass muster.
Why would antitrust and competition laws—and their government enforcers—care about companies that offer discounts? Antitrust law, after all, protects competition and, as the Supreme Court pointed out in Pacific Bell Telephone Company v. Linkline Communications (2009), “cutting prices in order to increase business often is the very essence of competition.”
Some courts, commentators, and agencies worry that certain loyalty discounts or rebates will eliminate or maim the price-cutters’ competitors, which ultimately will lead to less competition, as well as higher prices and lower quality.
This can happen in a few different ways, which I will describe in rough form below (as the economic models showing harm to competition are often complex). Please note that I am only discussing discounts for single products or services. Additional complications arise when a company offers discounts for a bundle of products or services.
Importantly, unless the company offering the discount has some form of market power (or is conspiring with a competitor), the discount won’t likely offend antitrust and competition laws. But please consider that market power doesn’t necessary entail that you are a large company, as there are many narrow geographic and product markets.
The form of the loyalty discount is also important to the analysis. The type that are most likely to offend antitrust law are market-share discounts, which are discounts that customers receive when they purchase a certain percentage of their needs from a particular company. These are usually first-dollar discounts (sometimes called retroactive discounts), which means that if the customer purchases, for example, a required seventy-percent of their needs from a company, they receive the discount on all of their purchases, not just the purchases above a certain threshold.
Also, loyalty discounts offered to end-users like customers of a sandwich shop rarely raise competition issues. Antitrust and competition authorities are typically more concerned about wholesale loyalty discounts.
As you have probably figured out by now, the competitive implications of a loyalty discount program are complicated, and go well beyond the basics that I am providing here.
I was recently privileged to be invited by the International Competition Network (ICN)—whose members are government competition enforcers throughout the world—to explore the depths of these issues for its membership by debating the test that antitrust agencies should apply to loyalty discounts.
So if you want more than the basics, you can view the slides from that debate and listen to the webinar here. I also published an article in the journal Competition on the topic a couple years ago, but the law has developed since then, so keep that in mind.
Below are a few different ways to review loyalty discounts. The problem for companies considering loyalty discounts is that it isn’t clear which of these tests courts and agencies will apply. And just as troubling, some of these tests are so dependent upon difficult-to-obtain data or varying assumptions that ex ante accurate predictions are highly unlikely.
The Predatory-Pricing Approach
I advocated to the International Competition Network that government agencies should apply this test.
The price cut from the discount or rebate could reduce prices so low that what is called an equally-efficient competitor cannot match it. The fear in that situation is that the price-cutter (with market power) will cut prices such that all of its competitors go out of business, then will raise prices when it has the market to itself.
This is the predatory-pricing approach, and claims premised on this theory are difficult to prove, particularly in the United States. The challenger must show that (1) the company cut prices below their own costs (the definition of which is still hotly contested), and (2) they have the ability to recoup this loss that they take, from pricing below their costs, when other competitors leave the market.
The likelihood that any particular loyalty discount will harm competition is very low, and the difficulty of satisfying this test reflects that. An advantage of this test is that companies considering discounts have the data they need—their own price and cost information—to determine whether the discount will survive antitrust scrutiny.
This is important because companies that can’t predict whether they will violate antitrust and competition laws—either because the law is unclear or they don’t have the data that a court or agency will use to determine whether they violated the law—are much less likely to offer the discount. And discounts are almost always pro-competitive, so the law would chill price-cutting, which is bad for competition (and consumers).
A Predatory-Pricing Variant Based Upon Contestable Shares
There are economic theories that support the notion that certain dominant firms are unavoidable trading partners such that their customers will always purchase a certain percentage of their products or services from that firm. That is typically because the customer is a retailer or distributor and customers demand the dominant firm’s wares.
Let’s say, for example, that retailers will always purchase at least fifty-percent of their baseball bats from a certain dominant manufacturer called Dominant Bats, which has market power for the sale of baseball bats. So, the theory goes, if the dominant firm offers a discount on every baseball bat to retailers that purchase, for example, seventy-five-percent of their needs from that firm, a competing bat manufacturer—Upstart Competitor Bats, for example—could have trouble staying in business, even if the discount doesn’t lead to below-cost pricing for Dominant Bats. It depends upon the numbers, of course.
The math is complicated, but it has to do with the fact that, according to the theory, the bat manufacturers are really only competing for fifty-percent of the market (as retailers will buy fifty-percent of their needs from Dominant Bats regardless), but the discount applies to all of the bats purchased from Dominant Bat.
The solution, i.e., the proposed test, is even more complicated. You add up the entire discount that Dominant Bats provides to a retailer, bat-by-bat, and attribute it all to only the “contestable” part of the market. Then you calculate whether the newly-calculated discount, attributed to the “contestable” share of the market, would have led to below-cost pricing in a made-up world.
I won’t go into more detail than that. But suffice to say, a company that is deciding whether to offer a loyalty discount is going to have a lot of trouble predicting whether it will violate the antitrust and competition laws. Part of the reason is that while you can state a particular “contestable” market share in hypotheticals, in reality, that number is not at all clear. So you are really just guessing. And your guess might differ from the guess that the Federal Trade Commission, or European Commission might make. If so, you are in trouble.
As you can tell, I don’t like this approach. Even if certain economic models say it is correct—which is still up for grabs—the predictability of it is terrible.
An Exclusive-Dealing Framework
An increasingly popular approach is to treat loyalty discounts not as a type of predatory pricing, but instead as a form of de facto exclusive dealing. Exclusive dealing, of course, is a contract whereby one firm agrees to buy all of its needs from another (or similar variant). This approach to loyalty discounts applies to market-share discounts where the seller offers a discount when the buyer, for example, purchases a certain percentage of its needs from a seller. It is, in essence, a partially-exclusive contract.
This approach applies the foreclosure test from exclusive dealing to market-share discounts, with adjustments for the fact that the it is really only partially-exclusive dealing because the market-share discount applies for less than 100 percent of the buyer’s needs (as in a fully exclusive-dealing contract).
Future blog posts will discuss the nuances of a foreclosure test, and how it might apply here. Importantly, however, companies deciding whether to offer certain loyalty discounts might (again) have trouble determining whether they are legal because the foreclosure analysis will often require data that they simply do not have, or that would be expensive to procure.
Notably, the Third Circuit in ZF Meritor, LLC v. Eaton Corporation recently applied this approach in a controversial opinion.
If you want more detail about this approach (and loyalty discounts in general), I recommend that you read FTC Commissioner Joshua Wright’s June 3, 2013 speech on the topic.
One disclosure: Before Commissioner Wright’s appointment to the FTC, he was our economic expert in a loyalty-discounts matter we defended. (He and his team at Charles River Associates in DC did brilliant work, by the way).
I also recommend that you read Steven Cernak’s excellent blog entry on the topic, and past discussions on loyalty discounts at the blog, Truthonthemarket.com.
If this blog post was useful or interesting to you, please send it to your friends. It was a little more technical than most of my posts, but I think that was unavoidable here.