Grail and its competitors are developing tests which will save perhaps millions of lives. They will detect many different types of cancer very early—if they ever exist. All these tests need Illumina’s instruments. The FTC, reversing an administrative law judge, said Illumina could not buy Grail. If it did, the FTC said, it would not let Grail’s competitors use its instruments. Illumina has appealed, saying, among other things, that since the tests do not exist there is, for antitrust purposes, currently no market. Yet while the tests may or may not exist in the future the Fifth Circuit has to decide this case now.
In his key finding, the Illumina/Grail Administrative Law Judge (ALJ), D. Michael Chappell, found that he was not sure what future products Grail’s competitors may make, or what the features of these products may be. He therefore refused to block this transaction. The Federal Trade Commission had alleged vertical foreclosure: that if Illumina (which makes DNA sequencing instruments), bought Grail then Illumina would not let Grail’s competitors use Illumina’s instruments and, as both sides agree, Grail and its competitors need Illumina’s instruments to make their revolutionary cancer screening tests work. And, as both sides also agree, these tests are indeed revolutionary—if they exist they will identify, very early, very many different types of cancers, and will save thousands, if not millions, of lives.
The full FTC disagreed with the ALJ, and reversed his decision. But the FTC failed to address the key issue of this case. The FTC never explicitly addressed the fact that no expert witness, no one, can ever know, with 100% certainty, what products may exist in the future, or what their features will be. The FTC just assumed the products will exist and will have features which raise competitive concerns.
The Fifth Circuit is now reviewing the FTC’s decision. Illumina, not surprisingly, has said that since Grail’s competitors’ products do not exist there is, for antitrust purposes, no market. The FTC, which has not yet replied, must respond. The FTC must explain why the Fifth Circuit should act to protect competition regarding products which do not exist.
The ALJ, although he did not acknowledge it, analyzed competition in a Future Market. As I explain in The Future Markets Model Explains Meta/Within: A Reply to Herb Hovenkamp, I have, in 10 articles, reviewed the many cases in which not just the American, but also the European, competition authorities claim they protected competition to innovate. In these cases at least some of the relevant products did not exist. As I show, in these cases the authorities have, in reality, protected competition in Future Markets. Further, from all these cases I derived the Future Markets Model, the analytical tool all competition authorities always use when they analyze competition in a Future Market.
The Illumina/Grail ALJ analyzed competition in a Future Market. I show this, in detail, in Competition to Innovate and Future Potential Competition. As the following summary illustrates, the ALJ made his key finding when answering Question C of the Future Markets Model:
A. Does a current product exist?
One product exists. Grail has just begun, on a limited basis, to sell its product.
B. How many firms are trying to develop a future product?
5-7 competitors are trying to develop comparable products.
C. For each possible future product, is it sufficiently developed that the authority will consider it a possible future product?
The ALJ emphatically said, “No.” He said:
It is unclear when any other MCED [multicancer early detection] tests might launch, how many cancers the tests will screen for, what features the MCED tests may have, or what the prices of the MCED tests in development might be.
The ALJ then, in his key holding, applied this finding to the FTC staff’s vertical foreclosure claims. He said that in the future, if and when the other tests exist, they may be so different from Grail’s tests that Illumina may not have an incentive to withhold its instruments from these firms. These other tests may be so different that, if customers could not buy these other tests because Illumina withheld its instruments from their manufacturers, these customers would not necessarily buy Grail’s test instead, the ALJ said.
D. How broad will the authority define the Future Market? Will the authority consider future products which are similar, but not identical, as future competing products?
The ALJ said that he could not predict the future well enough to determine the features of these new tests. He therefore did not consider them competing products.
The FTC, when reversing the ALJ, simply explained that the Future Market it implicitly found, which is the same Future Market the ALJ found, met the standard market definition criteria. The FTC just assumed the products would exist, saying, “the market defined in this case must be broad enough to encompass competition to commercialize MCED tests.”
Again, the FTC ignored the crucial issue. All parties agree that the tests are indeed revolutionary and, if they exist, will form their own market. The key question is whether these products form a market now, while they are still being developed. Before the Fifth Circuit the FTC cannot ignore this issue.
Further, the FTC failed to explain what legal doctrine allows courts to protect competition in a Future Market. Since the ALJ did not block this transaction, he did not address this issue. Illumina believes no legal doctrine allows courts to protect competition in a Future Market. To prevail before the Fifth Circuit, however, the FTC must address this issue as well.
To convince the Fifth Circuit to adopt a legal standard which allows courts to protect competition in Future Markets the FTC should first point out that, as I show in my 10 articles, the enforcers have in fact been protecting competition in Future Markets for decades. They have because in the right case they must: clearly two firms cannot merge just before they start selling the same product, and thus create a monopoly.
The FTC should also point out that the National Cooperative Research and Production Act empowers courts to protect competition in “properly defined” R&D markets. And, as I have also shown, properly defined R&D markets are Future Markets.
To protect competition in Future Markets, I have proposed that courts adopt a third strand of the potential competition doctrine. I call this Future Potential Competition. As I explain in greater detail in Competition to Innovate and Future Potential Competition, Future Potential Competition asks courts to consider all relevant variables. It asks courts to consider, for example, how well developed the relevant products are, and thus how likely they are to exist. It also asks courts to consider how many firms are trying to develop future products; the more firms trying to develop a product, the more competitive the Future Market will be, even if it loses one competitor. The doctrine, in sum, recognizes that no one can predict the future; it asks courts to balance the harm of acting with the harm of failing to act.
Lastly, the FTC should point out that in another landmark case, another court of appeals acted to protect competition in a Future Market. As I explain in greater detail in Private Plaintiffs Can Easily Protect Innovation, the court in U.S. v. Microsoft did not know if Sun and Navigator would ever make full-fledged computer operating systems; both made “middleware,” which could have developed into operating systems. Yet the court said these companies were trying to make these future operating systems, and because these were possible future products Microsoft should not have stopped Sun and Navigator from developing these products. These possible future products, the court said, “reasonably constituted [a] nascent threat.” The court therefore acted. It acted to protect competition in a Future Market, that for new, and presumably better, computer operating systems.
Further still, the Microsoft court called its requirement “edentulous,” i.e. toothless, undemanding. In other words, the firms trying to enter the Future Market need not have developed their products very much; so long as the products constitute a reasonable threat, in other words so long as they have a reasonable probability of competing in the Future Market, then, the Microsoft court said, it would act to protect competition in the Future Market. Any court applying this standard would therefore act to protect competition in any Future Market so long as the relevant products, though still being developed, “reasonably constituted a nascent threat” of entering the Future Market. And this requirement, “reasonably constitute a nascent threat,” is easy to meet: it is edentulous.
Microsoft applied the Sherman Act. By contrast the Fifth Circuit will use the Clayton Act to review Illumina’s possible acquisition. Since, as the Supreme Court said in U.S. v. Penn-Olin, the Clayton Act allows courts to review transactions which the Sherman Act does not reach, if the Microsoft court could use the Sherman Act to protect competition in a Future Market, then certainly the Fifth Circuit can use the Clayton Act to protect competition in a Future Market.
Unlike in Microsoft, however, the issue in Illumina/Grail is possible vertical foreclosure: would Illumina, if it owned Grail, not let Grail’s competitors use its instruments? The FTC says it would not. The ALJ and Illumina say that the competitors’ tests are not sufficiently developed to determine if this is true.
Instead of simply declining to act because Grail’s competitors are still developing their products, the Fifth Circuit should, applying Future Potential Competition, balance the cost of acting with the cost of failing to act. As the FTC should say (assuming the confidential information bears this out), with 5-7 companies trying to make comparable tests, the probability that some tests will exist, and will need Illumina’s instruments, is sufficiently great that the court should block the transaction.
Using Future Potential Competition the FTC would certainly be asking the Fifth Circuit to answer a question different from the one the ALJ answered. It would be explicitly asking the court to consider the cost of not acting. Since the ALJ did not consider this, and ruled against the FTC, the FTC is more likely to prevail if it asks the Fifth Circuit to answer a question different from the one the ALJ answered.
Articles represent the opinions of their writers, not necessarily those of the University of Chicago, the Booth School of Business, or its faculty.
