out of step with the Department of Justice Guidelines. It gives me trouble in practice more than in theory because, as I noted above, in practice the application of the Stigler style of analysis tends to produce a list of things that must be done to enter and not an inquiry into the terms on which they must be done. It thus tends, in my judgment, to bias the outcome toward a conclusion that entry is easy.
In reading a number of these cases it also strikes me that one can detect an incremental shifting of the burden of proof. In several of these cases, it seems that, even though the merger considered would substantially increase concentration, the burden is on plaintiff to find and support barriers, not on defendant to show that entry is unusually easy. If the FTC and the courts had a presumption that barriers were important in the 1960s, so one did not even look at this question much, it now strikes me there is a movement toward a strong presumption that entry is easy unless somebody can find some barriers.
This is troublesome. I have taken some pains to go through what we know and do not know to make the point that we cannot prove very much very definitely. That means the party with the burden on this issue or the party who is against the presumption tends to lose except in extreme cases. This leads me to a couple of dicta of my own.
First, it seems to me that because the overall level of entry barriers can rarely be assessed confidently, precisely, and in a rigorous way, evidence pointing toward ease or difficulty of entry ought to be used with care. It is just not likely to be of the same reliability as other sorts of evidence.
Second, in light of this, if arguments about barriers or their absence are going to be used to offset other evidence, it seems to me that the burden ought to be on the party attempting the offset, and it ought to be commensurate with the weight of the evidence to be offset. Thus, in merger cases in particular, it seems appropriate to impose more stringent standards of proof on defendants using easy entry as a defense the more flagrantly the merger violates the Guidelines' concentration tests.