Thread by @ProfWrightGMU

Really LONG thread coming on vertical mergers responding to @TomValletti and @geoffmanne exchange --- which refers to me and to @GAI_GMU and questions my credentials as an economist at some point. But never quite finds a way to @ me or @GAI_GMU. First...a short version. (1/x)

TL; DR:
@TomValletti: the evidence is vertical mergers = bad.
@geoffmanne: Nope; see Lafontaine & Slade & @GAI_GMU.
@GAI_GMU: Most VI studies don't show net harm.
Valletti: GAI are not econs! They are bad. I am good. They are old.
Also Valletti:*most studies don’t show net harm.*

I woke up this morning to an interesting thread from @TomValetti. Tommaso was responding with sound and fury -- and a heavy dose of pedantry and misplaced credentialism -- to a post from @geoffmanne about the @GAI_GMU submission to the FTC hearings on vertical mergers. (3/x)

We’re going to keep our reply to substance. Let’s go back to original @TomValetti claim that started this whole thing off. His claim was that there is increasing evidence if anticompetitive foreclosure in vertical mergers. (4/x)

@Geoffmanne cited the @GAI_GMU vertical merger comment which purported to extend Lafontaine & Slade’s 2007 piece on the the competitive effects of vertical integration. Professor Valetti didn’t like that. Here is our (award-nominated) comment: bit.ly/2Cy0vxK (5/x)

First, Tommaso says the @GAI_GMU paper covers TOO MANY papers because we include vertical integration & not just mergers. The idea that the economics of vertical integration does not inform us about vertical merger policy is just wrong. (6/x)

L&S -- and decades of IO on this --- correctly note: “The empirical literature on vertical integration…are also important ultimately as input into the development of sensible vertical merger policy and related government intervention in vertical relationships” (7/x)

Tommaso does make one good catch. Thanks! Ashenfelter et al. should have been excluded. It is a horizontal beer merger and I had confused it with another beer paper on VI. Good catch! And thank you. But he thinks we should exclude more. Let’s talk about those. (8/x)

Tommaso also thinks we should exclude papers that focus on “transaction cost economics.” He wants to narrowly focus on the tradeoff b/w foreclosure effects and double marginalization. I'm not sure why those are the only econ tradeoffs that matter for welfare-based policy. (9/x)

Consideration of “transaction cost economics” are a critical component of evaluating merger efficiencies. If you are a credential snob, how about Williamson (1974, “The Economics of Antitrust: Transaction Cost Considerations,”) (10/x)

“Failure or refusal to make allowance for transaction costs, in circumstances where these are arguably nonnegligible, can lead to error. Not only is an understanding of the issues impaired, but incorrect policy prescriptions will sometimes result,” p. 1439)." (11/x)

Here’s a sample of results of papers and results Tommaso says agencies should be shielded from:

Forman & Gron: “Our results show that insurers with exclusive agents adopt consumer applications significantly more quickly" (12/x)

Malik: “vertical integration positively affects new product development” (p. 861). Increasing the rate of innovation hardly seems like an efficiency that we should be dismissing." (13/x)

Hanssen: “All this is consistent with the proposition that cinema ownership was part of a system that supported efficient ex post adjustments in the length of film runs” [emphasis added] (p. 521). (14/x)

More from Hanssen: “In this paper, I provide evidence that cinema-owning motion picture companies did not favor their own productions—days of exhibition were divided nearly equally across the firms of all producers” [emphasis added] (pp. 522-523). (15/x)

I could go on. But again, discounting or completely ignoring the value of studies that indicate more efficient firm structures with vertical integration is bad economics and bad antitrust policy. For an economist, this approach to excluding evidence seems rather lawyerly (16/x)

Professor Valletti also objects that a few papers are not “econ” enough for him or not in journals he finds good enough? In particular, this is his beef with including Baker et al. and Malik. What gives? (17/x)

He’s just wrong about Baker et al. They are “from Econ.” Baker, e.g., is a Ph.D. economist from Princeton who is currently the Chair of the Dept of Health Research and Policy at Stanford Med School). His co-authors have Ph.D's in from Wharton and MIT. Those seem good. (18/x)

Perhaps most importantly, we do not believe it appropriate to judge the quality of the evidence by the authors’ credentials. Better to show it all and be transparent than to hide papers or worse, to judge by them alone. (19/x)

Let’s get to the nitty gritty --- if you’ve hung in this long. Thank you. What are the welfare effects of the mergers evaluated in the literature and what to they say about how agencies should be thinking about likely effects when evaluating transactions? (20/x)

Again, Valletti is unhappy with the @GAI_GMU interpretation of results of individual papers. Let’s take Crawford et al. as an illustrative example, where he claims we “misrepresented” the findings as showing a positive welfare effect. We didn't misrepresent anything. (21/x)

How about we hear from the *authors themselves*:
“On net, we find that the overall effect of vertical integration in the absence of effective program access rules—allowing for both efficiency and foreclosure incentives—is to increase consumer and total welfare on average."(22/x)

Maybe the EU approach to vertical mergers is so cramped that it only takes into account foreclosure and double marginalization. I doubt it. And it is definitely not so in the IO literature. Perhaps that accounts for some of the differences in emphasis here? (23/x)

Updating L&S (and its 739 cites!) is more than “econ jargon from the 1980s." Gerrymandered definitions of econ that exclude TCE might telling about EU v US differences. Attacks on individuals like Judge Ginsburg who have spent a career bringing econ to AT are telling too. (24/x)

But at the end of the day. None of the sound and fury from Tommaso changes the fundamental result does it? The @GAI_GMU conclusion that most vertical integration studies show either no net harm or net benefits stands easily even on his own cramped terms. / END