No Facts, No Problem?
There has been, as is to be expected, plenty of casual analysis of the AT&T / T-Mobile merger to go around. As I mentioned, I think there are a number of interesting issues to be resolved in an investigation with access to the facts necessary to conduct the appropriate analysis. Annie Lowery’s piece in Slate is one of the more egregious violators of the liberal application of “folk economics” to the merger while reaching some very confident conclusions concerning the competitive effects of the merger:
Merging AT&T and T-Mobile would reduce competition further, creating a wireless behemoth with more than 125 million customers and nudging the existing oligopoly closer to a duopoly. The new company would have more customers than Verizon, and three times as many as Sprint Nextel. It would control about 42 percent of the U.S. cell-phone market.
That means higher prices, full stop. The proposed deal is, in finance-speak, a “horizontal acquisition.” AT&T is not attempting to buy a company that makes software or runs network improvements or streamlines back-end systems. AT&T is buying a company that has the broadband it needs and cutting out a competitor to boot—a competitor that had, of late, pushed hard to compete on price. Perhaps it’s telling that AT&T has made no indications as of yet that it will keep T-Mobile’s lower rates.
Full stop? I don’t think so. Nothing in economic theory says so. And by the way, 42 percent simply isn’t high enough to tell a merger to monopoly story here; and Lowery concedes some efficiencies from the merger (“buying a company that has the broadband it needs” is an efficiency!). To be clear, the merger may or may not pose competitive problems as a matter of fact. The point is that serious analysis must be done in order to evaluate its likely competitive effects. And of course, Lowery (HT: Yglesias, ) has no obligation to conduct serious analysis in a column — nor do I in a blog post. But this idea that the market concentration is an incredibly useful and — in her case, perfectly accurate — predictor of price effects is devoid of analytical content and also misleads on the relevant economics. Quite the contrary, so undermined has been the confidence in the traditional concentration-price notions of horizontal merger analysis that the antitrust agencies’ 2010 Horizontal Merger Guidelines are premised in large part upon the notion that modern merger analysis considers shares to be an inherently unreliable predictor of competitive effects!! (For what its worth, a recent Wall Street Journal column discussing merger analysis makes the same mistake — that is, suggests that the merger analysis comes down to shares and HHIs. It doesn’t.)
To be sure, the merger of large firms with relatively large shares may attract significant attention, may suggest that the analysis drags on for a longer period of time, and likely will provide an opportunity for the FCC to extract some concessions. But what I’m talking about is the antitrust economics here, not the political economy. That is, will the merger increase prices and harm consumers? With respect to the substantive merits, there is a fact-intensive economic analysis that must be done before anybody makes strong predictions about competitive effects. The antitrust agencies will conduct that analysis. So will the parties. Indeed, the reported $3 billion termination fee suggests that AT&T is fairly confident it will get this through; and it clearly thought of this in advance. It is not as if the parties’ efficiencies contentions are facially implausible. The idea that the merger could alleviate spectrum exhaustion, that there are efficiencies in spectrum holdings, and that this will facilitate expansion of LTE are worth investigating on the facts; just as the potentially anticompetitive theories are. I don’t have strong opinions on the way that analysis will come out without doing it myself or at least having access to more data.
I’m only reacting to, and rejecting, the idea that we should simplify merger analysis to the dual propositions — that: (1) an increase in concentration leads to higher prices, and (2) when data doesn’t comport with (1) we can dismiss it by asserting without evidence that prices would have fallen even more. This approach is, lets just say, problematic.
In the meantime, the Sprint CEO has publicly criticized the deal. As I’ve discussed previously, economic theory and evidence suggest that when rivals complain about a merger, it is likely to increase competition rather than reduce it. This is, of course, a rule of thumb. But it is one that generates much more reliable inferences than the simple view — rejected by both theory and evidence — that a reduction in the number of firms allows leads to higher prices. Yglesias points out, on the other hand, that rival Verizon prices increased post-merger (but did it experience abnormal returns? What about other rivals?), suggesting the market expects the merger to create market power. At least there we are in the world of casual empiricism rather than misusing theory.
Adam Thierer at Tech Liberation Front provides some insightful analysis as to the political economy of deal approval. Karl Smith makes a similar point here.
Filed under: antitrust, economics, merger guidelines, mergers & acquisitions, technology