November 16, 2011 3:13 PM

H&R Block, AT&T, and the Error of Recency

Last week, the U.S. District Court for the District of Columbia, released its opinion granting the Antitrust Division of the U.S. Department of Justice a preliminary injunction preventing H&R Block ("HRB") from acquiring the stock of "2SS Holdings, Inc.", the maker of TaxACT, a digital "do it yourself" ("DDIY") tax preparation software.  

Since this news was released on November 2nd, many have speculated, opined and hypothesized--without basis--that this decision does not bode well for AT&T's acquisition of T-Mobile USA from Deutsche Telecom.  According to just about everyone, the impact was decidedly positive for the DoJ's chances to win its case against AT&T/T-Mobile if it goes to trial next February in Judge Huvelle's courtroom.

The simplified reasoning was that because the government had just prevailed on its most recent horizontal merger challenge, it will likely prevail on its next horizontal merger challenge in the same district.  These stories were the predictable result of a well-known human behavioral bias, known as the "error of recency"--the notion that humans tend to overweight the value of recent actions. While the "hot hand fallacy" and the "gambler's fallacy" were identified with respect to how humans approach "random" (i.e., "unrelated") events over 300 years ago, these biases continue to persist.

What is interesting, though, is that the "hot hand fallacy" infected all major stories reporting, or commenting on, the event.  The result was that the "legitimate" news stories (e.g., Bloomberg and Reuters) were remarkably similar to the interest group blogs.  But, since no one has yet attempted to offer any perspective on the HRB case (unaffected by the error of recency), let's look at how the two cases might be perceived differently by a different court, with a different set of facts.

Market Definition

HRB

This is perhaps the most obvious difference between the two cases.  The HRB case was, strategically, much more of a traditional horizontal merger case.  The basic strategy in a horizontal merger case is for the plaintiff to seek to define the markets (product and/or geographic) very narrowly, and for the defendants to seek the broadest possible market definition.  For example, if the alleged market was soft drinks, defendants would want to argue that all non-alcoholic beverages, including tap water, should be included in the product market.  

It would not be going too far to say that HRB was all about product market definition.  In fact, the court devoted more than half of its opinion to evaluating each party's claims regarding the properly-defined product market.  Only if the court agreed with the DoJ's  contention that DDIY software was a discreet product market, would the government have been able to show sufficiently high concentration numbers to make its case that this merger would lessen competition.

Defendants, on the other hand, were arguing that the market also included professionally-assisted tax preparation, and (fatally) consumers that file their tax returns without any assistance.  The court found this definition impermissably broad, because the inclusion of "pen and paper" filers distorts the market because these filers were not purchasing any product or service, but merely performing a legally-compelled "chore."

Ultimately, the court found the DoJ most persuasively defined the relevant product market.  After adopting "DDIY" as the product market, the concentration numbers were substantial.  The market leader, Intuit, had around a 62% market share, with HRB and 2SS coming in second and third with approximately 15.5% and 13% shares, respectively. Now, let's compare the market definition facts of the AT&T case.

AT&T/T-Mobile

Unlike the HRB merger, there are unlikely to be many, if any, novel proposed market definitions presented by this proposed wireless acquisition versus any of the many others for which the DoJ has alleged the same product and geographic markets it is alleging in the present complaint.
 
Competitive Effects--HRB vs. AT&T/T-Mobile

Contrary to the assumptions of some commenters, however, mergers are not evaluated solely on concentration numbers.  The HRB court, relying on U.S. v. Baker Hughes, explained that "[t]he Herfindahl-Hirschman Index cannot guarantee litigation victories." Opinion at 53 of 86 (internal citations omitted).  So let's compare the alleged anticompetitive effects of both proposed mergers.

First, we have to recognize that under the worst case for AT&T/T-Mobile, almost every geographic market starts with much lower concentration numbers--and smaller increases in concentration due to the merger--than in HRB.  The higher initial concentration numbers, and the greater changes in concentration in HRB, make it easier to understand how the simple removal of one vigorous competitor (the court eschewed the term "maverick") could have an anticompetitive effect.

Second, consider also that the DDIY tax preparation market was a differentiated product market, in which the proposed acquiring/acquired firms were each other's closest substitutes.  Thus, it was fairly easy to understand the DoJ's unilateral effects theory--that HRB could raise the price on its "high end" DDIY services and still capture lost sales through its ownership of the "low end" acquired brand. 

On the other hand, it is unclear whether any true "differentiation" exists in the wireless mobile telecommunications service market.  Unlike in HRB, the government has not alleged that retail consumers perceive AT&T and T-Mobile to be each other's closest substitutes.  Thus, it seems unlikely that this merger will provide the post-merger firm with the opportunity to execute a unilateral price increase.

The Effects of H&R Block on the AT&T/T-Mobile Merger

Did you really think I'd try to answer that question?  I have no idea.  It's obvious that the facts of the HRB merger made it into a more traditional battle over a narrow vs. broad product market.  On the other hand, from the beginning the AT&T/T-Mobile merger has been about competitive effects.  All we really know is that, while it's a gamble to try to predict the outcome of a case based on oral arguments, it's an odds-against bet to try to predict the outcome of one case based on the near-term results of an unrelated case. 
 

Leave a comment