How Do You Judge Success of FTC Merger Enforcement?

A seemingly innocuous paragraph buried on page 21 of the Federal Trade Commision’s Strategic Plan for Fiscal Years 2006-2011 has sparked a lively discussion on the ABA Antitrust Section’s listserv about how the FTC determines the success of its merger enforcement efforts.

Last Monday, David Balto, a former antitrust enforcer at both the FTC and Department of Justice Antitrust Division, started the discussion by pointing out that the FTC has measured the success of its Hart-Scott-Rodino “second request” process “by assessing the number of second requests that result in a positive outcome (a challenge, consent, or the deal is dropped) against the total number of second requests.”

In an earlier strategic plan, the FTC set a goal of 60-80% "positive outcomes." In 2005, according to Balto, the FTC had 52% “positive outcomes.” Nevertheless, the FTC’s strategic plan for 2006-2011 raised the bar to 90%:

For each year 2006-2011:

Continue effective administration of the HSR program so that at least 90% of HSR request for additional information result in a positive outcome, which includes Commission authorization of a complaint for preliminary injunction in federal court, issuance of a an administrative complaint, acceptance of a consent agreement, the parties’ voluntary abandonment or restructuring of a proposed transaction based on FTC antitrust concerns, and closing of an investigation without subsequent events indicating that the transaction injured competition.

www.ftc.gov/opp/gpra/spfy06fy11.pdf

Balto asked a series of questions, including:

“Why would the FTC increase the goal from 60-80% to 90%?”

“If 90% is the goal, won’t the agency tend to miss some worthwhile investigations? Would that suggest that the agency is underenforcing?”

Mark Whitener, also a former FTC enforcer, observed that the apparent change may not reflect “an actual change in policy in favor of more deal challenges,” since the term “positive outcome” now includes “closing of an investigation without subsequent events indicating that the transaction injured competition.”

According to Denis Paul, including cleared transactions that did not later result in injury to competition renders the statistic completely ineffective for its purpose—namely, to measure the efficiency of the FTC’s second request decision-making process. The issuance of a second request on a transaction later proven to been competitively benign is an inefficient use of resources, Paul asserted, whereas the issuance of a second request on a transaction that leads to enforcement is a more efficient use of resources.

Jon Putman chimed in that the “success rate” is a function of what deals are proposed. “If you raise the standard for challenge, people propose fewer borderline deals. So you actually understate the effect of the tightening standards if you restrict the sample to the deals that are proposed once standards are tightened.”

Cecile Kohrs Lindell disagreed that the policy necessarily “raises the standard for challenge.” It seems to indicate only that there will be fewer challenges. She wondered about the value of the percentage figures. When considering the acceptances of consent agreements, “there’s no way to assess whether the remedy is broader or narrower than it might have been on any continuum of enforcement.” Finally, she is not sure what a “negative outcome” would be: “Only closings where it could be proved there was an anticompetitive result within what period of time?”

A negative outcome, John Pisarkiewicz wrote, would be “no challenge, no consent agreement, and the merger goes through.” The FTC wishes to minimize negative outcomes because "spending resources to investigate a proposed acquisition which ultimately poses no harm to competition wastes resources.”

Luke Froeb, former Director of the FTC Bureau of Economics, remarked that there are two mistakes that can be made in challenging mergers: you can challenge pro-competitive mergers (Type I error) or fail to challenge anticompetitive mergers (Type II error). “In general there is a tradeoff between Type I and Type II errors (a reduction in one means more of the other). The only way you can simultaneously reduce both is to gather more information (which increases the cost of enforcement and compliance).” He interprets the FTC 90% goal as “an effort to reduce the costs of wasteful enforcement and compliance—the kind of investigation and compliance that does nothing to increase accuracy.”

In response, Jon Putman wrote that he agreed with that statement of the policy objective, but questioned whether the proposed statistics “tell you anything about whether the FTC is getting closer to that objective.”

Ken Davidson, another former FTC lawyer, weighed in that the “benchmarks on HSR filings/Second Requests/Consents are a poor substitute for an evaluation of merger effectiveness.” He related how he was involved in meetings in which the benchmarks were being developed. “Everyone was aware that by counting second requests and consents, we would distort bureaucratic incentives, but no one could think of anything that could be done annually that would reflect the efficacy of the merger program.”