Statistical Programs Under Antitrust Scrutiny

Date: Apr 24, 2013

Information exchanges, such as statistical programs, drew considerable attention at the marquee Spring Meeting of the American Bar Association Section of Antitrust Law this month. Two cases merit consideration because there was no evidence in either case that the information exchanges lead to price fixing or anticompetitive effects, but government prosecutors used the information exchanges to support price fixing charges.

The takeaway for those of us that counsel or review information exchanges is the need for a broader, and more measured evaluation of the context in which the information exchanges will occur, coupled with an increased need to articulate, in the association's decision making materials, the procompetitive business reasons supporting the information exchange and why this information exchange is a practical necessity to achieve the business goals.

Ductile Iron Pipe Fittings

The first case was a Federal Trade Commission (FTC) proceeding involving manufacturers of ductile iron pipe fittings. Standing alone, the information exchange itself seemed quite acceptable. Based on guidance from experienced antitrust council, the Ductile Iron Fittings Research Association (DIFRA) worked through a third part accounting firm who gathered and disseminated aggregated data for the industry by tons shipped. The thousands of products were combined into a handful of broad categories. In the aggregated reporting there was no differentiation between domestic and imported goods. Shipments were reported on a nationwide basis while pricing in this particular industry is local. Prices are typically set months before the shipments, so the reported data was not new and not directly related to price. It appears that aggregated data was only reported in 2008-2009. As for behavior, one of the three largest fitting companies lowered their published prices after their first and last reports after 2008 and 2009.

The FTC's complaint alleged that a conspiracy began in early 2008 supported by a number of signaling mechanisms, including the association's statistical program. FTC attorneys argued that the information exchange was prima face anticompetitive and, therefore, the burden of proof shifted to the fittings manufacturers to substantiate an efficiency or commercial necessity defense. The FTC claimed that, since the fittings manufacturers had failed to substantiate their defense, they violated Section 5 of the Federal Trade Commission Act.

Further, the information exchange requirements imposed in the FTC consent order go beyond the standard safe harbor restrictions. For example, data needed to be more than six months old, rather than the usual three months safe harbor criteria. Similarly, the data for any three firms could not account for more than 60% of the aggregated transactions, and the statics could not be reported more than twice a year. Certainly the FTC action did not rest solely on an information exchange; other market dynamics were present including a highly concentrated industry where three companies represented more than 90% of the market. Nevertheless, if asked to evaluate the propriety of the fittings information exchange program as originally structured, many practitioners would have concluded that the statistical program was appropriate as designed.

The FTC cases are: In re Sigma Corp., Docket No. C-4347 (FTC Consent Agreement Jan. 4, 2012) and In re McWane and Star Pipe, Inc., Docket No. 9351. Sigma and Star Pipe entered into consent agreements with the FTC; McWane did not enter into a settlement and the case is currently pending before the FTC.

Detroit Nurses

In contrast to the ductile iron pipe fittings cases, the conduct in the Detroit Nurses case tended to ignore the standard guidelines regarding information exchanges. Besides third party surveys, the participants exchanged salary information directly with each other, through industry groups, and through ad hoc discussions including both current and future compensation plans. While the behavior may be lamentable, there was no evidence that these exchanges resulted in price fixing. Indeed, different actions were taken by each participant with regard to employee compensation.

In 2006, registered nurses filed five lawsuits alleging that the hospitals' information sharing was one basis for fixing the wages paid to registered nurses. Despite the lack of evidence that the hospitals actually coordinated or jointly fixed wages, the district court allowed one of the cases to proceed to trial. Cason-Merenda v. Detroit Medical Center ("Detroit Nurses"), 862 F. Supp. 2d 603 (E.D. Mich. 2012).

Evaluating Information Exchanges

Legal commentators expressed concern that, in a departure from prior precedent, neither parallel conduct nor parallel outcome was required to prove a per se price fixing conspiracy. Others are questioning whether the safe harbor zones for information exchanges first articulated in the 1996 FTC and Justice Department Health Care Guidelines are now morphing into an outer limit. See, FTC and the Department of Justice Antitrust Division, Statements of Antitrust Enforcement Policy in Health Care (1996) (describing safe harbor levels as five or more participants with no one company having more than 25% market share of each reported category).

Associations need to be more vigilant in developing carefully articulated, procompetitive business reasons why the information exchange is a practical necessity. The justifications should appear in the association's program development documentation or related minutes. Post hoc rationalizations will not carry much weight.

Regrettably, there appeared to be scant market analysis in either case. While only the act of fixing prices need be proven, the question of whether there was anticompetitive effect was not addressed by the government. In both cases, the connection between the statistical program and any actual price fixing was highly speculative.

While neither case directs us to change our current criteria for assessing the propriety of association statistical programs, both cases suggest that an analysis of a statistical program in isolate may be inadequate. The assessment must include consideration of the industry and the association's overall activities and structure. This can only be achieved if we are familiar with the industry, the association and its activities, staff and members. An oversimplified reliance on safe harbor numbers may leave one adrift on uncharted seas.

For more information, please contact Peter de la Cruz at 202.434.4141 or at delacruz@khlaw.com.