Mike Sykuta and I recently co-authored a short article discussing the latest evidence on, and proper legal treatment of, minimum resale price maintenance (RPM). Following is a bit about the article (which is available here).
Despite the U.S. Supreme Court’s Leegin decision holding that minimum RPM must be evaluated under antitrust’s Rule of Reason, the battle over the proper legal treatment of the practice continues to rage at both the federal and state levels.
At the federal level, courts, commentators, and regulators have split over what sort of Rule of Reason should apply. Some, like yours truly, have argued that because RPM is usually pro- rather than anticompetitive, challengers should bear the burden of proving likely anticompetitive effect (at a minimum, the structural prerequisites to such an effect) under a full-blown Rule of Reason. Others contend that RPM should be assessed using some version of “quick look” Rule of Reason, under which a challenged instance of RPM is presumed anticompetitive if the plaintiff makes some fairly narrow showing (e.g., that consumer prices have risen, or that the RPM was dealer-initiated, or that the RPM is imposed on homogeneous products that are not sold with dealer services that are susceptible to free-riding).
At the state level, a number of states have simply decided not to follow Leegin and to retain, under state antitrust law, the per se rule of Dr. Miles (the 1911 decision overruled by Leegin). At least nine states have taken this tack.
We advocates of a full-blown Rule of Reason for minimum RPM have generally emphasized two things. First, we have observed that while the structural prerequisites to RPM’s potentially anticompetitive harms (facilitation of dealer-level or manufacturer-level cartels, or exclusion by a dominant dealer or manufacturer) are rarely satisfied, the necessary conditions for RPM’s procompetitive benefits (avoidance of free-riding, facilitating entry, encouraging non-free-rideable dealer services) are frequently met. Second, we have shown that the pre-Leegin empirical evidence on RPM’s effects generally confirmed what theory would predict: Most instances of RPM that have been examined closely have proven output-enhancing.
In recent months, advocates of stricter RPM rules have pointed to an ambitious new study that they say supports their position. The study authors, University of Chicago economics PhD candidates Alexander MacKay and David Aron Smith, purported to conduct “a natural experiment to estimate the effects of Leegin on product prices and quantity.” In particular, MacKay & Smith compared post-Leegin changes in price and output levels in states retaining a rule of per se illegality with those in states likely to assess RPM under the Rule of Reason. Utilizing Nielsen consumer product data for 1,083 “product modules” (i.e., narrowly defined product categories such as “vegetables-broccoli-frozen”), the authors assessed price and output changes between the six month period immediately preceding Leegin (January-June 2007) and the last six months of 2009.
With respect to price changes, MacKay & Smith found that 15% of the product modules exhibited price increases that were higher, by a statistically significant margin, in Rule of Reason states than in per se states. In only 6.9% of modules were price increases higher, to a statistically significant degree, in per se states than in Rule of Reason states. With respect to quantity changes, 14.7% of modules saw a statistically significant decrease in quantity in Rule of Reason states versus per se states, whereas only 3% of modules exhibited a statistically significant quantity increase in Rule of Reason states over per se states. MacKay & Smith thus conclude that greater leniency on minimum RPM is associated with higher prices and lower output levels, a conclusion that, they say, supports the view that RPM is more frequently anticompetitive than procompetitive.
Mike and I contend that the MacKay & Smith study is flawed and does not justify restrictive RPM policies. First, the study provides very little support for the view that RPM has caused anticompetitive harm within the group of product markets examined. As an initial matter (and as the authors admit), the study does not demonstrate that actual RPM agreements have caused anticompetitive harm in the post-Leegin era. To make such a showing, one would have to demonstrate that (1) minimum RPM was actually imposed on a product after the Leegin decision, (2) the RPM policy raised the price of that product from what it otherwise would have been, and (3) the quantity of the product sold fell from what it otherwise would have been. The authors present no evidence that RPM policies were actually implemented on any of the product categories for which they identified statistically significant price increases and quantity decreases. As they concede, their study could show only that legal environments treating RPM leniently (not RPM agreements themselves) are conducive to anticompetitive outcomes.
But the authors’ data provide little support for even that claim. To prove anticompetitive harm stemming from an “RPM-permissive” legal environment, one would have to show that the transition from per se illegality to rule of reason treatment occasioned, for a substantial number of products, both a statistically significant price increase and a statistically significant output reduction on the same product. An output reduction not accompanied by an increase in price suggests that something besides minimum RPM (or even a “permissive attitude” toward RPM) caused output to fall. A price increase without a reduction in output is consistent with the view that RPM induced demand-enhancing dealer activities that mitigated the effect of the price increase, albeit by not as much as the producer may have hoped. (A price increase without an output decrease could also indicate that demand for the product at issue was inelastic, but MacKay & Smith presented no evidence suggesting that demand for any of the product categories exhibiting price increases but not quantity decreases was particularly inelastic.)
According to the authors’ list of “modules with significant price or quantity changes” (Appendix A of their study), only 17 of the 1,083 product categories examined—a mere 1.6%—exhibited both a price increase and a quantity decrease. And those effects were for categories of products (e.g., barbecue sauces as a whole), not necessarily particular brands of a product (e.g., KC Masterpiece or Sweet Baby Ray’s). It could well be that within the 1.6% of categories exhibiting both an average price increase and an average output decrease, there were no individual brands exhibiting both effects at once. Indeed, most of the seventeen product categories involve dealer and manufacturer markets that are neither cartelizable (so neither the dealer nor manufacturer collusion theory of anticompetitive harm could apply) nor dominated by a powerful manufacturer or dealer (so neither the dominant manufacturer nor dominant dealer theory could apply). To the extent MacKay & Smith’s findings provide any evidence that RPM-permissiveness occasions anticompetitive harm in household consumer products markets, that evidence is awfully thin.
Moreover, in limiting their examination to the product categories included in the Nielsen Consumer Panel Data, MacKay & Smith excluded most products for which one of the procompetitive rationales for minimum RPM—the “avoidance of free-riding” rationale—would apply. As the authors observe, only about “30% of household consumption is accounted for by the categories in the data.” That 30% is comprised mainly of groceries, other consumable household products, and small appliances. The study thus excludes data related to purchases of large appliances, complicated electronics projects, and other relatively expensive products that are frequently sold along with “free-rideable” amenities such as product demonstrations, consumer education, and set-up or repair services. Because the MacKay & Smith study systematically disregards information on transactions likely to reflect a procompetitive use of minimum RPM, it fails to establish the authors’ conclusion that “the harm to consumers resulting from rule-of-reason treatment of minimum RPM seems to outweigh its benefits.”
In the end, then, Mike and I conclude that the new RPM evidence provides no reason to reject the persuasive theory- and evidence-based arguments in favor of lenient, full-blown Rule of Reason treatment of minimum RPM. Of course, we welcome comments on our article.