On January 27, the U.S. Federal Trade Commission announced a competition law remedy in respect of the Albertsons / Safeway grocery merger, requiring the divestiture of 168 supermarkets in 130 local markets in numerous states. This remedy is of particular interest from a Canadian competition law perspective in light of the recent completion of two Canadian grocery sector transactions, namely, Sobeys’ acquisition of Canada Safeway and Loblaw’s acquisition of Shoppers Drug Mart, both of which resulted in divestitures.
Several comparisons may be drawn between the approach taken by the U.S. FTC and the Canadian Competition Bureau in reviewing these mergers.
The U.S. and Canadian authorities both considered the question of the appropriate product market definition. They reached very similar conclusions.
In Sobeys, the Bureau concluded that the relevant product market was the retail sale of a full-line of grocery products that allows consumers to purchase the complete range of their grocery needs in a single location, and that, generally, stores with less than 10,000 square feet dedicated to grocery products do not offer the breadth and volume of products necessary to be considered a full-line grocery store. In relation to the key question of whether “conventional” grocery stores (such as Loblaws, Sobeys and Safeway) compete in the same product market as more “discount oriented” retailers (such as Price Chopper and Walmart Supercentre), the Bureau concluded that most discount-oriented grocery stores act as effective competitors and should therefore be included in the relevant product market, provided that they offer a full-line of grocery products and have at least 10,000 square feet dedicated to grocery products.
In Albertsons, the FTC reached a very similar conclusion. It found that the relevant product market was the retail sale of food and other grocery products in “supermarkets”, a term which was interpreted to mean any full-line retail grocery store that enables customers to purchase substantially all of their weekly food and grocery shopping requirements in a single shopping visit with substantial offerings in a wide array of product categories. The FTC went on to state that supermarkets typically carry more than 10,000 items and typically have at least 10,000 square feet of selling space. The FTC also explicitly found that the relevant product market includes supermarkets within “hypermarkets,” such as Wal-Mart Supercenters. At the same time, the FTC found that other types of retailers – such as hard discounters, limited assortment stores, natural and organic markets, ethnic specialty stores, and club stores – are not in the relevant product market because they offer a more limited range of products and services than supermarkets and because they appeal to a distinct customer type.
The Sobeys and Albertsons relevant product markets were therefore very similar. One distinction that may be drawn is that the Bureau clearly considered whether it was appropriate to bifurcate the market into “conventional” supermarkets versus “discount oriented” supermarkets (and ultimately decided not to do so) whereas the FTC apparently did not engage in an analysis of different types of supermarkets (but was clear to explicitly rule out all non-supermarkets, which the Bureau did implicitly). In the end result, the conclusion on the relevant product market in Sobeys and Albertsons was essentially identical.
This may be contrasted with the Bureau’s approach in Loblaw, which raised novel issues from a product market perspective. This is because while the parties sold many overlapping products, there was clearly a substantial amount of differentiation between the parties’ respective formats. Loblaw operated full-line grocery stores (which often include drugstore products) whereas Shoppers Drug Mart operated drug stores. The Bureau therefore chose not to address the question of relevant product market definition directly, instead concluding that, owing to the differentiation between the parties and to the level of diversity in the overlapping products, its review would focus primarily on the assessment of competitive effects rather than on precise market definition.
This leads to fascinating questions regarding the different analytical framework applied in the Sobeys and Loblaw transactions. In the former transaction, Shoppers Drug Mart was not considered to be in the same product market as Sobeys (and therefore would not have been in the same product market as Loblaw, given that Sobeys and Loblaw have identical formats). In the latter transaction, however, the Bureau moved beyond the structural question of market definition to a more direct assessment of competitive effects. More particularly, in Loblaw, the Bureau focused on two broad categories of products: (i) pharmacy products, including prescription medicine, over-the-counter and behind-the-counter medications; and (ii) drug-store type merchandise including “health and beauty aids”, cosmetics and “drugstore-type food” (such as beverages, snacks, milk and limited dairy and bread products). Had the Bureau instead focused on the structural question of product market, it would have been extremely difficult to conclude that Loblaw and Shoppers Drug Mart belong in the same product market, in light of the conclusion reached in Sobeys, a few months earlier.
In terms of future guidance for future retail mergers, the implication of the Bureau’s different approach in Sobeys and Loblaw is that the merging parties should carefully consider the degree of differentiation between their stores. To the extent that they operate using the same format, then the Bureau may carry out a traditional, structural, product market definition exercise capturing the entire range of products sold by the parties in a single product market (which may exclude firms using different formats). To the extent that they operate using a differentiated format, then the Bureau may instead carry out a direct effects analysis in relation to specific overlapping categories of products.
In relation to geographic market definition, the FTC (in Albertsons) and the Bureau (in Sobeys and Loblaw) concluded that the relevant geographic markets are local, with the specific size depending on factors such as whether the market was urban, suburban or rural; the existence of natural barriers, such as rivers and highways; and traffic patterns. The FTC referred to relevant geographic areas ranging from a two to ten mile radius around the parties’ grocery stores. The Bureau did not refer to a specific radius, but did refer to customer surveys, loyalty card data and the use of geomatics software and computerized tools, and clearly engaged in highly localized geographic analysis.
In Sobeys, after having defined the relevant product and geographic markets, the Bureau’s position statement simply states that there were certain local markets where market shares were high, competition was limited, barriers to entry were high and that therefore the transaction would result in a substantial lessening or prevention of competition in the retail sale of a full-line of grocery products in those local markets. It required the divestiture of 23 stores.
In Loblaw, the Bureau’s position statement describes in more detail the extent of the analysis it conducted to assess competitive effects, including the use of internal and external economic experts to analyse pricing and sales data, and the use of a cross-sectional methodology to quantify the competitive impact of stores owned by each party. The Bureau states that this analysis resulted in the conclusion that there was material price competition between the parties. However, as in Sobeys, the position statement then simply concludes that the transaction would result in a substantial lessening of competition in 27 local markets (without describing the specific characteristics of those markets).
In Albertsons, the FTC describes its competitive effects analysis in purely structural terms. It relied heavily on post-merger HHIs and HHI deltas, and the presumptions of illegality contained in the U.S. Horizontal Merger Guidelines in relation to these metrics, thereby reaching the conclusion that divestitures would be required in 130 local markets. The FTC’s complaint also includes a table setting out the pre and post-merger HHIs, HHI deltas, and the number of effective remaining competitors for each of the divestiture markets.
In summary (but based only on the public disclosure of the two agencies), it would appear that the FTC applied a rigid, structural analysis in Albertson, whereas the Bureau may have engaged, to a much greater degree, in a more finely-tuned direct effects analysis, taking into account, based on cross-sectional data, the actual influence of each of the parties on each other in particular geographies. At the same time, however, this difference may simply be attributed to the fact that, at least in Loblaw, the Bureau was faced with a vastly more complicated analytical problem, given that the parties were considerably differentiated Finally, in terms of the presentation of their conclusions, the FTC was quite transparent in disclosing the structural market characteristics of those markets where divestitures where required, whereas the Bureau’s position statements do not allow the reader to fully understand what it is about those particular markets that led to divestitures being required.