Super-Premium Ice Cream

In 2003 the U.S. Federal Trade Commission (FTC) challenged the merger between Dreyers Grand Ice Cream, Inc. and Nestle Holdings, Inc., citing that it would reduce competition in the super-premium ice cream market from three firms to two and would give the new company control of 60 percent of the market. This dominance, the FTC argued, would “lead to anticompetitive effects . . . including less product variety and higher prices.” When regulators evaluate proposed mergers and acquisitions based on potential antitrust concerns, they carefully consider, among other criteria, whether the new company’s impact on its industry’s competitive landscape will hurt consumers. To evaluate this impact, regulators use sophisticated econometric models to determine whether the newly merged firm will raise its prices.

However, current models do not consider how the newly merged firm will determine its product variety. A lack of competition could cause the firm to reduce its product variety, further harming consumers; however, the newly merged firm may in some circumstances offer a greater product variety, greatly benefiting consumers. The lack of a comprehensive model begs the question: Are antitrust authorities estimating the new company’s product variety correctly? Working to incorporate optimal product variety choices into existing models, Michael Mazzeo (Associate Professor of Management and Strategy at the Kellogg School of Management), Michaela Draganska (PhD from the Kellogg School of Management and Associate Professor of Marketing at the Stanford Graduate School of Business), and Katja Seim (Assistant Professor of Business and Public Policy at the Wharton School, University of Pennsylvania) studied the U.S. ice cream industry. Their research provides insights into what conditions result in merged companies providing greater product variety.

Ice Cream Market Research
The premium segment of the ice cream industry, just a notch below the above-mentioned super-premium segment, proved to be an ideal subject for Mazzeo, Draganska, and Seim for two reasons. First, although there are many regional ice cream companies, the industry has only two firms that compete nationally: Breyers and Dreyers. Because the two firms compete in all of the same geographic markets, the researchers could control for the effects that different geographies and regional competitors have on their product availability and sales. Second, data on ice cream purchases is readily available from Information Resources, Inc. (IRI), an information aggregation firm that collects data on consumer packaged goods companies.

The researchers chose to focus their study on the firms’ offerings of vanilla ice cream. Each firm offers as many as six different types of vanilla, depending on the geographic market. Narrowing the focus was necessary because evaluating the many flavors offered by the two firms would be extremely difficult. In addition, vanilla is the most popular flavor of ice cream, comprising more than 25 percent of all ice cream sales; restricting the study to vanilla would still produce data that fairly represented consumers’ ice cream purchases in general. Moreover, the number of different vanilla flavors offered by each firm in a given market is representative of the number of total flavors, both vanilla and otherwise, offered in that market.

Using this IRI data, the researchers developed a new model that forecasts post-merger product variety and the resulting consumer benefits. With their new model, they ran simulations on various merger scenarios between two firms in an oligopoly.

Model Findings
Based on these simulations, they found circumstances under which the newly merged firm would alter its product variety. If the fixed costs associated with offering additional product variety are sufficiently low, the merged firm would not increase variety and in some cases would decrease it. Because offering additional products is cheap, firms can profitably offer enough product variety to reasonably satisfy all consumers in the market. Therefore, when a merger occurs, the new firm has no reason to offer more products, and in some cases decreases its product offering. Prices increase while variety does not, so the merger ends up hurting consumers.

Conversely, if fixed costs associated with introducing new products are high, product variety may increase following a merger. Even though additional variety would be valuable to some groups of consumers, the effects of competition would reduce either the price or quantity of new products that firms can sell. Thus, in a competitive industry, investing in additional product variety beyond a certain point is not profitable. However, with the reduction in competition following the merger, firms can profitably offer these new products to better serve those smaller consumer segments. While price increases as a result of the firm’s increased pricing power, product variety also increases; the increased benefits to consumers may outweigh the higher prices.

In the case of ice cream manufacturers, there are fixed costs associated with creating new flavors, as well as fixed costs associated with the distribution of additional flavors to various geographic markets. The industry’s direct-to-store distribution network allows ice cream companies to sell their products directly to supermarkets rather than going through grocery warehouses. This arrangement gives ice cream makers greater control over which flavors to offer and forces them to pay the cost of “renting” freezer space from supermarkets to display their products, which means they directly see the costs and benefits of their decisions to expand their flavor line-up. If an ice cream maker decides to offer a new flavor in a region, it would incur a rental charge for additional freezer space or sacrifice existing freezer space used to sell other flavors. Based on the level of these fixed costs, the researchers’ simulations suggest that a merger of Breyers and Dreyers would be likely to change the number of vanilla flavors in many regions.

Although the Dreyers-Nestle merger was eventually approved—contingent on the new company selling several super-premium ice cream brands—further improvements of demand estimation and merger simulation techniques could have made such evaluations additionally precise. The research of Mazzeo, Draganska, and Seim offers antitrust regulators the beginnings of a new econometric model that would allow them to forecast changes to product variety resulting from a merger and, thus, the impact on consumers. Moreover, they demonstrate that the concept of “plain vanilla” is a misnomer, at least in the ice cream industry.