This report analyzes the U.S. carbonated soft drink ("CSD") industry, with its primary focus on the 1980s and early 1990s, a period of rapid structural change that transformed the industry. In addition to documenting these changes, an empirical model is developed to evaluate the antitrust merger policies that were pursued by the Federal Trade Commission ("FTC") during this period -- the FTC challenged large horizontal acquisitions of Dr Pepper and 7UP franchises by Coca-Cola and Pepsi-Cola bottlers, but did not challenge vertical acquisitions of CSD bottlers by their franchisors or other horizontal bottler acquisitions. Our findings tend to support or are consistent with these policies, but also identify areas that seem to warrant further study.
Until 1980, the fragmented independent franchised bottling distribution system that had characterized the industry since before the turn of the century was still in place. Bottlers held perpetual franchises with exclusive territories and were bound by flavor exclusivity clauses, as is true today. Since 1980, the number of bottlers with franchises of the major CSD brands has fallen by more than one-half, as franchised bottlers were acquired and consolidated by their franchisors and by other bottlers. In addition to FTC merger enforcement activities in the CSD industry, the Department of Justice brought many price-fixing cases in the mid- to late 1980s against CSD bottlers affiliated with each of the leading concentrate firms. By 1990, it had obtained more than forty bottler and individual guilty pleas or convictions in ten states.
The bottler acquisitions that took place during the study period are the main focus of this report. Alternative theories for each type of bottler transaction are summarized. The specific hypotheses we test, using price and output measures of competitive effects, are (1) whether horizontal transfers of Dr Pepper and/or 7UP franchises to Coca-Cola or Pepsi-Cola bottlers were anticompetitive (i.e., associated with higher prices and lower output), (2) whether vertical acquisitions by the Coca-Cola Company and/or PepsiCo of their respective bottlers were procompetitive (i.e., associated with lower prices and higher output), and (3) whether consolidations of third bottler franchises (i.e., franchises not held by a Coca-Cola or Pepsi-Cola bottler) were procompetitive.
The empirical model developed to test these three hypotheses includes qualitative variables to examine the competitive impacts that these types of events have on CSD prices and per capita volumes. The model also contains other control factors that may affect CSD prices and per capita volumes. These control variables include sets of factors that would affect the demand, supply, and market structure for CSDs.
Three different cross-section/time-series data sets were compiled to estimate the empirical model. Each data set contains dozens of local areas, and together they span more than 10 years. Separate CSD price and per capita volume regressions were run for each of these three data sets. The application of the model to three different data sets permits us to evaluate the robustness of the parameter estimates, including those that have public policy implications.
This study represents a substantial improvement over earlier CSD research efforts because (1) it considers a variety of events corresponding to a wide range of policy questions, including horizontal acquisitions and third bottler consolidations, rather than being limited to vertical integration; (2) it examines CSD performance during three periods spanning more than ten years, rather than being limited to a single relatively short-term time horizon; (3) it uses both CSD price and per capita volume regressions (rather than one or the other) to evaluate CSD performance; (4) it examines local CSD performance across all of the major CSD brand groups, rather than relying exclusively on individual company (and individual package size) observations, or aggregating private label and warehouse brand sales with sales of major brands; (5) all of its regression results are based on data for dozens of local areas, rather than using a handful or fewer local areas to perform empirical tests; and (6) it includes a more complete set of explanatory variables.
Of the three types of events analyzed, the regression results were strongest for the horizontal Dr Pepper and 7UP franchise acquisitions by Coca-Cola and Pepsi-Cola bottlers. Our specific findings include:
- Horizontal franchise acquisitions by Coca-Cola and Pepsi-Cola bottlers led to higher CSD prices and lower per capita CSD volumes, as hypothesized. On average, these transactions were associated with CSD prices that were 3.5%-12.8% higher than otherwise, and per capita CSD volumes that were 12.2%-19.8% lower than otherwise.
- Vertical integration was associated with lower CSD prices for alternative measures of the degree of vertical integration (as hypothesized), but had mixed results in the per capita CSD volume regressions using the three data sets. On average, vertical acquisitions that resulted in both the Coca-Cola Company and PepsiCo controlling their bottlers lowered CSD prices by 4.3%.
- The results for third bottler consolidations varied with the local market shares of the franchises being acquired. On average, large franchise acquisitions were associated with lower CSD prices (1.2%) and higher per capita CSD volumes (14.0%). In contrast, small franchise acquisitions were associated with higher CSD prices (5.5%) and lower per capita CSD volumes (13.2%), on average.
Overall, the results are generally consistent with prior expectations and with recent antitrust policy in the CSD industry. However, some results, particularly those associated with vertical integration, suggest that further study is warranted.