Do Mergers Raise Consumer Prices?

Bhattacharya, Illanes, and Stillerman (2026) ask how mergers affect consumer prices and whether U.S. antitrust enforcement effectively identifies harmful mergers. They examine 129 consumer product markets involved in 47 mergers from 2006 to 2017, using NielsenIQ retail scanner data. They find that mergers increased prices by only 0.4% on average, but effects varied substantially. Twenty-five percent of mergers raised prices by more than 3.9%, while another quarter reduced prices by more than 2.1%. They estimate that antitrust agencies generally challenge mergers expected to increase prices by roughly 4.8% to 6.3%.

Why This Article Was Selected for The Policy Scientist

This article addresses a central policy question: whether antitrust enforcement can distinguish mergers that harm consumers from those that do not. That issue is especially timely amid renewed debate over whether merger enforcement has been too lax. The article makes an important contribution to the merger-retrospective literature by systematically examining a broad sample rather than selected, prominent mergers. The NielsenIQ scanner data are unusually extensive, and the statistical analysis provides a credible event-study framework with extensive robustness tests, although I generally prefer stronger causal inference designs. The findings may inform other jurisdictions, but differences in markets and enforcement institutions limit direct generalization.

Full Citation and Link to Article

Bhattacharya, V., Illanes, G., & Stillerman, D. (2026). Merger effects and antitrust enforcement: Evidence from US consumer packaged goods. American Economic Review. Advance online publication. https://doi.org/10.1257/aer.20240497

Central Research Question

Bhattacharya, Illanes, and Stillerman examine a fundamental question in antitrust policy: What happens to consumer prices and other market outcomes after firms merge, and how effectively do federal antitrust agencies identify mergers likely to increase prices? The authors study mergers among manufacturers of consumer packaged goods, including food, beverages, cosmetics, household products, and other items commonly sold in grocery and mass-market stores. Their first objective is to estimate how completed mergers affect prices, sales quantities, product offerings, and distribution. Their second objective is to assess the stringency of merger enforcement by estimating the price increase that appears to trigger greater scrutiny from the Federal Trade Commission and Department of Justice. The authors then simulate stricter enforcement regimes and examine the trade-off between allowing price-increasing mergers, blocking price-decreasing mergers, and increasing the number of cases agencies must investigate.

Previous Literature

The study contributes to a substantial literature on the economic consequences of mergers. Earlier researchers frequently conducted detailed retrospective studies of individual mergers or small groups of mergers. These studies examined industries such as airlines, consumer products, appliances, beer, hospitals, and gasoline. Although such research provides detailed evidence, the selection of prominent or controversial mergers for study can create publication and case-selection biases. A merger that generates large price increases may be more likely to attract researchers and appear in the published literature than a merger with little measurable effect.

More recent research has assembled larger collections of mergers. Studies have examined consumer goods, hospitals, pharmaceuticals, and other sectors. Some researchers have also moved beyond prices to study product offerings, efficiency, or changes in geographic service networks. Other studies have examined whether market concentration measures predict merger outcomes or have estimated the causal effects of antitrust enforcement itself.

The authors build particularly on Carlton’s argument that the average effect of completed mergers cannot, by itself, establish whether antitrust enforcement is strict or lax. Enforcement agencies attempt to screen out the mergers they expect to be most harmful. Consequently, observed mergers are already a selected group. Even an effective enforcement system could permit mergers with small average price effects while preventing mergers expected to cause much larger increases. The authors therefore connect observed post-merger outcomes to an explicit model of agency decision-making under uncertainty.

Data

The study uses a particularly extensive data set covering U.S. consumer packaged goods mergers from 2006 through 2017. The authors begin with mergers recorded in the SDC Platinum database and identify transactions involving manufacturers whose products are sold in retail stores. They restrict the analysis to transactions worth at least $280 million and identify cases in which the acquiring firm and target competed in at least one product market and geographic area. Their final sample contains 129 product market-merger combinations generated by 47 corporate transactions.

The primary outcome data come from the NielsenIQ Retail Scanner Dataset. Depending on the year, the database follows approximately 2.6 million to 4.5 million individual products and includes scanner information from 35,000 to 50,000 grocery, drug, mass merchandise, and other retail stores. NielsenIQ reports weekly store-level sales and average transaction prices for individual Universal Product Codes. The data cover more than half of total U.S. grocery and drug-store sales volume and more than 30 percent of mass-merchandiser sales volume.

The authors supplement the scanner data with Euromonitor Passport information to identify product ownership, Federal Reserve Economic Data commodity price indices to measure changes in input costs, and American Community Survey estimates of household income. They also collect enforcement information from FTC and DOJ case filings, including whether agencies required divestitures or scrutinized particular product markets. As a robustness test, the authors repeat much of the analysis using NielsenIQ’s Consumer Panel Dataset.

Methods

The authors use a two-stage event-study approach to estimate merger effects. For each product and geographic market, they first use as much as 36 months of pre-merger data to estimate the outcome that would be expected in the absence of the merger. The regression accounts for brand-specific time trends, persistent product and geographic differences, seasonal patterns, commodity input costs, and household income. They then compare actual outcomes during the 24 months following the merger with outcomes predicted from the pre-merger model. Estimates are calculated separately for merging firms and their non-merging competitors.

The central identification assumption is that, absent a merger, outcomes would have continued along their pre-merger trends after accounting for the included controls. The authors examine pre-merger patterns and find no systematic departures from estimated trends immediately before merger completion. They also find that important price changes tend to emerge shortly after mergers. Placebo analyses using randomly assigned merger dates generate price-effect distributions that are more narrowly centered around zero. The authors additionally experiment with dampening pre-merger trends and use synthetic controls for mergers with unusually large estimated effects. The synthetic-control results frequently reproduce the baseline estimates, although the findings can be sensitive to donor-pool selection.

The second part of the analysis develops a statistical model of antitrust enforcement. Agencies are assumed to receive an imperfect signal about a merger’s likely price effect and to challenge the merger when the expected increase exceeds an implicit threshold. The model uses observed merger characteristics, estimated realized price changes, and agency challenges to infer both the agencies’ effective price threshold and the uncertainty they face when predicting merger effects. The authors then simulate alternative enforcement thresholds and estimate changes in agency challenges and enforcement errors.

Findings/Size Effects

The average price effect of completed mergers is surprisingly small. Across alternative specifications, estimated average changes range from a 0.5 percent decrease to a 1.1 percent increase. In the baseline specification, the average increase is approximately 0.4 percent. Merging firms’ prices increase by about 0.2 percent, while non-merging competitors’ prices increase by approximately 0.6 percent.

These averages, however, conceal substantial differences among mergers. Twenty-five percent of mergers reduce overall prices by more than 2.1 percent, while another 25 percent increase prices by more than 3.9 percent. Among merging firms specifically, one-quarter of mergers raise prices by more than 3.7 percent and one-quarter lower prices by more than 3.3 percent. Prices charged by merging and non-merging firms are positively correlated, suggesting that competitors often move prices in the same direction following a merger.

Quantity effects also vary substantially. Aggregate quantities decline by approximately 0.3 percent in the baseline analysis, and 59 percent of mergers produce total quantity reductions. The merging firms themselves experience a considerably larger average quantity decline of 6.3 percent. For aggregate quantity effects, the bottom quartile represents a 9.7 percent decline, while the upper quartile represents a 5.8 percent increase. Quantity reductions among merging firms are associated with higher prices, smaller distribution networks, fewer product offerings, and the elimination of products nationally.

The study also finds evidence supporting the use of changes in market concentration and the merging firms’ combined market share as screening measures. Price increases are correlated with the change in the Herfindahl-Hirschman Index and the merging parties’ share of the market. These findings provide empirical support for some of the structural indicators used in federal merger guidelines.

The enforcement model estimates that agencies behave as though they generally challenge consumer packaged goods mergers when they expect price increases of roughly 4.8 percent to 6.3 percent. The precise estimates vary by data source and model specification. In the scanner-data analysis, the sales-weighted average threshold is approximately 5.1 percent to 6.3 percent, while the consumer-panel analysis produces an aggregate price threshold of approximately 4.8 percent to 5.8 percent.

The counterfactual simulations indicate that lowering the enforcement threshold to 2.5 percent would reduce the average price effect of completed mergers by approximately 1.4 percentage points. The share of allowed mergers classified as price-increasing enforcement errors would fall from roughly 50 percent under the estimated current threshold to about 35 percent. The probability of blocking price-decreasing mergers changes relatively little. The principal cost is administrative: agencies would have to challenge nearly four times as many mergers.

Conclusion

The authors conclude that there is no single typical effect of a merger in consumer packaged goods markets. Average price changes are modest, but this average combines mergers that substantially lower prices with mergers that substantially increase them. The paper’s most important empirical finding is this extensive heterogeneity: one-quarter of mergers reduce prices by more than 2.1 percent, while another quarter increase them by more than 3.9 percent. These differences remain evident across alternative specifications and robustness tests.

The analysis also suggests that current antitrust enforcement rarely blocks mergers that ultimately lower prices, but agencies permit a substantial number of mergers that later increase prices. More stringent screening could reduce the prevalence of these price-increasing mergers without producing a comparable increase in blocked price-decreasing mergers. The primary trade-off is a substantial increase in the investigative and enforcement burden placed on federal agencies. The study therefore shifts the analysis of merger policy away from the average effect of mergers and toward the more difficult question of how accurately regulators can identify harmful mergers before they occur.