How Competing in Multiple Markets Can Affect Merger Assessment

Tacit Coordination and the Role of Multimarket Contact^[i]

Companies in many industries find themselves competing against the same rivals across different markets. Car brands might compete in multiple countries, airlines on different routes and food companies across both snacks and beverages. Economists refer to such overlaps as multimarket contact (MMC), and research shows that MMC typically facilitates tacit coordination between firms.[ii]

Tacit coordination refers to a situation in which firms avoid intense competition by behaving in a coordinated way without communicating directly. Such a softening of competition raises their profit margins. Firms should prefer it, but stable coordination is difficult to achieve in practice. This is because firms must solve both a coordination problem and an incentive problem to engage in effective and stable tacit coordination.

The coordination problem arises because too many possible outcomes exist on which the participating firms could agree. If firms cannot talk to each other to coordinate their behaviour (on prices, output or market division), they face a guessing game in which each firm must anticipate how its rivals expect it to behave, and the plethora of possible strategies makes achieving coordination extremely difficult from the outset. This becomes even more acute in the presence of uncertainty that persists over time, for example stemming from unstable demand conditions or volatile input costs.

The incentive problem (for which it is less important whether the firms can talk to each other) arises because a single firm will always be tempted to deviate from the coordinated outcome if everyone else is sticking to it. By unilaterally deviating and offering customers more attractive terms, a firm can significantly increase its market share and profits. This temptation makes coordination inherently unstable. Stable coordination is possible only if firms believe that deviation will trigger harsh competition from others, thus eroding future profits and making deviation unprofitable overall.

MMC can make it easier for firms to solve the incentive problem by allowing them to “transfer” coordination incentives across markets. Strong incentives to sustain coordination in one market can help increase a firm’s willingness to coordinate in another market where its stand-alone incentives would otherwise be insufficient. The key mechanism is that deviation in one market can be punished by reverting to intense competition in other markets as well, thus forgoing the gain from coordination in those markets where it is highest. In this way, a firm’s incentive to coordinate in one market can help discipline its behaviour across the board. And if this is not enough to incentivize coordination in the “weaker” markets, firms can always treat each market separately and coordinate only where it is viable on a stand-alone basis. In this way, MMC can facilitate coordination in a given market, but never make it more difficult.

How does this “transfer” of incentives across markets work? To illustrate, consider two routes—London–New York (NY) and London–Dubai—which constitute two distinct markets. Airlines X and Y compete on these routes, while a small local airline is active on the London–NY route only.

It is well understood that tacit coordination is easier the smaller the number of firms. With fewer firms, coordinating the firms’ behaviour becomes easier as there are fewer competitors to align with and monitor. In addition, fewer participants share the gains from coordination, increasing each firm’s share of the gains, while the immediate profits following deviation remain relatively unchanged.

We therefore assume that coordination is impossible on London–NY but feasible on London–Dubai. The key insight: MMC can actually enable tacit coordination on London–NY by leveraging the strong incentives X and Y have to sustain coordination on London–Dubai. If coordination takes place in both markets simultaneously and any deviation by X or Y on London–NY is punished by also competing aggressively on London–Dubai, then deviating on London–NY would mean sacrificing the profitable coordination on London–Dubai. This increase in the cost of deviating on London–NY can help to keep X and Y in line in that market.

Stable coordination on London–NY also would have to include the small local airline active on this route. To achieve this, X and Y could offer the local airline a larger share of the London–NY market—say one-half instead of one-third—to incentivize it not to compete aggressively. With the small airline effectively on board, the remaining question is whether X and Y would still find it profitable to maintain coordination on London–NY, despite their reduced market shares. If the gains from coordinating relative to deviating on London–Dubai are large enough, X and Y still may prefer to maintain coordination in both markets. In this way, MMC enabled coordination on London–NY by transferring some incentives from London–Dubai.

How should competition authorities factor MMC’s potential to facilitate coordination into their assessment of horizontal mergers, where two direct competitors combine forces? Should MMC serve as a red flag, signaling an increased risk of anticompetitive effects?

The Risks of Mergers for Competition

When two direct competitors merge, competition authorities assess the competitive risks of the transaction by comparing what would happen to competition in the affected market(s) with and without the merger. One competitive risk is “coordinated effects”, which involves the risk that the merger will make tacit coordination among some firms in the market(s) where the merger takes place more effective or easier to achieve or maintain.

The key question in assessing coordinated effects is not how likely it is that firms will coordinate after the merger, but whether the merger makes coordination more likely or effective than in its absence. In other words, the authorities try to establish whether the new market conditions that arise after the merger are more favourable to coordination. What matters is the change brought about by the merger.

Does MMC Make Coordinated Effects More Likely?

Consider again our example. This time, suppose that X wants to acquire the small local airline that operates on London–NY. In our setting, the merger affects one market but not the other, alongside MMC linking the two. Importantly, the merger does not create additional MMC.^[iii]

A merger typically increases the likelihood of coordination. One reason for this is that there is one less firm in the market. As explained before, this mitigates the coordination and incentive problems.

In a recent paper, we studied the following question: assuming that the merger increases the likelihood of coordination between X and Y on London–NY, does the presence of MMC intensify or mitigate the merger’s effect on the likelihood of coordination?

If MMC makes it easier for firms to coordinate, it seems reasonable to expect that coordinated effects are larger in a merger involving MMC (all other things equal). However, our research suggests that this isn’t necessarily true. In fact, we find that the risk of tacit coordination often increases less if mergers involve MMC. That is, mergers are often less anticompetitive in the presence of MMC than in its absence.

In our example, in the absence of MMC, the merger drastically changes the incentives to coordinate on the London–NY route: coordination is impossible with three firms but becomes feasible with two. The merger enables coordination on this route. With MMC, however, coordination is already feasible even without the merger, so the merger’s effect is more limited. The merger itself adds little extra risk.

In short, because MMC can be an important reason for firms to coordinate in the absence of the merger, the increase in the likelihood of coordinated effects purely due to the merger is greater without MMC. Conversely, if MMC doesn’t affect the incentives to collude in the market affected by the merger (because the other markets, where there is no merger, do not offer enough stand-alone incentives for coordination), the coordinated effects of the merger are the same with or without MMC.

Implications for Business and Regulators

Our findings suggest that, even if MMC facilitates coordination, competition authorities should be careful not to automatically assume that it makes mergers more dangerous—quite the contrary. For firms operating in multiple markets and considering mergers, this is good news.

As already mentioned, our analysis assumes that the merger does not change the configuration of the MMC. It is therefore important to clarify that our results are most directly applicable to mergers that do not create new instances of MMC. In practice, however, mergers may change the pattern of multimarket overlaps, and these new overlaps could open additional channels for “transferring” coordination incentives across markets. As a result, competition authorities should begin their assessment by checking whether the merger gives rise to new MMC. If it does not, our analysis provides a relevant framework. But if new overlaps do emerge, a more detailed analysis is warranted, accounting for how the new MMC might influence coordination incentives across markets.

[i] The authors would like to thank Dennis Beling (Managing Director, BRG, Hong Kong), Laurent Eymard (Managing Director, BRG, Brussels and Paris) and Hugo Canau (Senior Associate, BRG, Brussels) for helpful comments and discussions.

[ii] For a theoretical analysis, see the seminal paper by Bernheim and Whinston, Bernheim, D., and M. Whinston, “Multimarket Contact and Collusive Behavior,” The RAND Journal of Economics 21(1) (1990), 1–26. For empirical evidence, see Ciliberto, F., and J.W. Williams, “Does multimarket contact facilitate tacit collusion? Inference on conduct parameters in the airline industry,” The RAND Journal of Economics 45(4) (2014), 764–791; Evans, W.N., and I.N. Kessides, “Living by the ”Golden Rule”: Multimarket Contact in the U.S. Airline Industry,” The Quarterly Journal of Economics 109(2) (1994), 341–366; Jans, I., and D.I. Rosenbaum, “Multimarket contact and pricing: Evidence from the U.S. cement industry,” International Journal of Industrial Organization 15(3) (1997), 391–412; Lin, H., and I.M. McCarthy, “Multimarket Contact in Health Insurance: Evidence from Medicare Advantage,” The Journal of Industrial Economics 71(1) (2023), 212–255; and Parker, P.M., and L. Roller, “Collusive Conduct in Duopolies: Multimarket Contact and Cross-Ownership in the Mobile Telephone Industry,” The RAND Journal of Economics 28(2) (1997), 304–322.

^[iii] It may well be that the merger induces more MMC, if for example the firm acquired by X overlaps with firm Y in a third market, say Dubai–Sydney, where X had not been present. In this case, more MMC may lead to a larger impact of the merger on X and Y’s incentives to coordinate on London–NY (and possibly other markets). This effect is identified as the main and only effect of MMC on a merger’s coordinated effects in the literature. See, e.g. Ivaldi, M., B. Jullien, P. Rey, P. Seabright and J. Tirole, The Economics of Tacit Collusion, Final Report for DG Competition, European Commission, Toulouse: IDEI (2003).