Some of these opportunities involve the nuts-and-bolts of evaluating competitive effects during the pre-complaint investigation phase, while others involve issues and arguments that can be raised in the litigation phase. For this reason, this article is organized around various competitive issues, not the phases of the process. This is because Agency analysis and decisions are carried out in anticipation of possible litigation.
Foreclosure Analysis
The VMGs’ foreclosure analysis focuses on input foreclosure, where the concern is that the upstream merging firm might raise price to, or withhold input supply from, one or more rivals of the downstream merging firm. That exclusionary conduct can harm consumers and competition by raising downstream market prices. The VMGs stress that a complete analysis would include the impact of elimination of double marginalization (EDM) on prices.
Several potential input foreclosure scenarios are flagged in the VMGs. One scenario is that the upstream merging firm totally withholds rivals’ access to its input. Total foreclosure reduces the upstream merging firm’s profits. If the targeted rivals are unable to replace the full input supply at the same cost, their costs will rise, which can lead some customers to divert to the downstream merging firm, increasing its profits. Total foreclosure increases the merged firm’s overall profits if the increase in the downstream profits exceeds the decrease in the upstream profits.
A second scenario is for the merged firm to raise the input price by a significant amount rather than totally foreclose access to the input. The incentive to raise rivals’ costs in this way is similar to total foreclosure but also depends on several other factors.
A third scenario is for the bargaining leverage of the upstream merging firm to increase when negotiating input prices. In this scenario, there is no upstream profit sacrifice. Even if the profits of the merged firm would fall in the absence of a negotiated agreement, the profits of the downstream rival could fall by more. If so, the threat of foreclosure would be predicted to lead to a bargaining outcome in which the rival is willing to pay more for the input.
The merging parties should generally place the total foreclosure scenario front and center in their advocacy. This is because it is the least profitable type of foreclosure and because it is the easiest scenario to analyze. In contrast, raising the price of the input under the second and third scenarios is generally more profitable. For example, threatening foreclosure may allow the merged firm to negotiate a higher input price, even if total foreclosure is unprofitable. The merging parties can stress the greater complexity of estimating effects under these other scenarios, or their inapplicability. The parties also can emphasize how courts have been skeptical of Nash bargaining theories generally, and the bargaining theory in the DOJ’s challenge to the AT&T/Time Warner merger specifically was criticized by Judge Leon as a “Rube Goldberg contraption.”
Downplaying Analysis of Accommodating Price Increases
The VMGs have only a very limited discussion of accommodating price increases by competing input suppliers in response to foreclosure (i.e., the possibility that other input suppliers will also raise price in response to a foreclosure effort by the merging firms). If other suppliers engaged in accommodating price increases, that would make foreclosure more profitable. Conversely, the lack of accommodating price increases reduces the profitability of foreclosure and the likelihood and magnitude of consumer harm. The VMGs’ downplaying of the possibility of accommodating price increases thus provides opportunities for the merging firms.
Because the VMGs fail to spell out the analysis, it is more likely that accommodating price increases will be assumed away by Agency staff. If Agency staff alleges such responses, the parties can remind the staff of the litigation risk because the Agency will not be able to rely on the VMGs for a clear explanation or support of its relevance. Furthermore, the parties can point to the fact that the FTC did not find accommodating price responses to be a concern in the Staples/Essendant office supplies merger, despite the fact that Essendant only faced one main competitor. However, Commissioner Slaughter flagged the issue of accommodating price increases in her dissent, so it may be given more emphasis if the Commission shifts to a Democratic majority in the future.
Furthermore, the VMGs’ discussion of the accommodating price effect is placed in the context of simulation modeling. This placement may signal to a court that the Agencies believe that there must be quantitative evidence of the effect, even if qualitative evidence of input market concentration, product differentiation, and capacity limits might be highly probative.
Advocates thus might press the Agency to frame its arguments solely with regard to the behavior of the upstream merging partner, not also its competitors, and to dismiss any concerns about accommodating price increases making foreclosure more profitable.
Emphasizing Quantitative Evidence
The increased role of econometric and other quantitative evidence in merger analysis also provides opportunities for the merging parties. Section 7 does not necessarily require that the Agency establish its prima facie case of competitive harm with quantitative evidence. The VMGs remark that the Agency may rely on qualitative evidence, but the VMGs appear to place far more stress on quantitative evidence, “where sufficient relevant data are available.”
Reliance on quantitative studies serves to raise the Agency’s evidentiary burden because such studies are subject to various shortcomings and criticism. Thus, the parties often should provide data to the Agency to ensure its availability. Staff focus on quantitative evidence also will be increased if the parties provide their own quantitative evidence showing lack of harm. Questioning the robustness of the Agency’s quantitative analysis will be a useful way to defend a merger.
A mandate to generally rely on quantitative evidence can make the Agencies more reluctant to litigate because the VMGs’ embrace of quantitative evidence can significantly benefit the merging parties in litigation. In court, the Agencies will have the burden of establishing significant harm. Given the language of the VMGs, the parties will be well-positioned to argue in court that robust quantitative evidence is required.
The VMGs state that the “likely merger-induced increase or decrease in downstream prices would be determined by considering the impact of both these effects, as well as any other competitive effects.” For example, if the parties make a quantitative EDM claim, the Agencies will find it necessary to balance the quantitative comparison of harms to estimate the expected overall effect. If the magnitude of the harm is not easily quantified (e.g., for coordinated effects theories), a court may discount the Agencies’ qualitative evidence.
Even highly rigorous empirical studies by the Agencies might be criticized by the parties. Because it can be difficult for courts to evaluate econometrics and simulations, and the relative importance of criticisms made of them (particularly in the context of a short preliminary injunction hearing), the parties can gain an advantage by demonstrating that the empirical analysis appears complicated and hard to understand. The complexity of the quantitative evidence in the AT&T/Time Warner merger trial illustrates this issue.
The parties also might try to tweak the Agency’s empirical estimates to reverse, or at least substantially alter, the results. In a complex model, this can be easier to do than the court might anticipate, and the VMGs invite such an approach. The VMGs state that “[t]he Agencies do not treat merger simulation evidence as conclusive in itself, and they place more weight on whether merger simulations using reasonable models consistently predict substantial price increases than on the precise prediction of any single simulation.” This emphasis on “robustness” can be another Agency hurdle in court because robustness often is a subjective matter.
Requiring quantitative evidence of competitive harms also allows the merging parties to raise the Agency’s evidentiary hurdle by exploiting an intrinsic anti-enforcement bias in conventional statistical analysis as used in the academic literature. Specifically, even if an econometric model predicts a price increase, merging parties may argue the estimated price increase is not “statistically significantly different from zero” according to the “standard” statistical test of a 90–95% confidence interval. A statistical estimate of a price increase is normally interpreted to mean that the “true” value of the price increase falls within a “confidence interval” that encompasses both negative and zero price increases as well as positive price increases. The standard 95% or 90% confidence interval used in most studies implies that the likelihood of the true value of the price increase lying outside that interval is only 5% or 10%. As a result, in this conventional approach, the “null hypothesis” that the “true” price increase is zero is said to be rejected by the econometric evidence only if zero lies outside the confidence interval.
This approach raises the Agencies’ evidentiary bar by focusing almost exclusively on avoiding false positives by requiring a high degree of estimation precision to reject the null hypothesis of “no price effect.” That is, the approach sets a standard that even if the actual price increase may well be positive or larger than the “point estimate,” and even though the point estimate may be the expected value of the price increase, those facts are not considered sufficiently certain to justify blocking a merger. Thus, the current methodology provides an opportunity for the parties to get their deal cleared, even if the expected (estimated) price increase is positive.
While economists have begun to pay attention to how the difference in context between academic literature and adjudication interacts with appropriate statistical tests, courts may be slow to adjust, particularly if the Agencies do not take the lead. The VMGs do not suggest an inclination for the Agencies to do so nor do they provide a justification for the Agencies to argue in court for a different standard for their quantitative evidence.
Coordinated Effects Analysis
Coordinated effects of various types can occur in vertical mergers. For example, the VMGs flag collusive information exchanges, which have been a common driver of consent decrees. They also provide an example of how the merger may foreclose a non-merging firm from acting as a maverick in the downstream market. However, the analysis of mavericks and other coordinated effects scenarios is limited.
For example, the VMGs omit mention of other relevant maverick scenarios in which a merging or non-merging firm might have been a disruptive buyer or seller in either market and might be hindered or eliminated by the merger. The VMGs also fail to discuss the potential for coordinated foreclosure by multiple vertically integrated firms, which was one of the allegations in the AT&T/Time Warner complaint. Coordinated foreclosure also was central to the economic analysis in the private JeldWen litigation.
The VMGs’ failure to discuss these other types of coordinated effects downplays their importance. These omissions also make it more likely that the investigating staff will overlook them, and more difficult for the staff to educate courts on the breadth of potential harms from a particular vertical merger. The parties might suggest to the court that these omissions are a telling signal that the Agencies viewed them as so implausible that they did not bother to include them in their VMGs.
Deletion of the Quasi-Safe Harbor
The explicit quasi-safe harbor was deleted from the final version of the VMGs in the face of criticism. However, there are indications that the Agencies remain amenable to safe harbor-type arguments. The VMGs now state that when conditions indicate an ability and incentive to engage in input foreclosure, those conditions “potentially raise significant competitive concerns and often warrant scrutiny.” This qualification is noteworthy because if the merger leads to the incentive and ability to foreclose, then one would expect that it always would warrant scrutiny. Thus, this language suggests that the policy underlying the safe harbor was not changed. At the very least, this language suggests a higher Agency evidentiary burden than a simple showing of incentive and ability to foreclose. Advocates might raise this point to staff and agency leadership. It also may serve to reassure clients that the Agency staff will have the quasi-safe harbor in mind even if it was formally deleted from the final version of the new VMGs.
Evasion of Regulation and Long-Term Contracts
A vertical merger might permit a regulated merging firm subject to cost-based price regulations to evade regulatory constraints. A similar evasion analysis also would apply when long-term contracts for inputs set price escalators based on some proxy for input or output market prices. In this situation, the downstream rivals might not be able to rely on their long-term contracts to prevent the merger from raising their costs, if the merging firm raises its prices in a way that affects the market price escalator. This effect would be exacerbated if other upstream firms make accommodating price increases. However, the VMGs address neither of these theories.
The omission of these theories of harm from the VMGs can lead the Agency to overlook them and make it more difficult for them to rely on the theories in court. That is, the VMGs provide no explanation for courts to follow. Therefore, if the Agencies were to litigate either theory, its absence from the VMGs would make it more difficult to win. The merging parties can use the omissions to argue that the complaint involves only a regulatory matter or contract dispute, not an antitrust issue. The merging parties also might characterize the allegation as speculative in light of the potential efficiency benefits of both the long-term contracts and the escalators.
Analysis of Elimination of Double Marginalization
Analysis of the elimination of double marginalization (EDM) provides very important opportunities for the merging firms. EDM can lead to downward pricing pressure by internalizing the pre-merger price/cost margin earned by the upstream merging firm. The Agencies may view EDM benefits as dispositive, even if the merger otherwise raises significant concerns. The VMGs’ tone suggests a significant procompetitive presumption for vertical mergers that is driven by potential EDM benefits. The VMGs clearly signal the importance they place on EDM benefits by the number of places where they are explained. If the parties have any plausible EDM benefits, they should emphasize those to the Agencies.
The nuts-and-bolts analysis of EDM also provides opportunities. Certain arguably relevant mitigating factors that might reduce EDM benefits are omitted from the VMGs. The VMGs also can be argued to place a low evidentiary burden on the merging parties to substantiate EDM and concede a high rebuttal burden.
Low EDM Evidentiary Burden on the Merging Parties and High Burden on the Agencies
While the VMGs nominally state that the parties must substantiate EDM, practitioners can make the point that the language suggests a very low burden of production on the parties sufficient to shift the burden back to the Agencies themselves to rebut EDM claims.
The Agency rebuttal burden could be satisfied with evidence supporting the following conclusions: either (1) EDM will not occur at all; or (2) quantitative evidence indicates that merger-specific EDM will not outweigh competitive harms; or (3) EDM will not be merger-specific. Each of those possibilities is discussed below.
Substantiation of EDM
The burden on the parties for demonstrating EDM benefits will occur is very low. The VMGs ask parties solely for substantiation that the EDM will benefit the merging firms, not consumers. The VMGs also state that a showing of a pre-merger price above marginal cost is “often the best evidence” of the price that would occur absent the merger, a point that goes to the magnitude of the EDM as well as the merger-specificity. The parties can argue that the “best evidence” language amounts to placing a very low burden of production on the parties sufficient to shift the burden back to the Agency to rebut the EDM claim. It essentially presumes that the upstream merging firm can expand its sales to the downstream merging firm, the downstream merging firm has the capacity to expand, and there is no coordination that would deter pass-on.
The VMGs omit the role of benefits to the upstream merging firm from a higher price set by the downstream merging firm that causes sales to divert to rivals that use the input of the upstream merging firm. This diversion and associated upstream profits reduce the downward pricing pressure from EDM. By contrast, this was a key factor in the DOJ analysis of EDM in the AT&T/Time Warner case. The draft VMGs explained in detail how incentives to pass-on EDM with downstream price decreases are mitigated or even reversed by offsetting incentives to increase prices. In fact, this source of upward pricing pressure is inherent in all vertical mergers, even if there is no foreclosure. Thus, the deletion of discussion of this issue in the final version seems to represent a deliberate and significant policy decision.
This omission of opportunity costs can be an advantage to the merging parties, which may choose to leave out this effect in their initial advocacy. It also can provide an advantage in court because this effect may not be obvious to judges. As a result of the omission, the Agency would face the burden of explaining this possibly complicated concept to the court, increasing the probability it will be treated with greater skepticism or ignored.
Merger-Specificity of EDM
In contrast to the HMGs, the merger-specificity of EDM appears to be presumed if pre-merger prices exceed marginal cost by the statement that a pre-merger price above marginal cost is “often the best evidence.” The parties also have an advantage because the VMGs prevent the Agencies from rejecting a claim of merger specificity “solely because it could theoretically be achieved but for the merger, if such practices are not reflected in documentary evidence.” The VMGs do not specifically demand that the parties explain the failure to achieve EDM absent the merger with credible evidence of transactions costs. And if asked, the parties can allude to technical economic terminology from bargaining analysis such as bargaining frictions.
The VMGs correspondingly appear to burden the Agencies with proving a high likelihood of a non-merger bargaining solution. Because of the caveat about documentary evidence just discussed, the parties can argue that the VMGs put a burden on the Agency to produce documentary evidence showing that the EDM likely could be achieved absent the merger. That may be difficult for the Agency to do. The best documentary evidence for rejecting merger-specificity would involve documents that reveal that the parties had recently completed negotiations of an EDM contract but then decided to merge instead. If there is no completed contract, the VMGs say that the Agencies would examine “contracting efforts considered by the merging firms.” But efforts considered are not nearly the same as efforts likely successful. In fact, if the negotiations had not been completed, executives at one (or both) of the merging parties can point to a fearful feeling that negotiations would fail, so they chose to negotiate a merger instead. It would not be easy for the Agencies to rebut this “fearful feeling,” if the executive is a good witness.
The VMGs also suggest the existence of “contracts between similarly situated firms in the same industry” as relevant documentary evidence. However, the merging parties can make the point that this evidence also is only theoretical. This evidence does not explain why the EDM contract would be negotiated between the merging parties, nor why EDM might not occur after the merger. Indeed, the parties can argue that the fact that these other EDM contracts were successfully negotiated by these other firms––but not between the merging parties for this input –––is not only the “best evidence,” but also conclusive proof that the situations or bargaining dynamics must have been different. The same explanation would apply to evidence that one of the merging firm has EDM contracts with others.
It similarly would be very difficult for the Agencies to provide documentary evidence if the true reason for failure to achieve EDM in the pre-merger world is the fear that EDM by contract (or by merger) might disrupt a coordinated outcome. Discussion of likely anticompetitive coordination would not be likely to show up in the documents. Even if the agencies can establish a high likelihood of pre-merger coordination, the VMGs’ emphasis on quantitative evidence may make the Agencies’ task more difficult. That is, there may be good qualitative evidence for why the merger would not eliminate coordination under an “appreciable danger” or “reasonable probability” evidentiary standard. But proving this effect in a quantitative simulation model is a much higher evidentiary bar.
The Role of Hypothetical But-For World Anticompetitive Contracts in EDM Analysis
The VMGs explain that “[t]he Agencies will generally take the same approach to evaluate the likely contractual arrangements absent the transaction as the one they use when evaluating raising rivals’ costs or foreclosure.” One apparent goal of this sentence is to reject the possibility that the merged firm might act as a unitary entity in making foreclosure or anticompetitive coordination decisions, but instead instruct executives to treat sister divisions as if they were separate companies in buying or selling inputs.
The sentence apparently also alludes to the theoretical possibility that if the parties would have been able to eliminate EDM absent the merger, then the parties similarly also could negotiate a vertical contract (absent the merger) with the same anticompetitive effects as would occur from the merger. Thus, Agency analysis attacking merger-specificity (i.e., EDM absent the merger) also must explain why the parties could not negotiate an anticompetitive agreement absent the merger.
The parties can draw either of two implications from this latter reasoning, both of which benefit the merging parties. On the one hand, the parties can argue that the lack of such an anticompetitive contract implies that such a contract would not have been profitable absent the merger, in which case the vertical merger also would not lead to anticompetitive effects. On the other hand, the parties can argue that enjoining the merger would not deter anticompetitive effects because the parties then could strike an agreement with an equivalent anticompetitive outcome. Either way, the merging parties can argue that there would be no consumer benefit from stopping the merger. That is, parties can argue that if EDM benefits are not merger-specific, then the claimed anticompetitive effects also are not merger-specific.
The Agency might explain to the court that such an anticompetitive vertical agreement would violate Section 1 of the Sherman Act. However, the absence of this point in the VMGs creates something of a credibility gap if the Agency attempts to walk away from its VMGs.
Coordinated Effects and EDM Analysis
EDM analysis is altered when the merger concern is coordinated effects. The VMGs flag the possibility that EDM may increase the incentives to cheat on a cartel, which may reduce the likelihood of successful coordination. This is a standard element of the economic theory of coordination, and the merging parties have the opportunity to make this point. The merging parties also benefit from the fact that the VMGs fail to flag the flip-side – that EDM analogously may increase the ability and incentives of the merging firm to punish cheating by non-merging firms, which may increase the likelihood of coordination. This latter effect can occur because the profits of the punishing firm (if coordination breaks down) increase if its post-merger costs are lowered, all else equal. If the Agency raises this argument, the merging parties can explain that non-merging firms would retain the incentive to punish or that the merging firm lacked punishment incentives in the pre-merger world. The parties also could point out the lack of support for this theory of harm in the VMGs.
The VMGs also fail to flag the circumstances in which incentives to coordinate would mitigate or eliminate incentives to pass on EDM. For example, suppose that the upstream firms were coordinating in the pre-merger world. If so, the upstream merging firm would resist a pre-merger EDM contract out of a fear that the resulting downstream price decreases would disrupt the upstream market pricing coordination. Fear of disrupting coordination may be reduced somewhat by the merger, but not by enough to incentivize the upstream merging firm to become a maverick or for the downstream division of the merging firm to pass on the EDM as a lower downstream price. Or the merged firm might limit the pass-on to avoid setting off a price war. As a result, consumers would gain little or no EDM benefits. At the same time, the merger could increase the foreclosure incentives of the upstream merging firm and other vertically integrated firms. If the Agency raises these arguments, the merging parties can question whether the Agency has credible evidence that these effects would dominate the effect of EDM to increase incentives to cheat, or whether the Agency concern is merely theoretical. As above, the parties also can point out the lack of support for these effects in the VMGs.
The parties may be subject to more extensive discovery on these topics than before due to the emphases discussed above on both documentary and quantitative evidence. The Appendix to this article contains some hypothetical document requests and questions regarding EDM that the parties might expect. These questions also might be useful for pre-merger counseling purposes. (An earlier article by Salop and Culley sets out questions relevant to evaluation of potential harms and other benefits.)
Conclusion
The new VMGs are unlikely to lead to an increased prospect of vertical merger challenges. However, the VMGs may alter the merging parties’ interaction with the Agencies because they introduce new factors that can affect both the pre-complaint and litigation phases. As a result, the Agencies may ask more questions and carry out more analysis before granting early termination or may include more questions and information requests in Second Requests.
