Increased Ambiguity in the Merger Guidelines
Since 1982, the Agencies’ Merger Guidelines have become increasingly ambiguous as to how the Agencies are likely to treat captive capacity in their analysis of market definition and market shares.
1982/84 Horizontal Merger Guidelines
Each version of the Guidelines has provided a method for identifying market participants as part of the market definition analysis. Under the 1982/84 Guidelines (“1984 Guidelines” for simplicity), identifying market participants began with a “focus primarily on firms that currently produce and sell the relevant product.” In addition, other producers may be included if “their inclusion would more accurately reflect probable supply responses.” Specific examples of other producers in the 1984 Guidelines include production substitution, durable products, and internal consumption or captive capacity.
The 1984 Guidelines specifically stated that the DOJ considered captive production and consumption of the relevant product by vertically integrated firms as part of the overall market supply and demand. The Agencies would include a producer of a captive good in their market definition analysis if that producer would be likely to respond to a “small but significant and non-transitory” increase in price in one of two ways—by (1) diverting the relevant product from internal consumption to selling it in the merchant market, or (2) increasing production of both the internally consumed good and the downstream products in which the internally consumed good is incorporated. According to the 1984 Guidelines, either response would likely frustrate collusion by firms currently selling the relevant product or an effort by the current sellers of the relevant product to exercise market power, and as such, these producers should be included in the market.
The 1984 Guidelines also contained specific instruction as to the amount of sales or capacity that should be included in the market for captive producers. The Agencies would only include the portion of captive production that would be diverted to the merchant market in response to a price increase. In other words, even if a vertically integrated firm would respond to a price increase by diverting captive production to the open market, the market definition would not necessarily include the firm’s total captive capacity—only that portion that would likely be diverted. The 1984 Guidelines did not describe a methodology to predict the amount of likely diversion.
1992 and 1997 Horizontal Merger Guidelines
The 1992 and 1997 versions of the Guidelines began an ongoing trend of offering less discrete, specific, and clear guidance for the treatment of captive capacity (as compared to the 1984 Guidelines).
The 1992 Guidelines dropped the specific terminology of “captive production” and “internal consumption.” The Agencies also stopped treating captive producers as a distinct category of firms that do not currently produce or sell the relevant input product, but might still be included in the market for the input product with currently active producers and sellers of that product, as the 1984 Guidelines had done. Instead, the 1992 Guidelines appeared to have included captive capacity with “firms that currently produce or sell in the relevant market” or perhaps with “firms not currently producing or selling the relevant product in the relevant area.”
The guidance on how the Agencies would analyze these two categories of firms (those that either are or are not deemed to be current producers or sellers) also became more ambiguous. The Agencies would include “vertically integrated firms” in the market (for the input product) under the former category “to the extent that such inclusion accurately reflects their competitive significance in the relevant market prior to the merger.” Firms would be included in the market (for the input product) under the latter category “if their inclusion would more accurately reflect probable supply responses”—these firms would be termed “uncommitted entrants.” Supply responses from firms included as uncommitted entrants “must be likely to occur within one year and without the expenditure of significant sunk costs of entry and exit,” in response to a “small but significant and nontransitory” price increase.
The 1992 and 1997 Guidelines also removed specific guidance in the 1984 Guidelines as to how much of captive producers’ capacity would be included in the market. The 1984 Guidelines stated that DOJ would include the portion of captive production that would be diverted to the merchant market in response to a price increase. The 1992 and 1997 Guidelines offered more general guidance, stating that the Agencies would measure the market shares of all firms identified as market participants based on the total sales or capacity currently devoted to the relevant market, together with capacity that likely would be devoted to the relevant market in response to a “small but significant and nontransitory” price increase. On the other hand, the Agencies would not include a firm’s sales or capacity “to the extent that the firm’s capacity is committed or so profitably employed outside the relevant market that it would not be available to respond to an increase in price in the market.”
2010 Horizontal Merger Guidelines
The 2010 Guidelines further pared down the specificity of guidance on captive capacity. These Guidelines now simply stated that “[a]ll firms that currently earn revenues in the relevant market are considered market participants” and that “[v]ertically integrated firms are also included to the extent that their inclusion accurately reflects their competitive significance.” Under the 2010 Guidelines, the Agencies might also include “rapid entrants” in the market, including (1) firms that are not current producers in a relevant market, but that would very likely provide rapid supply responses with direct competitive impact in the event of a SSNIP without incurring significant sunk costs, (2) firms that produce the relevant product but do not sell it in the relevant geographic market, or (3) firms that clearly possess the necessary assets to supply into the relevant market rapidly, such as a supplier with efficient idle or swing capacity. The 2010 Guidelines left it unclear which category would cover producers with captive capacity for input products or services.
On the question of how much of a producer’s captive capacity should be counted in the market, the 2010 Guidelines contained even less clear and specific guidance than the 1992 and 1997 Guidelines. The Agencies would “normally calculate market shares for all firms that currently produce products in the relevant market, subject to the availability of data,” and would “also calculate market shares for other market participants if this can be done to reliably reflect their competitive significance.” The 2010 Guidelines did state that “[w]hen market shares are measured based on firms’ readily available capacities, the Agencies do not include capacity that is committed or so profitably employed outside the relevant market, or so high-cost, that it would not likely be used to respond to a SSNIP in the relevant market.” However, it is doubtful that this instruction was intended to apply specifically to captive capacity, given that it was part of a discussion about markets for homogeneous products and how capacities or reserves may serve as a better measure of firms’ market share than revenues in those markets.
2023 Merger Guidelines
The 2023 Merger Guidelines similarly state that “[a]ll firms that currently supply products . . . in a relevant market are considered participants in that market” and that vertically integrated firms should be included “to the extent that their inclusion accurately reflects their competitive significance.” The 2023 Guidelines also include similar comments about “rapid entrants” as the 2010 Guidelines but replace the more specific and well-understood SSNIP test with a much more vague and more discretionary test based on “small but significant change in competitive conditions” to analyze which firms would qualify as a rapid entrant.
The 2023 Guidelines contain even less guidance for how much captive capacity might be counted in the market, and the standard for calculating market shares became even more vague. The Guidelines now state that the Agencies will normally calculate market shares for all firms that currently supply products in a relevant market, subject to the availability of data, and will “measure each firm’s market share using metrics that are informative about the market realities of competition in the particular market and firms’ future competitive significance.”
Review of Government Litigation and Investigation Involving Captive Capacity
The trend toward less specific guidance on how to analyze captive capacity in merger enforcement is surprising, given that the inclusion or exclusion of captive capacity can so often be a dispositive issue in merger cases.
Captive Capacity Included in Relevant Market
In United States v. Aluminum Company of America (1945) and Matter of B.F. Goodrich Company (1988), the DOJ and FTC staff, respectively, argued that captive capacity should be included in the relevant product market. In Aluminum (a non-merger case), the court sided with DOJ that defendant’s production of aluminum ingot for its own use should be counted for purposes of calculating defendant’s share of the aluminum ingot market. The court reasoned that ingot is almost always an input used to produce other end-products, and so any end-products made by the defendant using its own ingot would necessarily directly affect the ingot market and reduce demand for ingot itself.
In Goodrich, FTC staff argued that two vertically integrated firms’ captive production of an intermediate good, vinyl chloride monomer (“VCM,” a gas used in the manufacturing of polyvinyl chloride, or “PVC”), should be included with non-captive capacity in the market. In deciding in favor of the staff’s position, the FTC Commissioners cited the 1984 Guidelines, reasoning that the integrated VCM producers could respond to VCM price increases initiated by nonintegrated producers by increasing VCM production, either for sale or for producing additional quantities of PVC.
By contrast, in Matter of International Telephone & Telegraph Corporation (1984), California v. Sutter Health Systems (2000), United States v. Sungard Data Systems (2001), and most recently in FTC v. Tempur-Sealy & Mattress Firm (2025), the government enforcers (FTC staff, DOJ, and California) argued for the exclusion of captive capacity from the relevant product market. In International Telephone, FTC staff contended that the relevant product market should be limited to bread produced by wholesale bakers, whereas defendants contended that the relevant market should include bread produced by both wholesale bakers like defendants and “captive bakers” who produced and supplied bread to their own grocery stores, because captive bread competed with wholesale-baker bread. In Sutter Health, California argued that services provided by Kaiser hospitals should be excluded from the market because Kaiser’s services are “captive” in that Kaiser provided care only to its own health plan members and therefore was not a “practical alternative” for members of other health plans. In Sungard, the DOJ wanted to exclude from the market firms that provided their own internal “hotsite” services (remote facilities that serve as secure backup for computer systems), because these firms were unavailable to serve third parties. In Tempur-Sealy, while conceding that a third-party mattress producer who exclusively supplied its own retail stores was a major competitor to defendant mattress producer, FTC staff nevertheless wanted to exclude the vertically integrated third-party producer from its market definition and calculation of market shares and projected market foreclosure rates.
The FTC Commissioners and the courts ruled against the government in each of these cases, finding that captive production should be included in the relevant product market. In International Telephone, the Commission again invoked the 1984 Guidelines in its examination of whether a change in the price of wholesale bread would affect the quantity of captive bread supplied and vice versa. The Commission found that it did, because in the face of a wholesale price increase, captive bakers could readily divert their production to other customers and at least some stores would be capable of beginning in-house production. The court held in Sutter Health that the captive hospital services provided by Kaiser should be part of the market because Kaiser did serve as a viable substitute for services offered at other hospitals—if faced with a price increase, patients may choose to join the Kaiser network to access Kaiser hospitals. In Sungard, the court reasoned that internal hotsites should be included in the market because the customers that currently relied on outside vendors for hotsite services could switch to an internal hotsite in response to a SSNIP (that is, reverse integrate or backward integrate). And in Tempur-Sealy, the court ruled that “all methods for reaching the consumer in the alleged market must be considered, including through direct-to-consumer or vertical integration” and that “where a competitor such as [vertically integrated producer] shows that such model can be successful, it must be considered as one available to other rivals.”
Captive Capacity Excluded from Relevant Market
In United States v. Blue Bell, Inc. (1975), the court considered whether captive production of industrial uniforms supplied to integrated rental laundries was in competition with the sales of industrial uniforms made by non-integrated firms before the merger, and whether the garments sold to integrated and non-integrated rental laundries were considered to be distinct markets within the industry. The court concluded that the captive production in this case was not in competition with independent production because the integrated rental laundries generally purchased their garments from their own manufacturer whenever available, and thus “other manufacturers are…effectively foreclosed from competing.” Additionally, documents showed that the merging companies and the industry in general recognized a separate market and set of competitors for garments sold to independent rental laundries.
In Matter of British Oxygen Company Ltd. (1975), defendants, who were non-integrated producers of industrial gases, contended that captive production of industrial gases by integrated firms should be included in the market. The ALJ and the Commission both disagreed for two reasons. First, captive production was not marketed or sold in the open market for industrial gases; and to the extent any surplus production was available, such surplus was sold to other industrial gas companies who then resold it in the open market rather than sold to customers directly in the open market in competition with other industrial gas producers. Second, the threat of reverse integration was insignificant in this case and would likely have little effect on industrial gas prices in the merchant market.
In other litigated merger cases, the courts focused on the responses to a price increase in sales of intermediate products to determine whether captive capacity should be included in the market. In Matter of B.A.T. Industries Ltd. (1984), the FTC Commissioners cited the 1984 Guidelines and concluded that captive production should be excluded when “[t]here is no evidence that [captive producer] has responded to any price increase by selling in the open market or that it has any intention to engage in such competition in the future.” In FTC v. Cardinal Health, Inc. (1998), the court excluded captive capacity on the basis that the ability to self-supply was limited to a small segment of market participants, customers would not consider self-supply in face of a price increase, and documents showed that independent producers did not perceive self-supply as a competitive threat. Similarly, in FTC v. CCC Holdings Inc. (2009), the court declined to include captive production because a survey of customers revealed that the majority of customers would not begin in-house productions even if faced with a price increase of 10% or more.
The Agencies have also taken positions regarding the inclusion of captive capacity in mergers where, but for the inclusion of the captive capacity, there would have been no horizontal overlap at issue. For example, in 1999, bookseller Barnes & Noble and book wholesaler Ingram Book Group proposed a merger that was abandoned following an FTC investigation. The FTC expressed concerns that, prior to the transaction, Barnes & Noble had plans to potentially offer its own captive wholesaling services to third parties and that the transaction could eliminate a new competitive entrant for the intermediate service of book wholesaling.
Analysis of Government and Court Treatment of Captive Capacity as Shown in Merger Guidelines and Cases
As discussed above, the 1984 Guidelines offered specific and easily cited guidance on how the Agencies would treat captive or internally consumed production capacity for merger enforcement. However, subsequent versions of the Guidelines removed that terminology and distinct category of analysis, opting to no longer provide a clear position on when and to what extent the Agencies would deem captive capacity to be part of the relevant market. It is possible that the adverse court ruling in International Telephone contributed to Agencies’ decision to pare back the 1984 guidance—in that case, FTC staff argued for excluding captive bakeries from the market for wholesale bread, but the court disagreed, citing the 1984 Guidelines in support of its decision to include captive bakeries in the market. If so, this would indicate a troubling tradeoff that the Agencies have made between regulatory transparency and litigation strategy. In other words, it appears possible that the Agencies have provided merging parties with less guidance on how the Agencies analyze capacity simply because doing so might lower the Agencies’ chances of victory in court. Given the costs to parties, such strategic opacity would arguably fall into the category of bad government.
In our view, the Agencies should reintroduce specific guidance regarding the treatment of captive capacity, reverting to and further expanding upon the 1984 Guidelines in light of more recent court decisions and enforcement experience. This would increase government transparency and accountability, improve private parties’ efforts and ability to comply with antitrust laws, and help shape this legal doctrine towards greater consistency and predictability. Adding guidance on the treatment of captive capacity would also assist private parties in evaluating merger risks where no overlap would exist between merging parties unless captive capacity was taken into account.
Our review of the litigated merger cases shows that, prior to 1990, the Agencies argued for the exclusion of captive capacity from the relevant market in all but one case—Goodrich in 1988. In that case, the FTC Commission agreed with FTC staff’s position and closely followed the 1984 Guidelines. The Commission reasoned that captive capacity should be included where the integrated VCM producers could respond to an SSNIP initiated by nonintegrated producers by increasing its own VCM production either for sale or for producing additional quantities of PVC.
However, in years following the Goodrich case and since the Agencies omitted or revised the guidance on captive capacity in the 1984 Guidelines, the Agencies have taken the position in every litigated merger case we have identified that captive capacity should be excluded from the relevant market, even where the facts would seem to indicate inclusion under the principles delineated in the 1984 Guidelines. For example, in Sutter Health, California’s position was that Kaiser’s captive hospital services should be excluded, even though the court found that Kaiser hospitals had sufficient capacity to serve as an alternative for additional patients if those patients switched to Kaiser insurance in the face of a price increase. FTC staff has taken a similar position with respect to Kaiser’s captive physician services in recent investigations, where excluding Kaiser would mean a significantly more concentrated market and higher market shares for the merging parties. Yet the 1984 Guidelines provided that a captive producer should be included in the market when it would be likely to respond to an SSNIP by increasing production of both the internally consumed good and the downstream product in which the internally consumed good is incorporated—in this case, both Kaiser’s hospital services and health plans.
The more nebulous guidance provided in subsequent versions of the Guidelines would appear to afford the Agencies greater latitude to argue for the inclusion or exclusion of captive capacity to support an enforcement action. For example, our review of enforcement actions or investigations suggests that Agencies are likely to argue for the inclusion of captive capacity where defendants possessed such capacity, which would then increase defendants’ market shares or create a competitive overlap. This was the case in Aluminum, Goodrich, and Barnes & Noble. On the other hand, the Agencies are likely to argue for the exclusion of captive capacity possessed by third parties, which would then result in a more concentrated market and higher market shares for defendants. This was the case in International Telephone, Sutter Health, Sungard, Tempur-Sealy, and others.
Although rulings by courts and the FTC Commission in litigated merger cases show some general trends in the analysis of captive capacity, there is significant room and opportunity for the development of this doctrine towards greater clarity and consistency. For example, some courts appeared to hold that including captive capacity in the relevant market required evidence of captive producers’ ability to divert captive production for sale directly in the open market, while others did not. The Second Circuit noted in a private merger enforcement case that “[c]ourts have divided on whether in-house activity should be included in market measurement.”
There are also apparent inconsistencies between later court decisions and the 1984 Guidelines. For example, as just mentioned, some courts included captive capacity in the relevant market only if there was proof that the capacity would be used for sales of the intermediate (input) product directly in the open market, whereas the 1984 Guidelines (and other courts) included captive capacity in the market if the capacity would be used in this manner or for increased internal consumption to produce and sell more of the downstream goods that incorporated the intermediate product.
The 1984 Guidelines also do not clearly address reverse or backward integration, i.e., where customers who currently purchase the intermediate product from third-party suppliers have the ability to begin self-supply of that product in response to a price increase for the product. Courts that addressed this issue in litigated merger cases have reached differing conclusions based on case-specific evidence of whether customers would engage in backward integration and whether backward integration would result in significant competitive effects.
Agency guidance would be helpful on whether (and to what extent) to include in the relevant market the potential capacity for production of intermediate products that would result from backward integration in response to a price increase for that product. This would appear to be a distinct question when the ability and incentives for a current captive producer to divert existing production capacity could be very different than for non-producer customers facing a decision to begin in-house production. For example, the two groups likely face different tradeoffs and considerations in terms of recouping prior investments and managing risk to existing operations and sales, as well as different costs and timeline for entry or expansion. Notably, the court in British Oxygen appeared to address existing captive capacity and the possibility of backward integration separately, finding that (1) existing captive capacity should not be included in the market when vertically integrated producers of the relevant input did not compete with other producers of that input, and (2) the threat that additional customers will backward integrate in the future was insignificant in terms of likelihood, magnitude, and potential price impact. The court distinguished this case from Aluminum, stating that “[t]he vertical integration [in Aluminum] played a qualitatively different role in [the] aluminum industry compared with the backward integration of industrial gases users.”
Agency guidance may also be helpful on how to address other case-specific factors courts found important that were not specifically discussed in the 1984 Guidelines. These include (1) whether industry participants view vertically integrated producers that hold captive capacity as competitors of merchant players, (2) whether any excess production of intermediate products by captive producers is then sold to competitors for resale or to customers directly, and (3) whether vertically integrated firms used only internally produced inputs or would also purchase the input from independent producers.
Conclusion
Whether captive production capacity should be included or excluded in a relevant antitrust market is an important and often outcome-determinative question in analyzing whether a merger transaction would violate the antitrust laws. However, the Agencies have provided increasingly ambiguous guidance on how they would analyze captive capacity in their merger guidelines since 1984, and their positions in enforcement actions and investigations are difficult to reconcile with one another.
The Agencies should restore the relatively clear and specific guidance that the 1984 Guidelines provided on the treatment of captive capacity, and update that guidance to address important issues that have arisen since that time in court decisions and Agency enforcement experience. This would help the government, the courts, and private parties to better enforce and comply with the antitrust laws.
