Evaluating Innovation Theories of Harm in Merger Review: Economic Frameworks and Difficulties

A number of latest high-profile merger evaluation instances in the life sciences house—together with Celgene/Bristol-Myers Squibb, Illumina/Pacific Biosciences (PacBio), Roche/Spark—have targeted on so-called innovation theories of hurt—a priority that the merger might lower the extent of innovation actions by the merging events or their opponents and hurt shopper welfare. Whereas evaluating such theories of hurt is understandably of excessive curiosity, antitrust authorities ought to recognise that innovation is an space that doesn’t lend itself to generalizations of a single financial principle or mannequin. Whereas generalizing the manufacturing perform for widgets as a mathematical perform works effectively sufficient, improvements are usually not like widgets, and utilizing the identical method can result in three elementary issues. First, generalizations are essentially restricted by how effectively current fashions, which match some conditions properly, describe the broader universe of modern exercise. Second, proof in these instances is stricken by definitional and measurement issues which can be worse than normal for merger evaluation for the reason that underlying ideas to be measured or examined are usually not effectively specified. Third, treatment insurance policies are troublesome to establish for these instances with out essentially assuming some particular means in which innovation is produced. This text summarizes every of these classes of challenges; discusses some approaches and concerns employed in previous instances; and finally explains why a very good mannequin of innovation, rooted in the information of the {industry}, is the very best method.

I. Looking for a theoretical framework

Growing a framework for analysing innovation, as an alternative of value adjustments for companies or manufactured items, is difficult as a result of economists have a comparatively extra restricted understanding of exactly how improvements are created. As a result of manufacturing of widgets normally follows a deterministic course of with well-understood inputs and outputs, economists can extra confidently predict the potential impression of a merger on manufacturing portions after analysing data akin to substitutability of merchandise and price synergies of manufacturing. This in flip permits economists to foretell merger-specific value adjustments. However, the truth that economists can not readily predict innovation output from innovation inputs makes an innovating agency’s revenue maximization downside advanced, opaque, and troublesome to mannequin. Thus, even when an antitrust authority can establish overlaps in analysis targets and potential synergies between R&D groups, economists can solely say as a normal matter that combining the groups will get rid of one incentive—rivalry—for investing in innovation. And not using a normal mannequin of how combining analysis efforts would possibly have an effect on their productiveness, economists can not rule out different prospects, together with, for instance, that elevated productiveness from the mixture will dominate and that the inducement to take a position in innovation will improve. Thus, it’s pure that antitrust economists would seek for a theoretical framework that may enable them to extra confidently predict such adjustments for mergers of innovating companies.

A number of economists—famously Arrow[i] and Schumpeter,[ii] amongst others—have tried to mannequin innovation and its interplay with competitors; nevertheless, no consensus has been reached. Carl Shapiro’s handbook chapter on the subject summarizes themes frequent to those fashions in an effort to offer a unified framework for analysing whether or not innovation will lower after a merger.[iii] Sadly, this effort doesn’t handle the underlying downside—that the variability of processes by which improvements are created are usually not effectively generalized in anybody or two mathematic fashions of manufacturing. Figuring out the variations between the fashions we do have can create a false sense of completeness—the categories of variations between these fashions don’t seize all of the doable sorts of variations between innovation processes in completely different industries.

For instance, one theme the article emphasizes is appropriability—the extent to which an innovator is ready to seize the social profit of the innovation as revenue. One would possibly contemplate two financial fashions—one in which there’s a race to a single, remaining discovery, which will probably be patented and completely protected ceaselessly; and one other in which there’s a collection of doable discoveries, every of which allows opponents to ‘leapfrog’ rapidly whereas entitling the innovator to little or no safety—as bookending a spectrum on which appropriability is measured. Nevertheless, even contemplating simply this one explicit sort of innovation incentive—how a rival’s innovation success impacts the rewards of subsequent R&D—certainly the gathering of industries in the economic system covers a wider array of potential constructions than simply these mendacity completely in between these bookend fashions.

Sadly, specializing in this spectrum (or others talked about in the chapter) in the absence of an industry-appropriate, sensible mannequin can lead an antitrust authority—seeking to economics for normal ideas—astray. For instance, the authority would possibly rigorously measure proxies for appropriability, akin to patent power, which—in the absence of an applicable mannequin—will shed little gentle on how the actual merger at hand will change innovation outcomes. Do robust patents form the sort of R&D initiatives companies pursue? Does that affect how a lot rivalry they expertise or keep away from pre-merger? Is there a single race, or would possibly one agency be racing for short-term wins whereas one other performs an R&D lengthy recreation? Consequently, economists in search of to assist antitrust authorities ought to keep away from the generalization and as an alternative rigorously evaluation information to learn the way innovation works in the {industry} in query, try to mannequin that innovation, and then interpret empirical information inside that mannequin to foretell the results of a merger. It will likely be inadequate to depend on a preordained assortment of parameters for which one would possibly be capable of discover empirical proxies.

II. Proof and Measurement

 Even given a well-formulated mannequin for a selected {industry} beneath investigation, empirical evaluation of a merger’s impact on innovation incentives would nonetheless be sophisticated by measurement points. One may think that nearly any empirical evaluation would conceptually require some quantification of the innovation produced by the merging events and different opponents. Nevertheless, such quantification runs into no less than two principal difficulties.

First, as already mentioned, innovation is just not a homogeneous phenomenon, however moderately a time period used to explain all kinds of technological advances, all qualitatively completely different from one another.[iv] It can be crucial that the chosen mannequin be tailor-made to the precise sort of innovation that’s perceived to be in danger. For instance, if there are issues that the companies will produce fewer cost-reducing applied sciences, then a mannequin specializing in price discount could also be extra applicable. Nevertheless, if there are issues that the companies will abandon plans to develop new merchandise, then a mannequin specializing in product selection can be extra related. This alternative is said to proof and measurement as a result of it is necessary that any empirical tabulation of innovation inputs and outputs match these occurring inside the mannequin; not simply any quantification of ranges of or adjustments in innovation, broadly construed, will suffice. Returning to the examples above, if the mannequin focuses on price reductions, then empirical work must also concentrate on improvements that cut back prices. Equally, if the mannequin focuses on the event of new merchandise, then so ought to empirical work. Generalized measurements would danger measuring the improper factor totally in particular instances.

Second, innovation impacts shopper welfare rather more circuitously than costs do, and an antitrust authority faces trade-offs when deciding how a lot to concentrate on welfare. Measures that the majority intently approximate welfare penalties, akin to gross sales of new merchandise and decrease manufacturing prices, are additionally closely influenced by elements except for innovation, akin to demand for the brand new product in the market or shocks to the prices of uncooked supplies used to make the brand new product. Measures akin to funding, expenditure, and quantity of R&D staff could also be simpler to watch, however are rather more distantly associated to welfare. In the end, the very best method will depend upon the information of every case.

These points might be illustrated by contemplating a way that previous instances have used: tabulating patent citations. Of their evaluation of the Dow/DuPont merger, the European Fee (EC) pursued a range of such tabulations, citing tutorial literature: ‘One important finding of the economic literature is that citation-based indexes are informative on the technological importance of patents.’[v] Whereas investigating patents, weighted by some measure of their significance, akin to citations, might have some probative worth, the method additionally highlights the 2 points above.

First, patents cowl a range of differing types of innovation—manufacturing processes and new merchandise, for instance—which mustn’t enter any mannequin homogeneously.

Second, the connection between patents and true innovation or its welfare penalties is oblique and, in some instances, non-existent. Patenting is just not an inevitable manifestation of innovation however moderately a strategic alternative. On the one hand, patent thickets display that producing extra patents is just not at all times good or pro-competitive.[vi] Bennato and others (2018) argue, ‘it has long been recognised that patents are sometimes used to protect an incumbent’s market power.’[vii] However, sure companies select to not patent socially precious improvements, and as an alternative to take care of their worth as a commerce secret. Even for patents that defend socially precious improvements, they monitor neither the eventual output (precious items and companies) nor the sources invested ({dollars} or staff)—two of the categories of portions almost definitely to seem in a mannequin of innovation and rivalry.

One other empirical method, much less generally used, is evaluation of deal worth to find out whether or not a transaction in an innovation-related {industry} is perhaps a ‘killer acquisition.’ Cunningham and others (2019) outline ‘killer acquisitions’ as acquisitions the place ‘incumbent firms … acquire innovative targets solely to discontinue the target’s innovation projects and pre-empt future competition.[viii] The UK’s Competitors and Markets Authority used the deal-value method in PayPal/iZettle and discovered no proof that the transaction was a ‘killer acquisition.’[ix] Nevertheless, such an method is fraught with potential shortcomings.

First, as with patents, the transaction value is strategically chosen, not a direct measure of agency worth or anticipated innovation. Certainly, the Hart-Scott-Rodino Antitrust Enhancements Act of 1976 created transaction greenback thresholds beneath which firms are usually not required to file for a pre-merger evaluation with U.S. regulators,[x] and Cunningham and others (2019) discovered extra acquisitions happen just under this threshold than can be anticipated and, moreover, that these transactions have been more likely to contain a discontinuation of analysis initiatives.[xi] This means that the transaction costs are set with coverage in thoughts. Nevertheless, concentrating an excessive amount of on this relationship may additionally be deceptive. Acquisitions—‘killer’ or not—can be anticipated to keep away from valuations simply above the brink and bunch just under the brink, to keep away from the pricey course of of merger evaluation. Furthermore, the empirical discovering of ‘killer acquisitions’ with low sufficient values to be manipulated under the submitting thresholds is inconsistent with the idea {that a} ‘killer acquisition’ is perhaps recognized by the massive worth an entrant with noteworthy potential to disrupt an incumbent might demand as a cost to forego that potential.

Second, when a given goal operates in a number of geographic or product markets, the transaction worth will characterize an aggregation throughout these markets. If just some of them are candidates for the ‘killer acquisition’ label, it will likely be a lot tougher (or unimaginable) for a contest authority to find out the consideration paid particularly to the goal’s participation in such markets and then examine that towards an inexpensive benchmark. Basically, this can be a explicit manifestation of the issue that innovation is just not a homogenous phenomenon, and an applicable mannequin ought to separate differing types.

III. Treatments

Within the occasion that anticompetitive hurt is discovered to be probably, then a treatment—typically divestiture of sure belongings—have to be recognized.[xii] Figuring out acceptable divestiture packages for mergers involving innovation theories of hurt entails two points particular to these mergers: (a) whether or not to divest merchandise in the marketplace or merchandise in the pipeline, and (b) whether or not the divestiture of complete merchandise or analysis pipelines is important to make innovation competitors doable. Problem in discovering normal coverage solutions to those questions finally derives from the sooner commentary that economists have but to develop a persistently relevant mannequin of how improvements are created. Due to this fact, economists face problem in figuring out which alterations to the market construction will end result in extra innovation—i.e., which divestitures will work—and might generally want to decide on a safer, however suboptimal, path.

The primary subject talked about above arises in mergers that includes ‘product-to-pipeline’ competitors—that’s, one of the events has a product in the marketplace, whereas the opposite get together has a competing product in improvement. The FTC has expressed a desire for divesting the product in the marketplace in these instances.[xiii] Certainly, in the latest, high-profile Celgene/Bristol-Myers Squibb merger, the merged entity divested Celgene’s Otezla enterprise—the product already in the marketplace.[xiv] Nevertheless, in the Amneal/Impax merger, which concerned a number of overlapping product markets, most divestitures have been of the pipeline product; a number of merchandise at present in the marketplace have been ‘complicated to manufacture,’ and one other held a monopoly place. In each instances, the FTC acknowledged that divesting the market product would possibly result in its failure, which was an unacceptable danger to shopper welfare. Due to this fact, they selected the safer path given case-specifc information regardless of their normal causes to favor the divestiture of a longtime product.

‘Pipeline-to-pipeline’ mergers—these the place the 2 events have merchandise in improvement that may compete have been they each to come back to market—increase the query of the suitable extent of the divestiture or mental property licensing. In some instances, once more as a consequence of lack of a sturdy mannequin of innovation, competitors authorities fear whether or not divesting a selected analysis program will intervene with the innovation course of, or whether or not requiring patent licensure will probably be enough to spur innovation at competing companies. For instance, with respect to the Illumina/Pacific Biosciences merger, the CMA discovered patent licensure to be inadequate, acknowledging the significance of ‘know-how.’[xv] Equally, with respect to the Dow/DuPont merger, the EC required divestiture of ‘almost the entirety of DuPont’s world R&D organisation’ since doing so would ‘enable … a buyer to sustainably replace DuPont’s aggressive impact in these markets and proceed to innovate.’[xvi]

IV. Conclusion

Evaluating innovation theories of hurt is difficult as a result of (i) economists lack a sturdy, correct mannequin of how improvements are created; (ii) proof and measures used in the analyses should be chosen with care as a result of improvements are usually not homogenous, and the route between innovation and shopper welfare is advanced; and (iii) figuring out acceptable divestiture treatments entails predicting the innovation penalties of altering the market construction, which is troublesome as a result of of the dearth of a sturdy, correct mannequin. Due to this fact, mergers elevating innovation theories of hurt will probably require case-specific approaches to deal with these challenges.


[i] Kenneth Arrow, ‘Economic Welfare and the Allocation of Resources to Invention’ in The Charge and Route of Creative Exercise: Economic and Social Components (Universities-Nationwide Bureau Committee for Economic Analysis and Committee on Economic Development of the Social Science Analysis Councils), 609–26.

[ii] Joseph Schumpeter, Capitalism, Socialism and Democracy (Harper & Brothers, New York 1942), 82. Cited by Carl Shapiro, ‘Competition and Innovation: Did Arrow Hit the Bull’s Eye’, in Josh Lerner and Scott Stern (eds), The Charge and Route of Creative Exercise Revisited (College of Chicago Press, Chicago 2012), 361–404.

[iii]Shapiro (n 6).

[iv] For a pair of examples of methods to categorize innovation, see Rebecca M. Henderson and Kim B. Clark, ‘Architectural Innovation: The Reconfiguration of Existing Product Technologies and the Failure of Established Firms’, (1990) 35 Administrative Science Quarterly 9–30; Greg Satell, ‘The 4 types of Innovation and the Problems They Solve’, Harvard Enterprise Evaluate (21 June 2017) <[v] European Fee, Case M. 7932 — Dow/Dupont: Fee Choice of 27.3.2017 [27 March 2017] ¶ 389.

[vi] See, for instance, Stefan Wagner, ‘Are “Patent Thickets” Smothering Innovation?’ Yale Insights, interview.

[vii] Bennato and others (n 15) 6.

[viii] Colleen Cunningham and others, ‘Killer Acquisitions’ working paper [2019].

[ix] UK Competitors & Markets Authority, ‘Final Report’, Accomplished Acquisition by PayPal Holdings, Inc. of iZettle AB, ¶ 11.

[x] US Code, 2006 Version, Complement 5, Title 15 – COMMERCE AND TRADE, Chapter 1 – MONOPOLIES AND COMBINATIONS IN RESTRAINT OF TRADE, § 18a – Premerger notification and ready interval.

[xi] Cunningham and others (n 21).

[xii] US Federal Commerce Fee, ‘Statement of the Federal Trade Commission’s Bureau of Competition on Negotiating Merger Remedies’ accessed 29 January 2021.

[xiii] US Federal Commerce Fee, ‘Analysis of Agreement Containing Consent Orders to Aid Public Comment’, Within the Matter of Amneal Holdings, LLC, Amneal Prescription drugs LLC, Impax Laboratories, Inc., and Impax Laboratories, LLC, 4–5.

[xiv] US Federal Commerce Fee, ‘Analysis of Agreement Containing Consent Orders to Aid Public Comment’, Within the Matter of Bristol-Myers Squibb Firm and Celgene Company.

[xv] UK Competitors & Markets Authority, ‘Notice of Possible Remedies under Rule 12 of the CMA’s Guidelines of Process for Merger, Market and Particular Reference Teams’, Anticipated Acquisition by Illumina, Inc. of Pacific Biosciences of California, Inc., ¶ 28.

[xvi] European Fee, ‘Mergers: Commission Clears Merger between Dow and DuPont, Subject to Conditions’, press launch [27 March 2017].

Source link