This action might not be possible to undo. Are you sure you want to continue?
Maarten Pieter schinkel
Amsterdam Center for Law & Economics Working Paper No. 2007-05
The complete Amsterdam Center for Law & Economics Working Paper Series is online at: http://ssrn.acle.nl For information on the ACLE go to: http://www.acle.nl
Electronic copy available at: http://ssrn.com/abstract=1009573
Forensic Economics in Competition Law Enforcement Maarten Pieter Schinkel1
4 September 2007
forthcoming in the Journal of Competition Law and Economics (March 2008)
This paper delineates the specialty field of forensic IO as the application of theoretical and empirical industrial organization economics in the legal process of competition law enforcement. Four stages of that process which can benefit from forensic IO techniques are distinguished: detection and investigation; case development; decision making and litigation; and remedies, sanctions and damages. We survey the use of economics in such aspects as identifying potential forms of anticompetitive behaviour, screening markets for competition law violations, determining causality, advising on appropriate remedies and assessing antitrust damages. The paper discusses the role of expert economic witnesses in competition cases. It calls for an organization of forensic IO within the context of existing forensic institutes.
Keywords: forensics, antitrust, litigation, competition law JEL-codes: K21, L44, K41, K42
Department of Economics and ACLE, Universiteit van Amsterdam and CEPR. Roetersstraat 11, 1018WB Amsterdam, The Netherlands. E-mail: firstname.lastname@example.org. The papers in this special issue were selected from contributions to the ACLE workshop Forensic Economics in Competition Law Enforcement, held March 17, 2006 at the Universiteit van Amsterdam. John Connor, Andrew I. Gavil and Franklin M. Fisher made outstanding keynote contributions to the event. I am indebted to them, as well as other participants to the workshop, for stimulating discussions. I thank the editors of the Journal of Competition Law and Economics, Damien Geradin and Gregory Sidak, as well as more than twenty anonymous referees for their assistance in the editorial process. Iwan Bos, John Connor, Kati Cseres, Franklin M. Fisher, Andy Gavil, Vivek Ghosal, Jon McNally, Jakob Rüggeberg, Francesco Russo, Daniel Slottje and Jan Tuinstra gave constructive comments to an earlier version of this paper. Norman Bremer and Marie Goppelsroeder provided excellent research assistance. Opinions and errors remain mine.
Electronic copy available at: http://ssrn.com/abstract=1009573
1. Introduction Forensic science is the application of scientific disciplines to the legal process.2 Scientific methods and tools have found productive and popular application in criminal law enforcement. Techniques such as DNA-profiling, blood spatter projection and handwriting analysis, are used to establish suspects, methods and motives in homicide cases. They feature in television and documentary series, such as Forensic Detectives, which graphically describe the pivotal role of forensic techniques in police investigations and criminal law proceedings. Forensic tools have become important in the discovery of crimes, their investigation and presentation in court, and in establishing causality and damages. The field has its own professional journals, university curricula, respected national forensic laboratories and institutes.
There exist a wide variety of applied forensic methods. Saukko and Knupfer’s 3 volume Encyclopaedia of Forensic Sciences contains more than two hundred different alphabetically ordered entries, ranging from Airbag Related Injuries and Deaths to Wood Analysis. Most methods derive from natural sciences. They include bio- medics – to determine skeletal trauma, make time-since-death analysis and conduct morphological age estimation – chemistry and toxicology – including ink analysis in document forgery and counterfeits – as well as physics – informing enforcers with ballistics, bloodstain analysis, image processing and forensic IT – and engineering – understanding fire-scene patterns in buildings and motor vehicle accidents. The three volumes of the encyclopaedia also describe applications from psychology – such as criminal profiling, eyewitness reliability and methods to detect lies and deception. In addition, it has some entries on white-collar crime and forensic computing.
Economics is a science that has methods and tools to offer for application in legal processes. Insights from the discipline of industrial organization (IO) in particular are extensively used in the interpretation and the enforcement of the competition laws. Expert IO economists have made contributions to both the challenge and defence of mergers that raised competition concerns and alleged antitrust violations. They have
Lyle (2004), p.8. The Forensic Science Society defines it similarly as: “the application of science to the law.” See http://www.forensic-science-society.org.uk.
assisted in assessing antitrust effects and damages, as well as the identification of new types of behaviour that could potentially be anticompetitive. Their economic justifications have been taken into consideration in the judgements of the Supreme Court of the United States, as well as the European Court of Justice. 3
The aim of this special issue on Forensic Economics in Competition Law Enforcement is to delineate the speciality area of industrial organization economics applied in various aspects of the legal process of antitrust cases. It is an exciting area of work, in which a great many industrial economists are active. It generates a considerable turnover as well as large likely deterrence benefits. The area has an active international conference circuit and even its own society journal, the Global Competition Law Review.4
The area nevertheless lacks a systematic description as a discipline. Saukko and Knupfer’s Encyclopaedia does not contain a single entry related to economics – not even on forensic accounting, which is the analysis of financial and economic transactions for the purpose of assessing their legality. Likewise, there is no mentioning of topics related to economics in the description of forensic science given by the Forensic Science Society. 5
The American National Association of Forensic Economics (NAFE), and the Journal of Forensic Economics (JFE) which it publishes, has had a strong business orientation since its creation in 1986. In their inaugurating contribution to the JFE, John Ward and Gerald Olson set the research agenda for the field:
“The primary focus of the research of the forensic economists is the measurement of market loss (damages) arising from market failures, contract disputes or Torts.” (Wald and Olson, 1987, p.2)
See Adams and Brock (1991) and Kaplow and Shapiro (2007). See Slottje (1999), a collection of essays by leading economists that were active in high-profile antitrust cases and the round-table on expert witness work in antitrust cases in the 2003 spring issue of Antitrust. Contributions to both are discussed in the text below. Baker (2003) and Neven (2006) attempt to quantify the effects of competition law enforcement in the US and Europe respectively, concluding that the gains are likely to far outweigh the costs in either regime. 5 See http://www.forensic-science-society.org.uk. The society publishes the journal Science & Justice.
Antitrust is mentioned as a specialization of damage calculation in public interest disputes, but considered an area of minor importance compared to business valuation. 6 Indeed, the vast majority of the papers published in the volumes of the JFE is on the quantification of damages in individual tort cases. Topics include the appropriate discount rate, expected employment duration and the effects of progressive taxes in present value calculations of lost earning as a result of personal injury and wrongful death. Typically, causality in these accident cases is straightforwardly established and has nothing to do with economics. Only a hand- full of papers discusses applications of economics to competition cases. 7
In a further methodological paper in the Journal of Legal Economics, Ireland (1997a) recognizes the application of “econometric tests of causality” in antitrust, but keeps it outside the scope of forensic economics as “this element involves only a minority of forensic economists” (op. cit., p.7). Emphasis remains on damages assessments. Finally, Kaufman et al. (2005) define forensic economics as “the application of economics to litigation” (op. cit., p.xv). The volume of seminal contributions to the field claims to offer coverage of “all the major topics in this sub- field” (ibid.) but contains not a single paper on antitrust.
We define the specialty field of ‘forensic IO’– as the application of theoretical and empirical industrial organization economics in one or more of the various stages of the legal process of competition law enforcement. 8 Apart from antitrust economics, the area overlaps with law and economics and the economics of law enforcement. Forensic IO is much narrower than IO itself, which is only in part applied to antitrust. This latter part, antitrust economics, in turn is narrower than forensic IO, as it focuses on the economic interpretation of the competition laws, more than on the techniques and tools of their enforcement by institutions and courts.
This research focus on damages assessments is also maintained in Thornton and Ward (1999). An interesting early example is Einhorn (1989). There even appears to be some hostility towards antitrust applications. Slottje (1999) was received in the JFE as a “view from Mt. Olympus” (Vernarelli, 2001, p.101). According to this review, the volume had the “mainstream forensic economist” little insight to offer. Instead, the reviewer suggested that “…if, heaven forbid, Professor Slottje were to suffer serious injuries when the car he is driving is hit by a tractor trailer which runs a red light, then he would really learn what forensic economics is all about” (op.cit., p. 103). 8 One of the first references to the term ‘forensic economics’ may be Boudin’s 1984 book review of Fisher et al. (1983), based on the authors’ experience as consultants to IBM in the American IBM case, on which more in Section 4. See also Fisher (2000).
In his introduc tory contribution to this special issue, John Connor traces the use of forensic economics back to the first written record of an antitrust proceeding in the Roman Forum, more than 300 years BC. Connor discusses its subsequent role in modern cartel law enfo rcement in both the US and Europe. He establishes a considerably longer modern tradition of applied forensic IO in the US courts, but also sees an emerging field in Europe resulting from the European appellate courts increasingly requesting sound economic argumentation and emerging private antitrust suits before the national courts. The paper focuses on cartels and surveys a number of methods for the construction of “but for” worlds and the determination of cartel overcharges.
Forensic IO is not limited, however, to the assessment of damages in antitrust litigation. It makes some of its most fundamental contributions to competition cases by assisting to establish causality. Economic analysis can help to determine, for example, the relationship between a collusive agreement between incumbents to restrict entry and low product quality due to a resulting lack of effort. Economics can be applied to identify market power and how it was abused. Its tools can also be used to value claims of merger-specific efficiencies, as well as to determine the extent to which intermediaries and final consumers were damaged by the anti-competitively raised prices of their suppliers.
To determine likely causalities requires a complex process of building a relevant economic theory, deriving testable hypotheses, and corroborating them with the help of econometric tests. Precisely this is the scientific method for which academic experts are consulted to produce objective and reliable truths. 9 As a result, it is on issues of causality that many of the more interesting differences of opinion exist between experts applying industrial economics to antitrust. In fact, the Daubertqualification of admissible scientific expertise – discussed in Section 5 below – is essentially concerned with the question whether causation has been shown satisfactorily and by means of accepted scientific methods.
On methodology generally, see Chalmers (2004). On economics as a science, see Hands (2001).
The relevant elements of the legal processes in which forensic IO can assist should also be interpreted broadly. To begin with, forensic IO is not restricted to (criminal) cartel litigation and per se monopolization or abuse cases. In fact, the application of IO economics is often most powerful in investigations that come under the rule of reason. The same is true for merger investigations, in which the legality of acts under the antitrust laws is assessed ex ante. It is in these cases that the causality of economic effects often is an important part of the analysis. 10
Forensic IO is also not limited to litigation. Instead, its tools find application in investigation, litigation and adjudication, as well as the enforcement of remedies. In the US, Federal and State courts decide on cases that are prosecuted by the antitrust agencies. Appeals can reach the U.S. Supreme Court. In Europe, competition law enforcement is mostly an administrative process in which the agencies take formal decisions. These decisions can be appealed with the national tribunals and appellate courts, or, in cases affecting trade between Member States, with the Court of First Instance and European Court of Justice. Actual legal proceedings are a mixture of the elements of the court-based and administration-based models. 11 The US antitrust authorities have considerable discretionary powers in deciding to sue or settle cases, for example. Also, several parts of the European decision making processes involve litigation-style events, such as oral hearings. Furthermore, the emerging private antitrust damage actions in Europe are primarily before national courts. Forensic IO finds application in all of these aspects of competition law enforcement.
Forensic IO is an academic discipline that contributes to the advancement of economics as a science. The debate over important competition cases in many cases spreads from competition agencies and the courts to academia. Frequently, the stakes of the parties in competition cases are high and they are willing to invest in a good representation of their case. Likewise, the forensic experts testifying to the case in court have valuable reputations to protect and know that their contributions will be publicly scrutinized by the opposing side. As a result, the debates are often of very high quality. Economic reasoning between theory and empirical evidence helps to
For example, the contributions on antitrust in a special issue of the Journal of Econometrics on “Statistics and Econometrics in Litigation Support”, edited by Robert Basmann and published March 2003, are concerned with estimation techniques for merger control. See Basmann (2003). 11 See Fingleton (2005), Neven (2006), Section 4 and Borrell (2006).
support and sharpen the legal analysis of the facts and circumstances in the case. In return, the involvement in actual cases provides access to internal corporate documents that normally would be confidential and unavailable to the academic economist. The economic principles of landmark cases can so be an inspiration for academic contributions in the abstract that can be developed beyond the litigation deadlines and budgets.
As a result of this, in a growing number of competition policy regimes around the world, IO economics and the interpretation of the competition laws work and develop in a productive symbioses. Novel business strategies cast up anticompetitive concerns that subsequently inspire fundamental academic research. The recent software and credit card cases, for example, stimulated the research area of network economics and two-sided market theory. 12 At the same time, new economic insights and techniques are being translated into revisions of guidelines and priorities in enforcement. The adoption of the efficiency defence in the 2004 EU horizontal merger regulation was likely influenced by the seminal work of Oliver Williamson. 13 Similarly, leniency programs have been designed on the basis of a game-theoretical understanding of the problem of internal cartel stability. Finally, courts have become increasingly open to considering economic justifications in their rulings.
It is possible to distinguish four (related) stages of the process of competition law enforcement in which forensic IO has a role: first, detection and investigation, second, case development, third, decision making and litigation, and fourth, remedies, sanctions and damages. In the following four sections, each of these roles is discussed in some depth. Section 6 offers some concluding remarks on forensic IO as a discipline.
2. Detection and Investigation
Competition law violations are different from more traditional crimes, such as murder or burglary, in the sense that anticompetitive acts often leave no obvious traces. There is not necessarily a body, signs of a break- in, or at all a crime scene. The fact that a
See Evans et al. (2000), Fisher and Rubinfeld (2001), Shy (2001) and Rochet and Tirole (2006). Williamson (1968). See also the introductory chapter in Ghosal and Stennek (2007).
cartel had raised prices, for example, can be difficult to prove with only time-series on prices and no data on costs. Often, however, cartel damages to society are more indirect; for example when reduced competition leads to lower quality of goods and services, less innovation efforts or the suppression of new technologies. Even people who are close to the cartel conspirators, including in- house counsel and in some cases higher management, may not know about the illegal practices that are going on in their organizations. Forensic IO therefore has an important role to play in the discovery of the very fact that a violation of law was committed, essentially in two ways.
First, economics can uncover anticompetitive aspects of behaviour previously not understood as restrictive of competition. In this sense, economics informs the interpretation – or even the choice – of competition law rules and standards of conduct. A business practice is only anticompetitive if it is shown to be restrictive of competition by some plausible economic argumentation. Sometimes, a new type of antitrust violation is discovered through independent theoretical and empirical academic research. An example of this may be the insight pointed out by Steve Salop that price- matching clauses, although they appear genuinely competitive can in fact sustain collusion. 14 Another example is the more or less accidental discovery of a suspicious lack of odd-eight quotes for actively traded NASDAQ securities by William Christie and Paul Schulz. The research uncovered the market- makers conspiracy, which led to an antitrust class action that was eventually settled in the late 1990’s. 15
Often also, the understanding of possible anticompetitive strategies develops over the course of an antitrust investigation. A prominent example is the US Microsoft case, in which the insight was developed, primarily in the testimony of Franklin M. Fisher, that Microsoft had attempted to levy existing monopoly power from the market for operating systems to that of internet browsers by means of a predatory tie between the two products. 16 This type of forensic assistance in antitrust cases is highly innovative, drawing on theoretical and applied developments such as game theory, computer
See Salop (1986). Corts (1997) challenges the view that price-matching clauses are anticompetitive. See Christie and Schultz (1994) and In Re NASDAQ Market-Makers Antitrust Litigation, MDL No. 1023, (S.D.N.Y.) (R.W.S. 94 CIV 3996). 16 See Evans et al. (2000).
simulation techniques, and experimental economics, thus advancing both competition law enforcement and academic research.
Second, forensic IO can help find recognized types of antitrust violations by systematically screening industries and firms. A suspicion of an anticompetitive act will typically be aroused by some kind of irregular market behaviour. A purchaser, for example, may be alerted to the presence of a wholesale cartel by a sudden increase in prices, or, if the cartel has divided the market geographically, by a refusal to supply all but his local dealer. More sophisticated “tell- tale signs” of complicated abuses of dominance or collusion may only be detectable by specialized detectives who monitor markets to find violations. 17 Examples may be sudden atypical changes in sales conditions or product quality captured by the data, a decrease in price volatility over time, odd bidding behaviour in an auction, or correlated capacity investments. 18
Monitoring markets for behavioural patterns indicative of collusion can help target further inspections of companies that display suspicious behaviour. An emerging literature develops such ‘live forensics’ methods to systematically screen markets for antitrust violations, in particular cartels. Practitioners as well as academic economists contribute to the development of sophisticated antitrust screens. 19 They typically apply a combination of two types of indicators of cartel likelihood: structural and behavioural indicators.
The first type of indicator is based on structural characteristics of industries and markets in which theories of collusion predict cartels to be particularly sustainable and in which cartels were found in the past. 20 These characteristics include relatively stable demand for a low-tech homogenous product sold in large volumes by a relatively small group of more or less identical suppliers that frequently interact in a
Effective collusive markers are suggested by contextual and detailed cartel studies such as Simpson (1993), Griffin (2000), Connor (2001; 2007), Eichenwald (2001) and Mason (2004). 18 See, for example, Porter and Zona (1993) and Bajari and Summers (2002). The US Department of justice publishes a leaflet on its website that advises the public on how to identify collusive schemes. 19 See Porter (2005) and Harrington (2007a; 2007b). The distinction between ‘live’ and ‘dead’ forensics stems from forensic IT. Live forensics takes place in real time and analyses running IT systems in use, whereas dead forensics analyses a conserved snapshot of system. It is more complicated, but less invasive and allows for monitoring industries without the market parties being aware of it. See Volonino (2003) and Adelstein (2003). 20 See Grout and Sonderegger (2007).
transparent market. Likewise, relatively little buyer power, or common crossownership may be reason to zoom in on certain ind ustries rather than others. The second type of indicator uses behavioural screens of the kind discussed above. 21 It is applied to pre- identified high-risk industries in order to systematically identify potential antitrust concerns.
Forensic IO can assist the antitrust agencies to actively discover and assess the illegality of certain business strategies. The US Department of Justice Antitrust Division, for example, has developed ‘cartel profiling’ techniques based on its experience that cartel activity may take place in markets adjacent to industries under investigation for collusion, or to merger inquiries. 22 The division pre-emptively monitors industries in which convicted price- fixers are active as “mentors” in “cartel trees”, or that are vulnerable to cartelization because they have the kind of market structure characteristics discussed above.
Academic research can provide insights into uncovering such patterns. The paper by Vivek Ghosal, who worked for the Antitrust Division for several years, focuses on the genesis and taxonomy of cartel investigations. Ghosal examines the dynamic interrelationships between civil and criminal investigations. His empirical findings suggest that prior criminal prosecution of firms and individuals, as well as information gleaned from merger and monopolization investigations, provide useful leads for future cartel prosecutions. Ghosal also finds that cartel prosecutions rise following economic downturns, which he relates to the stability and breakdown of cartels.
Independent and active detection of antitrust violations is important for an effective enforcement regime. It allows for an efficient allocation of limited enforcement budgets over various antitrust priorities. This helps to create a sufficiently high probability of discovery by the authorities, which in turn is essential to destabilize active cartels through the leniency programs. 23 In their paper in the issue, Hans Friederiszick and Frank Maier-Rigaud argue that ex officio inspections remain crucial
See Harrington (2007b). See Barnett (2007). 23 See Rey (2007).
in an era where m any cartels apply for leniency. Their contribution reflects a highlevel strategy debate in the European Commission’s Directory General Competition. Friederiszick and Maier-Rigaud make a strong case for a mixed instruments methodology in which complementarities in enforcement are exploited.
Another perspective from inside the agencies is offered by Peter van Bergeijk. In his paper he revisits the Dutch construction cartel, which was a complex and wide ranging case of collusion that involved nearly all major construction companies in The Netherlands, which lasted for more than ten years. Van Bergeijk is one of the people responsible for the Dutch cartel “Paradise Lost” and worked for the Netherlands Competition Authority (NMa) as an economic expert on the construction cartel case. His cartel post mortem offers intriguing insight into this Dutch case, as well as useful suggestions for tell- tale signs of collusion.
Companies take a keen interest in forensic IO methods to detect antitrust violations as well. To know what collusive markers can trigger an investigation is useful for hard core repeat cartel offenders of the type the US DoJ profiles. In particular, it can help them to hinder or even avoid detection. It has been documented how specialized cartel consulting firms organize secretive meetings, making sure no physical evidence leaves the smoke-filled room. 24 Allegedly, it is possible to reserve server space in Switzerland, where the inspection rights of competition authorities are limited, to store sensitive documents. Just like a sophisticated thief makes sure to wear gloves to avoid leaving finger prints, so will white collar crime constantly professionalize. It is crucial that the antitrust agencies make sure to continue to update their use of the latest detection methods in their struggle to stay ahead of this game of hide-and-seek.
Private-eye antitrust detection also serves a legitimate demand. In larger companies, it may well be the case that the owners or senior management disapprove of anticompetitive practices, but have insufficient oversight over their complex organizations to know of the existence of the latter. Lower managers in subdivisions or remote parts of the world may use collusive schemes to fix the division targets to
Case COMP/E-2/37.857, PO/Organic peroxides. Commission decision of 10 December 2003.
which their persona l bonuses are tied, thus leaving the firm with future liabilities.25 Higher management that deals insufficiently with cartel problems may even risk criminal prosecution. Those involved in cartel conspiracies often even manage to hide the cartel from in- house counsel, which has a duty to report irregular activity. 26 Alternatively, in planned acquisitions or mergers, forensic IO detection methods can support due diligence inspections by the prospective buyer. Internal audits for corporate-risk analyses can find potential antitrust problems before the authorities do and clean them up. 27 Leniency programs offer companies a way to resolve their unwanted antitrust problems, if they manage to discover a cartel in their organization before anybody else does. Compliance programs can subsequently be put in place to help prevent future violations. 28 As the market for these forensic services develops, so do the methods of detection available for public policy.
3. Case Development
A second role for forensic economics in competition law enforcement exists in providing assistance in the development of cases prior to litigation, either on the agency or prosecutor side, or on the defence side. Partially, this concerns forensics issues such as collecting physical evidence of monopolization strategies or cartel meetings, such as paper minutes, e- mail correspondence, or conference call records that goes beyond initial detection. 29 This type of investigations aims to build a profile of individuals and their networks similar to the traditional white collar crime forensics methods mentioned in the introduction. It can benefit from the analysis of information that is readily available within the business environment through employment records, expense claims and network data. Forensic accounting and IT techniques can be used
On corporate governance issues and corporate crime generally, see Alexander and Cohen (1999), Paul (2000) and Green (2006). Applications to antitrust are offered in Jamieson (1994), Garoupa (2000), Spagnolo (2005), Buccirossi and Spagnolo (2006). 26 Sophisticated methods to hide cartels internally have been revealed in several of the cartel cases recently documented, including the auction houses cartel. See Mason (2004). 27 See Anastasi (2003) for an account of the role of forensics in corporate fraud detection. Internal company data often allow for analysing evidence of personal benefit, such as preferential treatment or bribes, and on-the-job surveillance. The industry has developed highly sophisticated tools, including packages such as ISYS, DTSearch and Attenex that assist in extracting relevant intelligence from background noise. 28 See Beckenstein, Gabel and Roberts (1983) 29 In U.S. vs. Microsoft Corp., Civil Action No. 98-1232, for example, discovered internal emails on the ‘jihad to win the browser war’ played an important role in establishing the intend to monopolize.
to recover deleted files. 30 It may further involve covert surveillance, handwriting analyses and voice recognition to identify responsible individuals and their network. Forensic IO can advise here on targeting inspections in ‘dawn raids’ on the premises of companies by identifying the type of suspected antitrust violation and the likely evidence that it leaves.
The main function of forensic IO, however, is in building the economic logic of a case. The economics applied in this is closest to the material covered in the IO textbooks, and so I confine myself here to only a brief discussion. 31 The extent to which industrial organization economics can be used in a competition case depends very much on the nature of the case. In merger investigations, structural models and econometric specification are important in determining post- merger effects and matching divestitures. IO theory and empirics also underlies monopolization cases in the US and abuse of dominance cases in Europe. Here economics can point out the principles that are at work to subvert competition, or the other way around, show how behaviour that appears anticompetitive really is just genuine competition. In per se cartel proceedings, the finding of an infringement will be based on direct evidence. Nevertheless, economics can help determine whether the market structure is conducive to collusion, whether there is an incentive to collude, or if a cartel could at all be sustainable. 32
In any case of substance, both sides are likely to benefit from employing economic advisors and both sides typically decide to do so. The standard of economic work in many antitrust investigations is high, often leading to additional insights and a deeper understanding of economics. An advanced form of economic consultancy has developed that applies cutting edge economic techniques, reasoning and evidence. These competition practices employ PhD’s in economics and industry experts. They maintain international and academic networks, which allows them to put together teams of experts with a division of labour that is tailored to the case at hand.
See Volonino (2003) on the admis sibility of electronic evidence in computer forensics cases. Tirole (1988), Carlton and Perloff (2006) and Motta (2004). 32 See Milne and Pace III (2003), Werden (2004) and Johnson (2007). An early contribution in the JFE is Einhorn (1989).
A number of routine economic analyses, such as application of the SSNIP-test for the determination of the relevant market, HHI calculations for merger assessments, the Pivotal-Supply- index (PSI), or the Elzinga-Hogarty-test have become standard procedure. 33 Yet, diversity across industries assures that many competition disputes have one or the other fundamental issue of general economic interest. In some of the landmark antitrust cases, such as Microsoft, VISA/Mastercard and Leegin, the contributions of economists involved have been crucially innovative. 34 They generated important positive externalities for our general understanding of the workings of these special markets and the potential monopolization, abuse and cartel strategies to which they are exposed.
The same is true for some of the larger merger investigations. In GE/Honeywell, which was cleared in the US after thorough investigation, the European Commission blocked the merger on the basis of a rather adventurous conglomerate merger theory. 35 It feared that post- merger product bundles of GE and Honeywell would be so attractive to customers in comparison to the separate components offered by riva ls that the competition would be forced out of business. The decision caused considerable debate. 36 It further highlighted the uneasy relationship between law and economics in European competition law enforcement when the Court of First Instance rejected most of the Commission’s analysis while upholding its decision to block the merger. Other merger inquiries in which forensic IO made competition law enforcement history are Staples/Office Depot in the US and Volvo/Scania in Europe.37 In both cases, sophisticated econometric analyses were entered to predict post- merger market developments to support the challenge of the proposed merger. 38
Obviously, in each case the choice of modelling has to go with the structure and type of competition in the industry and the particular competition concerns at hand. The
Bishop and W alker (2002) provides a useful overview of the economics toolkit for European competition cases. 34 In re Visa Check/Master Money Antitrust Litigation, Master File No. CV-96-5238 (E.D.N.Y.) and Leegin Creative Leather Products, Inc. v. PSKS, Inc., No. 06-480, June 28, 2007. 35 General Electric/Honeywell Commission decision 2004/134/EC  OJ L 48/1. 36 See Patterson and Shapiro (2001). See Vickers (2007) on the issue of transatlantic con- and divergence in competition law and economics. 37 Federal Trade Commission v. Staples, Inc., 970 F.Supp. 1066  and Volvo/Scania Commission decision 2001/403/EC  OJ L 143/74, Case COMP/M.1672. 38 See Baker (1999) and Ivaldi and Verboven (2005).
more interesting landmark antitrust cases centre on questions about the specification of the nature of competition. A theory of harm related to margin squeeze, for example, requires a dominant wholesaler who is vertically integrated in part of the retail market. An allegation of price predation would need to be sustained by showing pricing below long-run average costs, as well as the predator’s likely ability to recoup initial losses. Crude observations on the industry should fit such model specifications. 39 There is a danger in applying off- the-shelf instruments rather than bespoke models. In particular this is so in the use of merger simulation models, some of which are readily available as user- friendly software package with an appearance of generality, whereas outcomes are very much dependent on the details of structural and empirical specifications. When the models remain undisclosed to opposing sides, which often is the case when the algorithms are proprietary to the consultants, they can obscure rather than enlighten the understanding of merger effects. 40
At regular intervals, Jonathan Baker has reviewed the use of empirical methods in antitrust cases with various co-authors. 41 Baker and Bresnahan (2007) emphasises that difference among industries matter, although similarities with related industries are an important source of learning. The authors discuss a number of methods tested in past cases for the measurement of the magnitude of economic relationships and effects. These include various ways to identify buyer power substitution and market power, using different types of tests that are related to structure, conduct and performance. An important concern in the use of econometric techniques is the quality of the data. The interpretation of empirical findings is further relative to a proper market specification. A high market concentration (few firms) in a bidding market, for example, may not be so much of a concern, whereas it can signal dominance in a regular production market.
In light of the many degrees of freedom with regards to model specification and data collection unfolds a discussion on “bright line tests”. Simple tests were called for, amongst others, by Easterbrook (1984), to help reduce enforcement costs and create legal certainty. Baker and Bresnahan (2007) suggest that where such tests presently
See Wish (2007). See Goppelsroeder and Schinkel (2005) and Walker (2005). 41 See Baker and Bresnahan (1992, 2007) and Baker and Rubinfeld (1999). Bishop and Walker (2002) offers several chapters on the use of empirical methods in European competition cases.
do not exist, economics should further categorize generalizations and develop a standard toolkit that is sufficiently reliable for antitrust enforcers to use. In his contribution to this issue, Fisher is not so optimistic. Based on his extensive experience in economic consultancy and as an expert witness in some of the definitive antitrust cases of our time, Fisher warns against the dangers of giving in to the pressure from attorneys and judges to oversimplify economics. To do so is likely to lead to mistakes in antitrust decisions and generally harms the discipline. In his paper, Fisher gives several examples of this that are both entertaining and alarming.
One particular pitfall in measuring market power that Fisher addresses, here as well as in his earlier work, is to use evidence on profits. Paul Grout and Anna Zalewska explore the boundaries for doing so in their paper. Grout and Zalewska carefully qualify excessive rates of return to confirm they have limited reliability, certainly for assessing standalone excessive pricing cases. The authors argue, however, that when handled with care, case specific rate of return measurements can help inform antitrust decisions.
4. Decision Making and Litigation
The process of discovery and case development is obviously geared towards final decision making. Competition law is grounded in the economic insight that a workable competitive process is socially beneficial. As a result, competition law enforcement is fundamentally based on economic reasoning and evidence. The agencies apply economic analysis and interested parties will go at great lengths to present their case in the best possible light, using cutting-edge economic analyses, often presented by reputable economic expert witnesses.
Differences between the US court-based system and the European administrationbased approach, however, can have important implications for the type and quality of forensic IO analyses that are applied to a case. In the court-based model, parties present their case in a public arena before independent judges. Administrative law
processes, on the other hand, can be less transparent. 42 High quality economic reasoning and quantitative analyses have historically played a more prominent role in US antitrust than in European competition law enforcement. European practice is converging towards American practice, however, with the adoption and promotion of a “more economic approach” in recent years. 43
The quality of the forensic economics submitted is best safeguarded when the decision making process is fully public. There appear to be at least three reasons for this. First, an open process reduces the scope for factors other than substantive arguments to influence decision making. That is, favouritism, lobbying, political pressuring and corruption are exposed more easily and thereby partly deterred. Second, an open process allows for all parties involved to learn as case law develops. Enforcers, firm management, attorneys and judges can improve their understanding of economics and its application in antitrust. Third, the economists involved in competition cases can build – and destroy – their reputations. When economic analyses are available for critical examination by peers, sloppy or fabricated arguments are exposed. Unscientific analyses are weeded out, and thorough economics is supported by the possibility to establish a reputation for objective and high quality work.
The US system of expert economic witnesses in court proceedings is one of the most open forums of antitrust debate. Expert witnesses submit their testimony, give it in court and are subsequently cross-examined. In high-profile cases, this creates a competitive arena for discussion, in which industrial organization economics is truly put to the test. This is to the benefit of the field.
Posner (1999) summarizes the US law governing the use of expert witnesses. According to the original Federal Rules of Evidence (FRE), anyone who has relevant
The process of decision making in competition cases by the European Commission and many of the Member States does involve elements of an open court system, such as oral hearings, and disclosure of submitted reports. The European appellate courts, in turn, are often restricted in their freedom to consider economic arguments. See Neven (2006). 43 See Vesterdorf (2005), Roeller and Stehmann (2006) and Vickers (2007). DG Competition, and in its wake many national competition authorities in Europe, have created chief economist positions, supported by teams of PhD economists. The goal is to raise the quality of standard of in-house economic analyses, as well as the ability to appraise outside experts’ reports.
expertise “by knowledge, skill, experience, training or education” is permitted to be classified as an expert witness. The expert witness is expressly not a consultant to the parties, but an independent professional who testifies on the evidence to the courts.44 The expert has a moral and professional obligation to testify honestly to his or her professional opinion. 45 Posner does note that testifying experts tend to polarize over time, representing plaintiffs or defendants, but rarely both. This may reflect personal dispositions, but to the extent that it is a separating equilibrium, it may compromise objectivity and make it difficult for courts to choose between duelling experts.
The reliability and integrity of the expert’s testimony is safe- guarded by a number of mechanisms, including professional rules of conduct. 46 Members of the NAFE, for example, are held to work by The Statement of Ethical Principles and Principles of Professional Practice. It specifies in its first article that:
“1. Engagement Practitioners of forensic economics should decline involvement in any litigation when they are asked to assume invalid
representations of facts or alter their methodologies without foundation or compelling analytical reasons.”
On diligence, it further requires in the third article that:
“Practitioners of forensic economics should employ generally accepted and/or theoretically sound economic methodologies based on reliable data. Practitioners of forensic economics should attempt to provide accurate, fair and reasonable expert opinions. […]”
In addition, the code contains several paragraphs that are to assure transparency of information, methods and academic debate. The latter is also a legal requirement, as the expert must make the facts or data which he or she relied upon in forming opinions available for opposing counsel to cross-examine.
See also MacKenzie -Mason (2001). Posner (1999), p. 92. 46 See Piette (1991).
In case law, a higher standard for expert testimony to be admissible and sufficient to show causation developed seminally since the U.S. Supreme Court decision in Daubert v. Merrell Dow Pharmaceuticals, Inc. in 1993. 47 In Daubert, the Supreme Court considered the question what are ‘reliable’ scientific principles and methods and who qualifies as a scientific expert. Criteria to answer these fundamental questions on ‘the scientific method’ were sought in evidence based on falsifiability, replication, limitations and potential rate of error, peer review, publication and general acceptance by consensus in the academic community. The case was much debated at the time – involving several Nobel Prize laureates who in Amicus Briefs pointed out, amongst other things, the limitations of the peer review system as an indicator of scientific truth. The implications of the Daubert decision are still discussed today. 48
The methodological debate led to a revision of rule 702 of the FRE in 2000, which now specifies on expert witnesses qualifications:
“If scientific, technical, or other specialized knowledge will assist the trier of fact to understand the evidence or to determine a fact in issue, a witness qualified as an expert by knowledge, skill, experience, training of education, may testify thereto in the form of an opinion or otherwise, if (1) the testimony is based upon sufficient facts or data, (2) the testimony is the product of reliable principles and methods, and (3) the witness has applied the principles and methods reliably to the facts of the case.”
A literature developed on expert economic testimony in antitrust cases. Beyond the fundamental question whether there is sufficient consensus in the economics profession at all for it to satisfy the Daubert-test, Slottje (1999) collects expert witness account to provide insight into the “nuts and bolts of what forensic economists actually do” (op. cit., p. ix). In the same year appeared the Spring issue of the Journal of Economic Perspectives that includes Posner (1999). Gavil (2000) is an early
U.S. Supreme Court, Daubert v. Merrel Dow Pharmaceuticals Inc., 509 U.S. 579 (1993). See Solomon and Hackett (1996) for an insightful review of Daubert from the point of view of philosophy of science. See Bernstein (1996) for a general critique. Ireland (1997b) introduces a special issue of the JFE on the Daubert decision.
detailed legal study. The Spring 2003 issue of Antitrust on “Working with Economic Experts” contains various relevant contributions as well, including Mile and Pace III (2003) on successful and unsuccessful expert witness testimony in alleged cartel conspiracies. Werden et al. (2004) discusses the use of economic models in assessing the competitive effects of mergers in the context of the Daubert-criteria. Werden (2007) surveys the admissibility of expert economic testimony in antitrust cases.
In his critique to “bright- line tests” in this issue of the Journal of Competition Law and Economics, Franklin M. Fisher draws on his extensive experience as an expert witness in antitrust cases. In the process, Fisher gives various insights into how to be an effective expert economic witness. Several other leading economists had done so before in Slottje (1999), making for entertaining reading. The first contribution to the book is by Hendrik Houthakker, and largely about his involvement for the DoJ in the landmark United States v. IBM, which was filed in 1969. 49 In this case, Fisher was the opposing principal economic witness to IBM. It ran for more than 10 years before the DoJ withdrew it as “without merit”.
Houthakker describes how he was instrumental in advising the DoJ to “go after” IBM, “with its 70% market share in computers,” as a top priority in 1967. In his later role of witness, IBM’s counsel failed to intimidate him, because the lawyer who crossexamined him paled in comparison to the Gestapo tactics he had faced in his youth as a prisoner of war. Houthakker resigned as a witness when the DoJ scheduled his testimony during fall classes in 1977, a few years before the DoJ “threw in the towel.” According to Houthakker, the government’s decision to stop pursuing the case was political, and he criticizes the account of Fisher et al. (1983) that the allegations against IBM had been a mistake from the start. Houthakker experienced the adversarial element that litigation introduces into otherwise “generally peaceful” academia as not necessarily productive, which was a reason for him to select only few cases to get involved in since.
William Baumol’s account is brief but no less insightful, with several examples of how consulting experience contributed directly to his academic output, including
United States v. IBM, 69 Civ. 200 (S.D.N.Y. 1969).
contestable market theory. Top-level consulting allowed him to meet key players, Baumol says, and see things that otherwise would never have been disclosed. Frederic Scherer’s expert witness account is most instructive and amusing. He too acknowledged a cross-fertilization between research and litigation consulting. To that end, he accepted cases with substantial “precedential interest” and avoided cases in which he would be “reluctant to submit my testimony to critical judgement by my economist peers” (op.cit., p.132). Scherer would further not testify on behalf of firms that were likely to have committed per se law violations and only in cases that concerned a limited number of industries that he knew well.
Scherer distils some valuable lessons from “surviving” roughly 40 years in “a blood sport” ( op.cit., p. 129). The first is: “Know thy subject matter thoroughly.” This lesson Scherer explains he learned the hard way: being cross-examined in the IBM case. Good preparations include reading everything, but also visiting production facilities, talking to product users and preparing a chronology of key events. The second rule is: “Admit your mistakes and get on with the show.” Third is not to try to deny the significant weaknesses in your client’s case, as it may discredit the entire testimony. Fourth, “search one’s prior writings for statements that might be construed as conflicting with impeding testimony.” Such potential contradictions should be explainable and: “Altering views for specific testimony is not recommended.” Finally, Scherer advises to: “always tell the truth, as best you know it.” (ibid.)
A key concern with all the contributing experts is to guarantee independence and objectivity – or to avoid the “whose bread I eat, his song I sing”-attack, as Scherer calls it ( p. cit., p.130).50 The NAFE-code forbids forensic economists to accept o contingency fees, or fees that relate to the size of the court award or settlement. The large amounts of work that competition cases often are, are normally compensated on the basis of hourly fees, which may in part be paid into faculty funds. MacKie-Mason and Pfau (1999) offers an extensive discussion on advocacy issues, motives and pitfalls of the expert witness. The authors point out that since expert witnesses are repeat players there is a market mechanism that induces professionals to guard their
See also Meier (1986) and Mandel (1999). The latter discusses “the booming market for expert testimony” and calls for ethical standards to regulate it that include full disclosure by academic economists on the cases in which they are or have been involved, and a protocol to avoid conflicts with academic objectivity.
reputation for professional honesty and integrity, as these credentials – more so possibly than winning the case – generate future work. Whilst professional interests and collegial reputation may further help to encourage objectivity, experts should watch out for ego, obliga tion to the client, a deceptive team spirit and the risk of getting locked into position.
In MacKie-Mason and Pfau (1999), various pieces of advice to steer clear of these problems are given, including tips on how to choose a case and research it wisely. The authors recommend to provide lots of quantitative evidence, tie testimony to a publication commitment, and be conservative in drawing conclusions in the sense of trying to “err on the low side” (op.cit., p.218). In all of this, an open relationship with the attorneys is important to be able to discuss reservations and doubts. First and foremost, however, the authors advise due diligence in deciding to accept a case. They developed a two-stage approach to case selection based on the difference between testifying and non-testifying experts. 51 In the first stage, it is agreed that initial research is done as a consulting expert, and not yet as a testifying expert, at a separate fee. Only on the outcome of that research does the expert decide to testify in the case or not. This approach leaves an exit option and no up front commitment. The client benefits as well, in that the initial report and the materials shown to the non-testifying expert are not discoverable by the other parties and can be suppressed if they do not support the case.
For practicing lawyers, working with expert witnesses has become a speciality area in itself. Handbooks and consultants advise on trial techniques for how to select and present expert witnesses and prepare them to explain economics to jurors and judges. The expert can expect close scrutiny of his or her testimony and needs special skills and training to prepare for intense cross-examination. For that purpose, court sessions are sometimes extensively rehearsed. The detailed accounts in Slottje (1999) by Dennis Carlton and Hal Sider on Toys R Us, and by Lawrence White on the FTC’s challenge of the 1986 merger between Coca-Cola and Dr Pepper contain various
See Posner (1999), p.92. The requirements on discoverability of information shown to non-testifying expert are lower than to testifying experts. As a result, a division of labour has developed between testifying expert witnesses and their associates who help prepare the testimony. The latter are sometimes given a great deal of independency to avoid that irrelevant information or intermediate draft reports are discoverable by the other party. This, in turn, had led to case law in extreme cases on the question of who effectively is giving testimony. See also Keyte (2003).
details on their preparation for trial and deposition. The aspiring expert witness can furthermore consult Nolo’s Deposition Handbook or Smith and Bace’s 2002 Guide to Forensic Testimony: The Art and Practice of Presenting Testimony as an Expert Technical Witness. The latter draws extensively on U.S. v. Microsoft Corporation, and critically analyzes the video recordings of the deposition of Bill Gates, which the government released in April 1999.
This extensive industry of specialised and high-paid employment raises questions on the economic efficiency of the system. Apart from the public good of just law enforcement and fair compensation, a cost-benefit analysis of the competition law enforcement process is essentially a comparison between legal transaction costs and deterrence benefits. A key factor in this is the quality of decision making. In antitrust regimes that are open to erroneous judgements in the form of either false positives or false negatives, the incidence of competition law violations may be higher in anticipation, as shown in Schinkel and Tuinstra (2006). An improved quality of competition law decision making therefore has an efficiency effect as well. Competition and reputation help to sustain neutrality and a high quality of work. 52 To stimulate both, several reviewers of the expert economic witness system have proposed further disclosure of expert witness assignments. Posner (1999) suggested that the American Economic Association could maintain a public file of the involvements of its members in antitrust cases, including electronic links to the testimony. Such improved transparency would help to maintain high professional standards and thereby guard the reputation of the profession as a whole. In addition, Posner and others have urged the courts to make greater use of their ability to appoint their own independent experts. They could help decide in cases in which reasonable economic expert witnesses disagree in their findings. 53
Interestingly enough, in the more traditional forensic disciplines, models of competition are studied as a means to improve quality. Such competitive self-regulation for police forensics and forensic labs is discussed in Koppl (2005). 53 Scherer (1999) and Baker and Bresnahan (2007). In Australian courts, it is possible to have all experts on the stand at once and questioning each other in the presence of the judge. A version of this ‘hot-tub’ model is a meeting of the expert witnesses of opposing sides prior to the trial, in which they are to agree on their disagreements before giving testimony before the court. This latter version was used, allegedly quite effectively, in court proceedings following claims by Seven Network, an Australian television network, of anticompetitive conduct against a number of competing network. The Federal Court in Sydney dismissed these claims end of July 2007.
Most of the above accounts and best practice rules relate to the US antitrust practice. As remarked already, the European process of competition law enforcement can be less transparent than the US court-based model. In his contribution to the issue, Andrew Gavil relates the American experiences with the use of economic expert witnesses and the Daubert-rule to lessons for European enforcement. Gavil practises US antitrust law and teaches at Howard University School of Law. In his paper, he examines the preference expressed in the European Commission’s Green Paper on Damages Actions for court-appointed experts over party-retained testifying experts.54 Gavil questions some of the assumptions that underlie that preference and posits that reliance solely on court-appointed experts may be insufficient to realize the Green Paper's principal objective of promoting private enforcement of competition laws. He concludes that additional procedural tools to facilitate the development and presentation of economic evidence, including party-retained experts, may be necessary.
Currently, there does not seem to exist an association or a common code of ethics for expert economic witnesses to the European courts or agencies. Nevertheless, the role of academic economists in the EU competition law is growing, as the revolving-door position of Chief Competition Economist at the European Commission exemplifies. Other functions, such as those of the Economic Advisory Group on Competition Policy, also seem to have similarities to non-testifying experts’ work. The members of this group are all leading academics, who publish reports on competition issues in subgroups that influence decisions in a major way. 55 The national competition authorities in Europe increasingly add (part-time and/or temporary) academic economists to their advisory boards and staff. In addition, there is a large pool of competition consultants available to assist the parties. Furthermore, various policy makers have argued for court-appointed experts in recent contributions and an increasing number of judges specialize in competition matters and receive economic training and advise. 56 The role and quality of forensic IO in Europe is therefore likely to increase in the decades to come.
Commission of the European Communities (2005). See http://ec.europa.eu/dgs/competition/eagcp.htm. 56 Fingleton (2005) and Neven (2006).
5. Remedies, Sanctions and Damages
The fourth aspect of the process of competition law enforcement in which forensic IO can assist is in determining appropriate remedies in cases in which an antitrust concern or an infringement has been established. Parties found to have breached the competition laws face a number of possible consequences, including interventions in the structure of the firm, disciplinary actions against management, and obligations to pay corporate fines and private damages. In merger control, competition concerns with the agencies can be overcome when the merging parties divest sufficiently large parts of their business to competitors or new entrants in the market.
The principles that underlie punishment relate primarily to concepts such as fairness, vindication and compensation, which are in the domain of law, more so than of economics. Yet in antitrust, the law and economics approach of sanctioning in order to deter and prevent lawbreaking has had a strong influence on the way in which competition authorities have designed their sentencing guidelines and other remedies. In addition, economics can inform how to remediate harm and repair competition. Therefore, industrial organization economics plays an important role in advising on effective remedies. The same kind of analyses that helped to identify competition problems can assist in stopping and further avoiding them. Forensic IO therefore has an important preventative role.
Economists play an essential role in devising strategies for restoring competition, especially in cases of exclusionary conduct. In some cases, the obstacle to workable competition is structural and can be removed. An early case in which this was believed to be possible is US v. AT&T, which led to the ordained break-up of the company in the early 1980’s. 57 The structural intervention aimed at reducing the company’s ability to monopolize the telephone industry by erecting artificial barriers to entry and fighting entrants out of the market through predatory pricing. 58 Many smaller and independent telephone companies would compete more forcefully, reducing prices and stimulating innovation.
United States v. American Telephone & Telegraph Co., 552 F.Supp. 131 . See Evans (1983).
Similarly and more recently, the power of Microsoft Corporation to monopolize the platform market has been argued to relate to the company’s exclusive intellectual property right on its popular operating system and network software, refusing to disclose the source code to competitors. This makes it difficult for rivals to produce a viable alternative to Microsoft’s products that is compatible with the installed base of application software and can induce Microsoft clients to switch. As a result, competition is stifled and innovation with it. 59 The original proposals to split up Microsoft in a systems and an application software company were in part inspired by the idea that this would force the systems company to reveal all application interfaces. On the same logic, the European Commission has ordained Microsoft in its official decision to untie some of its software bundles and make its source code information publicly available to other software writers. 60
Structural remedy negotiations are more common in merger control. As noted in Section 3, Phase II merger inquiries involve extensive economic analyses. They are often cleared only with considerable structural remedies. Large merger investigations include advanced simulation models that combine oligopoly theory and econometric estimations. The models help ex ante assessments of the likely effects of the proposed mergers. As a result – and with all the caveats pointed out earlier – they can also advise on structural remedies on which the merger can conditionally be cleared, in particular the divestment of parts of the merging firms’ capacity. The detailed numerical scenarios that these merger models allow for can support divestiture negotiations between the agencies and the merging parties. The result is essentially a design of a post- merger market structure in which the competition concerns raised by the merger are reduced to a sufficient extent.
In the case of hard-core cartel violations or flagrant abuses of dominance, the law specifies what should be the appropriate sanction. For corporate fines, however, both US and European agencies have in recent years committed themselves in guidelines to
See Fisher (2000). See Evans et al. (2000).
punishing antitrust fines. 61 The calculation methods implied by these guidelines relate the effective fine to the volume of sales that was affected by the antitrust violation, with is further fine-tuned on the basis of several multiplicative factors that relate to the gravity of the infringement and mitigating circumstances. While crude and imperfect, this is a reasonable approach to setting fines with a bite – although more so for cartels than for abuses of dominance. 62 Forensic IO techniques can be used to estimate the appropriate base amount of the fine, which is not obvious in complex organizations.
In addition, enforcement regimes increasingly dispose of possibilities for criminal sanctions against individuals responsible for anticompetitive acts. 63 These sanctions can be more effective than punishing the company when they are targeted at the right decision makers within the organizations. Organizational economics in combination with more traditional forensic techniques discussed above in the context of cartel audits can have a role in this, laying out the de facto corporate governance structures. Similarly, economics can advise on the design of incentive compatible corporate compliance programs that can help prevent antitrust violations. Such programs can only be credible if combined with managerial incentive scheme, such as bonuses and stock option plans, that have safeguards against individuals seeking illegal profits. These are largely unexplored areas of corporate governance theory, which are only recently attracting attention as a result of several recent corporate frauds. 64 These issues have direct application in antitrust and, for example in combination with private antitrust detection, hold a promise for increasing deterrence.
Finally, forensic IO techniques are often used in the context of antitrust damages actions. This is the area in which members of the NAFE are most active, as discussed above. Part of the debate is on methods for quantifying damages and the design of an effective private enforcement practice. The latter includes questions such as whether
See, respectively, The United States Sentencing Guidelines, last amendments effective November 1, 2007 at http://www.ussc.gov/GUIDELIN.HTM., and Commission of the European Communities, “Guidelines on the method of setting fines imposed pursuant to Article 23(2)(a) of Regulation No 1/2003,” Official Journal of the European Union, 1.9.2006, C 210/2. 62 See Bos and Schinkel (2006) on the implied method for calculating fines of the European Commission. 63 See Cseres et al. (2006). 64 See Denis et al. (2006).
or not class actions are possible and in what form, whether the pass-on defence should be available, or whe ther punitive damages are an appropriate mechanism. 65 For the most part, however, the quantification of antitrust damages is involved practical work in preparing calculations for court proceedings or to support settlement negotiations. 66
An essential part of assessing antitrust damages is to determine what would have been the market situation absent of the anticompetitive acts – sometimes referred to as the ‘but-for’ world. The most basic aspect of this question is what would have been the price level if the industry would have been competitive. More complicated issues are what kind of industry dynamics – entry, exit and acquisitions – would have unfolded if not for the antitrust violation, what product quality and safety improvements would have been generated or what forms of innovation flourished. To reconstruct a but- for world with a sufficient degree of reliability requires an economic model of the market, in combination with a thorough econometric analysis of the relevant data – in the butfor price question, usually time-series of prices. With the qualification and quantification of the but- for world based on economic reasoning, the subsequent determination of the antitrust damage total is essentially an accounting exercise, but one that can turn on complex issues of discounting and probability calculus.
In their contributions to this issue, Connor, Fisher and Gavil cover several aspects of private antitrust damages actions. Connor surveys various methods to assess damages, including the before-and-after- method, the yardstick- method and cost-based approached. He draws amongst other cases on the Vitamins litigation. Fisher explains the case development and litigation aspects of the Visa Check/MasterMoney antitrust litigation, a private damage case in which he also was responsible for assessing the damage total that led to a landmark out-of-court settlement award. 67 Andrew Gavil discusses elements for the design of an effective private enforcement regime in Europe, where the practice is in its infancy and in need of further development.
6. Forensic IO as a Discipline: Concluding Remarks
See Rüggeberg and Schinkel (2006). See Hall and Lazear (1994). 67 See http://www.inrevisacheckmastermoneyantitrustlitigation.com.
The application of industrial organization economics to competition law enforcement is sufficiently multifaceted to be a speciality discipline of its own. The market for expert economic advise on competition issues and complementary forensic IO skills is highly specialized. Often the skills and knowledge necessary to address all issues in even a single antitrust case require the assistance of multiple experts. This necessitates considerable coordination and project management work. The market for competition advice is large and sustains an oligopoly of just a few advanced consulting firms. They employ economists trained in IO theory and econometrics at the best schools, highly specialized industry expertise and a broad experience. The various sides in competition law disputes, interested parties as well as the antitrust agencies, rely on the economic consulting firms for assistance. These firms produce institutional reports, but importantly also they form a platform for independent academic economists, who work with them on selected cases and on a freelance basis. 68 For reasons of identifiable reputation, the testifying individual expert witness remains central – although still much more so in the US than in Europe – and the consulting firms often act as non-testifying experts or sophisticated back-offices in the preparation of their eventual testimony.
Universities can supply the discipline with graduates trained in the relevant economic disciplines. Most universities offer theory and econometrics courses. In addition, a few have specialized courses in forensic economics. Slenick and Tinari (2001), in the Journal of Forensic Economics, survey forensic economics taught in university curricula. They report very few organized programs on the topic – mainly their own – and a just a few isolated individual courses, or elements of forensic economics introduced in traditional economics classes. The emphasis in the curriculum elements the authors found is firmly on damages assessment in tort litigation. An exception, which their study did not include, is a semester course ‘Forensic Economics’ that John Connor teaches at Purdue University, which is entirely on antitrust.
Antitrust agencies have forensic IO skills available at various levels within their organizations. Cartel and merger units each have their own specialists. As said, more and more competition authorities in Europe follow the US example of having a chief
See Short and Sattler (1996).
economist office. However, there presently does not seem to exist a public institution or other form of centralized organization of forensic IO. At best, agencies have a forensic IT unit. The more traditional forensic sciences, however, are organized in national institutes for the forensic sciences. 69 These governmental laboratories are typically part of the national departments of justice. They maintain close relationships with academia, having prominent scientific boards with renowned scientists from various disciplines. University professors closely cooperate with the institutes, are called upon to assist in particular investigations and teach in specialized masters programs on forensic science. 70 The institutes publish best-practice manuals, accredit independent experts and organize conferences, expert working groups and other platforms to advance the discipline.
It appears that the practice of forensic IO could benefit from a similar type of organization, for example on a European level. The agencies and the courts could rely on such an institute of society for expert advice in complex competition cases directly. It could further be a reference source to reputable specialists. If such an entity was set up as an economics division of existing national forensic institutions, there are obvious synergies across the sciences. The ideal team of forensic skills, including forensic IT and data analysts, could be put together around forensic IO specialists to solve complicated competition cases.
The traditional forensic sciences seemingly owe much of their popularity to suspense novels, drama series and television documentaries. Although it is likely that capital crime will remain more appealing to a broader audience, antitrust acts can be just as shocking and suspenseful. In fact, several internal cartel cases have inspired bestselling courtroom/law suspense novels. 71 The footage of the Lysine cartel members meeting in hotel rooms, which the FBI managed to obtain with cameras hidden in table lamps, shows just how exciting white collar antitrust crimes are. Lysine, just like the latest cartel decision of the European Commission in Dutch Beers, contain enough
Examples are the US Federal Bureau of Investigation (FBI), the National Institute of Forensic Science in Australia, and the Forensic Science Service in England. See European Network of Forensic Science Institutes 70 Key societies are the American Academy of Forensic Sciences (http://www.aafs.org) and the European Academy of Forensic Science (http://www.enfsi.org). 71 See Eichenwald (2001) and Mason (2004).
thriller and script for several episodes of Forensic Detectives.72 To attract the attention of the public in that way would not only make for effective marketing for the specialty discipline, but could also raise awareness of the serious harm that is often caused by competition law violations.
Abrantes-Metz, R.M., L.M. Froeb, J. Geweke and C.T. Taylor, “A Variance Screen for Collusion,” International Journal of Industrial Organization, 24, 2006, 467-486
Adams, W. and J.W. Brock, Antitrust Economics on Trial: A Dialogue on the New Laissez-Faire, Princeton University Press: Princeton, 1991
Adelstein, F. (2006), “Live Forensics: Diagnosing Your System Without Killing it First,” Communications of the ACM, 49 (2), 63-66
Alexander, C.R. and M.A. Cohen, “Why do Corporations Become Criminals? Ownership, Hidden Actions and Crime as an Agency Cost,” Journal of Corporate Finance, 5, 1999, 1-34
Anastasi, J., The New Forensics: Investigating Corporate Fraud and the Theft of Intellectual Property, John Wiley & Sons: Hoboken, 2003
Bajari, P.L. and G. Summers, “Detecting Collusion in Procurement Auctions,” Antitrust Law Journal, 70, 2002, 143-170
Baker, J.B., “Econometric Analysis in FTC v. Staples,” Journal of Public Policy & Marketing, 18, No. 1, 1999
Baker, J.B., “Economists’ Roundtable (moderator),” in: Garza, D.A. (ed.), “From Theory to Practice: Working with Economic Experts,” Antitrust, 17(2), Spring 2003
See Griffin (2000) and Dutch beer market, Case No COMP/B/37.766, Commission Decision of 18 April 2007.
Baker, J.B., “The Case for Antitrust Enforcement,” Journal of Economic Perspectives, 17(4), 2003, 27-50
Baker, J.B. and T.F. Bresnahan, “Empirical Methods for Identifying and Measuring Market Power, Antitrust Law Journal, 61, 1992, 3- 16
Baker, J.B. and D.L. Rubinfeld, “Empirical Methods in Antitrust Litigation: Review and Critique, American Law & Economic Review, 1, 1999, 386-435
Baker, J.B. and T.F. Bresnahan, “Economic Evidence in Antitrust: Defining Markets and Measuring Market Power,” in: P. Buccirossi (Ed.), Handbook of Antitrust Economics, MIT Press, Cambridge, MA, 2007
Barnett, T.O., “Seven Steps to Better Cartel Enforcement,” in: C-D Ehlermann and I. Atanasiu (eds.), European Competition Law Annual 2006: Enforcement of the Prohibition of Cartels, Hart Publishing, Oxford, forthcoming 2007;
Basmann, R.L., “Introduction to Statistics and Econometrics in Litigation Support,” Journal of Econometrics, 113(1), 1-2, 2003
Baumol, W.J., “Research Pure and Impure: Their Economies of Scope,” in: Slottje, D.J. (ed.), The Role of the Academic Economist in Litigation Support, Elsevier Science, Amsterdam, 1999
Beckenstein, A.R., H.L. Gabel and K. Roberts, “An Executive’s Guide to Antitrust Compliance,” Harvard Business Review, September/October 1983
Bergman, P. and A. Moore, Nolo’s Deposition Handbook, Nolo: Berkeley, 1999
Bernstein, D.E., “Junk Science in the United States and the Common Wealth,” Yale Journal of International Law, 21, 1996, 124-182
Bishop, S. and M. Walker, The Economics of EC Competition Law: Concepts, Application and Measurement, second edition, Sweet & Maxwell, London, 2002 32
Borrell, J.-R., “Fashioning Antitrust Regimes: One Legal Rule Does Not Fit All,” working paper, 2006
Bos, I. and M.P. Schinkel, “On the Scope for the European Commission's 2006 Fining Guidelines under the Legal Maximum Fine,” Journal of Competition Law and Economics, 2(4), 2006, 673-682
Boudin, M., “Forensic Economics,” Harvard Law Review, 97, 1984, 835-844
Buccirossi, P and G. Spagno lo, “Corporate Governance and Collusive Behavior,” in: D.W. Collins (ed.), Issues in Competition Law and Policy, 2006, Antitrust Section, American Bar Association
Carlton, D.W. and J.M. Perloff, Modern Industrial Organization, fourth edition, Addison-Wesley, New York, 2006 Chalmers, A.F., What's This Thing Called Science?, Open University Press, 3rd Edition, 2004
Christie, W.G. and P.H. Schultz, “Why Do NASDAQ Market Makers Avoid OddEight Quotes?,” Journal of Finance, 49(5), 1994, 1813-1840
Commission of the European Communities, Green Paper on Damages Actions for Breach of the EC Antitrust Rules, COM(2005) 672 final, 19 Dec 2005
Connor, J.M., “’Our Customers Are Our Enemies’: The Lysine Cartel of 1992-1995,” Review of Industrial Organization, 18, 2001, 5-21
Connor, J.M., Forensic Economics, Course Outline, August 2003
Connor, J.M., Global Price Fixing, second edition, Springer, New York, 2007
Corts, K.S., “On the Competitive Effects of Price-Matching Policies,” International Journal of Industrial Organization, 15, 1997, 283-299
Cseres, K.J., M.P. Schinkel and F.O.W. Vogelaar (eds.), Criminalization of Competition Law Enforcement: Economic and Legal Implications for the EU Member States, Edward Elgar, London, 2006
Denis, D.J., P. Hanouna and A. Sarin, “Is There a Dark Side to Incentive Compensation?,” Journal of Corporate Finance, 12 (3), 2006, 467-488
Easterbrook, F.H., “The Limits of Antitrust,” Texas Law Review, 63, 1984, 1-40
Eichenwald, K., The Informant, Broadway Books: New York, 2001
Einhorn, H., The Success of Economic Defenses in Price Fixing Cases: An Analysis, Journal of Forensic Economics, 2 (1), 1989, 89-99
Evans, D.S. (ed.), Breaking Up Bell: Essays on Industrial Organization and Regulation, North-Holland, Amsterdam, 1983
Evans, D.S., F.M. Fisher, D.L. Rubinfeld and R.L. Schmalensee (eds.), Did Microsoft Harm Consumers? Two opposing Views, AEI-Brookings Joint Center for Regulatory Studies, AEI Press, Washington, DC, 2000
Fingleton, J., “The Application of Economics in Court,” speech delivered at the Joint IBA and European Commission Conference, Brussels, 11 March 2005
Fisher, F.M., J.J. McGowan and J.E. Greenwood, Folded, Spindled, and Mutilated, MIT Press, Cambridge, MA, 1983
Fisher, F.M., “The IBM and Microsoft Cases: What's the Difference?,” American Economic Review, 90(2), 2000, 180-183
Fisher, F.M. and D.L. Rubinfeld, “U.S. v. Microsoft – An Economic Analysis,” Antitrust Bulletin, 46(1), 2001, 1-69
Garoupa, N., “Corporate Criminal Law and Organization Incentives: A Managerial Perspective,” Managerial and Decision Economics, 21, 2000, 243-252
Gavil, A.I., “Defining Reliable Forensic Economics in the Post Daubert-Kumho Tire Era: Case Studies from Antitrust,” Washington and Lee Law Review, 57, 2000, 569601
Ghosal, V. and J. Stennek, The Political Economy of Antitrust, Elsevier, Amsterdam, 2007
Goppelsroeder, M. and M.P. Schinkel, “On the Use of Economic Modeling in Merger Control” in: Bergeijk, P.A.G. van and E. Kloosterhuis (eds.), Modeling European Mergers: Theory, Competition Policy and Case Studies, Edward Elgar, Cheltenham, 2005
Green, S.P., Lying, Cheating, and Stealing: A Moral Theory of White-Collar Crime, Oxford University Press, Oxford, 2006
Griffin, J.M., “An Inside Look at a Cartel at Work: Common Characteristics of International Cartels,” American Bar Associations Section of Antitrust Law, April 6, 2000
Grout, P.A. and S. Sonderegger, “Structural Approaches to Cartel Detection,” in: C-D Ehlermann and I. Atanasiu (eds.), European Competition Law Annual 2006: Enforcement of the Prohibition of Cartels, Hart Publishing, Oxford, forthcoming 2007
Hall, R.E. and V.A. Lazear, “Reference Guide on Estimation of Economic Losses in Damages Awards,” in: Reference Manual on Scientific Evidence, Federal Judicial Center, Washington D.C., 1994
Hands, D.W., Reflection without Rules: Economic Methodology and Contemporary Science Theory, Cambridge University Press, Cambridge, 2001
Harrington, J.E., “Detecting Cartels,” in: P. Buccirossi, Advances in the Economics of Competition Law, MIT Press, Cambridge, MA., forthcoming in 2007a
Harrington, J.E., “Behavioral Screening and the Detection of Cartels,” in: C-D Ehlermann and I. Atanasiu (eds.), European Competition Law Annual 2006: Enforcement of the Prohibition of Cartels, Hart Pub lishing, Oxford, forthcoming 2007b
Ireland, T.R., “The Interface between Law and Economics and Forensic Economics,” Journal of Legal Economics, Spring/Summer 1997a, 60-70
Ireland, T.R., The Daubert Decision And Forensic Economics: An Overview, Journal of Forensic Economics, 10(2), 1997b, 121-126
Ivaldi, M. and F. Verboven, “Quantifying the Effects from Horizontal Mergers in European Competition Policy,” International Journal of Industrial Organization, 23, 2005, 669-691
Jamieson, K.M., The Organization of Corporate Crime; Dynamics of Antitrust Violations, Sage Publications, Thousand Oaks, 1994
Johnson, J.H., “Economic Evidence in Criminal Cartel Cases,” in: L. Wu (ed.), Economics of Antitrust: Complex Issues in a Dynamic Economy, NERA, 2007
Joshua, J. and C. Harding, “Breaking Up the Hard Core: The Prospects for the Proposed Cartel Offence,” Criminal Law Review 933, 2002
Kaplow, L., and C. Shapiro, “Antitrust,” in: A.M. Polinsky and S. Shavell (eds.), Handbook of Law and Economics, Elsevier, 2007
Kaufman, R.T., Rodgers, J.D., Martin, G.D., Economic Foundations of Injury and Death Damages, Edward Elgar: Cheltenham, 2005
Keyte, J.A., “A Risk-averse Guide for Working with Non-testifying Consultants or Experts,” in: Garza, D.A. (ed.), “From Theory to Practice: Working with Economic Experts,” Antitrust, 17(2), Spring 2003
Koppl, R., “How to Improve Forensic Science,” European Journal of Law and Economics, 20(3), 2005, 255-286
Levenstein, M.C. and S.W. Salant, Cartels, Volume I and II, Edward Elgar, Northhampton, 2007
Lopatka, J.E. and W.H. Page, “Economic Authority and the Limits of Expertise in Antitrust Cases,” Cornell Law Review, 90, 2005, 617-703
Lyle, D.P., Forensics for Dummies, Wiley Publishing, Indianapolis, 2004
MacKie-Mason, J.K. and R.A. Pfa u, “Inducements to Advocacy: The Economist as Independent Expert,” in: Slottje, D.J. (ed.), The Role of the Academic Economist in Litigation Support, Elsevier Science, Amsterdam, 1999
Mandel, M.J., “Going for the Gold: Economists as Expert Witnesses,” Journal of Economic Perspectives, 13(2), 1999, 113-120
Mason, C., Art of the Steal: Inside the Sotheby's-Christie’s Auction House Scandal, Putnam Publishing Group, 2004
Milne, R.A., Pace III, J.E., “Conspiratologists at the Gate: The Scope of Expert Testimo ny on the Subject of Conspiracy in a Sherman Act Case,” in: D.A. Garza (ed.), “From Theory to Practice: Working with Economic Experts,” Antitrust, 17(2), Spring 2003
Meier, Paul, “Damned Liars and Expert Witnesses,” Journal of the American Statistical Association, 81, 394, 1986, 269-276
Motta, M., Competition Policy: Theory and Practice, Cambridge University Press, Cambridge, 2004
National Association of Forensic Economics, “The Statement of Ethical Principles and Principles of Professional Practice,” Journal of Forensic Economics
Neven, D.J., “Competition Economics and Antitrust in Europe,” Economic Policy, October 2006
Patterson, D.E. and C. Shapiro, “Transatlantic Divergence in GE/Honeywell: Causes and Lessons,” Antitrust, Fall 2001, 18-26
Paul, R.D., “International Cartels in Crosshairs: Aggressive Enforcement Activity Presents Serious Risks for Corporations and Individuals,” New York Law Journal, September 11, 2000
Piette, Michael J., “Codes of Professional Ethics for Forensic Economists: Problems and Prospects,” Journal of Forensic Economics, 1991, 4(3), 269-276
Porter, R.H., “Detecting Collusion,” Review of Industrial Organization, 26, 2005, 147-167
Porter, R.H. and J.D. Zona, “Detection of Bid Rigging in Procurement Auctions,” Journal of Political Economy, 2001, 518-538
Posner, R.A., “The Law and Economics of the Economist Expert Witness,” Journal of Economic Perspectives, 13 (2), 1999, 91-99
Rey, P., “On the Use of Economic Analysis for Cartel Detection,” in: C-D Ehlermann and I. Atanasiu (eds.), European Competition Law Annual 2006: Enforcement of the Prohibition of Cartels, Hart Publishing, Oxford, forthcoming 2007 38
Rochet, J-C and J. Tirole, “Two-sided Markets: A Progress Report,” RAND Journal of Economics, 37(3), 2006, 645-668
Roeller, L-H, and O. Stehmann, “The Year 2005 at DG Competition: The Trend towards a More Effects-Based Approach,” Review of Industrial Organization, 29, 2006, 281–304
Rueggeberg, J. and M.P. Schinkel, “Consolidating Antitrust Damages in Europe: A Proposal for Standing in Line with Efficient Private Enforcement,” World Competition 29(3): 395-420, 2006
Salop, S., “Practices that (credibly) facilitate oligopoly coordination,” in: J. Stiglitz and F. Mathewson (eds.), New Developments in the Analysis of Market Structure, MIT Press, Cambridge, 1986
Saukko, P.J. and G.C. Knupfer, Encyclopaedia of Forensic Sciences, Vol. 1-3, Academic Press, San Diego, 2000
Scherer, F.M., “Economic Consulting Fire Fighting and Similar Adventures,” in: Slottje, D.J. (ed.), The Role of the Academic Economist in Litigation Support, Elsevier Science, Amsterdam, 1999
Schinkel, M.P. and J. Tuinstra, “Imperfect Competition Law Enforcement,” International Journal of Industrial Organization, 24, 2006, 1267-1297
Shy, O., The Economics of Network Industries, Cambridge University Press, Cambridge, 2001
Simpson, S.L., “Strategy, Structure and Corporate Crime: The Historical Context of Anticompetitive Behavior,” in: F. Adler and W.S. Laufer, New Directions on Criminological Theory, Advances in Criminological Theory, Volume 6, New Brunswick, 1993, 71-93
Slesnick, F. and F.D. Tinari, “Teaching Forensic Economics in the University Curriculum,” Journal of Forensic Economics, 14(3), 2001, 243-260
Slottje, D.J. (ed.), The Role of the Academic Economist in Litigation Support, Elsevier Science, Amsterdam, 1999
Smith, F. and R. Bace, Guide to Forensic Testimony, The Art and Practice of Presenting Testimony As An Expert Technical Witness, Pearson Education, 2002
Short, D.M. and E.L. Sattler, “The Market for Expert Witnesses,” Journal of Economics, 22, 1996, 87-89
Solomon, S.M., Hackett, E.J., “Setting Boundaries between Science and Law: Lessons from Daubert v. Merrell Dow Pharmaceuticals,” Science, Technology & Human Values, 21 (2), 1996, 131-156
Spagnolo G., “Managerial Incentives and Collusive Behavior”, European Economic Review, 49(6), 2005, pages 1501-1523
Tirole, J., The Theory of Industrial Organization, MIT Press, Cambridge, MA, 1988
Thornton, R. and J. Ward, “The Economist in Tort Litigation,” The Journal of Economic Perspectives, 13 (2), 1999, 101-112
Tomlim, J.T., “Distinguishing the Illegal from the Legal in Antitrust Damages Calculations: Lessons from Netscape v. Microsoft,” Journal of Forensic Economics,17(2), 2004, 223-239
U.S. Department of Justice, “Price Fixing, Bid Rigging, and Market Allocation Schemes: What They Are and What to Look For: An Antitrust Primer,” available at: http://www.usdoj.gov/atr/public/guidelines/211578.htm
Vernarelli, M.J., “Book Review of Daniel J. Slottje, The Role of the Academic Economist in Litigation Support,” Journal of Forensic Economics, 15(1), 2002, 101103
Vesterdorf, B., “Standard of Proof in Merger Cases: Reflections in the Light of Recent Case Law of the Community Courts,” European Competition Journal, 1(1), 2005
Vickers, J., “Competition Law and Economics: A Mid-Atlantic Viewpoint,” The Burrell Lecture of the Competition Law Association, given in London, 19 March 2007
Volonino, L., “Electronic Evidence and Computer Forensics,” Communications of AIS, 12 (27), 2003
Walker, M., “The Potential for Significant Inaccuracies in Merger Simulation Models,” Journal of Competition Law and Economics 1(3), 2005, 473–496
Ward, J.O., Olson, G.W., “Forensic Economics: A Perspective and An Agenda for Research,” Journal of Forensic Economics, 1 (1), 1987, 1-10
Werden,G.J., L.M. Froeb, D.T. Scheffman, “A Daubert Discipline for Merger Simulation,” Antitrust, Summer 2004
Werden G.J., “Economic Evidence on the Existence of Collusion: Reconciling Antitrust Law with Oligopoly Theory,” Antitrust Law Journal, 3, 2004
Werden, G.J., “The Admissibility of Expert Economic Testimony in Antitrust Cases,” ABA Section of Antitrust Law, Issues in Competition Law and Policy, forthcoming 2007
Williamson, O.E., “Economies as an Antitrust Defense: The Welfare Tradeoffs,” American Economic Review, 58, 1968, 18-36
Whish, R., “Oligopoly Theory and Economic Evidence,” in: C-D Ehlermann and I. Atanasiu (eds.), European Competition Law Annual 2006: Enforcement of the Prohibition of Cartels, Hart Publishing, Oxford, forthcoming 2007
This action might not be possible to undo. Are you sure you want to continue?
We've moved you to where you read on your other device.
Get the full title to continue reading from where you left off, or restart the preview.