Here is a Sunday afternoon conjecture about a common divide between Americans and Europeans when it comes to the emotional appeal of free markets. The Continental enlightenment was very much driven by philosophical rationalism, which in the administrative realm found its expression in the belief that only that which is planned is rational (Prussia, France). The English enlightenment, in contrast, had a more empirical bent from the get-go with an emphasis not merely on rational outcomes but also on rational processes (e.g., what would later become the scientific method). A culture shaped by the English enlightenment may thus be instinctively more comfortable with the idea of a rational yet unplanned market economy than more strictly rationalist traditions.

Share on Facebook

I am very much looking forward to the ABA 2009 Antitrust Intellectual Property Conference in my new hometown Berkeley, where I will be on a panel about different approaches toward “forced access” to IP in the US and in the EU. My remarks will focus on the exciting contributions that infrastructure theory has made to our understanding of what properly constitutes an essential facility.

Here are my slides. Happy New Year!

UPDATE: Here’s the link to the registration page.

Share on Facebook

Ever since Socrates demolished Gorgias to lasting effect, the universalism vs. relativism debate has had a prominent place in moral theory, in law, and (with some delay) in the natural sciences, too. In stark contrast, much of modern price theory proceeds in a strictly Platonic fashion. The rational actor is an abstraction liberated from the contraints of time and place, so is the firm and the idea of the market itself. The perfectly competitive market with its infinite number of atomistic buyers and sellers, each reacting to aggregated forces outside of their control but to which they each unwittingly contribute, trading perfectly homogenous goods in exchange for an ideal currency with constant marginal value, exists in a parallel universe of ideal forms. We observe its various more or less imperfect instantiations in the real world. The closer the real world approximates the ideal, the better. The implication is that basic economic laws are universal, and that there are thus objective normative criteria by which to judge economic systems, irrespective of their time and place. That belief in economics as a process for discovering universal laws of human action, either through mathematical formalization (”nature obeys mathematics”) or through the observation of a quasi-evolutionary process in which those real-world economies survive that better approximate the ideal, is a driving intellectual force behind the neo-conservative faith in the power of free markets, unlocking human potential everywhere.

The real world, however, does not reflect this hopeful (or plainly ideological) Platonism. As Max Weber pointed out, economic institutions and the theories reflecting them, are largely the product of political history, religion, and geography. For example, French and to a somewhat lesser extent German competition law still clearly reflect the administrative tradition, which has a longer history in centralized France than in decentralized Germany, where an expert government bureaucracy ensures that capitalistic enterprise remains alinged with the public good. Given the a priori embeddedness of the market in the broader body politic, the market never comes into view as a fully separate, autonomous system. As a result, there is no real need for a theory of the market as a self-regulating and self-stabilizing process, which is the defining feature of the perfect competition model. The perfect competition model describes, above all, a stable dynamic, a process moving towards equilibrium without any outside intervention. It is the theoretical justification for laissez faire, for why governments can, in fact, afford to leave businesses (and, not coincidentally, the rich) alone. The perfect competition model is expressly designed to explain why ideal free markets simply cannot spiral out of control.

In France and Germany, the market has always been (at least in part) a tool to be deployed by the state in furtherance of public good (however determined). The market is not an end in itself, as it contains no universal truth about human affairs, or at least no truth that, normatively, the body politic should necessarily encourage or embrace. Rather, the ends of the market process are expressly defined in political, not economic terms, and they include competing values beyond productive efficiency, such as predictable growth, limited inflation, and full employment. (Obviously, having those goals is one thing, achieving them is quite a different matter.) Similarly, competition as a key ingredient of the market process is “managed competition,” as opposed to unlimited competition in a hypothetical natural state. The same is true for property in rivalrous goods, another necessary ingredient of the market process. Property in critical sociatal infrastructure such as transportation, energy, and telecommunication, has never been an entirely private affair, irrespective of who created, planned, or financed the infrastructure, which is reflected both in the constitutional definitions of property (e.g., in Germany) and, in practice, in the still significant state ownership of such industries.

In other words, economic institutions and the building blocks of economic theory reflect in significant part the political structure of the society of which the economic system is a part. In the history of philosophy, countless Platonisms have turned out to be little more than theoretical universalizations of unquestioned contingent beliefs. Antitrust economics would greatly benefit from a more explicit historic and comparative perspective. Competition law certainly has.

Share on Facebook

One of the undercurrents driving the movement that led to Obama’s landslide victory is that people have finally started to reject en masse a public policy that by and large conceived of them only in the most limited, reductionist, and anemic way possible: as private, rational actors, pursuing primarily their own (usually material) advantage. Yesterday, people asserted their roles as citoyens, as political beings concerned not only with themselves but also with the public good. One of the hallmarks of Obama’s campaign was to ask people to do something, to get involved. Not just to contribute financially, but to reach out, organize house parties, call battleground states, take time off for poll watch, all of which meant doing things together with others, i.e., organizing. The message of change thus implies a belief in the possibility of politically meaningful choice, a notion that has been conspicuously absent for a long time, and which is categorically different from consumer choice. How will this re-assertion of the person over the individual play out in the legal world and the antitrust world in particular? I expect a gradual re-evaluation of how the legal system and economic scholarship relate to each other. The normative principle of law is justice. That of economics is efficiency. Taking rights seriously requires that judges and lawyers adopt an internal point of view and put justice first. Of course, this is not to deny that the legal system should import efficiency considerations here and there. For market regulation and marketplace design rules such as antitrust, efficiency is indeed the right default choice. Fortunately, over a wide range of issues, efficiency and justice are not in conflict with one another. It is, however, meant as a reminder that the legal system is fundamentally independent from economics. Normatively, the legal system belongs to a different domain, and in a democracy the legal system does not take orders from the economic system. The latter may very well constrain or enable actions of the former, but factual obstacles are very much unlike normative constraints.

Share on Facebook

Here is Larry Lessig’s assessment of McCain’s technology policy, which, by the way, was written and titled before McCain’s VP pick. Lessig chronicles how the U.S. has lost ground in terms of broadband infrastructure in the past eight years. Lessig’s summary of the positions in the net neutrality debate is worth quoting in full.

What is network neutrality? Network neutrality is basically the question whether the network owners get to pick and choose the applications and content that you as a user can have access to or use. If you think about these computer networks in the way that you might think about a cable television network, you probably think, “Sure, why shouldn’t they have that right, just like a cable television network has the right to choose what programs will play across their television network.” But if you think about the Internet in the way we traditionally thought about telephone networks, then you would say, “Of course, no way should the network owner have the right to choose with whom I connect or what I say.” That’s essentially the battle brewing around the question of the future of the Internet. Will the Internet increasingly look like cable television networks, where the owners of the network get to pick and choose the applications and content that run on that network? Or will the future of the Internet stay with the tradition of telephones, where it is the user that gets to pick with whom and what he says on that network?

There is an interesting connection between the significant decline the United States has suffered compared to other nations in terms of broadband penetration and net neutrality. Giving routing preferences to packets is necessary only if there is congestion. Historically, we have always dealt with that issue by throwing bandwidth at the problem, not only because it was the right thing to do from a policy and internet architecture standpoint, but also because it was usually less expensive than implementing and administering routing preferences. Why is that not the solution here?

Share on Facebook

This 1944 Simple Sabotage Field Manual from the U.S. Office of Strategic Services (the precursor to the CIA) contains valuable advice on how to run a large-scale organization. Highly recommended (cough).

(HT: BB).

Share on Facebook

Arguments about the merits of increased copyright protection usually proceed in three stages.

1. Pro: “IP is like any other form of property.” Con: “No, it is not, because information is non-rivalrous. The main justification for exclusive property rights, allocating scarce resources, is therefore lacking.”
   
2. Pro: “We need to drive up the price for accessing today’s information in order to create incentives for people to get involved in the production of tomorrow’s information.” Con: “The incentive effect today is likely outweighed by higher input prices tomorrow. Moreover, as culture is a collaborative enterprise, propertization increases the transaction costs. Lastly, a linear increase in fuzzy boundaries results in an exponential increase in litigation costs. The net welfare effect from increased protection is in all likelihood negative.”
   
3. Pro: Be that as it may. The fact remains that in the real world cultural production is IP protected. Books and songs are copyrighted, inventions are patented, etc.
   

I am always a bit puzzled by the “real world” argument, because while undeniably most cultural production is copyrighted, it is not at all clear that the copyright protection is a significant reason for the production. This may well be an endemic case of post hoc ergo propter hoc. For example, most organic search results on Google or Yahoo seem to come from either non-profit actors (Wikipedia, blogs, forums, universities, SSRN, etc.) or from for-profit actors whose core business model does not depend on IP protection (newspapers, consultants, universities, blogs, etc.). The online advertising revolution is certainly a major contributor to that trend, because the advertising-supported business model is the excact opposite of the IP-based paradigm in that it relies on increasing distribution, not limiting it. Not even the software industry is predominantly IP reliant, with almost two times the revenues from services than from software sales. (See, e.g., Software and Information: Driving the Global Knowledge Economy).

Of course, this is not to deny that some business models stand to profit from increased IP protection, they most certainly do, as Herbert Hovenkamp and others have described in detail. The point is rather to question the assumption that most creative production is in fact driven by IP-based business models. The more actual cultural production takes place outside the IP incentives model, the less credible — and justifiable as a matter of constitutional law — the call for increased protection.

Share on Facebook

The permit is one of the oldest business models. First, the government issues some prohibition. Then, for a fee, the prohibition is partially lifted. Usually, the government collects the fee. But in the days of public-private partnership in security matters, some private entity does. CLEAR is one of those business models. After 9/11, the government encumbered the freedom of movement for air travelers, inconveniencing millions in the name of heightened security at airports. Now, a CLEAR pass for $128/year, lets you cut in front of the security line. CLEAR thus monetizes the loss of civil liberties. The beauty of the business model is that it is entirely independent of whether it contributes to security or not. Its sine qua non is the existence of a government restraint on certain freedoms, not whether the restraint serves any useful purpose. CLEAR is admittedly a somewhat extreme example, because it makes no bones about its implicit assumption that everyone who doesn’t have a pass is a potential security risk. It is a harbinger of a world in which we live by permission, not by right, a world in which the default is control, not freedom, a world in which prosecutorial discretion, not the rule of law, protects liberties.

There are other business models that in more subtle ways require legislative restraints on freedoms. For example, prior to the enactment of the DMCA, reverse engineering, the “freedom to tinker,” was part of the set of civil liberties that everyone enjoyed and took for granted. In such a world, every technology is open in the long run. Users are free to break open black boxes and put them to uses that their designers did not intend and could not foresee. Discovering the unintended uses and thus the secret lives of artifacts is a creative act, and the history of art has long recognized it as such. Where the freedom to tinker is unrestrained, businesses spring up to support user demand for unlocking their technological environment. The issue is that in such a high-civil liberties environment, certain business models won’t work, or at least, they won’t work particularly well. Suppose a firm is selling a piece of computer hardware below cost to increase penetration and adoption, in the hope of making money by selling the rights to develop software for the platform to third party developers. This is a now common multi-sided business model, as described in detail by Schmalensee and Evans. The model is based on the premise that developers can’t access the platform without some form of an access payment to the maker of the box. That, of course, is perfectly fine. If a developer wants to use my copyrighted API, then they have to pay me for the privilege. However, another condition with much more significant impliactions is that users won’t hack their equipment, thus opening it up (en masse) for third party development, e.g., by installing a free OS on the device, for which developers could then code applications without access payments to the maker of the box. That second condition requires government intervention in the form of restraints on the rights of the users to unbox technology. It is the need for that particular enabling restraint, which brings us to the civil liberties v. consumer welfare trade-off. Note that the trade-off is, as always, marginal. That is, are we willing to give up some measure of liberties in order to make the environment more hospitable for certain welfare enhancing business models? Choosing between totals is easy, choosing between marginals is hard. Note, moreover, that I am perfectly willing to concede (at least arguendo) that a multi-sided business model as described above may increase consumer welfare by putting more hardware into the hands of people who could not otherwise afford it.

In the end, we are left with the realization that certain business models, crude ones such as CLEAR and more sophisticated ones such as some multi-sided platforms, are based on government restraints on civil liberties. Any complete economic or policy analysis of a practice should take its impact on civil liberties into account. Some trade-offs in favor of consumer welfare may well be worth making. Others are most certainly not. This is a discussion that we should be having — and it is one to which lawyers can make a genuine contribution.

UPDATE: Here’s a related post on TLF.

Share on Facebook

This chart from The Economist shows that for some important market changing products the time from innovation to market penetration has become progressively shorter.

The article explains:

The upshot is that technology is spreading to emerging markets faster than it has ever done anywhere. The World Bank looked at how much time elapsed between the invention of something and its widespread adoption (defined as when 80% of countries that use a technology first report it; see chart 1). For 19th-century technologies the gap was long: 120 years for trains and open-hearth steel furnaces, 100 years for the telephone. For aviation and radio, invented in the early 20th century, the lag was 60 years. But for the PC and CAT scans the gap was around 20 years and for mobile phones just 16. In most countries, most technologies are available in some degree.

One might question whether any use of a technology in a particular country should be sufficient to put it on the map as a country in which the technology is practiced. For example, there are undoubtedly PCs and CAT scans available in Kumpala, even though the technology penetration rate of Uganda is among the lowest in the world. Uneven distribution notwithstanding, Uganda, I assume, would be a country reporting the use of PCs and CAT scans. That said, the chart is certainly plausible for those parts of the world where infrastructure permits the rapid dissemination of technology. The results should be even more striking for electronically distributed products among world-wide internet users (e.g., adoption of PDF as a document standard, mp3 as a music standard, Firefox, etc.)

One implication of more or less instant dissemination is that the arguments for longer copyright and patent periods are further weakened. The shorter the lead time from publication to full scale distribution, the more immediate the financial reward for the creator (or, more accurately, the rights holder), and the greater the total reward over the life of the exclusive right. In other words, even if the exclusivity period remained constant, more immediate and broader distribution result in increased revenues and incentives.

Technorati Tags: economics, innovation, patents

Share on Facebook

If you look at what a lawyer makes (in a particular city), that’s what an escort makes.

SFGate $4,300/hour?

Share on Facebook

While everyone else is waiting in line to check out Kristen’s MySpace page, we at the Antitrust Review keep offering wholesome nerd fare such as Paul Krugman’s 1978 Theory of Interstellar Trade. (HT: /.) The paper

extends interplanetary trade theory to an interstellar setting. It is chiefly concerned with the following question: how should interest charges on goods in transit be computed when the goods travel at close to the speed of light? This is a problem because the time taken in transit will appear less to an observer traveling with the goods than to a stationary observer. A solution is derived from economic theory, and two useless but true theorems are derived. … It should be noted that while the subject of this paper is silly, the analysis actually does make sense. This paper, then, is a serious analysis of a ridiculous subject, which is of course the opposite of what is usual in economics.

The First Fundamental Theorem of Interstellar Trade is:

When trade takes place between two planets in a common inertial frame, the interest costs on goods in transit should be calculated using time measured by clocks in the common frame, and not by clocks in the frames of trading spacecraft.

Indeed. For further elaboration I suggest that the reader turn to Alastair Reynolds Revelation Space universe and to Ken MacLeod’s Engines of Light trilogy. Frank Herbert, on the other hand, cheated his way out of the problem with FTL travel, which Krugman rightly dismisses.

That brings us to the Second Fundamental Theorem of Interstellar Trade:

If sentient beings may hold assets on two planets in the same inertial frame, competition will equalize the interest rates on the two planets.

The upshot is that interstellar trade can be viewed as long-term investments, similar to 17th Century merchant capitalism. Interestingly, Krugman remarks that:

Recent progress in the technology of space travel … raise[s] the distinct possibility that we may eventually discover or construct a world to which orthodox economic theory applies.

But what if the rate of innovation within the inertial frame makes any estimate of the value of goods in transit a wild guess? In that case, interstellar trade might not be trade in new goods but rather in antiques, a possibility that Cory Doctorow considers in his short story Craphound.

Technorati Tags: economics, SF, interstellar trade

Share on Facebook

Merger analysis works best in markets with unchanging, undifferentiated products, and costs that are affected more by scale than by changes in production methods. This is mainly because merger analysis is forward looking and the more tomorrow will look like today, the more reliable our prognosis. However (fortunately!) few markets that I have dealt with fall into this category. Katz’ and Shelanski’s excellent paper on mergers and innovation deals with the twofold effects that “innovation” has on merger analysis:

1. Innovation disrupts market definition and competitive effects prognosis; and
2. Innovation may be threatened directly by the merger.

The first effect they refer to as “innovation impact,” the second as “innovation incentives.”

Innovation impact

As long as the technology that forms the backdrop of today’s products and thus market boundaries doesn’t change much, there is a robust positive correlation between the number of rivals, the intensity of competition, and consumer welfare gains. More rivals, more intense competition, lower prices. That’s the tried and true structure-conduct-performance or concentration-competition-welfare paradigm. But what if technology changes or is about to change? Disruptive innovation:

• Severs the continuity between pre- and post-merger market definition;
• Makes pre-merger market shares both less reliable and less significant, which, in practice, suggests that new customer sales or license revenues are often more telling than share of installed base;
• Makes it much harder to identify competitors and entrants;
• Renders the SSNIP test pretty much useless, because where innovation rules, “price” (of what?) is often not a useful reflection of value.

Innovation incentives

Here, the concern is that the merger itself might change the nature, pace, and direction of innovation itself. Consider the acquisition of a small electric car maker by a traditional gasoline car manufacturer. The small company has every incentive to disrupt the market and steal share. The large company has to balance incremental profits from electric cars against potential losses from selling fewer gasoline-powered cars and, more importantly, undermining the market itself. Cannibalization can therefore be a significant disincentive to radical innovation. Redeploying the small company assets to make better car batteries and starters might thus be the more “civilized” approach to innovation. (Here are slides from my antitrust/ip course, discussing the replacement effect.)

Katz and Shelanski discuss the two traditional approaches to “innovation incentives,” that courts and agencies have taken in the past.

1. Extension of the static rivals-competition-welfare model to innovation cases, assuming that more rivals lead to more intense competition for innovation, resulting in increased consumer welfare.
2. Schumpeterian competition, with a three phase model of monopoly, creative destruction through disruptive innovation, and a subsequent race to monopoly, repeated over and over.

The authors find that there is only a weak correlation between increased concentration and less R&D, except in cases of mergers to monopoly, and a somewhat weakened connection between R&D and welfare, considering that some R&D races are wasteful. For example, spending an incremental $1 billion to beat a competitor in a paten race by one day. The incremental day is hardly worth $1 billion to society.

Efficiencies

In “innovation cases,” companies often make claims as to increased post-merger innovation. Katz and Shelanski identify the following:

1. Combination of complementary IP and R&D assets
2. Better R&D funding through risk-spreading; and
3. Higher margins, which will allow greater R&D funding.

Personally, I have never heard anyone make argument (3), for obvious reasons. As to the other claims, Katz and Shelanski find (2) unpersuasive because of cannibalization and because there is no evidence that a larger firm invests in qualitatively better R&D. In contrast (1) is plausible, provided that the parties can show that cooperative R&D is more promising than parallel, competitive R&D. In my view, (2) is not per se unpersuasive. It all depends on whether and to what extent the acquired potential for innovation competes with the buyer’s present capabilities and its valuation of its present market position going forward. A large buyer looking to displace the status quo may well be a more powerful disruptive force than a small firm, for example, because customers are more likely to adopt a radically new technology if it is backed by a blue chip corporation.

There’s more in this great paper. Download it while it’s hot!

Technorati Tags: innovation, antitrust, mergers

Share on Facebook

A while ago we discussed Radiohead’s decision to sell its new album on a “pay us what you think it’s worth” basis. Today’s NYT has a nice article on the band and the album release. With respect to the donationware pricing, Jon Pareles reports:

The band and its managers are not releasing the download’s sales figures or average price, and may never do so. “It’s our linen,” Mr. Hufford said. “We don’t want to wash it in public.” A statement from the band rejected estimates by the online survey company ComScore that during October about three-fifths of worldwide downloaders took the album free, while the rest paid an average of $6. Factoring in free downloads, ComScore said the average price per download was $2.26. But it did not specify a total number of downloads, saying only that a “significant percentage” of the 1.2 million people who visited the Radiohead Web site, inrainbows.com, in October downloaded the album. Under a typical recording contract, a band receives royalties of about 15 percent of an album’s wholesale price after expenses are recovered. Without middlemen, and with zero material costs for a download, $2.26 per album would work out to Radiohead’s advantage — not to mention the worldwide publicity.

As a general matter, the tradeoff here is one between royalty-free dissemination and obscurity. The graph shows a simultaneous release of a for-pay good (P) and a free electronic download (F), e.g., books, music, etc. Some customers substitute F for P (x%). Others buy P because they became aware of it through F (think Google, costless recommendations, etc.) (y%). As long as y% > x%, the publisher is better off. This is very likely a winning strategy for writers in the “long tail. As to music, I’m not sure, because the free download is a perfect substitute for the for pay version. Hopefully Radiohead will release its data at some point, or others will repeat the experiment.

Technorati Tags: music, radiohead, law & economics

Share on Facebook

Modern tying law is concerned with foreclosure in the tied product market, not customer exploitation in the tying product market. As a result, we require market power in the tying product market, because without market power, there’s nothing to be leveraged into the tied product market. But doesn’t the same logic suggest a requirement that the tying product market be at least 30% of the size of the tied product market? (Assuming, for convenience’s sake, that foreclosure of less than 30% is unlikely to significantly distort the competitive process.) Or put differently: If the tying market is less than 30% of the size of the tied market, then there’s no (per se) tying concern, irrespective of the defendant’s market power in the tying market. Because even if the defendant had 100% of the tying product market, and it imposed an airtight tie on its customers, it could foreclose no more than 30% of the customers in the (larger) tied product market. The chart below illustrates the concept.

In this example, A is a monopolist seller of a specialized lab tool that uses CD-ROMs. A requires its customers to buy standard CD-ROMs only from A, i.e., imposes a tie (e.g., for metering purposes). Some CD-ROM sales from B, D, and E to customer C will thus be foreclosed. But even if A has market power in the tying product market, the tie won’t make a dent in the huge, competitive market for CD-ROMs. Am I missing something?

Technorati Tags: tying, antitrust

Share on Facebook

Here is a process for determining the truth of any statement. This process greatly simplifies the law of evidence. Relevance, admissibility? How quaint!

Thanks Sean Bonner for this revelation.

Technorati Tags: The Truth

Share on Facebook