The Lexis/Westlaw Duopoly and the Proprietization of Legal Research

The Political Economy of Legal Research

LexisNexis and Westlaw share duopolistic control of the U.S. market for legal information. Lexis/Westlaw's stranglehold on legal information is founded on the unfortunate circumstance that American law is not freely and publicly available. U.S. courts charge 8 cents per page. Some state courts offer recent opinions, but older precedents are mostly unavailable. Statutes and regulatory orders are generally available online, but even these sources are relatively difficult to access, and they lack the metadata offered by Lexis/Westlaw. To varying degrees, then, the law is virtually proprietized, sold at a profit by Lexis/Westlaw.

Of course, Lexis/Westlaw don't legally own the law. Statutes, regulations, and opinions are produced by agents of the democratic government and enter immediately into the public domain, regardless of public availability.

• Not of course. This is true of federal publications. But state legal publications can be copyrighted. Thomson/West/Reuters' predecessor in interest, West Publishing Company, was the proprietary publisher of state reports from the 1880s, which is how the company built its fortune.
• -- JohnPowerHely - There have been some halfhearted attempts to fix this. See H. R. 1584, 104th Congress. In the past 56 years only this and a companion bill have been proposed, and neither made it out of committee.

The value added by Lexis/Westlaw is research efficiency--an electronic search of an online database is infinitely cheaper than scouring filing cabinets at a courthouse or library. The Lexis/Westlaw duopoly does not face a serious rival entrant because the costs of building a minimally comparable database from scratch present a prohibitive barrier to entry. At the same time, digital databases operate at increasing returns to scale, so the existing incumbents can easily underprice rival entrants and put them out of business.

The Lexis/Westlaw duopoly's proprietary internment of legal information imposes grave social costs. From the standpoint of political theory, American citizens do not have public access to the laws that are supposed to be influencing their behavior and for which violation could mean punishment by the state. The costs of civil litigation are distorted upward, precluding the pursuit of socially optimal lawsuits. When a wide income gap separates the parties to a dispute, the wealthier party has a superior research advantage. Most ironically, courts must pay Lexis/Westlaw for access to the fruits of their own labor--and the American taxpayer foots the bill.

For the individual researcher, the use of Lexis/WestLaw produces platform-specific human capital (analogous to firm-specific human capital). By using a particular research platform, a user becomes habituated to and proficient with that platform, and the resultant efficiency surplus cannot be recovered on alternative platforms. To the extent that platform-specific human capital has been vested, Lexis/Westlaw can extract rents from their users; as long as the monopoly-rent tax is less than the costs of rebuilding platform-specific human capital for another platform, the user will remain loyal and pay the tax.

• This assumes a degree of difficulty in learning how to perform searches that could only be itself the result of ignorance. It is easier to learn enough about databases to make learning to search productively with either (or with something else) largely transferrable.

• -- JohnPowerHely - What of the fact that the two platforms are not completely equivalent in terms of database coverage? I would be interested in seeing some data on whether or not the difference in coverage of briefs and secondary legal materials has an effect on legal scholarship, and if individuals who are more proficient with one than the other will cite to sources from that product.

Neither Lexis/Westlaw offers an optional modification mimicking the other platform's interface, implying collusion. In the event of collusion, Lexis/Westlaw are incentivized to arbitrarily differentiate their platforms' interfaces. If the platforms had identical (or interchangeable) interfaces, habituated users could costlessly swap platforms, precluding the extraction of monopoly rents. This dynamic can be observed in the functionally equivalent but symbolically differentiated search terms implemented by Lexis and Westlaw. Relatedly, Lexis/Westlaw have a standalone incentive to complicate their platforms. The more complicated a platform, the more platform-specific human capital can be invested, and correspondingly the higher the monopoly tax that can be extracted. This hypothesis is confirmed by a passing glance at Lexis/Westlaw's cluttered interfaces (Cf. Precydent's 's parsimonious appearance).

• There are too many reasons for this sort of design to accept this analysis on faith, and you give no reason to believe that these are causal relations. It is true that making a generic user interface is not in either's interest, but the real point is that neither has an incentive to get rid of web interfaces altogether in favor of a semantic web approach that would be superior for all users.

The Role of the Law School

During my first semester at Columbia Law School, my Legal Practice Workshop instructor required his students to register with Westlaw and submit all graded work via [[lawschool.westlaw.com][TWEN], Westlaw's proprietary online courseware. Since registering on Lexis 13 months ago as part of my Legal Research course, I have received 56 emails from Lexis--about one per week. These emails offer me "Lexis Points" in exchange for using their service--not for schoolwork, but for playing around on an arbitrary legal-research task.

• Requiring use of proprietary services to deliver education is an objectionable practice. We should have discussed this with the instructor.

The purpose and effect of Westlaw's TWEN services and Lexis's bribes, of course, are to habituate me to working on their respective research platforms. By purchasing Lexis/Westlaw access for students and faculty--as well as participating in the aforementioned habituation schemes--Columbia is complicit in the services' later abuse of law graduates' platform-specific human capital. There are decent commercial alternatives? to Lexis/Westlaw, as well as irreproachable free services, but Columbia's legal-research curriculum ignores them.

The Free Alternative

The defining feature of online legal research is that database software and legal information are digital, infinitely replicable and distributable at near-zero marginal cost. As Eben Moglen's theory predicts and history has shown, functional digital goods (like databases and search engines) are superiorly produced anarchistically, while non-functional digital goods (like legal information) are superiorly distributed anarchistically. These insights imply a framework for a serious free alternative to Lexis/Westlaw's proprietary legal-research platforms: a free and open-source legal-research platform on the scale of Lexis/Westlaw, operating on an anarchistically developed search engine, distributing information via anarchistic peer-to-peer file-sharing networks, and composing metadata anarchistically through a public wiki.

The free legal-research platform requires two centralized structural components: 1) a central catalog , and 2) a uniform citation protocol. The catalog will be a centralized record of every document in the decentralized database for search purposes. The decentralized Legal information can be stored and distributed anarchistically via peer-to-peer file-sharing networks, with the centralized catalog keeping track of where the information resides.

• You are trying to invent the semantic web. You won't do better than all the people working on it have already done, and the lawyers won't need more than the others working with much larger problems need.

The uniform citation protocol will be a rationalized system of citing and locating legal documents, with authorial attribution, time of publication, etc., plus a digital code that allows instant access to the document from the platform's search interface. Reporter paginations--an obsolete relic from the old library days--can be phased out in favor of paragraph numbers.

Apart from the catalog and citation, other aspects of the platform can be decentralized and customized anarchistically. For starters, the search interface will be decoupled from the database searched, with the interface freely customizable. Users and groups will be able to release custom interface packages that meet particular objectives and circumstances. Interface mods mimicking the Lexis/Westlaw interfaces will become popular because they allow previously habituated users to recover the benefits of their platform-specific human capital--without paying the proprietary services' monopoly rents.

Meanwhile, the metadata linking legal documents (but not the documents themselves) will be publicly editable by the legal community through a document annotations wiki. This project will be undertaken anarchistically, with gradually greater structure and restrictions built in as existing materials are exhaustively annotated. _Nature_'s finding that Wikipedia is just as reliable as the Encyclopedia Britannica suggests that the free legal-research platform could plausibly be as reliable as Lexis/Westlaw. With so many smart people using the database, abuses and manipulations can be cleared up quickly and the perpetrators exiled. In the long run, it is probable that the free platform would become superior to Lexis/Westlaw in usefulness and usability.

Because the benefits of such a free platform would be widely distributed across society, costs would be most efficiently internalized were the project government-funded. But such a course of action faces serious obstacles, not least of which being Lexis/Westlaw's substantial lobbying efforts. It would be more feasible, I think, to establish the free legal-research database as an independent nonprofit in the tradition of Wikipedia. Funding would be provided mainly by leading law schools, supplemented by donations from cooperative law firms. For financial context, it bears noting that the founders of FastCase scanned and transcribed every federal and state case and statute for $7 million. The top ten (or fifty) law schools could easily fund a similar public-domain effort. Alternatively, the law-school coalition could pay Lexis/Westlaw/FastCase to transfer their documents en masse into the free platform's database. This latter route might be more feasible than one would expect; while the free platform would likely put all three companies out of business, the search company that sells first will get out with some profit.

The social benefits of a free alternative to Lexis/Westlaw are incalculable. The free alternative described above should be pursued with all deliberate speed.

• First, will allowing American citizens--many of whom assume that statutes are the entirety of what makes up the law--to view caselaw freely solve the problem of being influenced and controlled by laws they don't have access/understanding of? On the other hand, Wikipedia has entries for some cases where the cases are (minimally) synthesized. If your concern is about getting the citizenry information about what cases are out there, is that equally helpful? Differently helpful? Second, I'm pretty sure that Lexis and WestLaw? are forced to have non-mimicking platforms because they keep suing each other for (what they perceive as) copyright violations, not because of collusion.

• Why are you pretty sure, JoshS? ? Could you help us be pretty sure too?

• Some think that wikis would be well-used for, say, treatises. I think there's not much value in putting all our caselaw in wikis--a special-purpose annotator might be better, because there shouldn't be any reason for The Public to go around changing the text of opinions. -- DanielHarris - 15 Nov 2008

*-- JohnPowerHely - Daniel - I'd agree that The Public shouldn't be able to change the text of opinions, but just as WestLaw? has KeyCites? and Lexis has Headnotes (or is it Header Notes?), an open source wiki version could have its own case notes and synopses. - Elliot - "non-functional digital goods (like legal information) are superiorly distributed anarchistically" - legal information is a non-functional good more akin to music than coding? I'm not saying you're wrong or right, but I'm still a little shocked (and kinda wondering why we are all putting in so much time and money here at CLS if that is the case).