Articles Coase's Penguin, or, Linux and The Nature of the Firm Yochai Benklert CONTENTS INTRODUCTION........... ......... ........ .................... 371 I. PEER PRODUCTIONALL AROUND ...................................................... 381 A. Content......................................................................................... 384 B. Relevance/Accreditation...........................................................390 C. Value-AddedDistribution ............................................................396 D. Summ ary ......................................................................................399 t Professorof Law, New York UniversitySchool of Law. Researchfor this Article was partlysupported by a grantfrom the FilomenD'Agostino and Max E. GreenbergResearch Fund at the New York UniversitySchool of Law. I owe thanksto many for commentson this and earlierdrafts, including: Bruce Ackerman, Ed Baker,Elazar Barkan, Jamie Boyle, Dan Burk,Ben Damstedt,Bob Ellickson,Terry Fisher, Natalie Jeremijenko,David Johnson,Dan Kahan,Niva Elkin Koren,Tara Lemmey, Larry Lessig, Doug Lichtman,Jerry Mashaw, Eben Moglen, Mark Nadel,Helen Nissenbaum, George Priest, Peggy Radin,Carol Rose, ChuckSabel, Alan Schwartz, Clay Shirky,Richard Stallman, and Kenji Yoshino.I owe special thanksto Steve Snyderfor his invaluableresearch assistance on the peerproduction enterprises described here. I have gotten many questionsabout the "Coase'sPenguin" portion of the title. It turnsout that the geek culturethat easily recognizes"Coase" doesn't recognizethe "Penguin,"and vice versa."Coase" refers to RonaldCoase, who originatedthe transactioncosts theoryof the firmthat providesthe methodologicaltemplate for the positiveanalysis of peerproduction that I offer here. The penguinrefers to the fact that the Linux kernel developmentcommunity has adoptedthe image of a paunchy penguin as its mascot/trademark.One result of this cross-cultural conversationis thatI will occasionallyexplain in some detailconcepts that are well knownin one communitybut not in the other. 369 370 The Yale Law Journal [Vol. 112: 369 II. WHY WOULD PEER PRODUCTIONEMERGE IN A NETWORKEDENVIRONMENT? .................................................. 400 A. Locating the TheoreticalSpace for Peer Production ............... 400 B. Peer Production of Informationin a Pervasively NetworkedEnvironm ent ...............................................................404 C. Markets,Hierarchies, and Peer Production as Information-ProcessingSystems .................................................406 1. InformationGains . ................................................. 407 2. Allocation Gains.................................................................... 415 III. OF MOTIVATION AND ORGANIZATION:THE COMMONS PROBLEM ... 423 A. The "Incentives"Problem: Of Diverse Motivationsand Small Contributions .................................................. 423 1. Abstractingthe Effect of Diverse Motivations.......................426 2. Diverse Motivationsand Large-Scale Collaborations.......... 434 B. Integration:Problem and Opportunity. ......................................436. 1. Threatsto Motivation............................................................ 439 2. Provisioning Integration.......................................................441 CONCLUSION.............................................................................................444 2002] Coase's Penguin 371 INTRODUCTION Imagine that back in the days when what was good for GM was good for the country, an advisory committee of economists had recommendedto the President of the United States that the federal government should supportthe efforts of volunteercommunities to design and build cars, either for sale or for free distribution to automobile drivers. The committee members would probably have been locked up in a psychiatric ward-if SenatorMcCarthy or the House Un-AmericanActivities Committeedid not get them first. Yet, in September 2000, something like this actually happened. The President's Information Technology Advisory Committee recommendedthat the federal governmentsupport open source software as a strategic national choice to sustain the U.S. lead in critical software development. At the heart of the economic engine of the world's most advanced economies, and in particularthat of the United States, we are beginning to take notice of a hardy, persistent, and quite amazing phenomenon. A new model of production has taken root, one that should not be there, at least according to our most widely held beliefs about economic behavior. The intuitions of the late twentieth-century American resist the idea that thousands of volunteers could collaborateon a complex economic project. It certainlyshould not be that these volunteerswill beat the largest and best- financed business enterprisesin the world at their own game. And yet, this is precisely what is happeningin the softwareindustry. The emergence of free software2 and the phenomenal success of its flagships-the GNU/Linux operatingsystem,3 the Apache web server, Perl, 1. PRESIDENT'SINFO. TECH. ADVISORY COMM., DEVELOPING OPEN SOURCE SOFTWARE TO ADVANCE HIGH END COMPUTING (2000), at http://www.ccic.gov/pubs/pitac/pres-oss- 1lsepOO.pdf. 2. I use the terms"free software" and "opensource software" interchangeably in this Article. Those who considerthe phenomenonas first and foremostinvolving political values, to wit, freedom,use the former,in self-consciouscontradistinction to those who focus on the economic significance, who use the latter. Compare ERICRAYMOND, Homesteading the Noosphere, in THE CATHEDRALAND THEBAZAAR: MUSINGS ON LINUXAND OPEN SOURCEBY AN ACCIDENTAL REVOLUTIONARY65 (2001) [hereinafter THE CATHEDRALAND THEBAZAAR] (focusing on the economic significance),available at http://www.firstmonday.dk/issues/issue3_10/raymond/,with Free Software Found., Why "Free Software" Is Better than "Open Source," at http://www.fsf.org/philosophy/free-software-for-freedom.html(last modified Aug. 26, 2002) (focusingon the politicalsignificance). I have writtenand continueto write quite extensivelyon the normativeimplications of how informationproduction is organized,see, e.g., YochaiBenkler, The Battle over the Institutional Ecosystem in the Digital Environment, 44 COMM.ACM 84 (2001), butnot in this Article,where I generallyabjure disputations over the terms. 3. I describethe operatingsystem as GNU/Linuxto denote that it is a combinationof the kernel developmentproject initiatedby Linus Torvaldsin 1991 and of many other operating system componentscreated by the GNU project, which was originatedin 1984 by Richard Stallman,the fatherof free software.Throughout the Article,I referto GNU or Linuxseparately to denotethe specific developmentproject and to the operatingsystem as GNU/Linux.I departed from this practicein the title for stylistic purposesalone. The completeGNU/Linux operating 372 The Yale Law Journal [Vol. 112: 369 sendmail, BIND-and many other projects4 should force us to take a second look at the dominant paradigmwe hold about productivity. In the late 1930s, Ronald Coase wrote The Nature of the Firm,5 in which he explained why firms emerge, defining firms as clusters of resources and agents that interact through managerial command systems rather than markets. In that paper, Coase introducedthe concept of transactioncosts, which are costs associated with defining and enforcing property and contractrights and which are a necessary incident of organizingany activity on a market model. Coase explained the emergence and limits of firms based on the differences in the transactioncosts associated with organizing productionthrough markets or throughfirms. People use marketswhen the gains from doing so, net of transactioncosts, exceed the gains from doing the same thing in a managed firm, net of organizationcosts. Firms emerge when the opposite is true. Any individual firm will stop growing when its organization costs exceed the organization costs of a smaller firm. This basic insight was then extended and developed in the work of Oliver Williamson and other institutionaleconomists who studied the relationship between markets and managerial hierarchies as models of organizing production.6 The emergence of free software as a substantialforce in the software- development world poses a puzzle for this organization theory. Free software projects do not rely either on marketsor on managerialhierarchies to organize production. Programmersdo not generally participate in a project because someone who is their boss instructedthem, though some do. They do not generally participatein a project because someone offers them a price, though some participantsdo focus on long-termappropriation throughmoney-oriented activities, like consulting or service contracts.But the critical mass of participationin projects cannot be explained by the direct presence of a command, a price, or even a future monetary return, systemis whateveryone has in mindwhen they speakof the breathtakingsuccess of free software at makingexcellent high-end software. 4. For an excellent history of the free software movement and of the open source development methodology, see GLYNMOODY, REBEL CODE (2001). 5. Ronald H. Coase, The Nature of the Firm, 4 ECONOMICA386 (1937). 6. The initialframing in termsof the oppositionbetween markets and hierarchywas OLIVER E. WILLIAMSON, MARKETS AND HIERARCHIES: ANALYSIS AND ANTITRUST IMPLICATIONS:A STUDY IN THE ECONOMICS OF INTERNAL ORGANIZATION (1975) and OLIVER E. WILLIAMSON, THE ECONOMICINSTITUTIONS