Phantom authority, selfselective recruitment and retention of members in virtual communities: The case of Wikipedia by Andrea Ciffolilli
Virtual communities constitute a building block of the information society. These organizations appear capable to guarantee unique outcomes in voluntary association since they cancel physical distance and ease the process of searching for likeminded individuals.
In particular, open source communities, devoted to the collective production of public goods, show efficiency properties far superior to the traditional institutional solutions to the public goods issue (e.g. property rights enforcement and secrecy).
This paper employs team and club good theory as well as transaction cost economics to analyse the Wikipedia online community, which is devoted to the creation of a free encyclopaedia. An interpretative framework explains the outstanding success of Wikipedia thanks to a novel solution to the problem of graffiti attacks — the submission of undesirable pieces of information. Indeed, Wiki technology reduces the transaction cost of erasing graffiti and therefore prevents attackers from posting unwanted contributions.
The issue of the sporadic intervention of the highest authority in the system is examined, and the relatively more frequent local interaction between users is emphasized.
The constellation of different motivations that participants may have is discussed, and the barriersfree recruitment process analysed.
A few suggestions, meant to encourage long term sustainability of knowledge assemblages, such as Wikipedia, are provided. Open issues and possible directions for future research are also discussed.
- Club economies and virtual communities
- The case of Wikipedia
- Concluding remarks and open questions
Virtual communities are one of the building platforms of the socalled "new economy." Basically, they are associations of people able to meet and interact voluntarily on a network such as the Internet. A virtual community can be considered a potentially extraordinary new social structure in the sense that appears capable of guaranteeing unique outcomes in voluntary association. It cancels spatial distances and allows finding likeminded individuals outside the physical boundary of a circumscribed geographical area without modifying the basic principles governing voluntary association. So far, we have witnessed the outstanding effectiveness of virtual communities, especially in the creation of software; however, several other purposes can be reached by means of these organizations from political mobilization to the production of several kinds of public or club goods.
As all the actors belonging to the information society, virtual communities are involved in the provision and exchange of information through the use of IT. They can take advantage of new opportunities in order to achieve increasing returns in production, as well as engage in new valuable forms of social interaction. however, it should be borne in mind that virtual organizations, if mass produced, may lose their identity and hence cannot exploit all the sources of increasing returns that benefit other categories of actors that populate the new economy (Mansell and Steinmueller, 2000).
There are several kinds of virtual communities in cyberspace and a concise classification will be provided shortly. Despite such variety, in this paper I deal with a particular type of voluntary virtual associations: Purposebuilt virtual communities. Of course, the open source software movement represents the most famous and successful example of this typology. Drawing upon the literature on open source, I will focus on the economic issues that arise when that model is applied to an open content knowledge assemblage, such as a free available online encyclopaedia. A brief review will be made shortly of the main concepts introduced first by the economic theory of teams, followed by club theory. These concepts appear to be relevant for the analysis of purposebuilt virtual communities. The case of Wikipedia will then be applied in the context of these theoretical issues. In particular, the issues of authority, recruitment, reputation and retention of members are examined.
A useful taxonomy
As mentioned, there are several types of virtual communities that can be distinguished on the basis of the characteristics of their convenors. The classification, shown in Table 1, draws upon Steinmueller (2002) but aims at extrapolating and emphasizing the central issues of congestion costs (interpreted as the difficulty for the users to distinguish one community from another in a growing cyberspace) and coordination failure (interpreted as the difficulty to identify who is going to bear the responsibilities and costs arising when embarking upon the creation of a community). Moreover, an effort is made to recall the distinction between open and closed associations, as well as horizontal and vertical knowledge assemblages . The proposed taxonomy, which will prove to be useful in the following sections of the paper, is threefold.
Brandname Internet companies belong to the first type of communities, mainly involved in the provision of information services and resources. They are not specialized in the production of collective goods and the interaction of their members is not their principal concern, although they may encourage these activities in order to attract subscribers .
Second, in the market we find affinitybased communities, whose main goal is voluntary association. Affinitybased virtual communities cover their costs by means of advertising or subscription. These communities can be either open or closed based on how costs are distributed amongst members.
Finally, cyberspace is populated by purposebuilt organizations aiming to create public or club goods. These communities are mainly funded by governments, universities, foundations, or selffinanced by their convenors. Basically, they may be dedicated either to vertical or horizontal assemblages of information. Open source software constitutes the most popular and successful example of purposebuilt community, characterized by cumulative dependency.
Concerning the size of open source projects as a criterion for their classification, Krishnamurthy (2002) found that the community model is a poor fit for software production. Indeed, the top 100 mature projects on Sourceforge , are developed by one or a few individuals. However, Krishnamurthy does not argue that such result implies a sort of crisis in the open source community; as the author correctly recognises, large communities may exist and do things other than produce software. For example, communities may try out products and suggest new features.
Table 1: Virtual communities: A taxonomy
Community type Typical convenor Main purpose Dimension Main sources of funding Congestion costs and coordination failure Vulnerability to graffiti Popular examples Brand name Large "dot" companies Provision of information services and resources Large Subscription fees, sales and ads High High but technical and statutory means to face vandalism AOL Affinitybased closed membership Small groups of likeminded individuals Interest sharing, etc. Typically small but there are instances of large communities Advertising, subscription fees, selffinanced by convenor High Low the WELL open membership Small groups of likeminded individuals Interest sharing, etc. Typically small but there are instances of large communities Advertising, selffinanced by convenor High High but of small scale USENET, Slashdot, Napster Purposebuilt vertical assemblages of information Groups of highly skilled programmers Production of public goods or clubs (e.g., software) Mixed Governments, Universities, Foundations, selffinanced Low Low (they might be vulnerable but not very attractive) GNU/Linux, Apache horizontal assemblages of information Groups of highly motivated individuals (not necessarily programmers) Production of public goods or clubs (e.g., databases and encyclopaedias) Mixed Governments, Universities, Foundations, selffinanced Low Low (they might be vulnerable but not very attractive) Open Directory Project, Wikipedia
Club economies and virtual communities
The Theory of Teams represents a natural starting point for the analysis of the economies of virtual communities. Teams are defined as groups of people sharing identical preferences and hence aiming for the same purpose, but having at their disposal different information sets.
The first generation of team theory, developed by Marshak and Radner (1972), seeks to analyse how groups are efficiently built and coordinated, that is how a valuable structure is selected and maintained. Marshak and Radner propose a utopian interpretation of individual behaviour, in the sense that individuals that are part of organizations are not simply selfinterested but show a sense of social good and aim to maximize it (Kreps, 1990). Under the viewpoint of individual rationality, team theory deals either with completely rational people or with bounded rational individuals, those not able to anticipate all contingencies that may arise.
A map to locate team theory and other branches of research within economic theories dealing with types of individuals is drawn in Table 2. Two dimensions are taken into account: First, individual selfinterest orientation and second, degree of rationality.
Individuals may be completely rational, when they are able to foresee the future at no cost. Otherwise, they may show bounded rationality, when it is not possible to anticipate every future contingency and it is costly to maximize utility. Finally, individuals can be behavioural, when they follow a certain behavioural pattern of action characterized by relative profitability, rather than maximization of a utility function.
The second coordinate, on which the map is based, distinguishes between honest selfinterested individuals, opportunistic selfinterested and utopian individuals. People belonging to the last typology are a sort of illuminated seekers of social welfare, rather than selfinterested.
It is interesting to note that utopian individuals, characterizing team theory, in fact share similarities with members of some affinitybased, as well as, purposebuilt virtual communities. Indeed, these are moved by strong personal incentives, coincident with social interest to a certain extent when the goal of the community is the assemblage of a valuable collective good.
Basically, team theory analyses the problem of a decisionmaker that wants to select a network capable of maximizing a net expected payoff, with subtracted costs of observation, information transfer and computation.
Table 2: Economic theories and types of individuals
Source: Adapted from Kreps (1990).
Selfinterest orientation Opportunistic selfinterested Honest selfinterested Utopian Degree of rationality Complete Information economics General equilibrium Team theory Bounded Transaction cost economics Temporary equilibrium Behavioural Evolutionary economics
Club Theory and purposebuilt virtual communities: Is there a mismatch between theory and practice?
The first generation of team theory primarily aimed to answer two questions. First, what is the optimal decision function for a given structure of information and, second, what are the relative values of alternative structures of information (McGuire and Radner, 1972). In other words, what is the optimal organization involving several persons?
Team theory has "evolved" from this starting point. It still deals with group formation and with the provision of jointly consumed goods and services. However, it now considers alternative perspectives. For example, some research focuses on the case of groups that arise with a purpose of creating something valuable specifically for their members; this may happen through reciprocal conferment of externalities between participants.
Known as club theory, "club" indicates a public good which is nonrival but excludible. Clubs are made up of members who confer externalities on each other (Scotchmer, 2002). One externality that defines a club is the the way in which members share the cost of the production of a public good. There are several other externalities that tend to mark the boundaries of these associations.
First, there is a crowding effect that reduces the quality of goods and services provided and hence, the benefits of participation. Second, the cost structure of public goods production operates in a way that makes it more expensive to provide services to all parties in a large club. It is not obvious that both of these limiting elements are always at work in every kind of club or that they have the same importance in setting group boundaries. It may be straightforward to realize the effect of crowding and cost structure in defining and circumscribing the size of a gym for members of an athletic club. It is less clear to understand the effects of crowding and cost structure on the provision of an information club or public good such as television broadcast. In this case, crowding can be interpreted as the phenomenon where the larger the number of members accepted in the club, the lower the intellectual value of broadcasting becomes, due to the necessity to take into account everyone’s tastes, unless discriminatory methods like payperview are used.
Therefore, if we claim that club size is regulated by different categories of costs and benefits and we would like to compare those costs and benefits in order to decide the optimal size of a club, it is worth noting that it is not always an objective exercise.
As we will see, crowding, interpreted as high rate of participation, may produce positive externalities in some virtual communities like Wikipedia. The problem may be singling out a crowding threshold, above which congestion may arise and the ability to stimulate user interest collapse.
The premise that an association efficiently providing excludible public goods is bound in size has been challenged by Ellickson et al., (2001) who demonstrate the existence of competitive equilibrium in a large but finite club. They point out that, when no agent in the economy has more than a negligible impact on the utility function of other agents, perfect competition is possible.
In this paper there is another significant distinction to keep in mind, between anonymous and heterogeneous crowding (Scotchmer, 2002). In the first case, only the number of members matters but not the characteristics of the members themselves. In the latter, members are acknowledged as imposing different externalities due to their differing skills and characteristics. Such heterogeneity influences the problem of existence. When members are homogeneous, there might be a problem of existence because it is not guaranteed that the population size is a multiple of the small optimal club size. When the members are heterogeneous, there is not only a problem of scale but it may be also impossible to match the most desired combination of features in each group.
If heterogeneity is interpreted as variety of skills and needs, this feature is important for purposebuilt virtual communities because it confers mainly positive externalities. In Open source projects, a great deal of efficiency, reliability and quality is due to the vast number of individuals that check lines of code, fix bugs and provide every release with a detailed explanation. With software, productive skills matter much more than smoking preferences.
It is worth mentioning that, as evident and true for every kind of virtual community, cyberspace remarkably enlarges the boundaries of a location in which to find individuals with certain desirable characteristics (i.e. likeminded individuals).
In summary, club theory postulates primarily that a small size is better than a large one since the crowding effect is likely to hinder the development of an association. Second, the market is effective in producing clubs. However, with virtual communities, something different may be operating. Indeed, large can be beautiful and crowding, interpreted as massive participation, may be desirable, especially if heterogeneous.
In order to shed light on these issues, it is worth recalling an earlier distinction between openmembership associations and closedmembership ones. The latter are to be included in the category of clubs. The former are generally nonexclusive; some are open affinitybased groups while others are purposebuilt communities, such as open source and open content endeavours. Open content knowledge assemblages, while potentially valuable to the public, may not be appealing to corporations concerned with the bottom line. As a consequence, the financial support of foundations, universities and governments plays a big role in keeping many open content groups afloat.
Concerning crowding, in the case of open source and open content, the positive externality generated by increasing network size provides a stronger incentive that outweighs the assumption of club good theory that there is a value in exclusivity of access as a reason to avoid free riding.
The case of Wikipedia
Wikipedia.com was born in January 2001 as a complementary project of Nupedia.com, which was aiming to create a freely available online encyclopaedia . The publication of articles on Nupedia, in order to maintain highquality, was based on a traditional review procedure of the publishers responsible for coordinating the project. The result of such process was that the volunteer contributors had to face a long and deterrent itinerary of submission, review and, if necessary, negotiation that ended up in very few articles published. At this point, the idea of Wikipedia emerged as a laboratory in which the advantages of massive collaboration could have been exploited, with the intention of choosing the best articles and letting them take the hard and costly review path leading to Nupedia.
The project has been successful. Six months after the birth of Wikipedia, 6.000 articles were written. Currently, there are more than 170,000 articles .
The bandwidth and server was donated and is still provided by Bomis . On 20 June 2003, the existence a nonprofit parent organization, the Wikimedia Foundation, Inc. , was officially announced. The foundation would generate revenue through donations and grants, thus ensuring the continued growth of Wikipedia as well as other projects , providing information to the public free of charge and without advertising.
Wikipedia is based on the Wiki technology that characterizes many Web sites. A Wiki community is open in the sense that it allows anyone to participate, freely viewing information contained in a site, permitting editing of that information as well . Editing Web pages can be done without submitting changes to a publisher and negotiating for them.
Why does this approach work? One might assume that graffiti attacks would eventually frustrate an approach on this sort of large scale. However, Wikipedia has been successful with a great deal of notable content as well as content in a state of constant improvement. For the most part, content disasters in terms of quality have not occurred.
At the source of the boom: Low transaction costs and high disincentives to vandalism
Any transaction involves costs (Coase, 1937; Williamson, 1985). These range from the costs of writing a contract, to the costs of negotiating how to deal with unexpected contingences, from the costs of coordination to the costs of motivation (or commitment). Such costs are determined either by human characteristics (bounded rationality and opportunism), or by the type of transaction (frequency, uncertainty, asset specificity).
Teece (1988) claims that high transaction costs represent one of the main limits to the division of labour in the production and use of knowledge.
Wiki technology in a way literally cancels transaction costs for editing and changing information. Hence, this reduction in transaction costs acts as a catalyst for the development of the community. In turn, these reduced transaction costs means that there is full exploitation of massive collaboration economies. Hence, in the case of horizontal information assemblages, we might argue that any incentive that allows more authors to freely join in a given task, the larger the assemblage of information that is eventually produced (or in the case of Wikipedia, a larger number of articles is possible).
Another secret of the success of Wikipedia is related to the incentives that contribute to a "creative construction" of information, rather than a "creative destruction" of it. As noted earlier, I expected Wikipedia to be engaged in an endless war among reliable contributions and graffiti attacks that would have blocked the development of the Web site. In reality, that has not happened, basically because all changes made to any article are stored; it is possible to undo any unapproved modification with a single click. This makes the activity of littering a page extremely more expensive for an individual (in terms of time and reputation), than it is for anyone else. Therefore, also in this circumstance, it is a matter of costs.
Through this mechanism of editing and undoing meaningless changes or graffiti, an evolutionary process is fostered and only the best contributions survive the selection (Neus, 2001). Moreover, other factors contribute to Wikipedia’s success such as sources of authority and coordination.
Technology of social organization in Wikipedia: Accumulated reputation as a source of authority
Organizations exist to establish a certain degree of procedural and institutional authority (Steinmueller, 2002).
Procedural authority consists of incentives, social norms and power that define how decisions about practices, routines and procedures should be taken within an organization. It allows resolution of issues or disagreements among participants.
Institutional authority concerns the recruitment of members to an organization, assignment of roles, government of membership conditions and of expression.
In Wikipedia, some features that shape procedural authority are implicit of the Wiki software. The editing and undoing mechanisms, implemented by means of two pushbuttons on each article page, are all that a user needs.
Controversies about content occur. These disagreements are mainly resolved on discussion pages that accompany every article. Thus, every dispute is kept at a level closest to the issue that originated it. Moreover, discussion is not a conditio sine qua non for the changes, but it is only meant to guide edits. Only some procedures, such as editing the main page, are "not open" and hence, reserved for administrators (i.e. experienced and trusted users/developers). When this is the case, users can only participate to the discussion about the changes, without a capacity to edit directly the page.
Wikipedia has currently an unknown number of anonymous contributors and 29,853 registered users of which 143 are administrators and seven out have developer rights .
Whilst registered users can write articles, edit and discuss changes, administrators can exercise a certain degree of institutional authority. Indeed, they are allowed to ban IP addresses and permanently delete pages and their history. Such actions are undertaken when specific users are responsible for graffiti attacks, or when their writings and edits are not completely objective. Hence it is a general rule of the community that articles should be written from a neutral point of view. For example, one user was banned, who had written numerous contributions on German history . These articles were perceived as rightwing by most, hence ignoring this standard of neutrality. In this case, the decision for banning the user was secured only after a long and lively discussion on the mailing list.
A registered user may become an administrator. In order to become an administrator, a Wikipedian must send an email to the English Wikipedia mailing list (WikiENL), giving his or her login name and requesting access. This sort of access is granted to anyone who has been an active Wikipedian for a while, those individuals generally known and trusted in the community . In this sense, Wikipedia is characterized by a system of "distributed authority" (Mateos Garcia and Steinmueller, 2003a).
Final policy decisions are up to one of the founders, Jimmy Wales . However, if this sort of benevolent dictator attempted to deviate from a neutral and objective policy towards content (for example, in order to push a specific political agenda), then the license  provides a strong counterbalance to his power. The contributors may and should, in such a case, take the database and the software and set up a competing project .
Recruitment, retention and again, reputation
Individuals participate within Wikipedia voluntarily, with various degrees of involvement. The most active contributors, those who do not visit the site only for transitory curiosity, are likely to be pushed by strong incentives that given the nature of the project itself, cannot be directly related to money.
These motivations may be personal: Selfsatisfaction, selfefficacy, intrinsic drive to acquire knowledge. These motivations may also be social in nature: Passion and desire to take part in the production of a collective good, a need for belonging, a need to support a specific community. Motivations may also be ethical, or they may be related to reputation, which may become a source of authority, a font of fame and a voucher to play in the labour market . In general, selfsatisfaction and career concerns are better known as the economic incentive of signalling .
Certainly there is no space for political motivation, at least in Wikipedia .
Participation in the community is not characterized by a bouncerdriven selection, but rather by selfselection: Everyone may participate if he or she wishes, even without registering. At times, bouncers at the entry of a disco club have an extraordinary capacity to prime an adverse selection. Paradoxically, in order to be a good bouncer, one should be "bad boy". Unfortunately, "bad boys" have "bad" peers that eventually may reduce the "quality" of membership, turning the club into a den of rascals. It is only speculative whether the introduction of any kind of bouncer or barrier to entry could hurt the growth and bias the development of the free encyclopaedia. Perhaps, a low barrier may improve quality and further reduce graffiti, without any serious counter effect. Basically, a low barrier helps to avoid gathering an unfocussed group of people. A low barrier, for example, may be a compulsory registration and/or the filling a brief curriculum vitae.
The reputation of Wikipedians grows with the number of their contributions. By visiting the previous version pages of each article, it is possible to go back to its author or authors. However, the names or the nicknames of the authors, unlike print encyclopaedias, do not appear at the bottom of the articles. In my view, this is a pity because the possibility to directly show the names of the authors would represent a powerful incentive and a strong gratification for contributors. The current anonymity may represent, in the long run, a threat to retention of members. It is true that many users edit articles, but this does not preclude the possibility to show the name of the majority contributor (for instance the person or the persons that wrote more than 50 percent of a document). Such a possibility may also stimulate competition among users and this can only be healthy when the effort is directed to the construction of a collection of information.
It is worth noting again that, because of the Wiki paradigm, the community strongly supports innovation and seems free of some of the problems of scholarly publication .
Finally, Wikipedians have at their disposal a personal page. So far, this personal space has been used mainly by the administrators and few other registered users basically to recall articles that they wrote or strongly contributed to. It must be borne in mind though that personal pages can be crucial in strengthening a sense of trust and identity among community members (Neus, 2001). Indeed, community sustainability may largely depend on the complex interaction of a variety of factors, such as personal pages, authorship and sysop authority.
Concluding remarks and open questions
This paper has dealt mainly with the issues of authority, reputation, recruitment and retention of members in purposebuilt virtual communities, engaged in creating horizontal collections of information goods. Some key ideas, drawn from theories describing teams and clubs, have been explored.
The case of Wikipedia, a successful project committed to the creation of a free online encyclopaedia, was examined. The principal reasons for the success of Wikipedia namely, the drop in the transaction costs of submitting contributions and erasing graffiti were described. It was shown how procedural and institutional authorities work for this site. In particular, the importance of reputation, as a source of authority, was emphasized. Reputation is accumulated through participation and that shapes a system of distributed authority in which every participant potentially may have a role in the development of the project.
The free/open encyclopaedia an "impossible" public good challenges some economic premises. The claims of club theory that excludible goods are more efficiently produced in small associations, whose boundaries are necessarily circumscribed by crowding, and that the public sector should have no role in the provision of clubs, are contradicted by the observed case.
This apparent mismatch however may be explained by observing that club theory deals with exclusive efforts, whereas open source and open content projects of which Wikipedia represents an example are in general nonexclusive. In their case, the positive externality generated by increasing the size of the network provides a stronger incentive that outweighs the assumption of club theory that there is value in exclusivity.
While closed affinitybased communities are exclusive efforts and may be expected to be spontaneously provided, open source and open content projects are generally nonexclusive. Financial support from universities, foundations and governments appears to be crucial.
Despite their positive aspects, virtual communities such as Wikipedia are not absolutely free of troubles. While the processes of recruitment and retention have been working quite well so far, but will they continue to operate in the future? Are the mechanisms that allow accumulation of reputation and hence, authority distribution, strong enough to guarantee mediumterm sustainability, and to maintain the sense of trust and identity among members? Some instruments for entry selection, such as a compulsory registration, may further improve the quality of recruitment and therefore the outcome of the cooperative effort, without being detrimental to the community’s momentum. At the same time, more intense use of personal profile pages and a direct recognition of the contribution made by each user could foster a sense of trust and help retain retain participants. Finally, let me mention another problem related to the exercise of authority. If the number of administrators, retaining a certain degree of institutional authority, continues to grow over time, will a new complexity make it necessary to increase the number of hierarchical layers in the structure and discourage participation? This issue will need to be resolved at some point in the future.
There certainly is great charm exerted by successful projects of massive collaboration producing a public good or a club. Part of this charm may be attributable to the open source approach. In fact, an open source mode of dividing innovative labour may be useful to encourage knowledge sharing, breaking information bottlenecks.
About the Author
Andrea Ciffolilli is a PhD student in Economics at the University of Ancona, Italy. He has also studied Technology and Innovation Management as part of his Master’s degree at SPRU Science and Technology Policy Research, University of Sussex, Brighton, U.K. His current research interests revolve around the division of labour in online communities and socioeconomic consequences of strengthened intellectual property rights on production, access and distribution of electronic information resources.
I wish to thank Professor Ed Steinmueller for insightful comments on a previous draft of this paper. Many thanks are due to Judita Jankovic for proofreading. I am the only one to be blamed for whatever errors or omissions have survived in the text. Financial support from the Department of Economics, University of Ancona Italy, is gratefully acknowledged.
1. Vertical knowledge assemblages are collections of information characterized by cumulative dependency: The loss of a piece of information or a module is likely to render the assemblage unusable. This happens, for example, in the case of software. Horizontal knowledge assemblages are collections of information characterized by complementary dependency: The loss of a piece of information may reduce the value of the assemblage which however remains still usable. This is the case for example of a directory or an encyclopaedia. For a detailed discussion of these issues, see Mateos Garcia and Steinmueller (2003b).
2. MSN (Microsoft Network) constitutes an example of brand-name Internet community. The recent announcement of MSN to close down its chat rooms in Europe, apart from its declared motivation of restraining pedophilia, may be interpreted as a decision driven by the urge to prune an accessorial activity, considered no longer profitable.
3. SourceForge.net is an online platform that hosts open source projects and provides tools for their development. See http://www.sourceforge.net, last accessed 10 November 2003.
4. Jimmy Wales and Larry Sanger founded both Nupedia and Wikipedia. The project Nupedia started on 9 March 2000.
5. The total number of articles, reported on the main page of the English version of the encyclopaedia, is 171,951. See http://www.wikipedia.org, last accessed 11 November 2003.
6. Bomis is an Internet media company headquartered in San Diego, Calif. Jimmy Wales, co-founder of Wikipedia, is the majority owner of Bomis. See http://en.wikipedia.org/wiki/Bomis, last accessed 11 November 2003.
7. See http://wikimediafoundation.org, last accessed 11 November 2003.
8. The Wikimedia Foundation is the non-profit parent organization, not only of Wikipedia, but also of Wiktionary, Wikiquote, Wikibooks and Nupedia. See http://en2.wikipedia.org/wiki/Wikimedia, last accessed 11 November 2003.
9. As stressed, the extremely open nature of projects such as Wikipedia renders them "impossible" public goods rather than real clubs. The concepts introduced by the theory of clubs can be used however in this example to study an open mode of working.
10. See http://www.wikipedia.org/wiki/Special:Statistics and http://www.wikipedia.org/wiki/Wikipedia:Administrators, last accessed 11 November 2003. "Developer" is the degree of access for those who can make direct changes to the Wikipedia software and database.
11. The user’s nickname was "Helga Jonat"; his/her IP was banned in September 2002. See http://www.wikipedia.org/wiki/Special, last accessed 14 April 2003.
12. See http://www.wikipedia.org/wiki/Wikipedia:Administrators, last accessed 11 November 2003.
13. Larry Sanger, the other founder of the project and former chief editor, left Wikipedia on 1 March 2002.
14. The content of Wikipedia is released under the terms of the GNU Free Documentation Licence (GFDL). See http://www.wikipedia.org/wiki/GNU_Free_Documentation_License, last accessed 10 September 2003.
15. Jimmy Wales wrote: "[I]n order to hold the project together, and in order to keep the largest possible group of people working together on the central project, I must listen carefully to all elements of the community, and make decisions that are satisfactory to the best interests of the encyclopaedia as a whole". See http://meta.wikipedia.org/wiki/Wikipedia_Governance, discussion about the free encyclopaedia, last accessed 10 September 2003.
16. For a review of the motivations of individuals participating to open source projects, see for example Harhoff et al. (2000).
17. Lerner and Tirole (2000) define signalling as the sum of a career concern incentive and an ego gratification incentive. Signalling is, of course, informative about individual talent.
18. See the sort of policy statement draft, written by Jimmy Wales on http://meta.wikipedia.org/wiki/Wikipedia_Governance, last accessed 10 September 2003.
19. For an irreverent discussion of the problems related to the reference system in economics, see Frey (2002).
R. Coase, 1937. "The Nature of the firm," Economica, number 4, pp. 386405. http://dx.doi.org/10.1111/j.1468-0335.1937.tb00002.x
B. Ellickson, B. Grodal, S. Schotchmer, and W. Zame, 2001. "Clubs and the market: Large finite economies," Journal of Economic Theory, volume 101, number 1, pp. 4077. http://dx.doi.org/10.1006/jeth.2001.2795
B. Frey, 2002. "Publishing as prostitution? Choosing between one’s own ideas and academic failure," Zurich: Institute for Empirical Research in Economics, Working Paper number 117 (June); and at http://www.iew.unizh.ch/wp/iewwp117.pdf.
D. Kreps, 1990. A Course in microeconomic theory. London: Harvester Wheatsheaf.
S. Krishnamurthy, 2002. "Cave or community? An Empirical examination of 100 mature open source projects," First Monday, volume 7, number 6 (June), at http://firstmonday.org/issues/issue7_6/krishnamurthy/. http://dx.doi.org/10.5210/fm.v7i6.960
D. Harhoff, J. Henkel, and E. von Hippel, 2000. "Profiting from voluntary information spillovers: How users benefit by freely revealing their innovations," MIT Sloan School of Management, WP (July), and at http://www.inno-tec.bwl.uni-muenchen.de/forschung/henkel/FreeRevealing2003.pdf.
J. Lerner and J. Tirole, 2000. "The Simple economics of open source," Working Paper, number w7600 (March). Cambridge Mass.: National Bureau of Economic Research.
R. Mansell and W. Steinmueller, 2000. Mobilizing the information society. Oxford: Oxford University Press.
J. Marshak and R. Radner, 1972. Economic theory of teams. New Haven, Conn.: Yale University Press.
J. Mateos Garcia and W. Steinmueller, 2003a. "The Open source way of working: A New paradigm for the division of labour in software development?" Brighton: SPRU, Science and Technology Policy Research, Open Source Movement Research, INK, Working Paper number 1 (January), at http://www.sussex.ac.uk/spru/publications/imprint/sewps/sewp92/sewp92.pdf.
J. Mateos Garcia and W. Steinmueller, 2003b. "Applying the open source development model to knowledge work," Brighton: SPRU, Science and Technology Policy Research, Open Source Movement Research, INK, Working Paper number 2 (January), at http://www.sussex.ac.uk/spru/publications/imprint/sewps/sewp94/sewp94.pdf.
C. McGuire and R. Radner, 1972. Decision and organization. London: North Holland.
A. Neus, 2001. "Managing information quality in virtual communities of practice," In: E. Pierce and R. KatzHaas (editors). IQ 2001: Proceedings of the 6th International Conference on Information Quality at MIT. Cambridge, Mass.: MIT Sloan School of Management, and at http://opensource.mit.edu/papers/neus.pdf.
S. Scotchmer, 2002. "Local public goods and clubs," In: A. Auerbach and M. Feldstein (editors). Handbook of Public Economics, volume 4, pp. 1998#150;2042, and at http://ist-socrates.berkeley.edu/~scotch/survey.pdf.
W. Steinmueller, 2002. "Virtual communities and the new economy," In: R. Mansell (editor). Inside the communication revolution: Evolving patterns of social and technical interaction. Oxford: Oxford University Press.
D. Teece, 1988. Technological change and economic theory. London: Pinter.
O. Williamson, 1985. The Economic institutions of capitalism. New York: Free Press.
Paper received 22 September 2003; accepted 11 November 2003.
Copyright ©2003, First Monday
Copyright ©2003, Andrea Ciffolilli
Phantom authority, selfselective recruitment and retention of members in virtual communities: The case of Wikipedia by Andrea Ciffolilli
First Monday, volume 8, number 12 (December 2003),