martes, 1 de junio de 2010


Republished at the P2P Foundation

How can be explained peer-to-peer processes, the growth of Wikipedia in comparison with the Britannica Encyclopedia or the growing importance of Linux in contrast to Windows. Even the success of Google can be hardly explained within the classic economic paradigm.

The classic economic paradigm was tailored for industrial economies. Production was explained as the result of a given combination of capital and labour. But, can this two productive factors explain the productive process of peer-to-peer, Wikipedia, Linux, Google? Certainly they can be used in explaining the productive behaviour of a car-factory, a textile-production or a small regular shop of any possible kind, but the explanatory power of that paradigm is much lower when facing the productive processes present in many of the new economic productive entities that have appeared after the disruptive development of information technologies.

What is different? Highways vs rural paths.
How highways are produced can be easily explained in terms of the two single productive factors capital and labour. But is this explanation useful in explaining how a rural path is created? A rural path appears due to the repetitive use of a given route. The repetitive use of the route progressively eliminates vegetation establishing a marked track on the ground. If the path is not used enough, it just disappears. Obviously labour might have a role: a nearby village might reward a group of people to have some obstacles removed. Capital might also be relevant: a cliff might be avoided through the construction of a bridge. But the radical difference is that its production cannot be explained just through capital and labour.
The production of a rural path could not possibly be explained without taking into account its mere use (consumption) as a productive factor.

Rural paths, P2P, Linux, Wikipedia and Google: why are they similar.
The so called new information technologies are characterized by network-effects and network externalities, this is, "when the value of a product to one user depends on how many other users there are, economists say that this product exhibits network externalities, or network effects" (Shapiro and Varian 1998, p.13)1. This idea was originally related to the economic analysis of communication services, telephone and fax being the typical case studies2. Indeed the utility of having a telephone is zero if anybody else has one, and utility increases as more people is connected. Similarly, the utility of a rural path increases as more users are connected through it. The argument is also valid for virtual networks such as Google, P2P software users, Wikipedia or Britannica, Linux or Windows and so on.

There is however a subtle difference in how an extra user benefits the rest of the users in a telephone network or in a rural path:

  • "classic" network effect: like in the telephone network case, an extra user might benefit the rest of users of the rural path by expanding the network (so an extra user can be reached).

  • "peer-production effect": unlike the telephone case, this extra user collaborates with the maintenance/creation of the path by walking it.

Likewise a search in Google contributes to the production of Google´s searching engine3, reporting a mistake in Wikipedia improves it, reporting a bug in Linux helps debugging it, joining Bit-torrent network forces users to share files (at least those that are being downloaded4). Peer-production process are characterized by anti-rivalry (for a further analysis see), implying that even free-riders pose a positive effect on production (Raymond 2001, p.15, recasts the term free-riders into "outriders"), or more generally mere consumption entails a contribution to production, consumption becomes a productive factor.

"Peer-production effect": Consumption contribution to production as a comparative advantage
While for a highway to be built capital and labour are strictly necessary, for a rural path it is not the case, users are instead essential. Similarly, for Microsoft and Britannica Encyclopedia to deliver a new version of their product, labour needs to be recruited and capital resources obtained. They both benefit from net-work effects, but there is no "peer-production effect", anti-rivalry is not present. Differently in Linux and Wikipedia case, labour and capital requirements are softened by its ability to use users´ contributions in production

A rural path and a highway serve the same purpose in different ways, so usually usually there is no competition and each has its niche. However in the case of an encyclopedia (Britannica vs Wikipedia) or an operative system (Windows vs Linux), there is indeed competence
6. Open source and some platforms such as Google benefit from user´s contribution while most proprietary platforms don´t.

Does peer-production pose a comparative advantage?

Could this explain the success of open platforms?

  1. More literature on the topic is provided by Varian, Farrel and Shapiro (2004, p.35-36). See Liebowitz and Margolis (1994) for some critical views.
  2. See Rohlfs (1974) for an early approach.
  3. See Benkler (2006, p.33), for instance.
  4. See Strumpf and Oberholzer-Gee (2010, p.10) for instance.
  5. Economides and Katsamakas find in "The Economics of Open Source Software Development" (2006) the conditions for which investment in open platforms (Windows vs Linux) is greater than in propietary platforms.
  6. See Bitzerand and Schröder (2006) in "The Economics of Open Source Software Development" (2006) , for instance.

  1. Carl Shapiro y Hal R Varian, Information Rules: A Strategic Guide to the Network Economy, illustrated edition. (Harvard Business School Press, 1998).
  2. Hal R. Varian, Joseph Farrell, y Carl Shapiro, The economics of information technology (Cambridge University Press, 2004).
  3. Jeffrey Rohlfs, “A Theory of Interdependent Demand for a Communications Service,” The Bell Journal of Economics and Management Science 5, nº. 1 (Spring 1974): 16-37.
  4. S. J. Liebowitz y Stephen E. Margolis, “Network Externality: An Uncommon Tragedy,” The Journal of Economic Perspectives 8, nº. 2 (Spring 1994): 133-150.
  5. Yochai Benkler, The wealth of networks (Yale University Press, 2006).
  6. Felix Oberholzer‐Gee y Koleman Strumpf, “File Sharing and Copyright,” Innovation Policy and the Economy 10, nº. 1 (Enero 1, 2010): 19-55.
  7. Eric S. Raymond, The cathedral and the bazaar (O'Reilly Media, Inc., 2001).
  8. Jürgen Bitzer y Philipp J.H. Schröder, The Economics of Open Source Software Development, 1º ed. (Elsevier Science, 2006).

lunes, 19 de abril de 2010


Republished at the P2P Foundation

"Unlocking Knowledge, Empowering Minds" This is the lemma of the OpenCourseWare, or OCW, an initiative pioneered by the MIT University to provide access to university educational contents for free and ubiquitously. The project started modestly in 2002 with 50 courses published online. Now the OpenCourseWare Consortium entails hundreds of universities from all around the world, giving free access to thousands of courses´ material in different languages, English being this the main one, while Spanish is scaling up.

Last April 16Th took place in Madrid a seminar on Open Educational Resources. The seminar included presentations from Tíscar Lara (EOI), Stephen Carson (MIT OpenCourseWare Consortium), Edmundo Tovar and Jesús Jara (UPM -Universidad Politécnica de Madrid) and Salvador Ros (UNED-Universidad Nacional de Educación a Distancia).

There is a monthly seminar on the topic organized by the eMadrid net, a project for the promotion of
Technology-Enhanced Learning publicly funded (Comunidad de Madrid).This time it took place at the Escuela de Organización Industrial (EOI) whose strategic plan has been partially inspired by the OCW project. As Tíscar Lara (vice-dean of Digital Culture at the EOI) said, stating her support for public, free (not for free) and open education. And in consequence her support for tools, contents and methods open as well (see Whyfloss).

Stephen Carson presentation was focused on the impact of OCW. Originally from the three functions universities have -provision of content, learning/social experience and certification- the aim of th
e OCW initiative was only focused on making content accessible so educators could incorporate the MIT teaching materials. Surprisingly from the use of OCW, educators represent only a 9%, and from this percentage those incorporating material to their teaching is a minority (14%): educators use it mostly to improve their knowledge and teaching methods. Against the original expectations self-learning represents the principal use: from all the users 43% are self-learners (not involved in official/institutional learning), and from these percentage, 41% use it for personal enjoyment. Furthermore, from the students using OCW contents, 44% are self-learners, the rest use it for completing (39%) and for planning their course of study (12%). Stephen Carson announced that given these unexpected success there are future plans to improve contents in order to make them more suitable for independent learners.

In open initiatives there is usually a stress on the distinction between “`free´ as in `free speech´, not as in `free beer´": in this case the distinction is unnecessary, it means both. Being free underlies both its success and some of the problematics of its future: since OCW is offered for free it is largely dependant on donors. Even though that it is free, OCW benefits not only users but content providers. For Stephen these benefits include publicity, reputation, international engagement, collaborations, connection to students... From a financial perspective these benefits are translated in donations to the OCW project, greater recruitment for the universities and potential collaborations with funding institutions thanks to the publicity and reputation that the visibility of professors and universities engaged in the project implies. Besides, the success of the OCW has resulted in an increasing governmental support, specially from the Obama administration. However as the NewYorkTimes puts it "Still, someone must pay for these materials, and with the recession squeezing university budgets, open course programs are vulnerable".

Edmundo Tovar and Jesús Jara (UPM)´s presentation was focused on their own experience rather than on
the OCW users experience, an indication (in my personal view) of the very early stage of the project in Spain (in the UPM it started in the 2006). Even though the UNED project started later (2007) the views of Salvador Ros were focused on the user perception and perspectives of the future of education. UNED is the national university for distance learning and offering virtual access to educational resources is not a disruptive shift of the educational paradigm, it is as Salvador said their "natural environment".

Salvador agreed with the audience that openness is rather than an economic problem a problem of mentality.
Most of the contents taught at university are universal, there is no sense in using an appropriation scheme of the educational resources: changes are necessary and inmobilism is not a good policy. Aware that changes are costly Salvador pointed out that more students accessing the resources imply that providing them is -potentially- increasingly cheaper.
In many Spanish/Portuguese speaking universities Universia, a foundation of the Santander Group for the promotion of university education (in Portuguese and Spanish), sponsors many OCW projects (included UPM and UNED). Nevertheless funding is limited and expansion of OCW projects is constrained and dependant on the donor.

The original aim of the OCW was providing content, and Salvador wondered if the time has come to expand its functionality to social/learning experience and accreditation provision. Social nets like Facebook or Twitter could provide the former, it is only accreditation provision which remains out of the scope, Salvador argued. For both workers and employers knowledge is usually not enough, educational accreditation is used as a signal of expected/desired capabilities. Universities are not just educational centers, providing content and learning experience, they are also "credentialing agencies". This was one of the central points of Salvador presentation: access is not enough, for the student, credit is usually a must. In this sense, the data presented by Stephen showing that, unexpectedly self-learners are the main users of OCW, could be an indication that OCW might not meet the needs of other users.

A solution for the economic problem of open education was offered by David Wiley in OpenCourseWars, his
contribution to the book "Opening Up Education" (via: P2PFoundation). David Wiley addressed the sustainability of the OCW project and the proposed solution bridges the two claims of Salvador Ros, the economic issue and the accreditation needs of users. Charging for accreditation could solve the financial dependence on the donors making OCW self-sustainable:
“The first generation of OpenCourseWare projects (”OCW 1.0?) had essentially no sustainability plan. These first generation projects were funded by grants and had no means of supporting themselves once the grants ran out (...). A new generation of OpenCourseWare projects are built around sustainability plans. These second generation projects are integrated with distance education offerings, where the public can use and reuse course materials for free (just like first generation OCWs) with the added option of paying to take the courses online for credit (there is no way to earn credit from the first generation OCWs)."
Changes are coming in education. As Stephen Downes puts it in his “ten-year-after” update of his classic essay on The Future of Online Learning (via: P2PFoundation):
"Today, much of the value derived from the learning marketplace is based on an artificially imposed scarcity – a scarcity of seats in classrooms, a scarcity of credentialing agencies, and a scarcity of educational publications, for example. These scarcities will disappear as governments prefer to fund education directly, and at cost, rather than support such business models."
Similarly to the former Salvador declared that "closed educational platforms are about to disappear, technology will overcome them". Unfortunatedly these coming changes will certainly face opposition, and interests different than access to knowledge might be dominant threatening the project. Salvador pointed to the recent publication by the Spanish Ministry of Culture of the inform "The Electronic Book", which has been broadly criticized (see for instance). The book encourages property right administration organizations to enforce property rights remuneration upon works even if these have been licenced under copyleft or creative commons licenses (for a more detailed analysis):
"Esto significa que las entidades de gestión vienen obligadas por Ley a hacer efectivos estos derechos de remuneración incluso aunque el autor hubiera decidido regalar su obra o no cobrar las cantidades recaudadas a su nombre." (p.19)

Reposted at the P2PFOUNDATION

jueves, 15 de abril de 2010

P2P and the Role of Exclusion II: the case of Wikipedia

One of the crucial characteristics of P2P is equipotency of its participants, in consequence it is not exclusion but non-rivalry or even "anti-rivalry", with free-riders making positive contributions to production ("outriders"), what is essential (see previous post).

But exclusion is still present: Wikipedia, one of the mayor successes of peer-to-peer collaborative projects has included exclusionary policies, as Benkler (2006) already ascertained,

"(...) even Wikipedia includes, ultimately, a small number of people with system administrator privileges who can eliminate accounts or block users in the event that someone is being genuinely obstructionist. This technical fallback, however, appears only after substantial play has been given to self-policing by participants, and to informal and quasi-formal community based dispute resolution mechanisms." (p. 104).

And it is expanding: The expansion of exclusion in Wikipedia is what Bongwon Suh, Gregorio Convertino, Ed H. Chi, Peter Pirolli (2009)1, from the Palo Alto Research Center have found (they have an extremely recommended blog). They find what seems to be "evidence of growing resistance from the Wikipedia community to new content, especially when the edits come from occasional editors." (sc.5.3). This happens at the same time that Wikipedia growth rates have for first time started to decrease.

Source: Suh et al. (2009).

Is this slowdown due to the reducing number of contributors? or has the Wikipedia covered such an amount of information that it is increasingly difficult writing new articles?

Suh et al. (2009, 5.1.) argue that "(...) the population of Wikipedia editors is exhibiting a slowdown in its growth due to limited opportunities to make novel contributions." Indeed once an article has been written it might be improved, updated, rearranged, but it does not make much sense to write it again. As the article is improved there is less and less room for improvements. More individuals contributing to the same article result in decreasing marginal contributions and increasing overhead costs (see Suh et al., 2009, sc.6). This is interestingly similar to rivalry. As Benkler (2002, supra note 16) points out "While the reference to information as a public good is common, the reference to culture is not." Wikipedia´s articles are an embodiment of knowledge and culture around a specific topic and "Obviously, embodiments of culture, like a specific statue or building, are no more nonrival than embodiments of any other form of information, like a book or a corkscrew."

Wikipedia is then an antirival process but its product might become rival.

As the articles become rival, the condition for antirivalry production becomes more relevant: the rate of contributors over free riders increases, until there is no antirivalry any more.

"(...) article growth reached a peak in 2007-2008 and has been on the decline since then [Figure 5]. This result is consistent with a growth processes that hits a constraint – for instance, due to resource limitations in biological systems. Microbes grown in culture will eventually stop duplicating when nutrients run out."

Total percentage of edits reverted are on the y-axis, the figure (11) shows how the "disparity of treatment of new edits from editors of different classes has been widening steadily over the years at the expense of low-frequency editors. [1-9 edits per month]".

Source: Suh et al. (2009).

Knowledge requirements to make contributions becomes increasingly demanding and by analogy to Darwin, Suh et al. (2009, p.1) suggest that "(...) growth becomes increasingly constrained and limited, and under those conditions there will be increased evidence of competition and dominance."

This problematic poses a threat on the project,

"If a project cannot defend itself from incompetent or malicious contributions and integrate the competent modules into a finished product at sufficiently low cost, integration will either fail or the integrator will be forced to appropriate the residual value of the common project—usually leading to a dissipation of the motivations to contribute ex ante." Benkler (2002, 379)

However antirivalry might remain at three different levels of the process: first, new "events in the world (...) create new possibilities for write-up"; second, in the creation of new rules "a greater proportion of the overall edits is being devoted to overhead activities such as coordination, policy setting, and governance", these are also contributions to the growth of Wikipedia: it creates the coordination mechanism. Third and last, any contribution has a signalling effect pointing to deficiencies, needs of new topics to be covered, or non-solved conflicts.

The issue is relevant because differently from Suh et al. (2009), the objective here is not technological singularity, the objective here is how the shift in the growth pattern affects the idea of Wikipedia as a peer-to-peer production process.

And what is the importance of antirivalry? Can there be equipotency in the presence of exclusion?

Non-rivalry is a necessary condition for equipotency, but it is not a sufficient condition. Antirivalry implies that any agent by the mere consumption of the peer-produced goods is also collaborating in its production. Thereby antirivalry is a sufficient condition for equipotency.

Is this the underlying core characteristic of peer-to-peer?

Reposted at the P2PFOUNDATION


Bongwon Suh, Gregorio Convertino, Ed H. Chi, Peter Pirolli; "Singularity is not near: slowing growth of Wikipedia" (2009)

Benkler, Y.; "Coase's Penguin, or, Linux and the Nature of the Firm." (2002).

Benkler, Y.; "The Wealth of Networks" (2006).

via: blog.p2pfoundation


1 Suh et al. (2009) for the remaining.