Building a license better than the GPL

Copyright licenses seem to matter – millions are using Creative Commons and the perhaps biggest collaborative, collective effort of our times – Free Software – is based on the GPL. The General Public License has been the beacon for how to develop large scale software projects based on the freedom to cooperate, adapt and tinker. Now Open Business has done an interview with Marshall van Alstyne, professor at MIT’s E-Commerce centre, who has studied how licenses can be optimized. He says ” our plan is to offer a hybridized open license with a set of default parameters rather like the user preference settings that come with Unix shell scripts or most software applications. Tuning the parameters can make the license behave more like the best features of the GNU Public License (GPL), Berkeley Software Distribution (BSD) license, or even some of the best End User License Agreements (EULA).”
So in short he proposes a license which will be better than the GPL based on scientific insight. OpenBusiness also spoke to him about other insights into the interaction between copyright, openness and innovation.

1. You have worked on licenses that incentivice collaborative work in
an optimal way? What does this mean and how could it work in practise?

We really tried to figure out what implications various licensing terms have for innovation and user freedom considered over the long term. If proprietary incentives matter for innovation but public access to a reusable free/open platform also matters for welfare, then what scope of protection creates the greatest good?

Public debates over issues like openness, network effects, developer incentives, and reusability led us to try to analyze various assumptions formally. We can’t capture everything perfectly but we put a lot of these factors into a model and came up with a few interesting results.

I’ll share three of these here that directly address your question then invite anyone who’s interested to read more on one of the discussion groups < lists.ibiblio.org/pipermail/cc-bizcom/2005-October/000114.html >.

1) Decentralized innovation can increase a firm’s profits over going it alone. For profit motivated firms, this means it becomes privately rational to choose open licenses once reuse and network effects pass a critical threshold. This effect can even dominate subcontracts with targeted developers.
2) Regardless of whether an institution is freedom or profit motivated, the optimal license is more open than BSD in the sense that the length of proprietary protection on derivative works should not be arbitrarily long. Multi-period innovation is thwarted by overlong protection.
3) Even if no reuse is possible and innovation never occurs, conditions exist where a profit motivated firm prefers to allow limited piracy, interpreted as free user access, to a portion of its products. Roughly interpreted, word-of-mouth effects on sales can dominate lost sales.

So how would this work in practice?

Our plan is to offer a hybridized open license with a set of default parameters rather like the user preference settings that come with Unix shell scripts or most software applications. Tuning the parameters can make the license behave more like the best features of the GNU Public License (GPL), Berkeley Software Distribution (BSD) license, or even some of the best End User License Agreements (EULA).

The terms would be set once for a given software platform and be limited to parameter choices that fall within bounds that are socially optimal for that type of product.

Keeping the license open, and the license design process open, means you don’t need to be a lawyer to know what you’re getting (which has been a great virtue of GPL). Parameterization allows the license to adapt to slightly different contingencies (providing balance between conflicting interests).

2. Legal scholar Yoshai Benkler describes an emerging mode of
production, which he calls “commons-based peer production”. What does
this mean? Can you give examples?

Well, Yochai is probably able to give the best definition of his idea but a synopsis might be decentralized egalitarian development based on shared resources. A good but age old metaphor is academia.

In academia, we build new ideas on a foundation of existing ideas. The pool of published knowledge represents a common resource on which we all draw. We modify that pool as our individual needs and insights permit. And, we contribute back to the pool as we share through publication.

Now, despite the best intentions, a lot of academic work is crap! The modal citation count across all journal articles is less than two. But, among the sea of articles (including many by non-academics), are works of incredible brilliance, like Raymond’s Cathedral and Bazaar or Benkler’s Coase’s Penguin. And, best of all, these works are free for anyone to use, adapt, deploy, and extend.

Really good works also reuse ideas from the pool. Raymond applies ideas from software design to organizational design. Benkler applies ideas from transaction cost economics explain the locus of production.

Recent examples of commons based peer production are Slashdot, Wikipedia, Linux, and Google Mash-Ups.

3. OpenBusiness attempts to collect new business models, which could
be paraphrased under “the more you give, the more you get”. Other
speak about the “Long Tail” (Anderson) or wikinomics. Do you see a
trend toward more open practises in business? Where lie the dangers
and where are the opportunities?

In general, I do see a trend toward greater openness. Let me start with a one thought on business model openness, then move to dangers and opportunities.

What is it that openness buys you in a business model?

Loosely interpreted, the major benefit is information. You get information about what people really want and what resources are available. This is the “Long Tail” of preferences, and not just the mass market mainstream. You get information on prices and decentralized matching such as the pairing of buyers and sellers on E-Bay. But you also get information in the form of contributed content. This is knowledge added to Wikipedia, the friendship network and self expression web pages added to MySpace, and the imagery added to Google Earth.

With all this information, we need ways to sort through it, to judge its value, and redeploy it. This creates at least three opportunities including search, credentialing, and innovation. The great mass of information creates a fairly obvious opportunity in search that includes text, social networks, images, music. We also need credentialing of the participants and the information they provide. This usually takes the form of certification such as the reputation systems that rate the users (e.g. E-Bay and Slashdot). But this also takes the form of recommender systems that rate products (e.g. “customers also bought”) and cataloguing that provide taxonomies (e.g. Yahoo! and del.ici.us).

And, one of the single greatest opportunities of openness is to harness decentralized innovation. Often, your users can figure out ways to use your information products and services in ways you didn’t anticipate. They can’t do this, however, if you haven’t given them the permission to do so (see for example O’Reilly’s Web 2.0). Google mash-ups is a great example of users adding value to Google’s platform.

I do see a few dangers associated with business model openness. The first is business hijacking. As you open up, competitors learn more about you and can either use your generosity against you or they can bootstrap their own initiatives without having to start from scratch. For example, Sun Microsystems has, from time-to-time almost lost control of Java and IBM has move aggressively into the Linux product space.

Second, the wealth of information generated on an open platform requires a means of separating high quality from low. When there is simply too much information, users need to be able to separate tell what is wheat and what is chaff. Editors choose which news posts to add to Slashdot, for example, and people who offer comments are often rated for the quality of their comments.

4. There are some heavily distributed collaborative projects, relying
on thousands of volunteers, like the Wikipedia; can they be sustained?

Absolutely yes.

As I hoped to imply by using academia as an example of commons based peer production, these systems can be quite robust.

Critical success factors for online collaborative projects are having (i) low transaction costs (ii) high interactivity and (iii) a solid reputation system. Low transaction costs make it easy for people to find and acquire what they need (having originated offline, academia previously lacked this one). High interactivity means people can easily volunteer content and evaluations. A good reputation system helps identify high quality and can even recommend material that a user did not pre-specify.

5. OpenBusiness.cc looks into how some of these “open projects” can be financially sustained. Did you come across, in your research, any interesting, new business strategies?

One of the more novel open strategies involves use of “two-sided” networks. These are open platforms with distinct user groups that attract one another. For example, game developers want to write for the platform with more users while game players want to buy the platform with better games. Job seekers want to go where there are more employers while employers want better pools of candidates.

Often the platform owner opens part of the platform, giving it away for free to one of these user groups, in order to make money from the other user group. So for example, buyers pay nothing to use the E-Bay service while sellers pay posting and sales fees. Job candidates can post resumes at recruiting sites while firms pay for access. Google gives away search for free in order to attract advertisers.

The trick is figure out how to match the groups of non-paying users with paying users and segment the market. If it’s done well, this also invokes network effects in ways that tend to lead to large dominant players as each side attracts more of the others.

For anyone specifically interested in “two-sided” business models, we’ve written a bit:
1) Parker & Van Alstyne “Two-Sided Network Effects: A theory of information product design” Management Science (2005) v51 n10 pp. 1494-1504.
2) Eisenmann, Parker & Van Alstyne, “Strategies for Two-Sided Markets” Harvard Business Review (forthcoming Oct. 2006).

6. Can copyright block these open forms of collaboration – how?

Yes, copyright can block collaboration and yet, when carefully used, it can also foster collaboration.

The current duration of copyright law in the US and most of the EU is long by the standards of typical software lifecycles. It can be long but less troublesome for most film and literature. In terms of computer games, it’s an eternity. Long terms give the copyright holder considerable power to exclude copies, distribution, and reuse over any economically interesting period. If the copyright holder chooses to exclude 3rd party copies, the potential for open collaboration declines.

Usually, this implies that the copyright holder prefers to charge for access. If anyone wants access to the copyrighted work, for collaboration or other purposes, then they need to pay for it.

So how can copyright foster collaboration?

This depends on the intentions and enlightened interest of the copyright holder. In fact, “copyleft,” the Free Software Foundation’s clever grant of rights to use, access, modify, and redistribute, really operates under copyright. That is, the long duration of copyright gives FSF the right to set the terms of others’ downstream use, which is quite distinct from having an information resource enter the public domain. If a creative work enters the public domain, then anyone is free to change it, copyright the changes, and restrict access to these changes. The copyleft provisions prevent this from happening but ironically only for so long as the work is under copyright. When copyright ends so too do many of the freedoms granted under copyleft.

This is one of several issues we have sought to model formally with an interesting result: if a creative work represents a “platform” on which others can build, then the owner of that platform needs long term tenure precisely to help ensure others “play nice” with their downstream use of the platform. Without no one to enforce good downstream behavior, access to new development can become choked off (ssrn.com/abstract=639165).

Basically, there is a tradeoff. Increasing copyright power increases both the chances of innovation and good resource husbandry but also abuse of power for private gain. Decreasing copyright power decreases incentives for innovation and resource management but limits abuses of power by the copyright holder.

7. There is a tricky economic question when it comes to free (as in beer) distribution of media – books, text, music, film – how can, in this environment, artists be remunerated?

There are at least two answers to this.

First, if we really assume that distribution is free in the sense of money rather than liberty, then one focuses on contributions like Wikipedia and YouTube and Flickr. In these cases, remuneration is the satisfaction of sharing and of greater personal reputation. In fairness to the professionals, what’s posted to YouTube and Flickr is not always the quality of Ridley Scott or Ansel Adams. But again, just as in academia, the desire to share and/or be recognized can be sufficient to cause artists to produce. So, for media that are distributed free, artist compensation is generally also non-monetary.

There is one major exception in which free distribution does lead to real compensation. This almost always takes the form of an economic complement that becomes more valuable as something else is given away for free. If peanut butter suddenly become free, then everyone selling jelly might see a rise in income.

My favorite personal example is “Download this Song” by M.C. Lars. You can download the MP3 free on MySpace. Lars makes his money by selling complements such as concert tickets, shirts, and extended play versions of the album. The song itself is so good that people freely distribute it to friends, with Lars’s blessing, prompting them to find out more about him and possibly also buy one of the economic complements.

Second, let me question one assumption underlying this question. I don’t assume that free in the sense of liberty necessarily implies free in the sense of price. Having access to an artist’s work is not the same as having the right to redistribute that work. If an artist maintains certain rights of re-distribution, then it is possible to set terms that involve limited forms of compensation.

Apple’s iTunes has answered the question of whether people would pay for music, for example, with a resounding “yes!”. Given this realization, the issue need not be the magical alchemy of converting grit to gold. Rather, it involves designing the shared resources and pricing the parts in such a way that everyone wins. It involves getting the business model right.

Paying artists directly for their works, rather than indirectly for economic complements, will almost surely involve a form of bundled pricing with two features. On the consumer side, a modest flat fee will provide access to a lot of personal choice. Then, on the artist side, compensation will be a fair claim to a fraction of these flat fees, based on the popularity of the works they provide.

A lot of economic thinking has gone into figuring out how to do this so it’s ironic to see how much effort existing firms put into preserving existing business models, such as the CD, rather than moving to newer ones where everyone might be better off. In fact, resistance by the Motion Picture Association of America (MPAA) to the VCR is not unlike that of the Recording Industry Association of America (RIAA) to digital music. The right business model, however, led to video rentals and an increase in consumption. The same will happen for artists’ work in general. It just depends on the right information economics.

Comments

Responses are currently closed, but you can trackback from your own site.