Collective cooperation: The phenomenon of open source

Josh Lerner interviewed by Viv Davies, 18 February 2011

Unfortunately the file could not be found.

Open in a pop-up window Open in a pop-up window


Download MP3 File (16.7 MB)




View Transcript

<p><em>Viv Davies interviews Josh Lerner for Vox</em></p>
<p><em>February 2011</em></p>
<p><em>Transcription of a VoxEU audio interview []</em></p>
<p><strong>Viv Davies</strong>: Hello, and welcome to Vox Talks, a series of audio interviews with leading economists from around the world. I'm Viv Davies from the Centre for Economic Policy Research. It's the 14th of February 2010, and I'm speaking to Professor Josh Lerner of Harvard Business School about a book he has recently co authored with Professor Mark Schankerman, titled, &quot;The Comingled Code: Open Source and Economic Development.&quot; The book presents a rigorous economic analysis of the impact of open source software on consumers, firms, and economic development. I began the interview by asking Professor Lerner to explain exactly what open source is and why the authors suggest in the book that the impact of open source software has generated more heat than light.</p>
<p><strong>Josh Lerner</strong>: We might begin by just thinking about what open source is. And when you think about the key elements that make a software an open source project, certainly one of the most important aspects is that the code itself is available, in other words, that as people make contributions to these projects, there is an expectation that much of the contributions will be placed into the public domain for others to see. Now, the specific rules and the ability of people to take the software code and reuse it for commercial purposes varies, but the essential principle is that these projects are ones that are being done for the common good, and that for contributors to be able to take and use under the licensing rules is certainly a critical element. It's fair to say that open source has several other aspects, which may not fit into the formal definition but are certainly very important as well, which have to do with, often, the sort of strong sense of community, the self organizing principles behind them that lead to a very strong sense of commitment, in many cases, to the open source development process.</p>
<p>And I think this brings us to the second half of your question, which is why these things have proven to be controversial. Certainly, it's fair to say that in many instances, established software companies have found these things to be somewhat threatening, because it represents a very different way of developing software. And it's also fair to say that in many cases, people who are enthusiasts about open source make very broad claims about the superiority of the software development.</p>
<p>So, despite the fact that this is a really economic phenomenon, and one which has a lot of economic importance, a lot of the discussion has been in more what you could describe as ideological or emotional tones, rather than really taking a critical look of what's involved in open source and how does it work and what are its key features.</p>
<p><strong>Viv</strong>: You describe the main purpose of the book as being to try to understand the role of open source software in economic development, its relationship to growth and innovation and so on. And I guess this is why the subject will be of interest to economists. Could you perhaps expand a little on this, and also explain how you approached the research and writing of the book?</p>
<p><strong>Josh</strong>: I think that you can sort of begin by going back to one of the very famous essays in the economics of technological change, which was by Ken Arrow in 1962, and in it he really laid out the fundamental paradox associated with the development of new technologies. On the one hand, the new technology is a non rival good, which means to say that if one person figures out how to use it, it's not like a piece of property, a house, which only one person can live on. Essentially, this technology can now be used by any number of people. It's essentially a recipe that can be broadly disseminated. The problem is that to provide people with incentives to develop the recipe, or to do this, we have to give them some degree of exclusivity, or else they won't be inclined or tempted to go out and develop new recipes. So one has this paradox of: here we have this thing which is sort of very broadly applicable and could be used in many, many different settings, yet we, in some sense, have to limit its spread to provide incentives to develop the next generation of technologies.</p>
<p>And I think if we now turn to open source, at least at first blush, it seems that this resolves Arrows paradox, because, here we have something that, on the one hand, is certainly a technology which can be broadly applied in many different settings, and yet the people who are contributing to it seem to be, at least in large part, immune from the sort of traditional concerns of saying, &quot;I need to protect my contributions here.&quot; In point of fact, in many instances it seems the more widely spread a project is, the more people want to contribute to it, because of the recognition they'll get from it.</p>
<p>So, in a way, the power and the appeal of open source is that it essentially allows a new technology to be developed without the traditional mechanisms of intellectual property protection and keeping these things closed in, as we so often see among new technologies. So it is a fascinating area, and certainly one which is extremely intriguing for economists to think and write about.</p>
<p>Now, in terms of the methodology of the book, I think we're at a very early stage in terms of understanding open source. And in particular, I think, if you think back about alternative methods of developing technologies, like patents, there have been literally thousands of both empirical and theoretical articles looking at this. So, I very much feel that this book is a first look at a large and complicated phenomenon.</p>
<p>What we largely did is combine both careful thought, and particularly taking some frameworks developed elsewhere in economics, and applying them to the open source context, with the development of some new data, and particularly two surveys we did, of both software developers and software users, to understand the extent to which they were involved with open source software. So those were really the two key elements or two key aspects of the recipes. We also did a number of case studies to better understand the issues in the field.</p>
<p><strong>Viv</strong>: Would you say it's fair to say that the most successful examples of open source software are those that have been originally conceived and designed by one person who has then opened out the idea for others to contribute to? And if that's the case, do you think that this will continue to be so, or could we start to see genuinely collective innovations in open source development?</p>
<p><strong>Josh</strong>: I guess I would say that it's clear that an open source project needs structure to be successful. When you look at the successful ones: well disciplined, well organized efforts. And it's certainly clear as well that there is an important role for a charismatic leadership, the kind of leader who, when there is a fork in the road, when there's two different approaches and feelings are running high, can step in in that kind of situation and say, &quot;Well, we could do it this way, we could do it that way. There's a good case for going either way, but we're going to do A. And I think we should all follow this way because we're better off all sticking together than being in a situation where we splinter and end up with two much smaller projects.&quot; So I think that leadership is something that's very important. Now, it's interesting, when you look across the groups, that I think you can see a variety of different models there. Certainly, you'd point to Linux as a sort of classic example of a charismatic leader having very much put their initial stamp on the project and then institutionalizing it. I guess you could also point to some other projects, like Apache, which became institutionalized much earlier in their development and which were less reliant on one key leader in the way that, for instance, Linux was.</p>
<p>So I think there's certainly not a one size fits all element, but the importance of having a clear structure and having a clear leadership process is a certainly very essential element of the success of these projects. Going forward, I think it's a really interesting question, and I think you could imagine it going either way, becoming, as you say, more corporatized, or else still having a sort of important role for entrepreneurs in terms of creating these projects.</p>
<p><strong>Viv</strong>: You touched a little earlier on the issue of patenting. I'd like to ask you a little bit more about that now, and particularly how open source software code is protected from theft from companies who may want to profit from integrating the code into their products. We've recently seen, for example, how this can create quite significant problems, and even major lawsuits. I'm thinking here in particular of the recent Oracle versus Google case. First of all, it's important to emphasize that the treatment of corporations taking, or others taking, software and essentially privatizing it very much varies with the license that's being used. For instance, if you look at the boot up of a typical Apple device, they will acknowledge in the lines of code that stream in front of you as it boots up, that they've got BSD software. Which refers to an operating system often known as the Berkeley Software Distribution, which is essentially licensed under a relatively permissive license that allows companies to essentially privatize it to use the code, as long as they acknowledge that they are doing so in the program itself.</p>
<p>Certainly in some cases, there has been a deliberate decision made to make the software available under permissive licenses which allows corporations to use it. But, certainly the most famous of the open source licenses, the general public license or GPO, is a much more restrictive license and particularly has the ability or the requirement that companies cannot, except under very, very limited circumstances, essentially take the code and incorporate it into their commercial products. Unless they're willing to essentially take their commercial product and open up all the source code to the public and make it available at nominal cost, which of course the typical company would be very unwilling to do.</p>
<p>Now, as you can imagine, this does pose a variety of issues. There&rsquo;s clearly the issue around observability &hellip; even if a company does this kind of process of illegally appropriating some software code, is it going to be possible for people to see it. In other words, how difficult is it to be policed.</p>
<p>This, of course, is not unique to open source. It is also a problem with process patents as well, where essentially you may see that a competitor is producing a chemical and you may think that it's infringing your process patent. But unless you can get inside the factory, it may be very hard to prove it.</p>
<p>The other issue, of course, is one of litigation. This has been an area that still relatively unproven, unlike patent law where patents have been litigated for literally centuries. This is a very new area of the law. And in most cases where there have been disputes like this, they've ended up in settlements.</p>
<p>So really understanding the scope of these licenses, where the boundaries are drawn and so forth, still remains a little bit of a murky area with some very real challenges.</p>
<p><strong>Viv</strong>: I guess some of those settlements have been settled by trading patents. In the case of Oracle and Google, for example, Google have far less patents at their disposal than Oracle.</p>
<p><strong>Josh</strong>: Right. Certainly cross licenses are a common way many of these disputes do get settled.</p>
<p><strong>Viv</strong>: What does your research reveal about the compatibility of open source and proprietary software, and to what extent would you say that firms are co mingling the two?</p>
<p><strong>Josh</strong>: I think this was certainly one of the big surprises for us in terms of what we found in our research. Which was really across the board, the extent to which there was co mingling between the commercial and open source software was really quite striking. This was certainly true among users. That essentially in many cases, users would have when you looked at the various scales of corporations, the governments that we talked to had very mixed set of applications on their machines, including both open source and proprietary. But what's even more surprising was the extent to which one saw the mix on the part of developers. That in many, many instances, we saw firms basically doing both open source and commercial development, often with closely related products. I think we all knew of examples along these lines.</p>
<p>For instance, when you walk to the offices of something like Sendmail and there's basically right next to, essentially both the commercial and the open source project living right next to each other. But the extent of that co mingling, particularly on the developer side, was really quite striking and certainly very much came through both the surveys as well as the studies.</p>
<p><strong>Viv</strong>: What do you think the implications of open source software are for low and middle income countries? Do you think it helps the sector grow more rapidly in those countries, allowing them to catch up more quickly with developed countries?</p>
<p><strong>Josh</strong>: I think that you can say that the answer is undoubtedly the potential benefits of open source software in developing countries are going to be substantial. Essentially, it clearly is able to - given not always our large software companies are able to price discriminate in a way that price at a level where users can afford them - it certainly provides an avenue for broader diffusion of legal software in a very attractive way. You also have the appeal of the training aspect - that people can actually get into the software code and actually work with it. Which may position people for opportunities down the road, and so forth. At the same time, there is a little bit of a caution which is that I am not... I don't think we're comfortable with some of the policies which have sought to really help table in a way that very dramatically favor open source vis a vis proprietary software.</p>
<p>I think when we look, given what we see in the world out there which is this really strong co evolution between the two forms of software, it probably shouldn't be seen as &ldquo;either or&rdquo; but as&rdquo; and&rdquo;, where these two forms of development are really complimentary with each other rather than being substitute.</p>
<p><strong>Viv</strong>: I guess you're answering the question I was going to ask next. That is, that one of the questions posed in the book is how governments can develop a policy framework that enables competitive interactions between open source and proprietary software in a way that boosts efficiency and innovation. What is the answer to that?</p>
<p><strong>Josh</strong>: Well, as I was sort of hinting at, it probably does not involve very one sided things; &ldquo;we're just going to favor this kind of software or the other kind of software&rdquo;. The crucial elements seem to really lie, certainly in terms of allowing and encouraging the development of standards which are where the specifications are quite open and which as a result, allow the sharing of technologies and allow software to work with each other, regardless of the corporation or the project that originates it. Of course, keeping a firm eye on antitrust considerations to ensure that essentially an established player isn't somehow shutting the door or making it very hard for competitors, whether commercial or open source, to get a foothold in the environment.</p>
<p>My sense is that if there is a right, the appropriate ground rules that in many instances, we'll see open source take off and grow in many of these economies without any added help in a way that will really be quite beneficial to the economy.</p>
<p><strong>Viv</strong>: What would you say are the main messages from your research for firms and for the research community more generally?</p>
<p><strong>Josh</strong>: Let's start with the research community because I think there, the lesson is pretty clear that there has got to be very much a focus on this area in terms of understanding these issues. In particular, while there has been a variety of stuff looking at motivation of contributors and open source project and so forth, the interaction between the commercial and the commercial and open source world remains very, very poorly understood. It's also clear that our understanding has largely been derived from looking at the experience from the US and to a certain extent, Western Europe, and that much more work needs to be done in terms of understanding the experience of open source in emerging economies. What both the barriers and the opportunities that it poses are.</p>
<p>In terms of firms, I guess the most important thing I would emphasize is the extent to which open source has to be seen as a really critical, strategic decision and thinking about how it fits in with the commercial strategy.</p>
<p>Now, for the very largest of software firm, this may be nothing new. But, for many other firms, it seems often that open source strategy is almost an afterthought and hasn't really gotten the prominent, key strategic variable that it really deserves.</p>
<p><strong>Viv</strong>: Josh Lerner, thanks very much for talking to us today.</p>
<p>: I really appreciate the invitation. Thanks again.</p>

Topics:  Competition policy Productivity and Innovation

Tags:  intellectual property rights, Competition policy, Open source software, patents technology

See also

The Comingled Code: Open Source Software and Economic Development by Josh Lerner and Mark Schankerman, at


The Economist review of The Comingled Code

Jacob H. Schiff Professor of Investment Banking, Harvard Business School


CEPR Policy Research