Rishab Ayer Ghosh  
 
 
 

Good afternoon, today we are going to talk a lot about the economics of open source, which is essentially a software model. It's interesting to see how people who are programmers or technical people have suddenly become economists and that's interesting for me personally, because I used to write in C++ and Perl, and now I discover that it's easier to discuss many of my ideas in English. So I do a lot of that of course, but I think it's the issue of what makes an open source economy, or what is economics or what is of economic value, that is very important right now from two points of view.  

One is that lots of companies and others in the traditional economic world are worried about the things that are happening without money. Because economics has generally been understood as how to make money. And there are lots of people who may or may not be making money right now, but they're producing things and they are doing a lot of work, so that seems interesting, and whether or not they make money is worrisome for people who don't understand it. Among those who don't understand it are governments, and that's partly because they don't know how to account for it. There is a story about rural America in the early twentieth century where people used to stitch their own clothes and then Sears came out with their catalog by mail order clothes, and so people started ordering their clothes by mail order and paying for it instead of stitching them. Eventually in the long run people must have bought more clothes than they used to stitch for themselves and the division of labor must have lead to the greater increased production of clothes and all that, but immediately I think people who stitched maybe three shirts a year would buy three shirts a year, and there was no increased production. However there was an increase in the amount of economic, or measurable economic activity, because people were paying for something that they were earlier not paying for. So without any increase in production there was an apparent increase in economic activity. That's clearly a case where something was not treated as economic when in fact it was, because the same amount of production was going on but it seemed as if there was a smaller economy.  

A similar thing is happening in the opposite direction now when a lot of people spend time on line doing all sorts of things. Now we don't have to get into arguments over whether spending time in a chat group on AOL is a productive activity, but it's no more or less productive then watching a soap opera on TV, which is clearly part of the economy. But when you spend time in a chat room on AOL or when you answer questions on music in a newsgroup or mailing list or when you write Linux -- that's not being measured as part of the economy. And since a lot of people are doing that, and more and more people are spending more and more time online, they are disappearing off of the economic face of the world, as it were. So, nobody can see them, you don't know where they are -- they are becoming invisible. And that won't actually lead to a drop in international GDP figures or anything, but clearly there is a lot of economic activity going on which is not measured. Now that's the reason, essentially, why there is so much interest from so many different sides on the economics of what is free on the Internet.  

I'd like to answer that by posing a multiple choice question. There are say, a hundred million people on the Internet and they do a lot of things, they produce a lot of things, they interact, they consume a lot of things, they create free web pages, they spend time answering questions, they write millions and millions of lines of code which they don't get paid for. And there are three possible reasons for this. One is that this is a revolution and that this is a whole new world that has been created, or has spontaneously emerged, and everybody over there acts in a different way, they are some wonderful new Internet being, and they are very generous, and very open about things, and they give away all their work without getting anything in return. And this will be the basis of a wonderful new society. That's one view.  

Another view is that, well, these people aren't charging for it, and we can't put any price tags on what they do, therefore it must be work-less, therefore they are all hobbyists and eventually they will need to pay for their food or pay for their houses or whatever else they have to do, and as they all grow up, they will eventually start charging money for things, and so its an unsustainable system, and since its an unsustainable system and we don't understand it, lets ignore it. That's a pretty common view, especially among economists. Now both these views assume that the motive for doing this goes against the traditional economic motive of people. In economics you don't deal with real people, you deal with rational economic actors, who are not people, but they are a model of people. And these two views assume that everything that goes on free on the Internet is not with rational economic actors.  

Now the third view is when you do things for 'free' on the Internet, you are not doing it for free as in your not just giving things away for the good of the world or out of an affection for humanity, or the people in your Linux developer group, many of whom you probably don't know, or certainly don't know when you start off. But you are doing it because you are aware that you are creating value, and you are aware that you are getting value in return. That is my view, and clearly I think that's the right view, and that also is different from the other two views in that it assumes that everybody is a rational academic actor in this world of free Internet products and services. You don't have to get money in return for what you do, to be a rational economic actor. You are getting something for what you do, the fact that you are getting something decides what you do and you choose to write a free web server rather than a free open source version of Microsoft Word for a specific reason. And these choices are made rationally, and that's what makes it possible to call it an economy I think, and possible to call it an economy in a fairly traditional sense of the term: the rational allocation of resources among people. And I believe that when you put someone who thinks rationally, normally in front of a computer keyboard, you don't turn him or her into a saint or a crazy communist, but all you are providing is a new media of expression of the same type of human behavior. The Internet is not a new world, and the people who use it are the same people who live in brickspace as in cyberspace. So the Internet is not changing how people are, its changing, or allowing them to express themselves in a different way. There may be some things that they can express on the Internet which they can't otherwise, and that goes for economic activity as with everything else.  

The economics of the Internet then, is essentially the economics. It's not Internet economics, or knowledge economics or anything, it's just economics. Except that its got to do with a set of activities that happen on the Internet, because that's where people express themselves in the area we are talking about. With the focus on open source software, because it's easier to measure, it's easier to see that Linux, with six million copies as an installed base, which is about the same as Windows NT, and Bill Gates saying that he's very worried about Apache, it's easier to take open source software seriously, because of all these people with money talking about it.  

But it's not necessarily software alone that is of value, the very notion of value is something that is interpreted variously in the real economic world. Anything that anybody will pay for is thought valuable, and there are lots of people who think that it's not valuable to pay for, well, Soap Operas on TV, but people do that. Similarly, on the Internet, there are a lot of things that are treated or would normally be treated as not valuable. They would be treated -- those chat rooms on AOL, or those newsgroups -- would be treated as non-economic things. They would be treated as hobbies. But in fact, because people value them, even if they don't pay for them, they are economic. I make the distinction between what is economic and what is not economic only in that economics is a tool, economics is a way of doing something and understanding something about the way people interact. There are other ways of understanding how people interact, but economics is particularly good at understanding interactions between people who have no other ties. There was a study done by one of the UN agencies (I think this was about 15 years ago) on how women in homes don't get paid for taking care of children or managing a house, or whatever. And if you set a price tag to it it would be something like $100,000 a year. Now if you actually had to pay them a hundred thousand dollars a year there wouldn't be that much money and it wouldn't work, that was a valuation given. That does not mean that we immediately have a new economics of women in the seventies in homes, because they are not seeing themselves as doing something for a hundred thousand dollars. Because there are other things such as family ties and similarly there are other activities such as living in a community, what you do for your neighbor, or what you do for your friend is not necessarily treated best economically, because there are other motivations for it which have nothing to do with rational behavior.  

But on the Internet, the interesting thing is that friendships -- there are of course friendships within some programmers -- and many people in the open source or any developer community end up having bonds that are much stronger than the software they write alone would allow for. What initially brings them together is the work they are doing. And they might not know each other at all so then we have to look at something beyond traditional non-economic ties such as family-ties, neighbor-ties, things like that. And the only thing that works for the anonymous cooperative creation of value is economics. That's essentially why you need an economics, an economic way of looking at these things.  

The economic model that I have developed for looking at this is something that I call a "Cooking Pot Market" model because I drew an analogy to a hypothetical tribal cooking pot where people, which is not a barter economy, its where I say I have a chicken and you have a potato, and instead of bartering economy where I would give you the chicken and you would give me the potato, and we would still have a chicken and a potato, we put both of them into a cooking pot and share the stew. Now in a real brick space economy with real chickens and potatos you would have the same amount of produce at the end of it. So you might have the added value of the cooking, there is still only as much stew as the chickens and potatos that went into it. And therefore the division of all these things has to be according to what went in. So if I put in two chickens and you put in one potato, I would get, if I was being a rational economic actor, want two thirds of the stew (assuming chickens and potatos are of equal value). And if I don't then I am a communist.  

I don't know too many free software programmers who are communists and most of them are rather against the idea of giving things away. Most programmers I know are pretty strongly libertarian or anarchist generally somewhat right of center in many of their views. They are not even what you would call team workers or collaboratorive people, generally prefer to work on their own on a little bit of project, which somehow fits in with what everybody else is doing and that's generally how most open source projects work, because you don't actually have a structured team in the same way as you do which is a company (which is in fact supposed to be capitalist), you end up in a company, paying a number of programmers and telling them to share what they do and put everything together. Whereas in the free software community, which is supposedly collaborative, everybody is very conscious of what they are doing as opposed to what others are doing and they are very protective about their territory, as it were. And, that is something that I think, makes it all the more important to see how it's possible to have, to see how far the analogies can go. So this cooking pot sounds very communal and friendly, but it would be in a real world, it would be altruist. In order to produce different amounts and allow others to take more out of it than they put in, you have to be altruist because you are giving something away. That generally does not work, which is why you don't see the cooking pot markets in the real world generally.  

On the Internet the reason it works is because the cooking pot multiplies things, and it multiplies things an infinite amount of times. When I create a product, when I write an article or write a hundred thousand lines of code, it costs me the same whether one other person uses it, or a hundred other people use it, or six million other people use it. So, if I'm Microsoft, I want to make more money if six million other people use it, rather than if a hundred people use it. Well that's because I actually think that what I have written is worth 20 billion dollars. And the only way I can make twenty billion dollars is by charging a hundred dollars for each copy. Which does not make economic sense because each copy does not actually cost that much, what costs a lot is a single act of creation and after that you've got distribution, which on the Internet does not cost. So given that the marginal cost of reproduction is zero or near zero and the marginal cost of creation is very high because there is only one creation of any specific product, the cost structure is very different on the Internet, or in any knowledge oriented system from, say, the production of each car, where you can actually charge for each car. Whereas you can't actually charge for each copy of something on the Internet, it doesn't make economic sense, and that's what I call the problem of infinity. And there is a solution, which is the Cooking Pot market.  

When I create one product, I can give it away to a million people. But I am not giving it for nothing. When I take one of those free products that are available on the Internet, I am not taking it for nothing either, there is an exchange. The exchange is that when I write this article, say on Cooking Pot Markets, and I give it away -- so called give it away -- I've only given away one article. In exchange for that one article, I have access to several hundreds of other articles. I give away a hundred copies of one article, and I get one copy of a hundred articles. To me, it does not matter that I gave away a hundred copies or a million copies, because I only wrote it once, I didn't write it a hundred times. Or I hope I didn't. But when I get copies of your articles or your lines of code or whatever else you happen to do that interests me, even if lots of other people are getting access to that, it doesn't matter to me. I get value out of your line of code even if you are giving the same line of code to lots of other people. I get value out of it simply because to me it's only my copy that matters.  

So the problem of infinity, where the cost of production and the cost of distribution has split up so differently, over here becomes the benefit of consumption of variety. For me, what is valuable as a consumer is access to diversity and as a producer is originality, so what costs me to produce is not volume, but producing original things and I need to be paid for producing in some way for producing original things, and what I get in return is not one original thing or many copies of one original thing, what I get in return is access to a diverse array of things. This works whether you talk about individual contributors to a discussion group who are willing to answer one question because they are also able to ask and get answers to several other questions. This works in the Internet as a whole, and it works in the open source community where you write one program or one utility or one little bit of something and in exchange for that you also get the right to run Linux on your system.  

I was talking with one of the more traditional economists who tends to think of open source as a hobby and Linux is something that will disappear, by saying that if you want to put a price tag on the copies of Linux that are sold -- or if I were a Linux developer the amount of work I put into it -- if you want to put a price tag on that and if you say that well I'm giving away five thousand dollars a week of my productive time, then you also have to look at the other side of the equation. In exchange for those five thousand dollars of productive time in writing the device driver for a sound card or something for Linux, in exchange for that, what I get is an operating system, a server, a developer environment, a debugger, libraries and all sorts of other things. Now say we were to price these at the commercial prices for any Unix development system, it would run into several thousand dollars. What every Unix developer is, or what every programmer is in the open source community is a self employed software company. A one person software company. And if you were to actually do accounts for that one person software company, and put a price tag on what that person was creating and say that he's doing that much work he should be paid this much, you also have to do it the other way around and say, he's using all this software, which he should be paying for. And you will discover that he will probably be running into losses for several years if you did that. So it works out better for him to write five thousand dollars worth of productive code every week without getting paid the five thousand dollars because he also does not have to pay for any of the development tools he uses.  

When you start putting money price tags on things, you have to have an actual transaction. On the Internet, or in this cooking pot model, you don't have a transaction. It's something that I call, its a non-monetary model for implicit transactions. And the reason for that, is that in a transaction, in an exchange, whether it's monetary or whether it involves barter, you've got person A gives something to person B, person B gives something to person A -- that has to take place in one go. And what person B gives in return is usually money. That's a subset of something which I call "value flow." Person A gives something, somewhere. You call it the cooking pot market, you call it the Linux development community, you call it anything, it doesn't matter. Person A produces something, person B, person C, lots of others produce something. Person A gets something, it doesn't matter from where, person B gets something, it doesn't matter from where. And as long as for me as an individual, I am getting something which is more or less, in my mind, equivalent to what I am giving, I'm okay. If I'm getting more than I am giving, I'm making a profit. If I'm getting less than I'm giving, I'm making a loss and I'd stop doing that -- because I'm a rational economic actor, remember?  

What you have -- and this would work in terms of money as well, if I were to write articles and publish them for free and I get these checks from somewhere -- maybe Tim O'Reilly -- it doesn't really matter whether the place where I am publishing is paying for it, or someone else is paying for it. Now that in fact happens at research institutions or colleges where researchers are paid money. They aren't paid for what they actually produce, they might be paid for one thing, they might be paid for teaching, whereas what they think is more valuable is publishing papers for free. They are getting money from somewhere, and they are doing something else for free so it doesn't matter to them that they are not being paid for what they are doing, they are being paid for something else, its okay, because as far as an individual is concerned, its only the fact that the value flow is balanced that matters. This notion of value flow is something that you can develop further into a lot of areas and since we are talking about open source we'll stick to that. Where does the point come, when you are not getting enough back for what you are giving? I'm very clear that -- and here I'll use an example that's sort of on the borders of the open source community.  

I'm sure many of you have heard of the Internet Movie Database, IMDB.com. I think its now the number one movie site on the Internet, and when the Internet Movie Database started, it was a collaborative newsgroup of movie reviewers, into which came this programmer called Needham who was working at Hewlett Packard and Rob Hartell who was a Perl programmer. They both loved movies but they were not great reviewers, so what they contributed to the group in exchange for reading movie reviews was an increasingly sophisticated system to database, index, and otherwise organize the movie reviews that were there. As that got more and more complicated -- initially it was a sort of well balanced system, where these people were writing movie reviews and they were discussing them amongst themselves, while these people were writing programs and contributing and everybody felt happy about it. But as they started writing more and more and the programs became more and more sophisticated, there was a sort of hike in the curve, as it were, where they were producing more than they felt they were getting in terms of movie reviews alone. What they got of course, was a reputation, which like money in the real world acts as an index into value of the future, they got reputation that, in turn, got them good jobs in other companies because they happened to write this wonderful software. But that wasn't enough for them, they kept improving the system more and more, and spending more of their time on the movies database than on their jobs. I don't know if that was what lead them to quitting their jobs, but eventually the system became so big, and the number one site on the Internet for movies that people from everywhere were coming there. And it was no longer that well-knit balanced system anymore. They were getting reviews from some people yes, but they were adding much much more. And at the same time, their consumer base was vastly expanded. And so they were concentrating and giving far more than what they were getting. That was the time when, predictably they left their jobs, started a company and haven't yet done their IPO, but are taking advertising and presumably will become billionaires soon.  

The same thing won't happen, incidently, with Linux, because although Linux is used by six million people, and it may be used by more and more people, that concentration is not there. I've done a study on the ownership of Linux, and you find that Linus Torvalds' share of the code has been declining from a hundred percent, it's currently 9 percent, and the distribution is vast, so there is no such tremendous ratio or imbalance in the value flow where there are people contributing too much and not getting enough back. So I think, Linux is likely to be free forever, but the same need not be the case for other things. And I think it doesn't really matter because -- as long as the value flow gets balanced eventually -- whether people are getting paid for it or whether they are getting their value in other ways is up to them really. That's where I'll leave it, Thank you.