Economics and Open Source Projects

david_christie writes "Dan Gillmor has a piece on the economist Yochai Benkler's paper "Coase's Penguin, or Linux and the Nature of the Firm" which examines open source projects asan example of an emerging general model of economic behavior that is neither market nor company based. A previous version of the paper was noted in slashdot back in October, but it's been revised for upcoming publication in the Yale Law Review and is well worth a second look. Benkler attempts to explain why open source projects succeed, without falling back on theories about the special nature of software projects or hacker culture. He suggests that more general economic principles are at work, which are displacing the traditional motivations (market prices and employee relationships) that economists use to quantify individual behavior. If he's right the open source model could spread to other forms of creative work where the output is information or culture (music production comes to mind). The author thinks deeply about the information flows characterizing collaborative projects like free software development ("commons-based peer production"). That distinguishes this paper from the usual economist mumbo-jumbo about price points and such. Like Larry Lessig on the legal side of things, this is a guy who gets it and has thought deeply about how his field relates to it."

Comments

Open-source music?

[MaestroSartori]

But then, most open source programmers are, I would guess, full-time programmers. Which helps pay for all those neat toys. None of the professional musicians I know (and I know quite a few, session & orchestral players) would record music and give it away.

What does that leave us? Amateur musicians like myself pimping their home-grown stuff. Which in some cases will be as good as or better than the pros, but the vast majority of it will be as cruddy as all those non-updated open source projects on Sourceforge...

Re:Open-source music?

[Damek]

You know, people are always saying that the problem with amateur (i.e. non-corporate) music/art/software/whatever is that the vast majority of it is cruddy. Pick any artist from MP3.com at random, or any project from Sourceforge at random, and chances are the music will be pretty cruddy, or the project will be lifeless...

But you know what all this is? It's called CHOICE. And it has always been there. Corporate sponsorship of creativity and ideas makes it easy for people who don't like to deal with choice to find a couple of good things here and there. But it doesn't particularly foster the best creativity or ideas, only the easiest to market. It doesn't make the world a better place, it only makes shareholders and CEOs richer.

Choice is good, but it does take effort to appreciate.

As the parent acknowledged, in some cases amateur stuff is as good as or better than the pros. People have become so used to being force-fed their ideas that they can't believe that the good ones might also rise to the top in a truly free idea market. But they will - in a free idea market, without corporate sponsorship, the most visible art and ideas will be the ones you hear about most from other people, some of whom you will trust more than others. Much like internet memes, a good artist will quickly and easily become well known in an absense of corporate interest. Unfortunately, as things stand now, corporate interests drown out amateur efforts.

Re:Open-source music?

[SirSlud]

Making music, is, by definition, and Open Source process.

Open Source. Free as in speech, not free as in beer (or free as in CDs, if that helps.)

Please know what you are talking about before you comment on it.

It's not about giving it away, its about your music being trasparent. Your musician friends are Open Source friendly whether they like it or not - I can go to a show, transcibe their music, make some modifications, alter it. Now, I can't sell it, but that doesn't mean that I'm not able to know how the music was built. I can find out just by listening to it.

Compilers are like (neccessary) noise-makers .. you obfuscate the 'sound' of the software such that I can't transcribe it and use it as imput/inspiration to another creative work. In music, the 'source' and the 'compiled version' are usually, in practice, synonymous.

And the reason your musician friends dont give music away for free is because if anybody gave all their product away for free, nobody would make a living. Duh. That doesn't mean that they should try and prevent people from knowing how their sound/songs were built, because that kind of information is what *drives* culture. No artist has ever created anything original - art is just a history of creative "patches" to others' work.

This is Market Economics, plain and simple.

[BitGeek]

Open source is the definition of Market Economics. It does not need its own theory- it proves the Marekt theory in the most divergent context imaginable.

If you have an idea and you open source it, you get free engineering. People contribute their engineering and get the utility of yours and others in response. This is a free exchange of value in a free market.

IF your terms suck, or you change direction in mid stream, the others are free to leave your project and start their own, or go along with you if they like your direction.

Linux is a consistant linus kernel because people like linus's direction-- not because he "owns" it. That is direct market feed back to linus.

If you are selling your product to people who are contributing nothing but money to the process, and are just using it, then you are in the traditional software model. but Open source works here as well-- you can incorporate open source into your product and leverage others work to make more money off of your work. At the same time, the MARKET FORCES (not the GPL) will force you to contribute your improvements back to the community. And finally, you're not profiting unfairly from others work because the others contributed freely, and were compensated... and also can sell the results in the market place against you, so if your product is PURELY reselling open source, then you'll loose the inevitable price war-- its hard to beat free.

If, however, you actually add value to the product, on top of the open source, then you CAN charge for that value and everybody wins-- your customer gets a better product with more features and testing than you could otherwise do yourself, the other os developers get the benefit of your improvements and you get more money for selling a better product that cost you less to develop.

This is all free market economics.

The differnce between open source (free market) and communism is that under communism you are forced to work for the state against your will. Here in america, we are %50 forced to work for the state against your will, but they cleverly let us work for private companies and only took the product of half our work in taxes (fees, etc. And yes, last time I did my taxes, my total payments to the state were over %50, and I'm in a medium tax bracket.)

Even with the GPL, however, you are not compelled to work for the "State"... you can choose to not use the GPL for your code, or go make code to replace whats' in the GPL, or just use the gPL code and not change it. ITs a free market of licenses.

Since the government isn't (Yet) regulating software, the emergence of the open source movement proves that free markets work-- whenever one company gets to monopolistic, under free market theory, competitors emerge. Lots of competitors have emerged to Microsoft, but Open Source is the first one to really sustain a battle and change the terms of the war.

As long as the state doesn't mandate Microsoft control (As they may wit palladium) the free market will prevail and the products that offer the most utility value will succeed. For a long time Microsoft was able to distort the market with anti-market means, and also provide sufficent value to have locked up much of the market--- a great example of the market under a lot of stress.

But the emergence of the free market, the resurgence of a variety of MS competitors- from Sun to Apple to IBM to me, shows that the free market does work-- even with the governments help for microsoft, the market is beating them.

Not financially right now, but in terms of brainshare and technology, MS is currently loosing. In order to win, or even survive, they will have to deliver better value for their prices... and since open source software is free, the competition is stiff.

So, no, there isn't a new theory needed-- The Free market works and has been validated, yet again, by opensource. (So stop voting republican or democrat and become a libertarian already.)

Re:This is Market Economics, plain and simple.

[MattJ]

"Open source is the definition of Market Economics. It does not need its own theory-"

It might be *possible* to describe the Open Source movement in terms of standard market forces, but it is not parsimonious.

I wonder whether you've read Benkler's paper. It's long, and as he says, he spends "substantial space in this article explaining why peer-production processes appear to respond mostly to cues other than price signals." *That* is clearly one aspect that removes this from the realm of market forces. If you want to argue tautologies about how people always do what's in their best interests, etc., go ahead. But you're not describing standard market economic models unless actions are motivated by profits and regulated by prices. Free-as-in-beer software may be used by some companies to save money, but that's not why most programmers create it.

"[T]he emergence of the open source movement proves that free markets work-- whenever one company gets to monopolistic, under free market theory, competitors emerge. Lots of competitors have emerged to Microsoft, but Open Source is the first one to really sustain a battle and change the terms of the war."

Are you serious? This proves the opposite. It proves that in an unregulated market, monopolists can emerge that can't be dislodged by any competing firm. Only people who are not motivated by profits have a chance, and only by creating a product that has zero marginal costs (digital data) and zero price. Those assumptions overturn all of the first chapter of Econ 101.

Market forces do not explain all aspects of human behavior. Why do some libertarians get so defensive when that's pointed out?

Re:This is Market Economics, plain and simple.

[William Tanksley]

You're right and you're wrong.

Yes, classical liberal economics (the a priori discipline) encompasses open source, and explains its success quite well; but at the same time, there are broad financial models within that discipline which fail to explain open source, even though open source has financial consequences. This paper examines open source with the intent of fitting it into some of those models.

I'm using the term 'economics' to describe the science which examines the results of human actions and choices, while 'finance' is the subcategory of that which examines only intercomparable human actions and choices, and which uses money as the instrument of comparison. A better choice of words would be to use 'praxeology' for the first term, and 'economics' for the second.

See von Mises' text 'Human Action' for MANY more details. In that text he covers praxeology in general and economics in specific. Highly recommended, and available as a free ebook on mises.org.

Oh, and although Mises is the poster child of the Libertarians, he wasn't one and couldn't have been one (there weren't any at the time). So check your prejudices at the door. A lot of what he says is valid even under non-Libertarian assumptions, and much of the rest is valid as long as you don't try to apply it too far (as most Libertarians do).

Note the big L -- I'm a libertarian, but I'm not a Libertarian.

-Billy

I believe what he is saying...wrt coase.

[Bobzibub]

Coase described the size of the firm within the market, and claimed that with competition it would gravitate to the most efficient size.

So if one can call, for instance, the Linux kernel folks a firm, their fixed costs are fixed, but their marginal costs are zero. (barring Linus's scaling issues of course).

Marginal costs being the added cost of each extra unit of "firm size".

So organizations will scale to be quite big over the 'net because of low marginal costs.

And this is what we have witnessed.

I hope the paper has graphs. I like graphs.
-b

Same ol' economy, new face.

[MarvinMouse]

Open-Source doesn't really attack corporatism as it does attack Mass-Production Media.

Software, Music, Movies, Books, etc. Are all money makers based on the fact that they can mass produce a product that people will pay for. On an individual basis, the $16 or so made off of a single CD, book or movie doesn't matter unless they can product millions of these $16 products and sell them.

With the internet though, it has opened the possibility of distribution of IP products for free or near free prices. Thus the business model of these IP companies is not applicable anymore without forcing the public to play by their rules by legislating laws into place.

The Open Source Movement has a weird effect of showing what happens when people can produce the same products and share it with everyone else, allowing them to improve on it. Before hte internet, when I coded a small program I could only share it among my close friends easily. Now I can share it with everyone, and if it is useful, everyone can contribute to it.

In a way it is like the folksongs from way back. Somebody thought it up, and shared it among his friends and family, or in performance, thus making his money from his actual work and not a 'photocopy' of his work. Then other musicians got it, and would play with it, producing even better music. Some of the great classical pieces are basically open source folk songs that have been improved upon by the masters. Since folk songs could easily spread by word of mouth, and didn't cost anything to spread, these songs became the equivalent of Open Source Music. Everyone was able to enjoy it, and no one had to pay anyone for the right to hear, see, learn or play the song themselves.

Now, we can pass programs, books, poetry and more using the internet and allow others who may be better (may be worse) then us to improve on them and create a better product in the long run. It's not a new economic model, it's just an old one coming back in a new form.

I heard once that people don't like change, they like things to remain the same as long as possible. I think it would be more correct to say people with power and money don't like change, and will go to great lengths to prevent it.

Some interesting thoughts.

Re:Same ol' economy, new face.

[platypus]

I think you should differentiate two factors here:

- Obsolescence of classical distribution channels

This is indeed the problem of big media corporations and covers the distribution of the good itself and marketing logistics. The importance of both of them has suffered quite a bit since the internet emerged. I venture a guess the the biggest fear of big media might be not the fact that people can copy and distribute their stuff, but that artist in the long term will loose their dependence on a company to get their stuff to the people.

- Collaborative work on an open source product (whatever it may be).

I think the latter mainly will affect software production, because the process of creating "art" does not inherently profit from collaboration. Would Bach's work really be any better if he had done it with a couple of fellows over the internet? Would it really improve if someone took it and came out with Toccata & Fuge 1.1?
This would be possible today, because there is no copyright etc. on Bachs work, but nobody seems to think it's a good idea.

No Need To Turn Economic Theory Upside-Down

[istartedi]

There's no need to mess with economic theory to explain Open Source. There's nothing new there. Each programmer, as a rational operator, contributes for a number of possible reasons. For example, they may value creative control and consulting opportunities more than they value a salary. In other words, someone who waits tables at night and codes for free during the day isn't necessarily a radical leftwing crackpot--as long as they are doing it for the future hope of consulting $$$ and/or the right to maintain control of their work (witness the not insignificant number of people who have un-Opened their work).

Corporate sponsors have rational reasons too. IBM doesn't support Linux to join the lovefest. They think it's better for some applications, they want to offer consulting for it, they don't like being tied to a proprietary vendor, etc. Any contributions they make are made because they realize it's the price of doing business under the Linux model--they would lose business due to bad PR if they didn't.

As for software being "special", there isn't any need to appeal to such an idea. Coffee is a good example. Generic not-so-tasty coffee is often given away in waiting rooms, hotel lobbies, places like that. Same deal with those little mints on pillows. Same deal with free samples at the grocery store (I've known people who make a meal of free samples on Saturdays at Fresh Fields). In all of these cases, software included, there is a rational economic model that has given rise to support for some free riders. People still have to pay for these products. The payers have deemed that they are better off paying the free riders, much as society has decided that some taxation is better than none.

The OSS model could be regarded as a "natural tax". Once again, there is nothing irrational about it. Advocates just have to realize that neither model is "superior". The free market sometimes moves us towards paying for goods directly. Other times it moves us towards indirect payment (somebody pays for OSS, because TANSTAAFL).

Of course, I doubt that advocates will stop advocating. There is a demand for politics just like anything else, and they supply it. It's just that I hate to see it when the supply-demand for politics pushes the supply-demand for other things out of equilibrium.

Re:No Need To Turn Economic Theory Upside-Down

[mjh]

There's no need to mess with economic theory to explain Open Source.

No one is "messing with" economic theory. Benkler isn't saying that traditional economic theories are wrong. Just that they're not sufficient to explain a newly witnessed phenomenon: Open Source/Free Software.

I provide free software because I have no reason not to. I don't provide it for the hopes of some future consulting dollars. I don't provide it because I wish to maintain control of my work. I use the GPL because I don't want someone to take work that I've already done and make me pay for it in some slightly different form.

For Microsoft and Sun and ..., they only reason they produce software is to sell it. They're not actually concerned about using it to solve a problem that they are experiencing. They're not scratching their own itches. They're guessing at the itch that someone else has and trying to produce a backscratcher that will reach. But for me finished software is simply a scratched itch. I do NOT look at it as an oppurtunity to sell a backscratcher to others with the same itch. If they want it, fine. I've already solved my problem.

But if you want to get some idea of what I really want by releasing my software, it's this. I want access to the backscratchers that others have produced. But not only that with the ability to easily modify their backscratcher to reach my specific itch. This economy is not measured in dollars. It's measured in software, and that, I think is the point of Benkler's article. If you try to measure open source/free software using the traditional mechanisms, you're going to have to start looking for the money. For the *vast* majority of people who produce open source/free software, their just isn't any money involved. (IBM, Red Hat notwithstanding. They provide lots of opensource/free software, but not anywhere near the majority.)

So Benkler is just saying that there are other "enlightened self interest" factors that are involved. Not just money. Since all of the traditional economic production models have centered around money, they're not sufficient.

Re:No Need To Turn Economic Theory Upside-Down

[Anonymous Coward]

There is one advantage that OSS has over other kinds of software and is what I think gives it that aura of invincibility.
It can't be killed easily. Its distribution can't be stopped by bankrupting anyone in particular. It can't be sued out of existence because there is no one to sue.

OSS is equivalent to a meme and will continue to thrive as long as it is considered a valuable meme.

The net result of this is that there is no way for a corporation to snuff out OSS completely except by making it irrelevant through innovation - producing a better meme.

The economics are the same, it's the currency that's changed.

Is software a cottage industry?

[jlowery]

Perhaps we're seeing the reemergence of a cottage industry in software development.

Back in previous centuries, whole villages of craftsmen and women would do finishing work on mass-produced pieces that were then sold by a large retail company. The garment industry still operates this way in many instances.

As long as software remains a craft rather than a formal engineering discipline (it has elements of both, but each software project is pretty much unique to this day), then the economics of software will probably most resemble the crafts industry rather than industries based on mass production.

Re:He forgot something basic

[mccalli]

. In that community, as people realize there is no reward for contributing or punishment for not contributing, they lower their contributions as time goes on.

Hmm. Not really applicable to open source though, is it? I'd agree that this is the conventional model. However, no-one ever gave anything back in the first place to the developer, so whatever their incentive to start writing was - that incentive still exists.

Cheers,
Ian

Re:He forgot something basic

[merkel]

No, this is wrong...

The freeloader problem really manifests itself only in the Tragedy of the Commons. That is, freeloaders are only a problem when resources are scarce.

If we assume the marginal cost of distributing free software is 0 (which is probably true for the developer as there a many sites that will mirror popular software distributions), then why does it matter if 100 or 1000 or 1 million people download it?

I think most open source developers would be happier to have a popular application with 10mm freeloaders, rather than pulling a Bill Gates and bitching about all the ungrateful pirates out there.

The real viability issue for open source is whether it is possible to maintain a stable base of developers for an application -- not the number of freeriders.