DISQUS

Braden's post is interesting but he doesn't provide enough information to settle the issue one way or the other. The one substantial piece of data -- that people follow something like a power law in how much they contribute -- is not news, and is pretty much orthogonal to our question.

To nail down our question we'll need a definition of whether a developer is "in it for the money" (to use Richard Bennett's criterial phrase). Just because a developer makes money from open source related activities does not mean they are in it for the money.

To give an extreme example, if I'm walking down the street and I pick up a quarter, I'm certainly not walking down the street "for the money".

Open source developers will often get opportunities to make money because of their reputation, specialized knowledge, social network, etc. -- just as any of us will. However this does not imply that the activities that produced that reputation, knowledge, etc. were done so the developer could make more money.

We can say that someone is "in it for the money" if they choose the option available to them that maximizes their income.

If we can agree that that is the criterion, I'll think about how to find or generate data to find what proportion of Open Source contributions are made by people who believe that is their income maximizing option. I'm pretty sure most substantial Open Source contributors are nowhere near making income maximizing choices on how to spend their time, but I'd like to take this from an intuition to a substantiated claim.

The 2005 US median income was $46,326, less than half of what Richard Bennet says is "the minimum level required for basic subsistence in the West." I guess our society is in worse shape than I thought.

People who run large funded software projects know that a lot of the time, effort and money gets spent on "doing testing, and scripts, and other low-level crap" so those are very necessary contributions.

As for Richard's hypothesis that "the developers who contribute the bulk of the important, shipping code in major FOSS projects are in it for the money": I'd be very interested to see research on that. Two examples I know about, Larry Wall, who created Perl, and Robert Love, who wrote the current Linux scheduler, don't get paid for that work, and have no way to "take it public" and cash out.

Richard replies "the plural of 'anecdote' is not 'evidence'" but I did provide evidence via the link. Specifically, drilling down a couple of levels see this
survey of open source contributors.

I note that you are not posting anonymously, Richard. Can we conclude that you expect to make money by posting comments on blogs?

Citing some developers you know who make a living from Linux headers sounds pretty anecdotal.

Noel Le writes:

[D]on't propose to eliminate software patents, repeal the DMCA and try to hasten the demise of IP firms in support of peer-production.

These are all great examples of the divergence between ownership and entrepreneurship.

The DMCA is a way some groups get the government to help them extract money from others, with essentially no innovation on their part, and no overall social value. It helps some investors and media catalog owners, and hurts entrepreneurs.

The consequences of patents and other IP protection are more mixed, but in many cases they inhibit innovation and entrepreneurship. Certainly patent trolls are a pretty clear example of the conflict -- they buy patents not to produce anything, but to sue others who do produce something. Ownership vs. entrepreneurship.

Richard Bennett writes:

Open source software isn't really something that people do for the fun of it, it's a business.

This is by no means true in general; some do have monetary incentives, others do it for entirely non-monetary reasons. There is plenty of data on this.

Open source is capitalism one step removed

This is just false. Capitalism provides returns to owners of capital. In open source, no one gets returns from ownership. Contributors may or may not get monetary returns, but even if they do the returns aren't coming to them through investment or ownership.

A major source of ideological conflict is that in non-monetary coordination, the interests of capitalists and entrepreneurs are no longer aligned.

For example, Linus Torvalds is a great entrepreneur, and his management of the Linux community has been a key factor in the the success of Linux. Entrepreneurship is essential, and is valorized in non-monetary coordination.

By contrast, there is no way a capitalist -- someone who wants to get a primarily monetary return through investment decisions -- can benefit directly from non-monetary coordination. He or she may benefit indirectly, for a while, but the whole trend of the process is against monetary profits from monetary investment.

This isn't a problem, as long as we don't become confused, and think that a falling rate of return on financial investment is a bad sign.

Unfortunately knowing Windows, I'd guess the 100,000 will be needed just to keep it running. But that's not what I wanted to talk about.

I've recently been thinking about a close relative of the broken window fallacy, I don't know if it has a name. It was very common in venture capital reasoning during the late 90s boom.

The fallacy is thinking: "That technology is going to reduce the costs / friction / labor / whatever so much it will revolutionize that market. We can make billions!" Right about the first, wrong about the second. The hottest example right now is maybe CraigsList, where automating classified ads is cutting the heart out of the local newspaper business model, and CraigsList has three employees and will probably never go public.

Of course CraigsList could make lots of money if they started "monetizing their eyeballs" (yuck), but in a competitive environment they'd eventually be forced to run at their marginal cost which is approximately zero. The reason this won't happen is another set of paradoxes I won't discuss now.

So does this fallacy have a name? I think we'll be encountering this situation more and more often...

Good post.

I think cooperative relationships (such as e.g. open source software) actually involve less state coercion than market relationships. To see this, consider the state machinery needed to enforce proprietary copyright -- investigations, prosecutions, ultimately prison. OSS makes all this moot.

More generally, any kind of large scale market system requires lots of coercive protection -- think Brinks guards, counterfitting prevention, etc. etc. To me it doesn't matter a lot whether that is provided privately or by the state -- usually it is a mix.

Of course this sort of coercive structure is worthwhile for many purposes. But my point is that if we can get the same results without coercive protection and "stealable" property, we are better off.

Noel's claim that evolution has no predictive power is addressed in this summary.

I'm not sure why Noel believes that "population thinking is different from the 'market'". Evolutionary game theory arose from evolutionary theory, and is now heavily applied in economics. A detailed study of related explanatory mechanisms was done by Sunny Auyang in Foundations of Complex System Theories (a fascinating book about the mechanisms of spontaneous order, by the way).

As for the relative success of FOSS -- the remarkable thing is that FOSS development processes are a viable way of solving some very large, complex coordination problems, without institutional enforcement mechanisms -- either bureaucracy or contracts. This was considered impossible in conventional social and economic models. Evolutionary game theory approaches do better at modeling FOSS type coordination.

I call thinking that relies on these statistical mechanisms "population thinking" and I do believe that it is hard for people to accept, but conversely that these mechanisms are far more robust and effective in a lot of domains than deterministic mechanisms.

As a curious sidelight, the standard model for market economics, the neo-classical Walrasian synthesis (actually formalized by Arrow and deBreu), uses a deterministic mechanism, tattonement, rather than statistical trading. Ironically, this makes the Walrasian model very fragile and computationally intractable, while the statistical model has more robust and extensive equilibria (which are also closer to real world data) and is computationally trivial. The statistical methods existed when the Walrasian model was developed and formalized, and the economists chose deterministic ones instead.

So apparently even theorists of naturally statistical domains, like economics, have a deep aversion to population thinking.

Similarly AI and linguistics tried to use deterministic mechanisms, but they are finally coming around to a thoroughly statistical approach, since that actually works and the deterministic one didn't.

Seems like many responses conflate the problem (preferences vs. hedonic response) with decision making methods (individual decisions vs. government enforcement).

Suppose instead that we define "tricking" people into acting against their hedonic interests as fraud, and allow class actions suits against anyone defrauding people. (Note: this is not a specific policy proposal, so don't make arguments against it based on specific execution issues.) Then addictive drugs (including cigarettes), predatory lending, provably unwholesome food, etc. would be attacked by swarms of litigators.

No government policy required. Companies would have an incentive to educate consumers about how to make hedonically valid choices and to avoid marketing that would tend to generate divergence between preferences and hedonic satisfaction.

Lawyers, on the other hand, would have an incentive to push forward analysis of preferences vs. hedonic satisfaction, and to do extensive investigation of actual choices of groups of consumers.