The Simple Economics of Open Source

Neel Krishnaswami neelk at
Sat Apr 22 04:51:48 CEST 2000

Will Ware <wware at> wrote:
> Ed (elb at wrote:
> > ... I didn't think much of the article...
> > ...they gave the impression that programmers behave
> > somewhat like monkeys, motivated primarily by concerns
> > about group status and dominance...
> They also mentioned, but immediately belittled, the possibility
> that altruism might be a real motivation. They may have been
> projecting their own thinking upon programmers in doing so. 

One of the interesting things about this thread is that it has given
me a bit of insight into what economics looks like to people who
haven't internalized its assumptions. I'm afraid that I'm tainted
enough that it no longer looks like the dismal science to me. :)

Seriously, the rejection of altruism as a possible motive strikes me
as a very reasonable decision. Altruism is generally not sufficient
motive in other aspects of life -- and since there's no particular
reason that programmers in particular are more likely to be altruistic
than the rest of the population, positing that hackers have greater
virtue is at best a dubious proposition. Consider, for example, how
much better government would be if all the voters studied the issues
and took care to be informed voters, and that basically no one makes
the effort. Most people don't even bother digging up evidence to
support their prejudices, let alone try to come to a careful decision.
And searching the Census and BLS websites is *easier* than hacking on
the Python core.

Second, economists assume that everyone is basically self interested,
including themselves -- it's a standard analytic assumption, not
snobbery. I read an article last year where two economists developed a
model predicting that a disproportionate fraction of groundbreaking
academic papers should be wrong. Basically, they assumed that since
funding stems from status, and status stems from writing ground
breaking papers, there's an incentive to draw flashy, counterintuitive
conclusions rather than confirming prior work. (They did observe that
their work implied their paper was probably wrong, since it was
advancing a flashy, counterintuitive proposition. :)


More information about the Python-list mailing list