Subject: Re: economic efficiency of free software
From: L Jean Camp <jean_camp@harvard.edu>
Date: Mon, 9 Feb 2004 11:32:47 -0500


On Friday, February 6, 2004, at 10:57 PM, Stephen J. Turnbull wrote:

> An elementary exercise: applying this idea, why is Amazon.com
> successful (at least in terms of revenues, I dunno their P&L)?

Because they have chosen to target a small number of books and function 
as a organizational node for speciality book sellers and purchasers. 
You appear to be implying it is because they are willing to carry 
almost infinite inventory but they had a change in strategy.

> I think there is a real opportunity here.  Figure out a way to get
> people to pay for software indexing and review, and channel their bug
> reports and feature requests to the developers, and generate a revenue
> stream from that.  It would work like electronic publishing in
> academia, except that (see below) the problem is aggregating the
> signals of customers, not peers.

According to the ebusiness group at Sloan (and lots of other papers at 
WISE Seatlle 2003) people will provide this free information for 
reputation. The problem in many open code forums is culture class and 
feedback. The on-line personas of many open code new entrants  is one 
of assertion of competence through confrontation (and of course Eric 
Raymond).  No one is ever confronted at Amazon, ebay confrontations are 
carefully managed and slahdot is a free for all.

RedHat to my knowledge does not offer any perks to frequent bug 
contributors. If you have someone who carefully documents all their 
crashes and sends it in, they get no visible reputation. A "frequent 
reporter" system such as that used by Dell computing, where people who 
contribute accurate and widely used information get coupons or even 
visible public validation.  I think that RedHat could profit from such 
a system --offering only recognition.

>
>     rn> The open source world doesn't have the price signal.

I have come to believe the problem with open source is that not enough 
companies release their own custom code and thereby obtain value from 
others using and working on it. Ensuring returns to the producer from 
releasing the code can solve that problem, and there are licensing that 
do that. So, imho, education and time are the solutions too this 
problem.

So some dozen state are getting together to share code because as 
states they have very similar problems. One of the things they are 
developing is a community of users so that fifty clerks in the Sec of 
States office can talk to the other 49 (well now it is more like ten).  
The developers handling the package become part of the community of the 
clerks. In this closed environment there is no need for a price signal.

GE capital (according to the VP to whom I spoke) accepts user 
recommendations on what services would be useful. Then when enough 
users recommend something to be profitable they roll it out. Similarly 
when enough clerks recommend something to be useful then it will be 
rolled out.

I think the problem of economics of free software is not simply in the 
market - it is in the firms. The intra-firm transfer and creation of 
software is broken and that is why it has been outsourced (at very high 
profit margins) to MS.  Fixing the intra-firm problem is as important 
as fixing the inter-firm open market problem. So a solution that could 
work within an organization could in theory also work outside the 
organization. There is no way that GE and IBM and PwC could not have 
cooperatively develop an office suite for what they pay Microsoft. It 
is the inability of the big guys to coordinate both internally and 
externally that is the current barrier to open code.

>
>     Phil> Doesn't have or doesn't currently have.  I don't have a
>     Phil> solution here but it does seem like there must be some model
>     Phil> that would "work".
>
> First, I think you misunderstand what Russ means by "price signal"
> here.  It is a happy accident that (where markets are undisputably the
> best way to organize trade[1]) the same price signal that (1) tells
> the buyers and sellers how much to buy and sell also (2) extracts
> compensation from buyers for their use of social resources and
> compensates sellers for the resources they provide, and (3) allows
> outside observers to measure the social value generated by the market.
> All of this "falls out" quite directly from the so-called "marginal
> analysis", ie, differential calculus, and the identity of price with
> marginal value that the market enforces.

...

> What goes wrong _with the economic analysis_ in the open source model?
> Very simply, there is no constraint to pay market price, and most
> buyers don't.  This means that the buyer who actually pays market
> price is randomly selected, and I can think of no reason why market
> price will be pushed to marginal value.[3]

Of course this marginal cost is zero on information goods, information 
goods are non-excludable and non-rivalrous so all the assumptions fail 
miserably.  But don't let this disturb you in your tautology.

So the problem is not the open source model it is that the fundamental 
assertions do not work with information goods. The response has been to 
make them proprietary - to make them excludable and rivalrous. However, 
this has immense social cost. And now we are back to the point of the 
original poster - how to make information function in a market and in a 
way that is socially good. Because forcing information into the 
tautology of classic economics has a terrible price ---because it 
requires scarcity to work.  The market requires scarcity and thus 
denial in order to function.

So here is an interesting and applicable question - why does P2P work? 
It works because if I contribute to server C and you and I compete for 
server A and C I have reduced the relative load on A. Why does open 
code work? Because I either have to develop x, y and z or accept an 
environment where x,y, and z are both scarce and not optimized for my 
needs. How do we pay for it?

I think the idea of a NFP consortium is a great one, modeled on the 
cooperation that allowed for the creation of various industry xml 
naming standards. However, once you get enough companies in they want 
to create barriers to keep other companies out (to much exposure to 
classical economics or bad kindergarten experiences, I am not sure 
which).  It works for them in the very near term which is the only term 
of interest. But in the LT they could coordinate and release software.

The public sector can lead the way here because the public sector is 
being starved of resources. So they have to adapt and think long term.  
The "open source in government" and idiocy of  "let the market decide" 
are about understanding that if the states and feds get it right the 
market will follow. (Let the market decide usually translates to let 
the lobbyists decide in this battle.) Coordinated communities of users 
designing code with developers makes sense.

>
>     Phil> initial development would be funded
>
> How?  I mean, where does the money come from?  The users aren't
> compelled to pay, are they?  More important is, "Without the price
> signal [in its social value role], how do you know which projects
> deserve initial funding?"
>

If firms and organizations did the initial internal funding then there 
would be no need for imperfect price signaling and the false scarcity 
it requires.  Then the code could be made available so if someone else 
wants to pick it up and sell the excludable and rivalrous part (support 
and customization) it would be there.  Then the creating company can 
choose to out-source its own support or at least have competition. It 
is getting all the available code out the door so the information 
markets can work that is the current barrier.

I realize that alot of you guys are in the trenches and may not want to 
hear this but I will say it anyway - open code is changing the market. 
Open code is working in the market.  It is just taking time.

best regards,
Jean