[Disclaimer: When a lawyer writes about economics, things can go dreadfully wrong. My terminology may be way off the mark – and maybe my thinking too. I welcome constructive corrections in the comments.]
Harvard University’s Samuel W. Morris University Professor of Economics Dale Jorgensen kicked off this morning’s session of the PFF Aspen Summit with a talk entitled Whatever Happened to the New Economy? [ppt here – Shane Tews, reading over my shoulder, wants me to mention the pretty colors!] In it, he examined both growth in the IT sector and growth in other economic sectors thanks to their adoption of IT. One of the delights of IT is that innovation in this area propagates out across the economy, magnifying the benefits to society.
I was interested in a slide he put up showing the “Relative Prices of Computers, Communications, Semiconductors, and Software and Computer Services Industry Output, 1960-2005,” slide seven in the PowerPoint. Each area showed consistent decline in price over this period, except Software and Computer Services, which has remained essentially flat.
I was curious about whether and how open source software was measured in Professor Jorgensen’s data, so I asked him during the break whether it was.
Indeed it is, and he mentioned Red Hat as an example of an open source provider that is within the data.
Given that open source software isn’t sold, Professor Jorgensen treats it as “advertising” used to draw people into using service providers like Red Hat. I think that’s a sensible and orthodox approach. As Mike Masnick has documented, lots of digital goods are “advertising” of one sort of another for some complimentary good.
But I think price will tend to understate the economic value of things like open source software (a thing Professor Jorgensen didn’t purport to be capturing, so I’m not criticizing his chart or approach).
People often use price as a proxy for value. After all, the price people will pay reflects very well how much they value a thing. (Well, actually, they value it more than whatever price they pay – every transaction has at least a little consumer surplus.) The price decline in all these sectors, of course, reflects not that their abstract value – the improvement they bring over a baseline of zero – has dropped, but that their scarcity relative to other things has dropped. People need to pay less, and they get more and more surplus (again, over a baseline of zero). Life keeps getting better and better.
To use price as a proxy for abstract value, one would think that life hasn’t gotten better in the software area, but this is plainly untrue (and again, this is not a case Professor Jorgensen was making). Open source is a challenge to the measurement of economic value. Millions of software downloads have brought marginal improvements to billions of lives with no measured economic exchange taking place.
Take the example of two pieces of software that do exactly the same thing. A million copies of Program A are sold for a dollar a piece. A million copies of Program B are given away, drawing users to a Web site at a rate equivalent to what $10,000 in banner advertising would bring. Did Program A add $1,000,000+ in value to the economy, while Program B added only $10,000? It can’t possibly be.
Economics has a challenge before it that I don’t think it has addressed yet: measuring the value of digital goods that are not distributed and sold subject to traditional intellectual property protections.
Update: Added a link to Masnick’s Grand Unified Theory On The Economics Of Free