Network Externalities

Network externality, the concept that a product's value to a consumer changes as the number of users of the product changes...

-- ARE NETWORK EXTERNALITIES A NEW SOURCE OF MARKET FAILURE? by S. J. Liebowitz and Stephen E. Margolis, URL: wwwpub.utdallas.edu


Yikes! The Ginger Factor is very high with this thing. An example or two in conversational English would have been nice.


I've more often heard of it referred to as the Networked Effect although that could just be a misinterpretation on my part - I'll have to ask my wife the some-time economist. Anyway, the classical example is the telephone. A single telephone is quite useless. Two is better - especially if the owners know each other. The more people who own phones the more valuable each phone is to its owner. -- Phil Goodwin


An externality is an effect that is external to the market, that is, somebody profits or is hurt, but nobody pays/is paid for it (economically, that is).

Example: The production of your computer caused a lot of environmental pollution, almost as much as the production of small car. You profit from that, somebody else suffers, but you didn't pay them. If the price for the right to pollute the environment had been internalized to (?) the market, e.g. by pollution taxes, the computer might have cost a lot more.

[Injection: this is a false assumption. Environmental damage causes everybody to suffer eventually, so the "somebody else suffers" premise does not apply. No amount of pollution taxation will cure permanent ecological damage, and it's cumulative. Even though the cost is not readily apparent it is still there. Use another example.]

Is it true that producing a computer causes almost as much pollution as producing a small car? It seems very unlikely. Do you have any evidence? Or is this just a case of 'almost' meaning 'not?

Most externalities are negative, that is, the effect is hurtful to the third party not part of the market. Network Externalities are interesting because they are positive, so Conventional Economics Doesnt Apply Here.

Since an externality is an economic effect not covered by the market, it normally results in "market failure", that is, the market does not achieve optimum welfare. The article mentioned above argues that this might not always apply.

The effect doesn't have to be positive for everyone. The more people write word documents, the more valuable word becomes. Want to write .doc files on Linux anyone? The more people buy and build software as Sacred Intel Binaries, the more valuable those binaries become. (Hence the famous Bill Gates quote about other platforms: "Develop for it? I'd rather piss on it".) This can be the source of a Lingua Franca because of the value accumulating behind one standard. (See also Imbalance Of Power.)


In 1995, I presented a seminar and published a page that used Network Externalities and some other theories to study the economics of mass-marketed software:

http://www.csci.csusb.edu/dick/papers/rjb95b.one.size.html

A piece of this was presented at the 19th Int Conf on Software Engineering.


A good book explaining positive network externalities, lock-in, and a lot of other interesting bits of economics that are relevant to the software world is Information Rules:

ISBN:0-87584-863-X

(yeah, it's my amazon referer thing... so remove it if you don't like it)

Any good quotes, insights from the book that you can share with us, not in US and cash poor people?

See original on c2.com