I've got a terminology gap here. Somewhere along the line I seem to have conflated "network effects" with the notion that the maraginal cost of any addition to a network will typically be less then the value to the network of that addition. (Yes, "typically" is important, to avoid lumpiness quibbles, such as "a whole new server is a whole new server for one added user," but let's don't quibble.) It sounds like what we're talking about here involves what I would have thought of more as "network externalities." Education, please? MacN On Sat, 3 Mar 2001, Declan McCullagh wrote:
Phill gets it halfway right, or put another way is halfway wrong.
The question is not whether network effects exist or not; clearly they do in some form. Better questions include whether they have produced inferior results than "someone" desires, how strong they are, and whether state intervention is a good idea because of their alleged ill effects.