Two Letters To The Economist

I wrote the first in response to their 19 Oct 1996 leader. "Why The Net Should Grow Up"

As an Internet user for twenty years, an expert in several of its underlying technologies, and a close observer of the Internet industry, I was profoundly disappointed by your Oct 19th leader and special article. They were factually incorrect, logically flawed, and (for a magazine with The Economist's traditions) astonishingly illiberal.

The central economic contention in these articles rests on a myth, one which your writer might have usefully exploded had he investigated why the "barter" system you describe has been so effective to date. In fact (and despite the rather crocodilian tears shed in public by large national carriers) the central economic problem for ISPs has almost never been the cost of bit-piping; rather, it is the cost of the technicians and service people needed to maintain the network and support the customers. These costs are not proportional to bit volume; instead they rise (rather less punishingly) with network complexity. This is why peering and fixed settlements work economically, because the cost profile of large and small ISPs is much more alike than the disparity in their traffic statistics might suggest.

Far from being under threat, the Internet model of flat-rate pricing seems certain to spread to more traditional media as cheap bandwidth disrupts their traditional demand assumptions and technological improvements assimilate their cost profiles to the Internet's. Your leader could have observed that this is already beginning to happen to voice carriers.

From bad economics your leader segues to worse politics. Certainly the government "should not treat the Internet differently from other communications media, merely because it is new". The Economist betrays its own heritage, however, if it interprets this as an argument for regulating Internet content, rather than abolishing government control of older media.

On 21 Apr 1997 I followed up with this:

As one of the "wireheads" you cited in "Why the Internet failed to collapse" (19 April 1997), I found your editorial admission that the Internet is growing from strength to strength most gratifying. I invite you to publicly re-examine in this light the wrongheaded projections of your 19 October 1996 leader. "Why The Net Should Grow Up".

The text of my previous letter to you on this subject is available at http://www.tuxedo.org/~esr/netfreedom/economist.html. AOL's well-publicized troubles illustrate the fact that the Net's problems have little or no relation to the cost of bandwidth, and AOL's response demonstrates that abandoning flat-rate pricing is neither necessary nor sufficient to solve those problems.

Any cost or expected-outage model that is hypnotized by bit volume is doomed to miss its mark. Bandwidth is cheap. Even reliable bandwidth is cheap, and getting cheaper. Tellingly, the AOL outage and others like it didn't stem from undercapacity, but from configuration errors by their network administrators.

The Internet's central economic problem remains the cost of competent support people, including the salaries of "wireheads" like me. Volume pricing won't solve this problem; all it could do is raise the bar in the bidding war for experienced network hackers, a pool that is expanding at a rate relatively independent of network growth.

While I would find this result personally lucrative, I would prefer to see flat-rate pricing stay in place and salaries bid up only slowly, so that barriers to entry for new competitors remain relatively low. This will better serve both my long-term interests and those of the public.