[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Another Point Of View (Round 2)
At 08:02 PM 12/3/96 -0500, James Love wrote:
> A virtual flat gives people incentives to hang up the phone once in
>a while...(200 hours goes fast if you don't).
Yep. That's the ticket. But for users who need full-time availability, for
temporally sparse usage, there's probably more to be done with D-channel
packet. They could do very low volums stuff there, and turn on the B
channel when necessary. This is a hole in most router-vendors' offerings.
The other alternative is Frame Relay. US West strategically (under)prices
it a buck or so below Threshold-rated ISDN (I think around $68), for 56
anywhere in the LATA. This is useful for those who really need full time
connectivity. Bell Atlantic charges several times that rate, and cites that
as justification for its high ISDN rates!
> But the goal should be to have everyone nailed up... in the longer run.
>I think NATCO can do this now, with ISDN, for $17.90 per month. But then
>NATCO has a CEO with a PhD. in physics.
I disagree that nailed circuit-switching is a general solution. This is
known in ISDN-ese as "semi-permanent circuits" and is tariffed in some
countries as such. It's a leased-line substitute, with what are sometimes
somewhat more favorable costs and prices.
NATCO's the exception. It has, I think, one CO switch, a nice robust
DMS100, with a few remotes. The core of the DMS can handle very high
traffic, especially for a 7000-line telco. It doesn't scale to Bell
densities, where there are interoffice trunks, tandems, more concentration
in the switches, etc. Packet, frame, new low-cost intra-exchange leased
lines to an ISP switch, ADSL, and other non-circuit-switched services are
more appropriate ways to use the local loop for full-period connectivity,
when it's needed. One option might be for the CO switch to nail a B channel
to a Frame handler, for instance. B-channel packet (which is nailed to the
handler in the semi-lame US implementations) is too costly (around
$50-100/month plus packets) but Frame Relay was originally invented as the
"ISDN New Packet Mode Bearer Service". It just got away.
> Bill seems to be saying (he is saying) that high prices are actually
>good, because they encourage entry. Maybe Bill should lobby the PUCs
>for even higher rates. Hmmmmmm.... Maybe..... nevermind.
>
> Oh, one more thing. In game theoretic models, it is the expected
>prices after entry when are important... not the prices before entry.
Yes, and there's plenty of precedent to demonstrate it. Take, for instance,
intra-LATA toll. This has been preposterously overpriced (often 10x cost)
for a long time. Then as states began to authorize competition, the prices
fell precipitously. Bell South used to get over 50c/minute on some routes!
Now, with competition, some RBOCs have literally abolished intra-LATA tolls.
Bell South's Miami-LATA tariff now allows any intra-LATA call for 25 cents
untimed. This doesn't leave much room for competition. I'm not saying I'm
*opposed* to lower, if compensatory, rates. Just that the telcos hold rates
high *until* there is a real chance of competition, then undercut them in
order to preserve their monopoly. (An unprofitable monopoly is better than
no monopoly at all, I guess.)
Even if that weren't the case, it is (IMnsHO) bad policy to allow a
monopolist to take advantage of their monopoly at the public expense.
Mispricing (such as profit maximization absent competitive forces or
regulation) leads to inefficient allocation of resources, which is what free
markets are supposed to be best at. Gosplan wasn't successful as a state
enterprise, but neither are private Gosplans, the mediaeval-like franchises
that Bells effectively still have. They may, as Bill has said in a column,
irritate the public enough to make competition more welcome. But that's
like banging your head against a wall because it feels so good when you
stop. And the pain may linger.
---
Fred R. Goldstein k1io fgoldstein@bbn.com +1 617 873 3850
Opinions are mine alone. Sharing requires permission.