[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: Pak vs India - internet rates



---------------------------------------------------------------------
Please help make the Manifesto better, or accept it, and propagate it!
---------------------------------------------------------------------
Ruchir:

Since you welcomed a debate, I need no further invitation to engage you
in
one such--of course, all in the spirit of friendly repartee!

First, I don't know of the circumstances of your friend in New Zealand,
but
I'm sorry that his monthly Internet connection costs run to $500-600.
What
puzzles me is: why?  Based on communications I have had with the New
Zealand
Commerce Commission, most ISPs offer unlimited access for around $34.95
monthly (certainly Telecom NZ's ISP ("Xtra"), Clear's ISP ("Clear Net"),
and
the Internet Group Unlimited ("IHUG") all do).  Your friend must not be
referring to dial-up Internet access.  Of course, ever since Telecom
NZ's
"0867" service went into effect (BTW, it was this carrier, not the
Commerce
Commission--as I implied in my previous message--that started this
service),
customers of that service have had to pay nothing extra beyond their
monthly
ISP charge.  Those who have stuck with 7-digit local dial-up calling for

Internet access have had to pay an Internet dial-up charge ("IDC") of 2
cents/minute, mainly as an "incentive" to move to 0867 calling.  But,
even
at that rate (which comes closest to metered calling), to get to
$500-600,
your friend must have 25,000-30,000 minutes (or upto 500 hours monthly!)
of
calling.  Given that a 30-day month has 720 hours, that's a lot of
Internet
surfing.  Your friend must be almost "always on" (24/7) or what I care
not
to speculate.  Suffice it to say, if he's using dial-up, he's on the
wrong
pricing plan!

Second, the larger issue is whether 0867 calling that enables Telecom NZ
to
manage its data calls separately from voice calls is, in light of
resource
constraints, the smart thing to do.  I submit it is because some facts
about
Internet calls are undeniable:  they are, on average 8-10 times longer
(in
the U.S.) than ordinary voice calls, and therefore tie up scarce switch
resources far more than do voice calls.  From what data I have seen
(again
in the U.S.), the load distribution (busy hour CCS) of Internet calls is

very different from, and much less peaked than, that of voice traffic.
Experts are almost unanimous that Internet traffic is far more
efficiently
handled over packet-switched networks, rather than circuit-switched
networks
designed primarily for voice traffic.  And, where possible, dedicated
connections--such as those from DSL or other high-bandwidth
services--should
be used by customers that desire unlimited or 24/7 access to the
Internet.
The question is whether the pricing plans that are being served up
everywhere encourage movement in the direction of more efficient use of
resources, less costly or constrained service, and efficient traffic
management?

Third, it is axiomatic in economics (and as a practising economist for
the
past 25 years, if I have learned one thing it is this) that optimal
resource
allocation and efficient pricing requires the structure of prices to
reflect
the structure of costs.  I don't think you'll get any argument with
this--in
the U.S., practically all are agreed on this point, and the FCC has led
by
example several times on this.  Yes, there is often justification for
subsidizing a "network" industry because, to paraphrase your remark, the

value of a network to society grows with increasing participation (or
"interconnectedness") on it.  This is often called the "network
externality"
and has traditionally been the basis for pursuing universal service
(maximum
possible telephone subscribership) on the backs of unlimited local
calling
and subsidized flat-rated pricing plans.  Unfortunately, this policy,
while
laudable for its goals, is not a panacea for more fundamental problems
affecting subscribership.  Nor is it an unmitigated good for new
carriers
that wish to compete in the market.  In the U.S., where local service
has
been priced below cost as a matter of policy for several decades now (a
practice, which prior to the Bell System break-up, was sustained by
subsidies from inflated long distance charges), there is tremendous
political inertia on the matter of raising local prices for unlimited
calling to more cost-based levels.  This is an economic no-brainer, yet
a
political quagmire that regulators and legislators touch at their
peril.  If
you talk to representatives of new entrants who wish to compete with
incumbents like the Baby Bells and others of their ilk, you'll discover
that
the pitiful lack of service to residential (as opposed to business)
customers is almost uniformly blamed on regulated local service rates
that
are below cost.  How could an entrant possibly compete against that?
Yet,
subsidized pricing and unlimited local calling remains the keystone of
residential telephone service in the U.S.  I'm only anxious that other
countries, particularly India, that have launched ambitious programs for

gaining subscribers learn from the lessons implicit in the U.S. model,
and
not blindly copy everything that happened--and still happen--here.

Fourth, you raise the issue of economies of scale.  Pardon me, but that
is a
red herring and is hardly relevant to telecommunications today.  Up
until 20
years ago, telephone companies were pretty close to single-dimensional
(basic POTS local service or long distance service) and the economies of

scale argument was frequently applied to characterize those companies as

being "natural monopolies" and to justify their continued regulation as
exclusive franchises.  I have studied the extant literature on economies
of
scale in telecommunications, and the empirical evidence--particularly
since
the 1980s--is tenuous at best.  If anything, with the diversification of

telephone companies, economies of scope have become more likely,
although,
even there, it is hard to quantitatively prove what is considered to be
self-evident.  With the onset of modular and digital technology, it's
arguable that advantages once enjoyed by large phone companies from
scale
economies have been substantially dissipated.  If scale economies were
the
overriding factor then, even in a country as small as New Zealand, the
emergence of Clear and others to take on Telecom NZ would have been
unthinkable.  The very fact that India is likely to have "large network"

(relative to what?) has little to do with whether a "relaxed pricing
structure" --whatever that means--is appropriate.

Fifth, you claim that, in the U.S., the cost of a long distance call is
only
fractionally higher even when it is 20 times longer than another such
call.
I'm curious what your evidence on this point is.  You specifically use
the
term "incremental cost" although you appear to really mean "average
cost" in
which the generally fixed costs of billing and notification are folded
in by
averaging over call duration.  I have monitored cost data in this
industry
since 1988, and I am not aware that, say, AT&T's true incremental
network
cost to carry long distance calls actually falls with the length of the
call.  Perhaps you can draw my attention to what I've missed.  If what
you
claim is correct, then I would expect competing long distance companies
to
offer "call duration discounts," i.e., per-minute prices that get lower
with
every minute (or block of minutes) called.  I'm unaware of any such
practice.  Oh yes, call volume discounts are an old practice, and in
fact
the basis of all resale competition, but that is not the same thing as
call
duration discounts.  Your conclusion that this, and the possibility that
the
Internet customer does not care about his bill (a conjecture at best),
implies that a flat pricing structure would be viable escapes me.  I
don't
see how your "incremental cost" becomes any less incremental (i.e.,
sensitive to minutes of use) or more fixed in order to justify
flat-rated
pricing.

Sixth, your justification (almost a demagogic championing of the
"capitalist
system") for allowing Internet-based transactions escape taxation
completely
befuddles--and bemuses--me.  Capitalism, or free enterprise, is
definitely
not rooted in throwing largesse at any favoured group of institutions,
when
it has never been showed that somehow they use resources more
efficiently
than anyone else.  If that were indeed the case, then the kind of
subsidy
policy you've stressed for Internet use must itself have weak legs.  The

issue that I raised was not about efficiency, but rather about
competitive
equity.  If an Internet-based online company and a brick-and-mortar
company
can both manufacture and bring to market the same widget at the same
cost,
then they are equally efficient.  Both must be allowed to compete:  the
terms of their competition must not be artifically distorted by making
one
collect sales taxes and not the other.  That would, in fact, be a
complete
repudiation of the capitalist system, by allowing those who have the
power
to levy taxes to become the final arbiters with the power also to pick
winners and losers in the marketplace.

Finally, let me return to the main issue:  should Internet usage be
subsidized?  There are good economic and policy reasons for arguing both

sides of this issue.  If efficiency alone were the criterion, then no
subsidy would be justified.  On the other hand, if the network
externality
and the value of the network were considered factors, some subsidy could
be
justified.  There are two problems here.  In the process of creating and

sustaining a subsidy, privately-owned and operated telephone companies
must
not become unwilling handmaidens to the cause.  If efficiently-operating

companies are expected to enter the Indian telecom industry, then they
should not be saddled with the responsibilities of carrying the growth
of
Internet usage on their shoulders, particularly when it is those
companies
that are expected to build the infrastructure.  If a subsidy policy must
be
implemented, then it would be infinitely better to let the subsidy
payments
come out of India's general tax revenues (or, as happens in the U.S.,
from
proportional contributions from all telecom carriers), rather than let a

misbegotten flat-rated pricing structure for local calling carry the
burden.
With the latter, the subsidy would eventually be borne by telephone
customers who do not use the Internet, rather than by the wider public.
It
would also distort efficiency by encouraging those who do use the
Internet
to leave their connections at all times and even when not actively
surfing.
By not forcing those customers to face the resource use and cost
consequences of such consumption, valuable resources would be lost with
practically no gain elsewhere to show for it.  That is why, unlimited
Internet usage should not be encouraged by tinkering with the pricing
for
dial-up calling plans.  If someone desires unlimited usage, by all means

move them to dedicated connections or other more efficient network
arrangements.  No one should get a free ride, unless every member of
society
gets one.

I fully realize that economics alone cannot be the final arbiter of
social
policy.  But, economics can frame the parameters of a policy that is
sensible, and alert one to the adverse consequences of a policy that is
not.
Blindly aping a system in the U.S., based solely on the observation of
high
Internet usage growth here, is basic intelelctual dishonesty, not to
mention
an opportunity missed for the kind of introspection that must accompany
policy-making in India.

With that said, I shall hold my peace.  Ruchir, if you want to continue
this
debate, I would be happy to do so one on one.  Regards,

Andy





--------------------------------------------------------------------------
This is the National Debate on System Reform.       debate@indiapolicy.org
Rules, Procedures, Archives:            http://www.indiapolicy.org/debate/
-------------------------------------------------------------------------