[clug] [OT] The Register is trying to crowfund $100-$250K for indep study on NBN

steve jenkin sjenkin at canb.auug.org.au
Sun Jun 23 08:06:42 MDT 2013


Michael Cohen wrote on 23/06/13 7:45 PM:

> You dont need a study to look past the politicking with the NBN - its
> actually a very interesting case in free markets and government
> interference.
> 
> The entire project was estimated to cost about $46B at a current
> population of 22 million this would cost about $2k per person upfront.
> For example, our household might pay $10k for that. Now I am a
> technology professional and probably would be able to utilize and
> justify this, but my elderly retired family are unlikely to justify
> spending this kind of money on a service which, to them, is only
> slightly better than their current ADSL, HSDPE or LTE service which
> then can get for much less than the $100 per month the NBN is
> projected to cost. This just does not make sense.

Michael,

Thanks for a considered and reasoned reply.

My analysis differs from yours.

Around 12M households and $44B ~= $3,500/household.
I'm just one person, so it costs me a lot, while families of 2-3 kids
get a great subsidy from me.

I believe the economics are different. That the high-end users subsidise
the rest of us, handsomely.

NBN Co can sell the _same_ fibre connection at 5 or more prices, based
on the CIR you buy: 12, 25, 50, 100, 1000.

That's pretty nifty marketing. It reduces the consumer surplus by
matching a subscribers willingness to pay with a product.

The low-end price can be much lower, encouraging higher take-up rates in
that segment. That doesn't raise much revenue, but it _does_ increase
the utility to everyone of the network.

Secondly, the real money spinner for NBN Co is not AVC charges (Access
Virtual Circuits, or access rate).
It's Volume Download, reflected by CVC - Connection Virtual Circuit. The
charges ISP's pay for bandwidth out of the Points of Interconnect, PoI.

NBN Co's current plan embodies 30%/yr growth in download volumes, which
is conservative. I've seen estimates of 50%, 66% and 77% (ABS, CISCO and
Adelaide Uni).

BUT, the "slam dunk" for users, if ISP's pass on price reductions, is
that NBN Co will _reduce_ the real cost per GB by 19% each year (for
that 30% increase). For a total of ~5%/yr increase is charges (ARPU -
Average Revenue Per User). [calculated as: 1.3 * .81 = 1.053]

NBN Co will also decrease in real terms the cost of AVC's, 12Mbps might
not decrease much, but 100Mbps will drive down over 10 years.

The upshot is that:
 - _most_ of the NBN Co revenue will come from 10%-25% of their users
   (I haven't seen their forecast breakdowns)

 - while the _average_ usage is expected to rise, this reflects much
larger increases by that top 10-25%. Means are highly influenced by the
leaders.

 - _everyone_ gets the lower download rates and AVC charges, paid for by
the high-end users.

I see the NBN Co charging scheme and long-term reduction in $$/GB as a
subsidy for those at the low-end.

I agree that in the beginning, NBN Co charges won't be much different
for low-end users.

In 4-5 years, two things will be apparent:
 - if they wish to stay on entry-level plans, their real-costs will
decrease considerably.
 - the long-term trend of _all_ computing, especially networking, is
that demand rises to meet supply: people tend to use higher speeds &
download volumes as they use the Internet more.

I also have a different view on collaboration and competition between
Telcos in building infrastructure.

Take the Telstra/Optus competitive rollout of HFC for Cable TV.
Neither completed their planned rollout, neither hit critical mass, both
wrote off billion dollar amounts within a few years.

They _could_ have collaborated and in low-density areas (suburbs) run
their infrastructure in separate areas with "at cost" interconnection
deals. That way we wouldn't have ended up with ~25% of the country
covered, but closer to 50%. Which would've made FOXTEL more profitable.

Secondly, why do we overbuild both DSLAMs and Mobile Networks in
_sparse_ areas? In densely populated, high-profit areas, I absolutely
get it.

But in the suburbs and regional areas, why can't the Telcos co-operate
and like the rest of the Western world, allow roaming on each others
networks??

The "only compete" model of Australian Telcos is not in their best
interests, nor ours as customers. The proof is everywhere else: Telcos
co-operate in marginal areas and only compete head-to-head in areas the
market can sustain the high investments needed.

Imagine if only one Telco bid for radio spectrum in the low- and
mid-density areas. The government wouldn't play them off against one
another and we as consumer wouldn't pay those inflated costs in our bills.

This may sound wrong to those steeped in the "competitive markets are
the only efficient markets" model, but I *saw* this model work very
effectively for over 25 years in our second Telco: OTC.

Not only did we declare record profits every year, but reduced phone
rates every year as well.  Telstra could've done similarly, but it's
always been locked into a scarcity/monopoly model.

I think both Political Plans have their undesirable points and both
their strengths.

But overall, I don't know of _any_ other investment that can
single-handedly improve the productivity of _every_ business in the country.

The Productivity Commission has done a lot of work trying to explain the
record productivity growth in the 1990's, that continued past 2000, the
dot boom/bust.

I posit this record growth period coincides exactly with Internet
takeup, and tapered off as users, especially small business (50% of the
workforce) hit their ADSL maximum access rates. During this time, people
(& small/medium business) discovered that PC's are only really useful
when networked. Otherwise they are calculators and paperweights.

That's my analysis and it's different from yours.
I don't want or need to convince you of my view - this list has a very
egalitarian & accepting ethic.

But I've laid my view out to show that the same data can lead to
different conclusion.

Thanks again for your great contribution.

all the best
steve

-- 
Steve Jenkin, Info Tech, Systems and Design Specialist.
0412 786 915 (+61 412 786 915)
PO Box 48, Kippax ACT 2615, AUSTRALIA

sjenkin at canb.auug.org.au http://members.tip.net.au/~sjenkin


More information about the linux mailing list