[kictanet] Fwd: [Internet Policy] Universal service fund in the U.S. has failed badly. Despite $billions, U.S. has close to the worst coverage in the developed world.

[Barrack Otieno]

Listers,

In light of Walu's write up on Universal Access, this might provide an
interesting comparison.

Thank you

Best Regards

---------- Forwarded message ----------
From: Dave Burstein <daveb at dslprime.com>
Date: Wed, 13 Apr 2016 02:48:29 -0400
Subject: [Internet Policy] Universal service fund in the U.S. has
failed badly. Despite $billions, U.S. has close to the worst coverage
in the developed world.
To: Richard Hill <rhill at hill-a.ch>, "internetpolicy at elists.isoc.org"
<internetpolicy at elists.isoc.org>, Alejandro Pisanty
<apisanty at gmail.com>

Folks

I changed the subject to avoid getting entangled in the debate about the
particular report and focus on the issue of achieving access. Based on the
U.S. experience, I believe USF funding is inefficient at best.

The U.S. has about 5M homes that can't get landline broadband despite $7B
in our stimulus and billions more over the years in Universal Service
money. (FCC data) That's 7-10X the unserved rate in much of Western
Europe.The greater land area is only a partial explanation. See the work of
Professors Rosston and Hazlett as well as Scott Wallsten, who was chief
economist of our broadband plan. I did two workshops for our Broadband Plan.

I don't have depth on the issues in other countries but what I know
suggests USF isn't performing well most places.  I believe the Kurth
Solution, named after the German regulator, is strongly preferable.

Unserved rural areas by their nature are scattered and have limited local
infrastructure. In most places, only the incumbent has the local facilities
and backhaul required to serve the area at competitive cost. This became
obvious when almost none of the applications for our broadband stimulus
were from new entrants looking to connect the unserved, although that was
the primary goal of the funding.

The result was that the local phone company could and did claim their costs
were far higher than they needed to be. Add on the political power of the
phone companies (regulatory capture) and the system is spending what I
estimate is easily twice what an efficient subsidy would be. (Limited hard
data.)

Kurth in Germany covered most of the "white spaces on the map" at a much
lower cost. In a spectrum auction, the carriers were required to cover
white spaces before they use the spectrum in the big cities. While the
companies could reduce their bids, WIK Consulting told me the impact was
very small.

Vodafone and Deutsche Telekom, spending their own money covered those small
towns rapidly and what I believe was far less than any USF subsidy would
have been. LTE speeds are now 50-100 megabits, enough for most purposes.
The caps on the rural wireless were much too low but technology is rapidly
improving that. Today's LTE gear has 3-4X the capacity of the earlier
systems, mostly through carrier aggregation. More advanced MIMO and MU MIMO
are likely to deliver 10X the current capacity in the next few years,
according to leading wireless experts such as Stanford's Paulraj. That will
allow raising the cap to perhaps 100-150 gigabytes/month, probably with
unlimited access nights and mornings.

Portugal and others are working with a similar system.
----------------------------------
Unless you have a plan to educate the regulator and enforce reasonable
pricing in USF, I think the plans are somewhere between overly expensive or
ridiculous. The U.S. couldn't do it despite major efforts since the
Broadband plan.

The "Kurth Solution" - indirect competitive bidding as part of the spectrum
auction, seems to work well. It's not the only alternative.

Australia is building a National Broadband Network for everyone.
Unfortunately, the cost has gotten out of hand, as so often happens with
government projects.

India may have a better idea; the government is extending fiber to 10's of
thousands of small towns with the expectation private or cooperative
efforts can deploy the last mile. The backhaul costs are often the largest
expense in rural service. They require so much capital few but the largest
companies can finance backhaul.

With mobile phones and WiFi gear now under $50, local businesses or munis
can connect to that fiber and serve the area with a small investment. This
is very promising, although political entanglement has drastically slowed
construction. Some government projects work well (China), others we know
are unsuccessful.
-------------------------------

Satellite has become much more attractive than before. Mark Dankberg,
Viasat CEO, told our broadband plan today's satellites have an order of
magnitude more capacity and reduce latency by half. (They haven't solved
the 22,000 mile distance but improving the routers and caching apparently
makes a huge difference.) No politician would admit it, but the assumption
in the U.S. plan was satellite for the last 1%.

All current Universal access programs may have become obsolete on December
21, The Falcon 9 successfully launched a constellation of 11 Orbcomm-OG2
second-generation Low Earth Orbit (LEO) satellites. The first stage
returned to Cape Canaveral successfully, which Elon Musk's company says
will reduce costs by about 30%. The first two sea landings failed but April
8, 2016, Falcon 9 Flight 23 successfully landed a 1st stage booster on
drone ship Of Course I Still Love You.

Low Orbit satellites solve the latency problem but still have a way to go.
Small, inexpensive tracking antennas need to be developed. An engineer I
respect has studied the problems and expects LEO sats will be the way to go
in a few years.

Facebook's drones will probably never be cost-effective just serving
extreme rural areas, I believe; I'm anticipating they will need urban
customers to cover the costs. Their engineers may be good but the head of
the project didn't know the facts on the ground. I haven't researched it
but I expect the rapidly dropping costs of LTE will be much less than the
drone network. I would guess the same about Google's balloons but the team
seems far more knowledgeable.

I've bcc'd this to the experts mentioned. I hope they and any others catch
any mistakes I have.

The above is only about access. I'm very encouraged by the new U.S.
Broadband Lifeline, a $9.25/month subsidy for the poor where networks are
already in place. Broadband networks are a classic case of high fixed
costs/low marginal costs. The marginal cost of adding a customer to an
existing large broadband network in America or Europe is $4-8/month.

Dave Burstein

Editor, Fast Net News, Net Policy News and DSL Prime
Author with Jennie Bourne  DSL (Wiley) and Web Video: Making It Great,
Getting It Noticed (Peachpit)



-- 
Barrack O. Otieno
+254721325277
+254733206359
Skype: barrack.otieno