4. KNOWLEDGE FOR ACTION Prof. Kevin Werbach
“The Internet”
(The “Peacock Map”
of Internet
autonomous systems,
circa 1999)
5. KNOWLEDGE FOR ACTION Prof. Kevin Werbach
Personal Computers Phone Networks Information Goods
6. KNOWLEDGE FOR ACTION Prof. Kevin Werbach
≈1.5 billion PCs ≈1.3 billion landlines
1.65 billion CDs
The World, Circa 2012
7. KNOWLEDGE FOR ACTION Prof. Kevin Werbach
The World, Circa Now
> 1.6 billion
smartphones+tablets
7 billion mobile lines Streaming > 50% of
Net traffic
8. KNOWLEDGE FOR ACTION Prof. Kevin Werbach
Post-PC Devices
Converged Broadband Networks
The Cloud
The Next Internet
10. KNOWLEDGE FOR ACTION Prof. Kevin Werbach
So is the PSTN!
• U.S. Residential switched access lines
– 194 million in 2000
– 101 million 2012
• % of U.S. Households
with POTS
– 93% in 2003, 25% in 2013,
and…
11. KNOWLEDGE FOR ACTION Prof. Kevin Werbach
Not Just a U.S. Phenomenon
• >50% of OECD countries
experienced a drop in PSTN access
lines from 2009-11.
• In 8 OECD countries, already <20
PSTN access lines per 100
inhabitants.
– OECD Communications Outlook 2013
12. KNOWLEDGE FOR ACTION Prof. Kevin Werbach
1943 20001980 2020
0.004
0.000000002
0.1
6.5
Year
Coming Next: Internet of Things
13. KNOWLEDGE FOR ACTION Prof. Kevin Werbach
Cloud Platforms
are the New
Utilities
13
16. KNOWLEDGE FOR ACTION Prof. Kevin Werbach
(Share of fixed
connections still on
dial-up in the OECD
as of 2011)
%
Broadband is Here
17. KNOWLEDGE FOR ACTION Prof. Kevin Werbach
Why Go Beyond Broadband?
• Convergence
– Voice/video/data to IP
– Fixed/mobile/nomadic
• Applications
– Streaming media
– Real-time communications
– Telework/telepresence
– Cloud computing and storage
– Financial services
– Internet of Things
– Smart homes
17
18. KNOWLEDGE FOR ACTION Prof. Kevin Werbach
OK, But Why FIXED Broadband?
• Mobile still only 7% of traffic in 2017 (Cisco)
• Fixed is complementary to mobile
– Most “wireless” traffic quickly goes to fixed
– WiFi offload estimated at 70% of
smartphone data
• Most recent data suggests
per-device mobile data usage
may be peaking
– At least, until the wearable/IoT
explosion!
18
Has to plug in somewhere!
19. KNOWLEDGE FOR ACTION Prof. Kevin Werbach
From Here to Fibre
• Only truly future-proof
technology
• The Good
– GDP benefits of ultrafast networks
– Knock-on effects of “economics of
abundance”
• The Bad
– Up-front capex costs very high
– Heavily take-rate dependent
– Municipal obstacles like ROW
– Transit costs an issue in some areas
– Business case uncertainties
19
20. KNOWLEDGE FOR ACTION Prof. Kevin Werbach
…And The Confusing
• All modern fixed broadband access systems
incorporate some fiber
– Key is how far fiber is extended toward the end-user
• Claimed speeds aren’t necessarily representative
20
Netflix USA Speed Index, January-May 2014
21. KNOWLEDGE FOR ACTION Prof. Kevin Werbach
Big Variation in Fibre Adoption
21
0 10 20 30 40 50 60 70 80 90 100
Israel
Greece
Germany
Austria
Chile
Canada
Finland
Spain
Luxembourg
Netherlands
United States
Hungary
Czech Republic
Slovenia
Norway
Estonia
Korea
Fiber as % of Total
Broadband
Percentage of fibre connections in total broadband subscriptions, June 2013
OECD ranks by FTTP as % of Broadband Subscriptions
OECD Average: 15.75% (June 2013)
22. KNOWLEDGE FOR ACTION Prof. Kevin Werbach
Big Variation in Fibre Adoption
22
0 10 20 30 40 50 60 70 80 90 100
Israel
Greece
Germany
Austria
Chile
Canada
Finland
Spain
Luxembourg
Netherlands
United States
Hungary
Czech Republic
Slovenia
Norway
Estonia
Korea
Fiber as % of Total
Broadband
Percentage of fibre connections in total broadband subscriptions, June 2013
OECD ranks by FTTP as % of Broadband Subscriptions
OECD Average: 15.75% (June 2013)
23. KNOWLEDGE FOR ACTION Prof. Kevin Werbach
Cable Changes the Data Somewhat
23
0 10 20 30 40 50 60 70 80 90 100
Greece
New Zealand
Luxembourg
Germany
Finland
Spain
United Kingdom
Austria
Poland
Czech Republic
Slovak Republic
Belgium
Chile
Estonia
Portugal
Hungary
Japan
Fiber as % of Total
Broadband
Percentage of fibre connections in total broadband subscriptions, June 2013
OECD ranks by FTTP+Cable as % of Broadband Subscriptions
25. KNOWLEDGE FOR ACTION Prof. Kevin Werbach
Architectural Choices
• Fiber to the home vs. the node
– Cost vs. capacity tradeoff
– The Australia example
• May be multiple technologies deployed
within countries
– Esp. with municipal networks
– Hierarchy by density (FTTH/VDSL2/VDSL/Wireless)
• Path dependencies important
– High DSL/cable adoption may actually slow fiber
25
26. KNOWLEDGE FOR ACTION Prof. Kevin Werbach
Business/Regulatory Models
• Role of incumbents vs. new entrants
– Non-traditional entrants (Reggefiber, Google)
important in some countries
• Scope of public funding or provision
– Different models being used at the national, regional,
and local level
– Success stories (Stokab), failures (Provo, UT), and
incompletes (Australia)
• Requirements for open access or wholesale
– Prevalent in most of the world except the U.S.
– Wholesale model often chosen voluntarily
26
27. KNOWLEDGE FOR ACTION Prof. Kevin Werbach
Vectoring and G.Fast
• Potential game changers?
– “Fiber-like” speeds at “DSL-like” costs
• Challenges
– May make unbundling technical infeasible
– Heavily dependent on loop lengths
– Real-world performance and deployment pace lags
– Still tops out well below fibre speeds
27
28. KNOWLEDGE FOR ACTION Prof. Kevin Werbach
Israel Fibre Network
• An important global test case
• Israel Electric Corporation (IEC) deploying a
1 Gbps wholesale network nationwide
– First announced in 2011
– Partnership with a group led by Sweden’s Viaeuropa
– Service scheduled to begin this month, offered by
retail providers (10 so far)
– Plans to cover two thirds of the country by 2020, and
the remainder by 2033
• Bezeq/Hot investing in FTTH in response
28
30. KNOWLEDGE FOR ACTION Prof. Kevin Werbach
What’s the Goal?
• End-user maximum download speed isn’t
necessarily representative
– Interconnection, transit, caching, equipment matter
– Full capacity not always available at consumer prices
– Caps, tiers, usage-based prices also significant
• Importance and meaning of ubiquity?
• Technological neutrality may be impossible
– Investment decisions today lock in particular
configurations for many years.
30
31. KNOWLEDGE FOR ACTION Prof. Kevin Werbach
What About Competition?
• Virtually no business case to overbuild
fiber, except urban MDUs
– Natural monopoly?
– Will the cost dynamics change any time soon?
• Unbundling may be restricted
– E.g. with vectoring
– Potentially removes a major regulatory tool in much of
the world
– Choose between next-gen broadband and competition?
31
34. KNOWLEDGE FOR ACTION Prof. Kevin Werbach
Watch This Space Going Forward
• ETNO Proposal for “sender pays” rule
• Complicated arrangements
– CDNs, multiple end user fees, etc.
– Data caps, freezones, usage charges also significant.
• At high level, interests are aligned
– Preconceived idea of cost “causation” not realistic.
• Voluntary deals raise neutrality concerns
• Compulsion skews competition/investment
w/no guarantee of more infrastructure
34
35. KNOWLEDGE FOR ACTION Prof. Kevin Werbach
Surveillance and Governance
• Governments go where the information is
• Information goes
where the users are
• An ongoing governance
challenge
35
39. KNOWLEDGE FOR ACTION Prof. Kevin Werbach
“A new pronouncement by the
regulatory agencies of a
doctrine of free interchange of
signals across the boundaries
of individual systems would
be of tremendous
technological benefit.”
-- Paul Baran, “Communication Policy Issues for
the Coming Computer Utility”, May 1968
So let me ask you a question: what is the Internet? Where does its value come from? We take the Internet for granted, but we rarely stop to think what the Internet really means. I’ll let you ponder that question. I’ll give you an answer… but not yet.
Really, the Internet is a proxy for three developments that represent the information age.
Each of these developments is huge; together they define the digital world. There are some 1.5 billion PCs in use worldwide, 1.3 billion telephone landlines, and last year 1.65 billion compact discs were sold, as an example of the spread of information goods.
What this means is that digital is the new normal. It’s not the future, or the emerging sector of the economy. It underlies everything. It’s not universal, of course; there are many people in the world without access to these things. But for any major industry, whether we are talking about healthcare, or energy, or finance, or real estate, the digital revolution is now the baseline.
But we’re already moving beyond it.
Because as big as those numbers are that I showed you, they are being surpassed. There are already more smartphones and tablets in use than PCs. And the first successful smartphone, the iPhone, is only 6 years old. PCs aren’t going away, but the future is about connected mobile devices. By the way, the next category beyond tablets will be wearables: things like smart watches and fitness trackers. There are a few on the market today like the Samsung Galaxy Gear watch and the Nike Fuel wristband. Get ready for an explosion, as Google, Apple, and every major device vendor targets wearables as the next great platform.
In terms of networks, the landline, analog, circuit-switched phone system that connected the world and provided the foundations for the Internet – what we call Plain Old Telephone Service or POTS -- is dying. In the U.S., already, three fourths of homes get their phone service from voice over IP or wireless, and the trend is continuing. OECD figures show a similar trend across developed countries, and inevitably the same transition will happen even faster in the developing world. Already, the 1.3 billion landlines in the world are dwarfed by the 6 billion wireless service accounts. Those wireless and data networks are converging as Internet access moves from narrowband to broadband.
And finally, information goods are giving way to information services. Instead of CDs and DVDs, we have streaming audio and video. In fact, so fast has been the growth of services such as YouTube that already more than half of all public Internet traffic in the world is streaming video. And the same shift from products to services is transforming every content industry, including books, movies, and newspapers.
This leads to three fundamental trends that define the next Internet. I’m going to describe each of them, and then say a few things about what they mean for public policy, and for business.
These are US numbers, but similar patterns everywhere
These are US numbers, but similar patterns everywhere
Let me leave you with this thought. As I said, the fundamental business issue is change. And the reason it’s such a difficult challenge is that our brains aren’t wired to understand compounded change. This chart is not to scale, but it should give you an indication of that.
In 1943, IBM CEO Thomas Watson said he thought there was demand for 5 computers in the world. Based on the population at the time, that’s a ratio of devices to people of .000000002. In 1980, AT&T hired McKinsey to estimate potential demand for mobile phones. They came back with an estimate of 900,000 potential users in the U.S., in a population at the time of 226 million, so a ratio of .004. In the 90s, as I told you, Microsoft’s slogan was a computer on every desk and in every home, but that was an aspiration; by 2000 there were roughly 600 million PCs in a world of 6 billion, so a ratio of 0.1. At every point, our assumptions about the technology were wildly wrong. Today, at the dotted line, we’re at about one connected device per human on the planet. In 2020, according to Ericsson, there will be 50 billion connected devices. With a projected population of 8 billion, that’s a ratio of 6.5.
Nine orders of magnitude growth in less than 80 years.
Google just bought Nest for $3.2 billion
Broadband is here. What next
The most recent data collected by the Australian Bureau of Statistics showed that the total volume of data downloaded via mobile handsets between April and June 2013 was 19 636 terabytes -- a 43 percent increase from the previous period of October to December 2012. Given that the number of subscriptions increased by 13%, between December 2012 and June 2013, these data suggest usage is starting to increase per individual user. (Australian Bureau of Statistics)
GBP BENEFITS -- Ericsson, Arthur D. Little, and Chalmers University of Technology study concluded in OECD countries, increasing broadband speeds from 4 Mbps to 8 Mbps produced an increase in household income of USD 122 per household per month, and increasing from 8 to 24 Mbps produced a significant further increase.
Stokab in Sweden: estimated economic benefits, including direct revenues as well as municipal cost savings and benefits to users, at €1.9 billion (USD 2.57 Billion), or more than triple Stokab’s total investment. (Acreo analysis)
COSTS -- A 2008 survey in connection with the Australian National Broadband Network (NBN) found a range of estimates based on international comparisons from under USD 500 to over USD 2 000 per premises passed. More recent data from the NBN suggests an average cost per premises passed in Australia of AUD 1 100 to AUD 1 400 (roughly USD 1 000 to USD 1 300). Exact numbers depend on the architecture chosen (with FTTN networks roughly one quarter to one-half as costly as FTTH), local labour costs, and physical factors such as population density and topography. Thus, in Korea, where a very high percentage of the population lives in urban areas, per-home deployment costs for fibre are estimated at only USD 110 to USD 170.
GBP BENEFITS -- Ericsson, Arthur D. Little, and Chalmers University of Technology study concluded in OECD countries, increasing broadband speeds from 4 Mbps to 8 Mbps produced an increase in household income of USD 122 per household per month, and increasing from 8 to 24 Mbps produced a significant further increase.
Stokab in Sweden: estimated economic benefits, including direct revenues as well as municipal cost savings and benefits to users, at €1.9 billion (USD 2.57 Billion), or more than triple Stokab’s total investment. (Acreo analysis)
COSTS -- A 2008 survey in connection with the Australian National Broadband Network (NBN) found a range of estimates based on international comparisons from under USD 500 to over USD 2 000 per premises passed. More recent data from the NBN suggests an average cost per premises passed in Australia of AUD 1 100 to AUD 1 400 (roughly USD 1 000 to USD 1 300). Exact numbers depend on the architecture chosen (with FTTN networks roughly one quarter to one-half as costly as FTTH), local labour costs, and physical factors such as population density and topography. Thus, in Korea, where a very high percentage of the population lives in urban areas, per-home deployment costs for fibre are estimated at only USD 110 to USD 170.
Israel at the bottom (0%), but as you probably know there is the fibre project. I’ll talk more about it.
Factors -- Cable television is also available to 80% or more of the population in Canada, Belgium, Korea, the Netherlands, and Switzerland. In most countries, deployments of cable systems for multi-channel video programming were limited, especially once direct broadcast satellite (DBS) technology allowed for satellite-based systems to provide comparable services. Cable TV subscribership is below 20% of homes in countries such as France, Italy, Mexico, New Zealand, Spain, and Turkey. OECD Communications Outlook 2013, at 188.
Id.
Say here that evidence is mixed. Give the examples in the paper e.g. Reggefiber in the netherlands, Japan, Korea, etc.
Germany, BNetzA has proposed allowing operators to deploy vectored VDSL2 on a “first come, first served” basis. While this might limit competition, it could also create a strong incentive to upgrade in order to lock out other entrants.
These are the US companies, but similar debates everywhere. It’s a business model issue – high fixed costs vs. revenue and profits based on scale and low marginal cost. If you starve the network too much, limits investment. But if you allow the network too much control, limited innovation and growth at the edges.
Network neutrality, peering, interconnection.