Transcript of CITI Report presentation to FCC – Dec 10 2009

[report and more details: here]
Washington, D.C.
Thursday, December 10, 2009

Introduction of Workshop:

Panel 2 – Citi Report: ‘Broadband in America: Where It Is and Where It is Going (According to Broadband Service Providers)’

Director of Policy Research, Columbia Institute for Tele-Information (CITI)

Research Assistant Supervisor, Columbia Institute for Tele-Information (CITI)


Director, Pew Internet & American Life Project

  • * * * *

MR. KOUTSKY: Are we good? Okay. Okay, I’ll start the second panel of the day.

My name is Tom Koutsky. I’m with the Omnibus Broadband Initiative at the FCC. I’m seated here with Thor Kendall, who also works with me on the deployment team.

And we have the pleasure of welcoming Bob Atkinson and Ivy Schultz of the Columbia Institute of Tele- Information, who have put together a study, also at the request of the FCC, related to the broadband deployment and investment plans of broadband service providers.

They were actually requested to provide us expert information and advice with regard to either publicly state plans, or what other types of plans they could kind of dig up. Hopefully, they can give us an indication of their methodology on that, in terms of letting us know where broadband is today, where we expect it to be in the reasonably near-term future.

And it is actually — the report, “Broadband in America,” was posted a couple weeks ago on the FCC website and received a number of comments. And I think, based on the prior discussion, where we noticed that there was a fair amount of disagreement about the results of the findings, this is actually a study where I think a lot of people found lots of reasons to agree, in the comments, but sometimes for different reasons they were agreeing. That’s just one way of thinking about how FCC proceedings go sometimes.

Bob Atkinson is actually an FCC veteran, and a veteran of the telecommunications industry for several years. He has been with the — he served a Columbia Institute for Tele-Information since 2000. And prior to that, he was in this chair, basically s deputy chief of the FCC, what was then called the Common Carrier Bureau. And before then, like I said, he worked for a number of telecommunications companies and, in fact, some of the original competitive entrants, TelePort Communications Group, principally.

The study co-author, Ivy Schultz, is the research manager at the Columbia Institute for Tele-Information. And she worked the research assistants and the directors on a number of the projects, including the “Broadband for America” report. She received a master’s degree from Georgetown University, so she’s basically revisiting Washington, D.C., again.

And to comment on the report we have Lee Rainie — we’re fortunate to have Lee Rainie — who’s the director of the Pew Internet and the American Life Project, which is a nonprofit. And he has had that position since December of 2009 — or he’s had — I’m not sure how long he’s had that position, but since 1999, the Pew Center has examined how people use the Internet, and how that affects their families, communities, health care, education, civic and political life and workplaces.

Lee is a prolific author, and actually is a very important and a very significant independent voice on telecommunications and broadband adoption, on use policies, in the United States. So we’re very fortunate to have Lee here.

So I will turn it over to Bob and Ivy to give a presentation on their findings.

MR. ATKINSON: Thank you, Tom. First of all, sort of a disclaimer. CITI itself as an institution doesn’t “author” reports or papers.

So Ivy and I are the authors. We are responsible for it.

Ivy and her team of researchers did all the hard work, in terms of gathering the data.

And so all the good data she gets credit for. All the mistakes in interpretation, I’ll take — where there are mistakes. And if there are mistakes which, inevitably there will be, or errors or incomplete information — one of the things we’ve always said in any communications we’ve made on this project is please send us better information.

And I think, hopefully, one of the purposes of this kind of report is to smoke out better information.

So to the extent that any company, any trade association, et cetera, has better information, give us and give the FCC that better information. And we actually have some experience of companies’ giving us better information that Ivy will mention.

But if you have better information, here in the room or on the webcast, please send it to CITI-Broadband at GSB — that’s Graduate Business School — And I think more and better information is always better, and I look forward to seeing that.

Let me, in fact, turn it over to Ivy so she can explain our methodology and how we got all the good information.

MS. SCHULTZ: Thank you, Bob. And thank you for the opportunity to present here today.

I’d also like to acknowledge some of the research team. We had a lot of help from a lot of people, but specifically Max Muller, Chris Scheubel, and Harry Siebenweiber were key researchers in our efforts.

And I’d like to start with a summary of the tasks that the FCC requested. We had three main tasks, which became the three sections of our report.

The first was a listing of all the publicly- announced broadband plans. And I’ll show an example of that in a few slides.

The second was a comparison. And here we were asked to look at what was announced and compare it to the performance of the broadband deployment.

And, finally, we were asked to make a future projection summarizing the investment analysts’ forecasts.

Next I’ll talk about the data sources that we consulted.

We had, for the first two sections, mostly company data as a main source. And this includes quarterly reports, annual reports and company press releases.

For the third section, which was a future prediction, we relied mostly on analysts’ reports and record research reports.

We also made a choice not to use data already submitted to the FCC. And in the case where the data is similar, it validates both. And in the case where we have disparate data, it demonstrates the need, or signals the need for further analysis.

And this is a list of the broadband plans that we reviewed. There are 33 companies listed here. And the last three, or last four or so, are industry associations, which provided a way for us to get aggregated data of the smaller companies.

And this next slide is an example of a company in our appendix. The categories here are somewhat difficult to see, but we looked the announced timeline, the coverage, deployment and footprint — the expected coverage, deployment and footprint — the states covered, the capital expenditures. The expected broadband performance, which came through as transmission speeds and, finally, the expected average revenue per user.

And once we completed the research for each category we sent a draft of this chart to the individual companies for verification. Some companies responded with additions or corrections and some did not. The example shown here is Time Warner Cable. And this is an updated submission, subsequent to our report’s release.

We think that this information is useful. And if it is, indeed, we think it would be useful for the FCC to keep receiving updates from companies on their broadband plans.

In this chart we developed a way to show the performance of a plan, compared to the announced completion date. The diamond that you see represents the project’s goal, and the red shading represents a late completion, while the green shading represents early completion.

And as you can see, the companies are separated by industry, with cable at the top, followed by the wireless, wireline — sorry, wireline, wireless, and satellite companies. And although the chart doesn’t represent all of the companies or industries, there is a general observable trend that the performance varies on the type of technology project. So, for example, the cable industry’s recent upgrades from Doxis 2 to Doxis 3 have been early or on time, while the deployment of entirely new infrastructures have been less like to be completed on time.

And I’ll turn this back over to Bob to talk about some of the other broadband trends we found.

MR. ATKINSON: Thank you. Our report itself has 21 figures, 15 tables, and I’m not going to go through all of them. I’m just going to try to hit some of the highlights.

And I think this chart, which shows a forecast of wired broadband subscriber growth to be one of the most important ones, because if it were flipped over would show the end of the proverbial S-curve for broadband adoption. And I think quite important, the forecast out at 2012 is that you’re looking at less than 2 percent subscriber growth for wired broadband.

And presumably, if that trend continues

— and I think the analysts would say it’s likely
to do so — you know, you’re looking at basically little or no real growth in wired broadband.

More importantly, what’s left of wired broadband — to the right of the chart — in subsequent years, it’s the least attractive customers. It’s customers in the what we are often, I guess, calling the “unserved areas.”

It’s people who are “price sensitive,” i.e., you know, they want a really low price. It’s the low or slow adopters, non- adoptions, the “non-tech” people, people who have difficulties with computers. It’s — you know, there’s a whole variety of subscribers who aren’t subscribing but, from a service providers point of view, they’re probably not really, really attractive.

But at the same time, you see this low growth for the number of subscribers, there are other data that indicates the existing subscribers are going to be demanding ever more capacity.

Morgan Stanley, in our report — that’s page 50, if you’re following it in the book — but a 360 percent growth in per-subscriber usage of broadband from 2008 to 2013. That’s from less than 20 gigabits a month per subscriber to over 80 gigabits a month per subscriber.

So while you’re not seeing a growth, a substantial growth, in the number of users, you’re seeing very high growth in the usage per customer.

And that, of course, has its implications in terms of prices, in terms of capital investment, and those two trends are, in a sense, in many cases, pulling in different directions.

So the new demand, by the way, obviously, is, you know, video induced. And that obviously seems to be the trend for broadband.

Will broadband support video, and ever more video

— and even 3-D video as we look further and
further ahead.

Another, I think, just sort of base line chart is, you know, a snapshot of Internet penetration of U.S. Households. And since it’s measuring households, this is basically wire service. Because the other thing that becomes obvious in our report is that households are the measure for wired broadband, and “people” are the measure for wireless services, generally.

And what this chart looks at — 2012, the last bar — 27 percent of American households, according to this forecast, would not be using broadband, with 18 percent simply not using the Internet at all, and 9 percent still using dial-up or perhaps some fixed wireless or some other kind of things. And then you have the major suppliers, cable and broadband.

What that shows, of course — and, you know, there are different reports that kind of show some slightly different numbers, percentage-wise, about the non-adopters, the non-broadband users. But there’s still a substantial market opportunity to sell broadband services, 25 percent 27 percent. It’s still a lot of additional growth available.

So the challenge for both the operators and, I guess, from government policy is how to get more people to use this existing plant.

Let me just quickly summarize the two basic points of the report. This is in there, from our Executive Summary.

At the end of the day, basic conclusions: 95 percent of U.S. homes will have access to low-speed broadband — you know, a few megabits, 5 megabits, roughly, or more, but, you know, moderate speeds. And 90 percent will have access to 50 — advertised speeds of 50 megabits in 2013, 2014. And a lot of that high speed is going to come fairly quickly cable Doxis 3.0, 92 percent of the homes by 2013 — and that’s a 50 megabit service, basically. Even a lot of the cable companies are doing Doxis 3.0 upgrades as we speak, and much of the country will be covered by Doxis 3.0. You know, your two largest telephone companies, AT&T and Verizon, just between them, 50 million homes, 10 megabits or better by 2011.

So, you know, in the relative near term a lot of homes are going to have pretty decent broadband availability from, in most cases, at least two suppliers, a cable company, a telephone company, nothing new.

Wireless broadband — thinking of wireless, it’s the 4G, 3G, includes WiMAX in here.

Speed’s getting better. We’re looking at, if you believe the companies that are putting out these systems, 12 megabits is probably a reasonable number that they seem to be talking about in terms of subscriber service. But it could be less, could be more. These are shared bandwidths. But that’s 94 percent of the population by 2013 — the

94 percent being Verizon coverage. They cover —
they claim to cover 94 percent of the POPs in the country, and they claim that they will have their LTE 4G service up by the end of 2013. So that’s the big picture.

And throughout our analysis, the downstream speeds are always emphasized by the service providers, because that’s the nice headline number. Downstream speeds, I think, I guess, inevitably seemingly forever, are always going to be faster than the upstream speeds. Some of that’s for, obviously, technology reasons. But when we listen to and think about speeds the headline is on the downstream side. Upstream is being improved, however.

So, you know, the good story — we just heard the good news.

The bad news. You know, you get up into the 90 percent availability, that still is leaving

10 percent — maybe 5 percent. But it’s still a
substantial number of households in this country are going to have an inferior choice in broadband, with “inferior” for some people meaning no choice, and other people, a slower speed. And the choice may be satellite, whether or not satellite is fully substitutable for a wireline broadband services — in many applications fully substitutable, in some others not. But generally, satellite service should be available, you know, over the contiguous 50 states, Puerto Rico, Alaska, I think, an Hawaii, et cetera.

The interesting thing that I got out of this, though, in a sense was, you know, the conventional wisdom often thinks of rural America as the place that has no fiber, has no broadband.

And one of the things I learned doing the report was, there’s actually a fair bit of — a surprising, to me at least — amount of fiber in rural America. That’s not to say everyone in rural America has fiber. But there is a fair amount of fiber in rural America.

For example, the Tier 3 telcos, which tend to include, you know, the really small independents, telephone coops, et cetera, have over half a million fiber-to-the-home subscribers

— page 15 of our report. And analysts who are
looking at this actually are suggesting that in

2009, 2010, in this sort of down economic time,
the two organizations that are going to be still doing fiber to the home are Verizon and Tier 3 telephone companies.

And then, of course, you know, I think that part of the conventional wisdom also tends to overlook WISPS — wireless Internet service providers. And there’s at least 2 million locations — homes, I guess — locations served by WISPS. And that’s not an insignificant number.

And so while rural America, I’m sure, has less ubiquitous, less serviceable broadband than much of urban America, or well-populated America, the thing I’ve got out of the report so far is that there is a surprising amount of fiber in rural America, nevertheless. And I think that’s a useful — to me, an observation that I had not really focused on up until that point.

Adoption, which really is perhaps the bigger problem. Because if we’re looking at 94 percent availability of good speed broadband in the relatively foreseeable future, you know the forecasts are that adoption is going to continue to lag behind the availability — by a fairly wide margin. The numbers we have or the numbers we mashed together from a variety of sources and sort of average out to about 69 percent of the households are going to subscribe to a wired broadband service by 2015, and a little over half of the population will have a broadband wireless service by 2013.

Again, one of the things I sort of observed, a little observation going as we gathered the data and I looked at it and thought about it, I did see, you know, pricing obviously has to have an impact on adoption. And one nugget that I thought was very illustrative was Cablevision, major cable TV company in the New York metro area, and very much focused solely in the New York metro area — it has the lowest, as far as we could tell, broadband pricing amongst the cable industry, about $37.00 a month, with the average cable industry being in the $40- plus, mid to low 40s.

And it faces competition from Verizon’s FiOS, head to head. It faces competition from RCN, a complete over- builder, offering fully, complete bundled package. And there are quite a number of other independent suppliers and competitive providers in the New York metro area.

And yet, by having the lowest price, Cablevision has a 52 percent penetration for broadband of its customer base versus the, you know, 37 percent average penetration for cable as a whole. And I said, oh, they give low prices, better penetration rate, so maybe there’s a correlation there.

But pricing might be a problem. One thing we point out in our report — page 61, if you’re keeping track — analysts are actually looking at pretty stable broadband pricing. In fact, one analyst said, “2010 should represent an inflection point, with a turnaround in the price deflation we have seen.” And another analyst basically saying — looking at about a 1 percent increase in broadband pricing over the next few years.

So you’re not seeing pricing initiatives and certainly the analysts are not anticipating pricing being a driver for adoption.

So, putting what I just said into a graph, these show the curves, at least for the broadband, excuse me, wireline broadband availability and adoption. This is, again, an average of data supplied to us by a number of analysts, as a well as a number of market researchers and support. So it’s largely a mathematical average. So there are some reports that have it slightly higher, some slightly lower.

But all of the reports are pretty consistent in terms of the big numbers — you know, the 60s, 70s in the percentages, and certainly in the trends.

One thing I’d point out here which drove us crazy, so I think it’s worth — you know, when the FCC looks at these numbers, particularly when they’re looking at adoption and deployment, deployment is generally described as “percent of homes in the area,” total homes. Adoption, for some people, is expressed as “percent of occupied homes.” But in other cases I’ve seen data where they use the same metric, total homes, for adoption.

This chart, we’ve attempted to get it down to adoption of occupied homes. There’s about, from what I saw, 7 or 8 percent difference in the number of homes if you’re looking at all homes or occupied homes. So it’s, first of all, if you see some data inconsistencies, that would be one thing to check. But I think also it’s probably — I would recommend, you know, that “occupied homes” is the right metric, but, in any case, have a consistent metric, whichever you choose.

Wireless broadband penetration — again, this is an average of the various analysts forecasts and from other information that we have.

It doesn’t include short-message service. That’s not a broadband service. And it does include laptop wireless cards in the metrics.

And what you’re seeing is a pretty strong growth forecast through 2013, getting up to over half the population age 14 and under. We cut it at age 14. It happened to be a population number that came, that the Census Bureau has. It didn’t seem to me that, in the same way that you shouldn’t really be looking at “unoccupied times” in terms of adoption, you know, three-year-olds with data services just didn’t seem quite the right metric. But it’s, again, something to think about in terms of how you actually — what you measure, your denominator, numerator always make a difference.

Boy, now this is a tough subject, capex.

First of all, the service suppliers themselves, the broadband network operators, for some very logical reasons, business reasons, don’t give much forecast of their capital expenditures. And that’s for competitive reasons, for Securities and Exchange Law reasons, in terms of future looking numbers. And also, frankly, you know, investors keep an eye on that forecast. And until the company really knows that it’s going to incur a massive capital expenditure, it probably isn’t going to say much about it.

And looking ahead, you know, three or four years in this business, it’s kind of hard to make really strong predictions about — unless you’re going to do something like a FiOS project, where you’re literally going to be tearing out one kind of network and rebuilding, putting in a new one. Absent that kind of a mass initiative, you’re probably not going to see too many forecasts.

The result of that, the reason I emphasize that is, the tendency for analysts and for investment looks at broadband or capex, telco capex, it’s always going to have a tendency to be pretty flat in the out-years, unless there’s some well known inflection, change, coming.

I mention that because, in fact, our forecast that we have here, which is based on, in the early years, various analysts forecasts and other data, but extrapolated, it tends to be flat.

And I will explain, I think there are some reasons why flat may not be the — or there are things that could change where substantial growth could happen. But generally speaking, we’re looking at flat numbers.

So, you know, the carriers themselves, the service providers don’t — rarely, unless they’re only in the broadband business, they’re multi-product enterprises, at most, they talk about their total capex, and then you have to kind of derive what is broadband capex.

Now, recently, AT&T, for example, did say that two-thirds of its capex, both wired and wireless, was for broadband. So that was an interesting data point. A market researcher that supplies market intelligence to the manufacturers of broadband equipment, and therefore provides market data and surveys, did a study, and that’s the — for telco capex for broadband, the assumptions, so the 48, 52, 54 percent, 58 percent, numbers you see in the “telco” section, the third like down, that was a market research estimate of how telco broadband is being set out.

But all of these are, in a sense, fairly subjective, rough allocations in a multi-product line business. And these are best estimates, I think, by the analysts, and best estimates by our team on what to expect for broadband capex. And there’s a whole bunch of footnotes and explanations in the text of the document. I will not summarize them. But if you’re interested in actually knowing how we derived each of the numbers, I believe it’s self-explanatory.

That chart, in a graphical form basically shows flat, relatively, broadband capex

— in the $30 billion a year range, though.
That’s not chump change. And the total capex declining, but then flattening out, declining to in the mid-50s. So the legacy, if you want to call it that, networks are not being funded.

They’re not growing. That makes sense. And the broadband networks are growing, are being funded.

Now, my earlier comment was that the actual numbers of subscribers for wireline broadband isn’t growing very much, but volume is.

So, for example, if we go to the next chart here, which breaks down the total capex into the three industry sectors, wireline, broadband — cable broadband, and wireless broadband — I’m colorblind, so all I can tell you is the telco is the diamond, cable is the square, and wireless is the triangle. For those who have color, you’re fortunate.

What does this show? It’s another iteration of what I’ve basically said — but showing wireless, which is the growth business, we saw that both wireless adoption still has a long way to go in the next few years. So there’s obviously going to be a lot of wireless investment — 3G, 4G, is coming out being — 4G presumably is almost exclusively a broadband investment, not really a “telephone” investment.

How much of 3G investment you would want to allocate between broadband and telephone might be debatable. But you see the wireless is growing because wireless capex is growing, broadband capex is growing, because wireless broadband is a fast growing business.

Cable broadband, telco broadband, pretty flat. No new, there’s no new big Verizon announcements in the telco sector. The cable industry itself has a very low capex requirement for broadband. The major — Comcast expects to have its Doxis 3.0 project finished by the end of this year? 2010, I think. That’s 50 percent of the country, right there. And the upgrade for a cable company to get up to Doxis 3.0, you know, is $15 to $30 for the network upgrade, and then another $65, roughly, for the modem at the customer premises. You know, a hundred bucks, plus or minus, for a cable upgrade. And most of that — 92 percent of the cable industry — is going to be fully complete by the year 2013, which is that little dip on the rest of the cable. And then after that, presumably you’ve got maintenance and just general upgrades and things like that for the capital.

But, but, but flatness may not. It may turn out that the capex isn’t flat. The question really is, is 50 megabits, where most users are going to be by 2015, which is sort of the end of our time horizon — because the flatness in the forecasts could well be just the natural place in the cycle. We’ve invested from, you know, the

1990s and the early 2000s, billions and hundreds
of billions of dollars in broadband. It was a big peak. That is a cyclical — these things are cyclical. It comes down and it kind of flattens for a while. All these cycles flatten for a while because the investor is now trying, the organization and investors trying to reap some of the financial benefits of that investment. But then, for a variety of reasons, a new cycle begins.

So the flatness that our forecasts show may be attributable to simply because it’s the companies and the analysts don’t yet want to forecast capex increase for financial reasons and strategic reasons. Or it could be it’s just the flat part of a cycle.

And what I gathered from, you know, reading the materials and trying to digest them was that, you know, I think there’s sort of a —

50 megabits becomes probably a pretty important
breakpoint. Because DSL, you know, DDSL-2, 4, you know, whatever, you know, DSL can be stretched and can be improved. And I guess the good news for the telephone industry is that there are a steady loss of POPs lines is freeing up an awful lot of copper lines that can be redeployed as DSL services.

But it gets to be pretty capital intensive, labor intensive to, you know, keep doing DSL, DSL, DSL. So maybe 50 megabits at least seems to be a limit that I’m hearing about, or reading about, from the technology for DSL.

And similarly, Doxis 3.0, it can go higher than 50 megabits per subscriber, but only by continuing to make the Ethernet rings shorter and shorter and shorter.

And so at some point, I think both the telephone companies, the DSL-oriented phone companies and cable, if you go much beyond 50 meg as the demand, the baseline demand, fiber to the home starts to look like a pretty logical way to go, or they get pushed to do that.

So when and if — my conclusion sort of is, when and if 50 megs, the 50-meg barrier has to be broken for competitive reasons, for marketing reasons, or because it’s simply time to deploy a new technology because the old technology is getting obsolescent and two hard to maintain, et cetera, et cetera — until you get to that point, you’re going to probably keep these sort of flat capex.

Personal view? Longtime CITI view? We think the future of the world is ultra-broadband, which we define as a gigabit to every home. So we think in the long run, you’re going to get a lot of capacity to most homes in America. By the year

2015? Probably not. But I don’t know anything
about 2016 or ’17.

I think another useful thing to think about when we talk about capital is we’re in the middle of what we think is about a trillion dollar decade. And a trillion dollars has been invested or spent on broadband in this country, not only by the service providers, but by the service consumers. And we, as individual consumers, have bought modems, we’ve bought routers, we’ve bought computers, we’ve bought software. Businesses have done the same, all to take advantage of broadband.

And all of that really has — the total network broadband infrastructure, the whole broadband ecosystem, has to include what individual consumers have spent. And it’s about

— you know, we did some math and looked at the
numbers of various of these devices that have been purchased over these years, and they’re in the millions and tens of millions. And, you know, we come out to about a billion — excuse me, a trillion dollars, both what has been spent in the last five years, and looking ahead in the next five years. And that’s not an insignificant amount of money.

So, you know, for a country this — you know, I think we’ve done pretty well spending a trillion dollars, actually.

Some quick observations. We have eight observations. I’ll only just quickly mention two.

And I’ve already mentioned this one, so it hardly bears repeating.

Rural America may not be as underserved as expected. I see — as I say, I was surprised to see so much fiber, so much WISP activity, in rural America. This is not to say there isn’t a big problem in rural America, but there are parts of rural America that are doing very well in terms of broadband. And I was originally quite skeptical of the whole idea of the broadband mapping program. I’m now, after looking at this, going, yeah, it’s probably not a bad idea to figure out really where broadband is very sufficient in rural America and where it isn’t, so we can really focus on just those places. Because there’s a lot of places in rural America that have pretty good broadband.

I’ve pretty much covered all of my other observations. Anyway, they’re in the last two pages of the report.

The only other observation I’ll mention is another relatively traditional CITI view of the world, which is we just can’t quite — as much as we would like — and I come, in my background, from the competitive side of the industry, with the (inaudible) industry in the late ’80s and throughout the ’90s. Professor Eli Noam has been a longtime advocate of competition, both academically and when he was a commissioner at the New York Public Service Commission. And we, unfortunately, can’t figure out how there can be many more broadband infrastructure competitors in this country. And we would like it to be so, we just haven’t figured out how it’s possible because of, you know, basic economies of scale.

We see that — and the reason I bring this up, this was not — the report reinforced our previous thinking about this. There’s nothing we saw in our research that indicates there is, you know, a new major thing coming. Clear-wire would probably be as close as you’d come to a new infrastructure operator. WiMAX services. But only in, I think, 40-odd markets across the country and kind of affiliated with the incumbents. It’s not truly a new entrant. It’s a wireless, it’s the wireless arm of cable incumbents and Sprint.

So while we would like to see infrastructure entry, for lots of reasons, we just don’t think it’s that likely, and we think the report kind of heads that way. And we also always note that further concentration is always possible.

Thank you.

MR. KOUTSKY: Thank you, Bob. And now I’ll turn it over to Lee Rainie of Pew for a few minutes of response.

MR. RAINIE: I’ll take my cue from Tom Hazlett and do three things.

First of all, I think this is an impressive report. The assignment from the FCC was shrewd. I mean, you have lots of data from lots of other places, and this kind of direction was useful, and produced useful information that is sort of unchallengeable, at least at the level of (inaudible) reporting and doing the analysis of it.

My one quibble is the projections that the analysts gave about the penetration rate itself. Right now, or as of the spring of 2009, we saw 63 percent of people — so our unit of analysis is people rather than households. But 63 percent had broadband at home.

We’re collecting data in the field now.

It looks like it’s a little bit higher in December than it was then. I gather that there are independent surveys going on that also put it in the mid to high 60s in some places.

So my guess that the projection that penetration of broadband in households won’t even reach 70 percent by 2015 could be low. It’s what you heard. I absolutely know that that’s the facts you’re reporting. I just don’t see it so.

And partly, it’s because — MR. ATKINSON: And you are actually looking at occupied homes.

MR. RAINIE: Well, we’re just talking to people who talk to us on the phone. They’re occupied bodies, for sure. So, just alone, just looking at sort of simple projections.

We also have not ever seen the kind of substitution that’s been referred to here and elsewhere, where there is no net new gain. We saw massive substitution through 2000s, in people moving from dial-up to broadband, obviously. And we are seeing sort of new entrance into the field coming through mobile connections. There are people, particularly in minority communities, and particularly in relatively lower-income households, who are starting their Internet experience on their handheld devices. And it’s very likely to be a pathway to the richer experiences that you get on more robust devices, with more robust connections.

And so there’s a portion of the population that we are not really sufficiently even capturing now with our data that I would guess would be going on-line.

There’s always sort of churn in the Internet population. But interestingly enough, there has been less pullback from the Internet in the period of the recession, according to our data, than there has been from other telecommunications things. There are more people who are changing their cable plans than are changing their Internet plans.

So to the degree that they are sort of voting with their pocketbooks about what matters more to them in the long term, it seems like the Internet is the bet that they’re placing, rather than other kinds of information expenditures.

And so I dispute the sort of — or I wonder if we’ll be sitting here in 2015 and having as low a penetration of broadband, just by the natural force of things as you are documenting.

What I did deeply appreciate about that chart — the Figure 17, where the availability curve was so different from the adoption curve — is that it highlights things that we all know about Internet users, and have consistently seen in our data from the time we began to study non-users — which is there are significant numbers of people for whom availability is not the issue. They are not Internet users for a variety of reasons. And we’ve broken them down into four buckets.

So, you know, right now we’re saying 63 percent of American adults have broadband at home.

Among those non- users, about half of them say that it’s simply the relevance of the technology that matters to them. They say they don’t want it. They say they don’t need it. It wouldn’t improve their life in any measurable way, or they’re happy with the life circumstances and technology that they’ve got.

About a fifth of them cite price itself.

It’s too expensive. They lost their computer.

Their capacity to pay providers is at issue.

Seventeen percent of non-users say that availability is an issue. So there’s something going on when there’s that much availability and people don’t know it. With all the marketing that’s being done, with all the ways that sort of our culture is aware of what’s going on in this space, there are just people who have it in their communities and don’t know it.

And so I would hold that out as one of the sort of public policy challenges that the FCC will face and be smart in addressing as it’s thinking about bringing broadband to more people.

The other sort of related point — to digress just a little bit — it’s fine to think of the household as the unit of analysis, particularly when you’re looking at investments, because that’s where you’re taking it to. But interestingly enough, in the user population, there are a notable portion of people who live in households with Internet connections and consider themselves non-Internet users.

Since 2001, we’ve looked at this 4 times, and the most current data we have are 13 percent of self-described non-users of the Internet live in homes with an Internet connection. Other members of their family are going online, oftentimes as a secondary sort of situation going on here, where they are getting members of their family to do searches and e-mail exchanges and document gathering and research and stuff like that for them. But I would just point out that bringing it into homes isn’t necessarily going to get you everything that you want in terms of the benefits of broadband.

And the final way that we see people, non-users, being reluctant to embrace these technologies is usability issues themselves.

They’re scared of the technology or they’ve had bad experiences and they don’t want them replicated because it’s expensive when you’re computer goes down. And it’s hard to fix sometimes, and it’s not very consumer, you know, friendly technology at times.

In addition to thinking about non-users, there’s a third point I would make is that beyond the demographics that Yochai was talking about before, we also see non-users falling into two other groups of people who are independently sort of disconnected from the Internet in general, and broadband in particular — the disabled. Those who have significant disabilities in their life, even if they have the economic resources and, potentially, the education, are sometimes not users just because it’s too hard to master, or the extra expense of the technology to make it usable is too great for them to bear.

And in America, those who don’t feel comfortable speaking English are much less likely to be Internet users. Language alone, language proficiency alone, if you hold steady for other demographic factors like education and income and stuff like that, language proficiency alone is also an independent predictor. So it’s something for you guys to consider.

And I guess, for my last point — point

3.a, maybe — would be the most interesting stuff
to me in this is the degree to which we’re thinking about communities, and households, as the recipients of policies and changes in the way we are deploying these technologies.

It’s still true that a lot of people don’t yet know all of the benefits that these technologies can bring to them. It’s certainly true of non-users. And when we ask them sort of what do you think the Internet is, very often they cite all the problems they’ve heard about in the media: It’s full of pranks, it’s full of fraudsters, it’s full of predators, it’s full of people who are going to steal your money. Why would I want to bring that into my life is their basic argument.

But there’s another portion of people that we haven’t quantified but we talk to them often enough to know that they’re out there, who simply do not know that there’s quality health information to be gotten online that will change your relationship with your doctor, and change the nature of the health care you can bring to yourself.

There are people who don’t know they can interact with their government online. There are people who don’t know that they can embed themselves more directly in their communities.

And so I would hold that out as another piece of the public education effort, or even the commercial education effort that goes into encouraging people to think that these are a set of tools that can bring real change to their lives.


MR. KOUTSKY: So both reasons for optimism, and reasons for concern.

Do you have a couple questions?

MR. KENDALL: Yes, so first off, I’d like to thank you guys for I’m sure what was a lot of work and a great job, building a great fact base for us to use in the National Broadband Plan.

I think it’s going to be very helpful.

But I did have a couple of questions, more about your opinions of what you think is the reason for some of the conclusions.

So one of them was, why are Tier 3 operators deploying fiber and not Tier 2 operators deploying fiber? Is that because there is more favorable economics in their areas due to lack of cable competition? Additional USF support? Is it just a different capital structure.

Is there anything that you would put forth?

MR. ATKINSON: Well, I flipped to my report. Page 16, the research firm that, you know, sort of gave some data on rural Tier 3 ILECS said, “drivers for the rural independent telcos to deploy fiber to the home include aging copper lines in need of replacement.” But that probably also applies to Tier 2, and even Tier 1, “the opportunity to deliver video, given a more robust platform.” Well, that also applies to other telephone companies. “A pioneering tradition,” that’s probably pretty unique for the Tier 3 companies. These tend to be almost family-owned or cooperative telephone companies, so these are people who have, you know, a real stake in the enterprise and, you know, they were, their parents or grandparents laid the wires, and strung the wires when the telephone service first came. So they have that sort of a family or community tradition of that.

And then it says, “And in some cases, subsidies, such as rural broadband loan programs and universal service funds.” I have to think that those two factors are what distinguish the Tier 3 phone companies from a lot of the Tier 2, and certainly the Tier 1 phone companies. And in a sense, you can say that the government programs

— to the extent that rural loan programs and the
universal service funds, in fact, are the “different” factors that apply to those Tier 3 companies, the government has done a tremendous success. But maybe it’s time to stop, for some of them.

I mean, I did a BTOP review for the NTIA, and it was for a — I can’t get, obviously, into the details, name of the company, et cetera, because it’s all confidential, but it was from a rural coop. And in its plan — and this was for a public computer center grant. It was very proudly saying it’s a hundred percent fiber to the home.

And their finances are — I would like to own that company. They are doing really, really well.

And so I — you know, I just looked at it and went — I was stunned, actually, to read this company’s financial information.

MR. KENDALL: Did they have really high penetration?

MR. ATKINSON: They have 100 percent penetration. I mean, they are the only — they had the cable company and the phone company. They are the only thing in town — and it’s all fiber, and they’ve got a great — it’s a number of towns in a very rural part of America.

MR. KENDALL: So I had one other question about — as you had the second section, I believe, of the report that talks about announcements, and how well companies have done in meeting those announcements.

And just where do you think LTE might fall, as you look at some of the — you were saying — is that more like a DOCSIS 3 that is an upgrade, and so clearly they’re going to hit that?

Or is that more like a facilities-based upgrade, so it might fall more into the category of the satellite or like WiMAX, or some of the 3G?

MR. ATKINSON: (inaudible) it’s going to be a bit of both. To the extent that it’s just putting more radios on existing towers, then it’s just a facility upgrade, largely. And, you know, many of the towers are going to be fiber-fed. And if fiber-feeding to a tower is critical, then there could be delays. And if new towers are needed, there could be delays.

And I have some recollection, in one of the reports, of reading that, you know, the basic

4G rollout is going to require quite a few
thousands, tens of thousands, possible, of new towers.

So new towers, and even running fiber to old towers, you get into all those right-of-way issues, siting issues, et cetera, et cetera, et cetera. And those would be, I would think, the risk to the schedule for LTE or any — and also for WiMAX or anything else.

If the tower exists and is fiber-fed, no problem. If it doesn’t exist, or probably needs fiber to accommodate the increased capacity off that tower, siting and right-of- way problems could be a problem.

MR. KOUTSKY: I have two very quick ones. One that was actually an outgrowth of the rural income (inaudible) point. You mentioned that there were instances where rural telephone companies were also the cable company. Did your research ever — did that reveal or give us any indication to (inaudible) with that?

I’ve seen or heard of situations that perhaps as many as 30 to 40 percent of smaller cable companies may be affiliated with the incumbent LECLETs in town.

MR. ATKINSON: Not — no. I mean, we got some data from the American Cable Association.

And, you know, I think they have 483 — I never remember numbers, but that one popped in. So, I think 483 small cable companies. And then we have the NTCA — yes — NTCA, small rural. And they, you know, there may be overlaps. They may be operating in parallel. I guess I would expect there to be a fair bit of overlap, where the ACA member might be part of a rural co-op. Certainly a lot of the co-ops are providing video, but I don’t have any data on that.

It’s probably worth trying to look at the ownerships and see, or go co-op by co-op, and independent phone company by phone company and try to map that out.

Maybe that’s one thing the broadband mapping will show.

MR. KOUTSKY: And my other question, actually related to this, the capex percentage slide that you had, which actually I thought — I just had some questions, because it showed that cable spent about approximately 20 percent of its capex on broadband, I think was the number that you had. ILEC was around 45 to 55. And wireless was at 60 to 80.

Is this because — I actually found that to be, I haven’t seen it presented before until I saw your paper, and I thought it was interesting to see that. Is it because we’re at different deployment, stage of the deployment cycle there?

Or is it a function of the fact that wireless is a fast-growing business? Because they’re the ones that kind of leap out here as being, you know, significantly higher than the wire-line competitors.

Is it something about wireless that requires them to plow more of their capex dollars into broadband, as opposed to wireline?

SPEAKER: (inaudible) MR. ATKINSON: Well, I will tell you that we looked at the numbers. The cable numbers, I’ll be honest, they look way too high. If you took, you know, if you said, well 50 percent of cable is now broadband, I mean, what percentage of cable is television service versus telephone service versus broadband service? And that’s a tough allocation.

We tried to do a bottoms-up, though. We took the numbers that we had on the cost per subscriber for the network upgrade, which was the $15 to $30 per subscriber, and then also added forecasts for replacing every cable, or giving every Doxis 3 subscriber a modem at their home.

And we used a figure of like $65.

And so we did a little bit of a bottoms-up and said, okay, that’s going to be all done in the timeframe 2009 through 2013. And so we said, doing it that way you’re actually way less then 20 percent.

But then one of our other sources said that the cable industry spends around 13 percent of their capital on network improvements, upgrades and extensions — replacement. And so we added that to the mix. And I guess that’s how we ended up with approximately in the 20-odd percent range for cable for the broadband.

It seems low, because then I sit there and go, well, that means 80 percent of cable capex is for television and telephone. Again, a lot, 50 percent of cable capex is for customer-premises equipment: Set-top boxes, DVRs, cable modems.

And maybe DVRs and set-top boxes are way more expensive than cable modems, so maybe that’s part of the reason that you can expect these relatively high numbers. Or that the remainder is apt to be allocated somehow between telephone and television.

But that really goes to the whole point that these are allocations.

SPEAKER: And even (inaudible), on the wireless side, the consumer purchases the handset, so that — because there isn’t a capex component to that.

MR. ATKINSON: Actually, I believe cell companies include — because they’re really leasing you the phone, the handset, I think their handset costs go into capital, although I’m not an accountant and don’t pretend to be.

The other thing I would say is that, you know, with — as I think I mentioned in the remarks, you know, 4G I would allocate almost 100 percent to broadband. That’s its purpose. So the allocation to broadband for wireless in the high numbers, in the 80s I guess we get up to, don’t we

— yes, 85 percent — that made intuitive sense to

The cable 20-odd percent, I go, hmm.

I’d be interested to hear from the cable industry, if you’d like to provide more detailed information about capex, please send it.

MR. KOUTSKY: Questions from the audience, and we’ll try to keep this to just a few minutes. But let’s go there.

Him first. He beat you.

SPEAKER: Yes, a clear point of differentiation between the two reports this afternoon related to FTTH, Yochai said that, you know, he kind of equated the Verizon FIOS build-out with FTTH in the United States. And you also mentioned the rural component, which is quite active. I think Michael Render, in fact, said that there’s more than 400 providers in one form or another providing FTTH.

If I could get you to think outside the box maybe just a little bit, what kind of implications would you see in terms of broadband policy going forward, knowing that there’s an energetic, active dimension in the rural community bringing FTTH?

MR. ATKINSON: Well, if I were to be having the horrendous job that these guys have, I would actually try to figure out what makes, why are some of these companies moving ahead, why is rural, this co-op that I looked at 100 percent FTTH and, you know, the one down the road, or across the state zero? What is different about them?

Because clearly some of them have the money, the will and the market demand to push out fiber to every rural home. It may be demographic.

It may be financial. It may be management will.

It may be — you know, there’s hundreds of co-ops, there are hundreds, thousands. I think if you add them up there’s probably, what, a thousand small independent phone companies in America? Co-ops and truly independents.

Each one of them is different. They’re going to have a very different — but they, hopefully, they would fall into some boxes and so you say, ah, look, these fiber guys have these characteristics, now how do we try to get these laggards to have the same characteristics?

Particularly if it’s financial or management, or something that government policy can work on. You know, the government can’t change mountains and the geography aspects probably, but there’s a lot of things a government policy could affect. So that’s what I would do first.

MR. KOUTSKY: Actually we have a question from the Internet that’s somewhat related, so then I’ll ask that one and then go.

Actually, through the magic of the Internet, Professor Bronwyn Howell from New Zealand has submitted us a question. And I last saw her here at TPRC. So I was glad that she’s online and listening to us.

She mentions something very, a similar type dichotomy. And we’d like the panelists’ comment on the extent to which New Zealand experience can inform U.S. Policy.

And she mentions that in New Zealand, there was relatively widespread availability of ADSL, with some entry, but generally had a slow broadband uptake on the demand side. So, in essence, she summarizes and says that there is negligible supply-side problems, but considerable demand-side problems in New Zealand.

And she asserts that regulation has been impotent in addressing the demand-side issues, but has arguably led to some substantial demand-side distortion, such as withholding investment, or strategic gaming between participants, and slipping on absolute and relative performance.

I’ve actually — that’s a rather educated question, but it seems to be somewhat related the question Bob just raised, which is how do we begin to think about whether companies will make investments. You know, do you agree that there’s the potential for some distortion or some mismatch between the economics of supply and the economics of demand, particular in some of the more rural and insular areas of this country?

MR. ATKINSON: Well, the first thing is, I would have to go to New Zealand and do at least a three or four week on-site analysis.

Do businesses react, any business react to regulations and the possibility of regulations?

Sure. But they are ultimately wanting to, you know, thrive and survive. And I think what we’ve seen, I’ve seen, over the last 30 years in the telecom industry is that, you know, the companies are pretty agile and very creative. And, you know, both new entrants and the incumbents, in dealing with whatever government throws at them.

They want to survive.

And I don’t think people — the company I worked for didn’t make our investment decisions on the government policy, per se. We made our investment decisions, I mean, we had to be authorized to provide the service. But after that, once we’re in, had the certification, all of our investment decisions were simply made on business, you know, customers, serving customers.

And the government was like, yeah, yeah, yeah, we’ll work around it. We’ll do whatever we have to do.

I think most, in a competitive marketplace, you really can’t spend a terrible amount of time worrying about the government, because that’s the environment, you’re dealt the hand you’re dealt. You’ve got to play in that hand.

Will, more or less? I guess it’s possible. And I’ve certainly seen a lot of the studies and arguments that government policies can incent or disincent investment by companies. I personally think that companies do what they have to do. And what that tells me is they will not invest if they — that’s why this 50-megabit breakpoint to me is, why I emphasize it is, you know, there is no incentive for any company right now to make major new investments in, you know, a new, in fiber to the home, or big investments, as long as they’re doing, getting as much revenue as they can from the existing plant. When you have an existing sunk- cost plant, you want to milk it for as long as you can, and make it pay for as long as you can.

And until the existing plant — and that little quote I just read about rural America. You know, if your outside plant is deteriorating, the operating expenses are extremely high, then you start saying, in order to make more money, make higher profits, I need to replace it with something bigger and better. And then at that point you simply say, I’m not going to make an incremental addition, I’m going to make a big change, because I can see financial benefit to me.

Then they’re going to do it for business reasons.

MR. KOUTSKY: Thank you.

SPEAKER: (inaudible) substance of my question is how to get past that fork in the road.

What’s going to spur companies to do that, pass that 50 megabit per second (inaudible)?

MR. ATKINSON: Well, some companies, of course, have the ability today to offer more than

50 megabits. The question is will customers flock
to those services? So if Verizon, for example, you know, pushes the envelope and offers some really high-speed services at a pretty low, relatively low price, and customers march with their feet and their wallets to 80-megabit or

100-megabit or 150- megabit services that are
priced only 10 percent, 20 percent more than — that’s going to cause everybody else to go, darn, it looks like we’re going to have to really spend

— particularly if I’m a cable company in an area
where Verizon also serves.

At the same time, if you’re a DSL-oriented telco, for example, and you see that it really can pay, well, there would be a tendency to do it themselves and start rolling out the speed.

So there needs to be, someone has got to start a cycle going. And all these, the businesses and the marketing and consumer adoption are a series of cycles within cycles. And right now there’s no impetus, no one has yet started the really high-speed cycle. And we’ll see where it goes.

MR. KOUTSKY: Do we have further questions? We have time for one more question, so Stagg gets it.


MR. NEWMAN: (inaudible) markets where they’re already operating (inaudible).

In markets where they’re really offering these really high speeds, is there any evidence of different, substantively different, usage patterns, applications use or anything like that?

MR. ATKINSON: I don’t have anything like that. No. It would be useful, but don’t have it.

MR. KOUTSKY: All right. Thank you, we are well over our time. But it was a very interesting afternoon, and I appreciate the participation of Bob and Ivy and Lee today. And also appreciate everyone in the audience for participating.

Just to note, both reports are posted online. And the Commission does have, has put them in the record, and parties are free to submit their comments, responses, supplements, gripes and platitudes into the FCC record, as well.

And so we definitely appreciate the vibrant public debate that we’ve had about these.

Thank you. (Whereupon, at 3:42 p.m., the PROCEEDINGS were adjourned.)