High Speed Broadband Matters, But "Why?"

Policy advocates are right to argue that broadband access matters. But “what” matters is likely more than simple matter of making high speed access widely available and affordable. What really matters is whether people and businesses figure out ways to use broadband in a productive way.

Faster broadband, and affordable broadband is important, most would agree. The issue is why faster broadband brings benefits, what the benefits are, and how they can be realized. “The other half of the story is usage, the extent to  which people are active with digital technologies and applications, incorporate them into their lives and work,  and gain benefit from them,” Booz & Co. analysts have said.

In other words, how broadband changes life and work is the issue, not simply the fact that broadband is available.  

That is not to say studies do not show a correlation. Some studies of broadband deployment have estimated that a 10 percent increase in broadband penetration (adoption by people)

contributes a per capita GDP gain of just 0.16 percent to 0.25 percent.

The Booz & Company Digitization Index, which measured both the direct and indirect economic impacts of “digitization,” found that an increase in the “Digitization Index” score of 10 percent correlates with a 0.50 percent to 0.62 percent gain in per capita gross domestic product. Note that the phrase is “correlation,” not “causation.”

But that is largely beside the point. The consensus is virtually unanimous that broadband is an “essential” sort of infrastructure, without which an economy cannot grow as fast. Whether the relationship is causal or only correlational cannot fundamentally be ascertained.

But it doesn’t really matter. People act as though broadband does matter, and people will act on that belief. What probably matters is the way the “acting” takes place.

If people use broadband to watch TV, that’s a rather trivial reason to invest too much in broadband access, especially when there are other compelling places to invest capital. Also, the history of technology innovations suggests it can take quite a long time for people and organizations to learn how to use new technology effectively.

In fact, studies of  productivity are a hazardous undertaking. Some would note, for example, that U.S. productivity growth has been in long term decline since the early 1970s. You might argue that application of computing slowed the rate of decline, but that is not what people generally think.

In fact, studies of productivity in the 1980s, when computers first became ubiquitous in U.S. businesses and organizations, do not show positive changes in productivity. The Internet, on the other hand, does arguably seem to have changed the productivity curve.

On the other hand, some might point to a productivity gain from 1996 to 2006, and there is some thinking that a shift to Internet processes might explain the temporary gain in productivity rates.  

In fact, the value of high speed broadband might be overstated, as an economic development driver.

How Much Does "Ditching Phone Subsidies" Actually Change?

As T-Mobile USA and Verizon Wireless move to decouple recurring service plans from sales of devices, one question some might have is how those policies will affect service providers and consumers.

It isn’t so clear, even though some service providers tout “savings.”

For the carriers, separating device sales from recurring service plans allows them to market service plans at “lower cost,” since any installment payments or “device subsidies” are separated.

If installment plans are substituted for “device subsidies” to any great extent, that move alone will not markedly affect service provider operating costs or potential profit margins, though, since the carriers still will have to purchase the devices and then recover the costs over time.

It isn’t even so clear that most consumers will save money when device installment plans replace sales of subsidized devices. Recurring out of pocket costs might not be less under the new plans, compared to the older plans

It’s hard to ascertain the real prospects for actual movement by the U.S. federal government to mandate mobile phone unlocking in the U.S. market. Some argue that the benefits to consumers would be significant, as “phone unlocking” would create a  competitive situation more analogous to Europe.

Some might say the European experience with device unlocking is less useful in the U.S. market simply because the U.S. market has two major and incompatible air interfaces.

Even if all devices were sold “unlocked,” users would have less choice than in Europe, where all carriers use the GSM air interface.

Unlike many other countries, subscribers and networks are at the moment split into GSM (AT&T and T-Mobile USA) and CDMA (Verizon Wireless and Sprint) air interface camps. An unlocked CDMA device cannot be used on the AT&T and T-Mobile USA neteworks, while a GSM unlocked device cannot be used on the Verizon or Sprint networks.

Presumably, the value of unlocked devices will becomes a bigger actual value to consumers only when Long Term Evolution fourth generation networks are fully established in the U.S. market, and all LTE devices can be used on all U.S. networks.

But the matter is more complicated than often seems to be the case, in part because sale of unlocked devices might, or might not, affect end user behavior all that much. Carriers seem to recognize this by switching to installment plans, instead of requiring outright retail purchase prices.

Not many consumers are going to prefer shelling out full retail price for the latest Apple iPhone, even if they can afford to do so.

If all devices must be sold at full retail, basic economics suggests that people will trade down to cheaper devices, or stretch the amount of time they keep an existing device, if a switch to fully unlocked phones were to occur.

Installment plans now available from T-Mobile USA and Verizon Wireless address that issue.

The point is that the assumption mandatory unlocking would lead to “more choice and lower prices” is not as clear cut as might seem to be the case. 

It is the ability to get a new and “better” device every 18 months that might be the value consumers get from subsidized phones, whether unlocked or not. The installment plans might provide such value, at the same time heading off pressure to mandate device unlocking.

Nor is it clear that consumer device replenishment rates or demand for “used devices” will change, either way.

ReCellular is said to be one of the largest U.S.-based mobile phone refurbishers, reselling or recycling 5.2 million mobile devices in 2010.  ReCellular sells about 60 percent of its phones in the U.S. market and the rest mostly to dealers in Asia, Africa, Latin America and Eastern Europe.

But global sales of used phones total only a few hundred million units a year, estimates Andy Castonguay, an analyst at consulting firm Yankee Group. That compares with the 1.6 billion new devices sold worldwide in 2012.

Also, according to one U.S. study, the typical mobile device is used 18 months before being replaced. Whether that would be different, with separated device and service plans, is unclear.

At least for the moment, demand for sales of “unlocked” devices should be dampened by the growing trend to separate device subsidies from the cost of mobile service. The theory behind sales of unlocked devices is that consumers would see lower prices, and gain more freedom to switch providers.

The former should happen, in one sense. Consumers will see lower retail prices, if only because the hidden device subsidies are removed. For a while, the ability to switch providers, in the U.S. market, will be sharply limited because of differences of air interface.

Perhaps in an attempt to head off regulation that would force sale of “unlocked” devices, service providers are shifting in a direction that could provide greater buyer transparency, potential for some cost savings and greater freedom to switch providers, though not as much as backers of unlocked devices would prefer.

Mobile Offload Might be a Top-3 Value Provided by Fixed Broadband Networks

By some estimates, as much as 60 percent of all mobile broadband traffic consumed by users of smart phones will be offloaded to the fixed network, by about 2017,

By that point, only 40 percent of the 90,000 Petabytes of data generated by smart phones, feature phones and tablets  will reach the cellular network by 2017, as majority of the data traffic will use a Wi-Fi network, according to a study by Juniper Research.

That statistic could lead one to conclude that a vital role for fixed networks of the future is  mobile offload, since, if Juniper Research is correct, the overwhelming majority of mobile device traffic will move over a fixed access network, with a Wi-Fi “tail circuit.”

One might conclude that video entertainment also will be among the top-three reasons for buying high speed access. Video entertainment likewise will be one key reason for people to buy high speed access, even if they do not use PCs or the Internet. 

They might not use PCs or the Internet, but they do watch TV. In fact, it long has been true that the long term way to drive high speed access penetration to nearly 100 percent is to shift TV distribution to the Internet.

That explains why Fon, the global Wi-Fi network, will be working with Deutsche Telekom, the leading German telco, to build Germany's largest Wi-Fi network, which will launch in the summer of 2013.

“WLAN to go”  will create what is said to be the largest public Wi-Fi network in Germany. Separately, U.S. cable operators are working to bulld a seamless national network of cable-affiliated hotspots, while Verizon and AT&T both offer nationwide hotspot services as well.

KPN also has partnered with Fon in the  Netherlands. As part of that deal, KPN customers will share a portion of their own home bandwidth in exchange for free access on shared broadband connections of other KPN broadband customers in the Netherlands and Fon customers in other countries.

Belgacom , Belgium’s largest telecommunications company, has a similar deal with Fon. In addition to WiFi access in Belgium, Belgacom internet customers will get global access to the hotspots of Fon’s global WiFi network.

Oi, the Brazilian service provider, is working with Fon on a project in Rio de Janeiro.

The point is that TV delivered over the Internet will provide a reason for people to buy high speed access, even when they don't use PCs or the Internet, as such. On the other hand, one already can note that many people who "don't use the Internet or PCs" actually rely on smart phones.

Add smart phone connectivity to TV and high speed Internet access obviously becomes the foundation service for all fixed networks.

Are Tablets "TVs?" Sweden Says "Yes."

In most countries, voice communications, Internet communications, broadcast media, cable TV and broadcasting have been governed by distinct sets of regulations. That made more sense in an era when each type of service was provided by a distinct and purpose-built network.

These days, as all media types can be delivered by all or most networks, there will be a bigger discontinuity between the older forms of regulation and the ways services are created and delivered, across networks.

Sweden's new policy of taxing use of PCs and tablets to watch the state-owned TV service, and a German decision on copyright fees, neatly illustrate some of the regulatory challenges that accompany changing communications and entertainment ecosystems.

Traditionally, Swedish households owning televisions have paid a monthly tax of SEK173 ($27) per month to support Sveriges Television, Sveriges Radio and educational broadcasting known as Utbildningsradion.

But Sweden's Radiotjänst collection agency now is collecting the fee even from Internet-connected computers. The logic is that, in some cases, PCs are used as “TV devices.”

German lower house of parliament separately approved a copyright bill that protects Internet search firms from payment of  fees to newspapers and other print publishers when snippets of stories are included in search engine results.

The bill's original draft would have allowed newspapers and other print publishers to stop search companies from showing text snippets, unless they paid licensing fees. The bill still has to win approval in the upper house, which is expected to oppose the current version of the legislation.

The other angle is that the bill does not fully settle the issue of whether search engine applications might have to pay publishers if news aggregators publish bigger amounts of content.

The move by Radiotjänst effectively makes a key form of broadcasting regulation applicable to PCs, notebooks and tablets, in a real sense, even when owners of tablets or PCs  do not watch TV. The tax is applied to a households that own Internet connected PCs, but not TVs, whether or not people in the household actually watch television or not.

Smart phones have been exempted from the law, at least for the moment, on grounds that the primary function of a smart phone is communications, not “watching TV.” Obviously, that distinction will be virtually impossible to maintain over the long term. But there is an existing principle that the “TV tax” applies to a “household,” not devices.

Presumably, that means Swedish households without TVs, but using Internet-connected PCs or tablets, will pay the fee only once, and will not have to pay for smart phone use, in addition to tablet or PC access.

Households that do not own PCs or tablets (possibly only a small fraction of all households), and do use smart phones, might ultimately be forced to pay the fee as well. The point is not whether it is “fair” or “right” for Sweden, the United Kingdom or Denmark to tax owners of TVs.

The point is that rapid changes in user behavior and device capabilities are changing the actual environment within which regulatory policy is conducted. Any nation that has distinct regulatory regimes for broadcasting, communications, Internet and print media will increasingly have to confront the growing contradictions and irrationality of older forms of regulation.

Order of Magnitude More Bandwidth Needed as Soon as 7 Years From Today

A scramble by mobile service providers to acquire more spectrum is rational, even if some think attempts to maintain artificial scarcity is the real reason for talk of spectrum scarcity. But some think mobile bandwidth demand is going to explode.

Where today 3 Mbps to 16 Mbps would be a normal range of speed for a U.S. consumer user, by 2020, it is quite possible that half of consumers will be buying and using services offering 100 Mbps, as crazy as that might seem. Video is the reason.

Ooyala says the hours spent watching streaming video on tablets and mobile increased 100 percent in 2012.

That mobile viewing of video grew that much should not come as a surprise. For example, a new study sponsored by inMobi found that 50 percent of the average global mobile web users now use mobile as either their primary or exclusive means of using the Internet.

If people are using mobiles as a primary or exclusive way of using the Internet, and if video is one of the most common Internet media types, it is not too surprising that mobile video viewership would be growing, especially as fourth generation networks and bigger screens now make mobile video a more enjoyable experience.

In 2008, Cisco forecast that video would represent half of global bandwidth will be used for consumer video apps of one sort or another by 2012. Those forecasts were not far from the mark.

Video streaming traffic represented 42 percent share of all global bandwidth in the second half of  2011, up from 35 percent in the first half of 2011, Allot Communications.

Overall, global mobile broadband traffic grew by 83 percent in the second half of the year, with a compound annual growth rate (CAGR) of 234 percent during the year.

This bandwidth is being dominated by a small percentage of users, though: Arieso has estimated that one percent of mobile subscribers now consume half of all downloaded data, with a third of those subscribers using a smart phone.

In the second half of 2011, YouTube accounted for 57 percent of all global video streaming traffic, meaning that YouTube alone held 24 percent share of global bandwidth. Overall, YouTube traffic grew by 143%, while video streaming traffic rose 88 percent.

Ooyala also says live video is starting to drive viewing not pre-recorded video. Ooyala's data also shows viewers watch live video longer on all devices, suggesting live video is more engaging.

On desktops, viewers watched live video 18 times longer than video on demand content in the fourth quarter of 2012, for example.

About a third of the total time spent watching tablet video in the fourth quarter of 2012 was engagement with  premium, long-form content running more than 60 minutes, Ooyala says.

The percentage of time spent watching  long-form video (over 10 minutes) on tablets

increased 37 percent from the first quarter to the fourth quarter of 2012.

The share of tablet video viewing more than doubled in 2012 as well.

Consumers have demonstrated a clear preference for engaging with content on smart phones using apps, which account for about 80 percent of U.S. mobile time spent interacting with mobile apps, rather than the mobile web, according to comScore.

Mobile channels now account for about 37 percent of all digital media consumption “minutes of use.” PCs might still account for about 63 percent of media consumption, but the mobile share is growing.

Faster mobile access speeds seem to be having a clear impact as well. Typically, users consume more total data on faster connections, compared to slower networks. By the end of 2012, 97.7 percent of U.S. smart phone owners used 3G or 4G enabled devices.

While 3G users still represent the wide majority of the  smart phone market, the number of 4G users grew in 2012 to  33.1 million, up 273 percent from 2011 levels of adoption. 

But 42 percent of all smart phone content consumption happens on a Wi-Fi connection, not the mobile network, comScore also reports.

But the main point is that all current notions of how much bandwidth, and how big a data cap needs to be provided, will be broken, and soon, by video consumption habits and preferences.

So-Net Launches 2-Gbps Broadband Access Service

Tokyo and six prefectures in the Kanto area (Kanagawa, Chiba, Saitama, Gunma, Tochigi, Ibaraki) are getting 2 Gbps high speed Internet access, with 1 Gbps upstream. "Nuro" was launched April 15, 2013 by the Sony-backed ISP. 

Provo, Utah is Google Fiber City Number 3

Google Fiber is coming to Provo, Utah, where Google Fiber will get a headstart by buying  iProvo, an existing fiberoptic network owned by the city.

The deal has to be approved by the City Council on April 23, 2013. It would be accurate to describe the network as "troubled," at least historically.  The network was sold to Broadweave Networks.

Broadweave then merged with Veracity Networks, which defaulted on its agreement with the city. In 2012 Provo took control again, and has been looking for a buyer ever since. 

Competing Internet service providers will feel more heat now that Google Fiber operates gigabit networks in three cities. That might not mean that major ISPs start upgrading to 1 Gbps everywhere. 

As typically is the case, major ISPs will upgrade where they have to, because of competitive conditions. So unless Google wants to build everywhere, and most seriously doubt that, gigabit networks will take some time to arrive. 

But gigabit networks seem likely to become the new benchmark. That alone will put increasing pressure on ISPs to upgrade.

Technology Futures, a firm with an extraordinary record of broadband predictions, now argues it is reasonable to expect that half of U.S. broadband access users will be buying 100 Mbps connections by about 2020.

Technology Futures also predicts that about 10 percent of customers will be buying 50 Mbps connections, while 24 percent will still be buying 24 Mbps service.

That might seem a crazy amount of bandwidth for “many typical users,” but standard technology forecasting techniques have, for more than a decade, actually suggested that would happen.

In 2001, for example, Technology Futures predicted that by year-end 2004, over 25 percent of U.S. households will have adopted broadband services, up from about five percent at the end of 2000. The actual U.S. broadband penetration rate was 30 percent, according to the Pew Internet and American Life Project.

“By 2010, we expect that the percentage will exceed 60 percent,” Technology Futures predicted in 2001. The actual penetration wound up being 66 percent.

Google Fiber's gigabit networks might actually push 100 Mbps access, in the near term.