What is the Economic Contribution of "Free" Services and Apps?

“Free” is an interesting conundrum for economists. it is hard to value something valuable, when it is given away. Sure, it is possible to measure the derivatives, such as advertising revenue generated by providers of free apps and services.

Traditional economic approaches based on measuring prices and quantities do not work well for goods with a price of “zero.” Though economics often is counter-intuitive in its implications, most people could agree that it seems disjointed in some way to maintain that any economy is worse off because people use Skype, Google Hangouts, free web browsers or most Internet apps.

People might agree that some of that usage cannibalizes some product provider’s revenue. But most also would agree that use of those replacement products also leads people to use “paid” products more than they otherwise would have.

And the common sense notion likely also is that such products make people and companies “more productive” than they had been before.

But the value of “no incremental cost” services and apps is hard to measure, since the non-monetized value cannot be captured in conventional ways.

Ever since Netscape “gave away” its browser, the value of "free" has been a source of debate, despite underpinning what is now called the “freemium” business model.

The Massachusetts Institute of Technology economist Erik Brynjolfsson points out that, according to government statistics, the “information sector” of the U.S. economy, which includes publishing, software, data services and telecom, has barely grown since the late 1980s.

That will strike many as a puzzling finding.

Using a new methodology, economists at MIT estimate the increase in consumer surplus created by free Internet services to be over $30 billion per year in the United States, about 0.23 percent of average annual gross domestic product between 2002 and 2011.

Perhaps that is a conservative estimate, but being conservative in applying such new measures is wise, since the methodology involves assigning a “hard dollar” value to a “soft dollar” metric of “time spent online” (on the assumption that time is money). Being conservative is wise since most of us can think of lots of ways that time online is not really productive.

The same sort of caveat might apply to other forms of “attention,” such as watching television. Most people would probably agree that not much of that form of attention adds measurably, if at all, to gross national product, aside from the value to networks and broadcasters of their created ad revenue streams.  

Will Most ISPs Eventually Offer Free 5 Mbps Service?

What happens to the retail rates of lower-speed broadband access if the market standard winds up being a gigabit for $70 a month, or a gigabit for $80 a month, or perhaps less? At $80 per month, a gigabit service implies a per-gigabit price of eight cents per megabit.

So the retail price of a 5 Mbps service “should be” 40 cents per month, or, in other words, nearly at the level of Google Fiber’s “free for seven years” 5 Mbps access service.

The caveat is that lower speed services always cost more, per megabit of service speed, than higher speed services, on a cost per megabit basis.

Still, even with that caveat, experienced practitioners from the network access business quickly will say that a fixed network cannot be built, under conditions where Internet access is the financial underpinning, when per-user revenue is just 40 cents a month, even when other current and projected revenue sources are available.

So the hope, or expectation, would have to be that almost nobody wants Internet access service at 5 Mbps, making the demand for such service virtually nil.

If that is the case, and most people want to buy services closer to the $50 to $70 a month level, revenue per user is not the problem, especially when other services can be sold in significant quantities.

The business model problem then shifts to costs, both capital investment and operating cost.

In fact, based on current levels of revenue “per user” generated by the largest Internet app providers, it seems fairly obvious that the whole idea behind “free access for seven years” is that, within 10 years, nobody will want such a service.

Some calculate that Facebook presently earns about $1.63 per user, per year, while Google earns $10.09 per user, per year.

At such levels, no rational ISP would expect to cover the cost of building gigabit networks based solely on advertising and other revenue generated by advertising and commerce transactions.

Still, the essential point is that if gigabit access, for $70 or $80 a month becomes anything close to a market-dictated reality, the cost of slow-speed Internet access (5 Mbps, at the moment, higher numbers in the future) might in fact be something very close to “free,” with ISPs expecting to earn higher revenues from sales of additional services of various types.

It might be reasonable to expect an ISP to charge for installation or use of required access equipment after the first outlet, as Google Fiber does. But the possibility remains that in a world where gigabit service is fairly standard, it might make sense to offer low-speed access essentially for free.

For starters, almost nobody would want such service. For the few that do, there likely will be other incremental revenue streams an ISP might be able to create, once that Internet connection is in place.

Video Mergers Might Challenge FCC Horizontal Concentration Policy

Regulators in the U.S. market might soon find they need to take one more look how to determine when the U.S. communication markets are too concentrated, or might become too concentrated.

The reason is the growing possibility that some possible merger or acquisition efforts in the U.S. market will test formal and informal rules about market concentration in the cable TV, satellite TV, voice and broadband services markets, both directly and indirectly.

Though the Federal Communications Commission and U.S. Department of Justice determined that AT&T would become too big if it acquired T-Mobile USA, negatively affecting competition in the U.S. mobile market, new tests likely are coming in the fixed network and satellite businesses.

The latest rumors either involve Charter Communications making a bid for Time Warner Cable, or perhaps even Comcast trying to buy Time Warner Cable. Then there is new talk about a potential merger between DirecTV and Dish Network, the only two U.S. video entertainment providers.

DirecTV and Dish Network tried to merge in 1992, but that merger was blocked by antitrust authorities, because of reduced competition concerns. Whether the markets now are sufficiently changed, with telcos now providing video entertainment, is the question.

The possibility that Comcast, the largest U.S. cable TV provider, might add Time Warner Cable’s assets might also challenge FCC rules about maximum market share in the cable TV business, which had been set at a maximum of 30 percent share of customers in “cable TV” business.

But that concentration rule was overturned by an appeals court in 2009. Over time the original rule that no single cable operator could pass more than 30 percent of homes was modified to mean that no single cable operator could have more than 30 percent of actual subscribers in the broader market, including satellite and telco providers.

Some would note that Comcast at present already has perhaps 43 percent share of video customers in the cable TV business, but only about 23 percent share of U.S. video entertainment subscribers, when the whole market, including satellite and telco providers, is examined.

A combined Comcast with Time Warner Cable would serve about 35 percent of U.S. video market customers.

Were DirecTV and Dish Network to merge, the new company would have 36 percent share of the traditional U.S. video entertainment business. That would also pose a challenge to the current rule that no single provider can serve more than 30 percent of total subscribers in the broader market.

At least so far, over the top services such as Netflix have not been considered part of the multichannel video subscription business, though that could change at some point in the future, especially if Netflix were to secure broad rights to distribute full versions of channels, if providers such as Comcast or Viacom were to acquire rights to sell, over the top, the same content they now sell as packaged services on their owned networks, or if the entire business shifts to new forms of delivery broadly disconnected from traditional delivery by access providers.

Africa Broadband Adoption Low, but Poised for Big Change

While just 16 percent of the continent’s one billion people are online, that picture is changing rapidly.

While Internet penetration is just over 16 percent across the continent as a whole, it is much higher in urban areas, where more than 50 percent of residents use the Internet regularly.

A recent McKinsey report found that 25 percent of urban Africans go online daily, led by Kenyans at 47 percent and Senegalese at 34 percent, according to McKinsey.

Still, levels of broadband service remain relatively low. Overall broadband penetration, on a per-capita basis, is about 2.5 percent in Algeria, and 0.1 percent in Angola, for example. Per-capita measures understate fixed network broadband penetration, however, to the extent that a connection is to households with multiple residents.

If typical household membership is five, all the per-capita measures need to be multiplied by five, to derive penetration of households, which is the more meaningful metric, some would say.

Country	Per-Capita Broadband Penetration	Per-Capita Mobile Penetration
Algeria	2.5	103
Angola	0.1	49
Cameroon	0	64
Cote d’Ivoire	0	96
Egypt	1.8	115
Ethiopia	0.8	24
Ghana	0.2	100
Kenya	0	72
Morocco	1.6	120
Mozambique	0.1	33
Nigeria	0.1	68
Senegal	0.6	88
South Africa	1.5	135
Tanzania	0	57

Will FCC Formally Modify its Historic Cable TV Industry Market Share Rules?

Something potentially more interesting than smaller Charter Communications buying Time Warner Cable are afoot. 

The wild card at the moment is that Comcast might consider buying Time Warner Cable, a move that Comcast, the biggest U.S. cable company, would have avoided a decade ago, to avoid running afoul of Federal Communications Commission and Department of Justice anti-trust scrutiny and oppostion.

Though some would note that the rules were informally modified in a key way some time ago, the U.S. cable industry basically continues to operate under FCC rules that no single cable company can serve more than about 30 percent of U.S. subscribers. Some confusion exists, though.

In 1999, the FCC, in a unanimous vote, redefined the concentration rules and relevant markets to allow AT&T to buy Mediacom, theoretically allowing AT&T to serve potentially 40 percent of U.S. homes. But that is a slightly different issue. 

AT&T represented new telco industry competition, and was not a legacy cable operator.

The FCC cable market share rule remains in effect, formally, but the thinking was that AT&T would be able to offer competition to the incumbent cable TV companies.

Given the many changes in the video market over the last decade, might the FCC be willing to take a different look at concentration? 

The FCC once barred a merger by DirecTV and Dish Network, as that would result in one U.S. satellite provider. But in markets that structurally are different now, does it make sense to regulate industry by industry, or look at the whole market.

Even the definition of the "market" might be enlarged to include over the top providers such as Netflix, as well.

Although the old 30 percent cap was technically not changed, the revisions to the definitions of markets and cable ownership effectively raised the ownership cap for cable television to the rough equivalent of 36 percent and even more for video service, albeit only when it was a telco that could gain that much share (AT&T has not done so yet).

In other words, regulators were willing to look at competition and concentration not simply in terms of "industry silos," but service by service. 

So the interesting new challenge is, what if Comcast does make a bid to buy Time Warner Cable, pushing Comcast into a range that exceeds the traditional FCC market share rules?

Will Comcast be allowed to do so, and if so, will it be because its video market share is shrinking, while its voice and high-speed Internet access share is growing? At the time, the FCC thinking about competition and market share was that it makes a difference whether the market share being considered is cable in video services, or telcos in voice, not just inter-industry or intra-industry concentration measures. 

Though it would be notable if Charter bought Time Warner Cable, the bigger policy ramifications would come if Comcast were to try and buy Time Warner Cable. 

APT 700 Creates World Band for LTE

The Asia-Pacific Telecommunity (APT) band plan for 700 MHz spectrum harmonized globally is viewed by many as important for Long Term Evolution, as it creates a worldwide band of spectrum capable of supporting LTE in most countries, and also therefore allows suppliers of network gear and handsets to build more affordably and efficiently.

It therefore is not surprising APT 700 got at least some attention at the recent International Telecommunications Union Telecom World.

“This is the first time that a technology in one band is available almost globally,” said

Margit Brandl, Nokia Siemens Networks Global Head of Telecoms Policy.

But APT 700 also reflects a couple of other important reasons, including the issues of repurposing, refarming and sharing spectrum in new ways. As the APT 700 plan is based on repurposing former TV broadcast spectrum, APT 700 is akin to the TV white spaces effort, which likewise finds new communications applications for unused TV broadcasting spectrum.

In some cases, perhaps, the APT 700 spectrum could allow new or weaker competitors to obtain valuable ways to extend coverage to rural areas more affordably. In many markets, the carriers that now are the incumbents mostly were able to acquire frequencies that propagate significantly better than the later 2 GHz blocks upon which many competitors have based their businesses.

In Europe, SFR has already started refarming of the 1800 MHz band, according to SFR CFO Thomas Welter. Telstra likewise is an advocate of spectrum refarming, said Mike Wright, Telstra executive director.

The APT 700 MHz band offers the opportunity for European bands to overlap with Asian and South American bands, tapping into a potential pool of in excess of two billion people.

LTE in Unlicensed Spectrum?

Qualcomm wants to enable use of Long Term Evolutiion in unlicensed radio spectrum used for Wi-Fi or Bluetooth. That this can be done, in terms of technology, is probably not the biggest issue. 

The biggest issue is that virtually all blocks of allocated spectrum specify what purposes, and what air interfaces, can be used in those bands. 

Such multiple air interfaces approaches will necessarily require regulatory action.

To be sure, many have argued that innovation is promoted when regulators do not issue licenses with specific air interfaces or specific applications intended for use of that spectrum. How much chance Qualcomm might have in getting such regulations changed is the issue.

"Europe Falling Behind" is Temporary, as was "U.S. is Falling Behind"

Far too much is made of “leadership” in communications. For what seems like a decade, observers “worried” that U.S. consumers lagged Europeans, the Japanese or the South Koreans in use of mobile phones or text messaging. One might simply note that the gaps closed, arguably once service providers got serious about making mobile services a mass  market product, and once consumers grew to understand the value.

Similarly, some observers decried, or sometimes continue to decry, lagging U.S. Internet access speeds, prices or both. Ignore for the moment conflicting evidence about whether that remains the case, or to what extent there are differences.

Still, European observers now worry that “Europe is falling behind” in next generation fixed broadband or mobile Long Term Evolution.

European service providers lead the deployment of a succession of technologies, from GSM to early deployments of LTE in the nordic countries, for example. And then there was the leadership of firms such as Nokia. Nokia’s fall is part of the present worry.

With the advent of the smart phone era, innovation and perceptions have changed. the success of Apple and Android's devices, operating systems and apps moved the center of innovation to the United States, while Asian companies are more relevant than ever, according to Analysys Mason.

As once too much attention was paid to European leadership and U.S. “gaps,” now too much is made of U.S. leadership and European lagging.

It might be reasonable to say that since U.S. firms tend to lead with software, a shift to a software-driven mobile Internet would potentially shift fortunes. One might also argue regulatory regimes in North American and Europe have had direct impact on both the level of competition and investment.

LTE connections as a percentage of mobile connections, selected regions, 2010–2017 [Analysys Mason, 2013]

One might also note that European operators adopted faster versions of 3G networks at level that matched customer demand for speed, resulting in less need to upgrade to 4G.

The point is that investment, consumer demand, software and hardware supply will shift, from time to time. Over time, adoption, supply and impact tends to equalize, from one developed region to the next.

Also, at some point, the issue is what value a given nation or region can wring out of its investments, not the mere fact that investment in faster networks has been made. That always will be tough to measure.

Though the hand wringing about Europe “falling behind” is real, it has to be kept in perspective. Many would have noted U.S. “lags” in broadband and mobile adoption as well. Sustainable “advantage” seems to be as elusive in communications technology as it is in the life of firms.

Google Builds First "Middle Mile" Network

Google has for some time operated one of the world’s biggest global long haul networks. Google also operates access networks in several U.S. cities. Now Google also provides a regional fiber backbone network that apparently can be used by access providers in Kampala, Uganda.

Google’s Project Link is a regional fiber network that apparently will sell capacity to any local mobile operator or Internet service provider operating in and around Kampala.

Project Link’s network is one more way Google has built its own infrastructure, including a global long haul network, local access in three U.S. cities and now a regional backbone network in Kampala. Of those networks, the Kampala network is the first of the “middle mile” type, and also the first intended to sell capacity to third party access providers.

Is Privacy an Anomaly on Social Networks?

Google's chief internet evangelist, Vint Cerf, suggests that privacy--at least in the context of social networks--might be an anomaly. 

That is not a blanket statement about government data collection or privacy in some other contexts, only a recognition that the whole point of social networks is sharing. 

As was true for most people in prior historical periods where most people lived in small villages, and where "everybody knew everything" about other people in the community, something of that same process affects online social networks as well. 

Will 90% of Mobile Traffic Be Terminated on Wi-Fi?

Public Wi-Fi hotspots will represent about 22 percent of tier-one mobile operator mobile Internet facilities in 2013, according to a survey of 200 executives in the Wi-Fi ecosystem. The survey respondents also believe that 78 percent of new small cell capacity will feature public Wi-Fi capabilities as well.

Public Wi-Fi capacity added by mobile service providers will grow at a 13 percent compound annual growth rate between 2012 and 2018, representing an additional  10.5 million sites in 2018.

The Asia-Pacific region will account for 55 percent of the deployments by 2018.

The total installed base of public Wi-Fi hotspots will reach over 55 million, globally, in 2018.

Residential access points where some capacity is left open for public access, is likely to top 100 million nodes by the end of 2018, the survey finds.

Outdoor hotspots were expected to drive most of the deployments, 46 percent of the respondents said.

As you might expect, the business model was a key issue for executives considering wider use of public Wi-Fi as part of the mobile network infrastructure, according to the Wireless Broadband Alliance.

If those forecasts are correct, one wonders how much traffic actually will be terminated using the mobile networks, since some mobile service providers already find that up to 80 percent of total traffic is terminated on a Wi-Fi network, mostly at-home or at-work fixed networks.

If just 10 percent of mobile network traffic is shifted to public Wi-Fi, it conceivable that 90 percent of total mobile handset traffic will be terminated on Wi-Fi networks, public and private.

The  survey of about 200 Wi-Fi ecosystem players was conducted by Maravedis-Rethink.

Czech 4G Auction Fails to Bring New Competition

Communications regulators and lawmakers frequently have incentives to “do something” about competition in communications markets. Whether or not such efforts succeed in the long term, there is value in being seen to “do something” in the near term.

In other words, what is eminently rational in political terms does not have to be equally rational in economic terms. It still is quite rational to “do something” widely viewed as promoting competition, even if there are reasonable expectations that the policies will not work.

Set asides for new competitors, spectrum caps or spectrum limits are some of the steps regulators can take to promote market entry by new competitors.

It isn’t so clear whether such tactics work, even in the short term.

The Czech Telecommunications Office, for example, had set aside rules for some blocks of Long Term Evolution spectrum in recently-held spectrum auctions. Only new entrants could bid for those blocks, initially.

Still, the auction, which raised CZK 8.5bn (EUR  311 million), was won by the three existing mobile phone companies (Spain's Telefonica, T-Mobile--the mobile arm of Germany's Deutsche Telekom AG and the U.K.'s Vodafone Group).

Telefonica CR and T-Mobile CR bought two 800 MHz blocks, while Vodafone bought one 800 MHz block.

The three mobile companies also bought spectrum in the 1,800 and 2,600 MHz bands.

Revolution Mobile and Sazka Telecommunications dropped out of the bidding, even though

a 2x10 MHz block of spectrum was made available exclusively for new operators.

Will Amazon Web Services Be Worth $76 Billion in 2015?

If this revenue forecast is close to accurate, then Amazon Web Services, at a multiple of about eight times revenue, would be worth about $72 billion. Seriously. 

Mobile-Accessed Sites More Important in Some Industries Than Others

Mobile sites arguably are more important in some verticals than others. Travel comes to mind. But restaurants and bars might be among the industries most amenable to value created by mobile apps and websites.

After analyzing  traffic statistics of 14,000 websites over four years, Let's Eat, a builder of mobile app sites for the restaurant industry, notes that about half of all visits to restaurant websites consist of mobile traffic.

During the week, about a third of visits to restaurant-related sites come from mobile devices. On weekends, the percentage shoots up to nearly half of all visits. You can probably guess why that pattern exists.

During the week, when many people are at desks, it arguably is easy enough to check for a place to go eat from a PC. On weekends, when people are more likely to be out and about, using a mobile will often make more sense.

Mobile traffic has increased at a fairly regular rate over the last four years, from just three percent of traffic in October 2009 to 43 percent of restaurant site traffic in October 2013.

You also might not be surprised that population density affects use of mobile devices to get restaurant information. In urban areas associated with higher population density, people are more likely to be out and about during hours when restaurants are open, and needed.

Also, in smaller communities, it is more likely that people generally know where to find a particular restaurant or type of food.