Some Pro-Competitive Policies Just Don't Work

Political rationality and economic rationality sometimes are in conflict. In other words, public policy sometimes (perhaps often) is applied in ways that actually are counter productive, whether that is communications policy or social policy.

One example: nearly two years after the official end of the "Great Recession," the U.S labor market remains historically weak. Counter intuitively, dramatic expansions of unemployment insurance might be prolonging the problem.

In other words, not only do our efforts to ameliorate distress do very little, those efforts actually cause the problem to become worse.

To be specific, a study by the National Bureau of Economic Research suggests that

unemployment insurance extensions had significant but small negative effects on the probability that the eligible unemployed would exit unemployment, concentrated among the long-term unemployed.

In other words, UI benefit extensions raised the unemployment rate in early 2011 by about 0.1 to 0.5 percentage points.

National policies to promote competition in telecommunications markets suffer from similar dangers. What “seems reasonable” to promote consumer welfare might in fact lead to the opposite effect, namely a reduction in long term consumer welfare.

The practical example is policy affecting the number of providers in any market segment. And just how few service providers are necessary to provide meaningful competition in any segment of the telecommunications business is a thorny question.

Many observers would say the empirical evidence is fairly clear when the number of suppliers is “one or two,” based on the history of monopoly fixed network communications, or sluggish adoption, high prices and limited innovation when just two mobile service providers operated in any single market.

But there is more controversy about the minimum number of contestants required in the satellite TV segment, fixed network business and mobile business, under present circumstances.

Intermodal competition (competition from other suppliers outside the segment) is the difference. A few decades ago, one might have argued, as did U.S. antitrust regulators, that the satellite TV market would be insufficiently competitive if the two suppliers merged.

These days, satellite TV competes directly, and successfully, with cable TV and telco TV suppliers, at least for the video product. But satellite providers are at a clear disadvantage in the areas of broadband Internet access, voice and interactive services generally.

So “two becoming one” in the satellite segment might not be as challenging as in the past, in terms of impact on consumer welfare.

The other key challenge is the minimum number of service providers necessary to maintain effective or reasonable levels of competition in the core fixed network and mobile service segments.

In the fixed network access market, that minimum number today is “two.” Google Fiber will provide a key test of whether the long-term number of sustainable providers can become “three.”

In the U.S. mobile communications business, the developing issue is whether three major providers will provide sufficient sustainable competition. To be sure, there will be a common sense belief that four providers provides more competition than three providers.

That, in fact, is a common belief for regulators in some European markets.

To the extent that is true, the issue is whether competition is sustainable over the long term. Highly fragmented markets can be relatively stable over long periods of time, so long as capital intensity is low. Most packaged consumer products categories provide examples.

But access networks are capital intensive, limiting the number of viable providers over the long term, even if, in the near term, competition can temporarily support more competitors. And that’s the conundrum.

It is difficult to say what the minimum number of providers must be to provide the benefits of competition, beyond the number “one.” One is tempted to argue that “more providers” provides greater benefits than “fewer” providers.

That might even be the case, in the short term. Over the long term, sustainable competition might feature fewer competitors. The reason is simple enough: capital intensive businesses require enough profit margin to allow robust investment in the business.

Having “too many” providers in a market tends to reduce profit margins so much that no providers, at least theoretically, can earn enough to sustain themselves over the long term.

So although “more” sounds like a better recipe for competition than “fewer,” fewer might be the way to sustainable long term competitive benefits.

Sure, it sounds crazy that “fewer” competitors might produce better consumer outcomes than “more” competitors. But the problem is that a highly capital intensive business requires methods to earn enough money to build the next generation of networks. And “excessive” levels of competition might be quite detrimental in that regard.

In the end, perhaps political rationality wins, at the expense of economic rationality. But there is no reason to pretend that some policies designed to promote competition and consumer welfare actually will do so.

Some policies designed to ensure competition might actually do so at the expense of the ability to invest in the next generation of networks. In that sense, some touted pro-competitive policies might lead, in the long term, to sub-optimal consumer welfare.

Verizon Wireless, T-Mobile US Want to Swap Spectrum

Verizon Wireless and T-Mobile US have asked the U.S. Federal Communications Commission to exchange blocks of spectrum, generally on a one-for-one basis, in hundreds of U.S. counties.

Such spectrum swaps are not unusual in the mobile business. In 2012, five mobile service providers agreed to trade blocks of spectrum, acquiring spectrum from Cox Communications.

The Verizon and T-Mobile US exchanges would both firms to operate more efficiently, since after the exchanges each firm would have larger blocks of contiguous spectrum. In some cases, the additional spectrum is contiguous to spectrum each carrier already is operating.

In either case, each carrier would benefit from using larger blocks of spectrum, and in some cases also benefit from contiguous spectrum.

The moves are mostly tactical, allowing each service provider to operate more efficiently, since the deals do not change the aggregate amount of spectrum holdings of either carrier.

The FCC’s initial review of the applications indicates that, after the transaction, Verizon Wireless would hold 67 MHz to 149 MHz of spectrum and T-Mobile would hold 30 MHz to 100 MHz of spectrum in the 518 counties covering parts or all of 133 different cellular markets.

Since the swaps generally are one for one, those holdings reflect the initial amount of spectrum licenses held by each mobile service provider.

The exchanges will not affect any current subscribers of either network, and involve blocks of spectrum not yet activated by either mobile operator.

In the case of the intra-market exchanges of equal amounts of PCS spectrum, Verizon Wireless
and T-Mobile would exchange 5 MHz to 20 MHz  of PCS spectrum in 153 counties across 47 market areas, the FCC notes.

In addition, in 11 counties across three markets in Texas, Verizon Wireless would assign 20 megahertz of PCS spectrum to T-Mobile, and would receive 10 megahertz of PCS spectrum in return.  

Also, Verizon Wireless would assign 5 to 10 megahertz of PCS spectrum to T-Mobile in an additional 34 counties across 13 market areas.

In the case of the intra-market exchanges of equal amounts of AWS-1 spectrum, Verizon
Wireless and T-Mobile would exchange 10 to 20 megahertz of AWS-1 spectrum in 285 counties across 59 CMAs.  

In addition, in the Vineland-Millville-Bridgeton, NJ market, T-Mobile would assign 10 MHz to Verizon, and would receive 20 MHz of AWS-1 spectrum.

In the Oxnard-Simi Valley-Ventura, Calif. market, as well as the Eugene-Springfield, Ore. market, T-Mobile would assign 40 MHz and  would receive 30 megahertz of AWS-1 spectrum.  

Further, Verizon Wireless would assign 10 MHz of AWS-1 spectrum to T-Mobile US in 16 counties across four markets.

T-Mobile US would assign 10 MHz to 20 MHz of AWS-1 spectrum to Verizon Wireless in 26 counties across nine markets.

The swaps reflect a rationalization of spectrum each carrier had acquired in various auctions, but do not, in and of themselves, change market dynamics in the local markets or nationally. The swaps instead allow each mobile service provider to operate more efficiently, wringing more bandwidth out of the same amount of licensed spectrum, compared to the original set of holdings.

Google Launches Connected Car Initiative

Some might argue the automobile is the most important piece of new “hardware” that will become an important platform for software and apps.

Google has teamed with Audi, GM, Honda, Hyundai and Nvidia to form the Open Automotive Alliance (OAA), a global alliance to grow the connected car business, integrating all Android-based appliances.

The connected car is projected to generate US$282 billion in 2022, created by deployment of 1.5 billion machine-to-machine connections in the sector.

“Factory-fit vehicle platforms” such as GM’s OnStar or BMW “Connected Drive” will represent about 36 percent of connections in 2022, and aftermarket application-specific devices will account for the remainder.

Some US$32 billion will be generated by devices, US$20 billion by connectivity services and US$231 billion for applications and services that make use of the M2M connectivity. In other words, as has been true of the mobile and Internet business overall, the bulk of new revenue is generated by apps and services, not devices or access.

Learn more about the OAA at openautoalliance.net.

Will End of Smartphone Subsidies Actually Help Mobile Service Providers and Ecosystem?

It might seem self evident that smartphone subsidies are a burden for mobile service providers. If that is the case, getting rid of device subsidies should be financially helpful. 

And there is some evidence that operating income does improve when subsidies are ended.

After all, if a carrier buys devices from Apple at $660, on average, then requires consumers to pay $200 for the device, while recovering the balance of the device cost over the life of a two-year contract, then the carrier has to amortize the device cost over time, which has the effect of lowering operating income (some portion of revenue simply reflects a recovery of the upfront $460 difference between what the carrier paid for the device and the price the customer paid.

That is one reason why T-Mobile US has abandoned smartphone subsidies, and why other carriers are adding plans that achieve similar objectives.

Of course, the matter is more complicated. It is not clear how consumer behavior changes, if and when all consumers are required to pay full retail price for their devices, even when the advantage is lower monthly recurring fees for service.

Certainly many customers would find they do not want to spend $600 to $800 for a smartphone, or might not upgrade as often. Lower smart phone sales means lower data plan revenue.

The issue is how big the effect might be, though, now that a majority of users already use smartphones. Still, it is easy enough to predict that more users would shift to less-expensive devices and upgrade less often.

That would affect the fortunes of device suppliers and likely encourage suppliers to produce more models that are less costly.

That could lead to lower rates of software innovation and application development as well. Other ripple effects could include lower service provider equity prices (lower rates of revenue growth would lead to less robust retail share prices) and possibly less investment in the industry (rates of return would drop).

Perhaps it is indisputable that smartphone adoption and innovation have benefited from subsidies. The issue is how policies might change once adoption is nearly universal. 

Small Merchant Adoption of Mobile Credit Card Readers at 40%

With the possible exception of the Starbucks mobile payment system in the end user segment, retailer use of mobile credit card readers connected to smartphones and tablets are the standout winners in the emerging mobile payments business.

In fact, some 40 percent of small and mid-sized businesses surveyed by BIA/Kelsey said they accept payments at the point of sale with a mobile credit card reader attached to a smartphone or tablet.

About 16 percent of surveyed retailers said they were planning to add this capability in the next 12 months, according to BIA/Kelsey. 

In fact, such payments already eclipse the amount of contactless payment activity by quite some measure. 

Although mobile POS proximity payments made up just 0.01 percent of total retail POS volume in 2012, mobile devices (smartphones and tablets) will help propel mobile payments to $5.4 billion by 2018.

Mobile Now More than 65% of All U.S. Internet Access Connections

Of 262 million U.S. broadband access connections, there were almost 65 million fixed and 64 million mobile connections with download speeds at or above 3 megabits per second (Mbps) and upload speeds at or above 768 kbps as compared to 51 million fixed and 31 million mobile connections a year earlier, according to Federal Communications Commission data.

In other words, fixed and mobile networks supply an equal number of Internet access connections 3 Mbps and faster. To be sure, mobile and fixed access services are not equivalent in cost per megabyte or size of usage allowances.

But mobile has become a significant supplier of “faster” connections. For example, of connections offering 6 Mbps or faster service, fixed networks supply about 41 million connections, while mobile networks supply about 32 million connections.

For a historically bandwidth-limited sort of network, that improvement on the mobile front is significant.

To be sure, mobile Internet access speeds are underrepresented at 6 Mbps and faster, and over-represented among connections of 3 Mbps and lower speeds. But mobile Internet connections already represented 65 percent of all Internet access connections in the United States, at the end of 2012.

In December 2012, 21 percent of reported fixed connections (19.3 million connections) were

slower than 3 Mbps in the downstream direction, 16 percent (15.2 million connections) were at least 3 Mbps in the downstream direction but slower than 6 Mbps, and 63 percent (58 million connections) were at least 6 Mbps in the downstream direction, the FCC reports.

It might not be clear from the FCC statistics, but progress, measured in terms of typical Internet access speeds, has grown surprisingly fast in the U.S. market. That might come as a shock to some.  

In fact, Internet service provider speeds have grown at about the rate you would expect for a Moore’s Law driven product. That should be a surprise, since access networks are notoriously expensive and take some time to build. That noted, from 2000 to 2012, the typical U.S. access connection speed grew by about two to three orders of magnitude.

Retail prices also now provide dramatically more bandwidth per dollar.  In fact, people now pay less for a 40 Mbps access service than they used to pay for a 512 kbps access service.

Though the FCC report does not highlight the rapid changes in access speeds, progress has been rapid.

In August 2000, only 4.4 percent of U.S. households had a home broadband connection, while  41.5 percent of households had dial-up access.

A decade later, dial-up subscribers declined to 2.8 percent of households in 2010, and 68.2 percent of households subscribed to broadband service.

Though it perhaps is understandable that people expect more, and now, a bit of perspective probably is in order.

Internet access connections that essentially double speed every three to five years, while also featuring lower prices per unit of speed, are impressive. At those rates of change, gigabit connections will be common by about 2020.

How Big a Business Can "Exposing Network Services" to Business Partners Grow to Be?

Most major mobile service provider executives probably would agree that providing mobile network services to third parties is an important revenue growth opportunity.

One example is provided by AT&T and Verizon efforts to sell mobile network services to banks who host banking apps.

AT&T makes available a number of hosted services ranging from call recording to videoconferencing to geo-location services to mobile identity verification.

The telecom launched its first product in this vein, a mobile identity toolkit, in late December. In a way, it is following the footsteps of Verizon.

It's hard to say for certain how big such revenue streams might become, other than to note that immediate prospects should be rather modest. Even the established content delivery network business, a similar line of business generating revenue from business partners, earns perhaps several billion a year, globally.

Most efforts to sell third parties mobile network features will proceed industry by industry, necessarily implying a rather muted growth rate.

WhatsApp Takes OTT Messaging Lead

According to Portio research, text messaging traffic globally is about to reach a peak. After 2015,  text messaging volume will start a downward trend. 

With the caveat that usage or volume is not revenue, the forecast suggests text messaging will, in some markets, start to decline by then, as it already has begun to dip in Western Europe.

At least in part, the product maturation is caused by use of substitute products such as instant messaging such as WhatsApp.

Generally speaking, text messaging revenue growth will be strongest in many developing markets, where mobile Internet access is not yet something most mobile users buy and use.


source: Portio Research

"Micro-Basic" Subscription Video Tiers in 2014?

Though program networks generally dislike the concept, video subscription providers are well aware many customers think their monthly costs are too high.

That is leading a steady but slow stream of customers to abandon video altogether. Then there are the younger consumers who simply never have acquired the habit of purchasing linear video. 

So some might predict that worried programmers will consent to "micro tiers" of programming that might entice some customers to stay, and others, who never have bought the product, to try it.

These possible smaller packages of channels will cost less, and possibly appeal to many consumers who otherwise would not buy anything, or would be tempted to disconnect.

We'll see. As much effort as will go into constructing tiers that add incremental customers without triggering significant downgrades on the part of existing customers will be tricky. 

One More Example of How Internet Apps Can Grow ISP Revenue

With the caveat that a mobile or fixed network Internet access provider has to be able to charge for greater consumption by end users, Internet apps possibly have harmed telecom revenue less than imagined.

In fact, even cannibalization of text messaging by over the top services might not be as big a hit to revenue (globally), as often is assumed.

A shift to greater amount of video streaming actually should help Internet service provider revenues, assuming some rather direct correlation between consumption and revenue is possible.

For the sake of argument, assume a household shifts consumption from a linear video TV subscription to some form of online delivery.

Ignore for the moment any revenue the Internet service provider can earn from providing TV content that once was consumed using a cable TV, satellite TV or telco TV connection, and focus only on the impact on purchased Internet access service.

Assume a one-hour TV show streamed to a TV requires 1GB at standard definition, and 2 GB, for an hour of HDTV. Assume you are a typical users and consume five hours of video a day. Assume half your consumption if HDTV and half is standard definition.

That implies 75 hours of standard definition TV consumed per person per month. At 1 GB per hour, that’s 75 GB of data. The 75 hours of HDTV represent 150 GB of data consumption, for a total of about 225 GB of data consumption a month, for linear entertainment television.

That might not be an issue for a single-person household with a monthly usage allowance of 300 GB. Assume the monthly cost is about $70 for such a plan.

The math gets tricker for multiple-person households, especially if many users are watching different programs. But the revenue logic is simple enough. If two residents watch five hours a day of TV, and that consumption is shifted (for the sake of argument) to online delivery, then that household has to buy an access plan with a bigger data allowance, to account for the 450 GB of video consumption by the household.

As this one example illustrates, demand for Internet apps drives enough incremental usage that one might argue the Internet sometimes directly drives ISP revenue.

It’s complicated, but not quite a zero-sum game, as some might fear.

How Much Text Messaging Cannibalization, Really?

Following up on the issue of how much the Internet has harmed telecom revenues, the amount of revenue destruction is rather complicated. While most wouild agree that some amount of telecom service provider revenue from voice and text messaging has been disrupted by over the top alternatives, the net revenue impact has to be balanced by creation of new revenue streams directly created by end user use of the Internet.

For example, Portio Research questions the actual amount of lost revenue. While in some markets over the top messaging presumably does cannibalize significant text messaging revenue, in other markets, where mobile Internet access is not widespread, dramatic growth of mobile adoption means more text messaging revenue is being created.

“During 2012 and 2013 we have seen many reports that operators are losing $20 billion to $30 billion in SMS revenue to OTT messaging apps,” said Karl Whitfield, a director at Portio Research. “We see reports that OTT traffic will be double that of SMS by the end of 2013,” he says. “This is wrong on both counts.”

It may be true that SMS revenues are levelling off and that OTT is on the rise, but SMS is still generating revenues of $15.3 million per hour, 24/7, that’s a massive $133.8 billion in 2013, Whitfield says.

Over the top apps generate about $3 million an hour, by way of comparison.

Worldwide SMS revenue has gone up year after year since the early 1990s and will continue to be above 2010 levels until 2017, Whitfield said.

In fact, in some markets, SMS and OTT apps are coexisting, serving end users in different ways.

There is a huge uptake of OTT messaging in Japan, particularly with local player LINE, yet the SMS market remains healthy and stable, he says.

The same goes for South Korea, where KakaoTalk is enjoying huge success; here again the SMS market remains stable and is not declining as many predicted.

Where SMS has seen a decline, in markets such as Spain and Greece, there has been an overall fall in subscribers and revenues at the same time.

“Our research into mobile messaging completely contradicts what some other industry observers are saying,” said Whitfield.

Global OTT and P2P Messaging Traffic (Billions)

	2010	2011	2012	2013F	2014F	2015F	2016F	2017F
P2P SMS	5,812	6,546	6,623	6,687	6,654	6,522	6,304	5,931
OTT Messaging	1,494	3,840	6,774	10,452	14,970	20,437	26,359	32,141

source: Portio Research

The point is that product substitution, while a fact, might not be as destabilizing a revenue trend as sometimes  believed.

FAA Authorizes Commercial-Drone Testing

Most rather exotic technologies take a while to reach commercial maturation. Artificial intelligence and robotics might provide another example. Though robotics have been a mainstay of manufacturing for 40 years, the extension of more advanced forms of applied artificial intelligence only now seem to reaching a new stage of commercialization.

Google's self-driving cars are one example. A more important development is that the 
U.S, Federal Aviation Administration has selected a handful of universities and state agencies to operate sites for drone testing, in a step toward eventually integrating commercial unmanned aircraft into the U.S. aviation system.

That might begin to happen as early as 2015, some think. That's a huge step forward.

The six operators of test sites for unmanned aircraft include the University of Alaska, the State of Nevada, New York's Griffiss International Airport, the North Dakota Department of Commerce, Texas A&M University - Corpus Christi and Virginia Polytechnic Institute and State University.

None of those tests seem yet to involve commercial applications, instead focusing on how such systems might work, safely. The next big steps will involve apps using such drones.

Economics Does Not Explain Everything Because "Irrational" Behavior Matters

Economics is a discipline rather rare in the public policy arena, including the communications business, at times.

Though political rationality also is at work when policies are created, political rationality ("what can be done; what is possible") is not always quite so rational in terms of how people, firms and markets will change, once any set of policies are implemented.

In fact, it is impossible to know, with certainty, how behavior will change, in unexpected way, once a set of changes is made. That also applies for incentives people have when creating products and services. 

"People can be really smart or have skills that are directly applicable, but if they don’t really believe in it, then they are not going to really work hard," says Mark Zuckerberg, Facebook CEO.

That intangible--commitment--cannot be modeled mathematically. Sometimes forces that cannot be measured or modeled--love, affection, idealism--indeed can shape behavior. At least sometimes, that means impossible things can, at least for a time, become "possible."

So "irrational" behavior sometimes can confound forecasters and predictions. And by "irrational," one does not have to imply "not rational and therefore destructive." Love is irrational, too. Passion and idealism likewise can sometimes cause behaviors we cannot quite anticipate.

If some firms seem to continually outperform others, at least some of the time that is because those organizations have harnessed "irrational" commitments. No bureaucratic structure can fully compensate for that.