Does your iPhone service suck? Blame city hall

commentary As the Department of Justice and now Sprint file suit to block AT&T's pending merger with T-Mobile USA, federal judges will begin looking closely at the deal and its potential impacts on the exploding mobile services market.

Let's hope that the kind of evidence-based, rational analysis of the courtroom does a better job separating fact from fiction than the court of public opinion, which is easily seduced by catchphrases and unrealistic scenarios. Many of the most outspoken critics of the deal, it seems, either don't understand the fundamentals of cellular technology, or assume at the very least that their audience--journalists, regulators, and consumers--doesn't care.

Justification for the merger, for example, rests largely on engineering and legal limits that define how mobile providers can operate, particularly having to do with spectrum and the physical infrastructure (towers, antennas) of the cell sites serving thousands upon thousands of iPhones, Android devices, and other mobile phones. AT&T argues that combining the towers, equipment, and spectrum assets of the two companies will quickly lead to improvements in mobile service quality for both sets of customers.

But how, the critics mock, can the combined assets perform better together than they do individually? "You can call it fuzzy math," scoffed Free Press, a Washington, D.C.-based advocacy group that lobbies for nationalizing the communications infrastructure, "or you can call it lying."

Or you can call it sound engineering. Cellular networks are characterized by increasing returns to scale, meaning the combined network will indeed perform better for all customers than the two existing networks. That's what "cellular" means, after all. The better the site density, the better it works. And improvements are not linear--the sum really is greater than the parts.

Related stories
• AT&T and T-Mobile: Wireless megamerger (roundup)
• Sprint files suit to block AT&T's T-Mobile merger
• AT&T fighting to save T-Mobile deal

Today, the cell towers and antennas of the two companies serve only their own customers. By routing all the traffic through a combined set of equipment, however, everyone's packets will go through an expanded and denser network of cell sites. That means fewer dropped calls and better reception--for all users.

At a recent hearing on the merger in San Francisco, spokesmen for the two companies noted that the placement of their respective equipment is highly complementary. In high traffic area, including San Francisco and Washington, D.C., denser cells are essential, so even nearby sites added to the network will help.

The engineers also explained that a great deal of available bandwidth is eaten up by network management. Each company uses as much as 10MHz of spectrum just for "control channels" on their networks. With only one larger network to manage, one of the control channels will be eliminated and the spectrum freed up for customer use.

And as the merged company uses their combined spectrum holdings to deploy 4G LTE service, the efficiencies will actually increase. Even though LTE networks allow customers to use more high-bandwidth data applications, they are also much more efficient than 3G and older protocols still in use. Merger opponents did not dispute these facts.

Whose fault is poor service? Yours
But future uses aren't the most pressing problem. The need for redeployed spectrum and denser cells, as every consumer knows, is hardly theoretical. Since the introduction of the 3G iPhone in 2007, AT&T reports increases to its data traffic of more than 8000 percent, creating tremendous strain on its existing network. Today, service in some parts of the country is poor or at best unpredictable.

Let's put aside hypothetical impacts of the merger on competition for a moment and get real. If improving its existing network is really AT&T's objective, the critics argue, why doesn't it just spend a fraction of the T-Mobile purchase price on upgrading and expanding its infrastructure?

The real reason, as many of them know, is that doing so requires breaking long-standing regulatory logjams. Additional spectrum can only be made available by federal regulators, and Congress is in no mood to give the FCC the authority it has sought for years to conduct new auctions. (T-Mobile has little hope, with its existing spectrum allocation, of ever providing LTE service.)

But why, then, does AT&T need T-Mobile's towers and antennas? If increasing cell density and adding equipment to existing sites can have such a dramatic impact on service quality, why hasn't the company simply upgraded their infrastructure? No one could have predicted the dramatic explosion in traffic from the iPhone, Android, and other new mobile devices. But who's stopping AT&T and other carriers from fixing the problem now?

The answer, surprisingly enough, is you. Or, to put it more precisely, your local zoning authorities, who delay or outright refuse to allow new infrastructure investments to move forward in many parts of the U.S., including jurisdictions with the most serious performance problems today.

That, in any case, was the conclusion reached by the Federal Communications Commission in its recently-released annual report on competition in the mobile industry. The agency noted that despite the economic downturn, the largest mobile providers continue to invest billions of dollars annually in capital expenditures or "capex." In 2009, for example, the four largest carriers together spent more than $20 billion, or 13 percent of industry revenue. (Only Sprint, which not surprisingly has filed suit to block the merger, significantly reduced its capex.)

Most carriers would have spent even more, the FCC found, but their applications to build or modify cell sites were systematically delayed or denied by city and county governments, who must give final approval under local zoning laws. By the end of 2009, delays had reached epidemic proportions. The FCC reported that "of 3,300 pending zoning applications for wireless facilities, more than 760 (nearly one quarter) had been pending for more than a year and 180 had been pending for more than three years."

These included not only applications for new sites, but also requests to replace an antenna already on a tower or other location. Also held up were applications to add equipment to existing sites ("collocation") designed for use by multiple providers. Even applications to place antennas on building rooftops and other invisible structures (church steeples, trees) were being stalled.

In Northern California and Washington, D.C.--areas with notoriously unpredictable mobile service--the FCC found that applications regularly took 28 months to 36 months to process, even for collocations. In Berkeley, Calif., one carrier waited a year just for a hearing and another year for a decision. Sprint reported two California counties had applications pending since 2005. In 2008, a Maryland county enacted a 10-month moratorium on any new facilities.

Delays were also increasing. According to CTIA, the wireless trade association, processing time for applications involving new construction went from six months in 2003 to more than a year by 2008. (Collocation review went from 15 days to more than 90 days in the same period.) T-Mobile reported that in Maryland, the zoning process went from two months to nine months in the course of a few years. An application to add an antenna to an existing site in LaGrange, N.Y., was denied after a five-year review. It took a federal court to reverse that decision.

The agency also found that some local governments had a policy of denying any application if service to their residents was already available from another carrier. Verizon noted eight recent instances where applications were denied on that basis, including three in California. In other words, local governments across the U.S. explicitly rejected industry efforts to increase competition in the mobile market.

In late 2009, the agency concluded these practices seriously jeopardized the deployment of mobile broadband and unduly delayed the potential of next-generation mobile services to compete with fixed broadband, both key policy goals of the FCC. So the agency implemented what it called a "shot clock," which went into effect last year. Local governments must now decide one way or the other on cell site applications within 90 days (equipment changes on existing sites) or 150 days (new sites, including new towers). Denying applications because an existing provider already offered service was ruled presumptively illegal.

Shot clock proves ineffective
The new rules were a good idea, but so far have failed to solve some of the most maddening problems for mobile providers and their customers. Sadly, a review of recent FCC filings and court cases suggests the shot clock has done little to change the counter-productive behavior of local governments, particularly in areas with the highest levels of customer complaints. Both CTIA and PCIA (which represents the mobile infrastructure industry) have provided several recent examples of local government interference and incompetence, including:

• In Sterling, Va., an application for a new site disguised as a church bell tower was finally referred to local board review after a two-year delay.

• In El Cerrito, Calif., plans for a new tower disguised as a tree in a Boy Scout camp were abandoned after the city enacted a two-year moratorium on new towers, two years after the application was first filed.

• The village of Muttontown, N.Y., delayed action for three years on an application to place an antenna inside a church steeple, leading Verizon to sue the village under the shot clock rule. (Carriers and infrastructure providers have been forced to sue local authorities who refuse to abide by the shot clock in nearly a dozen jurisdictions, souring relations and endangering future applications.)

Even where the shot clock is obeyed, denied applications often serve no rational purpose. For example, Mount Vernon, N.Y., and other communities continue to require applicants to demonstrate that new equipment will only be "used" by local residents, "frustrating tower siting without any conceivable public benefit."

Local authorities are also expanding their zoning restrictions to prevent additions or changes to existing mobile infrastructure. Even as users take advantage of mobile broadband to replace fixed cable and telephone service, for example, more municipalities have passed ordinances prohibiting or severely restricting the placement of equipment in residential zones. Limits on tower height, meanwhile, effectively require carriers to place more towers, erasing any aesthetic benefit from the height limits.

These delays and costs are often imposed without any justification. Zoning codes forbid changing antennas on towers that were originally approved but which no longer conform. Others require full review simply to add equipment, or require hearings for all collation applications. (Again, Berkeley, CA.)

Both CTIA and PCIA place particular blame on the widespread use by local zoning authorities of outside consultants, who are retained to help review applications. Often, consulting fees are approved by the local agency but paid by the applicant, encouraging delay, waste, and redundant requests by the consultant for the same information.

Last year, for example, a federal court found that the denial by officials in Mount Vernon of an application to collocate six antennas on a rooftop that already had similar equipment lacked any basis. The court also ruled that the consultant's fee of more than $13,000 was illegally padded. PCIA has compiled a list of more than 400 zoning districts that employ the most troublesome consultants, many in California, Virginia, New York, and Massachusetts.

Even after the FCC's 2009 ruling, municipalities continue to deny applications for new or collated equipment if service is already available from another carrier, and some courts have refused to recognize the FCC's clear intention to ban the practice. Some courts have likewise held there is no penalty for violating the shot clock, so long as a decision is eventually reached. This effectively means the shot clock is no clock at all.

The reality: Mergers happen
The shot clock, it seems clear, has done little to help the FCC advance its highest-priority agenda item: promoting high-quality, low-cost broadband throughout the U.S. Regulatory constraints on spectrum and cell site infrastructure have shown no sign of easing, despite the best intentions and efforts of many in Congress and the FCC.

This is the reality through which mobile operators must navigate to deliver faster and more dependable service to a growing customer base with an insatiable appetite for bandwidth. And these are only two of the problems wireless operators must overcome. Yet everyone in Washington agrees on the importance of high-speed mobile services in meeting crucial economic, education, public safety, and other policy goals.

PCIA estimates that providing mobile broadband to all Americans will require an additional 40,000 towers, representing some 53,000 jobs. So the future of the mobile industry, it seems, is largely in the hands of local zoning boards, often made up of volunteers and busybodies supported by nearly-bankrupt local authorities. Try sitting through one of their meetings, and then think again about the wisdom of combining AT&T and T-Mobile's existing infrastructure.

Over the last five years alone, federal regulators have approved more than a dozen mergers of mobile carriers. Yet prices for all mobile services continue to plummet even as product variety and innovation are exploding. Large-scale mergers may be distasteful to some, but there are few real world alternatives if the broadband revolution is to continue. Those opposing the AT&T/T-Mobile deal, in any case, have yet to propose a different solution that comports with basic engineering, constraints imposed by government from top to bottom, or common sense.