2014's #3 Most Read: Why Google Fiber will never come to Seattle
Google is bringing high speed fiber broadband networks to homes and businesses in three cities – Kansas City, Austin and Provo. In February, it announced 34 more cities it will approach for building fiber – Portland, Phoenix, Atlanta and more.
But not Seattle.
And Seattle won't be making Google’s list anytime soon.
The “Seattle Process” and a balky bureaucracy at City Hall stand squarely in the way.
It wasn’t always this way. We were on the short list in 2010, when Google solicited cities to apply and become its launch city. 1,100 cities applied. And Google actually came to Seattle and met with Mayor McGinn and my team when I served as Chief Technology Officer for the City.
Now we’re not even on Google’s “long list”.
But Smyrna, Georgia, and Morrisville, North Carolina, are.
Why won’t Google build here?
1. The Seattle Process.
When Google announced its launch city for Google Fiber – Kansas City – it was a sensation. And the very next day the Kansas City Council authorized a contract with Google for the service. Can you imagine the Seattle City Council keeping a secret like this and then acting on it in just one day? Of course not. We’d need to have endless community meetings and hearings and public floggings of Google Executives. Every citizen in a tinfoil hat who thinks fiber is just another cereal ingredient would have their three minutes in front of the Council.
We also love our lawyers — and haggling over every minute detail of contracts. Overland Park, Kansas, apparently has its own version of the Seattle process. It spent nine months arguing the Google Fiber contract, including an insignificant indemnity clause. Google finally just walked away.
2. Pole attachments.
Seattle has over 100,000 utility poles, most owned by Seattle City Light and many jointly owned with CenturyLink. Under FCC rules, attachments to these poles by others must be allowed. Indeed, that’s the way cable companies Comcast and Wave have built their networks — by stringing fiber and coaxial cables on these poles.
But doing so is not cheap. A City Light pole lease is $28.12 a year. If the pole is co-owned by CenturyLink, the lease is $14.06 a year to City Light. At these rates, building a network on 100,000 poles to serve every home and business would cost Google up to $2.8 million just to rent the pole space.
But leasing the space is not the real problem. There are a lot of wires already on these poles, and many of the poles are old. As poles age, they rot from the center. Adding more cables may cause them to break. So, if Google (or anyone else) wants to add cables, they must pay City Light to survey the route, cut back vegetation overhanging the poles, test each pole and replace them if necessary.
This is patently unfair. Why should the latest company coming in to string wires have to pay the entire cost of the pole replacement? Moreover, this is a process that takes forever. At one time, City Light’s backlog to do pole attachment surveys and “make ready” work was over 18 months long.
3. Permits and Rules.
Oh gosh we love permits. Attaching fiber cable to a pole in Seattle may require a pole attachment permit, a street use permit, and land use and environmental permits, among others.
And we love rules. The Seattle Department of Transportation (SDOT) has a hornet’s nest of them. Rule 2-2009, which restricts the size of the cabinets on Seattle right-of-ways that house fiber, is one of the most appalling. Typically, a few fibers from a central location run to the cabinet, and then fiber or copper cables run from the cabinet to each home.
SDOT requires that 60 percent of the homeowners within 100 feet of a proposed new cabinet must give written consent to allow the cabinet to be placed in the right-of-way. In many neighborhoods, of course, properties are inhabited by renters, making homeowners very hard to track down.
This rule appears to be unique in the nation. Certainly it is not used by any other city in the Seattle area, or by Phoenix, Denver or Minneapolis. As a result, telecommunications companies invest their dollars for improving broadband elsewhere, and cities like Graham, Washington, have much faster Internet speeds than Seattle.
Citizens’ groups have tried to change this rule. UPTUN (Upping Technology for Underserved Neighbors) has been working on it for four years to little avail. And again, this is a rule, not a law or ordinance, which means it could be changed with just a stroke of the SDOT Director’s or Mayor’s pen.
4. Build out requirements.
Build-out requirements are a standard feature of cable company franchises. A city gives a franchise to a cable company to serve a certain area, such as Seattle, north of the ship canal. But the company has to agree to build out and serve every premise in that area. This is a lofty goal because it means all neighborhoods, rich and poor, get served, although it increases the overall cost because the company builds cable on streets with few customers.
That’s not the way Google does it. In Kansas City there were 202 “fiberhoods,” but each such fiberhood had to have a minimum number of customers sign up for service, or the network would not be constructed. Few customers, no service. Google says that 180 of the 202 Kansas City fiberhoods have qualified for service, including 17 of the 20 with the lowest median incomes.
Given all these restrictions, why would Google ever build in any city?
Some cities recognize the value of high speed broadband and are willing to become partners. Kansas City wanted Google Fiber so badly it agreed, in its contract, to review all permits within five days. The city gave Google space, power and “related services” for its equipment at no charge. They also gave Google access to all its assets and infrastructure without cost. These assets included “conduit, fiber, poles, rack space, nodes, buildings, facilities … [and] available land”.
And the city did not charge Google for permit or inspection fees. There are no build-out requirements, although Google does consult with the city in determining which neighborhoods to serve. Kansas City gave Google a ten year contract on these terms.
Did Kansas City residents complain about these terms? You bet they did. According to the San Jose Mercury News, "they couldn't get the service soon enough."
Austin, Texas, is another impending Google Fiber city. Just the announcement of Google’s plans caused Time-Warner Cable to increase its Internet speed sixfold at no increase in cost.
Provo, Utah, is the third Google Fiber city. It enticed Google by selling its city fiber network, built at a cost of $39 million, to Google for just one dollar. Provo and Austin each have robust technology-based economies – and they are going to expand even further with competitive gigabit broadband.
So is all hope lost for Google Fiber in Seattle? What would it take to entice Google here?
This city is extraordinarily generous in investing in its future. It has passed housing levies and library levies and family-and-education levies. We’re building a $4 billion tunnel under downtown which will serve only a small segment of the population — freight and those driving Highway 99. We’ve considered spending the $700 million or more it would cost to build our own fiber network, which might provide a billion dollars in benefits each year.
Could we simply agree to pay for all the pole replacements and permitting as a city, and hire a few extra employees to expedite the process? Couldn’t we just hand over title to a few strands of the 500 mile fiber cable network we’ve built to Google Fiber? And eliminate archaic rules like the infamous SDOT 2-2009?
In return, we’d get a gigabit of connectivity to homes and businesses throughout the city. Each one of us would get 10 times the speed for half the cost. We’d be as connected as Gladstone, Missouri, and Olathe, Kansas.
There's just one question: Do we love our Seattle process too much?