The week of November 23, 2014

Google Fiber may be making the digital divide worse

By Aaron Sankin

For Google, the creation of Google Fiber was a response to a very specific problem: Even though the United States is the undisputed leader in cutting-edge tech, the country’s network of broadband services is shockingly lackluster. The average download speed in the United States is slower than that of Estonia, and residential customers often pay higher prices in the U.S than in countries like France or Japan for comparable service.

The issue, at least domestically, is a lack of competition. A 2013 report by New America Foundation found that the places in the U.S. with the best Internet service were generally the ones where consumers had a variety of Internet service providers (ISPs) to pick from; however, most Americans have few, if any, choices.

The reason Comcast is the most hated company in America isn’t just because of its legendarily terrible customer service—it’s also because a large percentage of their customers don’t have any alternatives if they want high-speed Internet access.

Some cities—such as Lafayette, La., and Chattanooga, Tenn.—have built their own, government-owned municipal broadband networks. Those systems offer some of the fastest service in the country. The network in the small, rural town of Ephrata, Wash., has actually been rated as having the fastest Internet of anywhere in America. But not only are municipal networks few and far between, they’re also often too expensive an undertaking for cash-strapped municipalities just trying to deliver essential services. There’s also been a concerted effort by some conservative legislators to make municipal networks difficult, if not impossible, for city and county governments to deploy.

“When there are more choices, users see the benefit of that. If people are seeing faster speeds that’s a good thing,” Google spokesperson Kelly Mason told the Kernel. “That’s part of the reason why Google Fiber started in the first place—we felt like there was this gap in faster speeds in the U.S. We realized that, if we cared so much about this, why don’t we do something about it?”

When President Eisenhower nominated General Motors CEO Charles Wilson for secretary of defense in 1953, Wilson uttered a line that’s a perfect route toward understanding Google’s motivation for creating Fiber. During a confirmation hearing before the Senate Armed Services Committee, Wilson was asked what he would do if, as secretary of defense, he was faced with a choice between making a decision that was good for interests of the United States military but bad for the bottom line of General Motors—a company in which he still owned a significant amount of stock. Wilson said it wouldn’t be an issue: “For years I thought what was good for our country was good for General Motors, and vice versa.”

The sentiment is easy to ridicule, but when applying that same reasoning to Google, it starts to make some sense. At its core, Google is an advertising company—over 90 percent of its revenue comes from ad sales. Google doesn’t just sell ads on its own properties; the company’s ad network has snaked its way throughout most of the Internet. At this point, regardless of how people use the Internet, Google makes money from it as long as they’re doing pretty much anything online. In a number of very significant ways, what’s good for the Internet is good for Google because the company has spent the better part of the past 20 years successfully making those two things synonymous.

The speedier a country’s Internet, the faster the websites carrying Google’s advertising will load, and therefore the more money Google can make from that advertising.

However, this situation isn’t necessarily true for Google’s newfound competitors in the broadband business. Comcast, Verizon, and AT&T aren’t just in the ISP game to prove a point. Google can win plaudits for insisting that, regardless of what the Federal Communications Commission (FCC) ultimately decides on net neutrality, the company will never charge any online content providers for a “fast lane” to consumers. (By contrast, Google’s competitors sued the federal government to end net neutrality and halt all new high-speed broadband construction after the President Obama voiced full-throated support for banning paid prioritization.)

“Google is popularizing the idea of building essential infrastructure with a market-driven approach. We don’t build roads like that—if we did, there’d be no roads in rural areas.” —Christopher Mitchell, Institute for Local Self-Reliance

Google can afford to stand on the moral high-ground because its revenue model isn’t based on broadband connections. During a question-and-answer session at a 2010 tech conference in Barcelona, Google Chairman Eric Schmidt insisted that the company wasn’t interested in becoming a serious competitor to Comcast. Rather, it was using its vast resources to serve as a model for the type of Internet its leaders would like to see its online services being used on.

“We are not going to be investing in broad-scale infrastructure,” Schmidt said. “We’re going to have the operators do it. Just this past week we announced a test around the one gigabit per second network to precisely do that.”

Google wants to change the type of Internet that Americans have access to largely by reforming the process by which Internet service providers deploy into an area. The company wants cities to start viewing high-speed Internet connections as something so crucial to their future economic growth—using that connectivity as a lure to attract job-creating high-tech entrepreneurs—that they start taking down whatever barriers to entry they had previously erected.

However, those barriers were often put in place to ensure that all of a municipality’s citizens, regardless of their income level, had the ability to purchase a high-quality Internet connection. With Fiber, Google is changing that equation. Google’s design is to shred that type of red tape. Adding competition into the country’s unpopular, slow-moving, and monopolistic broadband market is in and of itself a bigger advantage, the company implicitly argues, than any external obligations a city could place on an ISP.

Google Fiber can afford to, and actually willingly does, a lot of things in the cities where it deploys to close the digital divide—from working with community groups to build excitement around fiber-to-home installations and offering new models for paying for broadband service specifically tailored to low-income communities to giving free gigabit connections to schools, libraries, and nonprofit groups.

But the tightly market-driven system Google is holding aloft as a model for the rest of the industry may have long-term effects that are far more pernicious—effects that could ultimately exacerbate the digital divide between the haves and the have-nots rather than closing it.

‘The ISP for the poor’

When Google first announced in 2010 it would be bringing Internet service capable of pulling in a gigabit of information every second to at least one lucky city, the response from civic leaders nationwide was overwhelming. Google received over 1,100 applications from cities looking to get the nod.

When Rev. Michael Liimatta learned that Kansas City, a metropolis of over 600,000 people stretching from Kansas on one side to Missouri on the other, would be the first to receive Google’s blessing, he was ecstatic. As the chief academic officer of the Christian online university City Vision College, Liimatta had long been concerned about the digital divide that too often separated the rich from the poor—the people with high-speed Internet access from those without. In the Kansas City, Mo. school district, its been estimated that 70 percent of the kids don’t have Internet access at home. After Google’s selection of Kansas City, Liimatta founded a nonprofit called Connecting for Good with the intention of hooking Google’s gigabit connections up to Wi-Fi that could then be accessed by people in Kansas City’s low-income neighborhoods.

“We thought that maybe Google Fiber could become the ISP for the poor, but the idea of sharing the connection via Wi-Fi was totally out because it conflicted with Google’s terms of service,” Liimatta told the Kernel with a sigh.

Frustrated, Liimatta pivoted. He started working to sign low-income Kansas City residents up for the cheapest tier of Google Fiber service. Traditionally, people pay for residential Internet service on a monthly basis. Google’s top-tier service, costs $70 per month. But the company also elected to offer service at 1/20th of that speed for a single one-time fee of $300. Spread over a year, and that’s just $25 per month. Stretched out over the seven-year period that Google promised to support this level of Fiber service, the monthly price dropped to only a few bucks.

At this point, regardless of how people use the Internet, Google makes money from it as long as they’re doing pretty much anything online.

This is, by almost any respect, an amazing deal. Yet, Kansas City’s digital divide stubbornly refused to close.

A survey conducted by the Wall Street Journal earlier this year in a half-dozen low-income areas of Kansas City, Mo. found that only 10 percent of residents subscribed to Google’s gigabit service with another 5 percent subscribing to the slower one-time fee connection. In contrast, when the survey looked at five wealthier neighborhoods nearby, 42 percent of residents had adopted Google Fiber, with 11 percent opting for the slower, cheaper option.

“Someone making less than $20,000 a year isn’t going to put down $70 a month for a gigabit connection. The free connection is only 5MB but, for some folks $300 is way beyond their budget,” Liimatta noted. “Google Fiber brought faster Internet to people who already use the Internet, but it didn’t get a whole lot of new people using the Internet.”

In fairness, asking one private company to single-handedly solve the issue of income inequality is a ridiculous proposition, no matter how not “evil” that company prides itself on being. Google representatives have said that the $300 deal is about what it costs the company to run Fiber to someone’s home, meaning that Google is basically breaking even on service—something the company can only justifying doing because it’s able to monetize those users simply by them being online.

However, the question of cost is only even an issue in the neighborhoods where people can get Google Fiber in the first place.

Putting on the rally caps

Sometime next year, the some residents of Austin, Texas, will get to experience the joy of having Google Fiber deliver piping hot Internet, undoubtedly boasting some of the fastest download speeds of anywhere on the planet, all over every last inch of their homes.

But not everyone in Austin will necessarily be able to sign up for Google Fiber, even if they are able to spare $70 a month for ridiculously fast Internet. When Google selects a city to deploy Fiber, it’s not guaranteed that every part of that city will be eligible for that service.

“[The way that cable companies historically operated,] there would often be franchise fees or build-out requirements that would make a company start building out in certain neighborhoods first or they couldn’t activate without a certain portion of the geography,” explained Brian Dietz, spokesperson for the National Cable and Telecommunications Association. “Those rules were meant to avoid new providers entering the market from being able to cherry-pick the most profitable neighborhoods.”

Google can afford to stand on the moral high-ground because its revenue model isn’t based on broadband connections.

Dietz added that, while the obligations put on the incoming ISPs could vary widely from one case to another, other common requirements included carrying local cable access channels or providing connectivity to government buildings.

Google, on the other hand, uses a different model. Google slices up each city into hundreds of different “fiberhoods,” which can qualify for service if enough people within each one show interest through a process called a “rally.” In Kansas City, Google dispatched employees into each individual neighborhood and worked with local community groups, sometimes employing tactics like handing out free ice cream, to hit the requisite number of households in that area to justify deployment.

This Google promotional video details the process:

Google’s decision to pick and choose where it wants to deploy is understandable. Expanding into a given area requires installing expensive hardware. Knowing ahead of time there are enough people in a given neighborhood willing to buy service is a consistent way to avoid losing money on that investment.

Ensuring that everyone in the neighborhood has the ability to sign up for Google Fiber isn’t the only reason for community activists in a given fiberhood might want to go door to door signing up people for Fiber. Google has a program called Community Connections where Google teams with city governments to identify about 100 schools, libraries, and nonprofit groups that the company will give free gigabit connections to—as long as the fiberhood surrounding each of those institutions has reached its registration goal.

The Community Connection grantees are listed far ahead of time, like on this map of selected sites in Austin included below, so organizations can leverage their local neighborhood support to encourage sign-ups.

Despite this added incentive, when Google went through the initial rally process in Kansas City in 2011, the company found that far more fiberhoods on the west side of Troost Avenue, the city’s long-standing dividing line between black and white (alternately, middle-class and poor), qualified for service than did fiberhoods to the east.

To Google’s credit, it commissioned an independent study looking at the problem, staged a new round of rallies that focused on the neighborhoods that didn’t previously qualify, and allowed some multiunit apartment buildings—the type in which a lot of lower-income people tend to live—to waive the $300 sign-up fee entirely.

In November, the company unveiled a new program in Austin called Unlocking the Connection that takes this community service aspect to the next level. In a partnership with the Texas capital’s public housing authority, Google agreed to give free broadband connections to the approximately 4,300 residents of all 18 city-owned public housing projects—as long as the buildings’ surrounding fiberhoods demonstrate enough interest to quality for service.

It’s this last move that shows why Google’s methods for selecting fiberhoods is simultaneously noble and problematic. Google is able to go back, expending money and time, to the communities who were left out, even to the point of giving some of them free connections, because Google can still make money from those people just by their being online. If Verizon or Time-Warner Cable were put in the same situation, they’d have little incentive to pursue further action.

Neither Verizon nor AT&T responded to a request for comment.

“I think this [rally model] is going to catch on and it worries me greatly,” insisted Christopher Mitchell of Institute for Local Self-Reliance, a nonprofit group that advocates for the expansion of municipal broadband networks. “Google is popularizing the idea of building essential infrastructure with a market-driven approach. We don’t build roads like that—if we did, there’d be no roads in rural areas. We don’t build electricity like that—if we did, our economy could be far weaker. We recognize that those things are essential infrastructure.”

This demand-driven model is one that, even without Google, is starting to catch on nationally. With Google using Fiber as a publicity campaign for how the company would like to see high-speed broadband deployed across the board, it not only gives the company’s more traditional competitors greater incentive to discriminate based on socioeconomic geography, but it also pushes cities to let them do so.

“I actually think that what Google does is less bad [than what other ISPs do] because it at least gives people a chance,” Mitchell explained. “With AT&T’s U-verse and Verizon’s FiOS, they just pick the areas where they’re going to unilaterally. Some people are really on Google about this, and I think it’s unfair to give Google different expectations for serving neighborhoods when existing providers have ignored those neighborhoods for a long time and also aren’t meeting their needs.”

The way of the future

The mindset of letting ISPs deploy wherever they want within a given municipality wasn’t born with Google. Over the past decade or so, at least 25 states have passed laws mandating something called statewide franchising.

Starting at the dawn of the cable era, if a telecom company wanted to expand into a new market, it would have to draw up a contract with the local government that often required it to it to build out the capacity to reach pretty much everyone who lived there. However, as America’s once-diverse ecosystem of cable companies gradually consolidated into a small handful of massive providers (that just so happened to almost never actually directly compete with each other), there was a realization that signing all of these contracts was an expensive hassle.

Google wants to change the type of Internet that Americans have access to largely by reforming the process by which Internet service providers deploy into an area.

With the support of the American Legislative Exchange Council, a conservative group that writes pro-business model legislation that elected representatives can then introduce in statehouses across the country, states passed laws making it so that ISPs no longer had to deal directly with local governments. Instead, they could simply strike a single deal at the state level and then expand however they pleased.

“The city doesn’t have any authority to tell Google where it should build,” explained Rondwlla Hawkins, a telecommunications and regulatory affairs officer with the city of Austin. “We lost that ability with statewide franchising [because] there are no build-out requirements in the state regulations.”

Excluding Google Fiber, there are currently three different options for Austin residents when it comes to broadband Internet service. Time Warner Cable came in before Texas passed a statewide franchising law in 2005, which meant the company was obligated to build out infrastructure to serve the vast majority of Austin residents. Grande Communications came in the 2000s. It was initially going to deploy citywide, but when the state franchising legislation passed, the company elected to deploy in about 20 percent of the market. AT&T’s U-verse moved in after that with a state franchise and doesn’t cover the entire city.

If Google’s goal with Fiber is to inject competition into the market, not covering the poor areas of the city allows its competitors to avoid having to make changes to make their services more attractive—perhaps through pricing or eschewing credit checks—to people in those areas. By excluding some low-income areas, the needs of the people who live in those areas won’t be as directly addressed.

All of the cities where Google Fiber has currently deployed have statewide franchising, but the company insists lacking it isn’t a dealbreaker. Even still, the market-friendly sentiment of the statewide franchising laws remains.

Google has etched an agreement with the city of Portland, Ore., through local franchising that allows it to only serve the portions of the city that qualify through the rally process, even though other ISPs active in the city had more robust build-out requirements.

Everyone involved seemed to predict that this agreement would inspire jealousy and, as part of the contract, both parties pledged to join in for a joint legal defense of the deal in the case of lawsuits that Google was getting unfairly getting preferential treatment.

Even so, that jealousy is kind of the point.

After Google came into Kansas City, Time-Warner Cable said it would up its game there if it was granted the same deal Google received, including the ability to determine its own service area, to which Kansas City agreed.

Regardless of whether Google Fiber actually comes into a city, the company encourages every municipality applying for Fiber consideration to have some kind of agreement ready that could entice anyone, be it Google or another provider, to come in and deploy a super-fast fiber-optic network.

“Google has essentially validated the concept of cherry picking neighborhoods and extracting favors from cities and … [municipalities] that may hamper their financial positions. Now AT&T and CenturyLink are also following a similar course,” insisted Roslyn Layton, a fellow at the American Enterprise Institute’s Center for Internet, Communication, and Technology Policy. “So, left unabated, this policy would no doubt result in extremely bifurcated capabilities maps for these providers, leaving haves and have-nots.”

On the other hand, maybe ripping down barriers to entry is a good thing. Maybe anything that encourages more competition in the U.S. broadband market is a positive development. Maybe, if it’s easier for ISPs to enter into new areas around the country, some of those ISPs will compete on making high-speed Internet affordable for people who were previously priced out of the market entirely.

That was certainly the case with Liimatta and Connecting for Good.

When using Google Fiber as the backbone for a free Wi-Fi network fell through, Liimatta was undeterred. “We began to look at other options and became our own wireless ISP,” he said. “Now we use the Clear Network to give people discounted Wi-Fi hubs supplying free Internet to about 500 households in Kansas City.”

It may not be the specific model that Google was advocating for with Fiber, but seeing that it gets people online, it still serves the same end goal.

Illustration by J. Longo