Surveying the landscape of internet access, one could be forgiven for a single dank conclusion: Winter is coming.
We know that Big Cable’s plan for high-speed internet access is to squeeze us with “usage-based billing” and data caps, so as to milk ever-growing profits from their existing networks rather than invest in future-proof fiber optics. We are also seeing that Big Cable has won the war for high-capacity, 25Mbps-download-or-better wired internet access, leaving AT&T and Verizon to concentrate primarily on mobile wireless. Indeed, Big Cable’s share of new and existing wired-access subscribers has never been greater — cable got both all new net subscribers in the third quarter of 2015 and captured millions of subscribers fleeing DSL — and its control over this market is growing faster than ever.
Wall Street analyst Craig Moffett predicts that, in the end, unless things change, cable will have 90 percent of subscribers in areas where it faces competition from only traditional DSL and will have the lion’s share of subscribers in areas where cable faces competition from souped-up copper-line DSL and fiber-to-the-node (aka “fiber to the neighborhood”).
We’re already seeing the deadening effects of this. Pew reports that home adoption of high-speed internet access has plateaued, while the percentage of smartphone-only users in the United States is growing. Just 8 percent of Americans were smartphone only in 2013. That number is now 13 percent—mostly lower-income households, minorities, and rural Americans. What’s the reason for nonadoption? Mostly cost: The monthly fee for high-speed internet is the main reason most of these people don’t have access at home. Smartphone-only users just don’t have same the quality of access as home high-speed internet subscribers. We are amplifying and entrenching existing inequality by not taking on this country’s internet access problem.
So where does that leave us? Are we really going to wait for the Cable Guy to bring everyone in America the same modern internet access capacity at reasonable prices that other countries have had for years? Judging from the history of this industry, in which every DNA strand is encoded with monopoly genes (unaccompanied by any oversight), we’d have a better chance with Santa Claus.
We need a better plan — a better vision — if we want to unlock the full benefits that access to the internet can bring Americans. Right now, as a country, we’re investing inefficiently and in the wrong things at a time when we should be unleashing private capital to invest in smarter, faster, and cheaper ways.
First, Americans should invest only in technologies that can meet both our future and immediate needs. That means fiber optic networks, which not only can handle the advanced internet services of today, but also will be able to bring us internet applications that evolve over the next 30 years.
We should stop throwing money at old, copper-based technologies, like the ones Big Cable and the telcos rely on. Because of its limitations over long distances, copper can’t handle many current internet applications, like telemedicine, much less the applications coming in the next few years. And since much of the cost of a network lies in wiring the last mile to homes and businesses, we shouldn’t pay to dig up our roads to put in copper lines that will be obsolete almost as soon as they’re installed.
We should also stop thinking that satellite and mobile wireless technologies are the answer. They don’t provide the speeds and bandwidth we will need. And it’s awfully far from the ground to satellites, which means that satellite transmissions won’t be a substitute for what’s possible over fiber. We can’t change the speed of light. Anyway, mobile wireless, with its usage caps and high prices, has sharp limitations that will dampen Americans’ use of the Internet.
Fiber is the best — the only — alternative, and it could win subscribers in areas where cable exists. But we’re not building these networks. Verizon sold its FiOS lines in California, Texas, and Florida, and AT&T’s GigaPower fiber plans are (with small exceptions, where the company feels the pressure of genuine competition) mostly “fiber to the press release.” AT&T is most likely to continue playing defense by further squeezing the capacity of its existing copper lines, which means that only densely packed apartment buildings that are very close to AT&T’s central offices will see the advantage of the company’s work on its copper connections, as signals can’t travel very far over copper. (By contrast, signals can go for dozens of miles over fiber without being boosted.) And despite all the hype, Google Fiber is predicted to reach no more than 10 million homes over the next few years — although Google’s recently announced plans to consider Chicago and Los Angeles may change that part of the picture.
Some of the problem with building fiber networks is that the needs of these profit-grabbing companies diverge from the public good, which requires long-term investments where the gains accrue to the economy in general. A sensible fiber system would be part of an overall smarter, 21st-century infrastructure. We should integrate planning for essential energy, water, and transportation systems with our communications transport planning. We need to affirmatively plan for fiber systems as the backbone for smart grid and distributed energy systems as well as educational networks. If we want our infrastructure to be resilient in the face of natural disasters and terrorist threats, not to mention helping us to mitigate climate change and support the growing Internet of Things, we will need a fiber-optic communications system that can respond in real time as demands change. A sensible network will work in complementary fashion with WiFi — we’ll need fiber lines deep into the neighborhoods and buildings to which those WiFi hotspots are attached. We should set high standards for ourselves — the FCC’s definition of “broadband” at 25 Mbps for downloads and 3 Mbps for uploads won’t cut it for many interactive internet uses. And to avoid waste and inefficiency, we need to get it right from the beginning — and not just hope we’ll get there with our current patchwork quilt of federal, state, and local government agencies and private utility planners, each with different goals and motivated by different incentives.
It’s got to be a single vision. Otherwise it won’t be the smartest one.
We have the technology and resources to build these systems, and there’s no reason we shouldn’t produce them at, well, internet speeds. It’s really not that hard. Coordinated planning will help speed things up by avoiding the mistakes that short-term thinking always triggers. But we also need to lower the barriers that exist to getting fiber-optic access built quickly and efficiently.
One obvious place to start is to ease the process of getting permits and rights of way. (Google works on this every time it approaches a city for possible installation of Google Fiber.) These requirements differ from place to place and often result in a maze of separate applications, review periods, and negotiations each time a network expands, which add time and cost to every buildout. We need to streamline these processes and make them more consistent. We also need to have centralized, online ways of moving permits and rights of way through the bureaucracy.
Less obvious but just as important is the need to make it easier to share facilities like poles and underground conduits, which are essential to fiber-optic network construction. Right now, for example, fiber-optic network builders have to negotiate access to each telephone or electricity pole and each underground cable conduit needed to build out a network — sometimes thousands of them. What’s worse is that often these agreements need to be made with competitors — and most have no real incentive to act quickly or reasonably.
To fix this, we need a new system for shared facilities. Connecticut did a splendid thing in setting up a “single pole administrator” for the entire state. In addition to scaling up this kind of centralized clearinghouse, we need to allow any licensed technician to do the job of making poles and conduits ready for shared access — using standard industry practices that will promote consistency across networks. That way we can stop giving cable and telephone incumbents the right to impose arbitrary requirements and endless delays on building the internet access we need — all of which act as nothing but a dead weight on innovation and an unnecessarily burdensome way to preserve incumbent profits.
Faster also means making sure that funding is available to spread best practices and project management skills across the country, particularly to rural areas that don’t have infrastructure project management expertise on tap. It’s important to standardize project methodologies. It’s important to create a big market in pre-build services like planning, feasibility studies, community engagement, financing, construction planning, and marketing so they all work better for everyone.
Once we decide to build the right thing to the right standards, building these networks will require capital. Right now, not enough capital sources have incentives to get involved with fiber builds in the United States.
Here’s where the Obama administration could make important progress over the next year. And I have a suggestion for funding.
Fiber networks should be seen as a source of modest, steady, long-term returns for communications infrastructure, like the returns on investments in toll roads and bridges. Pension funds and other institutional investors want those kinds of returns because they need to fund obligations that stretch out for generations — and that’s exactly what investing in fiber infrastructure would give them. But right now we’re not giving them that chance.
Let me explain. When municipalities issue bonds, the interest income these bonds pay to holders isn’t subject to federal taxes — it’s excluded from gross income. That makes muni bonds very popular with individual U.S. taxpayers. But if you are a sovereign, a pension fund, a charitable foundation, or an overseas investor, you don’t pay U.S. taxes anyway, so you won’t accept less interest on tax-free municipal bonds when you can get more interest on taxable bonds. So these other actors don’t invest in the municipal bonds that we’ll need for fiber builds.
This makes no sense. We need to open the market to include all the long-term investors we can find. Why should we subsidize municipal investments in roads, bridges, and tunnels by giving individual investors tax-free interest on municipal bonds but not provide equivalent support to critical internet infrastructure by giving institutions reasons to invest?
Back at the start of the Obama administration, an exciting program called Build America Bonds (BABs) fixed this problem. Instead of indirectly subsidizing bonds by refraining from taxing their interest payments, the Treasury Department provided a direct subsidy to municipal issuers of infrastructure debt equal to 35 percent of the taxable interest on the bonds. In other words, BABs were taxable bonds on which Treasury paid a direct subsidy to the issuer to offset their higher borrowing costs. This substantially lowered the cost of borrowing. (Here’s the report on BABs that I’m using to support these points.)
Importantly, this feature also made BABs attractive to a broader group of investors than those who typically invest in more traditional state and local tax-exempt bonds — they had the cash passed on to them.
And it worked. According to a Treasury postmortem, in under two years, a total of $181 billion BABs were issued, with the lion’s share coming from the institutional investors it targeted. And it worked even better for state and local governments, which used these subsidies to save $20 billion in borrowing costs.
We need a program like BABs for fiber-optic internet infrastructure around the country, particularly in the last miles between residences/businesses and local internet access points. And we need to make sure it extends beyond state and municipal issuers to include the public/private partnerships that are building that infrastructure, which we can do by relaxing restrictions on subsidies for internet infrastructure projects that serve both public and private interests. (We should always prefer dark-fiber builds where possible.)
In addition to increasing the pool of investors, we need to reduce the cost of capital for internet infrastructure by shifting inappropriate risks away from investors. For example, the risk of changes in law and taxation should fall on the government.
Many ways exist to shift risks like this. One common technique is through partial guarantees against specific government actions. It won’t take much capital to back up these guarantees — after all, they’re only guaranteeing against some contingency happening, not funding up-front the entire cost of an infrastructure build. And these partial guarantees will allow a much larger group of investors to participate at lower cost. Again, long-term institutional investors are happy to exchange lower returns for lower risks.
That’s only a start to get these networks up and running. To fully pay for them, we’ll need additional lending as well as bonding. But we have the wrong actors involved. Traditional banks doing construction lending want to see their money back in three years. That doesn’t fit the fiber infrastructure model: These are long-term, 30-year investments. We need to change things.
So here’s another suggestion. If we’re really serious about having internet access infrastructure that will meet our future needs and keep us competitive in a fast-changing world, we need an institution—a dedicated body that can make the necessary long-term lending decisions to build a future-proof fiber-optic network infrastructure. This would be an institution that marries the benefits of a centralized system (don’t recreate the wheel) with the advantages of a regional perspective — and one that’s insulated from the tug of short-term politics.
That’s not all. (Hey, it isn’t easy creating an alternative to Big Cable.) We also need an second institution with deep expertise in building, financing, and maintaining cutting-edge internet access infrastructure, like the Overseas Private Investment Corporation (OPIC), a government agency that has been investing in infrastructure abroad for decades as part of our foreign policy efforts aimed at helping our allies.
So here’s my dream pairing: What if some OPIC evaluators were detailed over to the 12 regional Federal Reserve Banks to help them evaluate domestic fiber projects for lending and guaranteeing purposes? As we watch the Fed change our monetary policy by raising rates, let’s remember that these 12 regional institutions were set up to provide regional economic perspective and analysis, coordinated by a centralized board that can’t be fired for pursuing particular policies. The Fed can take the kind of long-term investment perspective that we need right now.
Smarter, faster, cheaper. Now.
That’s my recipe for providing America with the connectivity it needs and deserves.
Big Cable and the telephone companies will undoubtedly claim it will be too expensive to connect everyone with a fiber-optic network. But is fiber really more expensive in the long run? Fiber will last 30 years or more — and it can be upgraded by installing new electronics rather than digging up the streets every few years. Operating a system of flexible glass fibers costs a fraction of what it costs to operate a system that relies on aging, corroding copper wires.
What about the hidden costs to our economy in terms of lost productivity and innovation? Isn’t it more expensive at the end of the day for all of us to pay more than we should for a legacy cable TV system because we have no alternative for high-speed internet access? And every year we pay for it, we are adding the opportunity cost of going to fiber.
In the long run, a fully fiber-optic network will be cheaper. And it will be cheaper still if we can unleash long-term investment capital to fund it.
Or we could wait for Comcast and Google. And suffer through a very long winter.