A happy confluence of political circumstances has made rural broadband a hot topic and makes it possible to believe that perhaps, finally, the stars will properly align to do something more than the Connect America Fund. No offense to CAF, but everyone knows that CAF alone cannot provide quality, ubiquitous affordable broadband to all Americans. Not by a long shot.) Needless to say, Republicans and Democrats have rather different approaches to how they want to close the rural digital divide. I’ll save a comparison of what’s out there for a different post, because I want to take this opportunity to propose an entirely different approach than anything else out there at the moment.
It begins by recalling some wisdom I learned at my father’s knee. My father teaches tax law at Boston University. When grading student exams, he would often shake his head and sigh. “Answer the question asked,” he would say. “Don’t answer the question you want to answer because you have the answer, answer the question asked.”
What does that have to do with rural broadband? When we think about solving the rural broadband problem, nearly everyone tries to answer the question: “How do I find a carrier to serve rural areas.” But that’s not actually the problem we’re trying to solve. The problem we’re actually trying to solve is getting people access to quality broadband so they can participate in the modern digital economy and modern society generally. On the surface, that may look like the same thing. After all, you can’t get broadband access without some kind of carrier, right?
But if we start by framing the question in terms of a goal (get people broadband access) rather than a solution (find people a broadband carrier), we open a whole new world of solutions and approaches. As I discuss in more detail below, the reason rural communities don’t have broadband access is fairly straightforward: the communities in question are not sufficiently profitable to serve to justify the investment by profit maximizing firms (I’ll get to the importance of the word “sufficiently” below). If we then apply the skills we all (hopefully) learned back in high school math, we then break the problem down into solvable components. So we can either (a) raise the profitability of the target area; (b) lower the cost of deployment and operation; or (c) find entities that are either not motivated by profit or that are satisfied with much smaller profits.
We solved this one way back in the 20th Century. But the great virtue of the modern communications market, which allows us to break up the supply chain and bring in economies of scale from other markets, provides us with a bunch of new ways to solve the problem. Ideally, used in combination, we can have a solution that doesn’t lock rural areas in to a single, permanently subsidized provider, but instead closes the digital divide and enhances competition and potentially drives down everybody’s costs.
Short version — fund infrastructure, not carriers. And by “fund” I don’t just mean “throw money at,” although we need to be clear there is no way to avoid throwing money at this if we want to get the job done.
Lets break this out below . . . .
First, it’s important to understand why it isn’t profitable to serve rural areas. Turns out, for pretty much everywhere this remains a problem it’s not regulation. Yes, you can in theory drive up the cost of doing business so high that it has an impact. But if we take the statements of CenturyLink and Verizon at face value, and do a bit of digging into Comcast, it’s hard to find evidence that the current frenzy of deregulation at the FCC or in the states is going to move the needle much. Additionally, the states that have drastically deregulated their legacy voice service and prohibited regulation of IP-based services have not seen massive investment in rural areas as compared to states that were less deregulatory. To the contrary, where companies have been permitted to abandon legacy networks as a consequence of deregulation, they have stopped making investment in even basic maintenance.
Unsurprising to anybody who cares about real economics, its just a simple rate of return problem. As this old New America Foundation paper was pointing out back in 2010, the basic problem for rural broadband is that it has very low population density covering a lot of geographic area. Additionally, rural areas tend to have lower household incomes than average. As a result, unless you live near an Interstate (which gets covered by wireless because the highway gets traffic), it is simply not worth it for a carrier to serve the area. This changes to some degree based on size of the carrier and its business model and whether it is publicly traded. Verizon has offloaded most of its rural lines to buoy its revenue-per-subscriber. Companies like Frontier and Fairpoint have a different business model where investors expect lower average revenue per unit (ARPU), but even these companies have a limit to how low they will go. Eventually, you hit the point where there aren’t enough customers — or the customers can’t afford to pay enough — to justify the expense of build out. Even where it may be profitable, it isn’t profitable enough to interest traditional carriers.
To solve this problem, you have three options.
- Make the customers more valuable.
- Reduce the cost.
- Find a provider not interested in profit, or for whom the available profit is sufficient.
We used all three approaches to get telephone service (and other high cost networks, like electricity) to 97% of all Americans and created a national phone network that actual was the envy of the world.
How Did We Get Telephone Service Everywhere?
Back in the 20th Century, we had the same problem — how do we get a phone on every farm? We solved it by having a big monopoly provider (Ma Bell) required to serve the entire franchise area (which included profitable urban areas and less dense rural areas), guaranteed a (comparatively, for those days) high rate of return (15%) by overcharging the urban areas to subsidize the rural areas. This made the rural customers “more valuable,” because you couldn’t get the high return urban customers without taking the low return rural customers. Since you were allowed to overcharge the urban customers on long distance to make up for this (that rate of return was for the whole franchise area), it worked out.
But even this wasn’t enough. We also supported independent rural carriers by charging a high fee for an urban call to reach (“terminate”) on a rural network. Since that is exactly the opposite of how a real world negotiation works, we forced it by regulation. This provided a subsidy to drive down cost. Finally, we also relied on co-operatives (co-ops) and municipal systems. These providers don’t care about profits because they are not profit maximizing firms. They just need to cover costs.
Then we broke up Ma Bell. So the FCC replaced some of the implicit subsidies with an explicit subsidy which we now call the Universal Service Fund (USF). Through the Lifeline program, USF subsidizes customers so they can afford service (making them more valuable). Through the High Cost Fund, it subsidizes the build out, maintenance and operational cost of carriers (effectively reducing cost). The downside of this is that we all pay for it with a fee on our phone bill. If you are a Republican who doesn’t believe anyone should pay taxes because “taxes bad tree pretty,” or a Democrat who lives in fear of being accused of raising taxes, we have not managed to actually do anything to apply USF to broadband other than CAF. (Shout out to Reps. Doris Matsui and Anna Eshoo, who have tried for years to get legislation passed that addresses this.) Additionally, because the FCC lives in either mortal fear or outright hostility of classifying things as Title II (depending on the Administration), the transition away from traditional voice services has been steadily reducing the contribution base. But I digress . . .
More importantly, the biggest problem with this system is that it created permanent monopolies. It basically boiled down to “find a carrier to do this, then pay the carrier to serve the area.” When everyone had a monopoly provider, that was OK. When we were willing to rate regulate the monopoly rural providers, that was OK. But in our brave new world of Ajit “Crowley” Pai, where “do as thou wilt provider shall be the whole of the law,” condemning rural areas to monopoly providers is not a geat solution — if we can figure out a better one.
And That Better Solution Would be . . .
The beauty of modern communications networks is that we can actually break up the supply chain and target subsidies to be much more specific. We can subsidize infrastructure instead of subsidizing carriers. The advantage of this is that by subsidizing infrastructure, we can subsidize infrastructure for many potential competitors (or at least more than one), rather than basically having a monopoly provider we either need to regulate up to the eyebrows to make sure we actually get decent, affordable service in exchange for the subsidy. Additionally, we have a lot of different ways to lower cost that actually lower cost. If we do that, we can actually see local businesses and local institutions willing and able to provide service for profits that, while perfectly reasonable for a local business, would be utterly uninteresting to even a small traditional carrier.
That Sounds Like More Socialist Progressive Field of Dreams Poppycock. What Makes You Think That Could Ever Work?
The natural evolution of WISPs over 10 years ago, and their continued economic viability even under existing conditions.
Would You Care To Explain That?
WISPs are wireless ISPs. If you bop over to the Wireless ISP Association (WISPA) website, you will find that there are about 3000 WISPs serving about 3 million or so residential customers and businesses — largely (but not exclusively) in rural areas. Like the bumble bee’s supposed inability to fly, WISPs should not exist. Most are small businesses (some with no more than one or two employees), and hence should lack economies of scale. They are serving areas that should be insufficiently profitable to support them — or competing agains larger firms that should be able to crush them with superior economies of scale. Traditional network economics from the 20th Century makes it clear that businesses this small, and in these areas, should not be able to provide service.
But WISPs no more defy the laws of economics than bumble bees defy the laws of physics. What WISPs demonstrate is that under the right conditions, it becomes profitable enough for certain types of businesses to provide broadband services. Specifically:
- WISPs are small businesses, usually folks who live in the area they serve. They don’t need the kinds of profits that attract publicly traded companies. They simply need the kind of profit that works for your average small business. Sure, you have WISPs that are growing and folks with ambition. And Ray Kroc took a single burger place and turned it into one of the most iconic brands in the world. But most folks who run a restaurant are happy turning a decent profit even if they don’t grow to become an international chain.
- WISPs began relying heavily on unlicensed spectrum. While lots of WISPs through various types of licensed spectrum in the mix as well, relying on licensed spectrum means not paying tens of millions of dollars at the outset for spectrum access.
- Additionally, and most importantly, relying on unlicensed spectrum moves economies of scale from the carrier market to the consumer market.
Lets expand on that last point, because it is critically important. One of the big advantages large carriers have is economies of scale. Carrier grade equipment is, by its nature, sold only to carriers. As a result, the price reflects the fact that fewer customers will buy the equipments, so you don’t have the same kinds of economies of scale you do for consumer premise equipment (CPE). You need to be a big company to get the volume discounts that bring down the price.
But modern equipment relies heavily on chips and software and other off the shelf components. Once you develop a chip to operate in unlicensed spectrum, of the kind you put in a standard Wi-Fi device, you can modify it relatively easily to work in other ways. Since just about every device manufactured these days is designed to connect to other devices using unlicensed spectrum (either Wi-Fi or Bluetooth), the cost of chips for devices using unlicensed spectrum is dirt cheap. So even though WISPs use “carrier grade” equipment, and rely not just on Wi-Fi, but on other protocols as well, they achieve enormous economies of scale by taking equipment developed for the consumer market and then customizing it for the carrier market. (Nor are WISPs the only carriers using unlicensed equipment. Cable operators, private network operators, even operators of traditional mobile networks using licensed spectrum supplement their networks with unlicensed spectrum.) So WISPs get to enjoy similar economies of scale for their carrier grade equipment that consumers enjoy for consumer devices and don’t have to be giant sized themselves to get economies of scale.
So IF WISPs Are The Answer Why Do We Have A Problem?
WISPs aren’t the answer. WISPs are part of the answer. But, more importantly, WISPs provide a real life demonstration that we do not need to rely on the traditional “find a single carrier and pay the carrier” to bring broadband to rural America. If we focus on providing infrastructure, either indirectly by providing necessary inputs (like spectrum) or directly (for example, by building towers or backhaul fiber), we will see entities interested and eager to provide service in regions that traditional carriers do not find sufficiently profitable to be interesting.
How Do We Fund Just Infrastructure?
We did this already with the Broadband Technology Opportunity Program (BTOP), which we funded as part of the broadband stimulus. Much of the BTOP funding went to things like backhaul fiber networks open to any provider that wanted to use them. That form of infrastructure made it possible to bring more high speed providers to rural areas, since the speed of the last mile network is limited by the speed of the backhaul network. (That is to say, however good your local last mile network may be, it still must reach the rest of the Internet. If your last-mile fiber network can only reach the rest of the Internet through an old copper line that runs along the highway, then your local network speed is irrelevant.)
But as the WISP experience shows, we can fund infrastructure indirectly as well as directly. Increasing the availability of open spectrum, for example, lowers the cost of a critical input. And not just for one carrier, but for everybody. Making more spectrum available not only promotes rural broadband, it also promotes competition — and without appropriating a single new dollar in spending.
How About Some Non-Spectrum Examples?
While there are certainly differences between ow Democrats and How Republicans are approaching infrastructure (including broadband infrastructure), we can find some good bipartisan examples of the approach of encouraging broadband deployment by either funding infrastructure build out or otherwise lowering the cost to provide service to anyone who wants to provide service. Rep. Anna Eshoo (D-CA) and Rep. David McKinley (R-WV) recently introduced two bills that lower the cost of deployment to any entity willing to offer broadband service. The CLIMB ONCE Act makes it easier for broadband providers to run wires or attach wireless antennas to existing utility polls. The Broadband Conduit Deployment Act would require that when federal money is used for road construction, the construction include a plastic conduit that makes it possible to pull fiber without digging up the road. Taken together, these bills help to reduce the cost of deploying broadband networks, effectively subsidizing broadband deployment in a pro-competitive fashion and without spending any new money.
Another example is the AIRWAVES Act, sponsored by Senator Corey Gardner (R-CO) and Senator Maggie Hassan (D-NH). In addition to enhancing access to unlicensed spectrum, the AIRWAVES Act, requires that 10% of the revenue from future spectrum auctions be allocated to a fund specifically to build broadband infrastructure. While nothing prevents the fund from subsidizing a single carrier, the language of the bill would also allow subsidizing infrastructure that could be shared by multiple ISPs, such as towers or backhaul fiber. By providing the initial capital funding for basic infrastructure, the federal subsidy would reduce the overall cost of deployment without needing to pick as specific carrier to receive the federal funding. Furthermore, by funding the underlying infrastructure, the money would attract carriers capable of sustainable operation over time without the promise of additional federal money.
Conclusion.
To be clear, no one of these approaches will totally solve the rural broadband digital divide. Getting broadband to all Americans is a complicated problem. Like rural electrification, and universal service for telephony, no one answer will work everywhere. Rural New England is very different from rural Appalachia, which is very different from the rural Southwest. Furthermore, just as in every generation since the United States commissioned the first Post Master General, we will invariably need to spend real money to keep ourselves connected as one nation.
But just as each change in technology has required us to spend money, it should also inspire us to find new ways to bring essential services to everyone that take advantage of the new capabilities of these technologies. We should certainly learn from the past, but we should not slavishly repeat it. If we really intend to make good on our national commitment in Section 1 of the Communications Act to provide broadband to all Americans, then we must remember that the goal is to provide everyone with access to ubiquitous, affordable broadband, not to fund carriers.
Stay tuned . . .
