Ironically, we had too much good, adaptable infrastructure to make this picture evolve in some orderly, manageable fashion in which course corrections could be made within the sloth-like patterns of public agencies.
The smart phone and Internet access have become so essential to our lives that we forget how fast they rose, from nothing to necessity. Indeed, it is very hard to remember that when Congress took up the 1996 Act, in 1993, the fastest Internet access speed was 33 kbps, cell phones were the size of a butternut squash, and none of the following existed: HTML (mid-1993), Amazon (1994), Google (1998), Facebook (2004), Twitter (2006), the smart phone (2007), and smart phone apps (2008). From Apple’s introduction of the iPhone to 2018, eleven years, more than 4 billion smart phones have been sold, more than 23 billions devices are now connected to the Internet, and daily text messages passed 8.5 billion a day in the United States. Indeed, reading a novel written before 2008 feels like entering an ancient age—there are no smart phones, no texting, no tweets, no Instagrams, no restaurants with every patron’s smart phone placed next to the forks and orders placed on iPads.
The Past Is More Than Prologue
None of this happened without broadband networks. But rather than build new networks, as were required for electricity and telephone service, each taking many decades to reach half the country, broadband could be delivered over existing cable television and telephone networks using special modems that allow digital data services to share lines with analog television and telephone signals. Telephone line modems, called Digital Subscriber Line (DSL), were standardized in 1993 and were commercially available by 1997; Cable Modems followed a similar trajectory. Each was greatly helped along by Moore’s Law that produced ever more complex signal processors at ever-decreasing costs, enabling modems to realize ever-increasing data rates. (In 1965 Gordon Moore predicted that integrated circuits would double in capacity and halve in price every two years, a geometric progression with profound consequences.) By 2000 these modems were overhauling our copper-wire infrastructure without stringing a lot of new wire. By 2010 most of the country had been covered in one way or another. This would have been inconceivable without DSL and Cable Modems to repurpose ancient copper communications lines. (To rub it in, the foundation technologies enabling both DSL and Cable modems arose largely from Bell Labs, with DSL itself an invention of BellCore, a fragment of Bell Labs left to the regional Bell operating companies. However, DSL was standardized around technology developed by a tiny start-up out of Stanford University, which technology is now central to Cable Modems and 4G/5G mobile transmission. So the World Turns.)
The Speed Bump and Video
Unfortunately, our tastes went from text to video. Video is now 75% of Internet traffic. Even in compressed form video takes way more bandwidth than text or image transmission. Adding pressure was the widespread dissemination of WiFi routers in homes, attached to their modems, which enable up to eight devices to share a single line. (New ones have much more capacity.) Early data rates, measured in low megabits per second, suddenly became more than limited. In urban areas telephone companies adapted to the change by radically shortening the copper line with fiber to the node networks, realizing data rates with V.Fast DSL up to 50 mbps, or building out new fiber optic networks. Cable television networks did a similar trick with its modems and network, now able to realize downstream data rates up to 400 mbps in some markets.
When Analogies Fail
We opened our story by comparing broadband to the early days of electricity. But as with all analogies, this one breaks down at places. Electricity began at 120 volts and 60 Hz (the voltage actually crept up a few percentage points over time) and stayed there. The power we got over the lines installed in 1919 could be the same today if the wire held up. Broadband will not be so lazy. The FCC definition of broadband has grown from 200 kbps in 1996 to 4 mbps in 2010 to 25 mbps in 2015. It will keep growing. The pace creates two serious threats: (1) the copper networks currently supplying broadband to at least 80% of the American population will not have sufficient capacity to keep up; and (2) many regions of the country, mostly in rural areas, will be left behind as the cost of upgrades and in the end an entirely new network bringing fiber optics rather than copper lines to the home will be left behind for a considerable period of time under market conditions laid out in the 1996 Act.
Where the Act Fails Us
The 1996 Telecommunications Act recognizes the inherent conflict between fully competitive infrastructure markets and universal service. For example, the Act authorizes public ownership of radio to insure adequate access to information resources supplied over the air (and hence it formally authorizes NPR and PBS). But it makes no such gesture relative to broadband service. As noted above, Section 254 provides a fund to subsidize underserved communities, but its sums are miniscule relative to the need for broadband infrastructure, and declining as wire-line telephone connections decline. The Act’s only suggested mechanism to encourage universal access to “broadband” is the removal of barriers to competition and investment. In short, the Act sees the need but fails to see the speed with which the need would materialize, and hence leaves us with what we have today, a country divided between those with adequate broadband service and those without.
Aren’t We Saved by Mobile Networks?
So far we have ignored the amazing deployment of mobile telephone networks. While the basic hand-off technology for slipping from one mobile cell to another without call interruption was sketched out at Bell Labs in 1947, mobile phones awaited other developments to mature into a white-hot market. Moore’s law again contributes its share to the story (in 1980 an integrated circuit had a maximum transistor count around 10,000. Now it is almost 20 billion, and every year Moore’s law has been declared dead it breaths again). But the 1996 Act gets some credit. The major carriers invested billions soon thereafter in fiber optic lines connecting all their offices together, creating within the Public Switched Telephone Network and the Internet an enormous (and still not fully used) internal capacity in what we call the “back-haul network.” Towers sprouted up like weeds in spring and, voila, cell phone service became almost ubiquitous in a very short span of time. Then the iPhone hit the market in 2007, and never looked back.
So why don’t we just connect everything up to mobile networks, eliminating the need for all those wires pushed through walls in our houses? The simple answer combines capacity and costs, the former insufficient, the latter too much to bear. The full answer is more complex (they always are it seems), but the simple answer will be enough here. Today cell phones account for 50% of Internet traffic. But 75% of cell phone traffic goes through a home or office WiFi network with wire-line connections to the outer world. Therefore in the order of 85% of all Internet traffic today relies upon copper or fiber optic lines into premises, not mobile networks. Yet we know that cell phone data rates collapse often because antennas congest with more than a certain number of simultaneous users.
The only way the mobile networks can increase capacity for existing cell phones is through “densification,” meaning increasing the number of antennas serving a given area; everything else (higher frequencies, beam forming, carrier aggregation) requires new cell phones and new antennas. Verizon and AT&T have densification programs around small cell antennas in several areas of the country (including a handful of pole-mounted small cell antennas in Connecticut), but the number of such antennas required to cover the country number in the hundreds of thousands, all requiring new fiber optic wire to connect them back to a mobile Core Network. Meanwhile the cell phone market has saturated, meaning revenue growth has subsided to what might be expected in a mature market. Thus a rapid build of mobile networks just to catch up with existing demand probably offends the business calculus, too much capital cost for too little incremental revenue.
We should also mention that small cell antennas are not exactly arriving in communities with parades. They are ugly, and they radiate RF energy close to homes. Some communities have already outlawed them. Even those who are just suspicious have long procedures to obtain site approval. As we observe in our section on 5G, mobile networks are not the answer.
The Problem Facing Existing Carriers
Our country’s telecommunications infrastructure was built almost entirely by private businesses. The 1996 Act promotes the same solution for the dual problem of advanced communications and universal service. But when the Act moved away from internal subsidies created by legally charging some customers a lot so those areas losing money were compensated, it threw the baby out with the bathwater. Now every capital cost within carriers today must be justified on its own power to return revenue in some suitable proportion to produce a profit. Markets for telephone, television, and the Internet are now largely saturated. Should competition increase (the hope of the Act) prices will go down (the hope of the Act), reducing the territories carriers can afford to invest in when markets are mature (against the hope of the Act). When around 2005 Verizon and AT&T decided to attack the broadband market seriously, with different technologies, they banked on their television package to draw enough incremental revenues from CATV companies to justify the costs. It did not work out. Verizon stopped deploying its FIOS network in new territories in 2011, and the AT&T U-Verse program never reached anything like serious coverage (it is still not available in the wealthiest sections of San Francisco). From their point of view they cannot justify new networks when there is no incremental revenue, and in some cases a reduction of revenue, over what they obtain from the existing network. They now pour their billions of dollars in investment each year into mobile networks. The same is true for CATV companies, not in the sense that they divert capital to mobile networks, but in the sense they cannot see any incremental revenues from significant upgrades to its network, or expanding coverage to regions that were not profitable before.
When Analogies Succeed
The story of Jewett and its initiative to build its own electric plant rather than wait for private electric companies to wire its small circumference is now playing itself out for broadband around America. More than 100 small communities have installed their own fiber optic networks and manage the same today. Many more are in planning or construction stages. Some larger communities such as Chattanooga, Tennessee, with a massive municipal electric utility, have fiberized their city to great benefit. Many of the small networks had state subsidies. Indeed, the conditions for almost all municipal fiber optic networks today are the existence of a prior electric utility that already knows about poles and wiring and/or state subsidies.
Municipal Networks are Now Mandatory
In Connecticut we are impoverished on both ends of the broadband conundrum. We do not have universal service, and the absence of real competition means no carrier is moving to the next generation network, fiber optics to the home. To date no municipality has seen fit to follow Jewett to the broadband altar. Efforts by Manchester, New Haven, West Harford, and the collection of rural communities represented by Northwest Connect have yet to reach a deal that looks like a real network. Obstacles center around community commitments to spend the money and various legal impediments fostered by Connecticut law and opposition from our Public Utilities Regulatory Authority (PURA). The battle forward is not technical, it is financial, legal, and in the end community willingness to pay something for the future of the state.