Here’s a question: What is the Internet, exactly? A cloud? A grid? A highway for information, possibly super? A series of tubes? Canadians have a metaphor problem when it comes to the web, and although this might seem like a rarefied concern it has real consequences for the amount we shell out for access.
Over the past several months, an impassioned debate has erupted about how—and how much—we pay for Internet. Most Canadians buy their service from big providers like Bell and Rogers, which charge them partly by the amount of data they upload and download; this is called usage-based billing. To encourage competition, the Canadian government forces these big players to share their digital infrastructure with smaller companies, like TekSavvy and Acanac, who rent it from the major ISPs at flat rates and cheerfully pass the benefits along to customers in the form of all-you-can-download packages. Last year, however, the CRTC ruled that Bell and Rogers could slap small resellers with usage-based billing fees, too—depriving them, and their customers, of their unlimited downloading advantage.
Some of these users, who make up just a sliver of the market, screamed blue murder. The more temperate among them claimed that Canadians pay too much for data, and pointed out that Canada risks becoming one of the only countries where usage-based billing is the consumer’s only choice. This alarmed the general public, which had previously accepted usage-based billing without complaint—perhaps because it makes intuitive sense, or perhaps because most didn’t realize there were other options. Suddenly, Internet users were asking—and loudly—Is this fair? A new round of CRTC hearings on this precise question is scheduled for July, guaranteeing a fresh raft of media coverage.
A ubiquitous term’s forgotten—and unsettling—origins
Since the 1870s, “pork-barreller” has stood out as one of the most damning political epithets, indicating corruption and self-interest. Strictly defined, though, pork-barrel spending is perfectly standard. It refers to money allocated for projects in a particular electoral district, usually proposed by its member of Congress: dams, space programs, water filtration plants, national parks. The negative associations come from the phrase’s nineteenth-century origins in the American South, which are rather disturbing. “Oftentimes the eagerness of the slaves would result in a rush upon the pork barrel,” wrote journalist C.C. Maxey in the National Municipal Review in 1919. “Members of Congress in the stampede to get their local appropriation items into the omnibus river and harbor bills behaved so much like Negro slaves rushing the pork barrel, that these bills were facetiously styled ‘pork barrel’ bills and the system has thus become known as the pork barrel system.”
When we start to consider what a fair way of paying for Internet might look like, we run up against a problem: most of us don’t know much about how the Internet actually works. The battle over billing is one of metaphors, because what we believe to be a fair payment scheme varies according to what we perceive the Internet to be. If, for instance, we imagine it to be something like a natural gas utility, usage-based pricing makes sense. (CRTC chair Konrad von Finckenstein compared the Internet to a public utility before Parliament.) But if we imagine it to be something like a highway, or the hydro grid, or a cloud, or a big truck, then the picture shifts.
Unlike with most public utilities, Internet providers don’t deliver something with a price. Natural gas, for instance, is a finite substance and costs money to extract. Electricity doesn’t generate itself for free. Water needs to be treated, and the more of it you treat the higher the expense. This isn’t the case with data, which is nothing but a signal: a cat video streamed from YouTube does not deplete some store of cat videos (fortunately for humanity, cat videos are infinite). On the Internet, the issue isn’t how much data is transferred, but how many people are trying to transfer it at once. That’s how network congestion, the bane of the Internet, happens.
This helps explain why the Internet is more often likened to a highway system, with its attendant traffic jams. A highway system has smaller roads that feed into bigger roads, which link to make a network. On the Internet, little bits of data independently navigate the roads toward their destinations, and get stuck in bottlenecks when too many try to pass through the same point. But the metaphor breaks down, because on real highways cars move at different speeds; there’s always the jerk in the BMW weaving in and out of the passing lane (maybe, gentle reader, it’s you).
Not so online, where packets of information all travel at the same lightning speed; the only question is how many can squeeze through the network at a time. What we casually call the “speed” of an Internet connection is really how many packets of information can get through at once, not how fast they’re going (with one key exception: sometimes packets are deliberately held back, or “throttled,” by ISPs, but that’s an artificial speed limit). In that sense, the Internet is more like a network of pipes than a highway. Nobody talks about the “speed” of water in a pipe; what matters is the pipe’s diameter. And here’s the dirty secret of Internet metaphors: the Internet, when it comes down to it, is a series of tubes. The notion was popularized by Ted Stevens, the late pork-barrelling senator from Alaska, in a floundering attempt to explain the Internet to Congress. He never lived it down, but the description was actually pretty apt.
Canadians can draw a few lessons from this as we ponder the way we’re being billed. First, most of the cost of providing Internet service is sunk into infrastructure, as well as administrative functions such as customer support, maintenance, and billing (although those who have had the pleasure of dealing with Bell or Rogers in this regard might well question where their money goes). These expenditures may be considerable, but once a system is in place there’s very little incremental cost to transmitting gigabytes of data. To put it another way, even if you opened the digital spigots and poured forth a flood of cat videos, you wouldn’t be running up significant bills somewhere in the bowels of your service provider—costs it might then legitimately pass on to you.
Usage-based billing doesn’t reflect the real cost of pushing data through pipes. Rather, it’s a crude cudgel meant to keep too many users from trying to share the pipes at once. Those who pay extra for more data aren’t subsidizing the cost of transmitting that data, because it cost so little in the first place; they’re just getting whacked with a financial penalty for tying up the circuits.
This can’t be the best way to run an Internet. Other service providers, in other countries, do it differently, using clever switching techniques to better balance data demands, and reducing the flow of data to users who download excessively.
The Internet needs to be paid for somehow; there’s a cost to laying bigger pipes. But the more we think of data as a commodity, and the Internet as a utility that delivers it to us, the more we’ll end up paying. Metaphors matter.
This appeared in the July/August 2011 issue.