The Tech - Online EditionMIT's oldest and largest
newspaper & the first
newspaper published
on the web
Boston Weather: 36.0°F | Partly Cloudy
Article Tools

Suppose the Federal Communications Commission (FCC) stood up one day and declared, under a law written 76 years ago, it has the authority to regulate the delivery of newspapers. Shortly after making this proclamation, the FCC chairman announces he has decided how newspaper delivery will be regulated — rather than letting the use of our finite news transportation supply be prioritized by competitive, free market bidding, he will instead institute a one-size-fits-all rule: Delivery companies must deliver the news in the order it is received and charge all customers an equal price. The chairman calls his model “Newspaper Boy Neutrality.”

Newspaper Boy Neutrality is not without downsides. If some time-insensitive publication (say, a monthly magazine), is placed in the queue just before a time-sensitive one (say, the early edition of a daily newspaper), neutrality demands that the time-insensitive publication be delivered first. The daily news producers are unable to pay extra for the on-time delivery that their customers want, and the monthlies are forced to buy a level of priority delivery that does not add value to their customers. At the margins, news creators are shut down, forced out by the market inefficiency created by the neutral protocols. The loss of their demand means a smaller market for the delivery companies (as well as reduced investment in new bikes). Innovation is stifled as companies serving those niches of the news market not well-matched to a neutral protocol — up-to-the-minute news and high-quality, time-insensitive news analysis — suffer under the system’s inflexibility.

The outcome is ugly, but necessary, the neutrality advocates argue — unless we forbid delivery companies from offering contracts that prioritize one content source over another, how could we be sure that delivery companies wouldn’t refuse to transport the news of one political wing, or create their own newspaper company and give it preferential treatment? We cannot risk such a threat to our free speech. And why stop with News Boy Neutrality? Why limit the argument to just media — why not Truck Neutrality? Or how about Rail Neutrality? Nothing smells of “fairness” quite like a train car of vegetables rotting at the station while loads of scrap metal are being whisked along the tracks.

If “neutrality” sounds like something a fat person in an Ayn Rand novel would think of, then brace yourself — the rationale behind net neutrality is not much different. The distribution system is finite. There is content that requires priority delivery — streaming video or VOIP. There is content that doesn’t — that 10GB torrent, or an iPhone app download. And finally there are a group of people who claim free markets are just a licence for corporations to restrict your choice and that a new economic system is needed.

Of course, few net neutrality advocates claim that their logic should be extended to trucks or trains — we are told that the net is different, and thus must play by different rules as the rest of our economy. But most of the arguments as to why the net is different are hokum — goods are goods and content is content. Appeals to free speech, apple pie, and motherhood notwithstanding, it doesn’t matter whether they’re being delivered by a kid on a bike or through a fiber-optic cable.

There is one exception that deserves mention: Telecommunications companies are natural monopolies. It is impractical to run duplicate cables, and as a result, there is little opportunity for the sort of competition that keeps newspaper boys from forcing the New York Times to pay higher delivery rates than the New York Post. There is a legitimate need to regulate — without protections in place, rapacious monopolies would squeeze economic rents from their customers.

Viewed in this light, net neutrality makes some modicum of sense. When all customers in the system pay the same rates, it is difficult for a monopoly to extract rents without raising the eyebrows of an oversight body. Even the densest of regulators could simply estimate the cost of the network, divide it by the number of customers, and have a decent idea of whether a provider’s rates are fair or price-gouging. But when a monopolist begins to offer differentiated services, the regulator’s task becomes much more difficult. Not only does the increasing variety of a monopolist’s operations make it difficult to estimate their costs, but even if the regulator obtains an accurate picture of a monopolist’s total costs, there is still the potential for extracting rents by forcing one section of customers to cross-subsidize another — in the net neutrality context, this might appear as some subsidiary of Comcast receiving specialized treatment on the network, and paying for it by gouging everyone else.

Despite the anti-trust problem that tiered services present, there is little experience to suggest that regulators would be unable to oversee this new system. Fuel transportation networks are a natural monopoly, and they allow for priority delivery just fine. In fact, MIT has a contract that gives its provider the right to cut-off its heating fuel supply during peak periods in exchange for lower prices; where are the screaming neutrality advocates ready to claim that monopoly abuses are right around the corner? Electricity transmission networks are another natural monopoly, and yet, most net neutrality advocates have nothing but positive things to say about the “Smart Grid.” But what is the smart grid besides a way of throttling consumer electricity use during periods of congestion and charging users different rates depending on their time of use? Why would we adulate the coming of a smart grid in the electricity sector, and still mawkishly cling to a dumb grid in telecommunications?

For a long while, we have benefitted from a past over-investment in network infrastructure. Back during the dot-com boom, many companies, falsely predicting that internet traffic was going to double every three months, laid a lot of pipe in the ground. They went bankrupt, but their investments were auctioned off and form a large portion of the network we use today.

Unfortunately, the one-time gift from the bankrupted bandwidth backers is not enough to fuel the internet forever. Internet traffic may not be doubling every three months, but it is still doubling every nine months, and this growth may accelerate as we see wider adoption of online video and peer-to-peer applications. The future will not be, as some net neutrality advocates believe, one of limitless bandwidth forever.

This poses a conundrum for the Internet Service Providers (ISPs), who own the pipe and find themselves brushing against the ceiling of their aging bandwidth capabilities. On consumer networks, 0.5 percent of subscribers account for 40 percent of total traffic, while 80 percent of users use less than 10 percent. At the present, this disparity in usage of network resources does not, by and large, translate into different prices faced by these two groups.

Younger, wealthier, and better educated users are being subsidized by the old, poor, and less educated. As our holdover pipe from the dot-com bust gets utilized, and bandwidth becomes scarcer, the extent of that subsidy will increase. The ISPs, noting that they will lose customers if they raise rates uniformly, would like to pass the cost of revitalizing our network infrastructure onto those who are burdening it the most. The internet adepts, seeing the end of their free ride, now rail against the “un-neutrality” of this proposal, and seek to make it illegal for ISPs to perform such price-discrimination.

In the long term, both pay-as-you-go (per-kb) and priority pricing are necessary, both to offer low-cost internet to the poor, as well as to enable applications that require differentiated traffic.

We are told that differentiated traffic is the enemy. The common horror story given by Net Neutrality advocates is one in which two nearly identical services (say, Google and Bing) are treated differently by ISPs, and as a result, one withers and dies while the other ascends to monopoly status.

A chilling tale. But it is akin to supposing that under a smart electricity grid, Pacific Gas and Electricity would be able to cut off power to Google during periods of congestion. It is simply not something that could happen under any faintly competent regulator.

Closer to reality is the idea of Google voluntarily choosing to purchase a cheaper, lower priority tier relative to Bing, and the two competing by offering differentiated products — Google: slower, but with fewer ads and a higher quality search function, and Bing: faster, but with an otherwise inferior product. This diversity of offerings is not a horror story — it is one of the successful hallmarks of free market systems.

Tiered services work, and work well. They allow providers to better tailor services to customer needs, and bring the price of services in line with the cost of supplying them. With proper oversight to prevent monopolistic abuses, pricing innovation will improve the fairness of the system and ensure that future investments in network infrastructure are made optimally. Tiers are an encouragement not a hurdle — to innovation, and will better allow end-use consumers to decide, through the free market, what they want their internet experience to be.

The time has come for a smart telecommunications grid.