The Open Internet Order

After a year of hearings and considerable public discussion, the Federal Communications Commission (FCC) adopted the Open Internet Order on December 21, 2010.

Fireworks flared on the blogosphere almost immediately. Net neutrality advocates cried that the order betrayed and sold-out sacred principles, while Tea Party supporters heaped scornful criticism at government activism. Both sides made intemperate and grim forecasts about the Internet’s future.

Levelheadedness left the political sphere as well. Pushed hard by Tea Party sympathizers, the House of Representatives passed House Joint Resolution 37 in April 2011, largely along party lines, disapproving of the order. As of this writing, the Senate hasn’t yet taken up the measure. President Obama promises to veto it.

Frankly, this conversation needs a calm and considered middle ground, not utopian visions abutting practical considerations. The Internet has never lacked government oversight, and Internet participants have occasionally compromised on neutrality to function. There are subtle economic issues to debate here, and simplistic absolutes don’t contribute much to finding reasonable economic solutions.

What does the order do?

The order begins with a discussion about dominance. It’s the premise for everything else. Like any policy in the spirit of common-carrier and antitrust principles, the order seeks to limit distortions from the presence of market power in retail access markets.

The issue is nothing new for communications markets, nor is it a surprise for wireline broadband markets. The National Broadband Plan discussed concentrated supply in its March 2010 report, and many witnesses at the hearings leading to the order also pointed to concentration as a key concern.

Many US wireline broadband markets have a strong potential for sustaining concentration caused by high barriers to entry. In addition to high capital requirements, barriers to new entry arise from limited rights of way, limited backhaul, and intermediate mile providers in many locations, and franchise restrictions from local authorities. Additionally, wireless broadband is still a weak substitute in most markets, partly because of limited spectrum.

Despite that premise, the order adds a twist. It carves out a major exception for mobile wireless providers, largely on the grounds that the wireless market is more competitive. (Needless to say, the recent proposal to merge AT&T wireless and T-Mobile raises questions about the merits of this carve-out.)

As many commentators have pointed out, this carve-out looks arbitrary in practice. After all, there’s little difference between a wireless ISP sending data to an iPad and a 4G cellular phone carrier moving the data to the same device. Is there really that much difference between fixed and mobile wireless to merit a carve-out?

While I sympathize with complaints about inconsistencies, for the time being they have no practical importance. Every wireless ISP or fixed wireless business falls into a competitive market. Most are so small that none would fit any sensible definition of being dominant.Only a crazy regulator would ever apply this order to them. In other words, wireless ISPs (WISPs) should be able to go about their business unheeded.

Management practices

Now consider the three issues the order covers: transparency, blocking, and discrimination.

In professional circles, there’s no disagreement about the importance of transparency. Broadly speaking, transparency encompasses three activities: timely information about carrier processes and interoperability, timely notification about when those processes change, and general procedures for forecasting, allowing others to anticipate likely changes in the future.

The order codified the first two notions of transparency, and, frankly, it’s just a codification of something firms already do. The only practice that requires professional discussion is making forecasts about changes, which is costly to do well and has legal ramifications if done poorly. As far as I can tell, discussing this minor issue will keep a few lawyers employed, but will make little difference to life.

The order’s teeth lie in the provisions to prevent blocking and discrimination. Even there, however, there’s not much bite, because dominant broadband firms today largely don’t do either. At most, the order prevents dominant carriers from planning to block content or discriminate against firms in the future.

Why forbid such plans? The order seeks to preserve some of the best features of the present situation. Practically speaking, the Internet has never been perfectly neutral, but no carrier has ever blocked or discriminated in spades.

Why might that change? Dominance makes these actions more likely. Moreover, if these practices become common, the potential damage could be great.

A broadband firm with retail market power would have incentive to put any content firm at a disadvantage when that content firm’s activities competed with services it offered. Voice and video entertainment are, for example, services offered over broadband channels, and most sensible observers expect the latter category of services to grow over time. These would be natural targets for blocking and discriminatory routing.

In addition, if not allowed, discriminatory routing would become a matter for negotiations. Broadband firms would try to negotiate with content firms for different priorities, extracting extra fees for the fast lane. In short, broadband firms would use blocking and discrimination to become gatekeepers.

Such behavior raises the possibility that content firms won’t be able to easily reach customers, and that entrepreneurial content firms wouldn’t able to explore new services and build a viable business, unhindered by gatekeepers. That could hamper the Internet entrepreneurial engine that has blossomed in the last 15 years.

In summary, this order mandates a regulatory action that differs little from present practice at broadband firms. The order has little impact because most broadband firms—even the dominant ones—are transparent today. In addition, most don’t block traffic and don’t use discriminatory routing practices. At most, the order preserves the status quo.

So, why all the fuss? After settling a contentious year-long process, the status quo matters. Most cities in the US finished the first wave of building broadband. Only a fraction of homes in low-density areas still experience any significant build-out today. In short, by leaving just about every existing practice in place, the order didn’t alter the economic incentives to make upgrades.

Open questions

Many observers have looked at this situation and made a wishful thought: Why can’t competitive markets discipline wireline broadband firms?

Consider content markets, for example. While Google has built a good business today, it’s both a creature of, and threatened by, competitive innovation. After all, Google was started a dozen years ago by two dropouts from a computer science PhD program. Earlier in the decade, Google replaced Yahoo as the most popular website.

Recall how fast and recent that was.

Moreover, both Google and Yahoo continue to innovate out of fear from competition. Today that threat comes from Facebook, a firm started by a Harvard undergraduate seven years ago, and which climbed into the number two spot last year. Today, Facebook is taking advertising dollars from Google and Yahoo, placing pressure on both to innovate further.

Why can’t the US get that competitive innovation in wireline markets? I wish it were so, but the barriers to entry in broadband do bind.

A more thoughtful objection to the rules forbidding blocking and discrimination seeks to preserve a carrier’s ability to implement “reasonable network management practices”—a euphemistic phrase that stands in for multiple issues.

For example, most professionals believe networks with severe capacity issues should be allowed to experiment widely with prioritization of traffic to enable services for most users. Some of this is just common sense, like letting broadband firms find ways to get teenagers using BitTorrent late at night when nobody else is online. Some of it is hard to figure out, as innovation lets engineers find new ways to route data more efficiently.

The lack of a settled definitional boundary between reasonable network management and discriminatory routing is no surprise. No order can settle every issue, and the Internet is complex. Technological change will bring new questions.

That open-endedness worries many observers, myself included. Trying to settle the boundary holds the potential to tie the FCC in Gordian legal knots, leading the US Internet into a dystopian regulatory future. It also could lead the US down a path of regulatory ping-pong, with issues bouncing between courts, regulators, and Congress, never settling anywhere. Nobody wants to see that future.

Nonetheless, that’s no reason to dismiss the order. Rather, it’s reason to rethink the regulatory framework in which the Internet gets supervised, bringing it into the new era and away from the stale regulatory structure that developed for cable and telephony. In other words, the FCC might be a natural national regulator for the Internet, but it should be restructured for the job.

The future of broadband’s regulatory structure deserves a serious discussion, but HR 37 doesn’t encourage that. It’s important to police anticompetitive actions to enable the competitive market to thrive, but HR 37 disapproves of the entire order. It gets rid of the police, the supervisor, and the policy. How is that a fruitful starting point for discussion?

Perhaps HR 37’s worst feature is its lack of nuance. The order includes the provisions on transparency, as well as some for behavior that any sensible antitrust prosecutor or regulator would focus on. HR 37 disapproves of everything, throwing out several babies with the bathwater.

Copyright held by IEEE. To view the original essay click here.