July 18, 2017

Net Neutrality Supporters Should Actually Hate the Regulations They're Endorsing

Andrea O'Sullivan

Feature Writer

If you went on the internet at all last week, you could not help but miss some of the web's most popular websites publicizing their campaigns that defend the Obama-era telecommunications regulation known as the Open Internet Order (OIO). Last Wednesday, tech heavyweights like Google, Facebook, Twitter, Reddit, and even Pornhub held a "Day of Action" to support the controversial FCC rules. The websites bombarded users with blog posts encouraging folks to contact their representatives and popup messages bemoaning the future of a slow and tiered internet. But ironically, these websites' stated goals are in direct contradiction of the regulations that they ostensibly support.

Simply stated, the OIO does not in fact secure the principles of "net neutrality" like so many of these websites implied to their users. In fact, the OIO may have the adverse effect of actively discouraging the principles of net neutrality through a loophole that would exempt motivated Internet Service Providers (ISPs) from OIO regulations. This cannot be emphasized enough: The OIO allows and encourages ISP filtering, a huge no-no in the world of net neutrality.

This is a point that my Mercatus Center colleague Brent Skorup has made since shortly after the OIO rules were first introduced in February of 2015. It's a bit of a nuanced argument, and one that would not be immediately obvious to anyone who does not closely read all FCC reports and related court cases as a profession. But as the general public is whipped into a veritable frenzy to defend the OIO rules or risk Internet catastrophe, it's a critical fact to hammer home in the debate.

To understand just how muddled the discussion surrounding "net neutrality" and the OIO has become, we need to know a bit about: 1) how the concept of net neutrality developed and what it means; and 2) the political pressures and compromises that were made in the run-up to the introduction of the OIO.

First, the definition of "net neutrality" is incredibly hazy. You could almost say that more people agree that net neutrality is a good thing than can agree on any particular definition. The concept was first laid out by law professor Tim Wu in his seminal 2002 article, "A Proposal for Network Neutrality." Wu lays out hypothetical scenarios where ISPs block or throttle access to content for reasons ranging from cost to anti-competitive activities. His article attempts to distinguish content differentiation that he finds reasonable and should be allowed from those that he finds unjustifiable and should be prohibited. The article generated a fair bit of controversy even under this more limited framework—critics responded by pointing out some benefits of non-neutral Internet arrangements—but it was at least a relatively narrow and understood topic.

From there, the concept of "net neutrality" morphed into something that was both utopian and unworkable. If you type the phrase into Google, the top definition provided is the "principle that Internet service providers should enable access to all content and applications regardless of the source, and without favoring or blocking particular products or websites." Yet this definition stands in sharp contradiction to the vision outlined by Wu, who noted that "a total ban on network discrimination, of course, would be counterproductive." This kind of extreme understanding of net neutrality has been dismissed by early Internet pioneer and MIT computer scientist David Clark as a "happy little bunny rabbit dream" that would be both impossible and undesirable to implement.

Unfortunately, the unhinged understanding of "net neutrality" has since won the day. And it has fueled average people's nightmares about what the future of the Internet holds—even though it looks a lot like what we've always enjoyed. (After all, the OIO regulations were only proposed in 2015.)

The core of these fears is a future where consumers are forced to purchase tiered package for Internet access of specific websites, or face slow service or even a complete blackout. Consider this image macro, which is popular on sites like Reddit and Twitter. The meme evokes a world where ISPs would be free to filter traffic through various "fast" and "slow" lanes that could be accessed by special fees.

Of course, we did not see ISPs engage in this kind of behavior in the pre-OIO history of the internet—which is to say, for the entire history of the internet. But ironically, the OIO actually creates incentives for ISPs to do just that.

How? The original OIO language attempted to deal with any First Amendment objections to the new FCC regulations by differentiating between platforms that engage in "editorial intervention" versus those that are mere conduits for speech, such as most ISPs. But this created a problem because some ISPs do engage in editorial curation—specifically, religion-oriented ISPs that filter out content objectionable to certain faiths. In 2016, the US Court of Appeals crystallized this loophole in its interpretation upholding the OIO, which noted that ISPs that explicitly offer "'edited' services" to its customers would not become subject to the OIO.

In other words: Tell your customers that you filter content, and you're free from the shackles of the FCC!

Which brings us to our second point: The OIO was not based on years of measured research and debate, but was a politically-expedient compromise to rally the Democratic base. Skorup notes that former FCC Chairman Tom Wheeler and his staff explicitly did not favor the Title II approach of the OIO until President Obama exerted pressure on him to do so in 2014. According to Gene Kimmelman of the progressive advocacy group Public Knowledge, the move was an effort to secure a "clean victory for us to show that we are standing up for [the] principles" of the Democratic Party.

The result was a half-baked and rushed OIO that contained a glaring loophole which inadvertently enshrined in law what net neutrality advocates had been fearing the whole time. It's a remarkable outcome, and one that even OIO hardliners now bemoan.

(Although some of the more Machiavellian net neutrality advocates know about these loopholes but simply don't care. Instead of a narrow grant of FCC power, they got a broad swath of Internet regulations that can be gamed to benefit certain parties and inject government into even larger social domains.)

This comedy of errors has resulted in a situation where the OIO is untenable for both true net neutrality advocates and net neutrality skeptics like myself. On the one hand, as discussed, the OIO explicitly violates the principles that many net neutrality activists hold dear. On the other hand, the OIO erects a major new permissioned regime that threatens to stifle internet innovation and favor certain firms and industries over others. In particular, the OIO employs a vague "general conduct standard" that could prompt regulators to clamp down on any online activity that they don't like. This kind of catch-all "standard to rule them all" could further frustrate the stated aims of net neutrality advocates.

So with the OIO, we find ourselves in a rare scenario where diametrically opposed parties to a narrow policy question can beneficially work together to destroy a regulation that accomplishes neither of our aims.

Unfortunately, the emotional inertia surrounding the hot potato of pure net neutrality renders many people's rhetorical shields frustratingly difficult to penetrate. But those of both sides of the issue who understand the flawed premises of the OIO should shout this fundamental error from the rooftops. After this bad policy is gone for good, we can get back to fighting over the core issues that really matter.