What BITAG (and MCDowell) Can Learn From ICANN.

Unsurprisingly, Commissioner McDowell’s dissent to starting an inquiry into the FCC’s broadband authority contained a reference to the new “Broadband Internet Technology Advisory Group” (BITAG). For those who missed the press release, BITAG consist of engineers from a variety of broadband industry segments chaired by Dale Hatfield at University of Colorado at Boulder. Given that BITAG has a bunch of smart folks — especially Dale, who is one of the smartest and nicest guys in this field — and a reasonable cross-industry representation, it may actually come up with some interesting stuff (hence the cautious endorsement from my employer here). Predictably, however, a chorus of Libertarian True Believers, and the “government-can-do-no-right (but hurry up with my subsidy check!)” crowd argue that this represents “industry self-regulation” and that FCC should therefore hold off on doing anything for the foreseeable future. Despite explicit statements from some participants that this does not in any way, shape, or form replace the need for FCC oversight, it does not surprise me to see McDowell again championing the idea that private sector regulation through BITAG and similar institutions provides a better way to protect consumers and encourage innovation and investment than actual, enforceable rules.

I snidely tweeted at the time “One ICANN Mtg will cure McDowell of his love of self-regulation.” For those not familiar with the Internet Corporation for Assigned Names and Numbers (ICANN), it formed back in 1998 when the Clinton Commerce Department avowedly wanted to “privatize” management of the domain while simultaneously making sure their buddies in the trademark lobby got protected. ICANN, now debates such “technical” matters as the appropriate level of cross-ownership for new generic Top Level Domains (gTLDs), has a sprawling bureaucracy, a budget of more than $60 million (US), and generally makes the FCC look like greased lightning when it comes to actually getting stuff done.

ICANN provides a lot of good lessons for the BITAG in what to avoid and a caution about trying to make technical standard setting the equivalent of regulation . . . .

I have, in the past, compared ICANN to a bad sitcom that long ago “jumped the shark”. As one present at its founding, let me share a few reasons why it failed so colossally in its promise. I can distill it down to this:

1. If policy people show up, give up and go home. In other words, I must never join the BITAG (unless I get a relevant technical degree).

2. If intellectual property policy people show up, the show is really over. Run for the hills as fast as you can and pray you escape the grim embrace of those I only half-jokingly refer to as the “IP mafia.”

The reason for this is simple. You can do technical stuff, or you can do policy stuff. But you can’t do both (which is why technical organizations like this aren’t a substitute for bodies that actually set policy and rules). ICANN tried, and failed. ICANN was charged to act as a technical coordinating body with absolutely no rulemaking and no business but voluntary coordination and stability of the naming system. With one tiny exception. As part of determining how to add new gTLDs, ICANN was also supposed to consider trademark issues. ICANN was also explicitly charged with developing a “Uniform Dispute Resolution Policy” so that trademark holders could have a quick and easy way to resolve complaints about “cybersquatting” (registering someone else’s trademark for your domain name for nefarious purposes).

And that was pretty much it for ICANN as a technical body. Because once it had to take a non-technical issue like trademark into account, it needed to consider how it would implement this directive. And if it were going to consider trademarks, it needed to consider how policies to protect trademark would impose cost on non-infringing parties, or block free speech, or stifle or promote competition in the name space. This, in turn, attracted a cottage industry of advocates, for various interests. Somehow, ICANN also got involved in other forms of “competition,” which meant developing a split between TLD registries and second level name “registrars,” which created an entire industry centered around ICANN as its chief regulator. Since this also required revenue to support, ICANN started levying various fees. Since the Internet is global in nature, ICANN needed to hold meetings around the world. Since it was making important decisions about global resources, governments needed to be represented. And on, and on, and on.

I will add that no one in ICANN (except for a few cynics like myself others in the “non-commercial constituency”) believed for a minute this would happen. ICANN supporters believed that because ICANN was NOT GOVERNMENT and PRIVATE SECTOR and TECHNICAL and run by people with engineering and network backgrounds that it absolutely positively MUST work out better than any dumb old regulatory government thingy. (I, on the other hand, coined the phrase ‘ICANN recapitulates the FCC — but does it badly’ sometime back in 1999 and haven’t been proved wrong yet.)

It’s not that the people involved in ICANN were bad, or evil, or greedy, or stupid. Sure, there were some folks involved who were bad, and greedy, and stupid, but no more so than in any other human endeavor. Which is the point that worshipers of industry self-regulation keep missing. The same public choice theories about how government officials make decisions on factors that have nothing to do with the public interest apply equally well to industry organizations and other “industry self-regulatory bodies.” If the subject of the discussion is technical and of little interest to people, then the experts present can relax and talk expert stuff. But once it becomes a question that involves significant policy stuff — like protecting consumer rights, or restricting the flexibility of network operators, or whether or not network operators can block or degrade certain types of content — people act different. Engineers are as conscious of who signs their paycheck and the viability of their future career as any other human being, and will be influenced by the positions of their employers or hoped for future employers just like government officials.

More importantly, most of the questions around broadband policy are not inherently technical ones. As the good folks up in Canada demonstrated when they adopted rules on traffic shaping along the lines many of us want to see here in the U.S., setting rules is a policy issue that requires looking at all sorts of trade offs and possible injuries that have nothing to do with technology questions. Rules will impact the development of technology standards, absolutely. Rules will make certain behaviors easier, certain behaviors harder, and will influence how engineers design systems.

In other words, technical standards are best developed in response to a specific policy decision or desired goal, not the other way around. Organizations like IEEE succeed (where they do) because people get together and say “hey, we’d like to do X, where X is some specific thing. How can we do that?” For example, the IEEE standard setting could not have said “let’s establish a rule that you can operate at 1-watt EIRP at 2.4 GHz so that we can create WiFi.” Instead, once the FCC established the Part 15 rules, a bunch of companies said, “hey, we all agree we it would be nice to find a way to set up local area and wide area networks wirelessly. The FCC has opened up some spectrum for “unlicensed” use. Lets develop some standards for how to do that within the rules the FCC set up.”

Had IEEE been in charge of developing a generic set of “best practices” for “unlicensed spectrum,” it would have gone nowhere, because every industry that operates on licensed spectrum would have shown up to make SURE it went nowhere. Which is why ICANN is still debating things like “whois policy” with very little practical impact over the last 10 years, because the folks advocating for more privacy and the folks who want draconian measures for failure to provide the maximum amount of identifying information in a very public manner keep canceling each other other out. This is why CableCard has gone nowhere despite nearly 15 years since the 1996 Act was passed. And I guarantee that if BITAG tries to set real policies on “network management,” the various interests represented at BITAG will be utterly incapable of coming to any sort of “consensus” unless one interest becomes so dominant it imposes its will on the others.


One would think that after a financial meltdown created by “industry self-regulation” (“the financial sector has every market incentive to police itself and develop best practices. And they are so much more expert than the government could ever be!”), a massive spill in the gulf (“the oil industry has every market incentive to avoid a major spill and develop its own best practices.”) and now a report on how our food supply is unsafe (“the food industry is much better at policing its products than the FDA could ever be, and has every market incentive to avoid poisoning its customers”), folks might raise a weary eyebrow at the suggestion that industry self-regulation could solve all our problems. I would also point out that the last time we went through a major FCC process on network management practices during the Comcast/BitTorrent Complaint, we saw a similar technology standards/industry self-regulation announcement when Comcast and BitTorrent (the company) announced the formation of a similar technical advisory committee to develop best practices. Once the FCC issued its order against Comcast, however, Comcast apparently lost interest and no one has heard of the Comcast/BitTorrent best practices initiative since. This makes me a bit skeptical that industry folks will abide by “industry self-regulation” out of enlightened self-interest, alignment of incentives, and the rest of the cliches tossed about by worshipers of the Gods of the Marketplace on why we industry self-regulation will beat bad-old-inefficient government every time.

This is not to say that industry technology advisory groups can’t provide useful information. They can. But technical practices can no more tell me what policy to adopt “to preserve an open internet” or to “ensure sufficient investment in deployment” than having the NTSC standard for television tells me what to do about newspaper/broadcast cross-ownership.

Stay tuned . . .

Comments are closed.