« The Wrong Prescription? | Main | "Federalism and Faith" »

Friday, May 05, 2006

Network Neutrality

I'm no expert on the new movement for network neutrality -- and I'm not quite sure what I think of it.  But I've learned enough about it over the past few days (see Balkin on the issue here and Tim Wu on the issue here) that I feel qualified to post a set of questions about it. 

The gist of the movement seems to be an effort to devise theoretical, policy, or constitutional arguments to prohibit ISPs from discriminatorily offering faster bandwidth to favored content providers.  The hope is that the internet's speed should be available to all equally, regardless of how the internet pipeline owners judge your content.

Tim Wu has a nice set of questions to start thinking about the matter:

To take a strong example, would it be a problem if AT&T makes it slower and harder to reach Gmail and quicker and easier to reach Yahoo! mail? . . .  [And h]ow would you feel if I-95 announced an exclusive deal with General Motors to provide a special "rush-hour" lane for GM cars only?

And here's why we should be more uncomfortable with non-neutrality on I-95:

It's a lot harder to substitute for an interstate. And if highways really did choose favorite brands, you might buy a Pontiac instead of a Toyota to get the rush-hour lane, not because the Pontiac is actually a good car.  As a result, the nature of competition among car-makers would change.  Rather than try to make the best product, they would battle to make deals with highways.

That's what would happen if discrimination reigned on the Internet: a transformation from a market where innovation rules to one where deal-making rules.  Or, a market where firms rush to make exclusive agreements with AT&T and Verizon instead of trying to improve their products.  There's a deeper point here: When who you know matters more than anything, the market is no longer meritocratic and consequently becomes less efficient.  At the extreme, a market where centralized actors pick favorites isn't a market at all, but a planned economy.

So far so good.  But here are my questions:  Isn't there competition among pipeline owners?  Just as a consumer of fast food can go to McDonald's if she wants a Coke and KFC if she wants a Pepsi (Wu's example), can't a consumer go to Comcast, say, rather than Verizon if Comcast offers more information and faster, to boot?  If information really does want to be free and fast, won't ISPs have good reasons not to discriminate without being micro-managed and "planned" from above?  And while we're on the subject, isn't this story about the meritocracy of the internet a bit naive too?  Admittedly, it seems Google conquered the search market through the sheer effectiveness of its product.  But are all internet market leaders where they are because their products are the best?  AOL seems to me to have been an especially lame product that has achieved substantial market penetration.  And, finally, is it really easy to answer the question of neutrality in the abstract?  Don't we need to know more about what is getting discriminated against before we can take a position?  To take Wu's example of certain cars getting a free pass to the fast lane, some places allow hybrid vehicles to use the HOV lane.  I'm perfectly happy with that kind of discrimination and don't much want interstate neutrality because I value the way in which it discriminates.

Thoughts?  Mine are tentative and I welcome feedback.

Posted by Ethan Leib on May 5, 2006 at 09:43 PM in Current Affairs | Permalink


TrackBack URL for this entry:

Listed below are links to weblogs that reference Network Neutrality:


All interesting and illuminating. It seems that I missed this Legal Affairs Debate Club on the topic, which also helps distill some of the issues I posed: http://legalaffairs.org/webexclusive/debateclub_net-neutrality0506.msp.

Posted by: Ethan Leib | May 9, 2006 1:19:07 PM

Thank you Frank for mentioning my work on this topic. I also think it should be a major part of the debate, but of course, that's just me. I'll not repeat what I've written elsewhere in the comments here, but I do want to point out that some of the problems with allowing discrimination by infrastructure providers persist even in the absence of monopoly. Where infrastructure supports a wide variety of uses (and users) that produce spillovers (e.g., because the uses produce public or merit goods), common nondiscriminatory access mitigates market failures that may arise on both the supply and demand sides of the infrastructure market.

Essentially, network owners are not necessarily optimal suppliers of access/use rights because they have an incentive to investigate and support only those uses that generate observable and appropriable private returns, which may or may not be the uses with the greatest social value. Users, on the other hand, are not necessarily optimal purchasers of access/use rights, because if they are productive users – as will often be the case – they do not themselves capture the full social value of their use. The problem with allocating access and use rights in a discriminatory manner is that potential spillovers may remain unrealized if they cannot be easily valued and appropriated by those that produce them, even though society as a whole may be better off if those spillovers were actually produced.

The previous paragraph is a very abbreviated version of the story. I explain the points in much more detail in my scholarship (especially, An Economic Theory of Infrastructure and Commons Management, and Spillovers).

Posted by: Brett Frischmann | May 9, 2006 10:30:39 AM

Lack of transparency is yet another problem with an assumption that competition will resolve the access issues here. The cable companies will not provide menus of which sites they load quickly or slowly. Indeed, ISPs may well choose to hide the existence of disparate treament altogether.

To average users, some sites will just take a long time to load. They will have no idea that the reason is the content provider's refusal to pay the ISP -- most likely users will assume siteowners' technical deficiencies are to blame for such problems. If a consumer actually has a choice (Comcast vs. Verizon, let's say), only extensive research will reveal how the ISPs' various business deals with content providers might affect the user experience.

Posted by: William McGeveran | May 6, 2006 3:34:58 PM

The market certainly may be more monopolized than I acknowledge in the post. And cable access (which, I must admit, is the best form of access from where I stand) hardly presents a field of vigorous competition in most places. But I do think the average individual and business consumer who wants and can get high-speed access has a bunch of high-speed ISPs competing for business (whether through cable, DSL, higher-speed dial-up, satellite, wireless, etc.). And in the 20% of the country where there is no broadband, I suspect there are still a wide variety of dial-up options. And who knows how we'll be getting our internet in the future? One suspects that the future is in wireless broadband, not cable, so the worries about cable monopolies might be short-sighted. Maybe our cell-phone companies will control our internet access soon -- and there is certainly competition on that front. Of course, there may be good reason to regulate their content discrimination if any one company does get a true monopoly. I don't really know what the endgame looks like, truth be told. But I do agree that where there is a monopoly, it may make more sense to regulate. I'm just not sure we are living in that reality yet. As we get closer to it and internet access starts looking like local phone service and cable service, it seems to make sense to treat the pipes owners as "common carriers" with control over a public good and regulate the hell out of them accordingly.

But I think perhaps it makes sense to wait to allow the monopoly to emerge organically BEFORE commencing the regulation AS IF the monopoly exists. Although of course the market should not be "faciley assumed" to have all the answers, it might make sense to let it do its work before assuming it will fail. The political economy of this sort of regulation may actually lead to the very problem network neutrality is seeking to solve. I would think we want to keep incentivizing new technologies and strategies of access (say, like Google's planned free wireless network for all of San Francisco), which might be stymied by preemptive regulation. If Google provided a free wireless network to all residents of the city but discriminated against porn providers, would that really be a tragedy? Wouldn't that help close the digital divide and just make porn consumers pay a bit extra for the luxury? I could live with that kind of non-neurality.

I'm well outside my areas of expertise here, so pardon my naiveté.

Posted by: Ethan Leib | May 6, 2006 12:57:00 PM

It may be due to my cynicism, but I've always seen network neutrality as a red herring "cover" story to keep providers from charging high bandwidth users more money.

I agree that there should be no discrimination based on content, but is it really that unreasonable to expect high bandwidth content providers to pay more money than low bandwidth content providers?

One way or another consumers will pay - either by metered usage through their provider or through paying content providers for higher bandwidth content. It seems like that issue should be where at least part of the debate should be focussed. I'm not sure where I come out on it yet.

Posted by: Michael Risch | May 6, 2006 11:43:39 AM

There was a panel on this at the Computers, Freedom, and Privacy conference. Gigi Sohn of Public Knowledge and Tim Wu gave very compelling arguments for network neutrality, while Chris Yoo argued that there should be a diverse array of services (like we have the post office, fed ex, ups, etc. depending on how urgent package delivery is). Tim Wu responded by asking: do we want an internet where companies and sites succeed based on the quality of their content, or on how sweet a deal they've cut with Verizon or Comcast? Sohn noted that in about 20% of the country, there is no broadband; in 30%, one company has a monopoly; and in 50%, there is a duopoly. Not too competitive a market.

My view is that Brett Frischmann's work on digital infrastructure should also be a major part of this debate. Brett points out that so much of this infrastructure has so many positive externalities that it makes sense to treat it as a public good. To me, the problem of allowing the "pipes" to be totally controlled by private parties is that there are so many public benefits that can arise from the use of individuals who can't afford private charges. If we let private companies adopt the same "high margin/low volume" business model here that has been adopted in so many IP/IT areas, we risk exacerbating the digital divide, both among corporations and individuals.

One last point--what really needs to be done here is cross-national analysis of the many, many countries who've done a better job than we have at supporting internet infrastructure. Many of these, like S. Korea, have made that support a *public* priority, and not facilely assumed that the "market" will have all the answers.

Posted by: Frank | May 6, 2006 9:29:41 AM

Your last paragraph asks some interesting questions about the fear of an emerging "battle of the consortia" in the ISP market. This fear envisions the future of the internet without network neutrality as a series of consortia between bandwidth providers and content providers, with each consortium bound together by exclusivity contracts concerning access to the higher tier of service (if not outright access-blocking of competitors).

You pose the possibility that ISPs won't want to managed from above to enter these types of agreements, but this turns the situation on its head -- it's the bandwidth providers who own the pipes, and it's the bandwidth providers who will drive the move to the "battle of the consortia."

Add to this the concern raised by the prior poster and you begin to see how little competition will really exist in a future without network neutrality.

One final note: You suggest that since a consumer can go to McDonald's for a Coke and KFC for a Pepsi, consumers will be able to make the same kinds of choices with respect to bandwidth providers and content providers. The key problem with this suggestion is that, if you want to have a Pepsi with you Big Mac, you can just hit both drive-thrus. Not so with bandwidth providers, unless you're suggesting that consumers could simply subscribe to *all* available services and pick and choose their content that way.

Network neutrality will keep the internet the internet. Without network neutrality, the internet will become a competing series of Pathfinders (remember Time Warner's ill-fated portal) and GO Networks.

Posted by: anonymous | May 5, 2006 10:44:39 PM

the only problem i have is that a lot of municipalities (in my area of FL, at least) grant monopolies to specific cable companies. for instance, in gainesville there is only cox, in most of orlando there is only bright house (though some parts also have adelphia), etc. so that side of the competition might not actually be there.

Posted by: none_ | May 5, 2006 10:06:39 PM

Post a comment