Ev Ehrlich's Everyday Economics

The Rabbit Test

The latest entry in the ongoing analytic debate over net neutrality comes from the New York University Law School’s Institute for Policy Integrity – love the name, guys – and their report, Free to Invest: The Economic Benefits of Net Neutrality.   To its credit, this report uses the tools of economic analysis to approach the issue.  And, again to its credit, it correctly depicts the net neutrality issue as whether we should subsidize content providers at the expense of infrastructure providers on the Internet. 

I’m going to spend the rest of this entry talking about how screwed up this paper is and how these guys missed the boat.  In fact, I wonder if the authors themselves realize that their analysis depends on a misconception they execute early in the analysis and then carry through to their conclusions, much like a rabbit stuffed in a hat and then pulled out later by its plucked ears to the wonderment of the crowd.  Still, at least they’re trying to present the argument in economic terms, which is a refreshing change from the all-too-frequent stuff you hear about net neutrality will stop broadband infrastructure providers  who want to censor free speech, corrupt the content of competitors, stop you from listening to Pearl Jam, and otherwise pillage the Internet. 

A flawed analysis is better than bizarre rhetoric any day, I say.

The Rabbit Enters

Let’s get to the heart of the matter.  Here’s a quote from page 12 of the report (and my ellipses are only to spare you from reading and me from typing);

                The Internet market exhibits …the “network effect.”  The effect occurs when the value of a good or service increases as others purchase the same good or service.  A telephone...(is a good) example…a single person who purchases a phone does not reap benefits unless others buy telephones and use them…(I’m on page 13 now)…This means the Internet becomes more valuable to each user as additional users are added.

Now that is simple and straightforward, and all economists agree.  It’s one of the reasons the old Bell phone system was a regulated monopoly, and a reason why universal broadband access is a policy goal that’s shared by rational people everywhere.

But then the authors stuff the rabbit (page 4, right at the top of the Executive Summary):

                The Internet…(is).. both…the physical infrastructure as well as the content and

information moving along that infrastructure

Now wait a gosh darn minute.  No.  The Internet is the medium by which content is distributed to people who want it.  No infrastructure, no Internet.  No content – empty Internet.  Big difference.  The reason why the Internet has a “network effect” is that when I get access to it, it’s better for my friends Stan and Marv, because they can communicate with me.  We call this “network effect” an “externality” because when I get access, I create benefits that are “external” to me – Stan and Marv are better off without paying to be better off, because now they can do more with their connection. 

Now, let’s say there’s a new website – content – that shows videos of cats playing the xylophone – xylophonecats.com.  That makes the Internet more valuable to me, Stan, and Marv, to be sure, because who doesn’t like a video of xylophone-playing cats?  But there’s nothing “external” about that.  If I like cats who play like Milt Jackson, I’ll be willing to pay more for broadband.  That’s just about me.  And the entrepreneurs who created the site will get whatever they were after – let’s hope it was money.

And that’s the point – you can say that the Internet is one “eco-system” of infrastructure and content, but the content rides on the infrastructure.  The medium is the message, as they used to say in a different context.    The only “externality” is the connectivity of the infrastructure.  The Internet is like the phone system – if I get on, Stan and Marv can send me e-mail and share stuff with me.  But websites are only like stores – Amazon is like Sears, EBay is like the Stock Exchange, and Google is like the Encyclopedia Britannica, the Automobile Club, the Phone Book, and the Prophesies of Nostradamus, only much, much better.  If I use Google, Stan and Marv aren’t any better or worse off, because they can use Google whenever they want irrespective of what I do, just as they shouldn’t (and don't) give a rat’s behind if I go to J. C. Penney this afternoon.

The Rabbit Appears

But the NYU authors don’t get this, because they want to show that ­subsidizing content is important and desirable.  They want to do that because “net neutrality” is a way to subsidize content – it prohibits any website from buying a faster speed than its competitors, much like they can buy express mail rather than regular mail to send letters.  So the NYU authors need to show that “content,” like “infrastructure,” generates “externalities.”  If they can’t show that, then making content providers compete for the “best” Internet speeds makes good economic sense – just like letting people use express mail versus regular mail, or buying Sears’ “Good” versus Sears’ “Best.” 

To show that, the NYU authors must resort to saying this, (page 13);

                If a potential user does not connect to the Internet because she does not find enough valuable information on it, the value of connecting to the Internet is reduced for everyone.

Think about that statement.  NYU is arguing that it’s OK to subsidize content because more content will induce more people to get on the infrastructure.  Let’s parse that for a few moments.

First, it’s a straightforward admission by the authors that they’re flat-out wrong, that only infrastructure conveys net work effects.  It admits this by noting that the significance of content is that it leads people to access the infrastructure.  That is, a user finding valuable information doesn’t generate any “external” benefits unless she connects to the Internet.  The connection is everything.

Second, do the NYU folks really think that  subsidizing content – Google, etc. -- is the best way to get more people on the Internet?  How about putting out contracts to build broadband infrastructure in neighborhoods that don’t have it – you know, rural areas, and/or where poor people live?  The Obama Administration is for this – perhaps the NYU guys could support that instead of making Google richer.  Or maybe we could give tax credits to companies that build new high-speed broadband (whether cable, fiber, wireless, or whatever else works) in currently-unserved areas.  But nothing about that from NYU – letting Google and EBay send their stuff out for free is apparently a better way to build new infrastructure than actually going on out and getting the job done.

The NYU authors have to confuse the information (and products and services) on the Internet with the Internet itself because they want to support “net neutrality,” and net neutrality is about whether Big Websites can stop the infrastructure providers (Verizon, Comcast, ATT, Cox, and so on) from buying faster connections, the way it allows you to buy a faster connection at the other end.  In order to reach the conclusion they want, the NYU folks have to argue that’s bad;

(“Net neutrality”)…may increase Internet service providers’ ability and incentive to invest in …broadband lines…but would decrease content providers’ ability and incentive to invest in…the Internet’s content.

Not only do they argue that subsidizing content leads to more infrastructure, but they now are arguing that a rational economic system for allocating the Internet would be bad!  Let’s say somebody wanted to create an alternative to Skype that was faster and that never froze up, but charged you ten cents every time you made a call.  They could do that by offering Internet infrastructure providers a payment to assure that their “packets” (the zero and one “bits” that make a Skype call) always go ahead of xylophonecats.com.  That would increase both infrastructure and content.  But you can’t do that under “net neutrality” because it’s not neutral.  In fact, “net neutrality” prohibits this kind of new content.

There’s a reason why Google and the other Big Websites are paying for the net neutrality campaign.  It’s because they are already the biggest and fastest providers on the Internet – they’ve “cached” their content all over the Internet like squirrels hiding acorns, creating server farms somewhere near you so you’ll reach them more quickly.  For someone else to come along right now and pay the infrastructure providers for the right to reach you at a comparable speed -- by buying an express lane on the Internet’s tubes – is good for you, bad for Google.

The NYU report is right in noting that “net neutrality” is about a boodle of money being shipped from internet providers to content providers.  But they’re wrong to argue that we benefit from that.  And they’re wrong to argue that changing the system would produce less content – it would simply let new kinds of content challenge the old.  And they’re wrong to argue that the economics of content and infrastructure are the same.  Let’s just say they failed the rabbit test.