Cord makes some good points about the disadvantages of open networks, but I think it’s a mistake for libertarians to hang our opposition to government regulation of networks on the contention that closed networks are better than open ones. Although it’s always possible to find examples on either side, I think it’s pretty clear that, all else being equal, open networks tend to be better than closed networks.
There are two basic reasons for this. First, networks are subject to network effects—the property that the per-user value of a network grows with the number of people connected to the network. Two networks with a million people each will generally be less valuable than a single network with two million people. The reason TCP/IP won the networking wars is that it was designed from the ground up to connect heterogeneous networks, which meant that it enjoyed the most potent network effects.
Second, open networks have lower barriers to entry. Here, again, the Internet is the poster child. Anybody can create a new website, application, or service on the Internet without asking anyone’s permission. There’s a lot to disagree with in Tim Wu’s Wireless Carterfone paper, but one thing the paper does is eloquently demonstrate how different the situation is in the cell phone world. There are a lot of innovative mobile applications that would likely be created if it weren’t so costly and time-consuming to get the telcos permission to develop for their networks.
Most of the problems Cord cites with open networks are by-products of their greater usefulness. The reason computers crash more than cell phones is that computers have a lot more functionality. There is more spam on the Internet because the barriers to entry to getting an email account is much lower. You could easily build a dumb-terminal computer that never crashed or a closed email platform that had no spam, but few people would want to use them because they would be much less useful.
Which isn’t to say that regulations mandating open networks is a good idea. A poorly-designed open network will often be worse than a well-designed closed network, and the regulatory process is not known for making great technical decisions. Moreover, bringing politics into the process leads to wasteful rent-seeking behaviors and opens the risk of regulatory capture. But just as our opposition to drug prohibition doesn’t necessarily mean that it’s a good idea to use cocaine, we shouldn’t fall into the trap of reflexively arguing against open networks simply because our ideological opponents are for them. It’s perfectly consistent to believe that open networks are a good idea, but regulations mandating them are not.
One point that I think libertarians in particular ought to appreciate is that if you want the benefits of a closed network, you can always build one atop an open network. People can create websites like Late Night Shots to exclude the riff-raff. They can create virtual private networks to prevent their ISPs or others from snooping on their network traffic. There’s a close parallel here to Nozick’s concept of a “framework for utopias”: if you want to live in closed community, you an voluntarily set one up within a larger open society. But you cannot do the converse. Precisely the same insight applies to networks: you can build a closed network atop an open platform, but you cannot do hte reverse. Which means that all else being equal, it’s preferable for basic infrastructure to be open, leaving users with maximum freedom to make their own decisions about how much they want to interoperate with others.