A quarter century ago, there was a very influential paper that shaped thinking on how best to design what we now call the Internet. The article offered a design principle called “end-to-end.” The idea was to keep the inner part of a computer network as simple as possible and allow the “intelligence” to reside at the edges of the network closer to the end user.
Proponents of this grand design have pushed for net neutrality legislation, which would discourage access providers from placing any intelligence in the inner part of the network. Their ideal of a “dumb network” would be achieved by preventing access providers from charging content providers for prioritized delivery and other quality enhancements made possible by placing intelligence at the center of the network.
This essay examines the merits of the end-to-end argument as it relates to the net neutrality debate. First, we review the evidence on the current status of the Internet, concluding that all bits of information are not treated equally from an economic standpoint. Second, we demonstrate that because consumers and business place a premium on speed and reliability for certain kinds of Internet services, network owners and specialized service providers have responded with customized offerings. Third, we consider our findings in the context of the current legislative proposals involving net neutrality. Fourth, we consider some of the problems with regulating prices and quality of service, which is essentially what the net neutrality proponents propose. Our principle conclusions are that the end-to-end principle does not make sense from an economic perspective and that further regulation of the Internet is not warranted at this point in time.