The Tragedy of the Last Mile: Economic Solutions to Congestion in Broadband Networks

Abstract

The growth of the Internet has constrained broadband networks, forcing service providers to search for solutions. We develop a dynamic model of daily usage during peak and non-peak periods, and estimate consumers’ price and congestion sensitivity using high frequency usage data. Using the model estimates, we calculate usage changes associated with different economic and technological solutions for reducing congestion, including peak-use pricing, throttling connectivity speeds, and local-cache technologies. We find that peak-use pricing combined with local-cache technology is the most effective way to shift activity from peak to non-peak periods.

Keywords: broadband, congestion, nonlinear pricing, peak-use pricing

JEL Classification: L11, L13, L96