 Sending-party-pays policies—in which companies such as Netflix pay mandated prices to have their services delivered the last mile to customers—distort prices in the complex Internet peering and transit services market, resulting in inefficient traffic management.

 The optimal price of peering varies widely and changes rapidly depending on network effects and the price elasticity of demand. Regulating it is likely to raise prices for consumers while driving some out of the market for certain online services entirely.

 Broadband infrastructure is not a zero-sum burden because edge companies aren’t just users; they also build delivery networks and cache content to improve performance. And users both create and consume content, which gives Internet services their value.

 South Korea’s experiment with sending-party-network-pays has caused less-efficient traffic flows, higher prices, and lower content quality. It is a cautionary tale for policymakers considering sending-party-pays.

 Proposals in Europe are likely to produce similar results to those in South Korea, which may explain why European regulators have rejected such proposals in the past.

 Expanding universal service fees to edge services is the wrong way to offset the rising costs of U.S. subsidy programs and would raise Internet prices for consumers. Policymakers should instead reduce USF expenditures, especially for rural broadband.