San Francisco priced curb parking so one or two spaces always stayed open

City drivers spent an enormous share of traffic circling for a curb space, because meters were priced the same everywhere regardless of demand, so desirable blocks were always full and empty blocks rarely empty.

In 2011 San Francisco launched SFpark: live sensors report how full each block is, and meter prices rise on crowded blocks and fall on empty ones. The goal is a steady 60–80% occupancy, which keeps parking both well used and available.

It worked. The average meter price fell about 4%, the measured elasticity of demand was about -0.4, and after two years 62% of blocks sat in the target range. Cruising fell by roughly half, so the same street space produced far less wasted driving.

Price curb space to keep it available

A flat meter price made desirable blocks chronically full and others empty

Keeping one or two spaces open removes the need to circle, which itself pollutes

Price data reveal true demand, so the city can steer occupancy without building more parking

Average prices moved down, not up, so the reform was not a revenue grab

A scarce shared resource gets mis-used when its price is fixed; expose occupancy and let the price adjust until a target availability is hit, and the resource allocates itself.

SFpark became the leading real-world test of occupancy-targeted parking pricing, on which later demand-responsive pricing models are built. Its results were published in the Journal of the American Planning Association and are studied as the case for pricing a scarce public good by availability rather than by flat fee.