About ten months ago, I briefly chronicled the San Francisco area's strategies for battling congestion. Under the category of congestion pricing, tolls for area bridges and roadways went up when use went up. While many people look at roadway capacity as a supply-and-demand function, they don't like to think of those same economic forces are applied to roadways, if they actually have to pay for it. Which in that essence is why it actually relieves congestion.
The San Francisco Chronicle reported on a study that followed up on the strategy, and it appears to be working. While carpooling (a dubious attempt at congestion reduction) has decreased, so has main lane use. Transit ridership is up, vehicle miles traveled are down and travel time savings is up, tremendously in some cases. The extreme example is I-880, which saw its average travel times reduced by half!
The main problem with congestion pricing implementation on a wide scale is the lack of available case studies. However, of the ones that are out there, it appears that adding a market approach actually reduces congestion, more so than any other attempt that has been tried before.