The ever-geekalicious Todd Litman of the Victoria Transport Policy Institute had a great take on traffic congestion a few weeks back on Planetizen.

As Litman explains, most congestion studies (such as this annual study, which always gets a lot of press) consistently overestimate the costs of congestion. But even using these relatively high estimates, the costs of congestion are pretty modest, compared with the comprehensive costs of owning and operating a car.

In fact, a quick scan of Litman’s data suggests that congestion represents less than 5 percent of the total cost of car transportation.

Congestion costs vs. all other drivingIn the chart to the right, the "All other costs" bar includes car payments, parking gas, insurance, crash costs, air pollution, road building, road maintenance, and the rental value of land occupied by roads, among other factors.

Of course, some congestion "solutions" actually increase other costs of driving. Building lanes, for example, is costly in itself. Plus, more lanes means more total traffic on a region’s roads, which can simultaneously (a) increase congestion elsewhere in the road system, (b) increase overall parking costs, and (c) increase crash and pollution costs. Taking all costs into account, "curing" congestion by expanding capacity can be more expensive than the disease itself. And as the graph shows, you don’t have to increase the other costs by much, in percentage terms, before the extra costs overwhelm the modest, temporary benefits of congestion relief.