In an Unfogged thread the other day, there was some discussion about the importance of transit in allowing commuters to escape congestion, and a commenter remarked that New York City has the longest average commute times in the nation. I responded by saying it’s more important to focus on the efficiency of people moved, rather than individual commute times.

New York manages to deliver over two million people to its central business areas with an average commute time of about 38 minutes. Houston, by contrast, moves about 300,000 people to its central business district with an average commute time of around 26 minutes. Two things about those datapoints. First, if Houston were required to move two million commuters to a central area on a scaled up version of its current transportation system, one imagines that the gridlock would be horrific, and average commute times would be well above 38 minutes. Second, the productivity gains produced in New York City by the density of its urban agglomeration allow employers to compensate workers handsomely for the additional time they spend commuting.

I want to add a few other points, now. One thing to remember is that commute times for those living in the central city are generally well below the metropolitan or all-city average. Average commute time in the District is second lowest to Arlington among regional jurisdictions. This isn’t just a function of distance; Prince William County has the highest average commute time in the area, despite the fact that the vast majority of Prince William residents work in Prince William or neighboring Fairfax County, and not in central jurisdictions.

It’s still the case that transit has a high fixed time cost, which leads to a high overall time cost, relative to other modes. Trains move about as fast as cars, but when you add in the time to get to and from the train and meet connections, mass transit ends up being slower than driving in many cases. Still the density of areas well-served by transit probably cancels out the additional slowness, and maybe even generates overall time savings.

Downtown Washington is dense, and one of the main reasons it can be so dense is because a large share of downtown workers–a third, at least–arrive at work by transit. Density also means that a lot of different businesses can locate in a very small area, and that business density means that errands during the day take very little time. My bank, my drug store, my barber, my favorite coffee shop, my Metro station, my preferred book store, and twenty different lunch options (conservatively estimated) are all located within two blocks of my office. I routinely run to the bank, pick up something from the corner bodega, buy lunch, and return to my desk before the screen saver clicks on.

My neighborhood in Northeast D.C. is slightly less dense, but still accommodates most of my weekly needs in a five block radius–which means errands take very little time. I can hit the grocer for a six-pack of beer and run up to the Chinese restaurant to get take-out and be back home in just over the duration of a commercial break. When we visit our folks in Prince William County or down in Raleigh, however, I’m constantly aggravated by how long it takes to get anywhere. The distance is a problem, as is traffic, as is the constant battle with traffic lights. Time just evaporates.

My point is this: getting to dense areas in big cities generally involves more time than commutes in low-density areas in other cities. I suspect that drawback is well-worth it given the time saved on short, intra-city trips during the day. Overall daily time in transit could well be less for dense places than for sprawling ones. We don’t really have a name for that statistic, or measures of it, as far as I can tell, but it’s certainly worth remembering when comparing average commutes.