In one recent survey the U.S. has relatively short commutes; in another, they're relatively long
This month, the Organization for Economic Cooperation and Development released the results of a new international survey on well-being. Over at The New York Times Economix blog, Catherine Rampell points to an intriguing finding from the O.E.C.D. data: commute times in the United States are quite low compared with those around the world. The survey showed that U.S. workers clocked their commutes at 28 minutes per day. That's a full 10 minutes less than the overall average commute time, half as long as the commute in worst-ranked South Africa, and shorter than the work rides in all but three developed countries:
The finding is a bit surprising, to say the least. General wisdom holds that Americans have pretty bad commutes, owing largely to sprawling metro areas. It also seems directly at odds with recent data released by the American Community Survey of the U.S. Census [PDF]. Those numbers — which Rampell also noted a few weeks back — gives a "one way" mean commute time of about 25 minutes, or close to 50 minutes per day:
Understanding the precise length of time spent commuting is important to understanding well-being in general. As we've explained before, commute time and happiness are closely connected — so much so that you have to earn a lot more money to make up for the poor quality of life you suffer from a long commute. In one frequently cited study of well-being from a few years back [PDF], economists found that of 13 daily activities, including housework, cooking, and child care, commuting to and from work ranked thirteenth and twelfth on average enjoyment, respectively. Just about the only thing Americans say they dislike less than commuting is working itself — and only slightly less at that.
So why is the question of commute time so hard to determine? Well one reason the U.S. figure may be tough to pinpoint could be the sheer size of the country. That leads to great disparities between metro areas. As Brad Plumer points out at Wonkblog, the Census average combines commutes in the New York City region (the longest one-way commute, at nearly 35 minutes) with the likes of Great Falls, Montana (the shortest, at 14). It's tough to settle on a single "American" commute with such a wide-ranging population base.
A second complicating factor noted by Plumer is that automobile and transit commutes are lumped together. That's problematic for a number of reasons. For starters, well-being may be different for car and transit commutes, since the latter lacks the stress of driving in traffic. Additionally, average commute times, according to the Census, are double that of driving times. That's a bit misleading, going back to the city size caveat, since most long commutes are in big cities, where comparable car commutes are also longer than the national average:
Another big problem, writes Ryan Avent over at the Economist's Free Exchange blog, is one of methodology. Avent takes the 50-minute Census commute time and compares it with yet another recent data set, the European Working Conditions Surveys, which uses a collection method similar to the Census. This comparison offers a dramatically different picture from the O.E.C.D. chart. Here the United States demonstrates a much longer commute than in countries like Britain, Spain, and Italy — countries it gets the better of in the O.E.C.D. rankings. The reality, as Avent notes, is probably somewhere in between the Census and O.E.C.D. numbers.