Over at Arcane Gazebo, Travis talks about commuting, based on an Ezra Klein post:
Longtime readers know my obsession with the way we overvalue positional goods like money, prestige, and real estate and undervalue non-positional goods like social connections, walking to work, and health. But the evidence really is clear that you need to make a whole dump truck of money to outweigh the happiness offered by being only a 15 minute stroll from the office, and that that extra room for your old guitars isn’t going to make you nearly as ecstatic as you think it will.
Lo these many years ago, I saw a talk at NIST by a transportation expert. His main purpose was to promote the idea of high-speed maglev trains (and he wasn’t terribly convincing on that count), but he opened with an interesting observation, that he claimed was backed up by science: He said that if you look back over the last hundred-odd years, the duration of the average “commute” to work has been more or less constant, at something like forty-five minutes (I don’t recall if that was one way or both ways).
His claim was that a long time ago, people had to walk or ride horses everywhere, and thus lived very close to where they worked, whether that was a local farm or a factory. As transportation options improved (cars, trains, buses), people moved farther away from their place of work, because they could live farther away and still get there in the same amount of time.
It’s an interesting claim. I have no idea how you would go about substantiating it. It would also seem to contradict the source that Ezra cites (which eventually traces to a Business Week article) claiming that non-commuters are happier with their lives than commuters. Of course, I’m not sure how you’d go about substantiating that claim, either, at least not in a way that I would really trust…
Anyway, I thought I would throw that out there, and see if anybody recognizes the claim about commuting times. I don’t remember the name of the speaker or where he said his information came from, but if anybody knows more about it, I’d love to have a source.
I don’t see how the transportation expert’s claim is inconsistent with Klein’s. It could be true both that non-commuters are happier than commuters and that the duration of the average commute has not changed over time if in the past, people also made ‘bad’ decisions about where to live.
Longtime readers know my obsession with the way we overvalue positional goods like money, prestige, and real estate and undervalue non-positional goods like social connections, walking to work, and health.
I don’t know what fucking crack he’s smoking, but for a large majority of us in metropolitan areas, living close enough to work to walk isn’t the inexpensive option, it’s the unattainable one. I couldn’t buy anything within walking distance of my job for less than a million dollars. Sorry, I’ll live with my commute, and continue hating it, but being able to afford to eat.
Yeah, I have heard that claim before, with slightly different numbers (approx. 2hrs total travel time per day is the way I’ve heard it). Never bothered to track down a reliable source though…
I’m not sure the commute time theory works at all. A couple examples:
Back in the 1880’s and up to the early ’50s, many people commuted daily from southeastern New Hampshire to Boston. The train rides averaged in the 30-40 minute range. Now the commuter trains are gone and highway traffic is horrendous. I’m sure the current times are in excess of an hour.
The older suburbs of northern and central New Jersey (Montclair, Nutley, and even as far south as the Princeton area) were largely built by commuters to NYC — they had either a quick 20-30 minute train ride to Hoboken, ferries that ran every 10 minutes to the city, and then a quick walk or subway ride to work, or a slightly longer train ride direct to Penn Station in NYC. In the late ’50s to early ’70s, most of the residents of those suburbs were commuting to jobs in Newark instead of the city. Newark at that time had the highest percentage non-resident workforce in the nation.
So, in some cases commutes are longer (or have stopped being feasible at all) and in some cases the target of the commute has changed to retain a reasonable time.
I don’t know what fucking crack he’s smoking, but for a large majority of us in metropolitan areas, living close enough to work to walk isn’t the inexpensive option, it’s the unattainable one. I couldn’t buy anything within walking distance of my job for less than a million dollars. Sorry, I’ll live with my commute, and continue hating it, but being able to afford to eat.
I think the argument that Klein would make is that if you were willing to accept a different sort of lifestyle– renting rather than owning, living in shared housing, taking public transportation rather than owning a car, etc.– you would be able to walk to work, and be happier for it.
It’s not an accident that this sort of stuff largely gets promoted by young, urban hipsters, as it basically boils down to “if everybody were willing to live like a young, urban hipster, the world would be wonderful!” It tends to ignore the fact that being able to do that sort of thing is contingent on a great many factors, such as having the sort of job (like “magazine pundit”) that lends itself to being situated in a walkable residential area.
I imagine you won’t find a lot of people working in the manufacturing sector advocating that everybody live within walking distance of work…
Back in the 1880’s and up to the early ’50s, many people commuted daily from southeastern New Hampshire to Boston. The train rides averaged in the 30-40 minute range. Now the commuter trains are gone and highway traffic is horrendous. I’m sure the current times are in excess of an hour.
I think the counter-argument would be that the figure is for some sort of global average commute. The fact that commutes in some particular area may have gotten longer doesn’t necessarily mean that the average commute has gotten longer, provided you’re averaging over a big enough population.
The older suburbs of northern and central New Jersey (Montclair, Nutley, and even as far south as the Princeton area) were largely built by commuters to NYC — they had either a quick 20-30 minute train ride to Hoboken, ferries that ran every 10 minutes to the city, and then a quick walk or subway ride to work, or a slightly longer train ride direct to Penn Station in NYC. In the late ’50s to early ’70s, most of the residents of those suburbs were commuting to jobs in Newark instead of the city. Newark at that time had the highest percentage non-resident workforce in the nation.
Again, that doesn’t necessarily argue against the idea that the average overall commute has stayed mroe or less constant. The figure isn’t for the average commute to a particular place, it’s for the average commtue to anywhere.
In fact, I think he even cited the tendency of people to live farther away and only commute partway in as one of the factors leading to that figure. If the commute from a nearby suburb to an urban center becomes longer for some reason, then you can keep the same average commute by changing your work pattern so you commute from a more distant suburb to an inner suburb, for example.
I’ve also heard that “45 minutes” claim but don’t recall a source.
In the mid 1990s while in San Francisco I visited the Maritime History Museum, which had an exhibit on the ferry network that existed before the Golden Gate Bridge and Bay Bridge were built. One of the factoids I remember from that exhibit is that traffic speeds on San Francisco’s surface streets were higher ca. 1920 than they were ca. 1990, mainly due to less traffic congestion.
I’m lucky enough to be in Ezra’s coveted 15 minutes walk to work situation. The problem for many people is not that such housing is unaffordable (Jamie’s complaint), but that such housing is unavailable at any price. Look, for example, around the Beltway: you will see massive office parks and shopping plazas served by pedestrian-hostile multilane highways and frequently not near any residential developments. To take a specific case: the Greenbelt Marriott (where I have occasionally stayed on business) is about as far from the Greenbelt Metro station as my house is from my office, but while the latter is easy to walk (crossing just two local streets), the former is generally considered unwalkable because there is a major arterial road that must be crossed.
I think there is a barnyard aroma about this. I suspect that it is strongly location biased, which might account for an average if the locations were high in population density. I suspect that commutes in those locations are significantly longer than 45 minutes. On the other hand, a commute in my home town might be at most 30 minutes if I were going from one end of the suburbs (I use that term loosely) to the other. But then I can’t get a job paying a third what I make if I work in my hometown.
Jamie, I think the contentment factor might include a willingness to relocate and change jobs, not just trying to find housing near where you already work. For me, it’s not possible to find work remotely similar in type or pay to what I do now. But if certain other factors were different, that might not matter so much. For example, I have to accumulate significant money to pay for retirement and health care. If those could be provided for by other means, I could be happier working closer to home but not making as much money.
I also don’t think the historical 45-minute commute is inconsistent with Klein’s claim, but for a different reason. An individual decides where to live based on some personal utility function which takes into account commute time but also the other factors people have mentioned: housing costs, school districts, etc. Since not everyone is a young urban hipster, these utility functions aren’t all going to be peaked in dense urban areas and some people are going to choose to live some distance from work despite a long commute time (even correctly valued). Since some of these preferences are highly individualized, it’s hard to make universal statements about it.
However, there’s a contribution to the utility function due simply to the opportunity costs of sitting in your car for time 2T every day, and this should have some universal properties, since having extra time is valuable to everyone. The statement about one hour commute = 40% salary is a statement about this contribution to the overall function, and it’s this that Klein is saying people undervalue.
I don’t know what methodology the researchers in the Business Week article used–it sounds like they were doing some empirical studies. But the 40% salary number can be reproduced with a very crude argument about opportunity costs:
Consider a worker who makes hourly wage W and works for 8 hours with a one-way commute of one hour. In order to collect his daily salary 8W, he must invest a total of ten hours. If his commute time were reduced to zero, one way he could spend the extra two hours is simply to continue devoting them to work, collecting two hours of overtime pay at hourly rate 1.5W. This earns him an extra 3W per day, a 3/8 increase on his salary (i.e. almost 40%).
(Disclaimers: I am not an economist, and this is not intended to be a rigorous proof.)
Danny Kahneman, recent Nobel recipient, did some of the research about happiness and components of happiness, noting that commuting time was a net loss for overall happiness. That’s one of the sources of Ezra’s perception. See:
ideas.repec.org/p/iza/izadps/dp1278.html
for a specific focus on this, and
http://www.atypon-link.com/AEAP/doi/abs/10.1257/089533006776526030
for a more general approach to the overall topic. For what it’s worth, I knoe people who commute by bicycle for up to 45 minutes and report it is a much more satisfying experience than driving.
In the video that accompanied the Science issue about cities a few weeks back, some mention of a constant commute time of 35 minutes was mentioned. I haven’t (yet) read the articles, so I don’t know if it’s mentioned there.