Two words: balanced transportation.
By Eric Jaffe, June 20, 2014
People run to catch a train at Lisbon's Cais do Sodre station.
For as much as U.S. metros have grown over the past few decades, commute times have remained oddly stable. One-way commutes averaged
21.7 minutes in 1980 and 25.3 minutes in 2010. That's not nothing —
it's 30 hours more a year assuming 250 work days — but it's also not
nearly as much as decade upon decade of urban sprawl might suggest. As
bad as our daily commutes are, in many ways we're fortunate they aren't
So why aren't
they worse? Among the most compelling ideas to explain this phenomenon
is that people have daily "travel time budgets" of about an hour that
they refuse to exceed. (This budget is often called "Marchetti's Constant," though it's better attributed to Yacov Zahavi.)
But that alone isn't enough to explain what's happening. After all, if
traffic increases and jobs stay in the same place, then our hour won't
get us as far as it once did, whether we like it or not.
Then what else is going on? Two new papers target this classic
question in a fresh way. Their findings suggest that much of what's
going on is more people relying on public transit.
The first study comes courtesy of economist Alex Anas of SUNY Buffalo.
Using metropolitan Chicago as an example, Anas modeled projected growth
in the area from 2000 to 2030, including expected highway expansions.
He found that a 24-percent jump in population led to 19 percent more
urbanized land — a sign of sprawl — but only a 4.5 percent rise in
commute time. The average commute climbs from 30 minutes even in 2000 to
31.7 minutes in 2030.
Once again we see commutes stay the roughly the same despite city
growth, and once again we're left wondering why. The biggest reason,
according to Anas's analysis, is that in the face of terrible road
congestion, lots of people switch to public transportation. Transit
trips increased by 55 percent in the model by 2030. Some of these people
moved from the outer suburbs to inner areas better served by trains or
buses. Most of them simply gave up driving to work.
Two other key factors shortened
average commutes. People moved closer to work in the core (and jobs
moved out of the core and closer to people). And people took fewer (or
shorter) non-work trips to balance out the time spent commuting.
Altogether, writes Anas, these three changes kept travel times
relatively constant for all modes over the 30-year period even as
population and land area grew considerably.
The second study comes from MIT's Senseable City Lab. Using cell phone data to track commutes around the world — a novel if imperfect
approach — the researchers found travel time to be largely independent
of distance in several regions. In other words, no matter how close
people lived to work in these places, they managed to spend about the
same time getting there.
In Boston, for instance, morning commutes took 50 to 60 minutes
whether people lived 3 miles from their job, or 25. In parts of the
Ivory Coast, the morning trip took longer on average, about 80 minutes,
but this travel time was still consistent across many distances. And in
parts of Portugal, where commuters averaged about 70 minutes in the
morning, trip times were again stable even as trip distance increased.
Those times feel a bit high (likely because cell phones are imprecise
tools) but their consistency fits with the concept of travel budgets.
Where the MIT data gets interesting is in Riyadh, Saudi Arabia, where
commutes finally do increase with distance to work. The difference here,
according to the researchers, comes down to travel options — of lack
thereof. In Riyadh, only 2 percent of trips occur on public
transportation. (Data from car-only commuters in Milan produced similar
fluctuations.) For travel times to hold constant, travel choices need to exist.
The short lesson here is that commutes find a way.
In the service of a daily travel time budget, some people choose
alternative transport options. Others move closer to work or take jobs
that moved closer to them. Still others simply cut out trips that aren't
commutes. We evolve with our metros; as the MIT researchers conclude,
people "adopt their lifestyles (e.g. commute behaviors) in a way so that
they spend reasonable amount of time of their lives commuting."
But the key really seems to be the presence of balanced transportation systems. And these don't even have to be well-balanced
systems. Look at the disparity in commute times between drivers and
transit riders in the Anas chart above: it's huge. Some people have to
ride because they have no car, but others hate sitting in traffic so much they willingly endure a longer trip. That's a clear case for building transit systems that reach more job centers and arrive more reliably.
Commute times may find a way on their own, but we can help them find a better one.