Cities More Accessible in U.S. Than Europe

“Should the U.S. repair crumbling roads and highways to enhance car-based mobility or replace them with new public transit infrastructure that re-orients U.S. commuting systems away from their current car dependence?” asks a paper recently published by the National Bureau of Economic Research. To answer the question, the paper compared accessibility via transit and driving in about 50 U.S. and 50 European cities. If transit made European cities more accessible, the researchers reasoned, then it would make sense for the U.S. to emphasize transit as well.

Should the United States attempt to build as much transit infrastructure as is found in Europe even if doing so reduces people’s access to jobs and other economic opportunities?

Instead, the researchers — two economists from Yale and one from UC San Diego — found that U.S. urbanites had far more access to their cities than Europeans did in theirs. Moreover, Europeans using cars had far more access to their cities than those who relied on transit. This shouldn’t be a surprise to those familiar with the research published by the University of Minnesota’s Accessibility Observatory, but it seems to have surprised the people doing this research.

Still, they concluded that American cities should consider building more transit. This is based on several flawed premises. First, their basic premise — that crumbling highway infrastructure creates an opportunity to shift people’s mode choices by investing in transit — is simply wrong. Our highways were never in the dire shape that the media claimed and American urban areas offer few opportunities to trade highways for transit infrastructure.

Second, there was a basic flaw in their study: instead of measuring total job accessibility in each urban area, as is done by the Accessibility Observatory, the Yale/UCSD researchers focused on access to central business districts. This was based on the flimsy premise that CBDs “account for an outsized share of the United States’ economic growth.” In fact, downtowns have steadily declined in importance for the last century. Measuring total accessibility would have revealed that transit is even less important, in both the U.S. and in Europe, than the researchers found.

Third, the researchers claimed to have shown that “US cities’ car orientation comes at the cost of less green space, more congestion, and worse health and pollution externalities.” In the researchers’ minds, this must offset the accessibility benefits of automobiles. Yet they appear to be choosing data that supports their preconceived notions rather than objectively assesses such issues as pollution and green space.

They claim to have found, for example, that more auto-oriented cities suffer from more nitrogen oxide (NOx) pollution. To reach this conclusion, they didn’t compare U.S. vs. European cities; instead, they just compared all of the cities in their sample. But the fact is that the U.S. has dealt with pollution at the tailpipe, not by trying to influence how people travel as Europe has done. As a result, other studies have found, European cities on average have at least a third more NOx pollution than U.S. cities of comparable size.

In assessing health effects, the researchers found that people in more auto-oriented cities are more obese than people in more transit-oriented cities. The results were statistically significant but small: a 10 percent increase in auto accessibility was associated with a 2.9 percent increase in obesity while a 10 percent increase in transit accessibility was associated with a 0.9 percent decrease in obesity.

This is an old debate and the researchers’ error is in assuming that correlation equals causation. At least two research studies have found that any correlations between obesity and urban form are due to self-selection: that is, auto-oriented areas don’t make people obese, instead, obese people choose to live in more auto-oriented areas because they are more accessible to them.

Their claim that auto accessibility leads to more congestion is also dubious. The latest international congestion data found that the first, third, and tenth most congested cities in the world are in Europe. Ten of the 25 most congested cities are in Europe compared with eight in the U.S. Living in Europe doesn’t immunize travelers from congestion.

The green space argument is even more questionable. The Yale/USCD researchers compared green spaces with auto and transit accessibility on a county level and found that increases in both transit and auto accessibility reduced the amount of green space, but autos reduced it more: about 13.9 percent for a 10 percent increase in auto accessibility vs. just 5.7 percent for a 10 percent increase in transit accessibility. Apparently, we can have the greatest amount of green space if we are too immobilized to get to it.

More important, it appears that, when counting green space, the researchers failed to include people’s yards. Auto-oriented cities allow people to have larger yards, which means they have less of a need for parks or other open spaces. Most people regard their yards as green space, but the researchers apparently did not as they found that only 9 percent of U.S. cities were green space. If yards were included, this would have been closer to 25 percent. Counting yards, it is likely that auto-oriented places have much more green space than transit-oriented places.

Using counties rather than metropolitan areas as a measure of green space is also a problem. Most large U.S. metropolitan areas cover multiple counties, and the suburban or more auto-oriented counties in those regions tend to have a lot more green space. This green space would have been ignored if the researchers limited themselves to the counties in which the CBDs were located. Unlike transit, automobiles give people easy access to green spaces and other resources that are outside their county of residence, so limiting the analysis to counties biases the results towards less mobile regions.

Similar criticisms could be applied to the researchers’ other negative conclusions about auto accessibility. For example, they say that “car accessibility is a positive predictor of ‘food deserts.'” Of course, if you have a car, you don’t need to live next to a grocery store. Similarly, they compare auto accessibility with a walkability index, which is a concept made up by people who don’t like cars.

Finally, the researchers make some sweeping claims that are not supported by their data. “European workers are substantially less likely to use cars [for commuting] than are US workers,” they say, and the reduction in driving is equally split between transit or walking/cycling to work. “Workers in large cities are more likely to use public transit and less likely to drive.” Note that they aren’t saying that Europeans are more likely to use transit than to drive — they aren’t — but to some it might sound like that.

Their fundamental conclusion is that redesigning cities to be more transit-oriented and less auto-oriented can have large effects on people’s mode choices. Their focus on central business districts is the problem here: They found that Europeans who work in CBDs are about 20 percent less likely to drive, 10 percent more likely to use transit, and 10 percent more likely to walk/cycle than Americans who work in CBDs. But less than 10 percent of urban Americans work in central business districts, so whatever happens in those areas is irrelevant to most U.S. workers. In Portland, for example, 40 percent of downtown commuters took transit to work before COVID, while only 3.4 percent of commuters in the rest of the region relied on transit.

Even if the CBD data were valid when applied to urban areas as a whole, the latest census data show that just 3.7 percent of U.S. urban commuters are taking transit to work. Spending hundreds of billions of dollars on transit infrastructure to increase this by 10 percent, or to less than 4.1 percent in total, seems pretty pointless.

Moreover, infrastructure is not the only issue. Although European cities have their suburbs, they tend to be much denser overall than American cities. It would cost tens of trillions of dollars to rebuild American urban areas to be as dense as European ones, not to mention huge infringements on people’s property rights trying to force them to live in higher-density areas. Yet, if the Yale/UCSD data are believable, that would only reduce auto commuting by 20 percent. That’s ridiculous.

Despite the superior accessibility enjoyed by Americans, the researchers conclude by suggesting that cities consider investing more heavily in transit infrastructure rather than rebuild worn-out highways. Yet the costs of shifting to transit — both in the costs of new infrastructure and the loss of accessibility — are simply too great for this to be a viable or worthwhile option.