Are any cities becoming less car-oriented?

The CityNerd Youtube channel has a video on the ten U.S. cities “that are becoming most city-like.” One of Citynerd’s criteria grabbed my attention: he gives credit for cities having an increase in “car-lite households”, by which he means having one or no cars. But I didn’t really think this element makes sense: if a city has smaller households (due to plunging marriage or birth rates) more households will be one-car households even if almost everyone has a car. For example, in Washington, DC the average household size decreased from 2.24 in 2010 to 2.02 in 2024, and in Miami the average household size decreased from 2.68 to 2.37.

So I decided to replicate his analysis but for no-car households: that is, for the ten (mostly growing) cities in his video, did the percentage of households without cars increase or decrease? For seven out of CityNerd’s top ten cities, the percentage of carless households actually went down: the largest declines were in Boston (where 36.5 percent of households lacked cars in 2010, as opposed to 34 percent in 2024), Miami (decline from 20.9 percent to 18.6 percent) and Atlanta (decline from 17.5 to 14.9). So at first glance, one might say: “So much for peak car! Even in Boston more people have cars.”

However, there is a complicating factor: these data are based on surveys that, according to the Census Bureau itself, have a margin of error. For example, Miami’s vehicle ownership estimates are subject to a 1.7 (2010) or 1.8 (2024) margin of error, which means that the percentage of carless households could have been anywhere between 19.2 percent and 22.6 percent in 2010, and between 16.8 and 20.6 percent in 2024. As a result, I am not sure Miami’s changes were statistically significant. Data for the other cities in CityNerd’s video also had a margin of error (usually in the 1-2 percent range; for example, Atlanta and Boston had error margins of 1.5 percent in 2010, making their changes equally questionable). So I’m not sure the changes reflected in Census data are real.

What about the three cities where car ownership apparently decreased? There are three such cities: Madison, Washington and Seattle. In two of these cities, changes were clearly within the margin of error: in Madison, the carless percentage increased by less than half a percentage point (13.3 to 13.7) and in Washington, DC, the carfree percentage increased from 35 to 36.5, well within the 1.5 point margin of error). The only possible exception is Seattle, where the percentage of car-free households increased from 15.7 percent to 19.8 percent. With a 1 to1.5 percent error margin (the former in 2010, the latter in 2024), that means that Seattle’s car ownership rate was probably between 13.7 and 16.7 percent in 2010, and between 18.3 percent and 21.3 percent in 2024. Thus, it seems to me that the number of Seattleites living car-free almost certainly increased; as for the other nine cities, any changes are so minor as to be undetectable.

PS You might be wondering “what about long-run trends?” I looked at 1990 Census data, and found that the answer seems to depend on how car-dependent the city is. In Seattle, 16.7 percent of households were carless- fewer than today. In the most transit-oriented cities on the CityNerd video (Washington and Boston) car ownership patterns were not significantly different from today: the percentage of car-free households in 1990 was 37.4 percent in Washington and 38.3 percent in Boston. But in some of the more car-dependent cities, far fewer people owned cars in 1990: in Miami, 28.9 percent- and in Atlanta 28.6 percent, almost twice as many as today,