U.S. Per Capita Travel Was 18,000 Miles in 2021

Per capita travel in the United States was a little lower in 2021 than 2019, according to data recently released by the Bureau of Transportation Statistics (BTS). In 2021, the average American traveled 17,909 miles by air, auto, transit, and Amtrak, down 8 percent from 19,475 miles in 2019.

Photo by Ken Kistler.

To calculate that number, however, I had to correct an error in the recently released spreadsheet, which says that domestic airlines carried 172.8 billion passenger-miles in 2021, just 23 percent of the 754.5 billion miles they carried in 2019. This is an obvious typo, as another BTS table says that domestic airline flights carried 571.8 billion passenger-miles. (To view passenger-miles on that web site, click on “Revenue Passenger-Miles” at the top of the table.)

After making that correction, and comparing with Census Bureau population estimates for 2019 and 2021, we can calculate that the average American traveled 1,722 miles by airline in 2021, or 25 percent fewer than the 2,298 miles in 2019. This is only domestic flights; international passenger-miles were greater than domestic in 2019, but only 42 percent of domestic in 2021.

The change in driving was much smaller. Counting all “light-duty” vehicles (which the DOT divides into “long wheelbase” and “short wheelbase”), the average American traveled 14,000 miles by auto in 2021, 6 percent less than the 14,000 traveled in 2021. Even 14,000 miles is more than the per capita miles of travel by all modes of travel for any other country, before or after the pandemic.

Transit was down nearly 60 percent from 165 miles in 2019 to just 67 miles per capita in 2021. Amtrak declined more than 50 percent from 20 miles per capita in 2019 to 9 miles in 2021.

Transit agencies would be thrilled if they could get ridership back up to 165 miles per capita. But considering that is just 0.8 percent of total travel, I remain incredulous that anyone believes transit agency claims that transit was still a relevant form of transportation in the United States before the pandemic, much less after. Intercity passenger trains are even more irrelevant.

As I’ve noted before at length, I am skeptical of one row of numbers in table 1-40: that for “Bus” (which, as a footnote shows, means buses other than transit buses). The table indicates that non-transit bus travel declined from 362 billion passenger-miles in 2019 to 347 billion in 2021. But where are all these buses coming from?

Before the pandemic, the American Bus Association estimated that motorcoaches (the fancy buses used for intercity service) carried around 65 billion passenger-miles a year (which includes intercity buses, casino buses, tour buses, charter buses, and some commuter buses). Wendell Cox estimated that school buses carry around 50 billion passenger-miles a year. No one has an estimate for how many passenger-miles are carried by airport buses, but I can’t imagine it is close to the number carried by school buses. That leaves more than 200 billion bus passenger-miles in table 1-40 unaccounted for in 2019.

Federal Highway Administration estimates of highway passenger-miles come from thousands of vehicle counters located across America. The counters count vehicles and the agency has to somehow decide what share of those vehicles are heavy trucks, short- and long-wheelbase light vehicles, motorcycles, and buses. Whatever algorithm they use appears to overestimate buses, which means that automobile passenger-miles are probably undercounted. Those 200 billion ghostly bus passenger-miles equal about 10 billion ghost vehicle miles. If those were really automobiles, then auto passenger-miles would be about 17 billion greater than shown in the table, adding about 50 per capita miles a year to auto travel.

Table 1-40 has data for walking and cycling only for the years 1990, 1995, 2001, 2009, and 2017. These numbers are based on the National Household Travel Surveys that were conducted in those years. Using survey results, someone estimated miles of walking and cycling.

However, the survey only asked people about trips with destinations such as work or stores. Walking and cycling for exercise or recreation aren’t counted even though recreation travel by other modes is included in table 1-40. This means the numbers shown for walking and cycling in table 1-40 are significantly undercounted; I suspect triple the number shown is probably more accurate. Once recreation and exercise are added, walking alone probably accounted for more passenger-miles than urban transit in 2019 and certainly did in 2021.

The most recent household travel survey was done in 2022 and the DOT hasn’t published the results yet. When it does, it will be interesting to see how the pandemic changed walking and cycling, among many other forms of travel. I’m also curious to see if the pandemic changed average automobile occupancies, travel by people who work at home, and automobile ownership rates by income.