INRIX has released its 2018 traffic congestion numbers for more than 200 urban areas around the world. Unfortunately, the company changed its methodology from previous years, so the numbers aren’t comparable. It also isn’t clear how INRIX ranks congestion.
For example, the INRIX web page notes that, “In 2018, Bogota drivers lost 272 hours due to congestion — more than any other city in the world.” Yet Bogota is ranked number three behind Moscow (where drivers lost 210 hours) and Istanbul (where drivers lost only 157 hours).The only other data offered for ranking congestion is the speed of driving in the inner city: in Bogota it was 7 mph compared with 11 in Moscow and 10 in Istanbul. So if Bogota is worse on both criteria, why is it ranked only number 3?
The INRIX report only says, “The 2018 Global Traffic Scorecard not only analyzes time lost, but also the severity of congestion.” Yet I don’t find any measures of “severity” anywhere in the report.
(The report isn’t even clear how its main unit of congestion, “hours lost,” is measured. It it hours lost per capita? Per commuter? Per driver? Other documents indicate it is per driver, but nothing in the report itself says so.)
The report does have a good explanation of congestion. “The critical phenomena known as ‘facility breakdown’ is frequently underappreciated,” says the report. “It occurs when a road cannot effectively accommodate more vehicles, which causes a decrease in the roadway’s overall capacity as more vehicles try to force their way onto the roads.” Freeway lanes can move 2,300 vehicles per hour in free-flowing traffic, yet “the same roadway may carry fewer than 700 cars per lane during facility breakdown.”
This is why congestion pricing is so valuable: by making sure that highways never experience “breakdown,” congestion pricing effectively doubles or triples their capacity during peak periods. The report even mentions Singapore, whose “road network continues to facilitate high-speeds despite high urban density” due to congestion tolls. When measured by hours lost per driver, Singapore is ranked 106 (yet for some reason it is ranked 14 overall).
Unfortunately, the report undermines its breakdown argument by praising other cities that “actively pursue policies that reduce roadway performance and capacity in favor of public transport, cycling and walking.” It calls these the “most progressive” cities even though their policies effectively reduce urban mobility by forcing people to accept terrible traffic congestion or to switch to even slower modes of travel. (Of course, it’s remotely possible that the report uses “progressive” as a pejorative.)
Of the 220 urban areas evaluated worldwide, only three from the United States made it into the top 25, and none were New York or Los Angeles. Instead, Boston was ranked number 8 (even though it was 25 in hours lost per driver), Washington was ranked 19th (but 36th in hours lost), and Chicago was ranked 23rd (but 64th in hours lost). New York was ranked 40th (70th) and Los Angeles 47th (76th). At 58, Seattle was ranked well below New York and Los Angeles, yet it was just 61st in hours lost and its average inner-city speeds were just 10 mph compared with 14 in Los Angeles and almost as bad as New York’s 9.
Among American cities, the real anomaly is Seattle. When measured in delay per driver, it has the third-worst congestion after Boston and Washington. When measured by average inner-city speeds, it is tied with San Francisco and Philadelphia for the the second-worst, after New York. So why does INRIX rank Seattle’s congestion as only the sixth-worst in the United States?
According to the complete data set issued with Texas Transportation Institute’s last major urban-mobility study, Seattle in 1982 suffered just 11 hours of delay per commuter and Boston had just 15. Eventually, by TTI’s measure, Seattle rose to 56 hours and Boston to 64 hours per commuter.
Now INRIX estimates Seattle drivers waste 138 hours per year and Boston drivers waste 164. It is difficult to say how much this growth is due to the the change in measures — per commuter vs. per driver — versus worsening congestion, but since most commuters are drivers (68 percent on Boston, 71 percent in Seattle), most of the difference must be from increased congestion. Is this large increase because Boston and Seattle are among the nation’s most progressive cities?
That may be. Though Dallas-Ft. Worth — which few would consider “progressive” — is the nation’s seventh-largest urban area, it is ranked as having 20th-worst congestion (22nd when measured by hours of delay, tied for 25th by average inner-city speeds). Meanwhile, “progressive” Austin is a much smaller urban area yet is ranked 13 in hours lost, losing 36 percent more hours per driver than Dallas-Ft. Worth.
Columbus has some of the lowest transit ridership in the country — just 12 trips per capita in 2017 compared with a national average of about 38 — is ranked as having the 25th-worst congestion out of the 25 urban areas ranked. Indianapolis, whose transit system carried just 6 trips per capita, isn’t even on the list, which indicates it has even less congestion than Columbus.
In fact, for the 25 U.S. urban areas included in the INRIX report, there is a fairly strong correlation — 0.58 — between hours lost and per capita transit trips — the more transit usage, the more hours lost. To a large extent, this is because “progressive” urban areas that emphasize transit have deliberately made congestion worse to discourage driving.
This should be the real lesson of the congestion rankings. Urban areas might not be able to completely build their way out of congestion, but these data show they can choose between trying to minimize congestion or making it worse. Those that have tried to make it worse have succeeded, but in what universe is deliberately wasting people’s time considered a good thing?
This piece first appeared on The Antiplanner.
Randal O’Toole (rot@ti.org) is a senior fellow with the Cato Institute and author of the new book, Romance of the Rails: Why the Passenger Trains We Love Are Not the Transportation We Need, which was released by the Cato Institute on October 10.
Email this Story
