Top 10 reasons to take TTI’s Urban Mobility Report with a grain of salt

By Daniel Hertz, on City Observatory

Since the Texas Transportation Institute released its 2015 “Urban Mobility Report,”urban transportation experts and advocates have unleashed thousands and thousands of words poking holes at its methodology, assumptions, and political agenda. (We’ve pitched in our fair share of those words, and perhaps more.)

Not a kind of urban mobility you'll find in the Urban Mobility Report. Credit: Chad Kainz, Flickr
Not a kind of urban mobility you’ll find in the Urban Mobility Report. Credit: Chad Kainz, Flickr


As one last entry to this conversation, we wanted to put together a sort of Cliffs Notes version of the problems with the UMR, skimmable by people who just want a quick overview of the issues, without all the background. So without any further ado, the top ten reasons to be skeptical of the “Urban Mobility Report”:

  1. Both TTI and media outlets claim that the UMR proves traffic is worse than at any time since 1982—but a major methodological change in 2009 makes comparisons before that date completely invalid.
  2. The UMR, in the words of Victoria Transportation Policy Institute executive director Todd Litman, “ignores basic research principles,” including failing to allow other experts to review its data and findings—the sort of “peer review” that is foundational to social science investigations.
  3. As the Brookings Institution’s Rob Puentes notes, the UMR measures mobility, not access. That is, cities get high scores if you can drive really fast—not if you can actually reach jobs or amenities in less time.
  4. In many cases, the UMR counts an inability to drive faster than the speed limit as “congestion.”
  5. Since 2009, the UMR has used traffic data from Inrix—but their reports tell a very different story than the numbers that Inrix reports publicly, which suggest that congestion has actually decreased, while the UMR claims that congestion hasincreased. TTI hasn’t explained how it reached the opposite conclusion as the organization that provides its data.
  6. Despite claims that they favor a mix of measures including transit and more density, the TTI’s travel time index actually penalizes cities that undertake measures that shorten work trips. The TTI scorecard perversely incentivizes sprawl and ignores the costs associated with longer trips.
  7. The UMR’s prediction that traffic will get much worse in the coming years is based on a model that simply pretends the last decade didn’t happen.
  8. In fact, the US reached “peak driving” in 2005, and the average number of miles driven per day has fallen by nearly 7 percent since then.
  9. The UMR says that adding road capacity will reduce congestion—but 92 of the top 99 places where congestion increased according to the UMR increased their roadway miles per capita. Building more roads won’t reduce congestion or improve access without land use changes and other transportation investments.
  10. The “Urban Mobility Report” is in fact a report just about driving. Public transit, walking, and biking—essential parts of the transportation network in any city—arealmost entirely left out. If you walk, bike or take transit you simply don’t exist in the eyes of TTI.

City Observatory’s “Cappuccino Congestion Index.” And previous editions have suffered from methodological flaws. Joe Cortright’s post at City Observatory is a great one-stop shop for the main critiques. – See more at:

Taking TTI’s findings at face value, this year is no exception. Driving is on the increase (though likely more as a result of crashing oil prices than economic growth). And TTI’s figures show that congestion did get worse in most cities last year compared with 2013.  But over a longer time horizon, stretching back to the pre-recession years of the early 2000s, TTI’s numbers show that congestion has hit something of a plateau. Assuming that data from the early years of the TTI report can be accurately compared to more recent figures (something Cortright questions), the average number of hours spent in traffic by an auto commuter is no higher than it was in 2006.

Annual Delay per Auto Commuter (data: Texas Transportation Institute)

The marginal increase in congestion in 2014 is certainly worth noting, but where things get really interesting is when TTI attempts to forecast the impacts of congestion in the future by simply pretending that the last decade didn’t happen, basing its projection instead on trends that prevailed from 2000 to 2005 (the years highlighted in red in the chart above).  The authors explain it this way: “The period before the economic recession (from 2000 to 2005) was used as the indicator of the effect of growth. These years had generally steady economic growth in most U.S. urban regions; these years are assumed to be the best indicator of the future level of investment in solutions and the resulting increase in congestion for each urban area.”

It’s hard to be diplomatic about how wrongheaded this is. The period TTI uses as its best guess of what is likely to happen over the next five years 1) ended a decade ago; 2) was characterized by consistent economic growth that was 3) driven in large part by a massive speculative exurban housing bubble; 4) predated the smartphone, Uber and even (until 2004) Facebook; and occurred 5) during a time when the Highway Trust Fund was still solvent.  If you want to argue that gas prices will remain low and the economy reasonably robust forever, fine. But it’s not legitimate to simply assume that America will magically be whisked back to 2002 again.

We do need to give TTI credit on one score. In previous editions of the report, TTI has promoted capacity expansion – especially highway expansion – as the primary antidote to congestion, relying at times on ill-considered and out-of-context analysis to make its point. Today, however, TTI has seemingly recognized that you can’t simply build your way out of congestion. While the report still puts too much emphasis on capacity expansion, it now advocates for a “balanced and diversified” approach to addressing congestion that includes a wide variety of tools. It even acknowledges that, for some areas, congestion might not be seen as a top priority problem.

Which brings us back to the fundamental issue with TTI’s approach to measuring and talking about congestion: context. The $160 billion “price tag” for congestion estimated by TTI is a big number, but it pales in comparison with other costs of automobility, including the $836 billion in economic losses and societal harm caused by motor vehicle crashes each year, the $411 billion households spend on cars annually and $328 billion we spend on fuel to run them[1] — much less the costs imposed by air pollution and transportation’s contribution to global warming. Congestion is just one of many problems caused by our car-dependent transportation system, and not always the most important or the most costly one. The best responses to it are likely to be solutions that address many of the costs and dangers of transportation all at once.
[1] Based on 2013 Consumer Expenditure Survey data (PDF)