New analysis shows far more Twin Cities residents walking, biking to work

MPR News reporter Jon Collins covers our recent bike study: New analysis shows far more Twin Cities residents walking, biking to work. My quotes below, though most of  the work is due to PhD Candidate Jessica Schoner and Prof. Greg Lindsey. The full report will be published within a week.

Minneapolis’ recent investments in bicycling and walking infrastructure, such as trails and bike boulevards, may be one reason that rates in the city appear to have been increasing at such a rate, said David Levinson, a University of Minnesota professor of transportation and principal investigator of the study.

“If you want to attract people who want to bicycle, putting the facilities in will attract them,” Levinson said. “People who know they want to bicycle are more likely to live in Minneapolis than to move to the suburbs.”

Although women who bicycle took as many trips as men who bicycle, the proportion of women who bicycle even in Minneapolis is about one-and-a-half times smaller than men. Even though Minneapolis leads the region in bicycle-friendly development, Levinson said the gender gap may be due to the fact that the infrastructure still isn’t developed to the point where many people feel safe bicycling.

“Right now we allocate road space for moving cars, we allocate road space for storing cars, but very few places are reallocating road space for moving bicycles,” Levinson said. “Is moving bicycles more important than storing cars?

Elements of Access: Time is Mis-underestimated (Vierordt’s Law)

Reported trip durations by minute for work trips (2011 Twin Cities Travel Behavior Inventory) shows spiking on
Reported trip durations by minute for work trips (2011 Twin Cities Travel Behavior Inventory) shows spiking on “5s” and “10s” and “30s” (Log Scale)

Sometimes people think places are farther away than they really are, and other times they are closer. Freeways seem to take shorter than they really area, local streets longer. This in part has to do with task complexity, or the “mental transaction costs” involved in traveling.⁠1

When I need to make a lot of small driving and navigation decisions, like on a signalized route with lots of turns, I need to focus on driving more times. Each time I am engaging my conscious brain in traveling decisions. More brain-space is occupied by traveling thoughts.

Other factors include temporal relevance (is the trip important), temporal expectancies (what do I think the travel time will be) temporal uncertainty (how reliable is my estimate of travel time), affective elements (what is the emotional state of the traveler), absorption and attentional deployment (am I paying attention to the task at hand) and arousal (how physically activated am I, am I on drugs?).⁠2

When I can drive on an uncongested freeway, I can avoid many such thoughts. Driving is less salient. Time passes faster. As the expression goes, “time flies when you are having fun”

In travel surveys we have a common phenomenon of rounding reported times, and times are usually rounded up. So if a trip was 14 minutes, it would be rounded to 15 minutes. If it were 22 minutes, it might be rounded to 25 or even 30 minutes. This makes self-reported times significantly biased in travel analysis. Until recently, that was the only data available. But now with the advent of GPS devices and cheap sensors tracking traffic across networks, we can get much better speed and travel time estimates.

Vierordt’s Law claims people are more likely to over-estimate short times and under-estimate long times.

We did not corroborate this with a driving simulator study for waiting at a traffic signal.⁠3 Perceived and actual waiting time were virtually identical for the first 30 seconds, but for times greater than 30 seconds, actual waiting time was higher than perceived waiting time, up to 120 seconds. At 120 seconds, the trend was for perceived time to overtake actual time, but that was the cut-off for the experiment, so perception findings in this situation require more information. However, the annoyance level at 120 seconds of waiting was much higher than the annoyance of waiting 30 seconds. Further people hated stops.

Of course, with all of this, it depends on how you frame the question, what you ask, and what travelers were expecting. Comparing a computer-administered stated preference with one in which travelers were in a driving simulator completely flipped preferences for traveling (waiting for free flowing travel vs. muddling through in congestion).⁠4

1 Parthasarathi, Pavithra, David Levinson, and Hartwig Hochmair (2013) “Network Structure and Travel Time Perception.” PLOS ONE: 8(10): e77718.

2 Carrion, C. and D. Levinson (2012). Uncovering the influence of commuters’ perception on the reliability ratio. Technical report.

3 Wu, X., D. M. Levinson, and H. X. Liu (2009). Perception of waiting time at signalized intersections. Transportation Research Record: Journal of the Transportation Research Board 2135(1), 52–59.

4 Levinson, D., K. Harder, J. Bloomfield, and K. Carlson (2006). Waiting tolerance: Ramp delay vs. freeway congestion. Transportation Research Part F: Traffic Psychology and Behaviour. 9 (1), 1–13.

Vandalism or terrorism

David Levinson:

In light of Amtrak 188, this seems appropriate from England in 2007

Originally posted on Transportationist:

We were returning from Harrogate to London last night on the Great North Eastern Railway (GNER) after attending the well-run and interesting UTSG conference.
The trip from Harrogate to Leeds was uneventful. On the trip from Leeds to Kings Cross in London, we were interrupted by what I believe the announcer said was a Code 3 on Coach M. (We were on a different coach so at the time didn’t know what that was), though we did not stop there.
Later the announcer told us that there had been vandalism, a rock through a window, which needed to be repaired before we proceeded.
The coach was held at Peterborough for about 15 minutes while repairs were made to the broken window, and we arrived 15 minutes late. When we arrived, I could not locate the vandalism, so it must have been cleaned up fairly well.
Three observations spring to…

View original 621 more words

Budget Cuts to Blame? With Amtrak, Smaller May Be Better, Experts Say – US News

With the recent Amtrak crash, media has once again taken interest in the sorry state of passenger rail in the US.

I get quoted in an article by Alan Neuhauser for US News and World Report: With Amtrak, Smaller May Be BetterSlashing the rail service’s budget may be just the thing Amtrak needs. We talked for 17 minutes. My soundbite:

“Getting rejected by Congress might be the best thing for rail in the Northeast,” says David Levinson, a transportation engineering professor at the University of Minnesota. “It would have a short-term disruptive effect unless something else is put in place,” he allows, but “for real change, you have to change the governance: who’s getting the money, who’s spending the money, how much they’re getting.”

Biking and walking are on the rise in the Twin Cities

Throughout the past decade, the popularity and importance of biking and walking have increased significantly across the nation. Pinpointing exactly how people incorporate biking and walking into their travel behavior, however, is a persistent challenge.

In an extensive five-part research study sponsored by the Minnesota Department of Transportation (MnDOT) and the Metropolitan Council, U of M researchers are exploring a rich set of data generated through detailed travel behavior inventories in the Twin Cities region. The data allowed them to analyze changes in walking and biking behavior during the past decade.

“Overall, our team found that auto travel decreased between 2000 and 2010, while biking and walking increased during that time,” says Professor David Levinson, RP Braun/CTS Chair in the Department of Civil, Environmental, and Geo- Engineering (CEGE) and the study’s principal investigator.

Average bike and walk trips a day in Twin Cities

Bicycling grew from 1.4 percent of all trips in 2000 to 2.2 percent in 2010—an increase of 58 percent. Walking started with a larger share in 2000 (4.5 percent) and grew by a larger amount to 6.6 percent of trips in 2010—a 44 percent increase. One of the most important findings is that the actual bicycle mode share in the Twin Cities region is two to three times larger than reported by national data.

“Though the private auto still consistently dominates travel in the region, these relatively small mode shares translate to a substantial number of walk and bike trips on an average day in the region,” says Jessica Schoner, CEGE research assistant and lead author of the study. “We estimate about 12 million daily trips across the metro area, which means that on an average day people are making 190,000 bike trips and 735,000 walking trips.”

The researchers noted key differences between bicyclists and pedestrians and their walk and bike trips, including demographics, geography, trip purpose, and trip distance. For example, they found that while men and women choose to walk at an equal rate, a “gender gap” persists among bicyclists.

“Most of the growth in cycling came from increases in men commuting by bicycle, but closer examination of the gender gap in bicycling revealed some encouraging information,” Schoner says. “The gap appears to be in bicycling participation rates of men and women—there was no observed gap in the frequency of making bicycle trips among cyclists. This shows that programs encouraging women to try bicycling may help to further boost bicycling mode share.”

New bike infrastructure appears to have played a key role in the increase of bicycle mode share in the past decade. “While infrastructure was a significant factor in predicting bicycling in 2001, by 2010 the quantity of bicycle lanes around the home no longer differentiated bicyclists from non-bicyclists,” Schoner says. “This suggests that the Twin Cities’ expanded bicycle infrastructure has created pervasive and easy access throughout the city.”

Greg Lindsey, professor in the Humphrey School of Public Affairs and MnDOT’s first Scholar-in-Residence, is a co-author of the report. He was also the principal investigator of research conducted under MnDOT’s Bicycle and Pedestrian Counting Initiative.

“MnDOT is developing a program to monitor bicycle and pedestrian traffic that is modeled after the motor vehicle count program,” says Lisa Austin, MnDOT bicycle and pedestrian planning coordinator. “When used together, the traffic volumes and the travel diaries provide a clearer picture of how, when, and why people are traveling. This helps MnDOT and other agencies plan for better systems and analyze safety.”

The first project completed under the five-part study examined how changes in the accessibility of destinations have altered travel behavior in the past 20 years (see article in the February 2015 Catalyst). Additional parts of the study will look at the effect of transit quality of service on people’s activity choices and time allocation, changes in travel behavior by age group, and transportation system changes. The Catalyst will feature coverage of these projects as they are completed.

Related Links

Reprinted from CTS Catalyst.

Where can you walk to work? Study ranks cities for accessibility to jobs by walking 

When we think of the cities that are “most walkable,” it’s often in terms of pedestrian networks such as sidewalks and trails. However, there’s another factor at play in determining which cities are truly most pedestrian-friendly: accessibility.

In a new report in the Access Across America series from the U of M’s Accessibility Observatory, researchers ranked accessibility to jobs by walking in the 50 largest (by population) metropolitan areas in the United States. Currently, walking is the third most widely used commute mode after driving and transit, accounting for around 2.8 percent of commute trips nationally and 5 percent within large cities.

The rankings were determined using a weighted average of accessibility, giving a higher weight to closer jobs. Jobs reachable within 10 minutes were weighted most heavily, and then were given decreasing weights as walking travel time increased up to 60 minutes.

Top 10 metro areas: job accessibility by walking (2014)

  1. New York
  2. San Francisco
  3. Los Angeles
  4. Chicago
  5. Washington
  6. Seattle
  7. Boston
  8. Philadelphia
  9. San Jose
  10. Denver

The report—Access Across America: Walking 2014 —presents detailed accessibility values for each metropolitan area, as well as block-level maps illustrating the spatial patterns of accessibility within each area.

“Our report provides a snapshot look at how much economic opportunity exists within a reasonable walking distance in metropolitan areas,” says Andrew Owen, director of the Accessibility Observatory. “In addition, it offers a framework to further our understanding of accessibility by walking. For example, not all jobs are the same—some are higher paying, some are lower skilled, and jobs exist in a variety of industries.”

The cities that make up the top 10 all have a combination of employment density and fast, frequent transit service. “Employment density is a primary factor in whether an area is economically walkable or not, and employment densities are typically high in cities that employ heavy rail systems leading into a central core,” Owen says. “In this way, transit systems promote walkability and walking accessibility, without requiring a user to board a transit vehicle to experience the positive effects.

The research team included Professor David Levinson, RP Braun/CTS Chair in the Department of Civil, Environmental, and Geo- Engineering (CEGE), and graduate research assistant Brendan Murphy. The research will be useful for other studies, such as a Roadway Safety Institute project now under way, Murphy says. “It gives us a basis for estimating pedestrian traffic in places where we don’t have good data.”

The research was sponsored by CTS and CEGE. The new report is a companion study to previous Access Across America reports on transit and auto accessibility. Reports and accessibility maps from the latest study are available on the Accessibility Observatory website.

Accessibility map

Related Links

Reprinted from CTS Catalyst.

For whom is transit?

In the May 11 Finance and Commerce, Matt Kramer, a local Chamber of Commerce representative lobbying for additional public transit and transportation spending (currently being debated at the Minnesota Legislature) is quoted as saying “Every person who is riding transit is one less person in the car in front of us,”

This is a fascinating quote. First is the use of “us”. So the Chamber of Commerce (probably correctly) identifies riding transit as something someone else does (since “we” are still in the car). And goes on to imply that it benefits us because there will be fewer cars. (Actually he says fewer people per car, but I think he meant fewer cars, not that it would reduce carpooling). And I suppose he could mean he rides the bus, and the car in front has fewer people (or there were fewer cars in front), but I don’t think that’s what he meant, since the arguments in the legislature are mostly about building  and operating new facilities – such as LRT lines or freeway BRT, rather than supporting existing buses driving in traffic.

This evokes the famous Onion article: Report: 98 Percent Of U.S. Commuters Favor Public Transportation For Others.

But it also suggests transit reduces auto travel. The converse is almost equally true, building roads reduces transit crowding. But that is not an argument road-builders make. [It is an argument urbanists make against roads.]

Of course, some transit users would have otherwise driven, but many would have been passengers in cars, walked, ridden bikes, or telecommuted. No one really knows what the alternative untaken mode would be. We have models, but the form of those models dictates the answer. Logit models, which are widely used by travel demand forecasters to predict mode choice (and whose development resulted in an Economics prize for University of Minnesota graduate Daniel McFadden), have the property called “IIA”, which is short for Independence of Irrelevant Alternatives. In short, if you take away a mode, IIA means people choose the other modes in proportion to their current use. So let’s say there are 3 modes: walk 25%, transit 25%, drive 50%, and there is a transit shutdown (like in 2004). IIA implies the 25% of former transit users would split 1/3 (25%/75%) for walk and 2/3 (50%/75%) for driving. We all know that is not true, and there are various techniques to try to fix the models and use more complicated functional forms. But the question of what is true is not at all clear.

While there are surveys that have answered those questions, they are all context specific. For instance, Googling turns up a Managed Lanes Case Study report:

95 Express bus riders were asked how long they have been traveling by bus and what was their previous mode of travel before using the bus service. 92 percent of respondents (307 out of 334) mentioned they have been traveling the 95 Express bus before the Express Lanes started. Only, 8 percent respondents (27 out of 334) began using the bus after the Express Lanes opened. Among them, 50 percent (13 out of 27) had their previous mode as drive alone and none of them carpooled previously. Therefore, 95 Express bus ridership consisted primarily of those who have been using the service prior to Express Lanes implementation and the small mode shift from highway to transit was mostly from SOVs. Note that the number of respondents is too small to make any conclusions (Cain, 2009).

Undoubtedly other services would have different numbers. But mostly transit lines are not a direct substitute for driving.

The line of reasoning in the opening quote suggests the primary purpose of transit is reducing auto travel, rather than serving people who want to, or have to, use transit. In other words, building transit is good because it reduces traffic congestion (and almost no one argues building roads is good because it reduces transit crowding).

That is at best a secondary benefit, a benefit which could be achieved must more simply and less expensively through the use of prices as we do with almost all other scarce goods in society, even necessities like water.

Transit today is, in almost all US markets, slower than driving. People who depend on transit can reach fewer jobs than those who have automobiles available. Some people use transit by choice, for instance to save money (if they need to pay for parking), and the rest without choice. In my opinion, it is more important to spend scarce public dollars to improve options for those without choices than to improve the choices for those who already have alternatives. Perhaps ideally we could do both, in practice, one comes at the expense of other.

The idea that transit is for the other person is true for the 95.5% of people who don’t use transit regularly. But it warps thinking that the aim of public transit funding is to benefit those non-transit users.

Cross-posted as “Who Benefits from Other Peoples’ Transit Use” at

Regulating Aerial Drones: An Undiscovered Country

I tweeted with Randall Munroe’s XKCD

Steve Crandall (who you may know better as @tingilinde ) writes in:

Actually Randall may not be far off in this one.  The definition of air rights was left poorly defined by the Supreme Court in the 40s and FAA rules, whatever form they take, won’t be universal remedies.

I sometimes sit on my town’s technology board – most recently during the last cable tv license renewal and when there was a chance for a fiber overbuild.  We have quite a few ex Bell Labs types in town who show \ interest in society and technology issues so I threw the match into the kindling by convening a township drone policy committee in the wake of Amazon’s announcement.

It turns out there was a lot of interest. People started talking about various restrictions to make the town safe from photography, crashes, and other real and imagined nuisances. One of the proposals was allowing people to define their own air-rights over their territory up to the FAA limits.  Any crossing would constitute trespass and they would be allow to impound the drone and require a release fee. Most felt the drones would have to fly over the street, there would have to be proof that video from the drone was not recorded, some level of quiet would be required, and the drone would not be autonomous – a FAA licensed pilot would be required for each flight with no multi-tasking.  Speed limits, weight limits, noise limits and an insurance requirement.   I stress this was a very tech savvy group.

Some communities in the West have passed ordinances that allow people to shoot them down and one offers a bounty.  Ag-gag laws in many states would prevent the operation of drones in rural areas.

It may well be a mess until uniform code exists.  That took a few decades for commercial aviation.  I can’t see it taking much less than a decade this time around, but who knows…

People saw a need for regulation and invented their own. I think there is the potential for an explosion of questions until air rights are firmly settled. The noise issue was very contentious

So maybe drones will have to travel over public rights-of-way at below some height limit. Does this decrease safety (more likely to crash if confined to crowded drone lanes, more likely to fall on cars)?

Elements of Access: One-Way Streets

by Wes Marshall.



Why do traffic engineers seem to like one-way streets so much? The AASHTO Green Book points out a handful of efficiency advantages [1]. By removing the delay caused by left-turning cars, we increase traffic capacity and speed. Fewer intersection conflicts means more efficient signal timings and, in theory, fewer and less severe crashes (e.g. by eliminating head-on crashes). Medians are no longer necessary, so you can often fit in an extra lane of through traffic, which further increases capacity and speed. More mobility with better safety? What’s not to love?

Beyond the abundant advantages, AASHTO lists a few disadvantages as well. There is the potential for increased travel distances in cases when you have to travel almost around a whole block to reach your destination. When all lanes begin to back up at traffic lights, emergency vehicles may be blocked. Lastly, one-way streets may confuse visitors.

Given AASHTO’s list of pluses and relatively few minuses, it makes sense why so many of our streets send traffic in just one-direction. Then again, it’s not hard to argue that what AASHTO deems an “advantage” might be the opposite. If I lived or worked on a one-way street, I’d be pretty hard-pressed to believe that more cars moving at higher speeds is necessarily a good thing.

Appleyard’s early studies found many residential livability advantages on the two-way streets, but the one-way street he investigated had far more traffic than the two-way comparisons [2]. Denver converted a handful of one-way pairs to two-way operation in the early 1990s and found that residents preferred the change [3]. A recent case study out of Louisville looked at a handful of one-way to two-way conversions and found significant increases in pedestrian traffic, property values, and business revenue [4]. These benefits were accompanied by a significant decrease in crime. Other cities such as Charleston, SC and Lubbock, TX also found success in terms of two-way streets helping downtown revitalization [5, 6].

Such livability benefits are all well and good, but are they worth the increased road safety risks that AASHTO made clear? The research is beginning to suggest that the safety answer isn’t clear cut. Lubbock found no significant change in terms of traffic volumes or safety [6]. Another study from Jerusalem also found no difference in road safety [7]. Despite similar traffic levels on the Louisville conversions, crash rates dropped with the two-way streets [4]. Moreover, child pedestrian injury rates on one-way streets have been found to be more than double the rates on two-way streets [8].

More research is needed on the safety outcomes. However, it is also interesting to ask why the safety benefits of one-way streets be overblown. First, there are likely to be differences in driver behavior, most notably with faster speeds on one-ways. It is pretty easy to understand see why slower traffic – despite the noted increase in conflict points – might help reduce crash severity. Another ITE guide even says the following regarding the safety of one-way streets: “one-way pairs with good signal progression and high travel speeds seemed to elicit red-light running behavior” [9].

The image below comes from the ITE Traffic Engineering Handbook. It makes the case for better safety on one-ways by depicting the number of conflict point at an intersection for a two-way street as 32 and for a one-way street as only 5. This is a stark difference that could theoretically result in better safety. Beyond the fact that conflict points are not often well correlated with actual safety outcomes [10], the bigger issue is that they are comparing apples and oranges. This diagram compares an intersection where all four-legs have two-lanes to an intersection where all four-legs have one-lane. In reality, the one-way streets would have at least two lanes, if not three as in the image from Denver above or in cases where the median is removed. One-way streets with multiple lanes is a fairer comparison that would substantially increase the number of potential conflict points and deem the comparison in the image below as meaningless.

Moreover with regard to conflicts, AASHTO even suggests converting from two-way to one-way operation in situations where an urban street has too many pedestrian-vehicle conflicts [1]. The reduction in pedestrian-vehicle conflicts is supposedly derived from a simpler set of intersection movements. The real reason for the reduction of pedestrian-vehicle conflicts might be even simpler: fewer pedestrians wanting to cross the street in the first place.

So after all that, the only definitive advantage left for one-way streets is increased traffic capacity. However, this point is also up for debate. Taking into account the decreased accessibility of circuitous routes, one paper found that drivers make significantly more turning movements and travel greater distances given the same origins and destinations in a network dominated by one-way traffic patterns [11]. Another more recent paper suggests that a network of two-way streets actually has a greater trip-serving capacity – particularly for trips less than 5 miles – as compared to a network of one-way streets [12]. When also prohibiting left-turns in the two-way network, this capacity advantage of the two-way network included longer trips as well.


Not only do one-way streets hinder accessibility and livability, but the traffic engineering benefits don’t necessarily seem to hold. While one-way streets are still needed when relatively narrow cross-sections prevent two-way traffic, in most other urban contexts, it is hard to imagine why so many cities continue to preserve one-way streets. Some cities are changing their ways. The before-and-after images at the top are from Larimer Street in Denver where a mile-long stretch was recently converted from a one-way into a two-way. Instead of three high-speed lanes heading toward downtown, there is now one lane in each direction with accompanying bike lanes. With noticeably slower traffic and more active transportation use along this corridor, it makes sense why there so many new businesses seem to be popping up, especially when compared to the parallel streets that remain focused on one-way traffic. It might be time for cities to find a new direction – and more research is needed – but it seems like this this new direction will run both ways.



  1. AASHTO, A Policy on Geometric Design of Highways and Streets. 2011, Washington D.C.: American Association of State Highway and Transportation Officials.
  2. Appleyard, D. and M. Lintell, The Environmental Quality of City Streets: The Residents’ Viewpoint. Journal of the American Institute of Planners, 1972. 38(2): p. 84-101.
  3. Robert Dorroh, I. and R. Kochevar, One-Way Conversions for Calming Denver’s Streets, in ITE International Conference. 1996, Institute of Transportation Engineers: Las Vegas, NV.
  4. Riggs, W. and J. Gilderbloom, Two-Way Street Conversion: Evidence of Increased Livability in Louisville. Journal of Planning Education & Research, forthcoming.
  5. Baco, M.E., One-Way to Two-Way Street Conversions as a Preservation and Downtown Revitalization Tool: The Case Study of Upper King Street, Charleston, South Carolina, in Historic Preservation. 2009, Clemson University and College of Charleston: Clemson, SC.
  6. Hart, J., Converting Back to Two-Way Streets in Downtown Lubbock. ITE Journal, 1998. August.
  7. Hocherman, I., A.S. Hakkert, and J. Bar-Ziv, Safety of One-Way Urban Street. Transportation Research Record, 1990. 1270: p. 22-27.
  8. Wazana, A., et al., Are Child Pedestrians at Increased Risk of Injury on One-Way Compared to Two-Way Streets? Canadian Journal of Public Health, 2000. 91(3): p. 201-206.
  9. ITE, Toolbox on Intersection Safety and Design. 2004, Washington, D.C.: Institute of Transportation Engineers.
  10. Jacobs, A.B., E. MacDonald, and Y. Rofe, The Boulevard Book: History, Evolution, Design of Multiway Boulevards. 2003, Cambridge, MA: The MIT Press.
  11. Walker, G.W., W.M. Kulash, and B.T. McHugh, Downtown Streets: Are We Strangling Ourselves on One-Way Networks? Transportation Research Circular, 2000(501).
  12. Gayah, V.V. and C.F. Daganzo, Analytical Capacity Comparison of One-Way and Two-Way Signalized Street Networks. Transportation Research Record, 2012. 2301: p. 76-85.
  13. ITE, Traffic Engineering Handbook. 5th ed. 1999, Washington, DC: Intstitute of Transportation Engineers.