A bifurcation of the peak: New patterns of traffic peaking during the COVID-19 era

Recently published:

  • Gao, Yang and Levinson, D. (2022) A bifurcation of the peak: New patterns of traffic peaking during the COVID-19 era. Transportation. [doi]

This paper analyzes the emergence of two well-defined peaks during the morning peak period in the traffic flow diurnal curve. It selects six California cities as research targets, and uses California employment and household travel survey data to explain how and why this phenomenon has risen during the pandemic. The final result explains that the double-humped phenomenon results from the change in the composition of commuters during the morning peak period after the outbreak.

Traffic flow diurnal curve of Los Angeles in 2020
Traffic flow diurnal curve of Los Angeles in 2020

Eleven Sydney council areas where pedestrians die more than motorists

Nigel Gladstone in the Sydney Morning Herald wrote “Eleven Sydney council areas where pedestrians die more than motorists“. I got some choice quotes:

Professor of transport at the University of Sydney David Levinson said changing the way Sydney’s traffic signals give priority to cars over pedestrians in busy areas was one way to stem the flow of injuries.

“This inequality [in traffic light phasing] undermines many of the stated goals of transport, health and environment policy,” Professor Levinson said. “Creating an environment that is better for pedestrians with separated footpaths, easy and frequent safe road crossings, generally slower cars and trucks, better trained and more law-abiding drivers (via police enforcement) will reduce the likelihood of fatalities.”

Road deaths have been increasing in both NSW and the US recently, while most other OECD nations are reporting fewer fatalities.

One theory [Referring to: Chi, Guangqing, Jeremy Porter, Arthur Cosby and David Levinson (2013) The Impact of Gasoline Price Changes on Traffic Safety: a Time Geography ExplanationJournal of Transport Geography 28 1-11. [doi] ] is that as more people can afford to be on the road, drivers with less experience, who tend to be younger, and therefore more dangerous, are added to the mix of motorists.

“When there’s economic expansion, people are working more hours and they probably get a bit more aggressive,” Professor Levinson said.

Research [by Wes Marshall] comparing Australian and American drivers found the rate of fatalities was more than twice as high in the US, where more than half of drivers do not stop or yield to pedestrians at crossings.

In Hawaii’s capital Honolulu, fines for pedestrians who text while crossing the road at traffic lights began this year.

In the City of Sydney, one in three people crossing the road is using a mobile phone and it’s time pedestrians “start owning this problem as well”, Pedestrian Council chief executive Harold Scruby said.

“Pedestrian deaths and serious injuries are going through the roof,” Mr Scruby said. “There is nothing that we’re seeing that the government is doing to help pedestrians.”

Hitting pedestrians with a $200 fine for using a mobile phone while crossing the road, even on the green man phase, is on Mr Scruby’s agenda.

“There’s no barrier there just because the light’s green. Half of the drivers coming towards you are on the phone too,” Mr Scruby said. “If you’re hit as a pedestrian, the driver will be automatically drug and breath tested but that’s a box they [police] have to tick, no one then goes looking for the mobile phone, there’s no box to tick.”

Professor Levinson said fining pedestrians was “basically a form of victim blaming”.

“Distracted pedestrians don’t kill drivers or passengers. Distracted drivers kill pedestrians,” he said. “Deaths are due to high speed and high mass, and drivers of two-tonne machines have an obligation to be more alert.”

In the three years to 2017, one pedestrian was killed and 25 were seriously injured while distracted by a mobile phone. But this is likely to be an under-reported issue, as it relies on witnesses telling police and other forms of evidence.

There are no plans to introduce penalties for people using mobile phones while crossing roads in NSW, a Transport for NSW spokesman said.


I also got to be ABC Radio Illawarra Friday Morning (talking about traffic safety), and ABC Radio Adelaide Friday Afternoon (talking about Beg Buttons) (first few minutes of this).

When pushing the pedestrian button works and when it doesn’t | SMH

Nigel Gladstone of The Sydney Morning Herald answers: When pushing the pedestrian button works and when it doesn’t.

(The answer is it doesn’t work in the CBD during the day).

My quote:

Monetizing the beg button

Professor David Levinson from the School of Civil Engineering at the University of Sydney said traffic signals in the city should be shifted to be more pedestrian friendly to encourage more walking.

“Traffic signals give priority to motor vehicles over pedestrians. This inequality undermines many of the stated goals of transport, health and environment policy,” Professor Levinson said.

“Sydney uses adaptive signals so that they’re designed to maximise the throughput for cars and so they’ll extend the green light for cars but that results in there being more ‘don’t walk’ time for pedestrians.”

Spatiotemporal Short-term Traffic Forecasting using the Network Weight Matrix and Systematic Detrending

Recent working paper:

LookBackWindowsThis study examines the dependency between traffic links using a three-dimensional data detrending algorithm to build a network weight matrix in a real-world example. The network weight matrix reveals how links are spatially dependent in a complex network and detects the competitive and complementary nature of traffic links. We model the traffic flow of 140 traffic links in a sub-network of the Minneapolis – St. Paul highway system for both rush hour and non-rush hour time intervals, and validate the extracted network weight matrix. The results of the modeling indi- cate: (1) the spatial weight matrix is unstable over time-of-day, while the network weight matrix is robust in all cases and (2) the performance of the network weight matrix in non-rush hour traffic regimes is significantly better than rush hour traffic regimes. The results of the validation show the network weight matrix outperforms the traditional way of capturing spatial dependency between traffic links. Averaging over all traffic links and time, this superiority is about 13.2% in rush hour and 15.3% in non-rush hour, when only the 1st -order neighboring links are embedded in modeling. Aside from the superiority in forecasting, a remarkable capability of the network weight matrix is its stability and robustness over time, which is not observed in spatial weight matrix. In addition, this study proposes a naïve two-step algorithm to search and identify the best look-back time win- dow for upstream links. We indicate the best look-back time window depends on the travel time between two study detectors, and it varies by time-of-day and traffic link.

The End of Driving — Money and Politics Podcast

The End of Traffic and the Future of Access: A Roadmap to the New Transport Landscape. By David M. Levinson and Kevin J. Krizek.
The End of Traffic and the Future of Access: A Roadmap to the New Transport Landscape. By David M. Levinson and Kevin J. Krizek.

My podcast on the Ricochet Podcast: Money and Politics with Jim Pethokoukis: Episode 21: The End of Driving is now up

Opens the player in a new window so you can listen uninterrupted while continuing to browse the site in this window. Please allow Ricochet.com to open a pop-up window for this to work.

Direct link to MP3 file

It’s not just driverless cars. In the latest Ricochet Money & Politics podcast, economist, transportation expert, and blogger David Levinsonargues traffic is declining and will continue to decline dramatically in the coming decades. And that decline is not only the result of some deeper economic and technological trends, but will itself cause a radical restructuring of American society.

What happened to traffic? — The Transportationist

Traffic on Washington Avenue – Raw data edition | streets.mn

Cross-posted from streets.mn:  Traffic on Washington Avenue – Raw data edition.

“Why does this matter? By being “conservative” and adjusting traffic counts up, they are over-estimating the need for roadway capacity, that is, they are being “liberal” with the number of lanes required to ensure a particular level of service.”

Traffic on Washington Avenue – Raw data edition


Prompted by Brendon’s recent post (and a Twitter conversation with Ethan and Janne), I got interested in some of the traffic counts and engineering on Washington Avenue. I looked at the Washington Avenue Traffic Operation Analysis by Alliant Engineering for Hennepin County. On p.11 of this document are traffic “counts” the report said were collected in Spring of 2011. The traffic added up perfectly, and with some experience looking at traffic counts in a previous life (some people’s previous lives were as Cleopatra or the King of England, like Bill Gates I was a traffic counter), I had to believe some massaging was going on, data never comes out that clean, especially when it is collected on multiple dates. For instance, there is conservation of flow in traffic, every car that enters and intersection must leave it (unless it is raptured). A traffic count at one site on one date will ensure this. Similarly every car that leaves an upstream intersection must arrive at the downstream intersection, after controlling for driveways. A traffic count on one date is likely inconsistent with another date. Other reasons for massaging include inconsistent peak periods (the peak time at intersection A may differ from downstream intersection B).

In one of the great blessings of open data, Minneapolis makes its raw traffic counts available online. So I went to their website and looked for myself, under turning movement counts [TMC] on Washington Avenue South. These counts are summarized in Table 1 for Eastbound traffic (and the first 5 columns of Table 2 for Westbound traffic), along with the numbers from Alliant’s report. As you can see Alliant’s numbers (data column 3) are 10 to 35% higher than the counts in the City of Minneapolis database for the same period (data columns 1 and 2) (and I assume these are the source of Alliant’s resulting numbers, though the report is vague on the traffic count details). The ratios are given in data column 4.

Why does this matter? By being “conservative” and adjusting traffic counts up, they are over-estimating the need for roadway capacity, that is, they are being “liberal” with the number of lanes required to ensure a particular level of service.

Table 1: Eastbound AM flows on Washington from Hennepin to 11th Avenue

Cross-street Count Inflow Count Outflow Alliant Inflow (Fig 5) Alliant/Count
Hennepin 1061 906 1172 1.10
Nicollet 880 887 972 1.10
Marquette 831 860 988 1.19
2nd Ave 812 773 1044 1.29
3rd Ave 753 890 952 1.26
4th Ave 923 561 1130 1.22
5th Ave 537 642 620 1.15
Portland 640 462 744 1.16
Park Ave 474 578 532 1.12
Chicago 702 666 724 1.03
11th Ave 593 776 656 1.11


Table 2: Westbound AM flows on Washington from 11th Avenue to Hennepin

Cross-street Count Inflow Count Outflow Alliant Inflow Alliant/Count Dominant Dir. WB/EB WB/ln w/ Rev. EB/ln w/ Rev. Lane Split
11th Ave 1046 1107 1268 1.21 W 1.764 349 593 3/1
Chicago 1120 870 1262 1.13 W 1.595 373 702 3/1
Park Ave 875 998 1040 1.19 W 1.846 292 474 3/1
Portland 875 874 1184 1.35 W 1.367 437.5 320 2/2
5th Ave 921 1131 1068 1.16 W 1.715 460.5 268.5 2/2
4th Ave 1155 783 1356 1.17 W 1.251 577.5 461.5 2/2
3rd Ave 786 667 928 1.18 W 1.044 393 376.5 2/2
2nd Ave 607 522 756 1.25 E 0.748 303.5 406 2/2
Marquette 595 612 662 1.11 E 0.716 595 277 1/3
Nicollet 555 535 702 1.26 E 0.631 555 293 1/3
Hennepin 594 573 686 1.15 E 0.560 594 354 1/3


Reversible Lanes

I was also interested in some other aspect of traffic. Ethan said the traffic was balanced on Washington after I posited that it was unbalanced, and we could consider reversible lanes. In fact it is both, depending on where you are looking. The final columns of Table 2 identify the dominant direction, the directional ratio (WB/EB flow), and what flows would be with the lane split given in the final column. At 3rd Avenue, traffic is balanced, to the East there is much higher westbound traffic in the morning, to the West there is much higher eastbound traffic in the morning. Along Washington Avenue, the midpoint of downtown traffic is 3rd Avenue (not Nicollet as I would have supposed before looking at the numbers).

Is the imbalance sufficient to justify reversible lanes? The case is marginal. In general, with two lanes in each direction, left turn lanes, and good signal timing, I think a 2/2 split should work well enough. Near I-35W a 3/1 split (3 lanes westbound, 1 lane eastbound in the AM) is plausible. Similarly on the westside of downtown, a 1/3 split is also plausible in the reverse direction.

I am leaving the PM analysis as an exercise for the reader.

The Minneapolis Turning Movement Counts can be found here: WashAveAMFlows.pdf andWashAvePMFlows.pdf.


Atlanta traffic bad but predictable

I get interviewed about the reliability measures of the new Urban Mobility Report by Ariel Hart of the Atlanta Journal-Constitution: Atlanta traffic bad but predictable :

“‘People care about this,’ said David Levinson, a professor of civil engineering at the University of Minnesota who researches traffic psychology surrounding reliability. People will even accept more congestion to get more reliability, he said, and he has a mathematical formula to calculate how much.
‘It’s the surprises, the inconsistency of the delay that makes it difficult,’ costing people social capital with colleagues, clients and friends when they are unexpectedly late, he said.”

The reliability ratio, the ratio of the value of reliability to the value of time us about 1, depending on how it is measured. See

Does TTI underestimate historic congestion levels?

To read the Texas Transportation Institute’s Urban Mobility Report is to believe congestion has more than doubled since 1982 (really between 1982 and 2000). From one perspective, of course congestion must have risen, demand (Vehicle Miles Traveled, Population, etc.) increased significantly over this period while supply (Lane Miles of Road Capacity) did not increase at nearly the same rate.
But I was alive in 1982, I was in cars at that age (and driving myself the next year) (in Central Maryland). I remember congestion in the 1980s. To misquote Lloyd Bentsen, “Congestion was a friend of mine”, and TTI seems to be saying to 1982 “You’re no congestion”. But congestion doesn’t seem appreciably different from today. People complained about it then as much as now. Some bottlenecks have been fixed, new ones have emerged.
So I wonder whether congestion did, in fact, “double”.
Some hypotheses:
1. Measurement issues. Continuous roadway travel time measurements were a lot scarcer in the 1980s than today. Freeways now have loop detectors on every segment, whereas there might have been a permanent recording station every 5 or 10 miles in the 1980s, so a lot more had to be estimated and approximated. There are still no good arterial measurements, the most recent Urban Mobility Report uses GPS data from Inrix, and this will clearly come to dominate congestion measures. Notably, including this measurement forced TTI to re-estimate downward their historical congestion measurements.
2. Definition: As noted by Joe Cortright’s report Driven Apart, mobility is not accessibility. A city where I can reach everything in 10 minutes, but travel at 30 MPH (when freeflow is 60 MPH) is more congested than one where I can reach everything in 30 minutes, but can travel at freeflow conditions. The TTI in a sense penalizes efficient land uses.
3. Induced Demand: Highway expansion tends to get used up (this is not a bad thing of itself, just a thing), so much of road expansion gets eaten up in more traffic. Similarly highway reduction reduces travel. Duranton and Turner write “We conclude that an increased provision of roads or public transit is unlikely to relieve congestion.”
This does not explain why congestion is under-estimated in the past though.
4. Congestion vs. Speed: Travel times on journey to work increased only marginally over this period. Average distances for trips rose faster than travel times, indicating average travel speeds increased. So even with increasing congestion, if travelers shifted to relatively faster (e.g. suburb to suburb freeways) from slower (e.g. suburb to city arterials), congestion can rise on each link, but travel speeds still increase. See The Rational Locator for an example of this.
5. Perspective: This previous point about perception can be refamed as one of perspective. There are differences between spatial averages (which TTI uses) and person-based averages (which individual observers perceive). So the person based average for any metropolitan resident may be the same, but the amount of space (network) covered by congestion may increase if the total amount of space which is developed increases. Similarly, if there is peak spreading, congestion occurs over a longer duration.
However, TTI is not simply saying that the amount of area that is congested increased, they are claiming, for Washington DC the delay per person increased from 20 hours per year in 1982 to 74 hours in 2010.
I am willing to believe that with recent measurements, 74 hours per year for an average commuter in DC is plausible in 2010, since that is just under 10 minutes each way each day for 225 work days per year. 10 minutes of delay on a 30 minute commute means the freeflow time on that commute (un-delayed, e.g. Sunday morning) was 20 minutes. This seems about right for the “average” commuter. Rush hour is when everyone has to slow down.
But this implies in 1982 that delay was less than 3 minutes a day per commuter each way. That seems unreasonably small when you think about it, I could have spent 3 minutes at a traffic light in DC at the time, and that certainly constitutes delay. They are saying for every person who had a 10 minute delay, 2 people had 0 delay to get an average 3 minute delay, and that is not the metropolitan Washington I was familiar with. Congestion was sufficiently important than that radio stations had regular traffic reports, and traffic helicopters, it was not something insignificant.
Of course this is impossible to fully validate, as we cannot go back in time and accurately measure speed. The best I could think of was using the Google NGram feature to track mention of some keywords in books. This proves nothing unfortunately, and suggests a small uptick in the word “traffic” in the 1990s, but is interesting none-the-less.
One however can imagine the motivation for wanting congestion to appear lower in the past than it actually was. This means congestion is rising faster, and thus creates a greater claim on the public weal than if congestion were always with us at roughly the same level.

Why does the Institute of Transportation Engineers exist? 10 Ideas for Big Changes – Mike on Traffic

Mike Spack vs. ITE: Why does the Institute of Transportation Engineers exist? 10 Ideas for Big Changes. See his post for the list.

I am disappointed Mike took down his spreadsheet, though I understand why. If he were at a University, they wouldn’t dare. Frankly, the ITE trip generation data is mostly like a telephone book and can’t be copyrighted, though its specific presentation (and maybe the regressions, though those seem pretty damn uncreative to me) can be. An analysis of that data is certainly fair game. An alternative though would be to set up a Trip Generation Wiki or Google Docs which is open, letting people upload their own data and updating the regressions automatically (since it is a pretty trivial spreadsheet operation).
I am thinking of unjoining ITE, my last professional organization (I quit APA a long time ago due to their profit-maximizing behavior since I gained nothing from the organization and they wanted a non-trivial share of my salary) over their heavy-handed, anti-public, guild-like behavior. If Mike were President, I would reconsider. The backwardness of ITE is one of many reasons Traffic Engineers are becoming increasingly unpopular.