Wendover’s Traffic Fixes Look Better When You Leave Out the People They Fail
Wendover Productions makes a persuasive case that cities can reduce congestion and road deaths without endlessly widening highways. But its most striking examples are presented with more certainty than their evidence deserves. The Katy Freeway comparison does not establish that expansion caused longer journeys, Poynton's pedestrian safety record contradicts the video's account, and the celebrated 90% roundabout statistic is presented without the qualifications needed to understand it. These are not minor technicalities: they change what the proposed solutions actually promise.
The Katy Freeway Is Not the Proof the Video Claims
The video presents Houston's $2.8 billion Katy Freeway expansion as a spectacular failure. It says afternoon travel times increased by 55%, from approximately 41 to 64 minutes, between 2011 and 2014. The figures come from analysis of Houston TranStar traffic data, but there is a serious problem with the comparison: both years were after the expansion had been completed in 2008.
That is evidence of worsening congestion after the initial improvements, not a controlled measurement proving that the expansion itself made journeys slower.
The Texas A&M Transportation Institute's 2013 evaluation provides important missing context. Its researchers found that the managed lanes were still saving approximately five minutes during the morning peak and 14 minutes during the afternoon peak compared with general-purpose lanes. Traffic volumes in the peak direction had more than doubled since the facility opened. The report also identified latent demand, economic development, employment growth, fuel prices and traffic diverted from neighbouring routes as factors affecting travel times. Texas A&M Transportation Institute, 2013.
The distinction matters. The expansion initially improved journeys, then additional demand consumed much of the available capacity. That is a legitimate example of induced demand, but it is not evidence that the project delivered no benefit or that every additional lane necessarily makes an individual journey slower.
The “One-to-One Law” Is Not Quite What the Research Says
The video attributes its central argument to economists Gilles Duranton and Matthew Turner, whose 2011 paper, The Fundamental Law of Road Congestion, appeared in the American Economic Review. Their research is substantial: it examines 228 US metropolitan areas across 1983, 1993 and 2003, finding that vehicle-kilometers travelled (VKT) increased approximately in proportion to interstate highway lane-kilometers.
matthewturner.org
But Wendover changes what that finding means. The researchers measured total distance travelled by vehicles across metropolitan road networks, not the number of individual people using a particular road. Their findings include increased driving by existing residents, additional commercial traffic and migration into better-connected cities.
A city generating twice as much vehicle travel is not the same thing as an individual road attracting twice as many drivers when its capacity doubles. VKT also measures distance, not unique drivers or vehicles.
The distinction is important because the video calls this a “perfect 1 to 1 ratio” and applies it as though it were a universal physical law governing individual roads. The original research supports a powerful economic relationship, not that simplified rule.
The authors themselves warn that studies of individual roads cannot establish the wider effect of infrastructure changes across an entire city. Wendover uses precisely such an individual-road example to introduce its citywide argument.
American Economic Association
Induced demand is real. The problem is presenting a qualified empirical finding as an exact rule.
Poynton: The Pedestrians Were Not All Better Off
The episode's most consequential failure concerns Poynton, England.
Wendover describes a town that removed traffic lights, pedestrian crossings, curbs and almost every safety device. It then announces that pedestrian incidents and collisions decreased, presenting the project as evidence that driver uncertainty can save lives.
There is a substantial problem with that account: Poynton's actual pedestrian accident figures moved in the opposite direction.
The Chartered Institution of Highways & Transportation (CIHT) published a detailed review of the project in January 2018. Its comparison of three years before and after implementation found:
Reported accidents Before (2007–09) After (2013–15) All accidents 14 9 Serious accidents 5 2 Pedestrian accidents 2 6 Serious pedestrian accidents 1 2Overall accidents decreased, but pedestrian accidents tripled. The review explicitly identified the increase as occurring along Park Lane. It rated the scheme's safety outcome neutral, not positive.
ciht.org.uk
ciht.org.uk
These are relatively small samples, and they cannot establish that the redesign caused the increase. But they certainly cannot support Wendover's claim that pedestrian incidents decreased.
There is another factual problem with the presentation. Poynton did not remove all its curbs or pedestrian protections. The actual design retained 30 mm kerbs, widened footways, incorporated central pedestrian refuges and provided informal courtesy crossings approximately every 50 metres. The intersection handled more than 25,000 vehicles daily.
ciht.org.uk
This was not simply an experiment in removing safety infrastructure and trusting drivers to behave.
The accessibility concerns were equally serious. A 2017 seminar organised through Transport Scotland and Edinburgh Napier University recorded testimony from visually impaired people familiar with Poynton. They described feeling petrified by the mixture of vehicles and pedestrians, questioned the safety of the courtesy crossings and reported uncertainty about who had priority.
transport.gov.scot
That report records participants' experiences, not a statistical accident study. But those experiences matter. A street that makes a sighted driver more cautious can simultaneously make a blind pedestrian less able to navigate safely.
The slogan “Discomfort is saving lives” ignores this fundamental difference. Discomfort is not distributed equally between road users.
Wendover turns a complicated accessibility question into a clever psychological lesson. The people who struggle most with the design become almost invisible in the success story.
The 90% Roundabout Statistic Needs a Denominator
Roundabouts genuinely improve road safety. But the video's presentation of the famous 90% figure is incomplete.
The 2018 Federal Highway Administration (FHWA) research review traces the frequently cited figure to a 2001 study of 23 US roundabout conversions. That study found:
- 90% fewer fatal crashes.
- 80% fewer injury crashes.
- 40% fewer crashes overall.
A subsequent 2012 study focusing on conversions from signal-controlled intersections found a 21% reduction in total crashes and a 66% reduction in injury crashes.
fhwa.dot.gov
The difference becomes even clearer in a 2017 meta-analysis by researcher Rune Elvik. Examining 44 studies and 154 estimates, it found approximately 65% fewer fatal accidents and 40% fewer injury accidents.
PubMed
These are substantial improvements. But they are not a universal 90% reduction.
FHWA also identifies two important qualifications: benefits diminish as average daily traffic increases, and some improperly designed roundabouts have actually increased crash frequency. It notes that 61% of roundabouts in one national survey were installed at locations with previous fatal crashes, creating a potential selection bias in before-and-after comparisons.
fhwa.dot.gov
There is another technical simplification. The video describes 1,800 vehicles per hour as the maximum capacity of a single-lane roundabout, equivalent to a traditional two-lane signalized intersection. FHWA's roundabout guidance uses 1,800 as a threshold for circulating traffic at any point; actual entry capacity depends on conflicting traffic and intersection geometry. These are not interchangeable measures.
Federal Highway Administration+1
The real argument for roundabouts is already compelling. It does not need a headline number stripped of its conditions.
Stockholm's Success Was Real, but the Policy Was Not Just a $2 Toll
Stockholm is arguably the video's strongest example of demand management.
The congestion charge reduced traffic crossing the central-city cordon by approximately 20%. The OECD's subsequent review reports reductions of 30–50% in queuing time and travel-time variability on bottleneck links.
OECD
But the video leaves out three important conditions.
First, Stockholm already had substantial public transportation. Approximately 75% of peak-hour trips across the congestion-charge cordon were made by public transport. Drivers had a practical alternative to paying the charge. Stockholm's geography also helped: water provided natural boundaries for the charging zone, reducing opportunities for drivers to divert through residential streets.
OECD+1
Second, the charging system required substantial investment. The Stockholm system's initial investment and first-year costs amounted to approximately €200 million. Its second-year operating costs were €22 million annually. The investment was financially recovered in approximately 3.3 years, so this was not a failed expenditure, but it was hardly a costless intervention.
OECD
Third, congestion charges have distributional consequences. The OECD identifies them as regressive for low-income households that own cars, because those households spend a larger proportion of their income on the charge. Stockholm's relatively high public-transport use among lower-income commuters reduced this problem, but did not make the distributional question disappear.
OECD
The lesson is not that congestion pricing fails. It is that successful pricing depends on geography, public transportation, system costs and the people who must pay.
A $1–$2 charge is the visible part of a much larger policy.
The Diverging Diamond's Savings Are Not Universal
The diverging diamond interchange is another legitimate engineering improvement, but Wendover oversells the supporting evidence.
The reduction from 26 potential conflict points to 14 is supported by transportation engineering research. The design also offers real advantages when traffic volumes are high.
GDOT
However, the video says the US Department of Transportation found that diverging diamonds universally allow more cars through faster, regardless of traffic volume.
The FHWA's 2010 evaluation is more qualified. Its computer simulations found that the designs performed similarly at low and medium traffic volumes, with their clearest throughput advantages emerging under heavier traffic. At the lowest tested volume, both designs achieved identical modeled throughput; in one medium-volume scenario, the conventional design actually achieved slightly higher throughput, although the diverging diamond produced less delay.
FHWA-HRT-09-060
The construction-cost comparison also comes from one particular proposal.
The $11.334 million conventional design and $5.772 million diverging diamond estimate came from a Missouri Department of Transportation comparison for an interchange in Kansas City. The figures included construction, land acquisition and utility costs. FHWA explicitly cautioned that there were insufficient existing installations to draw more conclusive general comparisons.
FHWA-HRT-09-060
Later evidence demonstrates why that qualification matters. The National Academies' 2021 Diverging Diamond Interchange Informational Guide reported actual retrofit costs ranging from approximately $3 million to $8.5 million, while new installations ranged from $14 million to $22 million. Costs vary considerably according to the existing structures and site requirements.
connect.ncdot.gov
There are also practical considerations. FHWA identifies additional signage, lighting, signal equipment and pedestrian-crossing requirements. The newer guide notes that pavement markings require more frequent inspection and maintenance.
Calling the design a useful improvement is justified. Presenting one project's cost estimate as a universal saving and declaring that there are no major disadvantages is not.
Some of the Video's Strongest Evidence Survives
An honest attack must distinguish weak presentation from genuinely strong research.
Minnesota's ramp-meter experiment is a particularly good example. When the state disabled its ramp meters, the official evaluation found 9% lower freeway traffic volume, 22% longer travel times, 7% lower speeds and 26% more crashes. These results support the video's argument that controlling highway entry can improve traffic flow and safety.
MnDOT
Stockholm's congestion-pricing results also stand up to scrutiny, as do the broader safety benefits of roundabouts.
The problem is not that Wendover has chosen ineffective solutions. It is that the video repeatedly removes the qualifications that tell us where those solutions work, who benefits and who might be left behind.
The Verdict: Traffic Engineering Is Not a Magic Trick
Wendover's central argument deserves serious consideration: congestion cannot be solved simply by building more lanes, and intelligent road design can improve safety and efficiency.
But an educational video should not turn a citywide economic relationship into a perfect physical law, a selected safety statistic into a universal promise, or a complicated pedestrian environment into a celebration of discomfort.
Poynton is the clearest failure. Its overall accident numbers improved, but pedestrian accidents increased. Its design retained safety features the video says were removed, and visually impaired residents experienced serious accessibility concerns.
A road cannot be declared safer simply because cars move through it more efficiently. The people walking across it count too.
That is the standard the video repeatedly fails to apply. Its engineering examples deserve attention, but its conclusions deserve considerably more scrutiny.
References
1. How to Fix Traffic Forever — Wendover Productions, July 18, 2017. Original video and the claims examined in this critique.
2. The Fundamental Law of Road Congestion: Evidence from US Cities — Gilles Duranton and Matthew A. Turner, American Economic Review, October 2011. Establishes the relationship between metropolitan road supply and vehicle-kilometers travelled, rather than a universal doubling of individual road users.
3. Reducing Congestion: Katy Didn't — City Observatory, Joe Cortright, December 16, 2015. Provides the TranStar-based 2011–2014 commute comparison behind the 55% figure; it is an analytical commentary, not a controlled before-and-after study.
4. Katy Freeway: An Evaluation of a Second-Generation Managed Lanes Project — Texas A&M Transportation Institute, September 2013. Provides official corridor measurements, managed-lane travel-time savings and the multiple factors affecting congestion.
5. Creating Better Streets: Inclusive and Accessible Places — Chartered Institution of Highways & Transportation, January 2018. Documents Poynton's actual design, accident figures, pedestrian outcomes and accessibility limitations.
6. Shared Spaces Seminar 2017: Final Report — Scottish Disability Equality Forum, Edinburgh Napier University and Transport Scotland, 2017. Records visually impaired participants' concerns about Poynton and other shared-space environments.
7. FHWA Research and Technology Evaluation: Roundabout Research Final Report — Federal Highway Administration, June 2018. Reviews the underlying roundabout safety studies, their differing results and potential selection bias.
8. Road Safety Effects of Roundabouts: A Meta-Analysis — Rune Elvik, Accident Analysis & Prevention, February 2017. Examines 44 studies and 154 estimates, finding approximately 65% fewer fatal accidents and 40% fewer injury accidents.
9. Long-Term Effects of the Swedish Congestion Charges — Maria Börjesson, International Transport Forum/OECD, October 12, 2018. Establishes the long-term traffic effects, public-transport context, operating costs and distributional consequences of Swedish congestion pricing.
10. Alternative Intersections/Interchanges: Informational Report, Chapter 7 — Federal Highway Administration, April 2010. Provides the diverging-diamond simulation comparisons, Kansas City cost estimates and operational qualifications.
11. Diverging Diamond Interchange Informational Guide, Second Edition — National Cooperative Highway Research Program, National Academies of Sciences, Engineering, and Medicine, 2021. Documents actual construction-cost variation and ongoing maintenance requirements.
12. Twin Cities Metro Area Ramp Meter Study — Minnesota Department of Transportation. Provides the experimental results supporting the effectiveness of ramp metering.