There’s a Better Way to Win on Traffic Safety
Backlash has undercut the success of 30km/hour speed limits. New research finds that a less controversial policy can be more effective in slowing cars.
Milan has faced backlash to its law capping speed limits to 30km/hour.
Photographer: Piero Cruciatti/Anadolu via Getty Images
In recent years, modest urban planning ideas have sparked outsized backlash. The “15-minute city” concept — designing neighborhoods so amenities are available within 15 minutes by foot or bike — became a flashpoint in conspiracy theory circles. Now, new citywide speed limits of 30 kilometers per hour are the latest target in many European cities: a sensible safety measure turned culture-war fodder.
These reactions have put local and national governments at odds over policies that aim to tackle traffic safety and livability. But recent research results suggest such political battles may be beside the point: In studies we have conducted at our Senseable City Lab, we found that the posted speed limits often have little effect on how fast people drive. Instead, the policies that work involve the design of the street itself.
The movement for citywide speed limits of 30 kilometers (19 miles) per hour gained visibility during Paris Mayor Anne Hidalgo’s 2020 campaign. The logic was compelling: Slower traffic means safer streets and quieter neighborhoods. Furthermore, as our past research showed, “slow zones” attract nearly twice as many people, drawing diverse visitors and enriching the city’s social mix.
Once re-elected, Hidalgo enacted the policy, and other mayors followed. From Milan to Brussels, Bologna to Amsterdam, many began testing or implementing lower-speed zones.
Particularly in my home country of Italy, the backlash came quickly. To some motorists, a lower speed limit felt less like protection and more like control. In the clickbait economy, the measure was recast as suspicion of an authoritarian drift.
Even our lab at MIT got pulled into the storm. In July 2024, we presented results on the impact of 30 km/h zones in Milan. Using real vehicle telemetry from the Italian insurance company Unipol, we were able to accurately assess changes in vehicle speed and evaluate their consequences. We found that increases in travel times were negligible — at most an average of about 34 seconds per trip — while previous research had found a decrease in fatalities of around 37%. We also found that air pollution levels remained mostly stable, with only a minimal increase.
Despite the breadth of our findings, headlines in Italian news outlets focused entirely on an increase in air pollution. In fact, Italy’s Transport Minister Matteo Salvini seized on our study as “scientific evidence” that the 30 km/h policy created more traffic jams and pollution. Since making that comment, Salvini and his colleagues in Italy’s national government have worked to erode cities’ 30 km/h policies, imposing limits and banning the cameras used to enforce the policy in the cities that had enacted it.
In response, we dug deeper. This year, we analyzed Milan, Amsterdam and Dubai. We gathered over 73 million telemetry points and 1.2 million street-level images. Such data is now analyzable at speed and scale thanks to AI models — a theme we are also exploring at this year’s Venice Biennale Architettura.
What we found suggests something surprising: Speed limit laws aren’t actually the best way to get residents to slow down. Drivers slowed only 2–3 km/h when limits dropped from 50 to 30 km/h, while the design of the streets played a much larger role. Narrow, enclosed streets with high building density tended to slow traffic, whereas wide, open roads with long sightlines encouraged faster driving. This suggests that design, more than legislation, sets the pace.
There are likely a few explanations for this. In many countries, particularly Italy, speed limits are considered merely a suggestion, meaning that legal changes won’t usher in cultural ones. Milan and Bologna, where speed cameras were eventually banned, offer clear examples of how difficult it can be to rely solely on punitive measures — even when noncompliance is widespread.
Street configuration, by contrast, can change behavior in more palatable ways. Urbanists have long intuited that design can moderate traffic speed. With the first large-scale, cross-city uses of AI, those insights can be quantified and generalized across different contexts.
Raised crosswalks, narrower lanes, or tree-lined streets are examples of the kinds of design interventions our model takes into account. In short, our model can now predict how fast people are likely to drive just from street imagery, offering urban planners a predictive tool to assess the impact of different design interventions.
The MIT study focused on parts of Amsterdam where cyclists dominate traffic flow.Photographer: Selman Aksunger/Anadolu via Getty ImagesApplying our method across Milan, Amsterdam and Dubai showed how local context matters. In Milan, with UNIPOL, we studied traditional, mixed-use streets — compact urban corridors where drivers, cyclists and pedestrians share space.
In Amsterdam, with the AMS Institute, we analyzed a cycling-first network — areas where bikes dominate traffic flows and car access is secondary, offering insights into how multimodal coexistence shapes speed.
In Dubai, working with the Dubai Future Foundation, we tested car-centric streets under extreme heat. There, we found that heat itself can reduce average driving speeds by up to several km/h, showing how environmental stressors can influence behavior even in the absence of formal design interventions.
These examples show how street geometry — but also culture and climate — shape how people drive, often more than regulations do.
To be sure, changes to street design are not without controversy: Attempts to add bike lanes and pedestrian spaces or reduce road space dedicated to cars have faced pushback, too.
Still, our results may offer comfort to both sides of the political spectrum. Changing how people drive cannot be achieved by decree alone. But there is no grand conspiracy at work: It may depend, above all, on the configuration of our streets.
Carlo Ratti is a practicing architect and a professor at the Politecnico di Milano and at MIT, where he directs the Senseable City Lab. In 2025 he’s directing Venice’s Biennale Architettura.