Something we never consider as we bicycle…

Katja Kircher

Excerpted from “Hot take – Cycling Research is in Fact Pro Car”, presented by Katja Kircher 14 June 2026, part of Climate Talks for Everyone. View the complete recording HERE.

PART ONE

My name is Katja Kircher, I work at VTI, the Swedish National Road and Transport Research Institute, and I mainly work with bicycling in rural areas.​

I’d like to start with some quick statistics and background information.​

​Very often, “road traffic fatalities” in Europe in 2024 are reported like this. By mode of transport. What we see and are often told is this: 9 % of the people dying on Europe’s roads are cyclists. That seems like quite a lot, especially since the mileage done by bike is so much smaller than the mileage done with motor vehicles.​ So, do you assume that cycling is dangerous?​

There is another way to present such grim statistics, and that is in the form of a matrix. This matrix, published by the EU, shows the absolute number of fatalities for each road user type, depending on who they were in a collision with.

If we only look at collisions involving cyclists and cars, we see that they killed 830 people.​ We also see that 98.5% of those were cyclists.​ 

So, can we really say that cycling is dangerous? Does this mean that we have to research cycling safety? Do we have to look into measures protecting cyclists?​

Or should we say that motor vehicles are dangerous and that we have to research road violence? And that we have to look into measures that reduce this violence.​

My own case - Overtaking

​As “motor vehicles overtaking cyclists” is the “cycling research” area I am most familiar with, I would like to give you an example of how the topic can be discussed.​ 

Let’s look at the typical research. ​

It is common to measure the lateral clearance between the bike and the overtaking vehicle. This is typically done with a device mounted on the bike. In addition to the clearance, you’ll typically get the geoposition of where the overtake happened. From this, you can extract the road type, road width, and other factors that happen to be available on Open Street Maps. Maybe a camera was installed on the bike, too, then you’ll get a few more variables.​

​So, a typical thing to do is to model the expected clearance based on all the factors that could be measured. For example, a study might find that oncoming traffic reduces the average clearance. Or that the average clearance is lower when the road is narrower, and with the model, we can calculate the relationship between road width and overtaking clearance. ​

But it is not discussed whether ​

• People should not overtake at all in certain situations​

• why they still do it, even though they clearly endanger the cyclist​

• why they do not always overtake with the widest possible clearance​

• and whether the drivers' right to go fast outweighs the cyclists' right to ride in peace.​

Some studies actually log how the overtaking is perceived by the cyclist. But also there, it is taken for granted that overtakes will take place. There is no rating of how it feels not to get overtaken. The results of the ratings are then set in relation to clearance in metres and centimetres. But they are not analysed in relation to ”respect”. It makes a difference if I am granted 1.5 m clearance when that’s all that’s available on a narrow road, and the driver slowed down behind me and waited for me to be ready to get overtaken, or when I get 1.5 m on a road where I easily could have got four metres, and if the driver zooms past me with 100 km/h.​

The paradigms

When we, as “cycling researchers”, start investigating overtakes, we often do this within a framework of existing laws and regulations. These are mostly based on two paradigms that I will explain briefly. They are called the “compliance-based approach” and the “performance-based approach”.​

First out is the Compliance-based regulation, that is, prescribing exact limits: This is common practice, often through absolute lateral distances [e.g. 1.5 m clearance]. There are variations between countries and, in some cases, moderations by speed or location. It looks like it’s clear and scientific, but it is rigid, hard to enforce, not supported by evidence and does not give cyclists any agency. Also, it may send a wrong message, turning the “minimum distance” into a target distance.​

Another option is Performance-based regulation. Examples are phrases like “give plenty of room”, “as if overtaking a car”, or to leave space such that one would not drive over the cyclist, should they fall. ​

​Both compliance- and performance-based approaches are driver-centric. They focus on a “minimum”, regardless of how much space there actually is. Their main concern is to get the driver past the cyclist, with minimum inconvenience to the driver, but without hitting the cyclist.​

There could be different ways to deal with the topic. Less engineering-focused and more humanistic.​

The underlying paradigm should be to minimise harm and to grant the cyclists agency.​

The minimum harm requirement can be achieved by overtaking with a full lane change or a corresponding manoeuvre. This would be predictable for the cyclist.​ Combined with a speed reduction, this is also a possibility to communicate the driver's awareness to the cyclist.​

To provide agency, we should allow riding abreast and not require cyclists to keep to the very edge, but let them choose their position within their side of the road. This is written in the UK Highway Code, for example, where cyclists are recommended to take the so-called primary position, in the middle of the lane, as opposed to the secondary position, which is keeping to the edge. Here in Sweden, where we are so proud of our Vision Zero, we still require cyclists to keep as far to the right as possible and to only ride abreast if they don't inconvenience traffic. And we can assume that this implies motor traffic, not cycle traffic.​

But when I talk about these overtaking requirements, for example, in the media, but also in other contexts, I face harsh criticism and incredulity. The Swedish Traffic Agency was tasked a couple of years ago with evaluating the possibility of introducing a regulation based on a suggestion I had published, and which follows the minimum-harm guidelines as explained here. Their conclusion was that nothing needed to be changed from the status quo, and that introducing a regulation like that would delay motorists and lead to more frustration and aggression on their side. No research was provided to substantiate their claims.​

Interestingly, in spite of Sweden currently having a right-wing government that is not very interested in climate and sustainable transport, it commissioned the Transport Agency with a similar task again. The assignment needs to be reported by the end of August, and I hope the outcome will be better.​

This is just an anecdote on how cycling research, even if available, is dismissed by politics, if it doesn’t just so happen that a politician has a personal interest, and we do have one minister who is an active cyclist.​

What type of research are we supposed to do?​

By researching such a small detail as motor vehicles overtaking cyclists, I implicitly agree to having cars on the road and that they also, in the future, will go faster than cyclists.​ I am certainly not dedicating my efforts to the root of the problem. Even the best overtake is never better than not being overtaken at all.​

Can I work to ensure that cyclists get the most direct routes on roads free of cars? Can I focus on making the environment for cyclists healthy and pleasurable, that is, without motor vehicles zooming past them?​

Honestly, can I even imagine a world without private cars? A world that is not just like the world right now, only with cars removed, but a fundamentally different world where cars aren’t a thing? ​

Part Two will follow next month.