Subheadline: An IIT-Delhi study finds that police records capture only a small share of bicycle crashes, while motorcycles emerge as the most common striking vehicle in self-reported incidents.
Standfirst: A new IIT-Delhi study has exposed how little Delhi’s official road-safety record may reveal about bicycle crashes. Of 218 crashes reported directly by cyclists in the study, only 2.8% were also reported to police. The gap is not simply a question of incomplete statistics. It changes how the city sees risk on its roads: police records overrepresented crashes involving trucks, while cyclists’ self-reports identified motorcycles as the striking vehicle in half of all cases. The study also recorded a high injury burden, with knee and lower-leg injuries accounting for 60.4% of injuries. Drawing on surveys of 844 cyclists at 27 locations and follow-up data from 742 participants, this analysis examines what the findings reveal about reporting, road design and the limits of relying on police data to understand cycling safety in Delhi.
The central finding of the IIT-Delhi study is straightforward but consequential: most bicycle crashes in Delhi do not appear in police records. The report, titled “Characteristics and Incidence Rates of Self-Reported Bicycle Crashes in Delhi: A Cohort Study”, found that only 2.8% of bicycle crashes identified through cyclist self-reporting were reported to the police. That leaves 97.2% outside the formal police record examined by the researchers.
The study was conducted by Srishti Agarwal and Rahul Goel of IIT-Delhi’s Transportation Research and Injury Prevention Centre. It used on-road intercept surveys involving 844 cyclists at 27 locations in Delhi last year. Of these participants, 742 provided crash and exposure information in at least one follow-up and were included in a prospective cohort followed for an average of 4.8 months.
This method matters because it captures incidents that may never become police cases. A crash can result in pain, injury, medical treatment or damage to a bicycle without leading to a formal complaint. When those incidents are absent from official records, the resulting picture of road danger is shaped by what reaches the police rather than by the full range of risks cyclists experience.
The difference between the two data sources is particularly visible in the type of vehicle involved. Among the 218 self-reported crashes, motorcycles were the striking vehicle in half of the cases. Cars accounted for 25%, while electric three-wheeled rickshaws accounted for 11%. By contrast, police-reported cases had a higher proportion of trucks and a lower proportion of motorcycles as the striking vehicle.
That contrast suggests that official records and cyclist surveys are describing different parts of the same road environment. Police records appear more likely to include severe crashes involving larger vehicles, while lower-visibility collisions involving motorcycles may be missed. The supplied study does not establish why those cases go unreported, but the difference is large enough to affect how authorities identify dangerous vehicle interactions and locations.
The researchers connect the motorcycle finding to the structure of Indian urban streets. Motorcycles constitute the majority of vehicles in Indian cities and often occupy kerbside lanes where cyclists usually ride, particularly during congestion, according to the study. The overlap between cycling space and motorcycle movement therefore becomes a central safety issue—not as an abstract conflict between transport modes, but as a recurring interaction at the edge of the carriageway.
The study’s findings also challenge the assumption that bicycle crashes are mainly single-bicycle events. Only 2% of the self-reported crashes were classified as single-bicycle crashes. These included loss-of-control incidents, sudden braking, manoeuvres to avoid collisions, mechanical failures and crashes involving objects or potholes.
That result differs from findings cited in the study from Australia, Europe and the United States, where single-bicycle crashes account for 40% to 70% of bicycle crashes. The researchers attributed the difference partly to lower cycling speeds in Delhi. The data instead points towards a road environment in which interactions with other vehicles are a more prominent feature of reported non-fatal crashes.
The distinction is important for how safety problems are diagnosed. If crash data is interpreted mainly through single-bicycle incidents, the response may focus on cyclist behaviour, bicycle condition or isolated obstacles. If collisions with motorcycles, cars and other vehicles dominate the local evidence, the safety question shifts towards how different road users share space, particularly along kerbside lanes.
The injury data reinforces the seriousness of the problem. All police-reported non-fatal crashes in the study resulted in injury and received medical treatment. Among self-reported crashes, 78% resulted in injury and 67% of those injured sought medical treatment. Knee and lower-leg injuries accounted for 60.4% of injuries, followed by wrist, hand and finger injuries at 27.8%. Head injuries accounted for 14.8%.
These figures do not describe every bicycle crash in Delhi. They describe the cases captured through the study’s survey and follow-up design. But they do show why a narrow reliance on police data can be problematic. A crash does not need to become a police case to impose a medical burden on a cyclist, and a pattern of non-fatal injuries can remain invisible if the main administrative record is built around reported incidents.
The study calculated a crash injury rate of 28.6 per million kilometres cycled. It said this was 60 to 200 times higher than in Denmark, Germany and the Netherlands, countries with extensive cycling infrastructure. The comparison, as presented in the study, supports the researchers’ conclusion that infrastructure improvements play a critical role in reducing bicyclist injuries.
The supplied material does not identify which specific infrastructure interventions were evaluated or quantify the effect of any particular design. It does, however, establish a clear relationship in the study’s argument: Delhi’s injury rate is substantially higher than the cited rates in countries with extensive cycling infrastructure, while Delhi’s crashes frequently involve other motor vehicles. The evidence therefore points to the built road environment as a central part of the safety question.
The reporting gap also creates an institutional problem. Police records are often treated as the foundation for identifying crash locations, vehicle patterns and safety priorities. If only 2.8% of bicycle crashes enter that system, officials may be working with an incomplete account of cycling risk. The result can be a mismatch between recorded severity and everyday exposure, especially for collisions that do not result in a police case.
The study’s use of self-reported data does not eliminate the need for caution. Self-reporting and police records capture incidents through different processes, and the study followed participants for an average of 4.8 months rather than monitoring every cyclist in the city. The findings should therefore be read as evidence of an under-recorded pattern, not as a complete census of all bicycle crashes in Delhi.
Even with that limitation, the difference between the two sources is too important to dismiss as a data technicality. The study did not merely add more incidents to an existing count. It changed the apparent profile of risk. Motorcycles, rather than trucks, were the striking vehicle in the largest share of self-reported cases, and most of the recorded incidents involved another road user rather than a cyclist riding alone.
That has implications for urban mobility policy. Cycling safety cannot be assessed only by counting fatal or police-reported collisions. A more complete picture would need to account for injuries treated outside the police system, interactions in kerbside lanes and the conditions that cause cyclists and motorcycles to occupy the same narrow space. The study’s findings make the quality of that evidence a governance issue as much as a transport issue.
They also raise questions about how Delhi’s cycling infrastructure is planned and evaluated. The researchers’ conclusion is not that one vehicle category alone explains the city’s crash burden. Rather, the evidence highlights the need for context-specific safety research in settings where motorcycles are a dominant mode of transport. Infrastructure designed without accounting for that local traffic pattern may fail to address the most common forms of conflict.
The larger urban question is whether Delhi’s road-safety systems are measuring the risks that matter to people who cycle. The study shows a substantial gap between what cyclists reported and what police records captured. It also shows that the injuries were not marginal: most self-reported crashes caused injury, and a majority of those injured sought medical treatment.
What the evidence confirms is limited but significant. Bicycle crashes in the study were overwhelmingly under-reported to police; motorcycles were involved in half of self-reported collisions; injuries were common; and the calculated injury rate was far higher than the rates cited for countries with extensive cycling infrastructure. What remains uncertain is how these patterns vary across Delhi’s roads, locations and cyclist groups, and why most incidents do not enter the police system. Those are the questions that future safety research and official monitoring will need to address.
