Learn about female representation in automobile safety testing and design

IHSI Diversity Committee Monthly Resources

Each month, the IHSI Diversity Committee shares resources and learning opportunities with the rest of the IHSI staff. Curating and sharing these resources allow us to educate ourselves on various topics related to diversity, equity, inclusion, and justice. As an institute devoted to promoting all forms of health, we want to share these resources more broadly with our campus and community stakeholders. We hope that you will find them as useful as we have.

Female representation in automobile safety and design

This month, Aman Kaur, Research Biostatistician, shares resources to highlight how the populations missing from our data are overlooked in safety design and systems in the automotive industry.

"My personal experience is with the car seatbelt that sits uncomfortably next to my neck and collarbone, and the headrest (driver or passenger seat) that does not seem to be designed for someone of my body size. There are also a couple of other issues that I won't get into. Navigating these issues can be challenging for people, especially women," Aman shares. 

Representation in data collection shapes real-world outcomes, and when certain populations are overlooked, it leads to products and services that can fail to capture that population’s needs or, in some cases, be harmful. The automotive industry is one such domain. 

Historically, crash test dummies have been modeled after the 50th-percentile male body as the default standard for human safety design. Because female seating posture and biology differ, women involved in car crashes are significantly more likely to sustain serious injuries or fatalities in comparable collisions.According to a recent study (January 2026) by the National Highway Traffic Safety Administration (NHTSA), women have a statistically significantly higher risk of injury than men in 26% of the 150 crash injury models, and women have a higher risk more frequently for moderate injuries. Another research study (July 2019) shows that in the United States, women are 73% more likely to be severely injured in vehicle crashes than men.

Female crash-test dummies do exist, but historically, many tests used smaller versions of male-based models (that looked like 12-year-old little people) rather than anatomically representative female models. More advanced female-body models have been developed, such as the EvaRID FE Model (a 50th-percentile female) and THOR-5F (Test device for Human Occupant Restraint, 5th-percentile Female), but their adoption in regulatory testing remains limited even today!

Here are two interesting videos that highlight this issue (5-9 mins):

• In this first video, Caroline Criado Perez, author of ‘Invisible Women: Exposing Data Bias in a World Designed for Men’, explains how gender gaps in data collection and analysis in vehicle safety testing have led to car designs that make car accidents more fatal for women. ‘Why car design is killing women’ | Penguin Books UK
• This second video delves into the half-century history of automotive safety testing in the United States, breaking down why crash test dummies were modeled exclusively on male bodies and how that design choice directly impacts female injury rates today. Why Are Women More Likely to Die in Crashes? New York Times Opinion | The New York Times

While there are some emerging technologies, such as intelligent speed assistance, automatic emergency braking, and virtual crash testing with computational modeling, that can help reduce overall fatalities, but it is still crucial to design vehicles that can serve all populations and for that, data must be representative!

Some automotive companies do conduct research to protect all people, such as Volvo’s EVA initiative (Equal vehicles for all), Toyota’s research on seat belt fit for pregnant drivers (in collaboration with the University of British Columbia), and General Motors’ female occupant biomechanics research – but the question remains of standardization and regulatory testing!

Some questions to consider:

1. Have you encountered a vehicle feature that didn't seem designed with your needs in mind? What was your experience?
2. What are some other domains where gaps in data collection/representation have resulted in unequal outcomes for certain populations?