In Car Safety, Why Some Companies Merely Meet a Standard and Others Exceed It

Corporate embrace of safety and experience in technical measurement is crucial determinant

Government regulations are designed to protect us — keeping our roads safer, air cleaner and products less harmful. ​Many save lives. After the 1994 introduction of a federal rule that set a minimum level of protection in what are also known as T-bone collisions, fatalities from side-impact car crashes fell 44%.

Research published in Management Science by UCLA Anderson’s Jane Wu reveals important differences in how firms responded to that rule. Wu uses a specific feature of the side-crash safety regulation — its reliance on a single crash-test dummy as a testing metric — to show that companies responded in different ways, producing materially different patterns of protection across car manufacturers.

The regulation required cars to pass a side-impact test based on a single crash-test dummy representing one specific body type: a mid-1960s “average” male about 5 foot, 8 inches and around 170 pounds. Wu analyzed more than 29,000 real-world crashes between 1991 and 2007 and found that some companies tailored design changes closely to the mandated body type, while other companies used the regulation as a starting point and pursued improvements in side-impact protection that benefited a wider range of passenger body types.

Both approaches dramatically reduced fatalities for body types similar to the dummy. But Wu found that firms that chose the narrow approach to meet — but not exceed — the single-test metric delivered no statistically significant extra safety for other body types, while companies that took a broader focus delivered significant safety improvements for all body types.

“Seemingly technical choices, such as how to design a test dummy or specify a threshold, can shape not only firm behavior but also the distribution of benefits across consumers,” Wu writes.

Wu’s analysis suggests that the decision by only some companies to interpret the regulation more broadly — and ultimately save more lives — came down to corporate priorities and, notably, capabilities. She found that companies committed to safety that also invested in the tools and resources needed to conduct detailed testing were the ones that delivered broad protection for all body types, not just the average-male crash-test dummy. 

Safety officials have known for 40 years that male and female bodies respond quite differently in automobile crashes. In November of 2025, after years of debate, the Transportation Department signed off on the use of a new, female-like crash dummy, though it’s far from clear the new model will become widely used.

Regulation as End Goal or Starting Point?

Wu’s main analysis compared overall fatality outcomes for crash victims whose height and weight closely matched the crash-test dummy’s dimensions to those of other body types. She found that dummy-sized occupants experienced about a 72% reduction in fatality risk after the regulation was implemented, while the risk for individuals of other sizes — shorter, taller, lighter or heavier — dropped by only about 41%.

That disparity suggested not all firms were focused on expanding beyond the narrow metric set by the regulation. That led Wu to explore operational differences among companies as a potential explanation for how firms chose to respond to the regulation. She focused on firm-level commitment to safety — and ability (resources) to conduct detailed testing:

• Wu examined whether firms had a history of addressing safety problems proactively — before regulators forced them to act — signaling that safety was a genuine priority. Specifically, Wu looked at whether companies, notified of safety problems with a model, acted before the government issued a mandatory recall or waited for that order.
• To capture a company’s testing capability, Wu reviewed nearly 4,000 auto-related patents issued before 1994, tracking how often the patents cited scientific research on measurement and testing. She posits that companies with patents drawing heavily on measurement science were more likely to have the infrastructure and habit of testing how their designs performed across a range of body types, not just the single mandated dummy size.

Wu finds that companies with both traits were most likely to deliver broad side-impact protection to a variety of body types. These firms achieved fatality reductions of about 54% across all crash victims. Companies that didn’t focus on safety and lacked strong measurement capabilities did the narrowly assigned job of protecting occupants matching the dummy — delivering reductions around 63% for that group — but showed no statistically significant improvement in fatalities for anyone with a different body type.

A Weighty Difference in Design Choice 

Wu also found that a company’s commitment to research and development seems to have had an important impact on the design choices made to meet the side-impact test. Firms without a testing culture relied on what they knew would work: steel reinforcements and thicker padding. That indeed produced better protection for dummy-sized crash victims, but it also added about 39 pounds to a car’s weight, and Wu estimates reduced fuel efficiency by about 1mpg. That is not ideal for consumers or for environmental regulators.

Companies with stronger measurement expertise tended to avoid the weight penalty. In several specifications, they were able to show modest reductions in vehicle weight without statistically significant losses in fuel efficiency. Wu interprets this as consistent with these firms’ testing infrastructure allowing them to quantify the benefits of advanced materials — high-strength composites, reinforced polymers, treated aluminum — and thereby justify the higher research costs. Without that capability, firms defaulted to the familiar but less efficient steel options.

While Wu’s study centers on a crash test dummy from the 1990s, her research should be of interest across regulatory fields. Nearly every regulation imposes measurable targets: parts-per-million emissions limits for air quality, capital-to-asset ratios for bank stability, maximum response times for emergency services, disclosure thresholds for financial reporting. These metrics are essential — they transform abstract policy goals into concrete, enforceable standards.

But as this research shows, some companies view a given regulatory metric as the entire mission and design just to meet the specifics of a given rule. Others view a metric as a minimum requirement that can be improved on. Wu’s research suggests that’s not some random choice, but the result of corporate strategy and capabilities. 

If regulators want rules to deliver broad protection, they should consider how metric design interacts with the diverse cultures and capabilities of the firms tasked with following a given regulation, for example, by offering multiple metrics, encouraging industry engagement or providing support for firms to build in-house measurement capabilities.