Allowing Californians to opt into usage-based insurance would save lives | Opinion
What if there was a simple program that lowered the number of car crashes and the cost of insurance? What if it was available to everyone but Californians?
That program exists, and it’s called usage-based insurance. Every major U.S. auto insurer offers a usage-based insurance program, and one in four drivers choose to participate. But, because of a ballot measure passed almost 40 years ago, Californians don’t have that choice.
The typical usage-based insurance program involves installing a smartphone app that monitors driving for a few months. During this period, drivers get feedback about risky behaviors, such as hard braking. Insurers use this information to predict a driver’s risk of crash and offer a policy priced accordingly. Safer drivers get bigger discounts.
Growing evidence suggests that customers who enroll in usage-based insurance programs drive more safely, and our new research confirms that. We recruited drivers from across the country and enrolled some in a simulated program. They got driving feedback and could earn up to $100 for driving safely. We found that these drivers reduced their hard brakes and rapid accelerations by 20%. Their rate of speeding decreased too, by 12%.
What’s more: After the feedback and incentives ended, they kept driving safely. The program helped build lasting habits.
These safety improvements could reduce crashes by 4% and crash claims by 7%. If everyone in California participated in a usage-based insurance program, this would mean about 7,000 fewer crashes per year and $150 less in annual premium payments for the average driver.
Yet, California remains the only U.S. state where usage-based insurance programs are not permitted. Under Proposition 103, passed in 1988, insurance companies may use only three factors — driving safety record, annual miles driven and years of driving experience — to price policies. By limiting insurers to these three factors, the law prohibits price discrimination based on superficial characteristics like a driver’s age or sex.
But these limits also mean insurers cannot collect and use the best information now available to accurately predict crash risk. The law has also had unintended consequences. Insurers routinely base crash predictions on driver records of moving violations, and these records may reflect biases in traffic enforcement.
Resolution ACR 52, under consideration for the 2025-26 legislative session, seeks to change the status quo. It urges the Insurance Commissioner to interpret Prop. 103 in view of today’s technology. This would allow crash-predictive behaviors measured by usage-based insurance to count toward one’s “driving safety record.” The resolution stresses that the decision to enroll in a usage-based insurance program should rest with drivers, and that programs must maintain drivers’ privacy.
We recommend passing this resolution and making usage-based insurance programs available to the millions of drivers in our nation’s largest state. The safety and economic benefits are increasingly clear, and other states have shown that programs can be regulated to protect consumers and their data.
To maximize the benefits, we also encourage regulators to allow usage-based insurance programs the flexibility to provide feedback and incentives designed with insights from behavioral science. For example, our prior research suggests that while offering a discounted price on their upcoming policy can motivate drivers to reduce distracting phone use, offering them smaller, weekly rewards for doing so motivates even greater behavior change.
The stakes are high: Across the U.S., over 40,000 people die each year because of car crashes. This figure includes 4,000 Californians. If we want everyone who gets in a car to make it home safely, shouldn’t the state use all the tools available? Our research says it’s time to reopen the toolbox.
Jeff Ebert, PhD, is director of Applied Behavioral Science at the Penn Medicine Nudge Unit. M. Kit Delgado, MD, MS, is executive director for Transformation Science and Behavioral Insights at Penn Medicine and associate professor of Emergency Medicine and Epidemiology at the University of Pennsylvania Perelman School of Medicine who previously practiced emergency medicine in three California Level I Trauma Centers. The opinions expressed in this article do not necessarily represent those of the University of Pennsylvania Health System or the Perelman School of Medicine.