A 2026 National Bureau of Economic Research study ran a managed-charging experiment across EV households at Peninsula Clean Energy (PCE), a community choice aggregator on the San Francisco Peninsula with one of the highest EV adoption rates in the country, and found that even when households were paid $40 a month to take part, only 4.6% of households chose to enroll.
The finding comes from economists Fiona Burlig, James B. Bushnell and David S. Rapson. Burlig is at the University of Chicago; Bushnell and Rapson are at UC Davis. It threatens to upend one of the electricity industry’s central assumptions about how the grid will cope with millions of new charging loads.
What the experiment tested
Despite the importance of program participation for policy, treatment effects are often measured on self-selected samples. Prior work finds large impacts of managed charging among households who volunteer for a randomized controlled trial. In contrast, the NBER team tested managed charging with an experiment including all EVs within a California utility.
The researchers ran a randomized controlled trial in partnership with Peninsula Clean Energy, a community choice aggregator on the San Francisco Peninsula serving San Mateo County. PCE has one of the highest EV adoption rates in the country.
Enrollment was low even with high incentives, and the researchers were able to reject even modest intent-to-treat effects on electricity consumption. The paper, NBER Working Paper No. 35086, was dated April 2026 and has since drawn wide attention from utility planners and regulators.
The broader industry picture
Approximately 6.5 million electric vehicles are on the road in the United States, simultaneously driving the need for grid upgrades and offering utilities a potential tool to defer those expenses. Utilities are now working to confront this challenge. One option is for utilities to manage the charging itself, requiring the EV owner to grant the utility control over their vehicle’s charging behavior, allowing the utility to shift when the vehicle charges to times when demand is low.
But the NBER results suggest that voluntary programs cannot reliably deliver that benefit at scale.
The Smart Electric Power Alliance separately reports that where residential time-of-use rates exist at all, opt-in adoption rarely tops 10 percent. Combined with the low managed-charging enrollment, most utilities are actively managing only a small fraction of the EV load already on their systems, often in the single digits.
A related NBER study evaluated the effects of one of the world’s largest AI-managed EV charging tariffs using a large-scale natural field experiment and found that the tariff led to a 42% reduction in household electricity demand during peak hours, with 100% of this demand shifted to lower-cost and lower-carbon-intensity periods, though that program also relied on active financial incentives and participant enrollment.
Grid reliability on the line in high-growth scenarios
The stakes of low enrollment rise sharply as EV adoption grows. Using actual PG&E transformer capacity data for San Mateo County, the Burlig team simulated what current enrollment rates mean for local grid reliability.
Under a low-EV-growth scenario, the observed enrollment is sufficient to keep most feeders (56%) out of trouble. Under a high-growth scenario closer to what California’s electrification targets imply, it covers only 9%.
To meaningfully relieve distribution constraints once the EV transition scales up, the researchers found that majority enrollment among Level 2 charger households would be needed. That threshold is far beyond reach, and the cost of compensating customers to get there could be prohibitive.
What comes next for utilities
The authors conclude that voluntary sign-up is not a path to grid visibility, and that to act effectively, a utility must know how many EVs are on its system and, critically, where and when they charge. That points toward automatic enrollment schemes, direct load control, or rate designs that remove the opt-in step entirely.
Utilities that encourage EV charging in areas with surplus capacity can capture maximum rate benefit. Those that allow uncoordinated charging in constrained areas face infrastructure costs that transform a rate-reducing opportunity into a rate-increasing burden.
The debate over mandatory versus voluntary managed charging is now moving from academic papers into state public utility commission dockets, particularly in California, where the rapid build-out of battery storage is adding new complexity to how overnight EV load is managed alongside other flexible resources.
The paper, “If You Build It, They May Not Come: Willingness to Participate in Managed EV Charging,” NBER Working Paper 35086, is available at nber.org.
