AUSTIN, Texas — A joint study conducted by the University of Michigan Transportation Research Institute, the Texas A&M Transportation Institute and Columbia University found that half the Uber and Lyft users in Austin, Texas, used or bought a personal vehicle when the ride-hailing providers temporarily suspended services in that city last year.

Without specifically mentioning the retail automotive sector, lead author and UMTRI professor Robert Hampshire said his team’s findings prove the growing availability and utilization of Uber and Lyft have a discernable, quantifiable effect on consumer behavior.

“Our findings show that these ride-sourcing companies do change behaviors,” Hampshire said in a statement accompanying the study. “While this study isn’t generalizable to the entire city of Austin, or to other cities, it provides crucial insights that are relevant to policymakers.”

Uber and Lyft suspended services in Texas’ capital city in May 2016 after failing to convince city officials to waive a local ordinance requiring fingerprint-based background checks for new drivers. State officials intervened on their behalf and the ride-hailing companies resumed operations in Austin one year after departing.

After polling Austin residents, the study’s authors determined that 42% of respondents sought out other, smaller ride-hailing services during Uber and Lyft’s absence, but 41% said they went back to using their own vehicles, and 9% purchased an additional vehicle. Only 3% said they used public transportation.

The findings dovetail with “Shared Mobility and the Transformation of Public Transit,” a groundbreaking 2016 study published by the American Public Transportation Association. APTA’s analysts found that Uber and Lyft users owned, on average, 1.05 vehicles per household, compared to the average of 1.5 vehicles per household occupied by drivers who do not use ride-hailing services.