Total: 1
We study the impact of a major policy and design change at Lyft that altered both driver earnings and platform transparency, offering insights into how such changes affect stakeholders and platform outcomes. In February 2024, Lyft began a staggered rollout of a new policy that guaranteed drivers a minimum share of rider payments and increased transparency by displaying estimated earnings per ride upfront. This policy was first introduced in major urban markets, creating a natural experiment to evaluate its effects. Using data from over 47 million rides across urban and neighboring suburban markets, we apply dynamic staggered difference-in-differences and geographic border strategies to measure causal effects on driver behavior, rider experience, and platform performance. We find the policy significantly increased driver engagement-particularly among those with lower pre-policy earnings or higher income uncertainty-leading to more hours worked, higher utilization, and greater trip volume. These supply-side changes also generated positive spillovers on rider demand. We disentangle the separate effects of earnings guarantees and transparency and show that while both were beneficial, transparency may have also triggered strategic driver behaviors. In ongoing work, we develop a counterfactual simulation framework linking driver supply and rider intents to ride production, showing how small behavioral shifts could further amplify platform outcomes. We also train a self-supervised model on driver trajectories to detect multihoming, examining whether the observed supply increase reflects net expansion or substitution from other platforms. Together, our findings highlight the potential for platform-led policies to serve as alternatives to regulation and offer design insights for managing platform change.