Policymakers in several cities have tried to change that.
The goal was straightforward: ensure that the people bringing you your lunch earn a decent living.
We compared what happened to the earnings of drivers who were primarily working in Seattle before the law took effect with the earning of drivers working in other parts of Washington state, where nothing had changed. By tracking both groups over the months before and after the policy, we isolated the policy’s impact from broader trends affecting all drivers.
The other major change was that drivers started completing fewer deliveries.
Beginning in the second month after the policy took effect, Seattle drivers who had been consistently active on the apps prior to the change completed roughly 20% to 30% fewer monthly deliveries than they would have without the policy.
Importantly, these drivers didn’t leave the apps. They were still logging on and spending about the same amount of time working. They just weren’t getting as many delivery offers.
What were drivers doing with all that extra time on the app? Our data shows they were spending more of it waiting.
The share of on-app time spent actually performing paid deliveries fell substantially. Wait times between tasks increased by about five minutes, nearly doubling from pre-policy levels. And drivers went farther between deliveries – suggesting they were actively cruising toward restaurant-dense areas to find their next task, burning more gas without being paid for those extra miles they were logging.
Put those pieces together – higher pay per delivery, but fewer deliveries and lower tips – and they almost exactly cancel out. After a brief bump in the first month, monthly earnings returned to pre-policy levels.
To understand why this happened, it helps to think about how gig delivery markets differ from traditional employment.
In a conventional job, raising the minimum wage creates a clear divide: Workers who keep their jobs earn more, while others may struggle to find work if their employers cut jobs.
But in gig delivery, there’s no such divide. There’s no hiring or firing involved; anyone can download the app and start looking for work. Delivery tasks are distributed among everyone who is online, and there’s no sharp boundary between having a job and not having one.
When what drivers get paid per delivery rises, gig work becomes more attractive, drawing new drivers into the market. Meanwhile, higher costs to pay drivers are passed along to consumers through increased delivery prices, which can lead to fewer orders and lower tips. More drivers chasing fewer deliveries means longer waits for tasks.
This process continues until the higher pay per task is fully offset by the longer gaps between paid work.
Our data confirms this pattern.
While deliveries by existing drivers fell sharply in Seattle, new entrants arrived. Within three months, newcomers were doing most of Seattle’s deliveries.
To be sure, gig workers’ low pay is a real problem. The impulse behind Seattle’s law reflects legitimate concerns.
But our findings do suggest that efforts to directly regulate what gig workers earn per task they complete won’t easily fix that problem.
As long as anyone can join the platform and start competing for deliveries, the guarantee of higher pay per task will attract more drivers until the benefit is competed away through longer wait times.
Still, there may not be a solution that preserves all the benefits of the current system while also guaranteeing higher pay.
Seattle’s experience suggests all cities should proceed with caution and be aware of the limits of what per-task pay regulations can achieve when the door is always open to new workers.
