I was a Product Manager at Postmates and Lyft from 2016 to 2020 and I'm voting No on Prop 22

California Proposition 22 (Prop 22) would consider app-based drivers to be independent contractors and not employees or agents. Therefore, the ballot measure would override Assembly Bill 5 (AB 5), signed in September 2019, on the question of whether app-based drivers are employees or independent contractors.

Before I start on my opposition to Prop 22, I want to say that the people I worked with at these companies are some of the most compassionate and thoughtful people I’ve met. I credit a majority of my career growth thus far to the time I spent at Lyft learning from my colleagues. I truly believe the humans in the corporate office are doing their best to make the gig economy work better for the humans powering it as drivers.

With that said, I want to share my perspective on how this product works, where I think it’s failing our community, and why I think Prop 22 is not the right path forward. There are trade-offs here (which I lay out), and the decision isn’t easy, but I do think No on Prop 22 is the best call.

My argument against Prop 22 boils down to four points:

  1. Drivers currently lack control over most of their work and have more in common with employees than independent business owners

  2. Drivers don’t get paid for all of the time that they are engaged with their work

  3. Drivers and riders are BETTER OFF when more control is exerted over drivers’ work

  4. The minimum wage is a net-positive for society

I also want to be clear that nothing I’m sharing here is confidential. You can find out any of this through talking to drivers or working as one yourself.

1) Drivers currently lack control over most of their work and have more in common with employees than independent business owners


Imagine you just started a side gig as a writer. Or as a craft jewelry maker. Or as a piano teacher. You might decide that the best platform to monetize your work is Substack (writing), or Etsy (craft supplies), or Thumbtack (local professionals). You set up a page for customers to consume your product. You set whatever price you want. You write in whatever style you want about whatever you want. You make your jewelry look however you want. You decide how to best teach your piano students. This is what being a true independent contractor looks like, where you as the gig worker have control over how you do business with your customers.

In reality, being a Lyft or Postmates driver is closer to being an In-N-Out cashier. There’s an immutable price for their time spent selling customers their burgers. Same goes for each ride or delivery a driver is offered. If an In-N-Out cashier refuses service to customers placing complicated orders, they’d get fired. If a driver rejects or cancels too many rides or deliveries they are dispatched, they risk losing their ability to do work on the platform (official stancedrivers’ experience).

I believe these things should matter when determining if a worker should be classified as an independent contractor or an employee. While imperfect, AB5 gets us closer to that better world for gig economy drivers, and Prop 22 takes us farther away.

There are of course also many differences between a driver and a cashier, mostly around the ability to work anywhere you want whenever you want. There’s also the ability to work across multiple platforms at the same time. However, I believe these purported “benefits” are a headfake. The things that matter the most to most drivers, like how much they get paid, is out of their control.

2) Drivers don’t get paid for all of the time that they are engaged with their work


If you talked to a Lyft or Postmates driver about what parts of the job they get paid for, or if you signed up to work as a driver yourself, you’d discover that you do not get paid for your time spent online and waiting to be dispatched for a ride or delivery. In some cities, you may not even get paid for the time you spend driving towards the pickup location. The prevailing approach across all platforms is to pay you only once you have completed your drop-off.

This incentivizes drivers to not cancel pickups and to complete their drop-offs, which is good for riders. This also means Lyft and Postmates can afford to have more drivers hanging around waiting for dispatches, which makes your rides and deliveries significantly cheaper and faster than they would be otherwise. But because of this approach to payment, drivers are at best getting paid for ~67% of their time, and even with Prop 22’s partial wage guarantees end up making far less than minimum wage.

These privileges we enjoy as riders are predicated on drivers spending lots of time on shift waiting for demand they do not control to be ready for them. This is often incentivized in various ways by the platforms. I’d equate it to only paying an In-N-Out worker for every burger they sell, and not for their time waiting for customers to come in.

I do think it’s worth noting that the effect voting No on Prop 22 would have on the pricing and availability of ridesharing would disproportionately affect riders in communities underserved by public transit. This is bad. But in my view not as bad as the pain caused by unpaid driver labor. And also something we should aim to fix by solving the root problem: lack of public transportation.

3) Drivers and riders are BETTER OFF when more control is exerted over drivers’ work


As a customer


Would you want a ridesharing app where the drivers can charge anything between $0.01 and $1000 for a 15 minute ride? Where you have to sort through the price point each driver is offering a ride for before picking one? How about one where a driver can refuse service to whoever they want?

This would make for a confusing and onerous marketplace. Our experience as riders today is much better exactly because drivers don’t have full control over how they do their work. Lyft and Postmates exerting control over their drivers in these ways is a great customer experience.

As a driver


The two biggest differences between working as a Lyft driver and working as an In-N-Out cashier are ownership of work equipment and flexibility. I think that ownership of work equipment is a bad thing. I think that flexibility, while tempting, is in this case a net-negative.

On ownership of work equipment: In-N-Out provides their employees a uniform, spatula, cash register, and more, for free as a part of their job. Lyft’s model is predicated on drivers using their own vehicles, which significantly eats into their profits as a driver (UC Berkeley study).

On flexibility: While In-N-Out lets you schedule a shift ahead of time, Lyft lets you work whenever and wherever you want on-demand. This may feel good since it gives optionality to drivers, but since they have no control over demand and how many other drivers try to work at the same time as them, this often ends up a net-negative for drivers!

This research study by MIT and NBER shows that when drivers follow incentives to go online at the best times to work, their on-the-job utilization per driver decreases and as a result average earnings for each driver go down. This is because, as mentioned above, they only get paid for every ride or delivery they complete. It’s like if In-N-Out let a third cashier work during a shift that only needed two cashiers, and all three of them split the earnings that just the two cashiers would have otherwise made.

I’m arguing here that more control over drivers’ shifts would make their jobs better. If Lyft and Postmates had to pay their drivers minimum wage for all time worked, they would be incentivized to also plan out shifts better. Shifts would solve this tragedy of the commons, and Drivers would make more money commensurate to their effort.

It is worth noting that flexibility is, of course, not entirely negative. Because of flexibility, students can drive on the side whenever their schedule allows. Parents can take a month off with no one's permission. Lyft and Postmates can help some people make rent when they are behind with 5 days left in the month. This is a tough trade-off I’ve struggled with, but for every driver I’ve spoken to who enjoys the flexibility, there were many more who talked about driving into San Francisco from several hours away, who work around the clock and regularly sleep overnight in their cars.

4) The minimum wage is a net-positive for society


One of the biggest arguments in favor of Prop 22, and more generally against the notion of a holistic minimum wage for gig economy drivers, is that it would eliminate many jobs. These platforms are already cautioning that they won’t be able to provide work opportunities for a significant percentage of their existing driver pool if Prop 22 fails.

However, there is ample evidence that the minimum wage does not hurt jobs (herehere, and here). I see no reason to believe this will be different for drivers who work 20+ hours a week. Since this has never been tested in this specific context, it’s hard to be certain. It is likely that drivers currently working only several hours a week will not be able to do so anymore without committing to the job more fully, though I think we can be more creative here.

In my view, this entire issue ultimately boils down to whether or not we, as voters, think that the minimum wage should exist. As with all important decisions in life, there are trade-offs to consider here. To me, it’s obvious that the minimum wage debate isn’t just a yes or no question. There’s plenty of nuance. It does help some groups (the working class) and hurt others (business owners looking for more labor). There are likely locations or time frames, such as during a recession, where the effects of a minimum wage are different. But I believe it nets out as something we need given how much it helps the least economically privileged people in our country.


1) But don’t most drivers support Prop 22?


There are some surveys that suggest this may be true, but the only ones I’ve seen have a small and unrepresentative sample size. For example, this survey of just 600 drivers.


  1. Uber/DoorDash/others are exerting a lot of pressure on drivers to vote yes, which I worry leads to a conflict of interest (12)

  2. I expect that that while drivers who occasionally drive may prefer yes (let’s say, less than 10 hours per week), the drivers most dependent on the platforms for their livelihood would prefer No on Prop 22. I elaborate on this more at the end of section 3 of this post

2) Isn’t there a third way? Something better than Prop 22 and better than AB5?


Sure! But until I see it, I think it’s our duty to protect the population most harmed by our current form of ridesharing. Which is drivers who depend on the platforms for a majority of their livelihood (eg: working 20+ hours per week).

I’ll write about this in more detail sometime, but for now I’ll say I would recommend something roughly like:

  1. Guaranteed minimum wage and benefits for drivers working over X hours / wk

  2. For drivers working under X hrs/wk, the option to work as independent contractors

  3. Radical transparency on projected earnings including idle time, baking in variable cost of gas / car depreciation / etc., so that drivers can make an informed decision as to if being for example the 100th driver to circle a given neighborhood at 6:00pm on a Thursday evening is a lucrative decision

3) Would voting No on Prop 22 lead to more drunk driving?


I have seen zero evidence to substantiate this claim. And even if there were a marginal risk of this, is that worth perpetuating an underclass of partially unpaid drivers?