James Parrott tells the inside story of the first-ever pay floor for Uber drivers

"It was very clear from our early analysis that this was an incredibly inefficient industry"

Photo by Alexi Rosenfeld/Getty Images

In July 2018, economists James Parrott and Michael Reich published one of the most important reports ever written on the gig economy. Previous research had attempted to estimate the earnings and hours of ride-hail drivers, but “An Earnings Standard for New York City’s App-Based Drivers” went further, proposing a minimum living wage for the city’s ride-hail drivers and setting out the policies needed to make it happen.

At the time the New York City Taxi and Limousine Commission (TLC) commissioned the report, drivers were in crisis. The vast majority of ride-hail drivers in New York City are not part-timers but immigrants for whom driving can be one of their only work options and makes up most if not all of their income. Many of these drivers had joined Uber during its early New York City heyday, when bonuses flowed freely and the company teased annual earnings of $90,000, but several years later, as the explosion in drivers outpaced demand, found themselves struggling with low earnings, high operating costs, and in many cases, exploitative subprime vehicle leases.

The Parrott and Reich report proposed a wage floor of $17.22 an hour, designed for drivers to take home $15 an hour after on-the-job expense like gas and vehicle maintenance. That represented a 22.5% jump in net earnings, or an increase of $6,345 per year for the average driver. The key to the proposal was a pay standard that took into account the share of time a driver spent with a passenger in their car, rather than simply logged onto an app like Uber. Parrott and Reich called this the ‘utilization rate’ and, in accounting for it, devised an earnings formula that penalized companies that flooded the streets with drivers but failed to provide them with sufficient work. In other words, this was a twofer, setting a minimum wage for drivers while also taking a stab at the problem of urban congestion.

Frequent readers of Oversharing will know that I’m a big fan of this pay formula, which I think is incredibly clever. So I was extremely excited to speak with James Parrott, director of economic and fiscal policies at the Center for New York City Affairs at The New School, for this weekend’s subscriber-only interview. We talked about the origins of the gig economy and fissured work, the story behind that famous TLC report and the driver pay floor, and who came up with the utilization rate magic. We also chatted about his latest report on low-paid misclassified workers, which I wrote about on Thursday.

Oversharing interviews with smart people like Parrott are available exclusively to paid subscribers. To access this transcript, upgrade your subscription, or use this link to purchase a discounted group subscription for your company. This interview has been condensed, edited, and supplemented with links for clarity. Also, please know that the AI software I use repeatedly transcribed ‘misclassification’ as ‘Miss Classification,’ which if I didn’t know better I’d think was its attempt at a joke.

Oversharing: Thanks so much for doing this! It’s great to have you on. To start, I wanted to ask what got you interested in the gig economy?

I followed labor economic policy issues for decades and had always done a lot of work on minimum wage. I’d been involved in New York State on trying to improve unemployment insurance and workers compensation.

It had become clear that there was a growing extent of misclassification in New York State. This was long before apps. So even before apps came on the scene, in the mid-part of the first decade of this century, there was an explosion in misclassification in New York. We had a Republican governor at that time, who ultimately served 12 years in office. The [state] Labor Department's enforcement pretty much disintegrated over that time. Employers recognized that and saw that they could reap considerable labor cost savings by misclassifying workers. So there was an explosion in misclassification.

I worked with a number of Building Trades back in 2005-07, on reports drawing attention to the misclassification problem in construction. It was particularly severe in construction because workers’ compensation premiums in construction can be as high as 15-20% of wages just by itself. And of course, if a construction employer can get away with misclassifying workers as independent contractors, they’re not paying that workers comp. They're not paying unemployment insurance, taxes. They're not paying any benefits and they're usually paying lower wages. A lot of times they're paying workers off the books.

Was this a trend nationally at the time, or was it particular to New York State?

It was a trend nationally to some extent. It was more pronounced in certain states than in others. It was most heavily rooted in construction and transportation. Again, this was all before ride-hail. FedEx for many, many years tried to claim that their drivers are independent contractors. So they were on that kick long before Uber and Lyft came on the scene. I think Uber and Lyft learned a lot from FedEx. And they involve their narrative about how workers really want this flexibility and so on. But they took their cue to the fact that if you misclassify workers as independent contractors, you can save a lot. You can save 30-40% on your labor costs. So that trend was well established by FedEx and construction companies. And Uber and Lyft come along and see that and say we're going to build our whole business model and get the drivers to buy their cars to boot.

I think people also forget that this wave of gig companies like Uber and Lyft really got started during the Great Recession, when a lot of people had lost their jobs and savings and were desperate to have any source of income. So if you had a car, being able to log onto an app and make some money was very attractive.

The fissuring phenomenon really started to take off in the 1980s and ’90s in the U.S. with a lot of subcontracting and then firms misclassifying workers as independent contractors, and again, a lot of that first appeared in construction. Fissuring is an employer practice of undermining the traditional employee-employer relationship. It takes various forms and misclassification is one of those. It caught on in media and other industries, as well, because the labor cost advantages were so alluring to a lot of employers.

Because labor in the U.S. has been so weak politically since the 1970s, the labor movement has never been able to really fashion an effective political response to misclassification. That's why the battles have tended to play out at the state level, because organized labor doesn't have sufficient political clout at the national level. They haven't even been able to increase the minimum wage in the U.S. since 2007. They haven’t been able to pass labor law reform since 1977. So because labor is weak, this misclassification and the broader fissuring set of issues has really expanded over the years. So anyway, the long and the short of it is misclassification was well-established in the non-app world before Uber, Lyft, and other platform companies come along in the wake of the Great Recession.

When did you start paying attention to Uber?

I was living in New York City at the time Uber really took off. It became very clear that it presented a serious public policy challenge, because we have a well-developed mass transit system in New York City. We have a highly regulated medallion taxi system. And it was very curious that a company like Uber could come along and expand so rapidly in providing taxi-like services and not be subject to any regulation.

There was an effort, of course, on the part of Mayor de Blasio and the city council in 2015, to try and cap the growth in Uber and Lyft. Uber has always hired a lot of well-informed political operatives, and they mounted a very effective mobilization effort to beat back any hint of the city regulating Uber and Lyft. So that was in 2015. Bill de Blasio originally got elected in 2013. He is up for re-election 2017. My own reading is that he didn't want to risk taking on Uber again until he was safely re-elected. So nothing happened. Meanwhile, Uber continued to expand. That added to problems that the medallion drivers were experiencing.

I think the problem with medallion valuations is a separate issue. The valuations certainly were undermined by the rapid expansion of Uber. But Mike Bloomberg, who was mayor at the time the valuations of medallions really took off, should never have let that happen. It's inconceivable that any taxi driver could pay $1 million and still expect to service the debt needed to buy a medallion and survive. That's just a simple calculation. So it was very clear that medallions had turned into a speculative instrument.

But Uber’s unlimited expansion was creating lots of problems. And for the first time during an economic expansion, when employment was expanding, there was a decline in ridership on the subway system because of Uber. And in a climate-conscious world, that's just mind-boggling. Why wouldn’t you have a policy response? My sense is that de Blasio didn't want to do anything until 2018.

The best cap on Uber would be to fix the subway

You famously co-authored a report for the New York City taxi commission in which you proposed a way to set a wage floor for drivers, which at the time had not been done. People had toyed with the idea, but there was a lot of focus on reclassification. No one had really managed to tackle this problem of just how do we raise pay for drivers?

The TLC was getting a lot of complaints from drivers, both app drivers and medallion drivers, about low earnings. Somebody suggested that I testify at the hearing on driver pay in June of 2017. I did a little bit of research and saw how their earnings had fallen, and I just echoed what the drivers were saying—this is a serious, serious problem and the city should figure out how to better regulate this industry. We started talking about doing a study a couple of months after the hearing. We started the study in 2017, knowing that it wasn't going to be finalized until 2018, after de Blasio had been re-elected. It came out in July of 2018.

The idea for the pay standard was really the TLC’s idea. It was suggested by the drivers and the TLC sort of formalized the idea. Then we were asked to investigate what their earnings actually were, to help the TLC develop the formula for applying the standard, to make sure that all of drivers’ working time was compensated, that all of their expenses were compensated, and so on. We jointly worked with Meera [Joshi, then TLC commissioner] and her staff at the TLC to develop that recommendation, which then got enacted later that year. It first took effect in February of 2019. The city of Seattle followed a similar path a couple of years later and Michael Reich and I did a report for the city of Seattle in July of 2020.

As you probably know, I love your pay standard formula, I think it's brilliant. I've always wanted to ask you: who came up with the magic utilization rate component?

It was very clear from our early analysis that this was an incredibly inefficient industry. It worked for Uber in the sense that they could flood the city streets with cars, because they would make a portion of the fare as their commission all the time. But there was no incentive to limit the number of drivers or to utilize the stock of automobiles that was wholly financed by the drivers. It was incredibly inefficient. And that in order to instill an element of efficiency there had to be an incentive for companies to increase utilization and to increase the portion of the drivers’ time that they had a passenger in the car. And that that would also increase economic efficiency, decrease the number of miles driven, and it was part and parcel of increasing drivers’ pay. The drivers would have a better opportunity to earn a decent living if there were fewer other drivers out there competing for a limited number of trips at any one time.

It was heavily informed by Meera and her staff, and we went back and forth several times on that. But Michael and I really sort of held the line on keeping utilization in the formula, because without that the companies wouldn't have their own incentive to do that. And we also saw it as a way that the companies by becoming more efficient in their use of drivers could help pay the higher driver fare. That was premised on the idea that if they increased utilization, the amount that they paid drivers per trip eventually would go down. The TLC still hasn't actually utilized the utilization rate in the intended way. So if the companies did increase utilization, they were supposed to get rewarded with a lower effective amount that they had to pay the drivers. Because as utilization goes up, since it’s in the denominator, the amount they pay the drivers per trip would go down.

Right, it was that each company's actual utilization was supposed to be being updated in the formula, rather than stay at the 0.58 starting point.

Right. So that's never been done. And in the data that we saw before the pandemic, it looked like utilization was still in that neighborhood. It was much higher for Via of course. But it was around 0.58 for both Uber and Lyft. Then when the pandemic came along, who knows what utilization was at that point. But it's not clear if they're going to try and go back to that or not. And Seattle I think decided to not even build in utilization in their formula.

At the time that you were working on the TLC report, were you familiar with the working paper by Uber’s Jonathan Hall and economist John Horton on pricing in ride-hail markets? It basically argued that Uber couldn’t raise driver earnings because, in an open market, higher rates would lead more drivers to work more hours, increasing labor supply, while also making trips more expensive, decreasing rider demand. Drivers might make more per trip but ultimately their wages would revert to an equilibrium because there would be less work to go around and utilization would fall.

The irony was that if you took this paper and inverted it, it was a great argument for capping the number of Uber drivers on the road, to make the market less elastic. And that’s pretty much what your pay formula did with the utilization rate.

Yes, we grappled with that paper a lot. We felt like we needed to respond to their argument. They tried to model that if you did increase fares, that it would only benefit drivers for a short period of time, but then as you described, as other drivers realized the better earnings opportunity and entered the market, that would expand the supply of drivers and bring the wages back down. So you pretty aptly described it as ‘we can't raise wages because it won't have any long-term benefit for drivers.’

But we always recognized that you've got to do something to limit the number of drivers. Our preferred approach was to do it through utilization, sort of to make it in the company's interest. The Taxi Workers Alliance always thought that the best way to do that was just a moratorium on the number of cars. And they succeeded in getting the city council to adopt a moratorium which took effect in August of 2018.

But it did seem like the growth was starting to taper off in the latter part of 2018 anyway. In December of 2020 we evaluated the first year of the pay standard. We accessed some data that the city of Chicago had on its ride-hail companies. We could see in Chicago, where there was no compensation standard in place, the number of trips also started to level off in 2019. So it wasn't just the effect of the moratorium in New York City or even the utilization rate, but that in a lot of big cities they were close to saturating the market anyway, and the potential for further growth was limited.

You have a new report on misclassification in New York State, particularly among low-wage workers. One of the main points you make is that while we might think of this predominantly as a gig issue, there's actually more far more people who are affected by this problem that aren't app-based.

We found that whereas there are about 200,000 app-based workers in New York, altogether there are 875,000 what we would call misclassified low-paid workers. So that includes the 200,000 app workers, but means there are 675,000 non-app low-paid workers in industries like transportation, construction, janitorial services, personal services like nail salons, retail, and some other industries.

This problem predates the advent of the platform economy. It's an example of the fissuring of work in the U.S. beginning in the early 1980s and that really expanded in the 1990s and the first decade of this century. New York State has identified for the past 13 or 14 years, on average, 150,000 misclassified independent contractors in low-wage industries like construction, janitorial services, transportation. And this is without any enforcement [of the platform companies].

How are these misclassified workers being identified?

The Labor Department does audits of the payroll records of companies. If a company is using independent contractors, the state Labor Department will determine whether or not they're a legitimate independent contractor or whether they're in effect an employee who is misclassified as an independent contractor. In many cases, the Labor Department learns that employers were paying workers off the books entirely. So that's verboten, and that's considered a form of misclassifying workers.

This average of 150,000 cases, those are individuals. That's not an estimate. Then the Labor Department will send a bill to the company for the unpaid workers’ comp and unemployment insurance, and if there are violations of wage and hour laws then pursue back wages. This is an indispensable part of maintaining labor standards in this country and in New York it’s primarily done at the state level. The federal government years ago, going back to when Ronald Reagan was President, really decimated the federal wage and hour enforcement staff. There are something like 1,000 wage and hour investigators for the whole country, and estimates are that given how long it takes to investigate a given employer, it would take 400 years to investigate all employers.

In your report, you talk about how misclassified workers earn, on average, a third less than other workers—and that doesn't include other compensation in the form of workers’ comp, benefits, access to paid leave…

Or health insurance, or paid time off of any form. Misclassification, the reason employers do it is because it provides considerable cost savings. Now, it also gives the firms that do it an unfair competitive advantage against companies that play by the rules. It really is an untenable situation where law enforcement can't let the noncompliance go unenforced.

New York City and New York State have acted in recent years to have a Paid Family Leave law and a paid sick days requirement for all employers. But again, for workers who are misclassified as independent contractors, they don't even have that coverage. And of course, they're not covered by unemployment insurance or workers’ comp, and the employers are not paying into Social Security on behalf of the workers. In the U.S., Social Security is jointly funded by workers and employers. They each pay 7.65%.¹ Misclassified independent contractors, if they're complying with the law when they fill out their tax return, they're paying both the employer share and the employee share of Social Security.

We were talking about how this really came into focus during covid, because suddenly a lot of people were out of work, a lot of people needed health care or paid time off or sick leave, and many of them were independent contractors.

It became much better understood how important employee status was vis-à-vis independent contractor status. But it surfaced mainly in New York in a dispute over whether ride-hail drivers would receive regular unemployment insurance or the Pandemic Unemployment Assistance (PUA). During the pandemic for the first time ever, the U.S. provided some unemployment benefits for self-employed workers or independent contractors, whether they were misclassified or not.

In New York State, an issue came up for ride-hail drivers because there had been many cases of drivers who had been deactivated by the companies pursuing coverage under state unemployment insurance. And the state Unemployment Insurance Appeals Board had decided in over 270 cases that these workers were in fact employees. The Labor Department still had not acted to generalize that finding. They're still undergoing their own assessment of what to do about that.

But the New York Taxi Workers Alliance pressed the issue because regular unemployment benefits were more favorable to workers than PUA. The PUA offered a minimal amount, whereas the regular unemployment benefit program would have provided up to 50% of a driver's base period earnings. And since many ride-hail drivers before the pandemic worked a lot of hours, they qualified for the $504 weekly maximum UI benefit, whereas under pandemic unemployment assistance, they would get $200 a week or less.

The federal court decided in looking at the record in the decision of the Unemployment Insurance Appeals Board that these workers were employees, and that the state Labor Department needed to get the requisite information from the ride-hail companies to document the earnings that could be the basis for what the weekly unemployment benefits under the regular unemployment insurance program would be. Over 50,000 drivers in New York did end up qualifying for regular unemployment insurance benefits.

So these benefits were extended several times under federal law. But the state unemployment insurance benefits are funded by payroll taxes on New York employers. New York State borrowed $10 billion to pay benefits. This has to be paid back by employer payroll taxes. So even though Uber and Lyft drivers under this federal court decision were held to be employees and qualified for regular state unemployment insurance benefits, Uber and Lyft still, to my knowledge, have not paid anything into the New York State unemployment trust fund. So they really have been freeloading on the backs of all of the law-abiding employers in New York who were paying the unemployment payroll tax.

Another thing that we've seen happen with Uber in New York is the creation of the Independent Drivers Guild, a driver advocacy group that is explicitly not a union. I think we've seen this model become popular among gig companies, of effectively preempting unionization by sponsoring a non-union group instead. Why would the prospect of a union be so threatening to a company like Uber, and where do you start when it comes to organizing independent workers?

I think a real union is a threat to Uber as it’s seen as a threat to a lot of employers, because it basically means that the workers feel empowered to act collectively on their own behalf. That obviously is different than dealing with an individual worker. Companies in many industries have resisted unionization for that reason.

And traditionally contractors have not been permitted to organize.

Right. Because they're seen as entrepreneurs and small businesses, and it would be a violation of antitrust if they combined together to bargain on their own behalf. So the companies, they've been very active at the state level in trying to get favorable legislation and permissions. They've supported legislation responding to growing complaints on the part of workers and concerns on the part of consumers about the exploitation and the unfair treatment of workers. They've made some concessions to provide some benefits to workers and so on. But they certainly don't want to empower a real union to represent workers in bargaining with the company. Their legislation sometimes is seen as sectoral bargaining where it sets up committees that are largely under the control of the company that provide the appearance of allowing workers some voice, but without ceding any real control to the workers.

Any last thoughts?

You had said earlier that you thought the classification battle was basically over regarding the platform companies. I don't see it exactly the same way. We're not sure what's going to happen in California. Prop 22 did get declared unconstitutional. There are still some issues with how AB5 is going to be implemented. In Massachusetts, the state attorney general is pursuing a pretty significant misclassification suit against Uber and Lyft, and the ballot initiative got disqualified this time around. In New York, because New York's law is not as conducive to ballot initiatives, that avenue is closed off for the platform companies. And there is pretty strong Democratic control of both houses in the New York State Legislature. So we could see legislation in the next session on the classification issue. So, you know, if you have key states like California, New York, and Massachusetts where the companies’ efforts have been thwarted, I don’t know that I would say the classification issue is over.

Another argument in that direction is that we’ve seen a greater degree of organization at least in New York, the case I know best. The Taxi Workers Alliance has been very well organized. It’s an immigrant led, non-company-controlled worker organization. Los Deliveristas Unidos, the organization representing restaurant delivery workers in New York, has achieved major gains in the past year, including a compensation standard that’s being formulated right now and should take effect on Jan. 1, 2023. Together with progress in non-app organizing—think of the Amazon labor union and the Starbucks union—there’s a degree of organization among immigrant and other exploited workers to fight back. I think that’s going to be a big factor in the next year or two in how these app-related battles play out.

1. Updated 15 August 2022 to correct the amount workers & employers pay into Social Security.