California, the Gig Economy, and Future of Work
BRG Managing Director David Lewin joins Michael Whalen to discuss the effects that the California AB5 ruling has had on the gig economy in Silicon Valley and the tech sector. Dr. Lewin provides personal insights on these complex issues and his work with independent contractors to modify this ruling prior to the election.
TRANSCRIPT
MW [00:00:00] [music] In this episode of the ThinkSet podcast, we're talking about the gig economy. It's a contractor-based work model that was accelerated by apps, such as Uber, Lyft, Postmates, Instacart, and DoorDash. These apps offered network platforms for ride sharing, deliveries, and odd jobs. Surprisingly, the assault on work flexibility made possible by such technological innovations has been most concentrated in California, the home of Silicon Valley and much of the tech sector.
Last year, California passed a new law, generally called AB5, that aims to reclassify many gig economy independent contractors as employees.
We're talking to BRG expert David Lewin, who has been at the forefront of the analysis of the implications of AB5 and of the voter initiative to modify it. In addition to heading up BRG's Human Capital practice from his base in Los Angeles, Dr. Lewin is the Neil H. Jacoby Professor Emeritus of Management, Human Resources, and Organizational Behavior at the UCLA Anderson School of Management. David, thank you for joining the ThinkSet podcast to talk about these complex issues.
DL [00:01:01] Happy to do so.
MW [00:01:02] So David, how did AB5 change the classification of contractors and employees?
DL [00:01:08] Well, AB5 doesn't change the classification of employees. It potentially changes the classification of contractors to employees. And that is, in essence, its intent.
MW [00:01:23] And was this a big change? Because I understand that the California Supreme Court had already reinterpreted traditional worker classification, putting a greater burden on companies to demonstrate that contractors were not employees.
DL [00:01:39] Well, that's correct, Michael. In 2018, in a decision that's popularly known as the Dynamex decision, the California Supreme Court came up with a new three-factor test for companies—and for that matter other organizations, such as nonprofits—to determine whether independent contractors are in fact independent.
And that three-factor test has popularly become known as the ABC test. The second factor—the B factor—is the one that has attracted the bulk of attention from companies, because it basically says that an independent contractor must be doing work that is distinctly different from the work done by the company for which it is providing services.
Most contractors are doing work that is very similar to the basic work that a company is doing. So most employers believe that they could not meet just the B part of the new ABC test. The legislature then came along the next year—that is, 2019—and enacted the proposed AB5 into law. California Governor Newsom signed the law in September 2019.
Thus, California sought a "legislative solution" to this matter so it wouldn't be just left to the courts to decide on a case-by-case basis.
MW [00:03:11] So that seems like a very big change. A contractor has to do work, which is different from the type of work that your company performs. I mean, that would seem to sweep up a lot of people who do freelance or other type of work for hire.
DL [00:03:29] Well, I think that's basically correct, because if you're having a contractor perform labor services for your company, they have to be services in connection with the goods or services your company is providing. So yes, it would be a very large change indeed. If in fact the law is actually implemented, then under the three-factor test, the ABC test, most if not all contractors would, in essence, have to become employees.
MW [00:03:59] Do we have a sense of who's been affected by this change? I mean, how, how big is this gig economy?
DL [00:04:05] Well, that's two questions, and I'm going to try my best to answer both of them. Who was affected by the change? For me, AB5 is as or more notable for what it refers to as carve-outs. We used to call these exemptions, but now the language is carve-outs, meaning occupations to which AB5 does not apply.
And these are in fact much bigger in terms of occupations and industries then the gig economy. So, for example, all physicians are excluded; all attorneys are excluded; all accountants are excluded. All of the people and occupations associated with the provision of business services are excluded, but so too are salon workers; so too are tow truck drivers; and the list goes on and on. What this tells us is that the law is basically aimed at the so-called gig economy, and therefore its impact, if in fact contractors become employees, could be very large for the gig economy but not particularly elsewhere.
MW [00:05:22] You know, when I listened to people advocate in a number of matters in court, particularly issues of regulatory authority or issues of a constitutional question, there's always this discussion of, well, what's the underlying distinguishing principle here? So if so many people are carved out of the reach of this new approach on reclassification, what is actually the distinguishing principle?
DL [00:05:46] I don't think there is an underlying principle, except if you believe that the objective of AB5 is to increase revenue flowing to the state of California. And that is clearly what the state itself believes will be the result of AB5, because various agencies in the state for the last several years have been making claims about "lost revenue." And there, the focus is on payroll tax revenue. The argument is that contractors are not covered by payroll taxes. If they were employees, the employers would have to pay payroll taxes and, to some extent, so would the employees, and that revenue would flow to the state of California, in particular to its general fund.
So if one is looking for a principle here, it's basically an argument that this change, if it happens, is going to increase the revenue—tax revenue—to the state of California.
MW [00:06:50] Was that the explicit basis and the explicit objective of AB5? And if so, did it have a target in mind?
DL [00:06:57] I think that it's definitely the objective of AB5. Yes. Did it have a target in mind is a very interesting question. And about that, I can say the following: when we began our work on this matter over a year ago, we started with a claim made by the state of California’s Department of Industrial Relations that California was losing $7 billion a year because contractors weren't covered by payroll taxes. So, we did a bunch of sleuthing, kind of a bit of detective work, to try to find out, “Where did this $7 billion come from?” And we talked with all of the agencies in California. We went through a whole variety of documents. We couldn't find anything.
Eventually, we did find what we were looking for. It turns out that in 2012—so that is eight years ago—a report was put out by a commission that was actually appointed in the Obama administration to look at independent contracting. Ironically, that commission was headed by one Joe Biden. The report of the commission says that the nation, not California alone, loses $7 billion—not annually, but rather over a ten-year period, because contractors don't pay payroll taxes. And the report covered all contractors—not just drivers for rideshare and delivery companies or contractors in the high-tech area or the gig economy, but all people providing contracting services.
You can appreciate what that might mean, because almost all of the insurance agents in this country are independent contractors. Most of the equipment installers for cable and satellite companies are independent contractors, and so on.
So when we took that apart and analyzed the potential loss to the state of California annually because of independent contracting, the loss may not be zero, but it's pretty close to zero. That $7 billion putative annual revenue loss therefore has no basis in fact or in empirical evidence.
MW [00:09:19] So the $7 billion seems to have been calculated on a number, which was a nationwide estimate, was over a decade [old], and was for a broader class of people than affected by AB5. But tell me a little bit more about what you think the real fiscal impact might be and how you came about determining that.
DL [00:09:36] In our work, our analytical work, we developed two models aimed at determining the revenue flow on an annual basis to the state of California from contracting versus employment. In other words, the current contracting arrangement versus the proposed employment arrangement.
And those two models differed in only one respect. In the first model, we assume that one-third (or 33.3%) of the currently working independent contractors would not be converted to employment, because that would be too costly for the companies. In the second model, the percentage was one-half (or 50%). Our estimates under those two models are that, relative to employment, the state would gain about $91 million annually from the independent contractor arrangement over the employment arrangement. And under the second model, it would gain about $111 million annually from the independent contractor arrangement relative to the employment arrangement. So, our analysis indicates that the state will do better revenue wise by keeping an independent contractor arrangement than it would by switching to an employment arrangement. And the main reason is that these gig economy companies simply will not employ anything close to the number of people who currently do driving for them, whether they're rideshare companies or food delivery companies. In 2018, the last year for which we had complete data, there were more than a million Californians who drove to some extent for the rideshare and food delivery companies.
If we follow the business models of those companies, which we discuss extensively in our report, the percentage of drivers who will be employed will likely be considerably smaller than one-third or even one-half. Our best estimate is it's in the 10 percent to 20 percent range. That means that about 80 percent to 90 percent of the people who currently drive for these companies would no longer do so, because the companies would find it too expensive to employ the drivers on a regular basis. So that gives you some sense of the fiscal implications of AB5. And you can compare our $91 million or $111 million annual estimate on the positive side to the state's claim of a negative $7 billion.
I mean, we're not close here at all.
MW [00:12:36] What would the broader economic impacts be, in addition to those strictly to California's general fund?
DL [00:12:42] The broader implications are several. Probably the most notable would be a decline in consumer demand for rideshare and food delivery services, because the prices of those services would inevitably have to rise under an employment model. We estimate [that] for the rideshare companies, that increase would be in the neighborhood of 26 percent, and for the food delivery companies in the neighborhood of 35 percent.
So, basic microeconomic analysis tells you that when you get a price increase, the demand declines. And with these magnitudes of price increases, the demands could decline very substantially. People who used to use Uber or DoorDash for rideshare and for food delivery, respectively, well, some of them will now look for alternatives given higher prices. Another impact is that the companies would concentrate more on providing these services in high-density urban areas, which would make more economic sense than providing the services to outlying and rural areas. So, folks in the suburbs and especially in rural areas would have much less availability of the rideshare and the food delivery services. And then, everything else equal, which it never is, there would also likely be increased waiting time to get a driver to take you somewhere or to deliver your food, because now fewer drivers would be employed. These are among the likely consequences of the implementation, the actual implementation, of AB5 on the gig economy.
MW [00:14:29] David, I understand the analysis you did was on behalf of some of the companies most affected by the legislation. These would be the app providers, the ones that are providing the rideshare and door delivery services. So what has been the reaction of supporters of the legislation to this analysis?
DL [00:14:45] Well, we were initially approached by public affairs firms that had DoorDash as a client for what was termed a coalition to focus on this issue. DoorDash was later joined by Uber, Lyft, Instacart, and Postmates. So, the Protect App-based Services and Drivers [coalition], which is our client in this matter, included those five companies. We accepted the assignment only under the condition that our analysis, our findings, and our conclusions would be done independently, as is all of our work at BRG.
Sometime after our report was delivered, the coalition began to pursue an initiative to put a proposition on the California ballot aimed at preserving the independent contractor arrangement. That initiative eventually yielded the sufficient number of signatures required to place a proposition on the California ballot. This is Proposition 22 and will be voted on by Californians next month.
And stripping away whatever language that is contained in voter advice pamphlets or sample ballots or actual ballots, a “yes” vote on Proposition 22 is basically a vote to preserve the independent contractor arrangement, and a “no” vote is a vote not to preserve the independent contractor arrangement.
MW [00:16:16] Was there a large outcry for AB5 by independent contractors who were engaged in the gig economy?
DL [00:16:23] I'll give you a couple of statistics that may be germane. In 2017, the US Department of Labor conducted a survey of a random sample of independent contractors across all major industries in the United States. Their main conclusion was that 79 percent of the contractors preferred to remain contractors. And of the remaining 21 percent, a little under 10 percent said they preferred employment, and the remainder said they didn't know or it would depend. That's a pretty substantial finding on this matter. We have other data from surveys of drivers for these gig economy companies, and very large percentages of the drivers say that they want to remain independent contractors.
And if one asks why is it that they, in the main—with some exceptions, of course—prefer not to be employees, the answers run like this: "flexibility" is the first thing. And one might say: well, "flexibility" or "autonomy," maybe that's a buzzword. However, about 40 percent of the drivers for these gig economy companies hold other full-time jobs. That is not widely known. About another 25 percent hold part-time jobs. That is not widely known. Of the remainder, call it roughly 35 percent, about half are students looking to earn money when they're not going to class or studying so they can pay off their loans and pay their tuitions [and] living expenses. The other half are retirees. And these retirees typically are receiving a pension payment from prior employment, but it's not enough to cover their expenses, especially since these days lots of retirees either have their kids at home, or who have returned home and they have to pay for them, or they are caring for their parents who are still alive under the increase aging of the US population.
So when you consider the larger picture—this business of driving for one of these app-based companies that provides flexibility, that provides autonomy, you can work when you want, you can decline the work—that's not a throwaway. That is a real live factor and is obviously very important.
MW [00:18:56] David, I thought it would have been a shock for a lot of people who prefer the flexibility of choosing their hours and deciding what kind of work they want to do to be classified as traditional employees in which they would potentially lose control over work flexibility and their ability to decline certain types of work. So, what's your sense of the reaction of workers to this reclassification.
DL [00:19:20] That is a really interesting question. I'll give you two experiences that we've had this century so far. In one, involving a package delivery company, a group of drivers—delivery drivers for that company—filed suit claiming that they should be reclassified from independent contractors to employees. When we analyzed the data involving that matter, we found that the company very highly controlled the work of those drivers: from determining what went into the trucks for them to deliver, to when they should deliver, to the routes they should use for delivery, to being supervised closely by terminal managers who are in effect their supervisors, and to having their performance evaluated carefully just like you would find in many employment situations. They had little to no autonomy or flexibility in their work.
And the decisions of the courts since we began working on this matter years ago increasingly have required that particular company to reclassify those contractors as employees.
More recently, for a different company that is in the transportation business, mainly taking people to and from airports, there was a lawsuit similar to the first matter, filed by a set of van drivers who also claimed that they should be reclassified as employees. In that matter, we analyzed terabytes of global positioning system (GPS) data on the drivers’ use of their vans. We found that the van drivers could decide what routes they wanted to use, they didn't have to use prescribed routes, they could turn off their communications devices with dispatch, and they could turn down jobs from dispatch. In other words, they didn't have to accept the driving opportunities offered to them by dispatchers. That's pretty unusual in this industry. Further, the van drivers could hire sub-drivers and could use their vans for other, private purposes.
In other words, the first example was one of high control by the company and low control by the delivery drivers. The second example was one of low control by the company and high control by the van drivers. Here, in our analysis of the independent contractor drivers for the rideshare and food delivery companies, I previously mentioned that a majority of the drivers work full-time or part-time for other companies, but I didn't mention that they are also able to use the apps from two or three of these gig economy companies to drive for them. They have the flexibility not only of deciding to accept a job or turn it down and decide what route, but they can do that for more than one company. In an employment situation, you either can't do that or you can rarely do that. So, this illustrates low company control over the basic driver's decision on whether to work or not, how often to work, and for whom to work.
MW [00:22:17] Has the pandemic had any impact on public reaction to AB5 or Proposition 22, particularly since there seems to have been an uptick in activity of delivery platforms?
DL [00:22:31] You know that there are more than thirty million employees—and that's probably a large undercount—who have lost their jobs just in the seven months of this pandemic. That's employees. If you're an independent contractor, depending upon what company and industry you work for, you might have fewer driving opportunities than you did before the pandemic, but you might have more than you had before the pandemic. Just think of package delivery companies. The impacts of COVID-19 are very substantially negative on employment when taken as a whole. In this comparative view, the independent contracting arrangement is, I think, a far preferable one for those who provide labor services.
MW [00:23:20] So often when new elements of technology or culture or politics emerge in California, there's the inevitable question as to whether this is going to be a pattern that's exported to other states. But I am aware as well that there has been some similar legislation in Massachusetts since 1990.
What's your sense on this? Is this going to be a wave, the reclassification legislation efforts across the country, or is this something [that] is going to have a narrower impact than might be suggested by what's happened so far in California?
DL [00:23:58] Right. Another way to ask the question that builds on your question is, “Can you envision the US Congress contemplating, let alone enacting, AB5-type legislation? So far, no one has said anything about that, but it's a logical extension of the question you asked, because, after all, the Congress passed the National Labor Relations Act in 1935, when prior to that time, few people thought that was possible. The Congress passed the Occupational Safety and Health Act in 1970. Few people thought that was possible. And other legislation, such as the 1964 Civil Rights Act, the 1974 Employee Retirement Income Security Act (ERISA), and the 1990 Americans with Disabilities Act, illustrates the same point. So, things are often difficult to predict on a national scale let alone a state-by-state basis.
AB5-type legislation is also being considered in New Jersey, Illinois, and some other relatively large states. It's not presently being considered, to the best of my knowledge, in a majority of the states. A key implication in this regard is that when a single state such as California enacts legislation, such as AB5, it can affect companies’ policies and practices domestically and even internationally if they operate internationally.
Now, the Ubers of the world, the Lyfts of the world, the Instacarts of the world, they've been in existence for a short time. They are not yet profitable. But their business models are designed to generate profitability. What company's business model wouldn't be? Yet after a decade or more, profitability is still an elusive goal for these companies.
Hence, if a company’s costs rise in one state or area of the country, the tendency would be to move away from that state or area and move to others. We saw this in the automobile industry where during the 1970s, 1980s, and 1990s, production moved from the Northeast, the West Coast, including California, and from the Midwest to the South. So this is a fundamentally important consideration when considering Proposition 22 and, more fundamentally, AB5.
MW [00:26:27] And how have companies outside of California reacted so far?
DL [00:26:28] Well, they're not rushing to set up business in California. I think one could say that. But, you know, there are various regulations, a combination of state and federal, that affect companies’ decisions about where to situate their businesses. One development within California, which you probably know well, is that more and more of another set of gig economy company people, and these are mostly employees, working for the Googles and the Intels of the world and many other companies are, in light of the COVID-19 pandemic, moving out of high-cost, high-rent areas, like San Francisco and Los Angeles, to the suburbs and to rural areas.
And you asked me before about the impacts of COVID. One of the big ones on employees and companies is that both have learned that there is a lot of work that doesn't have to be done in a central location or an office building. Thus, if you could work from home, do you have to be in a high-rent, high-cost area, or can you move somewhere else where the costs are more moderate? The answer to this question seems to be “yes,” as reflected in what is going on right now in California. And if you extend these observations nationally, you can expect to see similar kinds of behavior.
MW [00:27:41] From a standpoint of a prospective client of BRG's, what would you think the course of action is right now in thinking about the implication of AB5 or similar legislation that might be contemplated elsewhere?
DL [00:27:58] Well, I think I would go back to a question you asked earlier about control.
If you are a company that provides workers, whether it's independent contractors or employees, some autonomy or flexibility in their work, including deciding where they work, then you can't be taken to task for being a high-control company in which only the traditional employment model supposedly fits. The independent contractor model could potentially fit such companies much better. And the large growth nationally during the century—this is twenty years now—has been in independent contracting. I mentioned before that insurance agents, equipment installers, hair and nail salon specialists, and many others that provide labor services are independent contractors.
Most of these businesses would not be viable if they didn't have the independent contractor arrangement. If they are viable, they can produce more goods and services. When that happens, prices come down, and therefore more people can demand those services, including people who are relatively low on the income scale.
So, there are aspects of this matter that have implications not just with respect to employment versus independent contracting, but with respect to equity for businesses and with respect to income inequality.
MW [00:29:20] That's fascinating. And given the fact that we're all, to a certain extent, working more independently and working remotely, it'll be interesting to see if those who otherwise would have preferred the parameters of a traditional employer-employee relationship might be willing to consider other work models.
DL [00:29:37] Let me add one final thought about this. The traditional employment model, which is so strongly reflected in most of our existing legislation, nationally and statewide, was based on the practices that developed over a forty-year period from roughly 1940 to 1980. That's a fairly short period in the history of the country.
That was the permanent employment period. You went to work for a company—usually a man, you had a job, whether an operative or an assembler or a supervisor or an executive. You had a defined benefit pension plan. You could work for that company for your career unless you screwed up badly. And when you were done with that career, you got a payment, a defined benefit pension payment, every month.
But, beginning in about 1980, when we had the onset of the triple revolution of greatly expanded global economic competition, deregulation, and technological change, most of the industries in the country, which had been oligopolies or monopolies in technical terms, became competitive. Being competitive sounds like a good thing in many respects, but it also increases risk—for companies and for individuals. To illustrate, consider IBM and Kodak, two examples of the best-known permanent employment-type companies. Faced with increasing competition, both companies began early in the 1980s to reduce their work forces, including through layoffs. These actions mushroomed during the rest of that decade and then continued into the 1990s. Relatedly, these companies and many others replaced their long-standing defined benefit pension plans with 401(k) plans. All of this represented a secular shift in risk-bearing from employer to employee or from a company to the people working for the company.
There is not an enemy—a particular company or executive—that one can point to in this historical development. Rather, it reflects the effects of competition on the traditional employment model. While millions of people in the United States (and elsewhere) work as employees and may be fortunate enough to have a 401(k) plan (not a defined benefit plan), millions of people are now working as independent contractors.
In summary, we tend to be anchored to past practices because they provide stability and a benchmark; these are good things. But the changes that we're undergoing are turning the traditional employment arrangement on its head—if not fully, then about halfway, at least. This is in part what's being played out in Proposition 22 and the matter we've been discussing today regarding independent contracting versus employment for gig economy companies.
MW [00:32:27] Well, David, thank you very much for joining us here on BRG's ThinkSet podcast. A fascinating conversation. It'll be interesting to see what voters decide in November on this matter. And thank you for sharing your analysis.
DL [00:32:39] Thank you very much. And just to be clear, I have not placed a bet on the outcome of Proposition 22.
MW [00:32:45] Very good. Thanks very much.
DL [00:32:47] Thank you.
MW [00:32:50] This ThinkSet podcast is brought to you by BRG. You can subscribe to the podcast and access other content from ThinkSet magazine by going to thinksetmag.com. Don't forget to rate and review this show on iTunes as well. I'm Michael Whalen. Thanks for listening.
The views and opinions expressed in this podcast are those of the participants and do not necessarily reflect the opinions, position, or policy of Berkeley Research Group or its other employees and affiliates.
This transcript has been edited by the interviewee and may not reflect the audio-recording exactly.