Four Recommendations for Analyzing the Department of Labor’s Proposed Rule on Employees vs. Independent Contractors

The US Department of Labor (DOL) has proposed a rule change that adopts an “economic reality” test for determining whether a worker is an employee or contractor for the purposes of the Fair Labor Standards Act (FLSA). I am grateful for the opportunity to submit a comment to the DOL in response to its proposed rule. The Mercatus Center at George Mason University is dedicated to advancing knowledge about the effects of regulation and policy on society. Accordingly, my comment is designed to aid the DOL as it considers the impact of the rule.

Throughout this comment, I refer to individuals engaged in independent contracting as “independent workers.” This term encompasses gig or platform workers, freelancers, contractors, and workers in other types of external or alternative labor arrangements.

My analysis suggests that the DOL is correct in proposing an alternative to codifying California’s “ABC” test in federal regulation. The ABC test was formulated by the California Supreme Court in Dynamex Operations W., Inc. v. Superior Court to distinguish between independent workers and employees, and it was codified in California state law through Assembly Bill (AB) 5 in 2019. However, I make the following four recommendations for improving the DOL’s impact analysis of codifying the ABC test nationwide:

1. The analysis should cover ways in which greater amounts of independent work are beneficial to women. Nationwide codification of the ABC test, which already reduces the number of independent work opportunities in California, could have an adverse effect on women, who generally benefit from the flexibility of independent work.
2. The analysis should also cover how nationwide codification of the ABC test could harm small, innovation-driven technology startups that rely on the flexibility of independent workers during the early stages of their development.
3. The analysis should consider new evidence on how nationwide codification of the ABC test could lead to job terminations for independent contractors, with job losses as large as 75 percent. Moreover, the analysis should cover recent studies showing that independent work is potentially an “intermediate job” for those who are coming out of unemployment and for those who may later become entrepreneurs with employees of their own.
4. The analysis should cover how the lack of health insurance and other benefits for independent contractors could be mitigated by changes to the “employee benefits factor” test that is used by other government agencies. This is out of the purview of the DOL, but I provide recommendations for how to address the problem of benefits for independent workers without the damage associated with reclassifying them as employees through a rule such as the ABC test.

Overall, the DOL is correct in proposing an alternative to codifying the ABC test. The DOL should investigate further the potential negative effects of the ABC test at the national level, especially as they relate to women who engage in independent work and to small technology startups that rely heavily on independent contractors. Lastly, although the DOL is constrained in adopting a common law control test, I suggest that lawmakers amend the FLSA to allow for codification thereof. Codification of a common law control test would aid the DOL in achieving its goals of reducing uncertainty and creating greater harmony across federal statutes and agencies on what makes a worker an employee or contractor.

Impact on Women

In its impact analysis, the DOL includes a section on regulatory alternatives, where it addresses the implications of codifying the ABC test nationwide. I strongly recommend that the DOL include information in its analysis about how the ABC test could harm women who seek independent work and how this harm would be amplified at the national level.

There is substantial research on women’s participation in alternative and external labor arrangements. Such arrangements allow women greater flexibility in structuring their days, which is crucial for women who are the primary caregivers in their households. Policies such as AB 5 in California, which reclassifies and restructures independent work into traditional employment, are therefore harmful for women who are unable to satisfy the requirements of traditional employment opportunities.

Survey Research

A 2017 study by HyperWallet reports the results of a survey of 2,000 women who use platform economy companies for work. The study finds that 96 percent of women indicate that the primary benefit of engaging in platform-economy work is the flexible working hours. Moreover, the study finds that 70 percent of these platform working women are the primary caregivers in their homes. A quarter of these women recently left their full-time employment for platform-based work, and 60 percent of them indicated doing so because they wanted flexibility, needed more time to care for a child, parent, or other relative, or both.

Reports by consulting group MBO Partners published in 2016, 2017, 2018, and 2019 similarly find that women prefer platform or gig work, freelancing, or other forms of independent work because those work arrangements allow greater flexibility. For example, in its 2018 report, MBO Partners finds that the primary motivations for women to engage in independent work are flexibility (76 percent) and the ability to control their schedules (71 percent). Contrast this with men, who said that the primary reason for engaging in independent or freelance work was that they enjoyed being their own boss (67 percent) and did not like answering to a boss (64 percent). The 2017 report by MBO Partners finds similar results: “Women were significantly more likely to note that flexibility was a more important motivator for independent work than men (74 percent vs. 59 percent).”

Furthermore, a 2016 study by James Manyika and coauthors reports the results of a survey of 8,000 independent workers, finding that 42 percent of US women and 48 percent of European women who participate in independent work are also caregivers. In fact, referring to the 17 percent of the total sample in their survey who reported providing care to an elderly dependent, the study authors state that “these caregivers participated in independent work at a significantly higher rate . . . than non-caregivers.” Moreover, the study indicates that caregivers engage in independent work for supplemental income (67 percent, compared with 54 percent for noncaregivers). Among their conclusions, the authors say that independent work “provides a way for caregivers [who are disproportionally women] to generate income while fitting their hours around the needs of their families. This type of flexibility can ease the burden on financially stressed households facing logistical challenges.”

In another survey, researchers find that about 75 percent of self-identified homemakers, or stay-at-home mothers in the United States, indicated they would be likely to return to work if they were to have flexible options.

This finding is consistent with previous studies, published even before the rise of the modern platform economy. For example, in 1995 and 1997, Janet Marler and George Milkovich surveyed women participating in alternative labor arrangements and found that women are particularly attracted to these opportunities because the flexible work schedules allow them to reconcile work outside the home with family commitments.

Measures of Women in Alternative Labor Arrangements

In addition to analyzing survey responses, several studies analyze more objective metrics of women’s participation in the platform economy. This approach is often overlooked because transportation platforms such as Uber and Lyft take center stage in discussions about the prevalence of women in this part of the economy, and those companies tend to have greater participation by men than women. A study by Diana Farrell, Fiona Greig, and Amar Hamoudi shows that women constitute a greater share of platform-economy workers if one omits transportation platforms such as ridesharing and delivery from the analysis. This is corroborated by the HyperWallet study: most women indicated participating in professional freelancing, direct selling, or service platforms (e.g., Upwork, TaskRabbit, Care.com); but only 22 percent of women indicated participating in ridesharing (e.g., Uber and Lyft), and only 8 percent indicated participating in food delivery (e.g., Postmates, Grubhub).

Using tax data, Brett Collins and coauthors find that, while independent work is more common among men, the participation in independent contracting since 2000 has grown significantly more among women than among men. Economists Lawrence Katz and Alan Krueger also find that between 1995 and 2015, the growth in alternative work arrangements was driven primarily by women. Katz and Krueger also show that women are more likely than men to be employed in alternative work arrangements. These more recent findings are consistent with older studies that also find greater prevalence of women in alternative labor arrangements. Economists Linda Edwards and Elizabeth Field-Hendry analyze differences among working women in home-based versus on-site work. As with other alternative work arrangements, they find that approximately 60 percent of home-based workers are female. Allan King finds that industries that allow for greater variability in the distribution of work hours allow women to better coordinate work activities with home activities, thereby increasing female labor force participation.

In two separate journal articles, Guy Standing reports how the technological revolution in the late 20th century led to a wide range of working arrangements that has allowed for greater labor market flexibility and thus greater female labor force participation. Véronique Genre, Ramón Gómez Salvador, and Ana Lamo find that a decline in the strictness of labor market regulations in some European countries significantly increases women’s labor force participation. They also find that an increase in the use of flexible forms of work, such as part-time employment, helps explain the female labor force participation developments between the 1980s and the 1990s in Europe.

Altogether, this research indicates that independent work may be important for women who require more flexible work arrangements. Proponents of policies such as California’s AB 5 overlook the consequences of such policies for the types of independent jobs that attract women. For example, gig platforms for transportation and delivery apps may be more male dominated and look different than gig or contractor jobs through Etsy and Care.com, which tend to be female dominated. Yet the proponents of AB 5 do not acknowledge these differences among platforms and how the impact of such policies could vary depending on the types of individuals and the platforms on which those individuals work. Thus, to the extent that specific platform companies such as Etsy and Care.com provide flexibility of work for those who need it and extend work opportunities to women who would otherwise be unable to take on traditional employment, challenges to the legal classification of independent contractors could disproportionately hinder women’s participation on those platforms. In fact, when debating the legislation, California did not compare the potential benefits and the potential harms of AB 5 specifically to women. Therefore, I recommend that the DOL investigate this topic in its impact analysis of different regulatory alternatives and include ways in which codifying the ABC test nationwide may negatively affect women.

Impact on Small, Innovation-Driven Technology Startups

The DOL includes in its report the potential costs of curtailing independent contracting for different groups, such as small businesses. I recommend that the DOL include in its impact analysis the potential of such policies to harm small, innovation-driven technology startups.

As a co-principal investigator of the NYU School of Law study “Startup Innovation: The Role of Regulation in Entrepreneurship,” I led an effort to interview and survey technology startup CEOs and other members of the startup community across the nation. The findings presented in this section come from both original interviews and an original online survey conducted in the United States. We conducted 88 interviews in the United States between May 2017 and December 2017. Forty-five were with technology startup founders and executives, and the remainder were with venture capital investors, accelerators or incubator directors, startup lawyers, and other members of the startup ecosystem. Boston, Los Angeles, New York City, San Diego, and San Francisco and its neighboring cites (Silicon Valley) were the primary focus of our study.

To complement the interviews, we conducted an online survey of technology startup executives in May 2019, restricting our sample to executives of technology startups that were founded no earlier than 2012 and had fewer than 200 full-time employees. A total of 406 executives responded. The startups represented by the executives in our sample are headquartered in 42 states and belong to a variety of industries such as MedTech, FinTech, information services, media, retail and e-commerce, software, and transportation.

From the interviews, we discovered that 71 percent of startups relied on independent contractors and thought it was necessary to use contract labor during their early stages. The interviewees explicitly discussed the reason that early-stage small startups prefer contractor labor over employee labor: during unpredictable times, when startups are trying to find their market and build their product, they need flexible labor and need to be able to hire and fire easily.

The online survey supports the fieldwork findings. We asked executives to enter the number of full-time and part-time workers, contractors, and remote workers employed at their startups. Approximately 80 percent of the executives in our sample indicated that they use contract labor. We asked some follow-up questions to better understand the use of independent contractors. For example, we asked, “How important is the use of 1099 contractors for your specific business model?” The executives in our sample responded as follows:

• 57 percent of executives of startups that use contractor labor indicated that the use of contractor labor is an indispensable or essential part of their business model.
• 39 percent indicated that contract labor is not essential but highly valuable.
• 3.5 percent indicated that the use of contract labor is not essential and is unimportant.

We also asked respondents to “rank up to 3 primary reasons why the startup uses 1099 contractors.” Executives gave the following responses:

• #1 Reason: They needed individuals for one-off projects or specialized talent that they could not hire full time.
• #2 Reason: They needed flexibility, given the risk associated with early-stage development.
• #3 Reason: They needed flexibility, given fluctuating demand for their product or service.

Indeed, we got the impression from our interviews that the primary concern for startups in terms of labor regulation or policy is mostly with regulation of independent contractors. Therefore, any impact analysis should include technology startups because they are young and fast-growing and thus of interest for their potential impact on job creation. Ryan Decker and coauthors find that high-growth businesses (which are disproportionally young firms) account for almost 50 percent of gross job creation. Their research describes the unique role of young and fast-growing startups: most new businesses tend to die within 10 years, and most surviving young businesses do not grow but remain small (these may be what most people imagine as the typical “Main Street” small business), but a small portion of young businesses exhibits very high growth and contributes substantially to job creation. Other studies also indicate that almost all net job creation in the United States has occurred in firms younger than five years old, and of these firms, a small percentage of high-growth firms is responsible for most of the jobs. Technology startups are also important because they often are highly innovative.

The importance of independent contractors for the early years of young, small technology startups should be considered in analysis of policies such as AB 5, which make it harder to classify workers as independent contractors.

Job Terminations and Other Recent Research on Independent Workers

The DOL’s analysis in section G, “Regulatory Alternatives,” provides some news stories on how AB 5 has led to the termination of jobs. I recommend that the DOL also include in its analysis new evidence on job terminations and the forecast by the California Legislative Analyst’s Office that some number of contractors will not be rehired as employees: “We cannot predict the exact number of contractors who will become employees due to AB 5. Although we cannot predict the exact figure, it is probably much smaller than the roughly 1 million contractors that AB 5 applies to. . . .” One of the primary reasons the office expects this outcome is that “businesses will comply with the law in different ways. Some businesses may hire their contractors as employees, while others may hire some, but not all, of their contractors. Other businesses may decide to stop working with their California-based contractors.”

The predictions of the California Legislative Analyst’s Office are consistent with other cost measures and predictions discussed in this comment.

The costs for companies of reclassifying their contractors as employees are substantial. One study finds that having an employee costs over 20 percent more than having a contractor. Another report indicates that if Uber and Lyft were to reclassify all of their drivers as employees in compliance with AB 5, they would be faced with an additional annual cost of $3,625 per driver. Given that there are approximately 140,000 Uber drivers and 80,000 Lyft drivers in California alone, the report estimates that complying with AB 5 would lead to more than a $500 million operating loss for Uber and a $290 million operating loss for Lyft.

In fact, Uber’s own estimate suggests that complying with AB 5 would lead to a 76 percent decrease in the number of drivers finding work on the Uber platform. This means that each quarter, the number of drivers active each quarter would fall from 209,000 to 51,000. In another report, Dara Khosrowshahi, the CEO of Uber, estimates the impact of a national rule requiring all US drivers to be employees: “If Uber instead employed drivers, we would have only 260,000 available full-time roles—and therefore 926,000 drivers would no longer be able to work on Uber going forward. In other words, three-fourths of those currently driving with Uber would be denied the ability to work.”

Uber explicitly acknowledged this possibility in its IPO report filed to the Securities and Exchange Commission: “If, as a result of legislation or judicial decisions, we are required to classify Drivers as employees (or as workers or quasi-employees where those statuses exist), we would incur significant additional expenses. . . . Further, any such reclassification would require us to fundamentally change our business model, and consequently have an adverse effect on our business and financial conditions” (italics added).

Lyft also emphasized this risk in its IPO report: “If the contractor classification of drivers that use our platform is challenged, there may be adverse business, financial, tax, legal and other consequences.” The company further states that “any legal proceeding that classifies a driver on a ridesharing platform as an employee may require us to significantly alter our existing business model and operations and impact our ability to add qualified drivers to our platform and grow our business, which could have an adverse effect on our business, financial condition and results of operations.”

Therefore, the DOL impact analysis should include more evidence about the job terminations that would result if the ABC test were to be codified nationwide. This possibility is especially important to consider during times such as the COVID-19 pandemic, when jobs losses are already significant.

Research also suggests that individuals turn to independent work temporarily after the loss of a job until they can find full-time employment. In a 2017 paper published in the American Economic Review, Katz and Krueger find that workers who “suffered a spell of unemployment are 7 to 17 percentage points more likely than observationally similar workers to be employed in an alternative work arrangement when surveyed 1 to 2.5 years later.”

Moreover, a high-profile study published in early 2020 reports a similar finding: “Individuals are significantly more likely to enter solo self-employment from unemployment than from traditional employment.” That study reports the responses of 45,000 individuals who were surveyed across the United States, the United Kingdom, and Italy. The authors also find some indication that independent work may be an “initial stage of a future entrepreneurial activity with employees” and that solo self-employed workers “are more likely than the unemployed or the employed to become self-employed with employees.”

This research should draw attention to independent work as a potential intermediate job for those who are coming out of unemployment and for those who may later become entrepreneurs with employees of their own. Codifying the ABC test nationwide could eliminate this option and thereby worsen individuals’ economic standing, especially during the COVID-19 pandemic.

The Problem of “Employee Benefits” in Employee vs. Contractor Tests

In section D, “Potential Transfers,” the DOL acknowledges that independent contractors, on average, may be less likely to have health insurance and may have less access to other benefits, such as retirement contributions and paid time off.

One potential obstacle to companies providing benefits to their contractors is that the law discourages them from doing so. I recommend that the DOL inquire about this barrier. If this legal barrier is removed, the negative impact of the loss of health insurance and other benefits for independent contractors may be mitigated. In this section, I provide my analysis, also available as part of a recently published report by the Center for Growth and Opportunity, which discusses in detail why companies face a legal barrier to setting up benefits for their contractors.

One fundamental problem with platform companies providing health insurance and other benefits to the contractors on their platforms is that this action gives further credence to the legal argument that the contractors “look like” employees and thus should be classified as such.

The IRS and some other government agencies use the common-law test as their basis but include their own additional factors for determining whether a worker is an employee or a contractor. One factor used by the IRS and others, the “type of relationship” factor, looks in depth at the employer-employee or employer-contractor relationship. On its website, the IRS discusses this factor, pointing to the presence of employee benefits as one indicator of an employer-employee relationship rather than an employer-contractor one.

The IRS says specifically, “Businesses providing employee-type benefits, such as insurance, a pension plan, vacation pay or sick pay have employees. Businesses generally do not grant these benefits to independent contractors.”

In other words, if an employer were providing health insurance and other benefits, government agencies would interpret the relationship between a platform company and its contractors as an employer-employee relationship. In highlighting the different types of factor tests for employees and contractors in relation to gig economy platforms, a 2015 law review article explains that “it is advisable that bona fide independent contractors not receive overtime, vacation, or holiday pay or other additional pay beyond payment for work completed or invoiced.”

There is some evidence that companies are acting in a manner to minimize risk to their business model. As one example, in 2016, Handy and its consulting company, Tusk Ventures, helped to draft legislation in New York that would allow gig platforms to provide workers with portable benefits. However, Handy and Tusk Ventures included one major condition: workers could only access these benefits if they were legally classified as independent contractors. This part of the plan became controversial, as the Taxi Worker Alliance and some policymakers viewed it as a major concession to gig economy companies. Handy and Tusk Ventures did not move forward with a draft that would remove the contractor protection.

The risk of gig economy companies providing health and other benefits has been recognized by scholars as they prescribe their various solutions. For example, former Deputy Secretary of Labor Seth Harris and economist Alan Krueger argue that platform companies should be permitted to pool gig workers for the purposes of purchasing and providing insurance and other benefits “without the risk that their relationship will be transformed into an employment relationship.”

Thus, one obstacle to gig economy platforms providing health insurance and other benefits is the danger that such a provision would trigger the reclassification of contractors as employees, thereby threatening the business model of many of these gig economy platforms.

In fact, in August 2020, Uber released a proposal that discusses how the company could provide benefits and protections to drivers and calls on governments to make legal changes to allow it to do this. The CEO of Uber wrote an editorial for the New York Times explicitly supporting laws that would allow for this change and emphasized the problems of the current system as follows: “Each time a company provides additional benefits to independent workers, the less independent they become. That creates more uncertainty and risk for the company, which is a main reason why we need new laws and can’t act entirely on our own.”

Therefore, there is some indication that companies are committed to providing benefits, but they need the law to change in order for them to do so.

Conclusion

The DOL is correct overall in proposing an alternative to California’s ABC test. I recommend that the DOL acknowledge and investigate further the potential negative impacts that the ABC test would have at the national level, especially as it relates to women who engage in independent work and to small technology startups that rely heavily on independent contractors. Lastly, while the DOL is constrained in codifying a common law control test, I suggest that lawmakers amend the FLSA to allow for codification thereof. Codification of a common law control test would aid the DOL in achieving its goals of reducing uncertainty and creating greater harmony across federal statutes and agencies on what makes a worker an employee or contractor