How Well Are Independent Workers Prepared for Retirement?

Overview

Online platform (“gig”) workers have received widespread attention from researchers, policymakers, and the media in recent years. But the independent workforce has long included self-employed workers, independent contractors, on-call and temporary help, agency workers, and others.

Some labor analysts express concern that growing reliance on independent work may have long-term economic consequences for workers. For example, independent workers are much less likely than traditional ones to have a workplace retirement plan. As a result, they don’t benefit from employer savings matches or from the behavioral “nudges” toward participation that employers have learned to incorporate into workplace retirement plans, such as automatic enrollment with default contribution rates.

Many independent workers may enter retirement without adequate savings or Social Security benefits. They may face impoverished retirements or may be unable to retire. They also may be more likely to turn to means-tested government programs such as Medicaid and Supplemental Security Income, straining state and federal budgets at a time of increasing pressure on government finances. Many independent workers say they enjoy flexible work hours, like being their own bosses, and report high satisfaction with their work arrangements. However, others may have joined the independent workforce reluctantly, following the loss of a traditional job. 

Labor experts agree that independent work is characterized by a lack of job security, but consensus often ends there. Varying definitions and different datasets give rise to wide differences in measuring the number of independent workers in America’s workforce. By one measure, 3.8 percent of working Americans perform independent work; by another definition, as many as 40.4 percent do.

With the goal of helping researchers and policymakers, this brief provides a short review of the literature that attempts to define independent workers and measure their retirement plan coverage. The discussion also highlights the limited information and data gaps analysts and policymakers face when addressing the demand for such savings opportunities. In particular, existing research does not adequately identify whether independent workers, or specific subsets of independent workers, would like greater access to retirement plan coverage. There are no data on what percentage of independent workers have coverage through another job. Studies and surveys to date also do not help identify which modes of retirement plan coverage (e.g., online platforms, apps such as those that facilitate transfers from bank accounts to retirement savings accounts, or other types of access) would be most helpful to different groups of contingent workers.

Who are contingent, alternative, and gig workers?

Several alternative definitions address the temporary nature of independent work, the nontraditional aspects of the employer-employee arrangement, or both. While the definitions have their own strengths, they may lead to different understandings about the scope and needs of independent workers.

Most researchers—from the U.S. Bureau of Labor Statistics (BLS) to not-for-profit think tanks such as The Brookings Institution’s Hamilton Project—agree that independent workforce definitions should include workers such as agency temps, direct-hire temps, on-call workers, and online platform workers such as Uber and Lyft rideshare drivers, as these workers lack job security or have variable or unpredictable work schedules. More debate centers on the inclusion of other types of workers. For example, some broader definitions include independent contractors (who are contracted to provide work or services to somebody else as a non-employee), self-employed workers, or long-term part-time workers. While many of these may work in temporary or variable situations, others may have long-term employment stability.

Several key characteristics may help distinguish among various work arrangements.¹ First, is the person paid a steady wage or salary,² or is he or she paid by the job or task? Second, is the work expected to continue? Continuity of employment is a criterion used by the BLS, a government agency that collects survey data on the American workforce, in its definition of “contingent workers.” Third, does the worker have a predictable work schedule? Fourth, are work earnings predictable? Finally, some workers who are paid a wage or salary may be contingent if on-the-job supervision is provided by a different firm than the one that pays the salary. This last characteristic describes temp agency and contract firm workers and has been associated with business outsourcing (known as fissuring) in the labor market.

Contingent employment

The BLS defines “contingent” workers as “workers [who] do not expect their jobs to last or who reported that their jobs are temporary,” putting the focus on the short-term nature of work arrangements. The BLS’ Contingent Worker Supplement (CWS) provides three estimates of contingent work, with the higher estimates adding more types of workers and relaxing the duration of employment criteria (see Table 1). Depending on the definition, the BLS reports that 1.3 to 3.8 percent of working Americans did contingent work as their main job in 2017.³ This represents a slight decline from 4.1 percent in 2005, when the CWS was last fielded.

Researchers have criticized the BLS’ definitions, which are several decades old, along several dimensions. The survey focuses on a worker’s main job which, by definition, excludes contingent work that supplements primary income, an important shortcoming. Evidence suggests that independent work is a “side hustle” or secondary source of income for many workers. The CWS also does not probe deeply into the nature of work arrangements and may not cue a worker to report work he or she doesn’t consider a traditional job, such as selling crafts on Etsy or eBay, or walking the neighbors’ dogs.

The U.S. Government Accountability Office (GAO) developed two definitions of contingent work based on responses to the General Social Survey (GSS). The GAO defines “core contingent” workers to include agency temps, on-call workers, and contract company workers; these short-term workers constituted 7.9 percent of all employed workers in 2010. The GAO’s “total contingent” workers include core contingent workers plus independent contractors, self-employed workers, and standard part-time workers. Total contingent workers represented 40.4 percent of workers in 2010. The GAO notes that these high figures may be due, in part, to the 2007-09 recession, when many workers lost their jobs and took temporary, contingent jobs to tide over their families.⁴

Alternative employment 

Another frequently used definition, “alternative employment,” shifts the focus from the temporary nature of work to other dimensions of employment instability and the employee-employer relationship, such as unpredictable shifts or hours, lack of employer-provided health and retirement benefits, or lack of protection under key labor laws. Workers in alternative employment may include those in a variety of nontraditional work arrangements—whether or not the job is temporary—such as independent contractors, on-call workers, temporary help agency workers, and workers provided by contract firms.

Research assessing alternative employment tends to find significantly larger numbers of workers than research assessing contingent employment. The BLS estimated from the 2017 CWS that the share of working Americans in alternative work arrangements as their main job was 10.1 percent.

Self-employment

A study by the U.S. Treasury’s Office of Tax Analysis⁵ focused on self-employed individuals. This study examined Schedule C (sole proprietorships) and Schedule SE (self-employment earnings) filings. In 2014, 17 million individuals—about 12 percent of all tax filers with earnings—were self-employed, a 32 percent increase from 2001. Almost all the increase in self-employment was due to increases in sole proprietors who had little or no business-related deductions, and who therefore appeared to almost exclusively provide labor services (including contractors and workers whose employers misclassified them as nonemployees). The share of filers that were small-business owners was essentially unchanged.

Federal tax returns generally produce more accurate data than the self-reported information in the CWS and other surveys. However, the amounts reported to the IRS are still subject to key ambiguities.⁶ For example, reporting thresholds limit the activity reported on tax returns (see Table 2). Not all self-employed workers file Schedule C with their tax return, and not all employers provide W2s or 1099s to all their workers. Moreover, apart from Schedule C, the dollar figures reported on most of the tax forms are gross; they account for income but not expenses, so it’s impossible to see net self-employment earnings. In addition, tax administrative data can suffer from underreporting and nonreporting of income.

About 65 percent of individuals who have evidence of self-employment income from their Schedule SEs or information returns—forms that businesses use to notify government agencies about taxable payments—do not report any self-employed income to the Bureau of Labor Statistics’ Current Population Survey.⁷ One possible reason for the discrepancy is that the BLS measure addresses only a worker’s primary or main job, and secondary jobs aren’t counted.

In 2015, the Pew Research Center used the U.S. Census Bureau’s 2012 Survey of Business Owners to measure the self-employed, defined as those who are sole proprietors of their own business or who own it in partnership with others. Pew found that the self-employed under this definition numbered 14.6 million, or about 10 percent of the nation’s workers.⁸

Online platform or ‘gig’ work

Available data suggest that workers providing labor or goods through the online platform economy are a small component of all independent workers. The BLS estimated from the 2017 CWS that in May 2017 about 1.6 million workers obtained short-term jobs or tasks through websites or mobile apps, such as driving for Uber, selling crafts on Etsy, or developing websites for clients they found on Upwork. Such workers accounted for 1.0 percent of total employment.⁹ The BLS asks workers only about their main jobs, however; evidence suggests that gig work is a supplemental or secondary source of income for many.

U.S. Treasury researchers identified “gig economy” workers by specific phrases used on their Schedule C and 1099 information returns (for example, “ride sharing,” or a 1099 provided by one of about 25 prominent online or platform-based services) and measured this group at about 0.7 percent of all workers.¹⁰

Surveys of independent workers and corporations’ internal studies

Other surveys and internal studies by corporations have applied a range of labels to different types of independent workers. Although many surveys gathered additional information, such as respondents’ demographics, the summaries below focus on the surveys’ definitions and measurements.

The Federal Reserve Board fielded a one-time survey in 2015, the Enterprising and Informal Work Activities (EIWA) Survey.¹¹ The Fed found that 36 percent of survey respondents engaged in enterprising and informal work (E&I) activities—defined as informal online and offline paid work—in the six months prior to the survey. Further questions probed the motivations and attitudes of E&I workers, revealing that respondents who are traditionally assumed to be nonworking (such as retirees, homemakers, and students) participated in online and offline informal work activities to various degrees.

The Freelancers Union and Upwork, an online platform that pairs clients with a wide variety of freelancers, including graphic designers, writers, and computer programmers, have partnered for a few years to sponsor an online survey of U.S. adults who had done paid freelance work in the previous 12 months. This research defines freelancers as “individuals who have engaged in supplemental, temporary, project- or contract-based work within the past 12 months.” The definition includes people with multiple sources of income from traditional employers and freelancing, independent contractors, moonlighters with a traditional primary job, freelance business owners with employees, and temporary workers (for example, a data-entry worker employed by a staffing agency on a three-month assignment). The survey found that about 36 percent of the workforce (57.3 million workers) freelanced in 2017. Only 10 percent of freelancers described themselves as being part of the “gig economy,” with half preferring “the freelance economy” to describe their work arrangements. Rapid growth in freelancing—three times faster than the U.S. workforce overall since 2014—led the authors to conclude that, if current growth trends continue, a majority of the workforce will be freelancers by 2027, less than a decade away.¹²

McKinsey Global Institute defines “independent work” by three characteristics: a high level of control and autonomy; payment by task, assignment, or sales; and a short-term job duration of less than a year. This definition excludes people who keep regular work schedules with little autonomy and have a continuous relationship with their employer, even if they are legally classified as contractors, such as the “permatemps” who are on long-term or continuously renewed short-term contracts. Under McKinsey’s definition, about 27 percent of the U.S. working age population (as many as 68 million workers) is engaged in independent work; about 46 percent of these independent workers derive most of their income from independent work.¹³

An NPR/Marist poll in December 2017 found that 1 in 5 working Americans identify themselves by the label of “contract worker.”¹⁴ Survey respondents were asked, “Thinking about your work arrangement, are you a contract worker, that is, someone who has been hired to work on a specific project or for a fixed period of time?” About 51 percent of contract workers had no on-the-job benefits.

Access to workplace retirement plans

Workers who aren’t part of a traditional employer-employee relationship may lack benefits such as employer-provided health insurance, overtime pay, collective bargaining rights, or unemployment insurance. Independent workers also generally don’t have access to retirement plans through their employers.

BLS data from 2017 show that 23.4 percent of contingent workers (using the BLS’ broadest definition of contingent work) were eligible for employer-sponsored pension or retirement plans in 2017.¹⁵ This is about half the rate for workers in more traditional and permanent employment arrangements, among whom about 47.6 percent were eligible for employer-sponsored retirement plans. Even when they are offered a workplace retirement plan, contingent workers were much less likely to participate than other workers. The same BLS data show that about 18.4 percent of contingent workers—about 78.6 percent of contingent workers who were eligible for their employers’ plans—were included in the plans. By comparison, 43.4 percent of “noncontingent” workers—or 91.2 percent of eligible noncontingent workers—participated in their employers’ plans.

The BLS data are broadly consistent with other research. The GAO found that contingent workers are two-thirds less likely than traditional workers to have a retirement plan at work.¹⁶ A December 2017 NPR/Marist poll found that 32 percent of contract workers reported having a retirement plan at their current job, while 14 percent reported having a pension.¹⁷

Federal tax data analyzed by the U.S. Treasury’s Office of Tax Analysis show that about 41.9 percent of wage or salaried (only) workers contributed to an employer plan or IRA in 2012. By comparison, 7.8 percent of primarily self-employed sole proprietors and 18.8 percent of gig (online platform) workers contributed.¹⁸

A partial reason for low participation by independent workers in workplace retirement plans is that those who do have access to a plan may find it difficult to contribute because their incomes are often lower or sometimes more volatile. The BLS’ report shows that, among full-time workers, contingent workers’ median weekly earnings were about 77.5 percent of noncontingent workers’ median weekly earnings.19 Earnings discrepancies may be due to the different demographic characteristics of independent and other workers and to the types of jobs each group holds. About 49 percent of contract workers told NPR/Marist that their income changed from month to month or seasonally, compared to 34 percent of full- and part-time workers overall.²⁰

Independent work as a second job or ‘side hustle’

Multiple surveys have found that independent workers often have another job. In many cases, the primary job may be in a traditional work relationship and the independent work is a “side hustle.” The Federal Reserve Board’s Survey of Household Economics and DecisionMaking found that one third of self-employed workers and 39 percent of gig workers are supplementing income from a primary job, while only 16 percent of gig workers see it as their primary jobs.²¹ The Federal Reserve’s EIWA Survey found that 72 percent of workers engaged in informal work reported having a formal job as well.²² McKinsey Global Institute reports that 54 percent of independent workers surveyed used independent work to supplement their income, rather than as a primary job.²³ The NPR/Marist poll found that 40 percent of contract workers have another paying job. About 31 percent of Uber drivers work full time on another job while another 30 percent work part time at another job.²⁴

JPMorgan Chase & Co. found that, among workers in the online platform economy (“gig” work), those earnings represented about 54 percent of total take-home income during months in which they actively participated in platform work during 2017 and the first quarter of 2018. Workers cycled in and out of platform work, however, over the course of a year. JPMorgan Chase found that platform earnings as a share of total take-home income during the prior year fell to about 21 percent for the relatively few families who had engaged in this work at any time during the prior year. These findings suggest that online earnings are important to the families who participate, perhaps to smooth or augment income, but are not the only source of family income.²⁵

Independent workers’ demand for retirement benefits

Scant data exist on whether independent workers want greater access to retirement benefits. Surveys generally indicate that many independent workers are dissatisfied with their benefits. But many surveys commingle retirement benefits with health insurance and vacation leave in a single question or two, making it difficult to gauge the demand for retirement benefits in particular.

More evidence on independent workers’ demand for retirement benefits would be helpful as policymakers and researchers consider proposals to expand retirement plan coverage. For example, policymakers would benefit from knowing which categories of independent workers are interested in accessing retirement savings plans. Some types of independent workers—perhaps those with steady clientele or higher earnings—may be more interested in retirement plan coverage than others. It would also be useful to know the extent to which independent workers already have a retirement plan through another job.

Some evidence shows unmet need for retirement savings vehicles among independent workers. For example, one-third of full-time independent workers told talent provider MBO Partners Inc. that planning for retirement is a top challenge.²⁶ About 34 percent of workers in on-demand jobs told Intuit that they had no retirement savings.²⁷About 56 percent of full-time freelancers told the Freelancers Union and Upwork that they had $5,000 or less in savings; only 23 percent said they were on track or prepared for retirement.²⁸ Of note, 7 in 10 freelancers told the Freelancers Union and Upwork that they would prefer to manage their own benefits from higher pay rather than receive less pay in return for employer- or client-provided benefits.²⁹ In contrast, however, more than three-quarters of respondents in a 2014 Pew survey of all types of workers (independent and traditional) said they would prefer to receive employer-provided health insurance or retirement benefits, rather than a cash pay raise of the equivalent value.³⁰

Looking at benefits overall, including health insurance and vacation perks, the Federal Reserve Board’s Survey of Household Economics and Decisionmaking found that just 3 in 10 contract workers were satisfied with their benefits packages.³¹ The GAO, using the 2010 GSS, found 28.4 percent of core contingent workers (agency temps, direct-hire temps, on-call workers and contract company workers) and 28.7 percent of independent contractors said it was “not at all true” that their fringe benefits were good. By comparison, only 12.9 percent of traditional workers in the GSS said their fringe benefits weren’t good.³²

In contrast, McKinsey Global Institute reports that those who derive their primary income from independent work and actively choose this work (“free agents”) were very slightly more, not less, likely than traditional workers to be satisfied with their benefits packages. But those who were independent by necessity were less likely to be satisfied with their benefits than traditional employees. McKinsey notes that the results about satisfaction for the U.S. are surprising, and speculates that low expectations for benefits, the introduction of Affordable Care Act marketplaces for health insurance, and/or eroding benefits for traditional employees may have “lessened the tradeoffs involved.”³³

Theoretically, some workers with unpredictable or volatile incomes may focus instead on short-term, rainy day savings. Others with longer-term or steady contracts, or with higher incomes, may be more likely to save for retirement. One survey’s findings bear this out: Freelancers were more likely to save for periods between jobs, for education, and for homebuying, while nonfreelancers were more likely to save for retirement.³⁴

Gaps in data lead to unanswered questions

Researchers have not yet measured the role of second jobs in securing workers’ financial stability and access to workplace retirement plans. One hypothesis is that second jobs may bring several types of benefits to households, providing additional and potentially more stable income and/or needed health and retirement benefits.

For many, independent work may be a side job, and the primary job may provide pension coverage, possibly at the same rates as other workers. This has not been documented, however, and it is also possible that many workers with second jobs have primary jobs that provide inadequate income and benefits. Data on the extent of participation in a workplace retirement savings plan through a different job, or through a spouse, is a significant gap in knowledge and points to the need for better data.

Of those who don’t have a retirement savings plan, how many independent workers want it? How would they want retirement savings packages to be delivered? And what priority do they give retirement benefits relative to other concerns, such as health insurance or smoothing volatile incomes? Again, existing survey data do not provide much insight into these questions.

Better data on employer attitudes, concerns, and hiring patterns also would be helpful. What are employers’ concerns about benefits provision, and for which workers?

Answers to these questions would help policymakers to develop appropriate policies, and financial industry benefits managers to develop appealing products, that can bring independent workers into the retirement savings system.

Conclusion

Employer-provided benefits, including retirement plan coverage, are usually not available to independent workers. As a result, most independent workers must save for retirement on their own, researching options and contributing systematically. For those who don’t save, the future may hold an impoverished retirement—or no retirement.  

These concerns point to a need to explore new ways to provide secure, widely available retirement plans for independent workers. New savings vehicles should leverage what researchers have learned from behavioral science to expand access and participation. 

Several data gaps and unanswered questions persist for policymakers and analysts in this field, however. More and better data on the phenomenon of retirement plan coverage through second or other jobs would help with developing policies and platforms. Understanding which categories of independent workers would like access to retirement savings opportunities, and in what form, would also help policymakers as they develop proposals to expand retirement plan coverage.

Endnotes

1. K. Abraham et al., “Measuring the Gig Economy: Current Knowledge and Open Issues” (working paper, 2017), https://aysps.gsu.edu/files/2016/09/Measuring-the-Gig-Economy-Current-Knowledge-and-Open-Issues.pdf.
2. Workers who receive a wage or salary generally enjoy worker protections under the Fair Labor Standards Act, unemployment insurance, workers’ compensation, and other laws and regulations pertaining to employees.
3. U.S. Bureau of Labor Statistics, “Contingent and Alternative Employment Arrangements—May 2017,” news release, June 7, 2018, https://www.bls.gov/news.release/conemp.nr0.htm.
4. U.S. Government Accountability Office, “Contingent Workforce: Size, Characteristics, Earnings, and Benefits” (2015), https://www.gao.gov/products/GAO-15-168R.
5. E. Jackson, A. Looney, and S. Ramnath, “The Rise of Alternative Work Arrangements: Evidence and Implications for Tax Filing and Benefit Coverage” (working paper, U.S. Department of the Treasury, Office of Tax Analysis, Washington, D.C., 2017), https://www.treasury.gov/resource-center/tax-policy/tax-analysis/Documents/WP-114.pdf.
6. Abraham et al., “Measuring the Gig Economy.”
7. Ibid.
8. Pew Research Center, “Three in Ten U.S. Jobs Are Held by the Self-Employed and the Workers They Hire” (2015), http://www.pewsocialtrends.org/2015/10/22/three-in-ten-u-s-jobs-are-held-by-the-self-employed-and-the-workers-they-hire/. The Survey of Business Owners was most recently fielded in 2012, see https://www.census.gov/library/publications/2012/econ/2012-sbo.html.
9. U.S. Bureau of Labor Statistics, “Highlights of the May 2017 Data on Electronically Mediated Work,” accessed Sept. 29, 2018, https://www.bls.gov/cps/electronically-mediated-employment.htm#highlights. The BLS included four new questions on electronically mediated employment in the May 2017 Contingent Worker Supplement. With the release of the survey data, the BLS stated that the new questions on electronically mediated work “did not work as intended” and that the bureau had manually recoded false positives, although insufficient information was available to recode all cases and there may still be limitations to the recoded data. Note also that the BLS asks workers only about their main or primary employment, with the result that workers engaged in gig work to supplement their incomes are not included.
10. Jackson et al., “The Rise of Alternative Work Arrangements.”
11. B. Robles and M. McGee, “Exploring Online and Offline Informal Work: Findings from the Enterprising and Informal Work Activities (EIWA) Survey” (working paper, Federal Reserve Board, Divisions of Research & Statistics and Monetary Affairs, Washington, DC, 2016), https://www.federalreserve.gov/econresdata/feds/2016/files/2016089pap.pdf.
12. Upwork and Freelancers Union, “Freelancing in America: 2017” (2017), https://www.upwork.com/i/freelancing-in-america/2017/.
13. McKinsey Global Institute, “Independent Work: Choice, Necessity and the Gig Economy” (McKinsey & Co., 2016), https://www.mckinsey.com/featured-insights/employment-and-growth/independent-work-choice-necessity-and-the-gig-economy.
14. Y. Noguchi, “Freelanced: The Rise of the Contract Workforce,” National Public Radio, Jan. 22, 2018, https://www.npr.org/2018/01/22/578825135/rise-of-the-contract-workers-work-is-different-now.
15. U.S. Bureau of Labor Statistics (2018), Table 9. The figures in the text use BLS contingent worker estimate three, the BLS’ broadest estimate, which includes all workers who do not expect their jobs to last. Eligibility varies with the type of work arrangement. For example, about 48.1 percent of workers provided by contract companies were eligible for employer-provided plans, but only 12.7 percent of temporary help agency workers and 35.4 percent of on-call workers were eligible. (The BLS did not collect this data for independent contractors.)
16. U.S. Government Accountability Office, “Contingent Workforce.”
17. NPR/Marist, “Contract Jobs Are the New Normal: NPR/Marist Poll,” news release, Jan. 22, 2017, https://www.npr.org/about-npr/579672112/contract-jobs-are-the-new-normal-npr-marist-poll.
18. Jackson et al., “The Rise of Alternative Work Arrangements.”
19. U.S. Bureau of Labor Statistics, “Contingent and Alternative Employment Arrangements.”
20. NPR/Marist, “Contract Jobs Are the New Normal.”
21. Board of Governors of the Federal Reserve System, “Report on the Economic Well-Being of U.S. Households in 2017” (2018), https://www.federalreserve.gov/publications/2018-economic-well-being-of-us-households-in-2017-preface.htm.
22. Robles and McGee, “Exploring Online and Offline Informal Work.”
23. McKinsey Global Institute, “Independent Work.”
24. J. Hall and A.B. Krueger, “An Analysis of the Labor Market for Uber’s Driver-Partners in the United States” (working paper, National Bureau of Economic Research, 2016), http://www.nber.org/papers/w22843.
25. D. Hall, F. Greig, and A. Hamoudi, “The Online Platform Economy in 2018: Drivers, Workers, Sellers, and Lessors” (JPMorgan Chase & Co. Institute, 2018), https://www.jpmorganchase.com/corporate/institute/report-ope-2018.htm.
26. MBO Partners, “The State of Independence in America: Rising Confidence Amid a Maturing Market” (2017), https://www.mbopartners.com/uploads/files/state-of-independence-reports/StateofIndependence-2017-Final.pdf. The 2018 update of this report does not follow up questions about retirement planning concerns.
27. Intuit Turbotax + Intuit Quickbooks, “Dispatches from the New Economy: The On-Demand Economic Worker Study” (2017), https://intuittaxandfinancialcenter.com/wp-content/uploads/2017/06/Dispatches-from-the-New-Economy-Long-Form-Report.pdf.
28. Upwork and Freelancers Union, “Freelancing in America, 2017.”
29. Ibid.
30. The Pew Charitable Trusts, “Toplines for the Survey of American Family Finances” (2015), https://www.pewtrusts.org/~/media/assets/.../fsm-poll-results-toplines_artfinal_v3.pdf.
31. Board of Governors of the Federal Reserve System, “Report on the Economic Well-Being of U.S. Households in 2017.”
32. U.S. Government Accountability Office, “Contingent Workforce.” In the General Social Survey, the term “fringe benefits” was left open to respondents’ interpretation.
33. McKinsey Global Institute, “Independent Work.”
34. Upwork and Freelancers Union, “Freelancing in America, 2017."