Testimony | Retirement

Testimony for Hearing on “Exploring the ‘Gig Economy’ and the Future of Retirement Savings”

Download PDF

U.S. Senate Committee on Health, Education, Labor & Pensions
Subcommittee on Primary Health and Retirement Security
Hearing on “Exploring the ‘Gig Economy’ and the Future of Retirement Savings”

February 6, 2018

Statement of Monique Morrissey, Economist, Economic Policy Institute

This hearing addresses an important concern: how nonstandard work affects retirement security. Flexible work arrangements can have positive and negative effects on income across the lifecycle. They can provide work opportunities for older workers who lose their jobs and are transitioning to retirement or who are taking longer to be reemployed than their younger counterparts. “Gig” jobs may also reduce the need to tap into retirement savings between jobs for workers of all ages.

In practice, however, nonstandard work is often low-paid and precarious work that is not conducive to saving or accruing retirement benefits. The bad news is that this problem is likely to be getting worse. The good news is that efforts to improve retirement security for all workers would disproportionately help nonstandard workers.

Narrowly defined, the “gig economy” is very small and is not growing as fast as people assume. Independent workers employed via online platforms represent some 0.5 percent of jobs.1 Nevertheless, companies like Uber, Lyft, Etsy, Airbnb, and TaskRabbit loom large in people’s imaginations, especially in cities with many young professionals.

The nonstandard workforce is much larger than the “gig economy” and includes both W-2 employees and self-employed workers. The former category may include part-time workers, on-call workers, and temporary workers (direct-hire and agency temps). The latter category may include business owners and full-time independent contractors (including misclassified employees), occasional contractors, day laborers, and on-demand platform workers.

These broad categories include contingent workers with little control over their hours and earnings as well as highly-compensated professionals and people who prefer to work part-time. Though our main concern should be for workers who rely on income from nonstandard work to make ends meet and who face hardship in old age, even moonlighters who receive benefits through a primary job will see a sharper drop in income at retirement if their supplemental earnings are not factored into retirement savings or benefits.

Distinctions between “gig economy” workers employed through online platforms and other contingent or nonstandard workers are often exaggerated. There have always been self-employed musicians, cab drivers, and house cleaners. The broad category of “independent contractor” is much larger than this small group of workers and appears to be growing,2 as many workers who in the past would have been W-2 employees are reclassified as independent contractors.

Statistics for nonstandard workers vary, with some sources showing growth and some declines in various subcategories, such as on-call, temporary, or contract firm workers. These differences may be due to different definitions of contingent or nonstandard work; whether the share of jobs, the share of primary jobs, the share of workers, or the share of hours worked is being measured; and problems with survey data, including respondent error, non-representative samples, limited topical questions, and out-of-date surveys. However, most sources agree that independent contractors are the largest category of nonstandard workers, and that this category has continued to grow in the wake of the Great Recession.3

Independent contractors are a heterogeneous group. Workers with college degrees are as likely to be independent contractors as workers with high school degrees. Hispanic and white workers are more likely to be independent contractors than African-American workers.4 These patterns may reflect differences in job quality or geographic variation. For example, Hispanic workers may be overrepresented among misclassified workers, while white workers may be more likely to have white-collar occupations that can become self-employment in the transition to retirement. White workers are also more likely to live in sparsely-populated states where multiple jobholding is more common.5

Another explanation for these patterns is that employment rates tend to be higher among older whites and Hispanics than among older African Americans, and older workers are overrepresented among independent contractors. Nearly 16 percent of 55-75-year-old workers are independent contractors.6 Older workers are as likely as, or less likely than, prime-age workers to be employed in other types of nonstandard work, such as temporary help service workers and contract firm employees.7

In the gig economy, technology helps solve coordination problems, reduces financial transaction costs, and enforces quality standards, among other efficiency gains. For example, in the ride-sharing industry, technology helps connects passengers with nearby drivers, enables surge pricing to increase the supply of drivers as needed, centralizes payments and recordkeeping, and facilitates customer reviews.

What is good for the customer (and perhaps the economy) may sometimes also benefit workers. Flexible scheduling can enable workers to piece together multiple part-time jobs or to moonlight, students to work around class schedules, parents to vary hours based on childcare availability, and seniors to transition into retirement.

However, nonstandard workers are more likely to face financial insecurity, including retirement insecurity.8 A report by the U.S. General Accountability Office (GAO) found that contingent workers—the estimated eight percent of workers in 2010 who lacked job security or had variable or unpredictable work schedules—earned 11 percent less on an hourly basis and 48 percent less on an annual basis than otherwise similar non-contingent workers. These workers, in addition to working reduced and fluctuating hours, were also much more likely to become unemployed or exit the labor force. The GAO report found that these contingent workers were two-thirds less likely than standard workers to have a work-provided retirement plan, a disadvantage that both reflected and compounded their financial insecurity.9

In theory, Social Security’s universal coverage and progressive benefit structure should partly compensate for contingent workers’ lower earnings and lack of access to employer benefits. However, nonstandard workers are more likely to be to be paid under the table or to be classified (or misclassified) as independent contractors. Self-employed workers have greater incentive to underreport earnings or inflate expenses in tax returns, shrinking the tax base as well as their future Social Security benefits.10

In his 2014 book, The Fissured Workplace, David Weil described how contracting and outsourcing allowed companies to reduce the number of workers who share in employee benefits without running afoul of nondiscrimination rules. By avoiding legal responsibility for these workers, fissuring also encourages contractors, including self-employed workers, to underreport income and avoid contributing toward social insurance benefits as well as evade other worker protections.

While nonstandard workers may be at greater risk of retirement insecurity, they are hardly alone. By one conservative estimate, half of American households are at risk of being unable to maintain their pre-retirement standard of living in retirement, and younger generations and lower-income workers are at greater risk.11 Moreover, retirement wealth has become more insecure and unequal with the shift from traditional defined benefit pensions to 401(k)-style contribution plans.12

Because few people save for retirement on their own, lack of access to an employer-based plan is a significant barrier for many Americans, especially lower-income Americans. Retirement plan participation is almost three times as high for individuals with incomes above 300 percent of the poverty line as for those below this threshold, and the biggest factors driving this disparity are lower employment rates and working for an employer that does not offer a plan.13 Despite the fact that the cards are stacked against lower-income workers, who have less disposable income and often receive little or no help from their employer or the government in saving for retirement, take-up rates are high (78 percent) for those who have access to a 401(k) plan.14

Most efforts to improve retirement security for all at-risk workers would disproportionately help nonstandard workers. These include expanding Social Security (and cracking down on tax avoidance), expanding the Saver’s Credit, and supporting state and local initiatives to offer low-cost portable benefits to workers who do not have access to an employer plan.

Another option is eliminating the commonality requirement for employers in multiple-employer plans (MEPS). This is worth pursuing, at least for “open MEPS” sponsored by not-for-profit and government entities acting as fiduciaries. As Michele Varnhagen of AARP has testified before this subcommittee, Congress would need to set strict standards to ensure that participants benefit from economies of scale and are offered appropriate investments.15

Just as most efforts to improve retirement security for at-risk workers would disproportionately help nonstandard workers, most efforts to shore up labor standards would improve the retirement security of nonstandard workers. This includes tightening rules and enforcement to prevent the misclassification of workers as independent contractors.

There are targeted interventions and innovations that might help at the margins, though it would be a mistake to focus narrowly on these. For example, ride-sharing companies Lyft and Uber have partnered with financial technology firms to help their drivers save for retirement. This has received a fair amount of attention because these drivers are already, by definition, “wired.” However, even if these initiatives are emulated by other companies, this is only likely to help a small share of these workers and an even smaller share of the nonstandard workforce.

Other, more ambitious, proposals targeted at nonstandard workers, such as a proposal by William G. Gale, Sarah E. Holmes, and David C. John to create a new type of portable retirement account that could accept both IRA and 401(k) contributions,16 would require a bigger political lift. While it is worth considering the pros and cons of such an approach for supplemental voluntary savings, we already have an efficient and portable benefit—Social Security—which the authors note shares many features with their proposed accounts.

In contemplating ways to help nonstandard workers, it is important not to fall into the trap of accepting these arrangements as inevitable and innovative when the situation might better be described as a race to the bottom. We should take with a grain of salt companies’ claims that labor standards and other protections are incompatible with new business models. As my colleagues Ross Eisenbrey and Larry Mishel noted, Uber claimed it could not adhere to minimum wage laws even as the company set minimum hourly rates when it suited them.17 Even when labor standards do clash with new business models, it does not follow that these should be weakened or scrapped, since these companies’ competitive advantage may derive from evading standards and taxes as much as or more so than from socially beneficial innovation.

The gig economy is a microcosm of the American workplace. Nonstandard workers and other low-income, self-employed, and small business workers face greater risk of retirement insecurity. We should look at ways of helping these at-risk workers, but we should be careful to do so without making a bad situation worse. A system that relies on upside-down incentives to encourage voluntary savings by workers in high-cost and risky plans that their employers may not even offer is a system that is not working for most American workers. This is especially true for lower-income workers engaged in nonstandard work who are less likely to have access to a good employer plan and are more likely to tap into savings to smooth income fluctuations before retirement.

Many of us in this room likely agree that we rely too much on single-employer plans and should look to expand more portable accounts or benefits. However, we likely differ on whether to focus on making it easier for workers to participate in IRAs or other voluntary accounts, or to require workers and employers to contribute more to Social Security or other mandatory or quasi-mandatory systems.

It would appear that the risk of the first approach is doing too little and the risk of the second is doing too much. There are limits to what we can do by nudging people to save more on their own. On the other hand, some argue that expanding Social Security can force some people to over-save: low earners who need the money now to support their families or invest in their educations, and high earners who might get a better return on their contributions elsewhere.

I believe the latter concern is largely hypothetical, especially in the wake of 1983 Social Security benefit cuts and the decline of defined benefit pensions. In any case, framing the debate as one between a libertarian approach on one hand and a social insurance approach on the other misses the fact that both systems need to be fixed. It is possible to do harm even in a voluntary (albeit taxpayer-subsidized) system by steering low-income workers’ savings to high-cost and risky accounts, especially if these workers receive little or no government or employer support yet face tax penalties for accessing funds before retirement. We should therefore focus on reducing costs (by eliminating conflicts of interest in investment advice, among other things), exploring ways to pool risks through annuitized benefits and other means, and fixing upside-down tax incentives.

Expanding Social Security would address these problems. However, this requires additional revenue, and any additional employee contributions would need to be offset for low-income workers by expanding the Earned Income Tax Credit or other means. We also need to address tax avoidance by independent contractors and other self-employed workers, which robs our public services as well as workers’ retirement futures.

There are no easy answers—and we need to be careful to “first, do no harm.” But the problem is serious and urgent and I believe there are areas of common ground.


1. Lawrence F. Katz and Alan B. Krueger, “The Rise and Nature of Alternative Work Arrangements in the United States, 1995-2015,” NBER Working Paper No. 22667, September 2016.

2. Katz and Krueger (2016).

3. Katherine G. Abraham, “What Do We Know About Nonstandard Work?” National Academy of Social Insurance Conference on Nonstandard Work and Social Insurance, January 30, 2018.

4. Minority (both Hispanic and black) workers and less educated workers are more likely to be on-call and temporary help service workers.

5. U.S. Bureau of Labor Statistics, “Multiple Jobholding in States in 2015,” Monthly Labor Review, February 2017.

6. Katz and Krueger (2016).

7. Katz and Krueger (2016); U.S. Government Accountability Office (GAO), “Contingent Workforce: Size, Characteristics, Earnings, and Benefits,” April 20, 2015.

8. See, for example, Shayna Strom and Mark Schmitt, “Protecting Workers in a Patchwork Economy,” The Century Foundation, April 7, 2016; Nancy K. Cauthen, Annette Case, and Sarah Wilhelm, “Promoting Security in a 21st Century Labor Market; Addressing Intermittent Unemployment in Nonstandard Work,” September 2015.

9. GAO (2015).

10. Elliot Schreur, “Social Security’s Disability and Retirement Protections for Independent Contractors,” National Academy of Social Insurance, Conference on Nonstandard Work and Social Insurance, January 30, 2018.

11. Alicia H. Munnell, Wenliang Hou, and Geoffrey T. Sanzenbacher, “National Retirement Risk Index

Shows Modest Improvement in 2016,” Center for Retirement Research at Boston College Issue in Brief Number 18-1, January 2018.

12. Monique Morrissey, “The State of American Retirement: How 401(k)s Have Failed Most American Workers,” Economic Policy Institute Report, March 3, 2016; Alicia H. Munnell, Wenliang Hou, Anthony Webb, and Yinji Li, “How Has the Shift to 401(K) Plans Affected Retirement Income?,” Center for Retirement Research at Boston College Issue in Brief Number 17-5, March 2017.

13. April Yanyuan Wu, Matthew S. Rutledge, and Jacob Penglase, “Why Don’t Lower-Income Individuals Have Pensions?” Center for Retirement Research at Boston College Issue in Brief Number 14-8, April 2014.

14. Take-up rates are likely to be somewhat lower among workers who currently do not have access to a plan, including many nonstandard workers. These workers are likely to have more barriers to participation, and perhaps less motivation, than workers who currently have access to a plan.

15. Statement of Michele Varnhagen before the U.S. Senate Committee on Health, Education, Labor and Pensions, Subcommittee on Primary Health and Retirement Security, on Retirement Plan Options For Small Businesses,

June 21, 2016,

16. William G. Gale, Sarah E. Holmes, and David C. John, “Retirement Plans for Contingent Workers: Issues and Options,” Brookings Institution Retirement Security Project, September 23, 2016.

17. Ross Eisenbrey and Lawrence Mishel, “Uber business model does not justify a new ‘independent worker’ category,” Economic Policy Institute Report, March 17, 2016.

See related work on Retirement

See more work by Monique Morrissey