View original document

The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.

For release on delivery
12:30 p.m. EST
November 17, 2016

The “Gig” Economy:
Implications of the Growth of Contingent Work

Remarks by
Lael Brainard
Member
Board of Governors of the Federal Reserve System
at
“Evolution of Work,” a convening cosponsored by
the Board of Governors of the Federal Reserve System,
the Federal Reserve Bank of New York, and the Freelancers Union
New York, New York

November 17, 2016

I would like to thank Bill Dudley and Sara Horowitz for inviting me to participate
in today’s conference. The subject we are tackling today--the evolution of work--is a
particular interest of mine and is a matter of great importance to the Federal Reserve.
The Congress has mandated the Federal Reserve to implement monetary policy so as to
promote maximum employment and stable prices. Our dual mandate recognizes the
importance of work in enabling people to contribute to the financial security of their
family and the prosperity of their community and the country overall. Moreover, there is
a long-standing recognition that secure and dignified work provides a key sense of
purpose and worth. Understanding the changing nature of employment in today’s
economy is not only central to the mission of the Federal Reserve, but also goes to the
core of who we are as providers for our families and productive members of society. 1
In contrast to traditional work arrangements, in which an employee has a durable
employment relationship with a single primary employer, a large and growing proportion
of the workforce is working through contracting, temporary arrangements, on-call
arrangements, or as freelancers being hired for episodic “gigs.” Broadly speaking,
contingent arrangements are more transitory than traditional arrangements, in the extreme
consisting of a single transaction or gig. They often provide considerably greater
flexibility than long-term employment contracts, allowing workers and employers to
move in and out of work relationships easily. Depending on the nature of the
employment relationship, this enhanced flexibility could have benefits and costs that
accrue to workers and employers very differently.

1

These remarks represent my own views, which do not necessarily represent those of the Federal Reserve
Board or the Federal Open Market Committee.

-2Although they have always been a feature of the American economic landscape,
there has been a sharp increase in the prevalence of contingent working arrangements
over the past decade, and it is too early to tell how much of this acceleration is a cyclical
phenomenon associated with the Great Recession or reflects a structural trend. The
growing share and variety of contingent work has important implications for policy and
puts a premium on data and research exploring this topic. For monetary policy, the
growth of contingent work affects the way we assess maximum employment and the way
we interpret important labor market outcomes, such as the level of part-time employment
and aggregate hours worked. Depending on the contractual arrangements, it may also
have important implications for economic security and the behavior of households as
consumers and savers. Richer and timely data and analysis could help guide employers,
workers, and public officials toward outcomes where benefits and risks are better
understood and managed.
What Do We Know about Contingent Work?
Official measures of the changing nature of work have not kept pace with the
evolution of the economy. But thanks to some cutting-edge researchers, several of whom
are present today, there is a growing foundation of analysis.
Last year, Larry Katz and Alan Krueger conducted a version of the Contingent
Worker Survey (CWS) to track alternative and nonstandard work arrangements using the
RAND American Life Panel. 2 Their findings are striking: Over the past decade,
contingent workers have increased by roughly one-half and now make up 16 percent of
the workforce. This rapid increase is in marked contrast to the preceding decade, when
2

The Contingent Worker Survey is the main survey used by the Bureau of Labor Statistics for tracking
alternative and nonstandard work. The survey was last fielded in 2005.

-3contingent workers remained at a relatively stable 10 percent of the workforce, according
to the Bureau of Labor Statistics’ (BLS) CWS. 3 They conclude that all of the net growth
in aggregate employment in the decade leading up to 2015 can be accounted for by
contingent work arrangements, which means there has been no net employment growth in
traditional work arrangements. Given that this period spanned the Great Recession and
the first five years of a slow recovery, it naturally raises the question of how much of the
large recent shift toward contingent work is attributable to cyclical factors, as opposed to
structural forces that may be here to stay. 4
Long before the advent of firms such as Uber and TaskRabbit, many individuals
worked in contingent arrangements, such as freelancing and contracting, and it is these
groups that have increased their share of the workforce most notably over the past
decade. The largest increase in contingent work over this time has been among workers
whose services are contracted out by another company, which rose from 0.6 percent to
3.1 percent of the workforce. There was also a large increase in workers who are
independent contractors, consultants, or freelancers, whose share grew from 6.9 percent
in 2005 to 8.4 percent in 2015. On-call workers expanded from 1.7 percent to 2.6 percent
of the workforce. 5

3

See Katz and Krueger (2016).
The U.S. Government Accountability Office (GAO) study of the contingent workforce shows an increase
of 5 percentage points in the share of contingent work arrangements, although this shift took place from
2006 to 2010. The GAO’s definition of contingent work covers a much larger share of the workforce,
40.4 percent in 2010, than the CWS. See U.S. Government Accountability Office (2015).
In the same vein, a 2014 survey by the Federal Reserve Bank of Boston found that roughly 44 percent of
its survey’s respondents participated in some informal paid work activity during the previous two years.
See Bracha and Burke (2014).
5
See Katz and Krueger (2016).
4

-4Over this same period, technological platforms have emerged that are
transforming the way people identify, schedule, and engage in contingent work and are
also helping people to monetize their assets to generate income. These developments,
under the rubric of the “gig,” “on-demand,” “platform,” or “sharing” economy, have the
potential to be transformative. Although the data are still relatively sparse, it appears that
the number of workers using online platforms to secure gigs is still small but is growing
quickly. 6 Katz and Krueger found that workers who provide services through three
online intermediaries accounted for only 0.5 percent of all workers in 2015. Diana
Farrell and Fiona Greig (2016) reach a similar conclusion, using a unique and rich data
set. 7
The Federal Reserve Board fielded a survey to better understand the many ways
adults are generating income, regardless of their employment status and the frequency of
these activities. The survey uses an expansive concept of contingent work, capturing all
of the activity individuals are undertaking to generate income. The Enterprising and
Informal Work Activity survey, or EIWA, suggests that more than a third--36 percent--of
the adult population undertook informal paid work activity either as a complement to, or
as a substitute for, more traditional and formal work arrangements. 8 Similarly, a survey

6

See Harris and Krueger (2015).
By contrast, studies conducted by the Federal Reserve Bank of Boston in 2013 and 2015 found that
approximately one in four respondents--or more than half of informal work participants--say they earned
money through informal work activities conducted online. See Bracha, Burke, and Khachiyan (2015).
8
See Robles and McGee (2016).
Some informal workers in the EIWA would not meet the definition of employment in the Current
Population Survey (CPS) either because they did not do work in the CPS reference week (the EIWA asks
about activities over the past six months, not during the reference week) or because they would not
categorize the covered activities (such as the occasional selling of goods or renting of property) as work.
7

-5undertaken by Upwork and the Freelancers Union that also uses an expansive definition
of contingent work finds that 35 percent of the labor force engages in freelance work. 9
The Upwork survey provides further insight by subdividing the 35 percent of the
labor force who have done some form of informal work over the past 12 months into five
categories. In the categories most comparable with the Katz and Krueger CWS
categories, roughly 12 percent of the workforce identify as independent contractors who
do not have an employer and engage in project-by-project work, whereas between 2 and
3 percent are traditional contract or temporary workers who have a single employer or
project at a time and whose work status is temporary. By contrast, roughly half of the
Upwork respondents do informal work in addition to traditional employment
relationships; almost 10 percent derive their income from multiple different types of
arrangements, including both traditional employment relationships and freelancing, while
about 9 percent report having a full-time traditional job and “moonlighting” to gain
additional income on top of that. The group that derives income from multiple different
types of jobs has evidenced the fastest growth over the last few years covered by the
survey. In addition, 2.5 percent identify as business owners.
Finally, the EIWA survey provides valuable insights into the use of recent
technologies in informal work. The EIWA survey suggests that over the previous six
months, roughly 11 percent of adults in the United States have engaged in paid services

9

The Upwork survey encompasses an expansive set of contingent work arrangements, even if only
intermittent, over the past 12 months, and the EIWA encompasses a similar scope over the past 6 months,
while Katz and Krueger confine their focus to the Current Population Survey reference week, similar to the
BLS CWS. In addition, the Upwork definition is similar to the EIWA in counting as informal workers
those individuals with traditional full- or part-time work who also engage in some informal work, in
contrast to Katz and Krueger and the BLS CWS, which try to differentiate those workers who
predominantly engage in contingent work.

-6using an online platform, such as arranging transportation activities with companies such
as Uber, arranging work opportunities on sites like Care.com, Amazon Mechanical Turk,
or TaskRabbit, generating projects through companies such as Fiverr, selling goods and
crafts on sites such as eBay and Etsy, and renting rooms or homes through services like
Airbnb, among others. 10 This figure is much larger than the Katz and Krueger finding of
0.5 percent, partly because the EIWA survey was fielded to all adults regardless of their
formal employment status, asked about all online paid services, and used a six-month
look-back period, rather than the past week.
Who Is Engaging in Contingent Work and Why?
Together, these recent surveys of informal work paint a varied picture of the
universe of informal work arrangements, along with participants’ motivations for
engaging in them. Nineteen percent of informal workers in the EIWA survey, for
example, were engaged in three or more online and/or offline informal paid work
activities in the prior six months, and 25 percent indicated that informal and contingent
work activities have been “very much” or “somewhat” a regular and consistent source of
their monthly income.
Results regarding the ages of workers attracted to nontraditional work vary: Katz
and Krueger find that the share of contingent work arrangements increases with age,
while the Upwork study finds that freelancing is more prevalent among the young. The
earnings experiences of contingent workers appear to span a wide range. Katz and
Krueger find that those engaged in contingent work tend to be concentrated in the highest

10

A study by the McKinsey Global Institute suggests that 15 percent of independent workers in the United
States and the European Union participate in online informal work activities. See Manyika and
others (2016).

-7income quintiles. But while independent contractors and contract workers earn relatively
high wages, on-call and temporary-help work tends to occur among workers in lower
wage quintiles.
One common theme that emerges is the desire to earn extra money, which is a
primary motivation for 74 percent of part-time, and 68 percent of full-time, freelancers in
the Upwork survey. Similarly, in the EIWA survey, 65 percent of informal workers
reported that earning income was their main reason for engaging in informal and
contingent work, and the EIWA provides many examples that corroborate this result. 11 A
full-time restaurant cook in his 50s, reports that he does yard work and sells items both
online and offline to earn extra income and to help family members. He averages 60
hours per month from side and gig work and reported that these earnings constitute 75
percent of his household’s monthly earnings. Similarly, a respondent in her 20s works
part time as a dental hygienist and supplements this employment with about 8 hours per
month of additional gig work. She said that the gig work somewhat mitigated the effect
of the Great Recession, which found her working fewer hours, with stagnant wages and a
loss of benefits.
For others, work-life balance considerations seem to play a significant role in
choosing contingent work arrangements. For example, a mother in her 20s reported
spending 12 hours per month selling self-crafted items online, babysitting, and
completing online tasks, collectively providing about 20 percent of her family’s income.
She reported that her main reason for participating in the gig economy is to be home with
her daughter.

11

The responses to the survey were anonymous and confidential.

-8Many others find themselves in highly variable work arrangements not by
preference but because of the requirements of their employers. As employers have
outsourced noncore tasks to contract firms and moved to technologically enabled “justin-time” scheduling of employees, contract workers and on-call workers have had to
adjust to variable and sometimes unpredictable hours.
As these stories indicate, there are a variety of reasons workers engage in
alternative and nonstandard work arrangements. Many individuals use the flexibility and
enhanced connectivity of new technology-enabled platforms to pursue expanded
opportunities. While some supplement their traditional job, using their free time to earn
extra income, others cobble together contingent work arrangements or gigs to generate
necessary income in the absence of traditional work opportunities. Still others move into
contingent work arrangements due to their employers’ work requirements. While some
workers seem to appreciate the greater opportunities contingent work arrangements
provide, others engage in this work out of necessity because they appear unable to obtain
a full-time traditional work arrangement that meets their needs. Overall, the Upwork
study finds that 63 percent of freelancers say they are pursuing this type of work by
choice, while 37 percent do it out of necessity. It is unclear whether this outcome would
hold true of the more tightly defined population in the Katz and Krueger survey.
Implications of a Growing Gig Economy
If market forces and technology are driving the growing prevalence of gig work,
these trends will likely continue, and policymakers must better understand these changes.
A natural starting point is measurement. There is some good news on this front, as the
BLS is preparing to field a CWS in May 2017 for the first time since 2005.

-9Better data should help deepen our understanding of how the growth of contingent
work is changing the behavior of workers and employers and what these developments
imply for the overall economy and household welfare. From a macroeconomic
perspective, we should be attuned to the possibility that the growth in flexible work
arrangements may alter the natural rate of employment and how the labor market reacts
to shocks as we assess the cyclical behavior of important aggregates--such as labor force
participation, employment, hours worked, and underemployment.
One possibility is that new technologies, by lowering the barriers to workforce
entry, could raise employment and labor force participation. Because there are fixed
costs to obtaining and maintaining a job--such as searching for the job, learning about the
job, and traveling to the job site every workday--the traditional work arrangement
embeds incentives for individuals to work only one job, thus minimizing these fixed
costs. And for those individuals who desire to work relatively few hours, these fixed
costs in the past may have led to a decision to remain out of the labor force. But if
technology makes it less costly for individuals to find work, manage multiple work
relationships, and flexibly work more or fewer hours as their schedule permits, this could
significantly increase workers’ options and have important effects on labor market
behavior. Importantly, any such increases in participation and employment would likely
be structural, not cyclical, enabling the economy to run at a sustainably higher level.
New technologies could also make it easier for individuals’ actual hours to match
their desired hours of work in a day or a week. Instead of seeing hours per week bunched
around 40, we may see greater variation in the hours that individuals work. Lower
barriers to workforce entry may make it more attractive for more individuals to work only

- 10 a few hours a week if they desire--and can afford--to do so. At the same time, by making
it easier to find additional work, new technologies may lead to more individuals working
greater-than-full-time hours every week. In addition, we may see individuals changing
the number of hours they work more frequently, as these changes become less costly.
Both the EIWA and Upwork surveys confirm that gig work does, indeed, lower barriers
to workforce entry and increase hours and flexibility for some individuals.
The increasing prevalence of gig work may also affect the unemployment rate and
productivity. To the extent that gigs provide an easy entryway to employment,
unemployment may decrease. However, if gig work is less stable, it may increase job
loss. The net effect on unemployment is, thus, unclear. Regarding productivity, gig
work could lower aggregate productivity to the extent that it requires less human capital
or specialized knowledge than traditional jobs, or if it primarily increases hours worked
by lower-skilled individuals. That said, gig work, especially when enabled by new
technologies, may allow hours to respond more flexibly to changes in demand and
individuals to more easily connect with many different clients or employers. As a result,
workers’ downtime and the time required to acquire new clients and manage existing
clients may decrease, in which case resource utilization and productivity may increase.
It is also possible that the increasing prevalence of gig work will cause the
cyclical behavior of unemployment, participation, and the workweek to change, with
implications for how we assess the amount of slack. We know that contingent work
increases when the economy worsens. If new technologies make it easier to find gig
work, then we could see unemployment rise less in recessions, to the extent that gig
workers are counted as employed. However, it could be that individuals who are able to

- 11 avoid unemployment through contingent work would still be underemployed if it is
difficult to cobble together enough gigs to achieve full-time employment. This would
likely show up as lower average workweeks and higher levels of involuntary part-time
employment during downturns. As a result, cyclical changes in resource utilization could
be reflected less in movements in the unemployment rate and more in variation in hours
per worker.
Beyond the behavior of macroeconomic variables, it is unclear how the growth of
different types of gig work affects the welfare of workers. Welfare should increase in
cases where gig work meets the needs of workers by providing a low-barrier means of
accessing employment and by allowing workers to better match actual hours worked with
desired hours of work, especially if the gig work is available at times and in places where
traditional work opportunities are in short supply.
There is some evidence that this has, indeed, been the case. For example,
24 percent of informal workers in the EIWA survey indicated that contingent work had
offset their spells of unemployment, loss of working hours, loss of benefits, or frozen
wages in their formal employment “very much” or “somewhat.” In addition, threefourths of Uber drivers say that the greater control over their work schedules that Uber
allows has made their lives better. 12
However, there are likely many workers who would prefer regular full-time
traditional work to contingent work, particularly if much of the power in determining
hours worked in alternative work arrangements belongs to the employer. Technological
advances have enabled firms to use just-in-time strategies for their employees, making

12

See Hall and Krueger (2015).

- 12 them in effect on-call workers. This is a rising trend in industries such as retail and food
preparation. Several recent articles describe the challenges faced by these just-in-time
workers, who must conform their hours to the daily and even hourly ebbs and flows of
business, often not knowing whether they will have work on a given day until they call in
that morning to inquire. 13 These arrangements can leave workers scrambling to patch
together child care, elder care, and transportation to meet the often unpredictable
demands of their workplace, while making it difficult to engage in regularly scheduled
activities to enhance their income and opportunities, such as a second job or career
training. While such workers often are not given full-time work, they often must make
themselves available to work full-time hours. According to one survey, 71 percent of
retail workers in New York stated that their hours fluctuated from week to week, while
half said their employers could change their hours at will. 14 It is also notable that the
increase in contingent work over the past decade has coincided with an increase of
one-third in the share of employees working part time but who would prefer to work full
time from 3 percent prior to the Great Recession to close to 4 percent today.
In addition, contingent workers may receive lower wages, less training, and fewer
benefits than their counterparts with traditional jobs. Typically, the wages of low-skilled
employees within a company are boosted by social norms regarding pay equity, and
nonpecuniary benefits are often equalized across a company’s employees, in certain cases
as mandated by law. 15 However, the wages that contractors receive are unlikely to reflect
the same equity considerations. Moreover, contingent work generally does not offer

13

See Ansel (2015) and Carrillo and others (2016).
The 2011 survey was conducted by the Retail Action Project, cited in Wessler (2014).
15
Regulations stipulated by the Employee Retirement Income Security Act and the Affordable Care Act
generally mandate equal benefits provision of pension and health insurance plans across employees.
14

- 13 employer-based benefits and workplace protections that come with traditional
employment opportunities, like overtime compensation, minimum wage protections,
health insurance, family leave, employer-sponsored retirement plans, workers’
compensation, and paid sick leave.
As a result, for some, contingent work may entail greater risks than in traditional
full-time employment, with more variable and less predictable hours and earnings. The
Upwork study notes that one of freelancers’ biggest concerns is managing income
variability and benefits. For lower-income workers, unpredictable fluctuations in income
can lead to severe hardship. The Federal Reserve Board’s Survey of Household
Economics and Decisionmaking, for example, finds that 46 percent of households report
that they would need to borrow money or sell something in order to pay an unexpected
expense of $400. 16
These findings suggest that employers, policymakers, and workers should seek
ways to help individuals better manage the risks inherent in most forms of contingent
work. For example, we may need to enhance social safety net programs, such as
unemployment and disability insurance, to better support some types of contingent work.
Another possibility is to make benefits, such as health insurance and retirement saving,
portable across different employers. We may also want to encourage the additional
saving that many contingent workers need to ensure that their basic consumption needs
are not sacrificed when demand for their work declines, perhaps by providing monetary
or other types of incentives.

16

The survey is available on the Board’s website at
https://www.federalreserve.gov/communitydev/shed.htm.

- 14 Closing Thoughts
The apparent trend toward contingent work has recently coincided with new
advances in technology that can potentially amplify this trend and push it in new
directions. We are still at an early stage in these developments, and it is too soon to say
how these changes will play out. But the effects on the labor market could be long
lasting and significant. Taking into account the potentially varied effects of the rising
prevalence of gig work on household welfare, public policy should strive to maximize the
benefits of the greater flexibility and lower entry barriers provided by advances in
technology, while addressing the risks that currently accompany many forms of gig
employment. Going forward, better data will be necessary, along with detailed research
and analysis, in order to enable workers, employers, and policymakers to guide these
changes in the labor market in a direction that ensures the benefits are broadly shared and
the risks are well understood and well managed.

- 15 References
Ansel, Bridget (2015). “The Pitfalls of Just-in-Time Scheduling,” Washington Center for
Equitable Growth, January 27, http://equitablegrowth.org/equitablog/pitfalls-justtime-scheduling.
Bracha, Anat, and Mary A. Burke (2014). "Informal Work Activity in the United States:
Evidence from Survey Responses," Current Policy Perspectives 14-13. Boston:
Federal Reserve Bank of Boston, December, https://www.bostonfed.org//media/Documents/Workingpapers/PDF/economic/cpp1413.pdf.
Bracha, Anat, Mary A. Burke, and Arman Khachiyan (January 2015). “Changing
Patterns in Informal Work Participation in the United States 2013-2015,” Current
Policy Perspectives 15-10. Boston: Federal Reserve Bank of Boston, October,
https://www.bostonfed.org/-/media/Documents/cddp/cpp1510.pdf.
Carrillo, Dani, Kristen Harknett, Allison Logan, Sigrid Luhr, and Daniel Schneider
(2016). "On-Call Job, On-Call Family: The Necessity of Family Support among
Retail Workers with Unstable Work Schedules," Working Paper 2016-11.
Washington: Washington Center for Equitable Growth, November,
http://cdn.equitablegrowth.org/wp-content/uploads/2016/10/25123457/110116WP-retail-workers-with-unstable-schedules.pdf.
Farrell, Diana, and Fiona Greig (2016). Paychecks, Paydays, and the Online Platform
Economy: Big Data on Income Volatility. New York: JPMorgan Chase Institute,
February, https://www.jpmorganchase.com/corporate/institute/document/jpmcinstitute-volatility-2-report.pdf.
Hall, Jonathan V., and Alan B. Krueger (2015). "An Analysis of the Labor Market for
Uber’s Driver-Partners in the United States," Working Paper 587. Princeton,
N.J.: Princeton University, Industrial Relations Section, January,
http://dataspace.princeton.edu/jspui/bitstream/88435/dsp010z708z67d/5/587.pdf.
Harris, Seth D., and Alan B. Krueger (2015). "A Proposal for Modernizing Labor Laws
for Twenty-First-Century Work: The 'Independent Worker,' " Discussion Paper
2015-10. Washington: Hamilton Project, Brookings Institution, December,
http://www.hamiltonproject.org/assets/files/modernizing_labor_laws_for_twenty_
first_century_work_krueger_harris.pdf.
Katz, Lawrence F., and Alan B. Krueger (2016). "The Rise and Nature of Alternative
Work Arrangements in the United States, 1995-2015," working paper, March,
https://krueger.princeton.edu/sites/default/files/akrueger/files/katz_krueger_cws__march_29_20165.pdf.
Manyika, James, Susan Lund, Jacques Bughin, Kelsey Robinson, Jan Mischke, and
Deepa Mahajan (2016). Independent Work: Choice, Necessity, and the Gig

- 16 Economy. San Francisco: McKinsey Global Institute, October, available at
http://www.mckinsey.com/global-themes/employment-and-growth/independentwork-choice-necessity-and-the-gig-economy.
Robles, Barbara J., and Marysol McGee (2016). "Exploring Online and Offline Informal
Work: Findings from the Enterprising and Informal Work Activities (EIWA)
Survey," Finance and Economics Discussion Series 2016-089. Washington:
Board of Governors of the Federal Reserve System, November,
https://www.federalreserve.gov/econresdata/feds/2016/files/2016089pap.pdf.
Upwork and Freelancers Union (2016). "Freelancing in America: 2016," survey results,
October 6, available at https://www.upwork.com/i/freelancing-in-america/2016.
U.S. Government Accountability Office (2015). Contingent Workforce: Size,
Characteristics, Earnings, and Benefits. Washington: GAO, April, available at
http://www.gao.gov/products/GAO-15-168R.
Wessler, Seth Freed (2014). "Shift Change: 'Just-in-Time' Scheduling Creates Chaos for
Workers," NBC News, May 10, http://www.nbcnews.com/feature/in-plainsight/shift-change-just-time-scheduling-creates-chaos-workers-n95881.