Self-Employment as a Contributor to Job Growth and as an Alternative Work Arrangement
CRS Report for Congress
Self-Employment as a Contributor to Job Growth
and as an Alternative Work Arrangement
Updated September 14, 2004
Specialist in Labor Economics
Domestic Social Policy Division
Congressional Research Service ˜ The Library of Congress
Self-Employment as a Contributor to Job Growth
and as an Alternative Work Arrangement
Self-employment is the major alternative to the leading type of employee-
employer relationship, which is a firm paying a wage to an employee. The self-
employed have accounted for some 7% of nonfarm jobs over the past few decades.
Self-employment recently has drawn attention as one explanation for differing
survey estimates of net job growth since the 2001 recession’s end. There also was
debate over the two prior decades about the contribution to net job creation of self-
employment or entrepreneurship (i.e., people starting businesses). During the 1992-
2000 economic expansion, interest in self-employment focused on the kind of jobs
being created (i.e., alternatives to secure “traditional” wage and salary jobs) and on
the possibility of misclassifying employees as self-employed consultants or
contractors, who then are ineligible for certain government benefits (e.g.,
Unemployment Insurance (UI), Workers Compensation, and Family and Medical
Leave). Since the early 1990s, the government has encouraged the establishment of
Self-Employment Assistance programs to help reemploy some displaced workers
likely to exhaust their UI benefits.
The contribution of self-employment to net job creation has not varied greatly
by decade, except for the 1990s. In the 1970s and 1980s, increased self-employment
accounted for 7% and 10% of nonfarm job gains, respectively. During the 1990s,
however, people more often became employees such that self-employment
contributed just 1% to net job growth. To attract employees when unemployment
rates were low in the second half of the 1990s, firms may have expanded benefits that
appealed to those who otherwise would have chosen self-employment (e.g., flexible
work hours). Thus far in the current decade, self-employment has been a much
greater contributor to net job growth, at 13%.
Persons laid off in a sluggish labor market may become discouraged about their
hiring prospects and turn to self-employment (e.g., during the slow rebound from the
last two recessions). Studies disagree, however, about whether high unemployment
pushes people into self-employment because of a dearth of other options or whether
it dampens entrepreneurship because of scarce opportunities. And, factors beyond
conditions in the labor market affect the propensity of displaced workers and others
to start businesses. Older workers’ greater likelihood of self-employment may reflect
a desire to phase into retirement without jeopardizing receipt of employer-provided
health insurance and level of employer-provided pension benefits. Women’s
increased self-employment rate may reflect the desire of mothers with young children
for hours flexibility to allow them to balance work and family roles. Thus, over 80%
of independent consultants in the February 1995-2001 period said they preferred their
situation to that of traditional workers. Their satisfaction appears related to the
comparatively high wages of male independent contractors, although the availability
of benefits lags for the group as a whole. Further, in that low unemployment period,
only 3% to 4% of independent contractors considered their positions insecure —
many fewer than others in nontraditional jobs. This report will be updated as
Why the Interest?..................................................1
Who Are the Self-Employed?........................................4
Race and Ethnicity.............................................6
Self-Employment and Job Growth Over Time...........................9
Self-Employment as a Contributor to Job Growth....................10
Reasons for Entering Self-Employment from Unemployment..........10
Self-Employment as an Alternative Work Arrangement...................12
Worker Preferences and Job Insecurity............................13
Misclassification of Independent Contractors.......................17
List of Tables
Table 1. Self-Employment by Selected Characteristics, 2003...............6
Table 2. The Trend in Self-Employment, Selected Years..................9
Table 3. The Job Insecurity Rate, February 1995-2001...................14
Self-Employment as a Contributor
to Job Growth and as an
Alternative Work Arrangement
Self-employment is the primary alternative to the leading type of employee-
employer relationship in the United States, which is an employer paying a wage or
salary to an employee. In 2003, according to data from the Current Population
Survey (CPS), the self-employed numbered a little over 9.3 million. They accounted
for fewer than 7% of all employed persons in nonfarm industries. Wage and salary
workers, at 126.0 million, represented more than 9 of every 10 employed persons in
the nonfarm sector of the economy.1
Why the Interest?
Self-employment recently has drawn attention as a partial explanation for the
differing survey estimates of net job growth (i.e., gross job gains minus gross job
losses) in the last few years.
!The CPS is a survey of 60,000 households conducted by the U.S.
Bureau of the Census for the U.S. Bureau of Labor Statistics (BLS).
It is the representative survey of the U.S. population that yields the
unemployment rate. One of the rate’s components is the number of
employed persons in the U.S. civilian noninstitutional population
age 16 or older. Employment in the CPS is further disaggregated by
class of worker: wage and salary, self-employed, and unpaid family
workers. BLS’ Current Employment Statistics (CES) survey queries
160,000 U.S. businesses (excluding agriculture) and government
agencies, covering some 400,000 establishments, about the number
of payroll jobs. Unlike the CPS, then, the “establishment survey”
captures only wage and salary jobs in the nonfarm sector.
Because of these and other differences between the two surveys,2 some contend that
the CPS has better measured employment gains since the recession’s end in
1 Unless otherwise noted, this report will be confined to analysis of statistics pertaining to
the nonagricultural sector. Both total employment and self-employment (family farmers and
ranchers) in the agricultural sector declined substantially over time. In addition, data from
the establishment survey (discussed above) covers only nonfarm industries.
2 For more information see U.S. Bureau of Labor Statistics, Employment from the BLS
Household and Payroll Surveys: Summary of Recent Trends, available at
November 2001 and more accurately portrays conditions in the labor market as it has
evolved.3 Others counter that, regardless of the argument’s veracity, greater self-
employment is not a positive development because some individuals effectively were
forced into working for themselves as firms had created few new payroll jobs until
There also was disagreement during the two prior decades about the contribution
of self-employment or entrepreneurship — an individual starting a business — to net
job creation.5 The proposition, as originally espoused by David L. Birch, held that
births of small businesses account for a disproportionately large share of net new
jobs. Supporters of this viewpoint asserted that critics were mischaracterizing the
database and methodology that underlie it and were confusing its validity with other
concerns (e.g., the quality of the jobs at small firms).6 They further contended that
a study refuting the allegedly superior job generating role of new (as opposed to
expanding) small businesses7 could not be generalized to the nation’s largely service
sector economy because the analysis was limited to manufacturing. The study’s
findings subsequently were confirmed by an analysis of establishments across the
entire private sector:
entrepreneurship ... is important to job growth in the United States, but not as
important as job expansion in existing firms ... However, small establishments
(1 to 19 workers) play a much larger role in job growth in the [fast-growing]
services industry [e.g., health and business services] than they do in8
manufacturing industries in the United States.
Additionally, David Birch later amended his proposition to say that the precise
contribution of new entrepreneurs to job growth varies by time period and that most
3 Tim Kane, Diverging Employment Data: A Critical View of the Payroll Survey, The
Heritage Foundation, Mar. 4, 2004, available at [http://www.heritage.org/Research/
Labor/CDA04-03.cfm]. (Hereafter cited as Kane, Diverging Employment Data.)
4 From the recession’s end in Nov. 2001 to Aug. 2003, the CES showed further cuts in wage
and salary jobs. Between Aug. 2003 and Feb. 2004, the monthly job increase usually was
below 100,000. In Mar. 2004, however, payroll jobs jumped by some 300,000 and
continued at about that elevated level of gain to date.
5 Less than one in four businesses start with employees other than the owners, according to
Wells Fargo/NFIB, Business Starts and Stops, Nov. 1999. (Hereafter cited as Wells
Fargo/NFIB, Business Starts and Stops.)
6 David L. Birch, “Small Business (A Special Report): Statistics — The Hidden Economy:
While Big Companies Languish, Little Ones Are Creating Jobs and Fueling Demand,” Wall
Street Journal, June 10, 1988; John Case, “The Disciples of David Birch,” Inc., Jan. 1989;
and William J. Dennis, Bruce D. Phillips, and Edward Starr, “Small Business Job Creation:
The Findings and Their Critics,” Business Economics, vol. 29, no. 3, July 1994.
7 Steven J. Davis, John Haltiwanger, and Scott Schuh, “Small Business and Job Creation:
Dissecting the Myth and Reassessing the Facts,” Business Economics, vol. 29, no. 3, July
1994. See also Maria Shao, “A Question of Size: Economists Poke Holes in Notion that
Little Companies are Big Engines of Job Growth,” Boston Globe, June 25, 1995.
8 Robert W. Bednarzik, “The Role of Entrepreneurship in U.S. and European Job Growth,”
Monthly Labor Review, July 2000, p. 15.
of the job growth is fueled by fairly few members of the heterogenous small business
During the 1992-2000 economic expansion, interest in self-employment focused
on the kind of jobs being created, namely, alternative and contingent work
!The phrases, often used interchangeably, were coined to describe
firms’ increased utilization of individuals on an as-needed basis
rather than directly adding workers to their payrolls with an implicit
or explicit promise of job security. Variations on the traditional
work arrangement include bringing on board employees of
temporary help agencies and retaining self-employed consultants for
as long or short a time as necessary.
Some endorse the restructured business model because it purportedly enables
companies to respond more flexibly to changes in market conditions and to reduce
payroll taxes as well as employee benefit costs.10 They contend the self-employed
consultants or “free agents” are able, in turn, to have the independence that is
important to them. Others view with disfavor all alternative work arrangements
because they allegedly reduce workers’ job security and access to benefits (e.g.,
health insurance and pensions) and in some instances reflect misclassification of
employees as independent contractors who as a result are not eligible for certain
government benefits (e.g., Unemployment Insurance (UI), Workers Compensation,
and unpaid family/medical leave).11
The federal government, for its part, has encouraged self-employment as a
strategy for reemploying some displaced workers likely to exhaust their UI benefits.
Section 507 of the North American Free Trade Agreement (NAFTA) Implementation
Act of 1993 (P.L. 103-182) authorized states to establish a temporary Self-
Employment Assistance (SEA) program. The program was made permanent in the
Noncitizen Benefit Clarification and Technical Amendments Act of 1998 (P.L. 105-
9 David Wessel and Buck Brown, “The Hyping of Small-Firm Job Growth,” Wall Street
Journal, Nov. 8, 1988.
10 Randall L. Sarosdy and Ryan T. Shelton, “Arrangements with Consultants and
Independent Contractors,” presented at the International Law and Technology Forum
session on Employee Relations and Technology, Nov. 4, 2003, available at
11 John J. Sweeney, AFL-CIO President on the New GAO Report Shows Contingent Worker
Incomes and Benefits Lag, July 26, 2000; and Mike Hall, Curbing Corporate Greed:
Winning Full-Time Rights for Part-Time Workers, undated. Both available at
[ h t t p : / / www.af l c i o.or g] .
12 The following reports evaluate the SEA programs that operate in several states: William
T. Kosanovich and Heather Fleck, Final Report: Comprehensive Assessment of Self-
Employment Assistance Programs, prepared for the U.S. Department of Labor, June 2001,
available at [http://www.ows.doleta.gov/dmstree/op/op2k1/op_03-01.pdf]; and Wayne
commissioned by the U.S. Department of Labor (DOL) to examine the impact of
misclassification on UI tax payments and benefits, among other things. The study,
released in February 2000, noted that a rise in the unemployment rate — such as
occurred during and after the 2001 recession — could heighten the administrative
burden of state UI staff due to an increased number of contested claims filed by
independent contractors who thought they were employees.13
This report begins by defining and examining the characteristics of self-
employed persons. It next analyzes the trend in self-employment, its contribution to
net job growth, and why some individuals become entrepreneurs. The report then
explores a group closely related to the self-employed, namely, independent
contractors, and compares them with those in other alternative work arrangements
and in traditional jobs.
Who Are the Self-Employed?
The definition of self-employment in the CPS may be narrower than some
people’s conception of working for oneself. The monthly CPS asks people whether,
in their primary job, they currently are employed by government, a private firm, a
non-profit entity, or whether they are self-employed. Individuals who respond
affirmatively to the first three choices are classified as wage and salary workers.
Those who reply they are self-employed subsequently are asked if their business is
incorporated. If they say it is, then they too are classified as wage and salary workers
because they legally are the employees of their corporations. Thus, the CPS estimate
of self-employment applies to those whose primary job is owner of an unincorporated14
business from whose profits they derive their income. The unincorporated self-
employed may hire employees, but they generally do not do so.15
Confusion on the part of CPS respondents may have led to understatement of
the number of unincorporated self-employed in the last decade or so when
independent consultants seemingly became more prevalent.16 During the dot-com
bubble, for example, in-demand network administrators and software engineers may
have leapt from their wage and salary jobs to start working as freelancers. If they
subsequently contracted to work for their former employers, they or their family
members may have misidentified the job to the CPS interviewer as payroll
employment rather than self-employment.
Vroman, Self-Employment Assistance: Revised Report, prepared by The Urban Institute for
Department of Labor, Dec. 1997.
13 Planmatics, Inc., Independent Contractors: Prevalence and Implications for
Unemployment Insurance Programs, Feb. 2000, available at [http://www.ows.doleta.gov/
14 Self-employment is composed of sole proprietorships and partnerships, the latter of which
can include limited liability companies.
15 John E. Bregger, “Measuring Self-Employment in the United States,” Monthly Labor
Review, Jan./Feb. 1996. (Hereafter cited as Bregger, Measuring Self-Employment.)
16 Kane, Diverging Employment Data.
The following discussion first focuses on self-employment as defined in the
monthly CPS.17 Independent contractors are addressed in the report’s last section.
But, by way of a preview, the two overlapping groups have much in common.
As shown in Table 1, self-employment is much more common among men
(8.0%) than women (5.7%). The disparity has lessened over time. Women today
account for almost two of every five self-employed individuals principally because18
relatively more women than men have become business owners over the years. One
often suggested reason for women’s propensity toward self-employment is it gives
them control over work schedules, which facilitates accommodation of their
traditional role as family caregivers.19
Self-employment is increasingly common as people approach the end of their
working lives. Some 15% of workers age 65 or older were self-employed in 2003.
A desire to phase into retirement likely explains much of the group’s comparatively
high self-employment rate. Persons eligible for Social Security benefits might want
to work less than full-time schedules.20 Even were employers to allow them to work
part-time, older workers might be penalized for doing so if their pension plans base
benefits upon recent salary levels. Moreover, firms less often extend health
insurance coverage to part-time employees and when they do, often require a larger
employee contribution. Faced with these circumstances, members of the oldest age
group might respond by retiring from their payroll jobs but “keep their hand in” by
starting a business or by making what previously had been a side business their
A lesser but still well above-average share of 45-64 year olds also are self-
employed. The rate among mid-career workers probably is higher because they are
more likely than younger workers to have accumulated the capital and managerial
skills needed to become an entrepreneur.
17 An individual is classified as self-employed in the monthly CPS if that is the job held at
the time of the interview. In the March supplement to the CPS, an individual is classified
as self-employed if that is the longest job held during the preceding year. Consequently, the
self-employment rate derived from the monthly CPS will be higher than the rate derived
from the March supplement because the latter misses short spells of self-employment.
Studies of the composition of the self-employed based upon data from the March
supplement report findings similar to those discussed above. See for example Yannis
Georgellis and Howard J. Wall, “Who Are the Self-Employed?,” Federal Reserve Bank of
St. Louis Review, Nov./Dec. 2000.
18 Theresa J. Devine, “Characteristics of Self-Employed Women in the United States,”
Monthly Labor Review, Mar. 1994.
19 Karen V. Lombard, “Female Self-Employment and Demand for Flexible, Nonstandard
Work Schedules,” Economic Inquiry, vol. 39, no. 2, Apr. 2001.
20 Full-time employment is defined in the CPS as at least 35 hours a week.
Race and Ethnicity
White men have long had the highest self-employment rate (8.4% in 2003).
Blacks, regardless of gender, have had the lowest self-employment rate (4.0%).21
The number of 16-to 64-year-olds who owned their own unincorporated or
incorporated business (and who demonstrated a commitment to these businesses
based on number of hours worked) increased at a lesser rate among whites than non-
whites during the 1980s and 1990s. However, the self-employment ranking by
race/ethnicity was stable over the period.22 Most of the increase in the number of
self-employed men is attributable to expansion of the male labor force as opposed to
a rise in their self-employment rate — the opposite of the case among women.
Table 1. Self-Employment by Selected Characteristics, 2003
Number of self-employedPercent of
Characteristicworkers (in thousands)nonfarm workers
Total M en Women Total M en Women
Total, age 16 and older9,3445,7363,6096.98.05.7
16-19 85 59 25 1.5 2.1 0.8
20-24 253 169 84 1.9 2.4 1.3
25-34 1,513 909 605 5.0 5.6 4.4
35-44 2,545 1,520 1,025 7.4 8.3 6.4
45-54 2,579 1,597 981 8.2 9.8 6.5
55-64 1,709 1,063 646 10.5 12.6 8.3
65 and older66041824315.317.712.5
Race and ethnic origin
White 8,160 5,034 3,126 7.3 8.4 6.0
Black 593 358 235 4.0 5.3 3.0
Asian 390 222 168 6.8 7.3 6.3
21 For more information on differences in self-employment rates by race/ethnicity, see
Robert W. Fairlie and Bruce D. Meyer, “Ethnic and Racial Self-Employment Differences
and Possible Explanations,” Journal of Human Resources, vol. 31, no. 4, fall 1996.
22 Robert W. Fairlie, Recent Trends in Ethnic and Racial Self-Employment, University of
California — Santa Cruz and Joint Center for Poverty Research, Northwestern University
and University of Chicago, Nov. 2000.
Number of self-employedPercent of
Characteristicworkers (in thousands)nonfarm workers
Total M en Women Total M en Women
Construction 1,717 1,611 105 16.9 17.6 10.8
Manufacturing 325 216 109 1.9 1.8 2.1
Information 152 101 51 4.1 4.9 3.2
Professional and business1,9081,17273613.714.812.4
Education and health1,1382858544.04.14.0
Leisure and hospitality68639129220.127.116.11
Other services (excluding1,07155052117.716.918.6
and financial operations
Professional and related1,7249597618.104.22.168
Sales and office1,9139249822.214.171.124
Production, 684 545 139 3.8 4.0 3.3
Source: U.S. Bureau of Labor Statistics’ data from the Current Population Survey.
As shown in Table 1, the self-employment rate is highest in the following
industry groups: construction; professional and business services, such as legal and
accounting; and other services (excluding private household) such as repair and
maintenance as well as personal and laundry. Part of the reason for these higher rates
of self-employment may be their fairly low financial barriers to entry.23
A wide gap exists between the earnings of self-employed men and women.
Self-employed men’s higher earnings may be due, in part, to their presence in more
financially rewarding industries (e.g., construction and professional specialties).
Women, in contrast, are more prevalent in relatively low-paying retail trade and
services industries. Women’s customary activities outside the labor market — time
spent in housework and childrearing — also adversely affect the relative wages of
female business owners.24
The highest rates of self-employment are in natural resources, construction, and
maintenance occupations (13.0%)25 and in management, business, and financial
operations jobs (8.6%). The lowest self-employment rates are in production,
transportation, and material moving occupations (3.8%). Only in service occupations
(e.g., child care workers and home health aides) is the share of women business
owners appreciably greater than the share of men (8.9% and 5.8%, respectively).
BLS projects self-employment in management, business, and financial
operations jobs could expand. The number of self-employed persons in professional
and related occupations also is expected to increase (e.g., computer occupations and
registered nurses), but there might be fewer self-employed doctors “as employment
shifts into incorporated group practices” and fewer lawyers, “reflecting the difficulty
in establishing new legal practices.”26 Similarly, total self-employment is projected
to increase in service and construction occupations despite decreases in certain fields.
In contrast, self-employment overall among sales (e.g., retail store owners and sales
workers) and production (e.g., apparel and wood workers) occupations could decline.
23 Ellen R. Rissman, Self-Employment as an Alternative to Unemployment, Federal Reserve
Bank of Chicago, Working Paper 2003-34, Dec. 12, 2003. (Hereafter cited as Rissman, Self-
Employment as an Alternative to Unemployment.)
24 Greg Hundley, “Why Women Earn Less Than Men in Self-Employment,” Journal of
Labor Research, vol. 22, no. 4, fall 2001.
25 Included within this occupation are farmers and ranchers.
26 Daniel E. Hecker, “Occupational Employment Projections to 2012,” Monthly Labor
Review, Feb. 2004, pp. 97 and 98.
Self-Employment and Job Growth Over Time
Never a large presence in the postwar period, the share of self-employed
nonfarm workers fell steadily from 12% in 1948 to under 7% thus far in the current
decade. While the ranks of the self-employed have expanded by 53% over the 55-
year period shown in Table 2, total employment across all nonfarm industries grew
more than three times as much (by 167%). In other words, members of the labor
force increasingly have become employees in someone else’s or their own
Table 2. The Trend in Self-Employment, Selected Years
People employed inSelf-employment in nonfarm industries
Yearnonfarm industriesNumber (in
(in thousands)thousands)Percent of total
1948 50,714 6,110 12.0
1959 59,065 6,222 10.5
1969 a 74,296 5,252 7.1
1979 95,477 6,791 7.1
1989 114,142 8,605 7.5
1999 b 130,207 8,790 6.8
2000c 134,427 9,205 6.8
2001 134,635 9,121 6.8
2002 134,174 8,923 6.7
2003d 135,461 9,344 6.9
Source: U.S. Bureau of Labor Statistics’ data from the Current Population Survey.
a. From 1967 onward, the self-employed include only owners of unincorporated businesses who thus
derive their income from profits rather than salaries.
b. Although there was a major redesign of the CPS in 1994, it had a fairly small effect on self-
c. Data for 2000 are not comparable with earlier years because of the incorporation of population
estimates from the decennial census, which greatly increased employment estimates for the year.
d. An upward population adjustment made in 2003, which was based upon higher estimates of net
international migration from 2000 through 2002, resulted in a much smaller increase in
employment estimates in 2003 compared to the 2000 adjustment.
27 The CPS was reworded in the mid-1960s to explicitly count as self-employed only those
who own unincorporated businesses, and U.S. businesses have exhibited a marked tendency
over time toward incorporation. Accordingly, about the same decline in the self-
employment rate likely would have occurred but without the sharp single-year drop between
Self-Employment as a Contributor to Job Growth
As shown in the above table, the contribution of self-employment to net job
growth has not varied greatly from one decade to another — with the exception of
the 1990s. During the 1970s, measuring approximately peak-to-peak in the business
cycle, the addition of 1.5 million self-employed persons accounted for about 7% of
net new jobs. The 1.8 million increase in self-employed persons during the 1980s
comprised a larger share of job growth, at almost 10%. During the 1990s, the very
limited net increase in self-employment (185,000) contributed just 1% to the
expansion in nonfarm employment.
While the absolute number of self-employed workers had climbed steadily over
the 1970s and 1980s, their number remained fairly flat (at or approaching 9 million)
during the long economic expansion of the 1990s. Researchers who study self-
employment cannot explain the aberration.28 As the labor market tightened and
health insurance costs rose over the course of the decade, companies may have
offered such things as flexible work schedules, telecommuting, and benefits “to lure
free agents onto payrolls.”29 Between 1991 and 1997, for example, the proportion of
full-time wage and salary workers with flexible schedules almost doubled — rising
from 15% to 28%.30 The robust economic times also may have dampened business
starts because fewer people worried about losing or actually lost their payroll jobs.31
For example, in 1998 when the national unemployment rate was a low 4.5%, only
14.5% of people who started businesses said they did so because they had been laid
off or feared that might happen.32
Self-employment thus far into the current decade has been a much more
substantial contributor to net job growth than during the 1990s, and exceeds its
shares during the 1970s and 1980s. The net addition of 139,000 self-employed
persons from 2000 to 2003 accounted for over 13% of nonfarm job gains.
Reasons for Entering Self-Employment from Unemployment
Arguably, some workers who lose wage and salary jobs when few are being
created — such as during periodic nationwide, areawide, or industrywide slowdowns
or during prolonged geographic or industrial declines — become discouraged about
getting hired by companies and try to earn money by starting their own firms. In the
case of cyclical events, this might have occurred after the last two national
recessions. Businesses did not add workers to their payrolls for such a long time
28 David Leonhardt, “Self-Employment on the Decline,” New York Times, Dec. 1, 2000.
29 Kim Clark, “The Myth of the ‘Free-Agent Nation’,” Fortune, June 8, 1998, p. 40. See
also Gene Koretz, “What’s with the Self-Employed?: They’re Rejoining Corporate Ranks,”
Business Week, May 11, 1998.
30 Thomas M. Beers, “Flexible Schedules and Shift Work: Replacing the ‘9-to-5’ Workday,”
Monthly Labor Review, June 2000.
31 Joshua Harris Prager, “Start-Ups Drop Despite Strong Economy — Entrepreneurship Ebbs
as Job Opportunities Abound,” Wall Street Journal, Jan. 28, 1999.
32 Wells Fargo/NFIB, Business Starts and Stops.
following the 2001 and 1990-1991 recessions that their aftermath commonly became
known as jobless recoveries.33 For example, an analysis of the New England
economy found that during and after the 1990-1991 recession, employment in the
region contracted. Household survey data showed a somewhat smaller decrease in
employment than establishment survey data, and the only group to report increased
employment was the self-employed.34
A recent empirical study of U.S. men aged 21 and older in the 1979-1998 period
concluded that a dearth of wage and salary job opportunities “pushes workers into
self-employment.”35 Another analysis estimated that the unincorporated self-
employed in particular are sensitive, with a lag, to local economic conditions: the
unincorporated, presumably because incorporation can be costly and would be less
attractive to persons who regard working for oneself as a way station between
salaried jobs; with a lag, because it takes time to start a business. The economists’
model explained almost the entire increase in self-employment that occurred since
the 2001 recession began. They speculated that “many of these new businesses are
likely to disappear when the wage and salary sector improves.”36 There is an ongoing
debate among researchers, however, about “whether high unemployment acts to
discourage self-employment because of the lack of available opportunities or
encourage it because of the lack of viable alternatives.”37
Demographic and economic factors beyond current labor market conditions
affect the propensity of job losers, among others, to become entrepreneurs. A
displaced worker’s age and prior household wealth were estimated to increase the
probability of entering self-employment.38 Other studies found that among women,
the likelihood of self-employment increased if there were young children in the
family and if there was a self-employed spouse who could serve as a role model and
who could provide capital for his wife’s business venture.39 In addition, the
independence and higher anticipated earnings of self-employment are job
characteristics that appeal to some individuals.40
33 For more information on the current and prior recoveries based upon CES data see CRS
Report RL32047, The “Jobless Recovery” From the 2001 Recession: A Comparison to
Earlier Recoveries and Possible Explanations, by Marc Labonte and Linda Levine.
34 Katharine L. Bradbury, “New England Job Changes during the Recession: The Role of
Self-Employment,” New England Economic Review, Sept./Oct. 1994.
35 Rissman, Self-Employment as an Alternative to Unemployment, p. 19.
36 Daniel Aaronson, Ellen R. Rissman, and Daniel G. Sullivan, “Assessing the Jobless
Recovery,” Economic Perspectives, 2nd quarter, 2004.
37 David G. Blanchflower, Self-Employment: More May Not Be Better, Working Paper
38 Rissman, Self-Employment as an Alternative to Unemployment.
39 Richard K. Caputo and Arthur Dolinsky, “Women’s Choice to Pursue Self-Employment:
The Role of Financial and Human Capital of Household Members,” Journal of Small
Business Management, vol. 36, no. 3, July 1998.
40 Mark P. Taylor, “Earnings, Independence or Unemployment: Why Become Self-
Employed?,” Oxford Bulletin of Economics and Statistics, vol. 58, no. 2, May 1996.
Self-Employment as an
Alternative Work Arrangement
An examination of independent contractors might shed additional light on the
self-employed particularly if, as previously mentioned, there may be some
misidentification in the monthly CPS of the self-employed as wage and salary
workers. Additionally, the definition of independent contractor that BLS developed
may accord more with the current perception of what it means to be self-employed
(e.g., computer consultants and freelance writers as opposed to retail store and
From February 1995 to February 2001, BLS collected statistics biennially on
alternative work arrangements through questions that supplemented the basic CPS
questionnaire. Independent contractors, also known as independent consultants and
freelancers, obtain customers on their own to whom they provide a good or service.
The designation of self-employed or wage and salary worker according to the basic
CPS questionnaire is irrelevant in the determination of who is in an alternative work
arrangement.41 Thus, almost 9 out of 10 independent consultants are (unincorporated
and incorporated) self-employed while the remainder are wage and salary workers.42
There were almost 8.6 million independent contractors who accounted for 6.4%
of total (farm and nonfarm) employment in February 2001. This does not differ
much from the 9.1 million unincorporated business owners who accounted for 6.8%
of nonfarm employment for all of 2001 as shown in Table 2, which is not surprising
because about half of all self-employed persons (incorporated and unincorporated,
farm and nonfarm) said they were independent consultants.
Since the supplementary questions were first asked, independent consultants
have been the largest of the four nontraditional worker groups, most recently
comprising a little under 70% of the total. The other groups are temporary help
agency workers, on-call workers, and workers provided by contract firms.43
Traditional jobs are any that do not fall into the four other categories.
41 All individuals who respond affirmatively to the basic CPS question about self-
employment, regardless of whether they are unincorporated or incorporated, are asked
whether they consider themselves to be self-employed as independent contractors,
independent consultants, or freelance workers in order to distinguish them from those who
operate establishments (e.g., retail stores). Individuals who respond affirmatively to the
basic CPS question about wage and salary employment also are asked whether they very
recently had been working as independent contractors, independent consultants, or freelance
42 BLS, Contingent and Alternative Employment Arrangements, February 2001, May 24,
43 Ibid. Note: On-call workers, typically drawn from a pool of qualified individuals, are
called in to work only when needed (e.g., substitute teachers and construction workers
supplied by a union hiring hall). Employees of contract companies typically are provided
to work at just one firm and to work on that firm’s premises. For information on temporary
help agency workers, see CRS Report RL30072, Temporary Workers as Members of the
Contingent Labor Force, by Linda Levine.
Worker Preferences and Job Insecurity
Unlike others in nontraditional jobs, most independent contractors — more than
four out of five — have said they prefer to be their own boss. The great majority —
at least three out of four — give personal reasons for wanting to be freelancers (e.g.,
flexible schedules and family obligations) rather than economic reasons (e.g., only
type of work available).44
The labor market prospects of independent contractors seemingly temper their
preference for the work arrangement.45 Thus, compared to the group overall, fewer
of those who recently had been laid off preferred freelancing. Nonetheless, a
majority (66.2%) of persons who lost jobs in the three years before being reemployed
as consultants in February 1995 preferred their alternative to traditional employment.
The high rate of satisfaction among those who conceivably might have
involuntarily become independent contractors through downsizing or some other
job loss [following the 1990-1991 recession] is consistent with anecdotal
evidence which indicates that, while these workers may not have been happy to
have lost their previous wage and salary positions, they still would prefer not to
return to a situation in which they would be a company employee. Instead, these
individuals may find that being an independent contractor affords them more job46
security than being someone else’s employee does.
Indeed, fairly few independent consultants consider themselves to be in an
insecure or contingent work arrangement (i.e., one that is not expected to be
longlasting). When the economy was expanding vigorously toward the end of the
1990s and firms were competing intensely to hire employees, the already small share
of freelancers who believed their jobs were insecure shrank further: 3.8% in 1995,
3.5% in 1997, and 2.9% in 1999.47 As shown in Table 3, the incidence of job
insecurity among independent contractors also was substantially below that of other
44 Sharon R. Cohany, “Workers in Alternative Employment Arrangements,” Monthly Labor
Review, Oct. 1996 (Hereafter cited as Cohany, Workers in Alternative Employment
Arrangements.); Sharon R. Cohany, “Workers in Alternative Employment Arrangements:
A Second Look,” Monthly Labor Review, Nov. 1998 (Hereafter cited as Cohany, Workers
in Alternative Employment Arrangements: A Second Look.); and Marisa DiNatale,
“Characteristics of and Preferences for Alternative Work Arrangements, 1999,” Monthly
Labor Review, Mar. 2001 (Hereafter cited as DiNatale, Characteristics of and Preferences
for Alternative Work Arrangements.)
45 Janet H. Marler and George T. Milkovich, Determinants of Preference for Contingent
Employment, Cornell University, Center for Advanced Human Resources Studies, Working
Paper 00-03. (Hereafter cited as Marler and Milkovich, Determinants of Preference for
46 Anne E. Polivka, “Into Contingent and Alternative Employment: By Choice?,” Monthly
Labor Review, Oct. 1996, pp. 64-65. See also Gideon Kunda, Stephen R. Barley, and James
Evans, “Why Do Contractors Contract? The Experience of Highly Skilled Technical
Professionals in a Contingent Labor Market,” Industrial and Labor Relations Review, vol.
47 See also Charles F. Manski and John D. Straub, “Worker Perceptions of Job Security in
the Mid-1990s: Evidence from the Survey of Economic Expectations,” Journal of Human
Resources, vol. 35, no. 3, summer 2000.
workers in alternative employment and was virtually equal to that of traditional
workers. In February 2001 — just one month before the official start of the latest
recession — the rate of job insecurity for consultants rose sharply to (a still low)
4.1%. The increase for the group as a whole may reflect a decrease in satisfaction
with self-employment reported by one segment, namely, independent contractors in
wage and salary jobs.48 In contrast, job insecurity among those with traditional jobs
fell slightly, as it did to varying degrees among persons in other alternative work
Table 3. The Job Insecurity Rate, February 1995-2001
Percent who felt insecure in their work
arrangement in February
Temporary help agency workers66.556.855.955.4
Contract company workers19.816.720.217.1
Source: U.S. Bureau of Labor Statistics’ data from the Alternative Work Arrangement supplement
to the Current Population Survey.
During the tightening labor market that encompassed the February 1995 through
February 2001 survey periods, most independent consultants appear to have had good
reason to perceive their situations as secure. They reported both the longest median
tenure of workers in alternative arrangements and longer tenure than workers in
traditional arrangements. In February 2001, for example, the typical freelancer had
been his/her own boss for 8.1 years while the typical traditional worker had held
his/her job for 4.4 years. As in the past, the tenure gap was wider between male
contractors (10.1 years) and male traditional workers (4.8 years) compared to female
contractors (6.0 years) and female traditional workers (4.0 years), reflecting the older
profile of male contractors.49
The relative pay of workers in alternative jobs is estimated to positively affect
the preference for nontraditional employment.50 The median weekly earnings of
independent consultants employed full-time in 2001 was $644 — well above workers
48 Jeffrey Wenger, Share of Workers in ‘Nonstandard’ Jobs Declines, Economic Policy
Institute, undated. Note: Although BLS also included these individuals in the independent
contractor group, it did not report separate findings for the relatively small subset.
(Hereafter cited as Wenger, Share of Workers in ‘Nonstandard’ Jobs Declines.)
49 Unpublished data from the Feb. 2001 CPS supplement on alternative work arrangements.
50 Marler and Milkovich, Determinants of Preference for Contingent Employment.
in all other alternative arrangements except employees of firms that provide on-site
services to other companies ($790).51
Earlier surveys show independent contractors had higher median weekly
earnings than workers in traditional employment arrangements. In 1999, the latest
period for which such a comparison can be made, freelancers employed full-time had
median weekly earnings of $640, $100 or 19% greater than full-time workers in
These earnings disparities are associated with individual and job characteristics.
Compared to traditional workers, independent contractors — like the unincorporated
self-employed previously described — generally are older. They also more often are
white, males, and college graduates who typically put in long hours. In addition,
some occupations in which consultants are disproportionately employed are relatively
high-paying (e.g., executive, administrative, and managerial positions). While the
typical male freelancer employed full-time earned more than males in traditional jobs
during the second half of the 1990s, the typical female freelancer earned less than her
counterpart in a traditional job.53 An analysis comparing the hourly wage of
“nonstandard” workers and regular full-time workers and taking into account these
personal and job characteristics estimated that all self-employed independent
contractors, female independent contractors in wage and salary jobs, and other male
self-employed (e.g., store owners) earned a statistically significant wage premium;
in contrast, other self-employed women suffered a statistically significant wage
penalty compared to regular full-time working women.54
51 BLS, Contingent and Alternative Employment Arrangements.
52 DiNatale, Characteristics of and Preferences for Alternative Work Arrangements. Note:
The direction and extent of earnings differences appears to be sensitive to the definition of
self-employment and to the measure of earnings. For example, an analysis based upon
Survey of Income and Program Participation from the mid-1980s found that the typical
unincorporated and incorporated male small business owner aged 18-65 (excluding farmers
and highly paid professionals such as doctors and lawyers) earned less than the typical paid
employee. The width of the pay gap depended upon the measure of earnings utilized. See
Barton H. Hamilton, “Does Entrepreneurship Pay? An Empirical Analysis of the Returns
to Self-Employment,” Journal of Political Economy, vol. 108, no. 3, 2000.
53 Cohany, Workers in Alternative Employment Arrangements: A Second Look; DiNatale,
Characteristics of and Preferences for Alternative Work Arrangements; and Steven Hipple
and Jay Stewart, “Earnings and Benefits of Workers in Alternative Work Arrangements,”
Monthly Labor Review, Oct. 1996.
54 Wenger, Share of Workers in ‘Nonstandard’ Jobs Declines. Note: This analysis of the
alternative worker supplement to the CPS categorized the groups differently than BLS.
Regular full-time workers were defined as wage and salary workers employed at least 35
hours a week and not included in other arrangements called “nonstandard” (e.g., regular
part-time workers, whom BLS does not consider to be in alternative arrangements, are
defined as wage and salary workers employed fewer than 35 hours a week and not included
in the other nonstandard arrangements; temporary help agency workers; on-call workers; day
laborers; and contract company workers whose definition, unlike that of BLS, does not
require employment at the client firm’s workplace). The analysis includes as nonstandard
those persons who reported themselves as self-employed, including shop or restaurant
owners. It further distinguishes between independent contractors who said they very
More independent contractors than those in traditional work arrangements are
employed part-time (24.8% and 16.8%, respectively, in 2001). This too may be
related to demographic and job characteristics. The share of freelancers at least 65
years old has been about three times the share of traditional workers (most recently
7.9% and 2.6%, respectively).55 As previously noted, members of this age group
could choose to bridge their full-time salaried careers and their nonwork years by
consulting for a few hours a week. In addition, female independent contractors are
more likely than female traditional employees to be raising preschool children, which
may partly explain why over three times the share of adult female than male
freelancers are employed part-time by choice.56 Relatively more adult male
consultants than male traditional workers are employed part-time for economic
reasons (i.e., slack work, poor business conditions or unable to find full-time jobs),
which may be partly linked to the former’s greater presence in industries with work
weeks over which they have little control (e.g., agriculture and construction).57
As with earnings, workers provided by contract companies have the highest rate
of health insurance coverage (80.1%) among persons in alternative work
arrangements. Independent contractors again come in second, with 72.5% having
health benefits in 2001. This contrasts with the larger percentage of traditional
workers who have health care coverage (83.1%).58 Freelancers may obtain health
benefits by independently purchasing policies or through a spouse’s employer-
provided plan. (Indeed, it appears that the availability of health insurance coverage
through a spouse is positively associated with becoming self-employed.)59 Other
consultants may have previously worked for companies that extend health benefits
The situation is quite different when it comes to retirement benefits. As might
be expected, just 2.3% of independent contractors had pension coverage through their
employers compared to almost 58.3% of traditional workers in 2001.60 Some 38.4%
of freelancers had established Individual Retirement Accounts (IRAs) or Keogh
plans. About three-fourths of independent consultants were employed full-time in
recently had been on a firm’s payroll (i.e., independent contracting — wage and salary) and
independent contractors who said they were in the arrangement at the time of the survey
(i.e., independent contracting — self-employed).
55 BLS, Contingent and Alternative Employment Arrangements.
56 DiNatale, Characteristics of and Preferences for Alternative Work Arrangements.
57 For more information on part-time employment see CRS Report 98-695, Part-Time Job
Growth and the Labor Effects of Policy Responses, by Linda Levine.
58 BLS, Contingent and Alternative Employment Arrangements.
59 Alison J. Wellington, “Health Insurance Coverage and Entrepreneurship,” Contemporary
Economic Policy, vol. 19, no. 4, Oct. 2001.
60 Unpublished data from the Feb. 2001 CPS supplement on alternative work arrangements
and BLS, Contingent and Alternative Employment Arrangements.
2001, and a minority of these individuals had pension coverage through their
employers (2.5%) or through IRAs and Keogh plans (38.1%).
Misclassification of Independent Contractors
Whether individuals are considered to be employees of a business or working
for themselves has important consequences under employment, labor, and tax laws.61
For example, as a result of an Internal Revenue Service (IRS) audit of Microsoft in
1989 and 1990, the corporation had to pay withholding taxes on the salaries of
freelancers whom the IRS determined to be common law employees.62
Independent contractors who worked for Microsoft between 1987 and 1990 then
brought suit against the firm, attesting that as employees they should have been
eligible for benefits (e.g., group health and life insurance, vacation and sick leave,
defined contribution (401(k)) pension plan, and stock purchase plan). The district
court denied the plaintiffs’ claims. On appeal of the decision concerning eligibility
only for the 401(k) and stock option plans, the U.S. Court of Appeals for the Ninth
Circuit held in Vizcaino v. Microsoft Corporation, 97 F.3d 1187 (9th Cir. 1996) that
the contractors were common law employees and therefore entitled to participate in
their employer’s plans.63 That decision was vacated and the case was reheard by the
full court of appeals. On the rehearing en banc, the court affirmed the panel decision
that the workers were common law employees, and that Microsoft’s reasons for
rejecting the workers’ participation in the two plans were invalid. However, the court
remanded the decision about whether the workers were covered by the 401(k) plan
to the plan administrator for determination in accordance with the terms of the plan,
and remanded the decision as to the appropriate remedy for excluding the workers
from the stock pension plan to the district court. Vizcaino v. Microsoft Corporation,
Ultimately, after several further appeals and remands, the parties reached a
court-approved settlement. The result leaves the question of whether misclassified
employees are covered under an employer’s benefit plans subject, in part, to the
plan’s definition of coverage and to the outcome of litigation as other courts have
reached different conclusions.64
61 For information on “class of worker” status see CRS Electronic Briefing Book Taxation,
page on “Worker Classification Issues: Employees and Independent Contractors,”by Marie
B. Morris at [http://www.congress.gov/brbk/html/ebtxr34.html].
62 Charles J. Muhl, “What Is an Employee? The Answer Depends on the Federal Law,”
Monthly Labor Review, Jan. 2002.
63 Susan L. Cosky, “Vizcaino v. Microsoft Corporation; A Labor and Employment Lawyer’s
Perspective,” Labor Law Journal, Feb. 1997.
64 Daniel M. Feinberg, “Independent Contractors, Leased Employees, and Other Contingent
Workers,” Practical Lawyer, vol. 47, no. 2, Mar. 2001.