Order Code RL32553
CRS Report for Congress
Received through the CRS Web
Union Membership Trends in the United States
August 31, 2004
Gerald Mayer
Economic Analyst
Domestic Social Policy Division
Congressional Research Service ˜ The Library of Congress

---

Union Membership Trends in the United States
Summary
Union membership in the United States has declined significantly in recent
decades. The number of union members peaked in 1979 at an estimated 21.0 million.
In 2003, an estimated 15.8 million workers were union members. As a percent of
employed workers, union membership peaked in 1954 at 28.3%. In 2003, 11.5% of
employed workers were union members.
Most studies find that, after controlling for individual, job, and labor market
characteristics, the wages of union workers are in the range of 10% to 30% higher
than the wages of nonunion workers. The wage premium is generally greater for less
skilled, less-educated, and younger workers and larger for private than public sector
workers. Union members generally receive better or more generous fringe benefits
than similar nonunion workers. Job tenure tends to be greater and quit rates lower
among unionized workers. However, the wage premium may have declined in recent
years.
Data from the Bureau of Labor Statistics’ (BLS) monthly Current Population
Survey (CPS) show that the level of union membership varies among different
groups. Union members are more likely to be male, white, middle-age, work in the
private sector, and have a high school degree or some college. The rate of union
membership is greater among men than women and higher among older than younger
workers. In 2003, 12.3% of men were union members, compared to 10.5% of
women; 14.7% of workers ages 45 to 64 were union members, compared to 5.0% of
workers ages 16 to 24 and 11.3% of workers ages 25 to 44.
Although the level of union membership is greater among white than black
workers, in 2003 15.6% of black workers were union members, compared to 11.0%
of white workers. Also, although union members are more likely to be employed in
the private than public sector, in 2003, 37.2% of public sector employees were union
members, compared to 7.2% of private sector employees.
In 2003, 12.6% of workers with a bachelor’s or advanced college degree were
union members, compared to 6.6% of workers with less than a high school education
and 11.9% of workers with a high school degree or one to three years of college. In
2003, almost three-fourths (73.6%) of union workers with a bachelor’s or advanced
degree worked in the public sector, mostly for state and local governments. The
largest percentage of these employees (43.6%) were teachers.
In 2003, unionization was greatest in New York, Hawaii, Michigan, Alaska,
New Jersey, and Washington. Unionization was lowest in North Carolina, South
Carolina, Arkansas, Mississippi, Arizona, and South Dakota.
Finally, in 2002, the most unionized occupations were precision production
workers and operators (18.3% and 17.6%, respectively). The most unionized
industries were public administration (32.3%) and transportation, communications,
and utilities (27.4%). This report will be updated periodically.

---

Contents
Major Federal Collective Bargaining Laws . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Legislation in the 108th Congress . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Governments and Collective Bargaining . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Government Intervention in Labor Markets . . . . . . . . . . . . . . . . . . . . . . . . . 4
Distribution of Earnings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
Collective Voice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Economic Effects of Labor Unions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Earnings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Private and Public Sectors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Gender . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Race . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Job Tenure and Quit Rates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Productivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Profits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Trends in Union Membership . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Characteristics of Union Membership . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Gender . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Age . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Race . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
Hispanic Origin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
Educational Attainment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
Private and Public Sectors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
Level of Government . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
Industry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Occupation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Region . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
Appendix A: Annual Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
Appendix B: Data and Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
Confidence Levels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35

---

List of Figures
Figure 1. Union Membership as a Percent of Employment, 1930-2003 . . . . . . . 11
Figure 2. Union Membership Rates of Men and Women, 1994-2003 . . . . . . . . 13
Figure 3. Union Membership Rates by Age, 1994-2003 . . . . . . . . . . . . . . . . . . 13
Figure 4. Union Membership Rates by Race, 1994-2003 . . . . . . . . . . . . . . . . . . 14
Figure 5. Union Membership Rates by Hispanic Origin, 1994-2003 . . . . . . . . . 15
Figure 6. Union Membership Rates by Level of Education, 1994-2003 . . . . . . 15
Figure 7. Union Membership Rates in the Public and Private Sectors, 2003 . . . 16
Figure 8. Union Membership Rates by Level of Government, 2003 . . . . . . . . . 18
Figure 9. Union Membership Rates by Industry, 2002 . . . . . . . . . . . . . . . . . . . . 19
Figure 10. Union Membership Rates by Occupation, 2002 . . . . . . . . . . . . . . . . 20
Figure 11. Union Membership Rates by Region, 2003 . . . . . . . . . . . . . . . . . . . . 20
Figure 12. Union Membership Rates, by State, 2003 . . . . . . . . . . . . . . . . . . . . . 21
List of Tables
Table A1. Union Membership in the United States, 1930-2003 . . . . . . . . . . . . . 22
Table A2. Union Membership in the United States by Gender, 1994-2003 . . . . 24
Table A3. Union Membership in the United States by Age, 1994-2003 . . . . . . 25
Table A4. Union Membership in the United States by Race, 1994-2003 . . . . . . 26
Table A5. Union Membership in the United States by Hispanic Origin,
1994-2003 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
Table A6. Union Membership in the United States by Educational
Attainment, 1994-2003 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
Table A7. Union Membership in the United States in the Private and
Public Sectors, 1994-2003 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
Table A8. Union Membership in the United States by Level of Government,
1994-2003 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
Table A9. Union Membership in the United States by Industry, 1994-2002 . . . 31
Table A10. Union Membership in the United States by Occupation,
1994-2002 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
Table A11. Union Membership in the United States by Region, 1994-2003 . . . 33

---

Union Membership Trends in
the United States
Many factors affect the level and distribution of employment and earnings.
Individuals with more education, work experience, and job training generally earn
more. Savings and investment and technological advances can affect labor
productivity and real earnings. Changes in consumer tastes can influence the demand
for workers with different skills. Employment and earnings may also be affected by
fiscal and monetary policy and by institutional factors. Institutional factors include
government regulation of industry, immigration and trade policy, and labor unions.
This report summarizes the major federal laws that give certain protections to
employees who organize and bargain collectively. The report reviews the economic
effects of labor unions and examines recent trends in union membership in the United
States.
Major Federal Collective Bargaining Laws
The National Labor Relations Act of 1935 (NLRA) is the basic law governing
relations between labor unions and private sector employers engaged in interstate
commerce. The act does not cover supervisors and managers, agricultural laborers,
domestic servants, and others.1 Separate federal laws apply to railroads, airlines, and
federal employees. The NLRA is administered by the National Labor Relations
Board (NLRB).
The NLRA requires an employer to bargain with the representative selected by
a majority of the firm’s employees. The act does not require secret-ballot elections.
If a majority of employees indicate a desire to be represented by a union, an employer
may voluntarily enter into collective bargaining. If an employer does not voluntarily
recognize the union chosen by a majority of employees, a petition can be filed with
1 Agricultural laborers include crop and livestock workers and farmworkers who perform
work that is incidental (e.g., sorters and packers) to the production of goods on the
employer’s farm. National Labor Relations Board, Basic Guide to the National Labor
Relations Act (Washington: GPO, 1997), p. 28, available at [http://www.nlrb.gov].
(Hereafter cited as NLRB, Basic Guide to the NLRA..) Commerce Clearing House, Labor
Relations, vol. 1 (Chicago: Commerce Clearing House, 2004), pp. 4168-4169. United States
General Accounting Office, Collective Bargaining Rights: Information on the Number of
Workers With and Without Bargaining Rights, Report No. GAO-02-835, Sept. 2002, pp. 12-
13. (Hereafter cited as GAO, Collective Bargaining Rights.) The GAO is now called the
Government Accountability Office.

---

CRS-2
the NLRB for a secret-ballot election. A petition may be filed by a union, a group of
employees, or the employer.2
The Railway Labor Act of 1926 (RLA) gives railroad and airline employees the
right to unionize. The act allows some supervisors (i.e., “subordinate officials”) to
be union members.3
Title VII of the Civil Service Reform Act of 1978 (CSRA) provides collective
bargaining rights to federal employees. The law applies to executive branch
agencies, the Library of Congress, and the Government Printing Office. The law
excludes supervisors, members of the armed services, and various agencies.4
According to a 2002 report by the Government Accountability Office (GAO),
26 states and the District of Columbia have laws that provide collective bargaining
rights to public employees. An additional 12 states have laws that give bargaining
rights to specific groups of employees (e.g., teachers, firefighters, or state workers).
Nine states provide bargaining rights to agricultural workers. Some state laws allow
supervisors to be union members.5
2 In order to have a secret-ballot election, it is not necessary for a majority of employees to
sign a petition or authorization cards (i.e., cards authorizing a union to represent them for
the purposes of collective bargaining). Employees may petition the NLRB for union
representation if at least 30% of employees express a desire for union representation.
NLRB, Basic Guide to the National Labor Relations Act, pp. 7-8. National Labor Relations
Board, The NLRB: What it is, What it Does, National Labor Relations Board, p. 3, available
at [http://www.nlrb.gov]. Workers may organize without the protections of the NLRA, but
the employer would not be required to bargain.
3 Douglas L. Leslie (editor in chief), The Railway Labor Act (Washington: BNA Books,
1995), pp. 118-119, 424, 428.
4 The CSRA excludes from coverage Foreign Service employees, the Federal Bureau of
Investigation, Central Intelligence Agency, Government Accountability Office, National
Security Agency, Tennessee Valley Authority, the Federal Services Impasses Panel, and the
Federal Labor Relations Authority. The CSRA also gives the President the authority to
exclude, in the interests of national security, any agency whose primary function involves
investigative, intelligence, counterintelligence, or security work. 5 U.S.C. § 7103. CRS
Report RL30795, General Management Laws: A Compendium, coordinated by Clinton T.
Brass, pp. 325-326.
5 Local governments may have laws giving local public employees collective bargaining
rights. GAO, Collective Bargaining Rights, pp. 8-9. Lloyd G. Reynolds, Stanley H.
Masters, and Colletta H. Moser, Labor Economics and Labor Relations, 11th ed.
(Englewood Cliffs, N.J.: Prentice-Hall, 1998), p. 460. (Hereafter cited as Reynolds et al.,
Labor Economics and Labor Relations.)

---

CRS-3
Legislation in the 108th Congress
Legislation has been introduced in the 108th Congress that, if enacted, may affect
union membership in both the private and public sectors.
S. 606, the “Public Safety Employer-Employee Cooperation Act of 2003” would
provide collective bargaining rights to public safety workers (i.e., law enforcement
officers, firefighters, and emergency medical services personnel) employed by state
or local governments. The bill was introduced by Senator Judd Gregg and was
approved by the Senate Committee on Health, Education, Labor and Pensions on
October 2, 2003. A similar proposal, H.R. 814, was introduced in the House by
Representative Dale Kildee. No congressional action has been taken on the latter
bill.
Legislation has been introduced that would allow employees to unionize if a
majority of employees sign authorization cards. A secret-ballot election would not
be required. This proposal is included, with other provisions, in S. 1513 and H.R.
3078, the “Employee Right to Choose Act of 2003,” and in S. 1925 and H.R. 3619,
the “Employee Free Choice Act.”6 S. 1513 was introduced by Senator Charles
Schumer; S. 1925 was introduced by Senator Edward Kennedy. H.R. 3078 and H.R.
3619 were introduced by Representative George Miller. No action has been taken
in the House or Senate on either proposal.
Representative Charlie Norwood introduced H.R. 4343, the “Secret Ballot
Protection Act of 2004.” The bill would require secret-ballot elections for union
certification. Employers could not voluntarily bargain with a union that has not been
elected by a majority of employees in a secret-ballot election. No action has been
taken on the bill.
The “National Right-to-Work Act” would amend both the NLRA and RLA.
Under this measure, union contracts could not require employees to become union
members as a condition of employment. The bill was introduced in the Senate by
Senator Trent Lott (S. 1765) and in the House by Representative Joe Wilson (H.R.
391). No congressional action has been taken on the proposal.
Governments and Collective Bargaining
By bargaining collectively, instead of individually, unionized workers may
obtain higher wages and better working conditions than if each worker bargained
individually.7 The protections that governments give employees who organize and
6 For an overview of S.1925/H.R.3619, see CRS Report RS21887, The Employee Free
Choice Act, by Jon O. Shimabukuro.
7 The threat of a strike can increase the bargaining power of unionized workers. Unions may
also be able to increase wages by limiting the supply of workers; e.g., by restricting the
number of persons enrolled in union-run training programs. Federal employees cannot
(continued...)

---

CRS-4
bargain collectively are intended to achieve different policy objectives. These
objectives include to increase the bargaining power of employees, to reduce earnings
inequality, and to provide a means for improved communication between labor and
management.
Government Intervention in Labor Markets
Governments may intervene in labor markets for a number of reasons. One of
these reasons is to improve competition.8 According to economic theory, competitive
markets generally result in the most efficient allocation of resources, where resources
consist of individuals with different skills, capital goods (e.g., computers, machinery,
and buildings), and natural resources.
In competitive labor markets workers are paid according to the value of their
contribution to output. Under perfect competition, wages include compensation for
unfavorable working conditions. The latter theory, called the “theory of
compensating wage differentials,” recognizes that individuals differ in their
preferences or tolerance for different working conditions — such as health and safety
conditions, hours worked, holidays and annual leave, and job security.9
If labor markets do not fit the model of perfect competition, increasing the
bargaining power of employees may raise wages and improve working conditions to
levels that might exist under competitive conditions. In labor markets where a firm
7 (...continued)
strike. The employees of most, but not all, state and local governments are not allowed to
strike. Where state and local government employees are allowed to strike, the right often
does not include public safety employees (e.g., policemen and firefighters). Daniel Quinn
Mills, Labor-Management Relations, 5th ed. (New York: McGraw-Hill, 1994), pp. 306-307.
(Hereafter cited as Mills, Labor-Management Relations.) Michael H. Cimini, “1982-97
State and Local Government Work Stoppages and Their Legal Background,” Compensation
and Working Conditions, vol. 3, fall 1998, pp. 33-34. Bruce E. Kaufman, The Economics
of Labor Markets, 4th ed., Fort Worth, Dryden Press, 1994, pp. 275-280. (Hereafter cited
as Kaufman, The Economics of Labor Markets.) Reynolds et al., Labor Economics and
Labor Relations, p. 406.
8 The following conditions are generally cited as the characteristics of a competitive labor
market: (1) There are many employers and many workers. Each employer is small relative
to the size of the market. (2) Employers and workers are free to enter or leave a labor
market and can move freely from one market to another. (3) Employers do not organize to
lower wages and workers do not organize to raise wages. Governments do not intervene in
labor markets to regulate wages. (4) Employers and workers have equal access to labor
market information. (5) Employers do not prefer one worker over another equally qualified
worker. Workers do not prefer one employer over another employer who pays the same
wage for the same kind of work. (6) Employers seek to maximize profits; workers seek to
maximize satisfaction. Reynolds et al., Labor Economics and Labor Relations, pp. 16-21.
9 Randall K. Filer, Daniel S. Hamermesh, and Albert E. Rees, The Economics of Work and
Pay, 6th ed. (New York: Harper Collins, 1996), pp. 376-390. (Hereafter cited as Filer et al.,
The Economics of Work and Pay.) Ronald G. Ehrenberg and Robert S. Smith, Modern
Labor Economics: Theory and Public Policy, 7th ed. (Reading, MA: Addison-Wesley,
2000), pp. 251-259. (Hereafter cited as Ehrenberg and Smith, Modern Labor Economics.)

---

CRS-5
is the only employer (called a monopsony) unions can, within limits, increase both
wages and employment.10
If labor markets are already competitive, however, economists maintain that
increasing the bargaining power of employees may result in a misallocation of
resources. In competitive labor markets, higher union wages may reduce
employment for union workers below the levels that would exist in the absence of
unionization.11 If unions lower employment in the unionized sector, they may also
increase the supply of workers to employers in the nonunion sector, lowering the
wages of nonunion workers.12
It can be difficult to determine the competitiveness of labor markets. First,
identifying the appropriate labor market may be difficult. Labor markets can be local
(e.g., for unskilled labor), regional, national, or even international (e.g., for
managerial and professional workers). Second, labor market competitiveness is
difficult to measure, and labor markets may change because of economic,
technological, or policy changes.
Distribution of Earnings
Competitive labor markets may result in a distribution of earnings that some
policymakers find unacceptable. Thus, governments may intervene in labor markets
to reduce inequality.13 Unionization may be a means of reducing earnings inequality.
According to some economists, greater equality may, under certain conditions (e.g.,
such as the Great Depression of the 1930s), also increase aggregate demand and,
therefore, reduce unemployment.
10 Kaufman, The Economics of Labor Markets, pp. 277-280.
11 In competitive labor markets, unions can offset the employment effect of higher wages by
persuading consumers to buy union-made goods (e.g., campaigns to “look for the union
label”), limiting competition from foreign made goods (e.g., though tariffs or import quotas),
or negotiating contracts that require more workers than would otherwise be needed to
perform certain tasks. Kaufman, The Economics of Labor Markets, pp. 276-277. Ehrenberg
and Smith, Modern Labor Economics, p. 493. Toke Aidt and Zafiris Tzannatos, Unions and
Collective Bargaining: Economic Effects in a Global Environment (Washington: The
World Bank, 2002), p. 27. (Hereafter cited as Aidt and Tzannatos, Unions and Collective
Bargaining.)
12 If unions raise the wages of union workers and lower employment in the union sector, the
supply of workers available to nonunion employers may increase, causing nonunion wages
to fall (the “spillover” effect). On the other hand, nonunion employers, in order to
discourage workers from unionizing, may pay higher wages (the “threat” effect). Ehrenberg
and Smith, Modern Labor Economics, pp. 504-508.
13 Governments may also intervene in private markets to produce “public” goods (e.g.,
national defense) or correct instances where the market price of a good does not fully reflect
its social costs or benefits — called, respectively, negative and positive “externalities.” Air
and water pollution are frequently cited as examples of negative externalities; home
maintenance and improvements are often cited as examples of positive externalities.

---

CRS-6
Collective Voice
Finally, an argument made by some economists is that unions give workers a
“voice” in the workplace. According to this argument, unions provide workers an
additional way to communicate with management. For instance, instead of
expressing their dissatisfaction with an employer by quitting, workers can use dispute
resolution or formal grievance procedures to resolve issues relating to pay, working
conditions, or other matters.14
Economic Effects of Labor Unions
This section summarizes the findings of selected research on the economic
effects of labor unions.15
Earnings
Numerous studies have attempted to measure the wage differential between
union and nonunion workers. The results vary. But, in general, most studies find
that, after controlling for individual, job, and labor market characteristics, the wages
of union workers are in the range of 10% to 30% higher than the wages of nonunion
workers.16,17
14 Richard B. Freeman and James L. Medoff, “The Two Faces of Unionism,” Public Interest,
no. 57, fall 1979, pp. 70-73. Richard B. Freeman, “The Exit-Voice Tradeoff in the Labor
Market: Unionism, Job Tenure, Quits, and Separations,” Quarterly Journal of Economics,
vol. 94, June 1980, pp. 644-645.
15 The summary is of research on U.S. labor markets, although some of the studies cited
include both the United states and other countries.
16 Filer et al, The Economics of Work and Pay, p. 489. Kaufman, The Economics of Labor
Markets, p. 609. Kay E. Anderson, Philip M. Doyle, and Albert E. Schwenk, “Measuring
Union-Nonunion Earnings Differences,” Monthly Labor Review, vol. 113, June 1990, p. 26.
Reynolds et al., Labor Economics and Labor Relations, pp. 515-517. Aidt and Tzannatos,
Unions and Collective Bargaining, p. 42. For a review of several studies of the union-
nonunion wage differential, see Javed Ashraf, “Union Wage Effects: An Overview of
Recent Literature,” Labor Studies Journal, vol. 19, summer 1994, pp. 3-24.
17 Most, but not all, studies that use cross-sectional data have found a larger union wage
premium than studies that use longitudinal data. (Kaufman, The Economics of Labor
Markets, pp. 612-14.) A cross-sectional survey collects data at a single point in time from
a sample of households or individuals. A longitudinal survey collects data at several points
in time from the same sample of households or individuals. Surveys generally do not collect
information on all personal characteristics that may affect individual pay; for example,
motivation or work effort. By comparing the wages of individuals who move from nonunion
to union jobs (or vice versa), longitudinal data can capture the effect of otherwise
unobserved personal characteristics. Some research has concluded that the lower union
wage premium found using longitudinal data is due to errors in measuring changes in union
status. Steven Raphael, “Estimating the Union Earnings Effect Using a Sample of Displaced
Workers,” Industrial and Labor Relations Review, vol. 53, Apr. 2000, pp. 504, 513-516.

---

CRS-7
Some evidence suggests that the wage premium for union workers has declined
in recent years.18 One study concluded that, among wage and salary workers, the
union wage differential in the late 1970s was approximately 21% to 23%. By 2000-
2001, the analysis concluded that the differential had fallen to 14%.19
Total compensation consists of both wages and fringe benefits (e.g., paid sick
leave, health insurance, and pension plans). Union workers generally receive better
or more generous fringe benefits than similar nonunion workers. Therefore,
estimates of the union wage premium may understate the difference in total
compensation between union and nonunion workers. On the other hand, if working
conditions are less favorable for union than nonunion workers, analyses that do not
control for differences in working conditions may overstate the difference in
compensation between union and nonunion workers.20,21
The wage premium for union workers is generally larger for less skilled than for
more skilled workers, greater for blue-collar than white-collar workers, larger for
younger than older workers, and larger for less educated workers (high school
graduates or high school dropouts) than college graduates. As a result, unions tend
to compress wages (i.e., reduce inequality) within unionized sectors of the
18 Barry T. Hirsch, “Reconsidering Union Wage Effects: Surveying New Evidence on an
Old Topic,” Journal of Labor Research, vol. 25, spring 2004, pp. 245-252. Peter Turnbull,
“What Do Unions Do Now?” Journal of Labor Research, vol. 24, summer 2003, p. 493.
Kaufman, The Economics of Labor Markets, pp. 617-19. July 13, 2004. David G.
Blanchflower and Alex Bryson, What Effect Do Unions Have on Wages Now and Would
“What Do Unions Do?” Be Surprised?, National Bureau of Economic Research, Working
Paper 9973, Sept. 2003, p. 9. (Hereafter cited as Blanchflower and Bryson, What Effect Do
Unions Have on Wages Now?)
19 The analysis is for wage and salary workers ages 16 and over. The analysis controls for
both worker and job characteristics (e.g., education, potential work experience, marital
status, race, gender, region, large metropolitan area, part-time employment, industry, and
occupation). Barry T. Hirsch and David A. Macpherson, Union Membership and Earnings
Data Book: Compilations from the Current Population Survey (Washington: Bureau of
National Affairs, 2003), pp. 1-2, 7, 19.
20 Filer et al, The Economics of Work and Pay, p. 493. Aidt and Tzannatos, Unions and
Collective Bargaining, pp. 73-75. Kaufman, The Economics of Labor Markets, pp. 607,
629. Reynolds, et al., Labor Economics and Labor Relations, pp. 517-19. Ehrenberg and
Smith, Modern Labor Economics, pp. 510-511.
21 Some research has concluded that, for blue-collar workers, unionized firms tend to have
more structured work settings, more hazardous jobs, less flexible work hours, a faster work
pace, lower job satisfaction, and less employee control over the assignment of overtime
hours. Therefore, part of the estimated union-nonunion earnings differential may
compensate union workers for unfavorable working conditions. Ehrenberg and Smith,
Modern Labor Economics, pp. 510-511. Kaufman, The Economics of Labor Markets, p.
613.

---

CRS-8
economy.22 Some evidence suggests that unions reduce earnings inequality in the
overall economy.23,24
Private and Public Sectors
The wage gap between union and nonunion workers is generally larger in the
private sector than in the public sector. Within the public sector, evidence suggests
that the wage premium for union workers is greater for local government employees
than for federal employees.25
Gender
Research has concluded that there is very little, if any, difference in the union
wage premium between men and women.26
Race
Some, but not all, evidence indicates that the union wage premium is greater for
nonwhites than whites. Some studies do not find a difference in the union wage
premium between blacks and whites; other research concludes that the wage
premium for black workers is 5 to 10 percentage points higher than the wage
premium for white workers.27
22 Blanchflower and Bryson, What Effect Do Unions Have on Wages Now? p. 8. David G.
Blanchflower, Changes Over Time in Union Relative Wage Effects in Great Britain and the
United States, National Bureau of Economic Research, Working Paper 6100, July 1997, p.
30. Aidt and Tzannatos, Unions and Collective Bargaining, pp. 53-54. Ehrenberg and
Smith, Modern Labor Economics, p. 509.
23 Filer et al, The Economics of Work and Pay, pp. 503-504. Reynolds et al., Labor
Economics and Labor Relations, p. 527.
24 For an examination of trends in the distribution of earnings among wage and salary
workers, see CRS Report RL31616, The Distribution of Earnings of Wage and Salary
Workers in the United States, 1994-2002, by Gerald Mayer.
25 Aidt and Tzannatos, Unions and Collective Bargaining, pp. 52-53. Ehrenberg and Smith,
Modern Labor Economics, p. 508. Kaufman, The Economics of Labor Markets, p. 626.
26 Blanchflower and Bryson, What Effect Do Unions Have on Wages Now? p. 10. Kaufman,
The Economics of Labor Markets, p. 612. Aidt and Tzannatos, Unions and Collective
Bargaining, p. 49.
27 Lawrence Mishel, Jared Bernstein, Heather Boushey, The State of Working America:
2002/2003 (Ithaca: Cornell University Press, 2003), pp. 191-192. Aidt and Tzannatos,
Unions and Collective Bargaining, pp. 50-51. Ehrenberg and Smith, Modern Labor
Economics, p. 509.

---

CRS-9
Job Tenure and Quit Rates
Job tenure tends to be greater and quit rates lower among unionized workers.28
Higher wages attract more applicants, resulting in larger applicant queues, giving
employers a larger pool from which to hire qualified workers. As a result, some
evidence indicates that the “quality” of union workers may be better than that of
nonunion workers doing the same kind of work.29 Quit rates may also be lower
among union employees if unions give workers and management a means to improve
communications and resolve issues.
Productivity
Unions can potentially have both beneficial and harmful effects on labor
productivity. Restrictive work rules may harm productivity by limiting the ability of
management to assign work or introduce new technology. Higher wages may reduce
investment in equipment and lower spending on research and development. On the
other hand, higher wages may attract better workers and cause employers to
substitute machinery and equipment for labor (i.e., increasing the amount of fixed
capital per worker). A lower quit rate may create an incentive for employers to
provide more firm-specific training. Union firms may also hire more professional
managers and adopt more efficient management practices.30
Evidence on the effect of unions on labor productivity is mixed. According to
some research, the effect of unions on productivity varies across industries.31 Some
28 Aidt and Tzannatos, Unions and Collective Bargaining, p. 65. Reynolds et al., Labor
Economics and Labor Relations, pp. 535-537.
29 Filer et al, The Economics of Work and Pay, pp. 284-285, 493. In theory, persons will
voluntarily change jobs if the expected gain from changing jobs (where one is offered) is
greater than the expected gain from staying in the job they have, less the cost of changing
jobs.
30 For a summary of the beneficial and harmful effects of unions on labor productivity, see
Chrisom Doucouliagos and Patrice Laroche, “What Do Unions Do to Productivity? A Meta-
Analysis,” Industrial Relations, vol. 42, Oct. 2003, pp. 651-655.
31 In office building construction, unionized workers were found to be more productive than
nonunion workers. A study of the cement industry concluded that labor productivity was
greater in unionized firms. (Filer et al, The Economics of Work and Pay, p. 514.) According
to one study, labor unions have no affect on productivity growth in manufacturing, but have
a negative effect on productivity in the construction industry. (Steven Allen, “Productivity
Levels and Productivity Change Under Unionism,” Industrial Relations, vol. 27, winter
1988, pp. 103-104, 107-108.) A study of underground coal mines concluded that unions
organized workers in more productive mines, which accounted for a positive relationship
between unions and productivity. After controlling for differences in mine productivity, the
study concluded that unions have a negative effect on productivity. (Brian Chezum and
John E. Garen, “Are Union Productivity Effects Overestimated? Evidence from Coal
Mining,” Applied Economics, vol. 30, July 1998, p. 918.) A study of western U.S. sawmills
concluded that productivity was lower in unionized than in nonunionized mills. (Merwin
W. Mitchell and Joe A. Stone, “Union Effects on Productivity: Evidence from Western U.S.
(continued...)

---

CRS-10
research has concluded that the effect of unions on productivity may depend, in part,
on the quality of labor-management relations. In particular, if unions improve labor
management communications, unions may have a positive effect on productivity.32
Profits
Finally, research suggests that unions reduce a firm’s rate of profit. Some
evidence indicates that the effect of unions on profits is greater in concentrated
industries where profits may be relatively higher because firms have the ability to
influence the prices of their products.33,34 Other research concludes that unions
reduce profits in general, regardless of the ability of firms to influence prices.35
Trends in Union Membership
Union membership in the United States has declined significantly in recent
decades. The number of union members peaked in 1979 at an estimated 21.0
million.36 In 2003, an estimated 15.8 million workers were union members. See
Table A1 in Appendix A.37
As a percent of workers, union membership can be represented in different ways
(e.g., as a percent of the labor force or as a percent of wage and salary workers).
Each approach has its advantages and disadvantages. The labor force includes both
employed and unemployed workers. Union membership as a percent of the labor
force would be the broadest measure of union membership. But such a measure may
be more sensitive than other measures to changes in the unemployment rate. Union
31 (...continued)
Sawmills,” Industrial and Labor Relations Review, vol. 46, Oct. 1992, pp. 141-142.)
32 Kaufman, The Economics of Labor Markets, pp. 631-634. Reynolds et al., Labor
Economics and Labor Relations, pp. 537-540. Ehrenberg and Smith, Modern Labor
Economics, p. 512. Aidt and Tzannatos, Unions and Collective Bargaining, p. 70.
33 Aidt and Tzannatos, Unions and Collective Bargaining, p. 68. Filer et al., The Economics
of Work and Pay, pp. 515-516. Ehrenberg and Smith, Modern Labor Economics, pp. 512-
513. Barry T. Hirsch, “Union Coverage and Profitability Among U.S. Firms,” Review of
Economics and Statistics, vol. 73, Feb. 1991, pp. 74-76. Paula B. Voos and Lawrence R.
Mishel, “The Union Impact on Profits: Evidence from Industry Price-Cost Margin Data,”
Journal of Labor Economics, vol. 4, Jan. 1986, pp. 105-109.
34 A common measure of economic concentration is the percent of industry output accounted
for by the four largest firms. Paul A. Samuelson and William D. Nordhaus,
Microeconomics, 16th ed. (New York: McGraw-Hill, 1992), pp. 170-171.
35 William F. Chappell, Walter J. Mayer, and William F. Shughart II, “Union Rents and
Market Structure Revisited,” Journal of Labor Research, vol. 12, winter 1991, pp. 35-37.
36 Beginning in 1977 a union member is a wage and salary worker who belongs to a labor
union or an employee association that is similar to a union.
37 The percentages shown in the graphs in this report are based on the estimates shown in
the tables in Appendix A.

---

CRS-11
membership is often represented as a percent of nonagricultural employment.
Although union membership in the agriculture industry is small, such calculations
may exclude from the denominator an industry that is included in the numerator.
Union membership is also represented as a percent of wage and salary employment.
One of the possible economic effects of unions, however, is that they may reduce
employment in the union sector of the economy and increase the supply of labor to
the nonunion sector of the economy. This is called the “spillover” effect. The
nonunion sector of the economy includes both nonunion wage and salary workers and
nonunion self-employed workers. Workers not in the union sector have the option,
therefore, of nonunion wage and salary employment or nonunion self-employment.
But self-employed workers are, in effect, both employer and employee and, therefore,
do not unionize.
Figure 1 shows union membership as a percent of three measures of
employment: (a) total employment, (b) wage and salary employment, and (c)
Figure 1. Union Membership as a Percent of Employment,
1930-2003

---

CRS-12
nonagricultural wage and salary employment.38 Union membership as a percent of
employed workers is lower than union membership as a percent of wage and salary
workers. Reflecting the relative decline in agricultural employment and the number
of self-employed workers, the three series have converged somewhat over the past
half century.39
As a percent of nonagricultural employment, union membership peaked at
35.4% in 1945. As a percent of wage and salary employment and a percent of total
employment, union membership peaked in 1954 at 34.8% and 28.3%, respectively.
In 2003, 12.4% of wage and salary workers, 12.1% of nonagricultural workers, and
11.5% of all employed workers were union members.
Some workers are represented by a collective bargaining agreement but are not
union members. In 2003, an estimated 1.7 million workers were covered by a union
contract but were not union members. From 1994 to 2003, the percentage of
employed workers who were represented by a union fell from 1.7% to 1.2%.40
If the union wage premium has declined in recent years (as discussed above),
lower union membership may account for part of this decline. The decline in union
membership may have also moderated some of the other economic effects of unions
discussed above (e.g., on profitability and productivity).
Characteristics of Union Membership
This section examines selected demographic, social, and economic
characteristics of union members in the United States. The analysis examines trends
in union membership from 1994 to 2003 (or 2002, for industry and occupation). The
analysis examines union membership as a percent of employed persons ages 16 and
over. The data are from the monthly Current Population Survey (CPS), which is
conducted by the Census Bureau for the Bureau of Labor Statics (BLS). See
Appendix B for a description of data and methodology.
38 The definition of wage and salary workers used by the Bureau of Labor Statistics (BLS)
includes incorporated self-employed workers. Self-employed incorporated workers are
employees of a corporation. In its calculations of union membership rates, however, BLS
generally excludes both incorporated and unincorporated self-employed workers.
39 From 1948 to 2003, the percentage of workers employed in agriculture declined from
13.1% to 1.7%. U.S. Department of Labor, Bureau of Labor Statistics, Labor Force
Statistics Derived From the Current Population Survey, 1948-87, Bulletin 2307, Aug. 1988,
p. 625. U.S. Department of Labor, Bureau of Labor Statistics, Employment and Earnings,
vol. 51, Jan. 2004, p. 219. For a discussion of the trend in self-employment, see CRS Report
RL32387, Self-Employment as a Contributor to Job Growth and as an Alternative Work
Arrangement, by Linda Levine.
40 Unless stated otherwise, the differences and changes in union membership or coverage
discussed in the text are significant at the 95% confidence level. See Appendix B for a
discussion of confidence levels.

---

CRS-13
Gender
Figure 2 shows that men are
Figure 2. Union Membership Rates of
more likely than women to be
Men and Women, 1994-2003
union members. In 2003, 12.3%
of men were union members,
compared to 10.5% of women.
From 1994 to 2003, union
membership declined among both
men and women. The decline in
union membership was greater for
men (2.9 percentage points) than
for women (1.3 percentage
points).
Table A2 in Appendix A
also shows theat union members
are more likely to be male than
female.
Age
Figure 3 shows that workers
Figure 3. Union Membership Rates by
between the ages of 45 and 64 are
Age, 1994-2003
more likely than younger workers
or workers ages 65 and over to be
union members. In 2003, 14.7%
of workers ages 45 to 64 were
union members. By comparison,
5.0% of workers ages 16 to 24 and
11.3% of workers ages 25 to 44
were union members.
From 1994 to 2003, except
for persons ages 65 and over,
union membership declined
among all age groups. The largest
declines were among persons ages
35-44 (3.8 percentage points) and
persons ages 45-54 (3.7
percentage points).41
41 From 1994 to 2003, union membership among persons ages 45 to 64 increased (from 6.2
to 7.1 million), while membership among persons under 45 decreased (from 10.3 to 8.4
million). However, during the period, the percentage of employed workers ages 45 to 64
increased from 28.1% to 35.2%. From 1994 to 2003, the percentage of union members who
were between 45 and 64 increased from 37.2% to 45.3%. See Table A3.

---

CRS-14
Table A3 shows that, in 2003, over half (58.0%) of union members were
between the ages of 35 and 54.
Race
A majority of union members
Figure 4. Union Membership Rates by
are white: 79.5% in 2003.
Race, 1994-2003
However, Figure 4 shows that
blacks are more likely than whites
or other races to be union
members. In 2003, 15.6% of
blacks were union members,
compared to 11.0% of whites.
From 1994 to 2003, union
membership declined for all racial
groups. Union membership
among blacks declined by 4.0
percentage points, and by 1.9
percentage points among whites.42
One reason for the higher
rate of union membership among
blacks is that blacks are more
likely to be employed in the public
sector, where union membership is greater than in the private sector (see “Private
and Public Sectors” below). In 2003, 15.0% of public sector workers were black,
compared to 10.0% of private sector workers. (See Appendix B for an explanation
of how individuals are categorized by race.)
Hispanic Origin
Figure 5 shows that union membership is greater among non-Hispanic workers
than among Hispanic workers.43 In 2003, 11.7% of non-Hispanic workers were
union members, compared to 9.9% of Hispanic workers.
42 The estimates for 2003 of the number of workers by race are not strictly comparable to
estimates for earlier years. See Appendix B.
43 Hispanics can be of any race.

---

CRS-15
From 1994 to 2003, the
Figure 5. Union Membership Rates by
number of Hispanic workers
Hispanic Origin, 1994-2003
increased from 10.8 to 17.3
million. During this period,
Hispanic workers as a percentage
of employed workers increased
from 8.8% to 12.6%. Reflecting
the increase in the number of
Hispanic workers, the number of
unionized Hispanic workers
increased from 1.4 to 1.7 million.
However, the decline in union
membership from 1994 to 2003
was greater among Hispanic (3.3
percentage points) than non-
Hispanic workers (2.0 percentage
points).44
Educational Attainment
Figure 6 shows that workers
with less than a high school
Figure 6. Union Membership Rates by
education are least likely to be
Level of Education, 1994-2003
union members, while workers
with advanced college degrees are
most likely to be union members.
In 2003, 6.6% of workers who had
not graduated from high school
were union members, compared to
15.4% of workers with an
advanced degree. But more union
members have only a high school
education or less (6.1 million in
2003) than have a bachelor’s or
advanced degree (5.1 million in
2003).
From 1994 to 2003, union
membership declined among all
educational groups. The decline
was greatest among workers with
a high school education or less.
The percentage of union members with a bachelor’s or advanced degree has
increased. In 1994, 46.6% of union members had a high school education or less;
44 The estimate of the number of Hispanic workers for 2003 is not strictly comparable to
estimates for earlier years. See Appendix B.

---

CRS-16
25.3% had a bachelor’s or advanced degree. By 2003, 38.7% of union members had
a high school education or less, and 32.1% had a bachelor’s or advanced degree.
In 2003, almost three-fourths (73.6%) of union members with a bachelor’s or
advanced degree were employed in the public sector, mostly for state (16.4%) and
local (52.4%) governments. The largest percentage of these employees (43.6%) were
preschool, elementary, secondary, and special education teachers.
Private and Public Sectors
Union members are more
Figure 7. Union Membership Rates in
likely to be employed in the
the Public and Private Sectors, 2003
private than the public sector. In
2003, an estimated 8.5 million
union members were employed in
the private sector, compared to an
estimated 7.3 million union
members employed in the public
sector. However, Figure 7 shows
that the rate of union membership
in the public sector is significantly
greater than in the private sector.
In 2003, 37.2% of public sector
employees and 7.2% of private
sector employees were union
members. In addition, from 1994
to 2003, the percentage of union
members employed in the public
sector increased from 42.4% to
46.4%.
From 1994 to 2003, union membership declined from 38.7% to 37.2% in the
public sector and from 9.2% to 7.2% in the private sector.
The relative increase in the number of union members who are employed in the
public sector may account for part of the reported decline in the union wage premium
(i.e., the union wage premium is smaller in the public than private sector). In
addition, some evidence suggests that the decline in union membership in the private
sector has contributed to rising earnings inequality.45
Several reasons have been given for the decline in union membership in the
private sector. Changes in employment by industry, occupation, and region are often
cited as contributing factors.46 Historically, unionization in the private sector has
45 Martin A. Asher and Robert H. DeFina, “The Impact of Changing Union Density on
Earnings Inequality: Evidence from the Private and Public Sectors,” Journal of Labor
Research, vol.18, summer 1997, pp. 426.
46 Henry S. Farber, “The Decline of Unionization in the United States: What Can be
(continued...)

---

CRS-17
been greatest in four industrial groups: construction; manufacturing; mining; and
transportation, utilities, and communications. From 1994 to 2002, the percentage of
workers employed in manufacturing and mining declined from 16.9% to 13.6%.47
Blue-collar jobs as a percent of total employment declined from 25.5% to 23.6%.48
And the share of employment in the mountain and southern states, where there tends
to be less unionization, has increased. From 1994 to 2003, the share of the workers
employed in the mountain states increased from 6.0% to 6.7% and in the south
Atlantic states from 18.0% to 18.5%. See Appendix B for regional state groupings.
Increased competition is also cited as a reason for the decline in union
membership in the private sector. In some industries (e.g., airlines, trucking, and
telecommunications), deregulation has increased competition among existing firms
and led to the entry of nonunion employers.49 Increased foreign competition has also
led American firms to look for ways to lower costs. Unionized firms may seek wage
concessions from unionized workers or move production to nonunion locations.50,51
In addition, employers may have become more sensitive to employee concerns,
resulting in greater job satisfaction among nonunion workers and reducing the
demand for unionization. On the other hand, management may have become more
sophisticated in opposing attempts by workers to unionize.52
Slower employment growth in union firms may have also contributed to the
decline in the proportion of the workforce that is organized.53
46 (...continued)
Learned from Recent Experience?” Journal of Labor Economics, vol. 8, no. 1, pt. 2, 1990,
p. S76.
47 From 1994 to 2002, the percentage of workers employed in construction and in
transportation, communications, and utilities increased from 13.1% to 14.1%. See Table
A9.
48 Blue-collar workers are defined here as the sum of “precision production, craft, and
repair” workers and “operators, fabricators, and laborers.” See Table A10.
49 James Peoples, “Deregulation and the Labor Market,” Journal of Economic Perspectives,
vol. 12, summer 1998, pp. 111-112.
50 Reynolds et al., Labor Economics and Labor Relations, pp. 421-426.
51 In the United States the total level of trade (exports plus imports) as a share of gross
domestic product (GDP) grew from 9% in 1960 to 22% in 2000. CRS Report RL32350,
Deindustrialization of the U.S. Economy: The Roles of Trade, Productivity, and Recession,
by Craig K. Elwell.
52 Mills, Labor-Management Relations, pp. 80-81. Reynolds et al., Labor Economics and
Labor Relations, pp. 427-428. For a discussion of management efforts to discourage
unionization, see Morris M. Kleiner, “Intensity of Management Resistance: Understanding
the Decline of Unionization in the Private Sector,” Journal of Labor Research, vol. 22,
summer 2001, pp. 519-540.
53 According to one review of the literature, studies have typically found that employment
growth in nonunionized firms is three to five percentage points greater than in unionized
firms. (Aidt and Tzannatos, Unions and Collective Bargaining, p. 64.) A study of
(continued...)

---

CRS-18
Finally, governments and employers may provide benefits formerly provided by
unions, reducing the demand for union representation. For instance, government
health and safety laws may substitute for demands formerly included in collective
bargaining agreements.54 Similarly, governments and employers may provide fringe
benefits — such as unemployment compensation or retirement benefits — that were
formerly provided by unions.55
Level of Government
The public sector includes
Figure 8. Union Membership Rates by
the federal, state, and local
Level of Government, 2003
governments. Figure 8 shows
that union membership is greater
at the local level than at the
federal or state levels. In 2003,
42.6% of employees of local
g o v e r n m ent s w e r e u n i o n
members, compared to 30.9% of
federal workers and 30.3% of state
workers.
From 1994 to 2003 union
membership declined at the
federal and local levels, but not at
the state level.
Local governments include
elementary and secondary schools
and fire and police departments.
In 2003, 42.6% of unionized local government employees were preschool,
elementary, secondary, and special education teachers, teacher assistants, and
education administrators. Another 13.7% were police officers, firefighters, and
correctional officers.
53 (...continued)
manufacturing plants in California concluded that the annual rate of growth in employment
was two to four percentage points lower in union than nonunion firms. Jonathan S. Leonard,
“Unions and Employment Growth,” Industrial Relations, vol. 31, winter 1992, pp. 82, 91.
54 George R. Neumann and Ellen R. Rissman, “Where Have All the Union Members Gone?”
Journal of Labor Economics, vol. 2, no. 2, 1984, p. 176.
55 Martin A. Ahser and Robert H. DeFina, “Has Deunionization Led to Higher Earnings
Inequality?” Federal Reserve Bank of Philadelphia Business Review, Nov/Dec. 1995, pp.
8-9.

---

CRS-19
Industry
Figure 9 shows
Figure 9. Union Membership Rates by
union membership by
Industry, 2002
major industry in 2002.
The least unionized
industries in 2002 were
farming (1.6%), finance,
insurance, and real estate
(2.0%), and private
household and other
services (3.6%). The
m o s t u n i o n i z e d
industries were public
administration (32.3%)
and transportation,
communications, and
utilities (27.4%).
From 1994 to 2003,
the largest percentage
decl i n es in union
membership were in
mining; transportation,
communications, and
utilities; and manufacturing. In mining, union membership fell from 15.0% to 8.0%.
In transportation, communications, and utilities membership fell from 32.8% to
27.4% and in manufacturing membership fell from 17.5% to 13.9%.56
Occupation
Figure 10 shows union membership by major occupation for 2002. The least
unionized workers are farm workers. (Recall that crop and harvest workers are not
covered by the NLRA and that most states do not have laws that provide collective
bargaining rights to farmworkers.)
In 2002, unionization was greatest among precision production workers and
operators (18.3% and 17.6%, respectively). Precision production workers include
car, truck, and aircraft mechanics, machinists, plumbers, electricians, carpenters,
mine workers, butchers, bakers, and others. Operators include machine operators,
truck and bus drivers, train operators, assemblers, laborers, and others.
56 From 1994 to 2002, the number of unionized workers in the construction and professional
services industries increased from 6.0 to 7.0 million. But the share of total workers
employed in these industries increased from 29.7% to 32.5%. See Table A9.

---

CRS-20
From 1994 to 2002, union
Figure 10. Union Membership Rates
membership declined in all
by Occupation, 2002
occupations except farming. The
drop in union membership was
greatest in the most unionized
occupations. Unionization among
operators declined by 4.7
percentage points and by 2.5
points among precision production
workers.57
Region
Figure 11 shows union
membership by region for 2003.
The nine regions are based on
state groupings used by the
Census Bureau. See Appendix B
for a list of states by regions.
Figure 11 shows that, in 2003, the
m o s t u n i o n i z e d
regions were the mid-
Figure 11. Union Membership Rates by Region,
Atlantic region, the
2003
Pacific region, and
the east north central
United States. The
l e a s t u n i o n i z e d
regions were the
south and mountain
regions. From 1994
t o 2 0 0 3 , u n i o n
membership declined
in every region of the
United States.
Within regions,
unionization varies by
state. In 2003, the 10
most unionized states
were New York,
Hawaii, Michigan,
Alaska, New Jersey,
Washington, Illinois,
Rhode Island, Ohio,
57 Union membership as a percent of employed workers declined among managerial and
professional employees, but the number of unionized managerial and professional workers
increased from 4.1 to 4.8 million. From 1994 to 2002, the percentage of workers in
managerial and professional occupations increased from 27.4% to 31.2%. See Table A10.

---

CRS-21
and Minnesota and California (Minnesota and California tied for 10th). See Figure
12.
The 10 least unionized states (from lowest to highest) were North Carolina,
South Carolina, Arkansas, Mississippi, Arizona, South Dakota, Utah, Texas, Florida,
and Louisiana.
Figure 12. Union Membership Rates, by State, 2003

---

CRS-22
Appendix A: Annual Data
Table A1. Union Membership in the United States, 1930-2003

Percent of Wage
Percent of
Percent of
Union Members
Year
and Salary
Nonagricultural
Employed
(In 1000s)
Workers
Workers
Workers
2003
15,776
12.4%
12.1%
11.5%
2002
16,183
12.8%
12.4%
11.9%
2001
16,315
12.9%
12.4%
11.9%
2000
16,334
12.9%
12.4%
11.9%
1999
16,477
13.4%
12.8%
12.3%
1998
16,211
13.4%
12.9%
12.3%
1997
16,110
13.6%
13.1%
12.4%
1996
16,269
14.0%
13.6%
12.8%
1995
16,360
14.3%
13.9%
13.1%
1994
16,740
14.9%
14.6%
13.6%
1993
16,598
15.1%
15.0%
13.8%
1992
16,390
15.2%
15.1%
13.8%
1991
16,568
15.5%
15.3%
14.1%
1990
16,740
15.5%
15.3%
14.1%
1989
16,961
15.9%
15.7%
14.5%
1988
17,002
16.2%
16.1%
14.8%
1987
16,913
16.5%
16.6%
15.0%
1986
16,975
17.0%
17.1%
15.5%
1985
16,996
17.4%
17.4%
15.9%
1984
17,340
18.2%
18.3%
16.5%
1983
17,717
19.5%
19.6%
17.6%
1982
N.A.
N.A.
N.A.
N.A.
1981
19,123
21.0%
20.9%
19.0%
1980
20,095
22.3%
22.2%
20.2%
1979
20,986
23.4%
23.3%
21.2%
1978
19,548
22.4%
22.5%
20.4%
1977
19,335
23.2%
23.4%
21.0%
1976
17,403
21.6%
21.9%
19.6%
1975
16,778
21.6%
21.8%
19.5%
1974
18,177
23.2%
23.2%
20.9%
1973
18,089
23.5%
23.5%
21.3%
1972
19,435
26.3%
26.3%
23.7%
1971
19,211
26.9%
26.9%
24.2%
1970
19,381
27.4%
27.3%
24.6%
1969
19,036
27.3%
27.0%
24.4%
1968
18,916
27.9%
27.8%
24.9%
1967
18,367
27.8%
27.9%
24.7%
1966
17,940
28.2%
28.0%
24.6%
1965
17,299
28.2%
28.4%
24.3%
1964
16,841
28.3%
28.8%
24.3%
1963
16,524
28.5%
29.1%
24.4%
1962
16,586
29.3%
29.8%
24.9%
1961
16,303
29.5%
30.1%
24.8%
1960
17,049
30.9%
31.4%
25.9%
1959
17,117
31.8%
32.1%
26.5%
1958
17,029
32.5%
33.1%
27.0%
1957
17,369
32.7%
32.8%
27.1%
1956
17,490
33.2%
33.3%
27.4%
1955
16,802
33.0%
33.1%
27.0%

---

CRS-23
Percent of Wage
Percent of
Percent of
Union Members
Year
and Salary
Nonagricultural
Employed
(In 1000s)
Workers
Workers
Workers
1954
17,022
34.8%
34.7%
28.3%
1953
16,948
33.8%
33.7%
27.7%
1952
15,892
32.3%
32.5%
26.4%
1951
15,946
32.8%
33.3%
26.6%
1950
14,267
30.4%
31.5%
24.2%
1949
14,282
31.6%
32.6%
24.8%
1948
14,319
31.2%
31.9%
24.5%
1947
14,787
33.6%
25.9%
1946
14,395
34.5%
26.1%
1945
14,322
35.4%
27.1%
1944
14,146
33.7%
26.2%
1943
13,213
31.0%
24.3%
1942
10,380
25.8%
19.3%
1941
10,201
27.9%
20.3%
1940
8,717
26.9%
18.3%
1939
8,763
28.6%
19.2%
1938
8,034
18.2%
1937
7,001
15.1%
1936
3,989
9.0%
1935
3,584
8.5%
1934
3,088
7.6%
1933
2,689
6.9%
1932
3,050
7.8%
1931
3,310
7.8%
1930
3,401
7.5%
Sources: The estimates of union membership and the total number of persons employed for 1994-
2003 were calculated by CRS from the monthly Current Population Survey (CPS). Estimates of union
membership for 1973-1993 are from: Barry T. Hirsch and David A. Macpherson, Union Membership
and Earnings Data Book: Compilations from the Current Population Survey, Washington, Bureau
of National Affairs, 2003, p. 11. Union membership data for 1930-1972 are from: U.S. Department
of Labor, Bureau of Labor Statistics, Handbook of Labor Statistics, Bulletin 1865, U.S. Govt. Print.
Off., 1975, p. 389. Estimates of the number of persons employed for 1930-1993 are from: U.S.
Department of Labor, Bureau of Labor Statistics, Employment and Earnings, vol. 50, Jan. 2003, p.
158 and U.S. Department of Labor, Bureau of Labor Statistics, Employment and Earnings, vol. 41,
Jan. 1994, p. 182. Estimates of the number of wage and salary workers are from: U.S. Department
of Labor, Bureau of Labor Statistics, Labor Force Statistics Derived From the Current Population
Survey, 1948-87, Bulletin 2307, U.S. Govt. Print. Off., Aug. 1988, p. 383; Handbook of Labor
Statistics, 7th ed., ed. by Eva E. Jacobs, Bernan Press, Lanham, MD, 2004, p. 75; and U.S.
Department of Labor, Bureau of Labor Statistics, Employment and Earnings, vol. 51, Jan. 2004, p.
219. Data on nonagricultural employment are from the BLS Current Employment Statistics (CES)
survey, available at [http://www.bls.gov].
Notes: The estimates of union membership for 1983-2003 are annual monthly averages based on the
monthly CPS. The monthly CPS has included a question about union membership since November
1982. The estimates for 1973-1981 are from the May CPS. The data for 1930-1972 include members
of AFL-CIO affiliates, unaffiliated national unions, unaffiliated unions with collective bargaining
agreements with different employers in more than one state, and members of federal employee unions.
Beginning in 1977, the estimates include members of employee associations. Because of changes in
the CPS survey, data for 1994 and later may not be comparable to earlier years. The estimates of the
number of wage and salary workers include self-employed incorporated workers. Because these
workers are paid employees of a corporation, BLS treats them as wage and salary workers. Finally,
the estimates of union membership and the total number of persons employed for 2000-2002 use
revised sample weights based on population controls introduced in Jan. 2003 (see Appendix B).

---

CRS-24
Table A2. Union Membership in the United States by Gender, 1994-2003
Union
Employed
Union
Employed
Members
Labor Force
Percent Union
Members
Labor Force
Percent Union
Total
Year
(1,000s)
(1,000s)
Members
(1,000s)
(1,000s)
Members
Employed
Labor Force
Men
Women
2003
9,044
73,332
12.3%
6,732
64,404
10.5%
137,736
2002
9,360
72,903
12.8%
6,822
63,582
10.7%
136,485
2001
9,546
73,196
13.0%
6,769
63,737
10.6%
136,933
2000
9,664
73,305
13.2%
6,671
63,586
10.5%
136,891
1999
9,949
71,446
13.9%
6,528
62,042
10.5%
133,488
1998
9,850
70,693
13.9%
6,362
60,771
10.5%
131,463
1997
9,763
69,685
14.0%
6,347
59,873
10.6%
129,558
1996
9,859
68,207
14.5%
6,410
58,501
11.0%
126,708
1995
9,929
67,377
14.7%
6,430
57,523
11.2%
124,900
1994
10,096
66,451
15.2%
6,644
56,611
11.7%
123,061
Source: Calculated by Congressional Research Service (CRS) from the monthly Current Population Survey (CPS).
Notes: Estimates are for persons 16 and over. Details may not add to totals because of rounding.

---

CRS-25
Table A3. Union Membership in the United States by Age, 1994-2003
Union
Employed
Percent
Union
Employed
Percent
Union
Employed
Percent
Members
Labor Force
Union
Members
Labor Force
Union
Members
Labor Force
Union
Total
Year
(1,000s)
(1,000s)
Members
(1,000s)
(1,000s)
Members
(1,000s)
(1,000s)
Members
Employed
Labor Force
16-24
25-34
35-44
2003
966
19,340
5.0%
3,097
30,357
10.2%
4,308
34,942
12.3%
137,736
2002
996
19,668
5.1%
3,177
30,288
10.5%
4,465
35,252
12.7%
136,485
2001
1,028
20,082
5.1%
3,240
30,849
10.5%
4,711
36,057
13.1%
136,933
2000
1,009
20,405
4.9%
3,444
31,560
10.9%
4,704
36,412
12.9%
136,891
1999
1,110
20,047
5.5%
3,415
30,862
11.1%
4,918
36,743
13.4%
133,488
1998
1,014
19,595
5.2%
3,332
31,407
10.6%
5,013
36,269
13.8%
131,463
1997
968
19,033
5.1%
3,434
31,832
10.8%
4,987
35,913
13.9%
129,558
1996
991
18,619
5.3%
3,536
32,094
11.0%
5,132
35,084
14.6%
126,708
1995
1,022
18,856
5.4%
3,596
32,356
11.1%
5,254
34,240
15.3%
124,900
1994
1,125
18,931
5.9%
3,769
32,255
11.7%
5,405
33,589
16.1%
123,061
45-54
55-64
65 and Over
2003
4,848
31,918
15.2%
2,300
16,595
13.9%
258
4,584
5.6%
137,736
2002
5,028
31,304
16.1%
2,264
15,658
14.5%
253
4,315
5.9%
136,485
2001
5,057
31,074
16.3%
2,033
14,625
13.9%
246
4,246
5.8%
136,933
2000
4,910
30,351
16.2%
2,026
14,004
14.5%
241
4,159
5.8%
136,891
1999
4,881
28,654
17.0%
1,932
13,331
14.5%
221
3,850
5.7%
133,488
1998
4,737
27,616
17.2%
1,923
12,874
14.9%
193
3,702
5.2%
131,463
1997
4,645
26,739
17.4%
1,894
12,289
15.4%
182
3,752
4.9%
129,558
1996
4,626
25,522
18.1%
1,795
11,750
15.3%
189
3,639
5.2%
126,708
1995
4,483
24,359
18.4%
1,801
11,431
15.8%
203
3,658
5.6%
124,900
1994
4,418
23,354
18.9%
1,807
11,265
16.0%
215
3,667
5.9%
123,061
Source: Calculated by Congressional Research Service (CRS) from the monthly Current Population Survey (CPS).
Notes: Estimates are for persons 16 and over. Details may not add to totals because of rounding.

---

CRS-26
Table A4. Union Membership in the United States by Race, 1994-2003
Union
Employed
Percent
Union
Employed
Percent
Union
Employed
Percent
Members
Labor Force
Union
Members
Labor Force
Union
Members
Labor Force
Union
Total
Employed
Year
(1,000s)
(1,000s)
Members
(1,000s)
(1,000s)
Members
(1,000s)
(1,000s)
Members
Labor Force
White
Black
Other
2003
12,535
114,233
11.0%
2,298
14,739
15.6%
943
8,764
10.8%
137,736
2002
12,988
114,048
11.4%
2,392
14,872
16.1%
802
7,565
10.6%
136,485
2001
13,170
114,489
11.5%
2,385
15,006
15.9%
760
7,439
10.2%
136,933
2000
13,111
114,422
11.5%
2,466
15,156
16.3%
757
7,313
10.4%
136,891
1999
13,349
112,275
11.9%
2,463
15,056
16.4%
665
6,157
10.8%
133,488
1998
13,118
110,936
11.8%
2,460
14,556
16.9%
633
5,972
10.6%
131,463
1997
13,088
109,847
11.9%
2,394
13,969
17.1%
627
5,742
10.9%
129,558
1996
13,232
107,801
12.3%
2,441
13,542
18.0%
596
5,364
11.1%
126,708
1995
13,149
106,491
12.3%
2,519
13,279
19.0%
691
5,130
13.5%
124,900
1994
13,515
105,195
12.8%
2,511
12,827
19.6%
714
5,039
14.2%
123,061
Source: Calculated by Congressional Research Service (CRS) from the monthly Current Population Survey (CPS).
Notes: Estimates are for persons 16 and over. Details may not add to totals because of rounding. Estimates for 2003 may not be comparable to previous years. Beginning in Jan.
2003, when answering the question about race, respondents may pick more than one race. Previously, individuals could only select one race. For 2003, this report follows BLS practice
and only counts blacks and whites who select one race category.

---

CRS-27
Table A5. Union Membership in the United States by Hispanic Origin, 1994-2003
Union
Employed
Percent
Union
Employed
Percent
Members
Labor Force
Union
Members
Labor Force
Union
Total
Year
(1,000s)
(1,000s)
Members
(1,000s)
(1,000s)
Members
Employed
Labor Force
Hispanic
Non-Hispanic
2003
1,712
17,314
9.9%
14,064
120,422
11.7%
137,736
2002
1,644
16,556
9.9%
14,539
119,929
12.1%
136,485
2001
1,692
16,183
10.5%
14,623
120,751
12.1%
136,933
2000
1,641
15,744
10.4%
14,693
121,147
12.1%
136,891
1999
1,525
13,719
11.1%
14,951
119,769
12.5%
133,488
1998
1,471
13,236
11.1%
14,741
118,228
12.5%
131,463
1997
1,407
12,724
11.1%
14,703
116,834
12.6%
129,558
1996
1,394
11,622
12.0%
14,875
115,085
12.9%
126,708
1995
1,357
11,135
12.2%
15,003
113,765
13.2%
124,900
1994
1,420
10,777
13.2%
15,321
112,284
13.6%
123,061
Source: Calculated by Congressional Research Service (CRS) from the monthly Current Population Survey (CPS).
Notes: Estimates are for persons 16 and over. Details may not add to totals because of rounding. Estimates for 2003 may not be comparable to previous years. Beginning in Jan.
2003, the CPS question on Hispanic origin was reworded to ask respondents directly whether they are Hispanic. Previously, individuals were identified as Hispanic based on their,
or their ancestors’, country of origin. Hispanics may be of any race.

---

CRS-28
Table A6. Union Membership in the United States by Educational Attainment, 1994-2003
Union
Employed
Percent
Union
Employed
Percent
Union
Employed
Percent
Members
Labor Force
Union
Members
Labor Force
Union
Members
Labor Force
Union
Total
Year
(1,000s)
(1,000s)
Members
(1,000s)
(1,000s)
Members
(1,000s)
(1,000s)
Members
Employed
Labor Force
Less than a High School Education
High School Graduate
1-3 Years of College
2003
1,096
16,499
6.6%
5,008
41,417
12.1%
4,606
39,529
11.7%
137,736
2002
1,174
16,552
7.1%
5,330
41,543
12.8%
4,641
39,177
11.8%
136,485
2001
1,237
17,020
7.3%
5,404
42,018
12.9%
4,855
39,603
12.3%
136,933
2000
1,324
17,450
7.6%
5,534
42,580
13.0%
4,710
39,193
12.0%
136,891
1999
1,323
16,829
7.9%
5,621
41,999
13.4%
4,734
38,179
12.4%
133,488
1998
1,378
17,097
8.1%
5,612
41,718
13.5%
4,696
37,333
12.6%
131,463
1997
1,377
16,773
8.2%
5,762
41,812
13.8%
4,534
36,831
12.3%
129,558
1996
1,503
16,257
9.2%
5,845
40,772
14.3%
4,626
36,615
12.6%
126,708
1995
1,497
15,868
9.4%
5,950
40,489
14.7%
4,692
36,608
12.8%
124,900
1994
1,550
15,807
9.8%
6,245
40,712
15.3%
4,709
35,530
13.3%
123,061
Bachelor’s Degree
Advanced Degree
2003
2,994
26,859
11.1%
2,071
13,431
15.4%
137,736
2002
2,954
26,245
11.3%
2,084
12,968
16.1%
136,485
2001
2,837
25,603
11.1%
1,982
12,690
15.6%
136,933
2000
2,815
25,302
11.1%
1,950
12,366
15.8%
136,891
1999
2,841
24,494
11.6%
1,958
11,986
16.3%
133,488
1998
2,686
23,714
11.3%
1,839
11,601
15.9%
131,463
1997
2,653
23,082
11.5%
1,784
11,060
16.1%
129,558
1996
2,547
22,297
11.4%
1,748
10,767
16.2%
126,708
1995
2,469
21,362
11.6%
1,752
10,574
16.6%
124,900
1994
2,515
20,879
12.0%
1,721
10,134
17.0%
123,061
Source: Calculated by Congressional Research Service (CRS) from the monthly Current Population Survey (CPS).
Notes: Estimates are for persons 16 and over. Details may not add to totals because of rounding.

---

CRS-29
Table A7. Union Membership in the United States in the Private and Public Sectors, 1994-2003
Union
Employed
Percent
Union
Employed
Percent
Members
Labor Force
Union
Members
Labor Force
Union
Total
Year
(1,000s)
(1,000s)
Members
(1,000s)
(1,000s)
Members
Employed
Labor Force
Public
Private
2003
7,324
19,710
37.2%
8,452
118,026
7.2%
137,736
2002
7,387
19,589
37.7%
8,795
116,896
7.5%
136,485
2001
7,095
19,308
36.7%
9,192
117,625
7.8%
136,933
2000
7,059
19,157
36.9%
9,254
117,734
7.9%
136,891
1999
7,058
18,938
37.3%
9,419
114,550
8.2%
133,488
1998
6,905
18,401
37.5%
9,306
113,062
8.2%
131,463
1997
6,747
18,147
37.2%
9,363
111,411
8.4%
129,558
1996
6,854
18,210
37.6%
9,415
108,497
8.7%
126,708
1995
6,927
18,358
37.7%
9,432
106,542
8.9%
124,900
1994
7,091
18,339
38.7%
9,649
104,722
9.2%
123,061
Source: Calculated by Congressional Research Service (CRS) from the monthly Current Population Survey (CPS).
Notes: Estimates are for persons 16 and over. Details may not add to totals because of rounding.

---

CRS-30
Table A8. Union Membership in the United States by Level of Government, 1994-2003
Union
Employed
Percent
Union
Employed
Percent
Union
Employed
Percent
Union
Total
Members
Labor Force
Union
Members
Labor Force
Union
Members
Labor Force
Union
Members,
Employed
Year
(1,000s)
(1,000s)
Members
(1,000s)
(1,000s)
Members
(1,000s)
(1,000s)
Members
Public Sector Labor Force,
Federal Government
State Government
Local Government
Public Sector
2003
1,004
3,247
30.9%
1,706
5,636
30.3%
4,614
10,827
42.6%
7,324
19,710
2002
1,070
3,297
32.5%
1,769
5,706
31.0%
4,547
10,585
43.0%
7,387
19,589
2001
1,033
3,317
31.2%
1,726
5,713
30.2%
4,336
10,278
42.2%
7,095
19,308
2000
1,027
3,275
31.4%
1,624
5,515
29.4%
4,409
10,367
42.5%
7,059
19,157
1999
1,047
3,264
32.1%
1,527
5,233
29.2%
4,484
10,440
42.9%
7,058
18,938
1998
1,105
3,269
33.8%
1,431
5,150
27.8%
4,370
9,982
43.8%
6,905
18,401
1997
1,030
3,217
32.0%
1,485
5,031
29.5%
4,232
9,899
42.7%
6,747
18,147
1996
1,040
3,284
31.7%
1,566
5,132
30.5%
4,249
9,795
43.4%
6,854
18,210
1995
1,117
3,447
32.4%
1,531
5,171
29.6%
4,280
9,739
43.9%
6,927
18,358
1994
1,181
3,518
33.6%
1,596
5,174
30.8%
4,314
9,647
44.7%
7,091
18,339
Source: Calculated by Congressional Research Service (CRS) from the monthly Current Population Survey (CPS).
Notes: Estimates are for persons 16 and over. Details may not add to totals because of rounding.

---

CRS-31
Table A9. Union Membership in the United States by Industry, 1994-2002
Employed
Employed
Employed
Employed
Employed
Union
Percent
Union
Percent
Union
Percent
Union
Percent
Union
Percent
Labor
Labor
Labor
Labor
Labor
Members
Union
Members
Union
Members
Union
Members
Union
Members
Union
Total
Force
Force
Force
Force
Force
Employed
Year
(1,000s)
Members
(1,000s)
Members
(1,000s)
Members
(1,000s)
Members
(1,000s)
Members
(1,000s)
(1,000s)
(1,000s)
(1,000s)
(1,000s)
Labor
Force
Farming, Forestry,
Transportation,
Mining
Construction
Manufacturing
And Fisheries
Communications and Utilities
2002
55
3,458
1.6%
39
491
8.0%
1,367
9,538
14.3%
2,510
18,100
13.9%
2,675
9,769
27.4%
136,485
2001
50
3,332
1.5%
63
565
11.1%
1,388
9,695
14.3%
2,717
19,295
14.1%
2,692
9,778
27.5%
136,933
2000
59
3,554
1.7%
56
527
10.6%
1,390
9,505
14.6%
2,881
20,271
14.2%
2,749
9,882
27.8%
136,891
1999
56
3,393
1.7%
58
557
10.5%
1,362
8,872
15.4%
3,037
20,186
15.0%
2,795
9,591
29.1%
133,488
1998
36
3,446
1.0%
73
624
11.7%
1,212
8,549
14.2%
3,137
20,597
15.2%
2,795
9,380
29.8%
131,463
1997
43
3,527
1.2%
84
637
13.2%
1,223
8,296
14.7%
3,270
20,795
15.7%
2,736
9,170
29.8%
129,558
1996
37
3,567
1.0%
76
568
13.4%
1,158
7,924
14.6%
3,412
20,504
16.6%
2,692
8,792
30.6%
126,708
1995
45
3,562
1.3%
84
649
12.9%
1,072
7,632
14.0%
3,469
20,505
16.9%
2,770
8,725
31.8%
124,900
1994
49
3,552
1.4%
102
681
15.0%
1,110
7,475
14.8%
3,533
20,170
17.5%
2,847
8,690
32.8%
123,061
Finance, Insurance,
Private Household
Retail and Wholesale Trade
Professional Services
Public Administration
and Real Estate
and Other Services
2002
1,158
28,151
4.1%
183
9,093
2.0%
599
16,836
3.6%
5,588
34,832
16.0%
2,010
6,217
32.3%
136,485
2001
1,232
28,112
4.4%
199
8,912
2.2%
692
17,249
4.0%
5,331
33,860
15.7%
1,950
6,134
31.8%
136,933
2000
1,261
28,278
4.5%
179
8,850
2.0%
665
16,961
3.9%
5,198
33,022
15.7%
1,898
6,039
31.4%
136,891
1999
1,291
27,591
4.7%
201
8,770
2.3%
638
16,177
3.9%
5,179
32,384
16.0%
1,860
5,966
31.2%
133,488
1998
1,295
27,192
4.8%
205
8,568
2.4%
578
15,716
3.7%
4,920
31,492
15.6%
1,960
5,899
33.2%
131,463
1997
1,329
26,759
5.0%
216
8,288
2.6%
524
15,418
3.4%
4,856
30,945
15.7%
1,827
5,724
31.9%
129,558
1996
1,343
26,636
5.0%
230
8,110
2.8%
589
14,755
4.0%
4,871
30,062
16.2%
1,861
5,790
32.1%
126,708
1995
1,410
26,130
5.4%
195
7,975
2.4%
571
14,060
4.1%
4,834
29,683
16.3%
1,909
5,978
31.9%
124,900
1994
1,392
25,618
5.4%
217
8,114
2.7%
613
13,854
4.4%
4,914
29,037
16.9%
1,963
5,870
33.4%
123,061
Source: Calculated by Congressional Research Service (CRS) from the monthly Current Population Survey (CPS).
Notes: Estimates are for persons 16 and over. Details may not add to totals because of rounding.

---

CRS-32
Table A10. Union Membership in the United States by Occupation, 1994-2002
Union
Employed
Percent
Union
Employed
Percent
Union
Employed
Percent
Total
Members
Labor Force
Union
Members
Labor Force
Union
Members
Labor Force
Union
Employed
Year
(1,000s)
(1,000s)
Members
(1,000s)
(1,000s)
Members
(1,000s)
(1,000s)
Members
Labor
Managerial and Professional
Technical, Sales, and Administrative
Service
Force
2002
4,812
42,546
11.3%
3,204
38,945
8.2%
2,264
19,250
11.8%
136,485
2001
4,658
42,162
11.0%
3,208
39,573
8.1%
2,277
18,752
12.1%
136,933
2000
4,566
41,223
11.1%
3,122
39,891
7.8%
2,242
18,513
12.1%
136,891
1999
4,594
40,351
11.4%
3,191
38,851
8.2%
2,151
18,089
11.9%
133,488
1998
4,252
38,845
10.9%
3,239
38,754
8.4%
2,209
17,895
12.3%
131,463
1997
4,208
37,738
11.2%
3,158
38,342
8.2%
2,141
17,491
12.2%
129,558
1996
4,196
36,437
11.5%
3,231
37,735
8.6%
2,103
17,227
12.2%
126,708
1995
4,116
35,187
11.7%
3,364
37,530
9.0%
2,112
16,947
12.5%
124,900
1994
4,102
33,772
12.1%
3,465
37,334
9.3%
2,222
16,909
13.1%
123,061
Precision Production, Craft, and Repair
Operators, Fabricators, and Laborers
Farming, Forestry, and Fishing
2002
2,570
14,587
17.6%
3,243
17,674
18.3%
89
3,484
2.6%
136,485
2001
2,730
15,139
18.0%
3,346
17,985
18.6%
96
3,321
2.9%
136,933
2000
2,800
15,107
18.5%
3,509
18,683
18.8%
95
3,474
2.7%
136,891
1999
2,800
14,540
19.3%
3,627
18,265
19.9%
113
3,392
3.3%
133,488
1998
2,708
14,364
18.9%
3,713
18,173
20.4%
90
3,432
2.6%
131,463
1997
2,723
14,112
19.3%
3,791
18,393
20.6%
88
3,483
2.5%
129,558
1996
2,648
13,578
19.5%
4,000
18,182
22.0%
92
3,548
2.6%
126,708
1995
2,692
13,504
19.9%
3,983
18,106
22.0%
91
3,626
2.5%
124,900
1994
2,716
13,478
20.2%
4,132
17,946
23.0%
103
3,623
2.8%
123,061
Source: Calculated by Congressional Research Service (CRS) from the monthly Current Population Survey (CPS).
Notes: Estimates are for persons 16 and over. Details may not add to totals because of rounding.

---

CRS-33
Table A11. Union Membership in the United States by Region, 1994-2003
Employed
Employed
Employed
Employed
Employed
Union
Percent
Union
Percent
Union
Percent
Union
Percent
Union
Percent
Labor
Labor
Labor
Labor
Labor
Total
Members
Union
Members
Union
Members
Union
Members
Union
Members
Union
Employed
Year
Force
Force
Force
Force
Force
(1,000s)
Members
(1,000s)
Members
(1,000s)
Members
(1000s)
Members
(1,000s)
Members
(1,000s)
(1,000s)
(1,000s)
(1,000s)
(1,000s)
Labor
Force
East South Central
Pacific
Mountain
West South Central
New England
2003
556
7,863
7.1%
3,326
21,786
15.3%
621
9,210
6.7%
763
14,955
5.1%
884
7,162
12.3%
137,736
2002
619
7,720
8.0%
3,338
21,187
15.8%
646
9,045
7.1%
782
14,557
5.4%
931
7,261
12.8%
136,485
2001
619
7,738
8.0%
3,137
21,441
14.6%
702
9,021
7.8%
839
14,646
5.7%
910
7,206
12.6%
136,933
2000
676
7,921
8.5%
3,070
21,535
14.3%
729
8,944
8.2%
820
14,654
5.6%
921
7,172
12.8%
136,891
1999
641
7,825
8.2%
3,203
21,193
15.1%
727
8,530
8.5%
867
14,589
5.9%
977
6,917
14.1%
133,488
1998
644
7,811
8.2%
3,105
20,705
15.0%
664
8,314
8.0%
815
14,312
5.7%
950
6,875
13.8%
131,463
1997
628
7,632
8.2%
2,994
20,192
14.8%
692
8,140
8.5%
826
13,978
5.9%
901
6,802
13.3%
129,558
1996
681
7,539
9.0%
2,938
19,570
15.0%
660
7,790
8.5%
871
13,703
6.4%
895
6,661
13.4%
126,708
1995
706
7,453
9.5%
3,092
19,230
16.1%
694
7,645
9.1%
823
13,472
6.1%
963
6,585
14.6%
124,900
1994
739
7,334
10.1%
3,130
19,066
16.4%
684
7,349
9.3%
852
13,131
6.5%
926
6,532
14.2%
123,061
South Atlantic
East North Central
West North Central
Middle Atlantic
2003
1,572
25,465
6.2%
3,478
22,222
15.7%
1,107
10,273
10.8%
3,467
18,801
18.4%
137,736
2002
1,515
24,954
6.1%
3,618
22,445
16.1%
1,137
10,305
11.0%
3,596
19,011
18.9%
136,485
2001
1,602
25,083
6.4%
3,682
22,745
16.2%
1,169
10,257
11.4%
3,656
18,796
19.4%
136,933
2000
1,627
25,165
6.5%
3,715
22,733
16.3%
1,158
10,074
11.5%
3,618
18,692
19.4%
136,891
1999
1,588
24,134
6.6%
3,729
22,327
16.7%
1,211
9,900
12.2%
3,533
18,074
19.5%
133,488
1998
1,582
23,664
6.7%
3,791
21,937
17.3%
1,140
9,891
11.5%
3,520
17,954
19.6%
131,463
1997
1,545
23,294
6.6%
3,738
21,800
17.1%
1,168
9,761
12.0%
3,617
17,960
20.1%
129,558
1996
1,607
22,786
7.1%
3,824
21,504
17.8%
1,197
9,657
12.4%
3,596
17,498
20.6%
126,708
1995
1,543
22,465
6.9%
3,746
21,307
17.6%
1,149
9,473
12.1%
3,644
17,269
21.1%
124,900
1994
1,694
22,103
7.7%
3,789
20,967
18.1%
1,135
9,269
12.2%
3,792
17,310
21.9%
123,061
Source: Calculated by Congressional Research Service (CRS) from the monthly Current Population Survey (CPS).
Notes: Estimates are for persons 16 and over. See Appendix A for a list of states by region. Details may not add to totals because of rounding.

---

CRS-34
Appendix B: Data and Methodology
The analysis in this report uses data from the monthly Current Population
Survey (CPS). The CPS is a household survey conducted by the U.S. Bureau of the
Census for the Bureau of Labor Statistics (BLS) of the U.S. Department of Labor.
The monthly CPS is the main source of labor force data for the nation, including
estimates of the monthly unemployment rate. The CPS collects a wide range of
demographic, social, and labor market information. Each month, approximately
50,000 households are contacted to be interviewed, either in person or by phone. The
CPS collects labor force data for civilians 15 and over. The official definition of the
civilian labor force is ages 16 and over. The monthly CPS sample is representative
of the civilian noninstitutional population; it does not include persons on active
military duty.58,59
Each month, one-fourth of the CPS sample — called the Outgoing Rotation
Group, or ORG — is asked questions about union membership and current hourly or
weekly earnings. The monthly CPS has included questions on union membership
and union coverage since November 1982.60 For the tables in Appendix A, the ORG
samples for each month from 1994 to 2003 were combined to calculate a monthly
average for the year.61
The analysis in this report examines employed persons ages 16 and over.
Employed persons include both wage and salary workers and self-employed persons.
Data on union membership and coverage exclude self-employed persons. Data are
for the sole or main job of full-time and part-time workers.
Several changes were made in the January 2003 CPS. In answering the question
about race, respondents may now pick more than one race. Previously, individuals
could only select one race. For 2003, this report follows BLS practice and only
counts blacks and whites who selected one race category. Also, beginning in 2003,
the CPS question on Hispanic origin was reworded to ask respondents directly
whether they are Hispanic. Previously, individuals were identified as Hispanic based
on their, or their ancestors’, country of origin. Hispanics may be of any race. As a
result of these changes, data for 2003 on race and Hispanic ethnicity are not directly
comparable to data for earlier years.
58 U.S. Bureau of the Census, Measuring 50 Years of Economic Change, Current Population
Reports, P60-203, Sept. 1998, p. D-1.
59 U.S. Department of Labor, Bureau of Labor Statistics, Basic Monthly Survey, available
at [http://www.bls.census.gov/cps/bglosary.htm].
60 U.S. Department of Labor, Bureau of Labor Statistics, Current Population Survey:
Design and Methodology, Technical Paper 63, Mar. 2000, pp. 2-4.
61 Households are in the CPS survey for four consecutive months, out of the survey for eight
months, and back in the survey for four more months. The questions about earnings (and
union status and hours worked) are asked of households leaving the survey (either
permanently or for eight months). During a 12-month period, the observations on earnings
are for unique individuals.

---

CRS-35
In addition, in January 2003, the CPS introduced population controls based on
the 2000 Census. Sample weights for January 2000 through December 2002 were
revised to reflect the higher population estimates from the 2000 census and the higher
rate of population growth since the census. This report uses the revised sample
weights for 2000-2002. The revised weights increase the size of the labor force but
have less of an effect on percentage calculations.
Finally, in 2003, new classification systems were introduced for industry and
occupation. Because of these new systems, data on industry and occupation for 2003
are not comparable to data for earlier years, and are not included in this report.62
The regional data in Figure 11 and Table A11 are based on state groupings
used by the Census Bureau. The Census Bureau divides the United States into four
regions and nine divisions. The nine divisions are as follows:
! East North Central: Wisconsin, Indiana, Ohio, Michigan, and
Illinois;
! East South Central: Tennessee, Kentucky, Alabama, and
Mississippi;
! Middle Atlantic: Pennsylvania, New York, and New Jersey;
! Mountain: Colorado, Arizona, Idaho, Montana, Utah, Nevada,
Wyoming, and New Mexico;
! New England: Massachusetts, Maine, Vermont, New Hampshire,
Connecticut, and Rhode Island;
! Pacific: Alaska, Washington, Hawaii, Oregon, and California;
! South Atlantic: Maryland, Delaware, Georgia, District of Columbia,
Florida, North Carolina, Virginia, South Carolina, and West
Virginia;
! West North Central: Minnesota, Kansas, South Dakota, Missouri,
North Dakota, Nebraska, and Iowa; and
! West South Central: Texas, Louisiana, Oklahoma, and Arkansas.
Confidence Levels
The comparisons discussed in the text of this report are statistically significant
at the 95% confidence level. Estimates based on survey responses from a sample of
households have two kinds of error: nonsampling error and sampling error.
Examples of nonsampling error include information that is misreported and errors
made in processing collected information. Sampling error occurs because a sample,
and not the entire population, of households is surveyed. The difference between an
estimate based on a sample of households and the actual population value is known
as sampling error.63 When using sample data, researchers typically construct
confidence intervals around population estimates. Confidence intervals provide
62 Mary Bowler, Randy E. I.G., Stephen Miller, Ed Robison, and Anne Polivka, Revisions
to the Current Population Survey Effective in Jan., 2003, Employment and Earnings, Feb.
2003, vol. 51, pp. 4-5, 7.
63 U.S. Department of Labor, Bureau of Labor Statistics, Employment and Earnings, vol. 49,
Nov. 2002, pp. 147-148.

---

CRS-36
information about the accuracy of estimated values. With a 95% confidence interval
and repeated samples from a population, 95% of intervals will generally include the
actual value of a population characteristic.

---