Study on Women and Privatization

INSTITUTE FOR
WOMEN’S POLICY RESEARCH
Why Privatizing
Government
Services Would
Hurt Women Workers
By Annette Bernhardt, Ph.D.
Laura Dresser, Ph.D.
Why Privatizing
Government Services
Would Hurt Women Workers
By
Annette Bernhardt, Ph.D.
Laura Dresser, Ph.D.
Center on Wisconsin Strategy
INSTITUTE FOR WOMEN’S POLICY RESEARCH
About This Report
Why Privatizing Government Services Would Hurt
Women Workers was made possible by generous funding
from The American Federation of State, County, and
Municipal Employees (AFSCME).
Acknowledgements
The authors wish to thank Catherine Hill, Ph.D., for
editing this report and writing the executive summary.
Additional thanks goes to Dennis Houlihan, AFSCME,
for his comments on an earlier draft of this report.
About the Institute
for Women’s Policy Research
The Institute for Women’s Policy Research (IWPR) is a
scientific research organization dedicated to
informing and stimulating debate on public policy issues
of critical importance to women and their
families.
IWPR focuses on issues of poverty and welfare, employment
and earnings, work and family issues, the economic
and social aspects of health care and safety, and
women’s civic and political participation.
The Institute works with policymakers, scholars, and
public interest groups to design, execute, and disseminate
research and to build a network of individuals and
organizations that conduct and use women-oriented
policy research.
IWPR, an independent, nonprofit organization, also
works in affiliation with the graduate programs in
public policy and women’s studies at The George
Washington University.
IWPR’s work is supported by foundation grants,
government grants and contracts, donations from
individuals, and contributions from organizations and
corporations. IWPR is a 501(c)(3) tax-exempt
organization.
IWPR # B237 $5.00
© Copyright 2002 by the Institute for Women’s
Policy Research, Washington, DC. All rights
reserved. Printed in the United States of America.
INSTITUTE FOR WOMEN’S POLICY RESEARCH
1707 L Street NW, Suite 750 • Washington, DC 20036
P: (202) 785-5100 • F: (202) 833-4362
www.iwpr.org
Executive Summary .............................................................................. iv
Background ............................................................................................ 1
Data and Measures ................................................................................ 3
Trends in Public and Private Sector Employment Over Time ................. 4
Worker Characteristics ........................................................................... 6
Teachers and Non-Traditional Jobs ........................................................ 8
Health and Pension Benefits .................................................................. 9
Wages .................................................................................................... 9
Wage Regressions by Sex ................................................................... 12
Wage Regressions by Race and Sex ...................................................14
“At Risk” Occupations ...........................................................................15
Summary ..............................................................................................17
References ........................................................................................... 19
Figures & Tables
Figure 1 Percentage of Workers Who Work for an Employer Who
Sponsors a Pension Plan Compared with the Percentage
of Workers Enrolled in a Pension Plan ............................................. 5
Figure 2 Percent of Wage and Salary Workers
Employed in the Public Sector, by Race and Sex ........................... 5
Table 1 Sample Sizes ................................................................................... 3
Table 2 Employment of Wage and Salary Workers, by Sector ..................... 6
Table 3 Characteristics of the Workforce, by Sector .................................... 7
Table 4 Access to Non-Traditional Job ......................................................... 8
Table 5 Pension and Health Care Coverage ................................................ 9
Table 6 Median Wages by Education, Sex, and Sector .............................. 10
Table 7 Wages by Race/Ethnicity ............................................................... 11
Table 8 Regression Results, by Sex .......................................................... 13
Table 9 Regression Results, by Sext and Race ........................................ 14
Table 10 Employment in Six “At Risk” Occupations, Women Only .............. 15
Table 11 Median Hourly Wages in “At Risk”
Occupations, Women Only ($1998)............................................... 16
Table 12 Wage Regressions for “At Risk” Occupations, Women Only ........ 17
Table of Contents
Why Privatizing Government Services Would Hurt Women Workers
iv
This report analyzes the implications of privatization for women workers, especially
those employed in low-end occupations. Data analyzed show that women
disproportionately depend on the public sector for jobs that pay decent wages and offer
benefits. This is especially true for African American and Hispanic women, and for women
who do not have a college education. In part, higher wages and better access to health and
pension benefits in the public sector can be attributed to higher rates of union coverage.
The evidence suggests that privatizing government services will have a negative impact on
women workers, especially those workers who are most vulnerable.
A Relative Decline in Public Sector Employment
Over the past two decades, employment in the public sector has grown much more slowly
than employment in the private sector. In 1979, 16 percent of working men held public
sector jobs, while in 1998 only 13 percent did. For women, the share employed in the public
sector dropped from 20 percent in 1979 to 18 percent in 1998 (See Figures 1 and 2).
• The proportion of workers employed in the public sector declined for both
women and men between 1979 and 1998, with especially pronounced declines
for African American and Hispanic workers.
• In 1998 almost one in five women held a public job (18 percent), a higher rate
than among men (13 percent). This was especially true for African American
women (22 percent).
Public sector employment declined for both women and men between 1979 and 1998
with a somewhat sharper decline among men.
Public Sector Employees Have Higher Wages
and Better Access to Health and Pension Benefits
Focusing on 1998, the most current year for which data is available, the report shows
that median earnings in the public sector are higher than in the private sector for most
categories of workers.
• Median wages for women without a four-year college degree are 15 percent higher in
the public sector. For women with a college degree, wages in the public sector are 7
percent higher than in the private sector.
• Among women, 72 percent of public workers participate in a pension plan and 69
percent have employer-provided health insurance. By contrast, in the private sector
less than half have either benefit, and in the case of Hispanic women, less than a third
do (see Table 5).
Thus, privatization is likely to erode the wages and benefits of women workers, especially
for African American and Hispanic women and those with fewer years of formal education.
Explaining Why Wages in the Public
Sector are Higher than in the Private Sector
Wage ratios shown above are useful descriptions of employment conditions, but they can
also be misleading because they do not account for differences in the public and private
Executive Summary
Institute for Women’s Policy Research
v
sector workforces. For example, workers in the public sector tend to be older and hence
could be expected to have higher earnings. This study uses regression analysis to distinguish
components of the wage differential that can be attributed to the public/private sector
distinction. Overall, controlling for race, region, and work experience, women in the
public sector are still more likely to have higher wages than their counterparts in the
private sector. While non-college educated men, particularly African American and
Hispanic men, also benefit from somewhat higher wages and benefits in the public sector,
the effect is more pronounced for women workers. For example, women without a
college degree in the public sector earn wages that are 5 to 6 percent above the earnings of
their private sector counterparts.
Once union membership and occupation variables were included, however, the differences
in public sector and private sector wages largely disappeared. In other words, for women
who have the same occupation, union status, education, work experience, and race, the
public sector does not, on average, pay better than the private sector. Unionization emerges
as a central factor in understanding why the public sector pays better than the private sector.
Gender Equity in the Public and Private Sectors
While women are paid better in the public sector in an absolute sense, a gap between
men’s and women’s wages remains (see Table 6). The size of the wage gap between male
and female workers in public and private sector employment varies by race and educational
background.
• The gender wage gap is smaller in the public sector, especially for women of
color, but this result is driven largely by education. Only women with college
degrees see greater pay equity in the public sector. For less educated women,
gender inequality is as great in the public sector as in the private.
The public sector is generally regarded as providing better access to professional and
managerial jobs for women. When teachers are separated from other professional and
managerial occupations, however, the public sector does not appear to provide greater
opportunity for women to hold managerial or professional positions.
• With the exception of teachers, women are only slightly more likely to hold
managerial, technical, or other professional jobs in the public sector than in the
private sector (26.8 percent and 25.3 percent, respectively).
Because women in the public sector have more education than women in the private
sector, we would expect them to be well represented in managerial and professional
positions. This is not the case, however. That educational wealth has not translated into
greater numbers of managerial jobs for women (while it has for men) indicates the
continued presence of occupational barriers for women. This report finds that the public
sector does not, in general, offer exceptional opportunities for women to hold managerial
and professional positions (although other research by IWPR suggests that the public
sector does, in fact, offer better opportunities for women of color than the private sector).
Risks of Privatization
The second part of this research focuses on occupations — such as health care and child
care workers, janitors, food preparation employees, and clerical and administrative staff
— that are considered to be “at risk” for privatization. This research finds that women
Why Privatizing Government Services Would Hurt Women Workers
vi
working in these “at risk” occupations have less education, with close to half holding a
high school degree or less. Because the wage differential in the public and private sectors
is largest for women without a college education, these women have the most to lose
under privatization.
• For women without college degrees, occupations “at risk” for privatization
constitute 63.9 percent of their public sector jobs.
• Even though “at risk” occupations are not generally considered exceptional job
opportunities, these jobs pay better in the public sector than in the private sector.
The Upshot
From a policy standpoint, there is good reason to be concerned about the continuing call
for contracting out public services to the private sector. On average, public sector jobs
pay better and are more likely to include pension and health benefits. When government
services are privatized, women—especially women of color and women who do not have
a college education—will likely experience significant declines in how much they earn
and in their health and pension coverage. Even though, the public sector is far from a
perfectly fair employer—glass ceilings and the gender gap in pay and benefits persist as in
the private sector—this analysis finds that privatization, and the de-unionization that
frequently accompanies it, is likely to prove detrimental to the economic welfare of
women workers.
Institute for Women’s Policy Research
1
Over the past two decades, privatization, in which businesses (or not-for-profit agencies)
provide public services, has changed the nature of public sector employment. A growing
number of public functions that previously were performed by government employees
are now performed by private businesses. While the privatization of public services is not
new, its prevalence has grown steadily over the past two decades, resulting in a decline in
public sector employment as a percentage of the overall workforce.
The impetus for privatization of government services is to lower the costs of public
services and improve their quality (Sawicky 1999; Mason and Siegel 1997). The literature
on privatization is dominated by case studies where cost reduction was the main goal.
More often than not, the researchers claim that cost savings were achieved.
There are two major problems with this research approach. First, as Sawicky (1999)
argues, studying only governments that have privatized their services does not tell us
what the impact would be under different circumstances and in different settings. Private
contractors often “cherry pick,” taking over lucrative services (such as well traveled bus
routes and safer districts) and leaving less profitable services to the public sector (Barnekov
and Raffel 1990). For example, a private school might not accept a disabled student, but
a public school would, because it is legally required to accommodate all children. As a
result, the public school might appear less efficient. In other words, the problem of
comparing “apples and oranges” looms large in the privatization debate.
A second problem with the literature on privatization is that studies rarely address how
cost savings are achieved. In particular, few analysts document whether cost savings are
achieved by lowering workers’ wages and benefits (as opposed to improvements in
efficiency). Case study research suggests that reductions in workers’ wages and benefits
can be an important, if not the only, component in the cost savings achieved by
privatization (Schlar 2000).
This study re-examines the impact of privatization on the wages and benefits of workers
with particular attention to women workers. The public sector has historically had smaller
gender and racial wage gaps, better overall pay, and higher numbers of women at
professional levels than the private sector (Sawicky 1999). It also has substantially higher
union density (Hirsch and Macpherson 1999), which translates into greater job stability
and due-process rights—the latter may be of particular use to women and minorities. In
addition, because benefits in the private sector have eroded in recent years, the public
sector advantage in benefit-provision has increased (Belman and Haywood 1997; National
Commission on Employment Policy 1989). Finally, Freeman (1996) found that employee
involvement and job satisfaction are higher in the public sector, partly because of a greater
willingness by managers to share power. Consequently, privatization may also affect nonfinancial
aspects of job quality.
Much popular rhetoric about public sector employment focuses on raw differences in
compensation and benefits between the public and private sectors, which can be substantial
(Sclar 2000). Raw differences, however, are potentially misleading because they fail to
control for differences in the composition of the workforce across the two sectors. The
public-sector workforce, for example, is substantially better educated than the privatesector
workforce.1 It is also substantially more unionized.
Background
1 Belman and Heywood (1993) found that approximately 20 percent of private sector workers have a college education,
compared with 44 percent of public sector workers.
Why Privatizing Government Services Would Hurt Women Workers
2
Because there are differences of opinion regarding what variables should be controlled
for, estimates of the difference between public and private sector compensation are often
not consistent across studies. For example, Belman and Heywood (1993) found that the
pay of federal public employees was roughly equal to the wages of comparable private
sector employees; other studies have found public pay levels to be higher (Cox and Brunelli
1992; Peterson, Davis, and Walker 1986).2 All approaches, however, find a decline in
public pay relative to private pay in recent years. The pace of this decline has varied
depending on macroeconomic and political conditions (Freeman 1985), and also varies
regionally (Belman and Heywood 1995). Another point of agreement is that wages among
workers in the public sector are more equal than they are in the private sector (see Sawicky
1999). Part of this is likely due to the greater presence of unions, which tend to equalize
wages among their members.
There is also agreement that well educated workers tend to earn more in the private
sector, whereas less educated workers tend to earn more in the public sector.3 Thus,
privatization is likely to affect less-skilled, low-wage workers more adversely (Sclar 2000;
Poterba and Rueben 1994). Low-wage government workers who have been laid-off and
who get jobs with a private contractor often lose their benefits, and their wages often
(but not always) decline (National Commission on Employment Policy 1989; Suggs
1989). Minority workers may be disproportionately affected, especially if they have less
seniority in the civil service system (Suggs 1989).
Research Objectives
This study addresses the following questions: What are the implications of privatization
for women workers, especially those employed in low-end occupations? Are there marked
differences by race or ethnicity, suggesting that privatization might have particularly negative
effects on women of color? What factors might explain differences between public and
private sector employment?
Direct answers are not possible because there are no datasets to identify and track all of
the employment changes that can result from privatization. The Current Population
Survey’s bi-annual supplement on “alternative work arrangements” includes contract
workers including independent contractors, temporary help agency workers, on-call
workers, and contract company employees. However, “contract workers” are not
synonymous with the workers impacted when government services are contracted. For
example, if mass transit spending decreases, private companies may emerge to fill the
void. While the mechanism is different than explicit subcontracting (and much more
nebulous), the effect is the same: the public sector shrinks relative to the private sector,
and wages may decline as a result.
This paper takes the simple route of comparing wages and other outcomes in the public
sector to those in the private sector in order to get at the potential impact of privatization
on women. This indirect approach has the advantage of addressing the potential impact
of all forms of privatization, not simply direct subcontracting.
This study has two components. First, public sector employment growth relative to the
private sector over the last two decades is documented based on data from the Current
Population Survey (CPS). Second, using the most recent year available (1998), the report
2 In the extreme, one study found that the public/private wage gap evaporates when the higher education levels for public
sector workers are considered (Gold and Ritchie 1992).
3 This study finds that college educated women earn more in the public sector than in the private sector.
Institute for Women’s Policy Research
3
“maps” the differences in job quality between the public and private sectors. Particular
attention is paid to clerical and service occupations, since these workers are likely to lose
the most from privatization (relatively speaking) and are the most likely to be subcontracted
in the first place. Throughout, close attention is paid to differences by sex and race, since
the public sector has historically provided especially strong opportunities to women of
color. (The focus on differences by race, however, is limited at points by insufficient
sample sizes.)
The Current Population Survey (CPS), the government’s monthly household survey of
employment and labor markets, yields a nationally representative sample of individual
workers with employment variables such as earnings, hours worked, industry, occupation,
education, and unionization, as well as background variables such as age, sex, race, ethnicity,
and geographic location. In order to achieve sufficient sample sizes within any one year,
the Merged Outgoing Rotation Group (ORG) version of these data is used.4 The exception
is the analysis of pension and health care coverage, where the March 1998 CPS Supplement
was used (since this is the only month where such data can be found). Sample selection
was straightforward. The study includes workers aged 18-64 who had positive earnings
and who were not self-employed. Final sample sizes are given in Table 1.
Some of the variables used in these analyses are basic and do not require description (for
example, sex), but others require elaboration. The key variable of interest, sector, consists
of three categories: public, private for-profit, and private non-profit. For convenience,
throughout this report the terms public, private, and non-profit are used. The key outcome
variable, hourly wages, is constructed as follows: if an individual actually reported an
hourly wage, that value is used; otherwise, usual earnings per week are divided by usual
weekly hours worked. The resulting measure is hourly wages at the respondent’s main
job, in 1998 dollars. Part-time work is defined as less than 35 hours per week, again using
the usual weekly hours measure.
Data and Measures
Table 1. Sample Sizes
Women Men
All All
Workers White Black Hispanic Workers White Black Hispanic
Public 13,497 9,599 2,370 519 10,546 7,815 1,458 801
Private 56,205 40,187 7,561 2,562 67,913 48,712 7,065 9,125
Non Profit 5,989 4,845 675 179 2,688 2,109 294 147
Total 75,691 54,631 10,606 3,260 81,147 58,636 8,817 10,073
Source: Authors’ analysis of the 1998 Merged Outgoing Rotation Group, Current Population
Survey.
4 The ORG files effectively combine monthly samples into yearly samples. Specifically, each household entering the CPS
is administered four monthly interviews, then ignored for eight months, then interviewed again for four more months.
Since 1979, only households in months four and eight have been asked their usual weekly earnings/usual weekly hours.
These are the outgoing rotation groups, and each year the BLS gathers all these interviews together into a single Merged
Outgoing Rotation Group.
Why Privatizing Government Services Would Hurt Women Workers
4
The race variable makes use of questions about race as well as ethnicity and differentiates
among the following groups: non-Hispanic white, non-Hispanic black, Hispanic, and
non-Hispanic others. In the text, the ‘non-Hispanic’ prefix is dropped, but the reader
should remember that the race categories have been defined in this manner. The education
variable uses information on highest grade attended, and consists of three groups: high
school graduates and drop-outs; workers with some college experience but no four-year
college degree; and workers with a four-year college degree or higher. The union variable
measures whether a worker is a union member. For occupation, most often the Census’
one-digit occupational coding is used, collapsed for brevity as needed. To identify
occupations most at risk for privatization, two-digit and three-digit codes are used. For
health care coverage, the question asks whether the respondent is currently covered by a
health plan provided through their current employer, former employer or union. For
pension coverage, the question asks whether the respondent is currently participating in
the employer’s or union’s pension plan. Both of these variables, therefore, measure actual
participation in employer-provided benefits.
At the end of the paper, several regressions are performed in which the natural log of the
dependent variable, hourly wage, is used. In these regressions, an indirect measure of years
of work experience is used, since no direct measure exists. “Potential” experience is calculated
as the respondent’s age, minus years of education minus six. This formula roughly
approximates the number of years that an individual has worked in the labor market,
though it does a better job for men than it does for women (who are more likely to take
time out to care for children or elderly parents). Finally, all analyses are weighted with
CPS earner weights, normalized to represent the actual sample sizes (rather than the
whole population).
Figure 1 shows the percent of all wage and salary workers employed in the public sector
from 1979 to 1998, for men and women separately.5 Women have always been more
likely to hold jobs in the public sector than men, and this pattern continues to the
present. Both sexes, however, have seen a slow decline over the past 20 years. The trend is
more pronounced among men. In 1979, 15.9 percent of men held public sector jobs,
while in 1998 only 13.2 percent did so, for a decline of 17 percent over the past two
decades. Women saw their public jobs go from 19.8 to 18.0 percent over that same
period, for a 9 percent decline over this same period. For both men and women, there is
clear evidence of the relative reduction in public sector jobs. Although the absolute number
of both private and public jobs has grown, as the former has grown faster than the latter,
the percentage of public jobs dropped. These trends cannot be directly tied to an increased
use of subcontracting, the drop in relative public sector employment is certainly consistent
with it.
Figure 2 takes a closer look at the experience of different racial and ethnic groups. Most
striking are the high levels of public sector employment for African American men and
women. At least in part, this is likely due to stricter adherence to equal opportunity laws
in the public sector. It is worrisome that African American workers have seen very strong
drops in public employment over time, stronger than those seen for white workers. For
example, the percent of public workers has dropped by 16 percent for African American
Trends in Public and Private
Sector Employment Over Time
5 We do not show trends for non-profit employment because this sector is not consistently identified across all years.
Institute for Women’s Policy Research
5
women over the past two decades, compared with seven percent for white women. And
the drop for African American men (18 percent) also outpaced that of white men (13
percent). These are disturbing figures for a minority group that has historically found
some of its best opportunities for employment in this sector (see also Sawicky 1999).
Hispanic workers have relatively low levels of public sector employment, both historically
and currently. Moreover, the drop in such employment for Hispanic men is the strongest
of any group, declining by an astonishing 35 percent (as compared with nine percent for
Figure 1. Percent of Wage and Salary Workers
Employed in the Public Sector, by Sex
0
5
10
15
20
25
30
1979
1981
1983
1985
1987
1989
1991
1993
1995
1997
Year
Percent
Men
Women
Women
0
5
10
15
20
25
30
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
W hite Black H ispanic
Men
0
5
10
15
20
25
30
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
Figure 2. Percent of Wage and Salary Workers
Employed in the Public Sector, by Race and Sex
Why Privatizing Government Services Would Hurt Women Workers
6
Hispanic women). Given these trends, the public sector is unlikely to prove a major
source of stable, well-paid jobs for Hispanic workers, especially men, in the future.
Finally, Table 2 takes a closer look at the most current information on employment
across the sectors, including the not-for-profit sector. Despite the shedding of public
sector jobs, women of all races and ethnicities are still more likely to hold jobs outside
the private for-profit sector than their male counterparts. African American workers of
both sexes are more likely to hold public jobs than whites. On the other hand, Hispanic
workers are consistently underrepresented outside of the private for-profit sector. It is
worth reiterating that of all these groups, African American women have the highest level
of public sector employment (22.4 percent). This figure will become especially relevant
when wage differences between the sectors are examined below.
Demographic and occupational characteristics of the workforce vary significantly across
the three sectors of the American economy. Table 3 reviews these basic characteristics of
the public and private labor force. Compared with workers in the private sector, those in
the public sector have significantly higher levels of education. For example, 45.9 percent
of women in the public sector hold a four-year college degree, compared with only 19.2
percent in the private. (Some, but not all, of this differential is due to the large numbers
of teachers in the public sector.) On the other end of the spectrum, about a quarter of
women in the public sector have a high school diploma or less, compared with nearly
half in the private sector. This pattern holds true for men as well as women.
Public workers also tend to be older, reflecting in part the job stability and seniority rules
that stem from greater union strength in this sector. In fact, the difference in union
coverage between the sectors is striking. For women, 36.1 percent are union members in
the public sector, compared with only 6.1 percent in the private sector. The numbers for
men are 41.2 percent and 13.3 percent, respectively. While the absolute figure for men is
higher, it is worth noting that women have the most to gain: they are six times more
likely to be unionized in the public sector than in the private sector, while for men
Table 2. Employment of Wage
and Salary Workers, by Sector
Women Men
Public Private Non-Profit Public Private Non-Profit
All workers 17.8% 74.3% 7.9% 100 % 13.0 % 83.7% 3.3% 100%
White 17.6 73.6 8.9 100 13.3 83.1 3.6 100
African American 22.4 71.3 6.4 100 16.5 80.1 3.3 100
Hispanic 14.0 81.9 4.0 100 8.0 90.6 1.5 100
Other 15.9 78.6 5.5 100 13.0 83.2 3.8 100
Source: Authors’ analysis of the 1998 Merged Outgoing Rotation Group, Current Population
Survey.
Worker Characteristics
Institute for Women’s Policy Research
7
unionization is just three times as likely. Much of this difference is likely a function of
high unionization rates for teachers (who are disproportionately women) in the public
sector.
Perhaps because of this strong union presence, part-time work is less prevalent in the
public sector, though the differential is not as large as one might expect. In the public
sector 17.7 percent of women work part-time compared with 22.6 percent in the private
sector. Of course, rates of part-time work are much lower among men, with 6.7 percent
of men working part-time in the public sector compared with 7.2 percent in the private
sector. Given the relationship between full-time employment and benefits, the greater
access to full-time jobs in the public sector suggests greater access to benefits as well.
In terms of occupational segregation, there are greater differences between women and
men. In the private sector, women tend to be concentrated in service and clerical
occupations and men tend to be concentrated in manufacturing-related occupations. In
the public sector, however, the pattern shifts somewhat. As noted above, public workers
tend to be well educated, and this is clearly reflected in the types of jobs they hold. For
men, public sector employment means far fewer manufacturing jobs and many more
professional, managerial, and protective service jobs. For women, it means greater numbers
of professional and clerical jobs, and fewer in service and sales (which are traditionally
found at the bottom of service industries). Overall, while occupational segregation is
apparent across these sectors, the public sector contains a higher proportion of managerial
and professional jobs for women than does the private sector.
Table 3. Characteristics of the Workforce, by Sector
Women Men
Public PrivateNon-Profit Public PrivateNon-Profit
High school or less 27.2% 48.2% 25.7% 27.6% 50.8% 22.6%
Some college experience 26.9 32.7 33.2 29.8 27.2 24.3
College degree 45.9 19.2 41.1 42.6 22.0 53.2
100.0 100.0 100.0 100.0 100.0 100.0
Ages 18 to 30 19.5 35.3 24.2 19.6 34.3 25.2
Ages 31 to 54 69.0 56.5 63.8 67.5 57.3 63.1
Ages 55 to 64 11.5 8.2 12.1 12.9 8.4 11.8
100.0 100.0 100.0 100.0 100.0 100.0
Percent working part-time 17.7 22.6 27.1 6.7 7.2 13.3
Percent unionized 36.1 6.1 8.0 41.2 13.3 7.3
Executive, administrative, managerial 11.9 13.3 15.0 14.3 12.5 17.4
Professional (including teachers) 39.5 9.9 37.0 27.4 9.3 42.9
Technicians & related support 2.6 4.0 6.7 3.7 3.1 4.8
Administrative support, including clerical 28.9 24.3 21.2 11.4 5.7 6.5
Protective service 2.6 .4 .2 17.2 1.1 1.5
Service and sales 11.5 34.4 17.3 8.0 19.6 13.5
Precision production, craft, & repair .4 2.5 .2 8.5 20.9 5.6
Operators, fabricators, & laborers 5.4 10.4 2.1 7.9 25.0 6.1
Farming, forestry, & fishing .2 .8 .2 1.7 2.9 1.8
100.0 100.0 100.0 100.0 100.0 100.0
Source: Authors’ analysis of the 1998 Merged Outgoing Rotation Group, Current Population
Survey.
Why Privatizing Government Services Would Hurt Women Workers
8
At first blush, it might appear that the public sector is a watershed of sorts for women,
breaking down barriers to non-traditional occupations. But a closer look reveals that
much of this is a function of the large numbers of teachers in the public sector. Table 4
summarizes our occupational data in a different way, by distinguishing between three
groups. The first consists of managerial, professional, and technical occupations except for
teachers (for example, administrators and officials, engineers, computer scientists, lawyers
and judges, and all management-related occupations). The second consists of teachers
(primary, secondary, and post-secondary). The third consists of all other occupations,
typically front-line and non-supervisory jobs (for example, administrative assistants, data
processors, welfare and childcare workers, and building maintenance occupations).
When viewed this way, it quickly becomes clear that the public sector does not generally
provide exceptional opportunities for women. Although the public sector contains a very
high percentage of teachers, which women fill disproportionately, it would be a stretch
to call teaching a non-traditional job for women. Occupations that have truly been outside
of the reach of many women—managerial, technical, and other professions—are nearly
as hard to get in the public sector as in the private (26.8 percent versus 25.3 percent,
respectively). It is men, however, who have the best access, with close to a third (31.1
percent) holding high-end, non-teaching jobs in the public sector, as compared with only
24.2 percent in the private sector.
This point, however, should not be overstated. After all, the general conclusion of Table
4 is that far fewer women have front-line, non-supervisory jobs in the public sector.
Many more hold jobs that are well paid, challenging, have greater autonomy, and require
greater skill. There is no doubt that women are better off, in an absolute sense, working
in the public sector. But it is worth remembering that much of this “boosting” is
accomplished through teaching jobs that women have held throughout much of the 20th
century. This is relevant because women in the public sector are much better educated
than in the private sector (as are men), with nearly half holding four-year college degrees
(even a bit more than men). That such educational wealth has not translated into greater
numbers of managerial jobs for women (while it has done so for men) indicates the
continuing presence of other barriers for women.6
Teachers and Non-Traditional Jobs
Table 4. Access to Non-Traditional Jobs
Women Men
Public Private Non-Profit Public Private Non-Profit
Managers, professionals, technicians 26.8% 25.3% 47.6% 31.1% 24.2% 56.0%
Teachers 27.2 2.0 11.1 14.3 .7 9.0
Front-line, non-supervisory jobs 46.0 72.8 41.3 54.6 75.2 34.9
100.0 100.0 100.0 100.0 100.0 100.0
Source: Authors’ analysis of the 1998 Merged Outgoing Rotation Group, Current Population
Survey.
6 To push the point, female-dominated professions such as teaching typically pay less than male-dominated professions
such as the law, even though comparable education levels are often required (Baker and Fortin 1999).
Institute for Women’s Policy Research
9
Another key measure of worker welfare is the extent to which workers receive employeror
union-provided health care and pension benefits. Table 5 shows that benefits provision
differs enormously across the public and private sectors.7 All workers, regardless of sex,
race, or ethnicity, have significantly higher rates of pension and health care coverage in the
public sector.
Among women, 71.6 percent of public workers participate in a pension plan and 69.0
percent have health insurance. By contrast, in the private sector less than half of women
have either of these benefits, and in the case of Hispanic women, less than a third do.
These are striking differences—public jobs clearly give a significant boost to worker
welfare.
The public sector advantage in benefit provision looks quite similar for men, but with
the difference that men are more likely to have pension and health coverage, regardless of
sector. To a large extent, this reflects men’s greater unionization rates in both sectors,
which, in turn, is a function of the types of jobs men tend to hold. Another central factor
limiting women’s participation in pension plans is their greater propensity to work parttime
(see Shaw and Hill 2002).
The final and perhaps most important question is how wages compare across the different
sectors of the economy. Educational and occupational differences aside, the bottom line
for most workers is how much their jobs pay. This section describes differences in the
wages earned by men and women in the public and private sectors as well as differences
between the two sectors for African American, Hispanic, and white workers. While the
Health and Pension Benefits
Wages
7 For this analysis we can only use the March Supplement of the CPS (see Data section). As a result, the sample sizes for
the non-profit sector are too small to report reliable estimates.
Table 5. Pension and Health Care Coverage
Women Men
Private, Private,
Public for-profit Public for-profit
Percent with pension plan 71.6% 37.7% 78.4% 45.3%
White 71.7 39.5 79.9 49.7
Black 73.5 39.3 75.6 40.6
Hispanic 67.9 24.0 77.1 26.8
Percent with health insurance 69.0 47.2 78.6 59.3
White 68.7 48.4 79.1 63.4
Black 71.3 52.9 81.6 55.2
Hispanic 67.2 32.4 73.5 42.8
Source: Authors’ analysis of the March 1998 Current Population Survey Supplement.
Why Privatizing Government Services Would Hurt Women Workers
10
degree to which this differential reflects the public/private sector distinction is best obtained
through regression techniques (presented in the next section), a broad overview of the
patterns in wages and benefits is useful.
As shown in Table 6, with the exception of college-educated men, all workers receive
higher wages in the public sector. For both men and women, employment in the public
sector pays off—and it pays off the most at lower education levels. Thus, the public/
private ratio is the largest for workers who do not hold a four-year college degree. By
contrast, the ratio is significantly lower for college graduates. In fact, men with a college
degree would be better served by finding jobs in the private sector, as their wages are
higher, on average.
The general conclusion cannot be overstated: public sector employment provides the
most benefits to “less-skilled” workers—those who do not have access to the labor market
power that a college degree confers. It is likely that high unionization rates in the public
sector, particularly for less skilled men, drive this conclusion. As noted earlier, unions
have been found to compress the wage gap between front-line and supervisory workers.
Thus, privatization is likely to hit hardest precisely those workers, male and female, who
are the most vulnerable in the labor market: high school dropouts, high school graduates,
and workers who have attended college but did not attain a four-year degree.
Wages and Sex
Another question concerns women’s standing relative to men. Here, the findings are a bit
more complicated. In the final column of Table 6, we show the ratio of women’s to
men’s median wages in each sector. All of the ratios are less than one, meaning that
women’s wages are consistently lower than men’s. This is not surprising. There are, however,
important sector differences. Among well-educated workers (those with a college degree
Table 6. Median Wages by Education, Sex, and Sector
Ratio of Ratio of
Median Wages ($1998) Public/Private Wages Women’s to
Education Women Men Women Men Men’s Wages
High School or Less
Public $ 9.23 $12.35 1.15 1.24 .75
Private $ 8.00 $10.00 .80
Non-Profit $ 8.03 $ 8.90 .90
Some college
Public $11.00 $14.00 1.15 1.14 .79
Private $ 9.60 $12.25 .78
Non-Profit $10.80 $11.05 .98
College
Public $16.78 $19.74 1.07 0.96 .85
Private $15.63 $20.65 .76
Non-Profit $15.37 $18.00 .85
Source: Authors’ analysis of the 1998 Merged Outgoing Rotation Group, Current Population
Survey.
Institute for Women’s Policy Research
11
or higher), women’s wages are 85 percent of men’s in the public sector, but only 76
percent of men’s in the private sector.
For workers with less education, however, the picture is less promising. The gender
difference in wages is about the same for workers with some college experience, regardless
of which sector they work in. And for those with a high school diploma or less, the
gender gap in wages is actually greater in the public sector than in the private. Why? It is
true that women with less education have the most to gain from getting public sector
jobs, in an absolute sense. But this is also true for men with low education levels, in fact,
even more so than for women. Both men and women benefit from public sector jobs,
but men gain more. They are pulling away from women in a relative sense, widening the
gender gap.
What is the upshot? When considering the likely effects of privatization on women
working in the public sector, the bottom-line consideration has to be in terms of “dollarsper-
hour.” First and foremost, we should be concerned about absolute wage levels. The
data indicate that women—especially those with less education—have much to lose from
privatization; their wages are higher in the public sector than in the private.
A second consideration is the likely effect of privatization on women’s standing relative
to men. Here the data indicate that women with college degrees will see a widening of
the gender gap if their jobs are moved into the private sector. Among less educated workers,
the gender gap will remain unchanged in the private sector, or even shrink.
Wages and Race
As noted earlier, there are strong differences in the extent to which different race and
ethnic groups depend on the public sector as a source of jobs. Table 7 addresses whether
there are also differences in the wages those jobs pay. The first part of the table shows the
ratio of public to private wages, broken down by sex, education, and race/ethnicity.
Several themes emerge.
First, it is clear that non-whites tend to benefit the most from employment in the public
sector. This is especially striking for African American and Hispanic men, who consistently
Table 7. Wages by Race/Ethnicity
Women Men
White Black Hispanic White Black Hispanic
Public/Private Wage Ratio
High school or less 1.16 1.11 1.18 1.15 1.22 1.30
Some college 1.10 1.23 1.22 1.11 1.28 1.40
College 1.05 1.17 1.24 .93 1.19 1.13
Ratio of Each Group’s Wages to White Men’s Wages
Public Sector .81 .73 .68 .83 .86
Private Sector .73 .60 .55 .73 .64
Source: Authors’ analysis of the 1998 Merged Outgoing Rotation Group, Current Population
Survey.
Why Privatizing Government Services Would Hurt Women Workers
12
see a larger difference in pay between public and private sector jobs than do white men. A
similar pattern can be seen among women, though it is not as striking or as pervasive.
These findings do not mean that by some magical twist, African American and Hispanic
workers are paid better than white workers in the public sector. Far from it—white men
still earn the most, whichever sector one examines. However, because minorities are paid
so much less than whites in the private sector, they have more to gain from employment
in the public sector.
The second part of Table 7 illustrates this point. For this analysis, we computed the ratio
of each group’s wages to white men’s wages (hence there is no entry for white men).
Significantly, nowhere is the ratio equal to one—white men always earn more. But for all
groups, the public sector does, in fact, reduce the wage gap. This is especially true for
Hispanic men and women, where the gap is reduced substantially. The reduction for
African American men and women is also sizeable. The reduction for white women,
while still sizeable, is the smallest.
Straightforward wage tables like those can be illuminating, but they can also be misleading
in that they do not adequately control for differences in the composition of the workforce
across the two sectors. The public sector workforce, for example, is older than the private
sector workforce, and thus has more years of work experience. Work experience is a very
strong predictor of good wages, so the public sector may pay better simply because its
workers have more experience. It is important, therefore, to control for age and education
as well as other variables, and we do so by using standard regressions.
Some researchers have also argued that unionization rates and occupational composition
should be controlled for. Again, the public sector may pay better simply because it contains
more professional and managerial occupations and because more of its workforce is
unionized. Higher unionization rates, however, are precisely one of the causal, substantive
mechanisms that lead to higher wages in the public sector. We don’t want to negate this
effect from the outset; rather, we want to measure it. Likewise, for occupation, we do
not want to erase these differences in composition at the onset. True, public jobs are
likely to pay better, in part because they are more highly skilled jobs. But that’s precisely
the point, and not one that should be erased, at least not from the outset.
We therefore conduct two sets of regressions.8 Model 1 estimates average hourly wages
for men and women separately, in the public and private sectors, by education group.
This model controls for experience (and its square), race/ethnicity, and region. We consider
the estimates from this model to be our true, baseline estimates of the public/private
wage differential. Model 2 then adds unionization and occupation into the regression to
get a sense of how much these characteristics of the public sector help explain its higher
wages. A critical point to remember is that education is already controlled for by the
time occupation is introduced into the model—so the occupation effect is not a proxy
for skill differences.
Table 8 presents the results from the two models. The table does not show all estimates
for all variables, but rather focuses on the issue of interest here: how much better or
8 Specifically, OLS regressions run separately for men and women, using weights, where the dependent variable is ln
(wages).
Wage Regressions by Sex
Institute for Women’s Policy Research
13
worse public wages are, relative to private wages. “Better or worse” are expressed in terms
of percentages. For example, the first entry in Table 8 indicates that for women, the
average wages of high school graduates and dropouts are 6.7 percent higher in the public
sector than in the private sector.9 Estimates that are statistically significant are marked
with an asterisk.
For high school dropouts and high school graduates, the public sector premium is 6.7
percent for women as compared to only 3.7 percent for men. For workers with some
college experience, the premium is 5.5 percent for women and only 3.9 percent for men.
And among the most educated, those with a college degree or higher, the premium for
women is positive, while it is actually negative for men, and strongly so.
Furthermore, women with less education benefit the most from public sector jobs, with
wages that are about 5 percent to 6 percent higher than in the private sector. The increase
for college-educated women, 3 percent, is not nearly as strong. These regressions therefore
confirm our descriptive analysis of wages in the previous section.
Model 2 adds controls for unionization and occupation, showing their dramatic impact.
For women, these two variables eliminate the public wage premium altogether. Thus
broadly speaking, women earn more in government jobs because they are more likely to
be unionized and because the occupations in the public sector are higher-end occupations.
More precisely, for women who have the same demographic characteristics (age, race,
and so on) and hold the same type of job (occupation, union status), the public sector
does not, on average, pay better than the private sector. This confirms results from
previous research (Belman and Heywood 1993; Smith 1977).
But what does this mean from a policy standpoint? Model 1 tells us that a woman
without much education – say, a high school degree – will earn significantly more in the
public sector. Model 2 then helps explain why. First, the public sector offers betterpaying
occupations, even for someone with only a high school degree, and second, those
9 These estimates were obtained with a full interaction between sector and education.
Table 8. Regression Results, by Sex
Percent Increase or Decrease in Wages in the Public Sector,
Relative to the Private Sector
Model 1 Model 2
Education Women Men Women Men
High School or Less 6.7* 3.7* 1.3 3.3*
Some college 5.5* 3.9* -1.1 5.0*
College 3.0* -8.0* -10.4* -15.0*
R-Squared .298 .339 .397 .401
*Significant at .05 level.
Note: Model 1 controls for experience, experience squared, race, and region. Model 2 also
controls for occupation and unionization status.
Source: Authors’ analysis of the 1998 Merged Outgoing Rotation Group, Current Population
Survey.
Why Privatizing Government Services Would Hurt Women Workers
14
jobs are more likely to be unionized. From the vantage point of an average woman
looking for a job, this sounds like a good bet—especially if, like most workers, she
doesn’t hold a four-year college degree. And from that same vantage point, the threat of
privatization looms large, because it means lower-pay occupations, less union coverage,
and, in the end, lower wages.
How do these overall trends play out by race and ethnicity? Table 9 provides the results
of estimating Model 1 within each race/ethnic group. Before examining the results,
however, a word is necessary about significance levels in this table. When each group is
considered separately, sample sizes are smaller and statistical significance becomes a bit
harder to achieve. Although this point should not be overplayed (the really large effects in
the table are all significant), it is important to keep in mind that some borderline effects
in this table could have been statistically significant with larger sample sizes.10
That said, what is remarkable in these results is the strong and significant wage premium
that African American and Hispanic workers garner in the public sector, regardless of sex.
Depending on education level, African American workers gain between 8.3 percent and
11.4 percent in wages when they hold public jobs. With one exception, Hispanic workers
gain between 12.8 percent and 16.3 percent (the exception being college educated men).
These are quite substantial increases for minority groups that have traditionally had a
much harder time than whites getting good jobs with good wages.
White workers, by contrast, experience much smaller wage gains from public sector
employment. In fact, white men don’t benefit at all; they do better in the private sector.
White women do have higher wages in the public sector, but the difference is on a much
smaller scale than African American or Hispanic women, ranging from 1.3 percent to 4.9
percent. Only for high school graduates and dropouts is this increase significant.
Wage Regressions by Race and Sex
Table 9. Regression Results, by Sex, and Race
Percent Increase or Decrease in Wages in the Public Sector,
Relative to the Private Sector
White Black Hispanics
Education Women Men Women Men Women Men
High School or Less 4.9* -0.3 8.3* 9.5* 14.6* 14.9*
Some college 2.5 -0.3 10.1* 11.4* 12.8* 14.7*
College 1.3 -11.0* 7.9* 11.3* 16.3* 8.9
R-Squared .285 .325 .314 .245 .260 .235
*Significant at .05 level.
Note: Model 1 results, controlling for experience, experience squared, and region.
Source: Authors’ analysis of the 1998 Merged Outgoing Rotation Group, Current Population
Survey.
10 We do not show the results of estimating Model 2 separately for each race/ethnic/sex group, because the effect of
controlling for occupation and unionization is the same for all workers combined.
Institute for Women’s Policy Research
15
Non-white workers have more to gain from public jobs than white workers, both in an
absolute and relative sense. The converse, of course, is that they have more to lose if the
decline of public sector employment growth continues. In fact, what is worrisome is that
African American and Hispanic workers have seen the strongest decline over the past 20
years in public sector jobs, precisely the ones that offer them the most wage gains.
Up to this point, the analysis has proceeded as if all parts of the public sector were equally
vulnerable to the threat of privatization, regardless of function or service performed. Yet
this is clearly not the case. The best information on which types of public services are being
privatized comes from the International City/County Management Association (ICMA)
survey. While this survey has clear limitations (federal agencies are not included, for example),
it provides a general guide as to which public services are most vulnerable.
ICMA data from 1988, 1992, and 1997 show considerable variability in privatization. For
example, waste management has seen significant increases in subcontracting over the past
10 years, while public safety and parks maintenance have gone virtually untouched. But
few women actually work in waste management. The analysis presented here focuses on
those occupations “at risk” of privatization that also employ significant numbers of women.
A close examination of the ICMA data reveal five such “at risk” occupational clusters:
health care workers; child care, welfare, and social workers; janitors and maintenance
workers; bus drivers and parking lot attendants; and food preparation and food service
workers. Another occupation, clerical and administrative support, has been added because
it employs very large numbers of women. While this occupation has not yet been privatized
to a large degree, there may be increasing pressure to do so in the future, since it is one of
the few remaining public service areas untouched by privatization.
Table 10 shows women’s employment in these six “at risk” occupations across the three
sectors. Some of the occupations are more prevalent in the private sector (perhaps reflecting
“At Risk” Occupations
Table 10. Employment in Six “At Risk” Occupations,
Women Only
Public Private Non-Profit
Health Services 1.2 % 2.6 % 5.9 %
Child Care, Welfare, Social Workers 4.6 1.2 6.3
Janitors and Maintenance 2.2 2.5 2.3
Bus Drivers & Parking Lot Attendants 1.6 .7 .3
Food Preparation & Food Services 3.3 6.9 2.2
Clerical & Administrative Support 27.8 23.6 20.7
Percent of women employed in
“At Risk” occupations 40.6 37.5 37.7
Percent of women without a college
degree in “At Risk” occupations 63.9 41.6 51.0
Source: Authors’ analysis of the 1998 Merged Outgoing Rotation Group, Current Population
Survey.
Why Privatizing Government Services Would Hurt Women Workers
16
privatization that has already occurred during the past several decades). Others are more
so in the public and non-profit sectors (the latter may again be evidence of past
privatization). In any event, this analysis targets the potential effects of future privatization
on those women who currently work in the public sector.
These “at risk” occupations constitute 40.6 percent of women’s public jobs, and 12.9
percent excluding clerical occupations. These numbers are somewhat misleading, however.
For women without a college degree, the “at risk” occupations constitute a much larger
portion of public jobs (63.9 percent), even if clerical occupations are not counted (19.2
percent). Thus, for the population that is of greatest policy concern (i.e., “unskilled”
workers), future privatization of the above occupations can have a truly strong impact.
Another way to make this point is to look at workers employed in these “at risk”
occupations and look at their educational attainment. Not surprisingly, women in “at
risk” occupations are significantly less educated than the overall population of women
working in the public sector. Close to half (47.2 percent) hold a high school diploma or
less, and fewer than one out of six hold college degrees (14.9 percent)—quite a contrast
to the results shown at the outset in Table 3.
The main question, of course, is how wages compare across the sectors. The top part of
Table 11 shows that, in general, public wages are higher in these “at risk” occupations
than in the private sector with a few exceptions. For example, women who work in
health services are actually paid better in the private sector, and in clerical occupations and
maintenance jobs the public sector advantage is not all that pronounced. In the bottom
of Table 11, we calculate the public/private wage ratio for all six occupations combined.11
While there clearly remains a public wage advantage, it is not as strong as for other public
sector jobs.
An accurate estimate of the public/private differential for these “at risk” occupations requires
that we again turn to regression techniques. In Table 12, we re-estimate Model 1 from
Table 11. Median Hourly Wages in “At Risk” Occupations,
Women Only ($1998)
Public/Private
Public Private Non-Profit Ratio
Health Services $11.55 $11.77 $12.50
Child Care, Welfare, Social Workers 11.05 6.50 9.00
Janitors and Maintenance 7.15 6.70 7.00
Bus Drivers & Parking Lot Attendants 10.08 8.00 7.40
Food Preparation & Food Services 7.00 5.75 6.24
Clerical & Administrative Support 10.40 9.83 10.00
All “At Risk” Occupations 10.00 8.75 9.93 1.14
All Other Occupations 15.38 10.00 13.40 1.54
Source: Authors’ analysis of the 1998 Merged Outgoing Rotation Group, Current Population
Survey.
11 Small sample sizes do not allow us to calculate the ratio for each individual occupation.
Institute for Women’s Policy Research
17
the previous section, but this time only for the six “at risk” occupations. Results show
that public wages are consistently higher than private wages, ranging from 3.6 percent to
8.2 percent higher. (Because of small sample sizes, we are hesitant to give any interpretation
to this variability or to significance levels.)
In sum, in public sector occupations that are most vulnerable to privatization, women
stand to see significant wage losses if their jobs are subcontracted or outsourced. These
potential losses are not as strong as for other occupations. It is worth pointing out,
however, that even a 5 percent difference in wages is not insignificant for jobs that, in the
private sector, are paying under $10.00 per hour. This low on the wage scale, any change
in wages can have a significant impact on the economic welfare of the worker and her
family.
Finally, we should note that race/ethnicity breakdowns were not possible in this analysis
because of small sample sizes. We suspect, however, that our findings on “at risk”
occupations would be especially marked for African American and Hispanic women.
Indeed, research at the Joint Center for Political Studies found that African Americans
tend to be employed in occupations that are more likely to be contracted out (originally
cited in Sawicky 1999).
Even though the public sector has been shrinking over the past two decades, women are
still more likely to hold public jobs than men. Therefore, examining how privatization
and subcontracting will likely affect women workers remains a critical issue.
We find that the public sector holds a significant advantage in both wages and benefits
over the private sector, especially for African American and Hispanic women. Wages are
higher in the public sector for two reasons: first, the public sector is more likely to be
unionized; and second, it offers better-paying occupations, regardless of skill level. From
the perspective of the average woman looking for a job, especially, if like most workers,
she doesn’t hold a four-year college degree, public jobs are good jobs. From her vantage
point, the threat of privatization looms large because it means lower-pay occupations,
less union coverage, and,in the end, lower wages and fewer benefits.
Table 12. Wage Regressions for “At Risk”
Occupations, Women Only
Percent Increase or Decrease in Wages in the
Public Sector, Relative to the Private Sector
High School or Less 5.3*
Some college 3.6
College 8.2*
R-Squared .159
*Significant at .05 level.
Note: Model 1 results, controlling for experience, experience squared, race, and region.
Source: Authors’ analysis of the 1998 Merged Outgoing Rotation Group, Current Population
Survey.
Summary
Why Privatizing Government Services Would Hurt Women Workers
18
If women are paid better in the public sector in an absolute sense, does this mean they
also gain ground relative to men? Not necessarily. True, the gender wage gap is smaller in
the public sector, especially for women of color, but this is driven largely by education.
Only women with college degrees see greater pay equity in the public sector. For less
educated women, gender inequality is as great in the public sector as in the private.
We also find mixed results in examining the common perception that the public sector
offers women better access to non-traditional “professional” occupations. In fact, much
of this perception is due to the preponderance of women teachers in the public sector.
Once teachers are taken out of the equation, women are no more likely to hold managerial,
technical, or other professional jobs in the public sector than in the private. This is an
important finding, because women in the public sector are unusually well educated. That
this educational wealth has not translated into more managerial jobs for women indicates
the persistence of occupational segregation and “glass ceilings” in the public sector as well
as the private.
Finally, we focus on “at risk” occupations—jobs that have historically been privatized and
subcontracted. Not surprisingly, they tend to be less-skilled and lower pay occupations,
such as health care and child care workers, janitors, and food preparation employees. For
women without college degrees, these occupations constitute the mainstay of their public
sector employment. Even though these “at risk” occupations are generally considered
low-rung, they still pay better in the public sector than in the private sector. So, even
when we consider only jobs that are the most likely to be privatized, we find that the
impact on women’s wages would be detrimental, especially for the less educated.
In sum, from a policy standpoint, there is good reason to be concerned about the
continuing call for leaner government and the contracting out of as many public services
as possible. Women depend disproportionately on the public sector for jobs that pay a
living wage and that provide benefits. If these jobs are privatized, women—especially
women of color and less educated women—will see significant declines in how much
they earn and in their access to health and pension coverage. This does not mean that the
public sector is a cure-all for inequality: glass ceilings and the gender gap persist in the
public as in the private sector. But, the bottom line is that privatization, and the
deunionization that invariably accompanies it, can only prove detrimental to women’s
economic welfare.
Institute for Women’s Policy Research
19
Baker, Michael and Nicole M. Fortin. 1999. “Topics in Labor Economics: Women’s
Wages in Women’s Work—A U.S./Canadian Comparison of the Roles of Unions
and ‘Public Goods’ Sector Jobs.” American Economic Review May:198-203.
Belman, Dale and John S. Heywood. 1993. “Job Attributes and Federal Wage
Differentials.” Industrial Relations 32(1):148-157.
Belman, Dale and John S. Heywood. 1993. “The Truth about Public Employees:
Underpaid or Overpaid?” Briefing Paper. Economic Policy Institute, Washington,
DC.
Belman, Dale and John S. Heywood. 1995. “State and Local Government Wage
Differentials: An Intrastate Analysis.” Journal of Labor Research 16(2):187-201.
Belman, Dale and John S. Heywood. 1997. “Changes in the Relative Provision of Public-
Sector Pensions.” Public Finance Review 25(4):426-441.
Cox, W. and S. A. Brunelli. 1992. American’s Protected Class: Why Excess Public Employee
Compensation is Bankrupting States. Washington, DC: American Legislative
Exchange Council.
Freeman, Richard. 1985. “How do Public Sector Wages and Employment Respond to
Economic Conditions?” NBER Working Paper #1653. National Bureau of
Economic Research, Cambridge, MA.
Freeman, Richard. 1996. “Through Public Sector Eyes: Employee Attitudes toward Public
Sector Labor Relations in the U.S.” In Dale Belman, Morley Gunderson, and
Douglas Hyatt, eds. Public Sector Employment in a Time of Transition. Madison,
WI: Industrial Relations Research Association.
Gold, Steven D. and Sarah Ritchie. 1992. Compensation of State and Local Employees:
Sorting Out the Issues. Albany, NY: National Commission on State and Local
Public Service.
Hirsch, Barry and David Macpherson. 1999. Union Membership and Earnings Data
Book. Washington, DC: Bureau of National Affairs.
International City/County Management Association. 1999. The Municipal Year Book.
Washington, DC: ICMA.
Mason, Maryann and Wendy Siegel. 1997. Does Privatization Pay? A Case Study of
Privatization in Chicago. Chicago, IL: Chicago Institute on Urban Poverty Policy.
National Commission for Employment Policy. 1989. The Long Term Employment
Implications of Privatization: Evidence from Selected U.S. Cities and Counties. A
study prepared by Dudek and Company for the National Commission for
Employment Policy. Washington, DC: National Commission for Employment
Policy.
References
Why Privatizing Government Services Would Hurt Women Workers
20
Needleman, Ruth. 1994. “The Devaluation and Privatization of Public Service: Race
and Gender Implications.” Prepared for the Fourth Annual Women’s Policy
Research Conference, June 3-4.
Office of Management and Budget, Office of Federal Procurement Policy. 1994.
Summary Report of Agencies’ Service Contracting Practices. Washington, DC.
Osborne, D. and T. Gaebler. 1992. Reinventing Government: How the Entrepreneurial
Spirit is Transforming the Public Sector. Reading, MA: Addison-Wesley.
Poterba, James M. and Rueben, Kim S. 1994. “The Distribution of Public Sector Wage
Premium: New Evidence Using Quantile Regression.” NBER Working Paper
#4734. National Bureau of Economic Research, Cambridge, MA.
Sawicky, Max. 1999. Privatization: An Exploration of the Literature. Washington, DC:
Economic Policy Institute.
Sclar, Elliott. 2000. You Don’t Always Get What You Pay For: The Economics of Privatization.
Ithaca NY: Cornell University Press.
Shall, Lois and Catherine Hill. 2002. The Gender Gap in Pension Coverage: What Does
the Future Hold? Washington, DC: Institute for Women’s Policy Research.
Smith, Sharon. 1977. “Government Wage Differentials.” Journal of Urban Economics 4:
248-271.
Suggs, Robert E. 1989. Minorities and Privatization: Economic Mobility at Risk. Washington
DC: Joint Center for Political Studies Press.
U.S. General Accounting Office. 1985. Comparison of Federal and Private Sector Pay and
Benefits. Washington, DC.
INSTITUTE FOR WOMEN’ S POLICY RESEARCH
1707 L Street, NW, Suite 750 • Washington, DC 20036
Tel: (202) 785-5100 • Fax: (202) 833-4362 • www.iwpr.org