)=7<5;;?@6@A@2+<>86;4%.=2>? )=7<5;&2?2.>055<:2=.42
(523320@?<3(2:=<>.>D'2>B602?.;1<;@>.0@6;4$A@<;!<C(523320@?<3(2:=<>.>D'2>B602?.;1<;@>.0@6;4$A@<;!<C
'869921+<>82>?B612;023><:A@<'A==962>?<[email protected]?.;1'869921+<>82>?B612;023><:A@<'A==962>?<[email protected]?.;1
%A/960'05<<9?%A/960'05<<9?
2<>42>608028
+)=7<5;;?@6@A@23<>:=9<D:2;@&2?2.>05
2>608028A=7<5;<>4
'A?.;#<A?2:.;
+)=7<5;;?@6@A@23<>:=9<D:2;@&2?2.>05
5<A?2:.;A=7<5;<>4
>;2! .992/2>4
(52);6B2>?6@D<3#<>@5.><96;..@5.=29699
)=7<5;;?@6@A@2+<>86;4%.=2>#<
6@.@6<;6@.@6<;
>6080282<>42'A?.;#<A?2:.;.;1>;2! .992/2>4(523320@?<3(2:=<>.>D'2>B602?
.;1<;@>.0@6;4$A@<;!<C'869921+<>82>?B612;023><:A@<'A==962>?<[email protected]?.;1%A/960
'05<<9?)=7<5;;?@6@A@2+<>86;4%.=2>#< .9.:.E<<"+)=7<5;;?@6@A@23<>:=9<D:2;@
&2?2.>055@@=?1<6<>4C=
(56?@6@926?/><A45@@<D<A/D@52)=7<5;;?@6@A@2<>:<>26;3<>:.@6<;=92.?20<;@.0@>2=<?6@<>DA=7<5;<>4
(523320@?<3(2:=<>.>D'2>B602?.;1<;@>.0@6;4$A@<;!<C'869921+<>82>?(523320@?<3(2:=<>.>D'2>B602?.;1<;@>.0@6;4$A@<;!<C'869921+<>82>?
B612;023><:A@<'A==962>?<[email protected]?.;1%A/960'05<<9?B612;023><:A@<'A==962>?<[email protected]?.;1%A/960'05<<9?
A@5<>?
2<>42>608028
+)=7<5;;?@6@A@23<>:=9<D:2;@&2?2.>05
'A?.;#<A?2:.;
+)=7<5;;?@6@A@23<>:=9<D:2;@&2?2.>05
>;2! .992/2>4
(52);6B2>?6@D<3#<>@5.><96;..@5.=29699
)=7<5;A@5<>?$&12;@6F2>
5@@=?<>061<>4
%A/96?521*2>?6<;
;!<C+.42:2>60.<C:=9<D2>?.>2&2?5.=6;4$==<>@A;6@D6;@52+<>8=9.02216@21/D6922;
==29/.A:;;2@@22>;5.>1@.;1&605.>1"A>;.;2#2C,<>8&A??299'.42<A;1.@6<;==
(56?C<>86;4=.=2>6?.B.69./92.@)=7<5;&2?2.>055@@=?>2?2.>05A=7<5;<>4A=-C<>86;4=.=2>?
The Effects of Temporary Services and Contracting Out
on Low-Skilled Workers: Evidence from
Auto Suppliers, Hospitals, and Public Schools
Upjohn Institute Staff Working Paper No. 03-90
George Erickcek, Susan Houseman, and Arne Kalleberg
April 2002
revised July 2002
JEL Classification Codes: J31, J32, J41, J42, J45, J51, J81, K31
Paper prepared for the Case Studies Conference sponsored by the Russell Sage Foundation and the
Rockefeller Foundation, May 2–3, 2002, in New York. We thank Lillian Vesic-Petrovic and Peter
Einaudi for their assistance with data analysis, and Claire Black for her transcription of all interviews
and assistance in preparing the document.
The Effects of Temporary Services and Contracting Out on Low-Skilled Workers:
Evidence from Auto Suppliers, Hospitals, and Public Schools
Abstract
We examine why employers use temporary agency and contract company workers and the
implications of these practices for the wages, benefits, and working conditions of workers in low-skilled
labor markets. Through intensive case studies in manufacturing (automotive supply), services (hospitals),
and public sector (primary and secondary schools) industries, we define the circumstances under which
these workers are likely to be adversely affected, minimally affected, or even benefitted by such
outsourcing. Adverse effects on compensation are clearest when companies substitute agency temporaries
or contract company workers for regular employees on a long-term basis because low-skilled workers
within the organization receive relatively high compensation and employment and labor law or workers and
their unions do not block companies from such substitution. Often, however, organizations only contract
out management functions or utilize agency temporaries for brief periods of time, with little direct effect on
in-house, low-skilled workers. Moreover, employers often use temporary agencies to screen workers for
permanent positions. Because temporary agencies lower the cost to employers of using workers with poor
work histories or other risky characteristics, agencies may benefit these workers by giving them
opportunities to try out for positions they otherwise might not have had.
Temporary help employment grew dramatically over the last decade, accounting for ten percent
of net employment growth in the United States during the 1990s. Although government statistics on
contracting out are not maintained, evidence from case studies and business surveys suggests that there has
been dramatic growth in the outsourcing of functions to outside companies as well (Abraham and Taylor
1996; Houseman 2001b; Kalleberg, Reynolds, and Marsden 1999). In the both cases, the workers who
provide services to the client firm are not the client’s employees, but rather are the legal employees of the
temporary help agency or the contract company. Through intensive case studies in manufacturing
(automotive supply), services (hospitals), and public sector (primary and secondary public schools)
industries, we endeavor to shed light on why employers increasingly are using these nonstandard
employment arrangements and their implications for wages, benefits, and working conditions of workers
in low-skill labor markets.
Because workers in these arrangements often receive lower compensation than they would if they
were employees of the client organization, the growth of temporary help and contracting out generally is
viewed as inimical to workers’ interests. We find, however, that the story is not that simple. Our case
study evidence points to circumstances under which workers are likely to be adversely affected by the
outsourcing of jobs to agencies or contractors. These cases entail the long-term substitution of agency
temporary or contract company workers for regular employees and the loss of wages, benefits, or union
status.
In other situations, however, the effects on low-skilled workers appear to be minimal or even
beneficial. Often, organizations only contract out the management function, which has little direct effect on
low-skilled workers. Agency temporaries typically do not substitute for regular workers on a long-term
basis. Moreover, employers often use temporary agencies to screen workers for permanent positions.
2
Because temporary agencies reduce the cost to employers of using workers with poor work histories or
other risky characteristics, agencies may benefit these workers by giving them opportunities to “try out” for
positions they otherwise might not have had.
Throughout our discussion, we point to the important impacts employment and labor law, unions,
and other institutional factors have on (1) employers’ incentives to use temporary agencies or contract
companies, (2) the constraints employers face in using these employment forms, and (3) the implications
of such outsourcing for low-skilled workers.
BACKGROUND ON ECONOMIC CONDITIONS
The primary data for these case studies were collected in 1999 and 2000. During this time period,
all three industries faced tight labor markets for low-skilled workers. Moreover—although the sources of
the pressure differed across industries—automotive suppliers, hospitals, and public schools were all facing
intense pressure to cut costs. Automotive suppliers made intensive use of temporary help services and
public schools contracted out functions, while hospitals used a combination of temporary services and
contracting out. We argue that the utilization of temporary services or contract companies was one strategy
employers adopted to reduce recruitment costs, increase productivity, or lower compensation or other
costs.
Automotive Suppliers
During the latter half of the 1990s, automotive suppliers witnessed tremendous growth as car sales
reached new highs. In 1999, car and light truck sales climbed to 16.8 million units (an industry record)
after enjoying three “good” sales years of more than 15 million units. Industry economists stated that the
1
The Balanced Budget Act of 1997 created various Diagnostic Related Groups, which specified maximum
reimbursement rates for medical procedures; these reimbursements were lower than those previously in effect.
3
industry had established a new sustainable annual level of output of 16 million units, which was an increase
from their earlier estimate of 15.5 million units. In 2000, car and light truck sales soared even higher,
reaching 17.3 million units. Finally, in 2001, despite recessionary conditions in the nation’s business
investment sector, car and light truck sales stayed at 17 million units.
These high sales volumes came at the expense of prices; prices on cars and light trucks had been
stable or declining during the preceding five years. Facing lower prices, automakers turned to cost
reduction strategies to increase profits. Automakers pressured their suppliers to take on more research
and development responsibilities and risks, to respond quickly to new orders and design changes, to deliver
their product in a just-in-time environment, and to do all of this for less money. In this environment, use
of temporary help agencies became an important staffing strategy in four of the five auto supply plants we
studied.
Hospitals
Cost pressures on hospitals began to grow in the late 1960s, when there was a decline in
reimbursement due to changes in programs such as Medicare. These cost pressures increased during the
1990s, primarily because of declining reimbursement from the federal government and insurance
companies.
1
Adding to cost pressures on hospitals was the growth of managed care organizations, which
increased competition within the market for health care, and often forced hospitals to offload non-profitable
businesses (such as nursing homes or dialysis services) and focus more on their core competencies. These
cost pressures led hospitals to undergo consolidation and other forms of reorganization.
2
U.S. Bureau of Labor Statistics, Employment and Earnings, 1998.
3
See, for example, the results of the tenth annual “Contract Management Survey” (2000) reported in
Burmahl (2001).
4
A substantial fraction of a hospital’s expenses are in the form of labor costs. An estimated 60
percent of a hospital’s expenses are typically labor costs, of which 40 percent are accounted for by nurses.
For these reasons, the nursing pool typically has been a target for cuts by hospitals seeking to contain costs,
and nurses have often been replaced with lower-skilled, low-wage workers (Egger 1999). “Nursing aid,”
whose wage is just 60 percent of the median for all workers,
2
is one of the fastest-growing occupational
groups in the United States. In addition, many hospitals have outsourced support functions, such as food
and environmental services, in an effort to increase efficiency and reduce costs.
3
Public Schools
The public schools in our study are located in Michigan. Two measures enacted in the mid-1990s
greatly altered the competitive and financial environment for public schools in this state. The first was the
licensing and funding of charter schools, which were intended to increase school choice for parents and
students, primarily in poor urban districts, and to stimulate innovative educational practices for at-risk
children. The second was the passage of Proposal A, which shifted public school funding away from local
property taxes and onto the state sales tax. The new tax base was intended to equalize funding of public
schools across wealthy and poor districts, and stabilize funding for poor districts that had suffered declining
property values. The vast majority of a public school’s funding now comes from a per-pupil allowance
from the state, set at $6,971 in 2001–2002. The per-pupil allowance represents an average cost of
educating a child, and the marginal cost is lower. An unintended consequence of the new funding
5
mechanism is to reward expanding school districts and penalize districts with declining enrollments.
Declining enrollments, in large part due to competition from charter schools, placed many school districts,
especially urban districts, in severe financial straits.
One response by school administrators in financially troubled districts has been to cut costs in
noninstructional services—such as bus, custodial, and food services—in order to maintain the quality of
instructional services. Workers in noninstructional services are commonly unionized in Michigan,
particularly in larger urban and suburban districts. A 1994 Amendment to the State of Michigan’s Public
Employment Relations Act, which precluded collective bargaining over the outsourcing of noninstructional
services, paved the way to outsourcing of previously unionized, noninstructional services.
CASE STUDY METHODOLOGY
We studied the use of temporary services and contracting out in connection with the low-skilled
workforce in five automotive supply establishments, six hospitals, and seven public school districts. The
five automotive suppliers were located in the Midwest; three of the hospitals were located in Michigan and
three in North Carolina; and all seven public school districts were located in Michigan. In selecting
automotive suppliers and hospitals for case study, we had no prior knowledge of the organizations’ use of
flexible staffing arrangements. In public schools, however, we intentionally selected districts that contracted
out at least one non-instructional function.
Although the sample was restricted geographically, we endeavored to select a variety of
organizations within each industry. Among the automotive suppliers, the workforce ranged in size from 430
to 2,100 employees. Two of the plants were foreign-owned subsidiaries, one was a branch plant of a large
6
U.S. company, and two were locally-owned plants. Production workers in two of the five plants were
unionized. The hospitals varied in size from a small Michigan hospital with 450 employees to a large North
Carolina hospital with 6,000 employees. Half of the hospitals were public and half were private, not-for-
profit institutions. One hospital was unionized. In public schools, we interviewed a wide range of districts.
Three of the districts were urban, one was a large suburban district, and two were small rural districts. One
of the case studies was an intermediate school district organized at the county level, primarily to provide
services for children with disabilities. The bus drivers in all of the school districts were unionized—even
bus drivers who were employees of a contractor. Of the six districts with food service and custodial
employees, food service workers were unionized in three and custodial workers in four.
In all three industries, we conducted extensive interviews with the organizations’ managers,
temporary agency or contract company representatives, and workers in order to gain various perspectives
on why organizations used these staffing arrangements and their impacts on workers. We developed a set
of questionnaires for each type of person interviewed (manager, temporary help agency, contractor, worker
representative) within each industry. Therefore, a core set of questions was asked at each site within a
particular industry. However, we asked follow-up questions to allow interviewees to expand upon or
clarify certain points. All interviews were tape-recorded and subsequently transcribed.
In the auto supply industry, we interviewed the human resource director of each of the companies,
at least one of their first-line production supervisors, and the on-site temporary employment agency
representatives at the two production facilities with such representatives. At one company, we were able
to conduct two focus group sessions, one with permanent workers and another with temporary agency
workers.
4
We conducted focus groups with regular and temporary nurses at two hospitals. Use of temporary nurses
was the most important and controversial use of temporary services in hospitals. However, because nursing is a
high-skilled occupation, we do not cover it in this chapter.
7
Within hospitals, we interviewed the human resources director and managers in key functional
areas: nursing, clerical and administrative support, laboratory, food services and housekeeping, and clinical
specialties (e.g. radiology, physical therapy). We also interviewed selected temporary help agencies
providing clinical staff to hospitals, and the regional manager of a national contractor that supplies food and
cleaning services to hospitals.
4
In each school district, we interviewed the person or people in charge of non-instructional personnel
and contracting out. In one very small school district, this was the superintendent. In larger districts, these
duties were performed by a business manager and human resources manager. We also conducted
interviews with food service and cleaning contractors that were used by these public school districts and
with union representatives in occupations affected by contracting out.
In addition to collecting information through interviews, we collected basic data on employment,
wages, and benefits by occupation of regular full-time, part-time, and on-call employees at each
organization studied. Where possible, we also collected similar information for contract and temporary
agency employees assigned to the organization studied.
LOW-SKILLED JOBS IN AUTOMOTIVE SUPPLY, HOSPITALS, AND PUBLIC SCHOOLS
In this study, we focus on workers in “low-skilled” jobs. By “low-skilled,” we primarily mean
positions that do not require post-secondary education and whose tasks can be learned on the job in a
relatively short period of time. This would exclude, for example, skilled-trade positions. We use the term
8
“low-skilled” rather than “unskilled” to denote the fact that, although the training time required to learn these
jobs is relatively short, the jobs often require non-trivial skills. We further distinguish between “low-skilled”
and “low-paid” positions. While workers in low-skilled positions typically receive relatively low pay, this
is not always the case, particularly when workers are unionized. In the latter situation, employers may have
strong incentives to outsource these positions to temporary agencies or contract companies, as is discussed
further below.
Low-skilled jobs comprised a large share of employment in all three of the industries studied. In
auto supply, production positions accounted for 60 to 75 percent of employment in the case study plants.
The overwhelming majority of these production positions were low- or semi-skilled; entry-level positions
were typically learned on the job in two days or less.
Although registered nurses (RNs) form the single largest occupation in hospitals, hospitals have
shifted work away from high-skilled, high-paid RNs to low-skilled, low-paid nurse assistants. Moreover,
hospitals utilize a large number of workers in clerical, food service, and housekeeping functions. The vast
majority of workers in these support functions are low-skilled. Among the hospitals we studied, between
40 and 50 percent of the staff were in low-skilled positions. Between 13 and 20 percent of staff were in
low-skilled clinical positions, 15 to 21 percent were in low-skilled clerical positions, and another 7 to 10
percent were in housekeeping positions.
Like RNs in hospitals, teachers represent the core occupations in public schools. Like hospitals,
public schools utilize many low-skilled workers in such areas such as food service, cleaning, and
transportation. Among the schools in our study, 15 to 20 percent of a district’s employment was typically
in these three non-instruction support occupations.
5
For auto supply, we report data for entry-level production workers. For hospitals and public schools, we
report data for workers in the most prevalent occupational title within a particular functional area, which coincides
with the lowest-paying occupation. The precise occupational titles and the number of occupational titles in a
particular functional area varied across organizations within an industry.
9
Most of the entry-level positions offer some possibility for career advancement. For instance,
production workers may learn additional skills or stations, for which they would receive higher pay.
Eventually, the best workers may be promoted to team leaders or line supervisors. The typical entry-level
position in food service entails assistance with simple food preparation and serving. A worker may advance
to cook or cashier, and eventually to a supervisory position. Custodians in schools may advance to be the
head custodian of a school. Some occupations, such as bus drivers and nurse’s aides, have no job ladder.
The low-skilled jobs in our sample are not always associated with low compensation. This point
is pertinent to our study because the incentive organizations have to outsource jobs to a temporary help
agency or contract company and the effects on workers of such outsourcing are likely to depend on
workers’ compensation. Organizations would be expected to have a greater incentive to outsource jobs
the higher the wages of their low-skilled workforce relative to low-skilled workers in other competing
organizations. In turn, relatively high-paid, low-skilled workers potentially have more to lose from
outsourcing.
Information on starting wages, availability of health insurance benefits, and union status of workers
in low-skilled occupations in automotive supply, hospitals, and public schools is reported in appendix table
A-1.
5
The wages and benefits for entry-level production workers in automotive supply tended to be better
than those for low-skilled occupations in the other two industries. Production workers always worked full-
time and received basic health insurance and retirement benefits. Perhaps most notable was the large
10
variation in starting wage across auto suppliers. Two of the plants in our sample were unionized and these
workers received relatively high wages, however one of the non-union plants paid the second-highest
wages. Historically, this employer intentionally pursued a “high-road” strategy, paying above-market
compensation to its workers. It believed that by so doing it would attract the best workers and thereby
reduce quality control and turnover problems. This employer had a reputation of being one of the best in
the area and had little trouble attracting workers, even during tight labor markets. A high compensation
strategy was also motivated by a desire to remain nonunion.
Wages for entry-level positions in food service, housekeeping, nurse assistant, and clerical positions
in hospitals tended to be quite low. Hospitals offered health insurance and retirement benefits to all full-time
workers. However, many low-skilled hospital workers worked part-time and received prorated benefits
with high copayments on health insurance, or no benefits. Interestingly, the one hospital in our study with
unions representing low-skilled workers had a high fraction of part-time workers in its low-skilled positions.
The wages for low-skilled workers in public schools varied considerably. Entry-level food service
positions in public schools were almost always part-time and the wage levels were quite low, even in cases
where these workers were unionized. None of the schools in our sample offered health insurance to their
part-time food service employees. Under state law, all public school employees, including food service
employees, are part of a state retirement plan. Payment into the retirement plan is made solely by the
school and amounts to between 12 and 13 percent of the workers’ earnings. Entry-level custodians, who
were often unionized, typically worked full-time and earned substantially higher wages and benefits than
did food service workers. Bus drivers bear considerable responsibility and require special training,
particularly in the area of safety. All of the bus drivers in our sample were unionized, even when they were
6
Temporary help agencies were only used occasionally for clerical positions in public schools.
11
employees of a contract company. The hourly wage of bus drivers tended to be similar to that of
custodians, but many school systems experienced great difficulty recruiting bus drivers. One reason is that
most positions were part-time, with hours concentrated in the early morning and mid-to-late afternoon.
In an effort to recruit bus drivers, some school systems offered these individuals full-time employment with
full benefits by utilizing them in other jobs in the intervening hours.
TEMPORARY HELP USE
The use of temporary agency help was common in our auto supply and hospital case studies.
6
Usage was particularly high among some auto suppliers. Although no statistics exist on use of temporary
agency employment at a detailed industry level, our case study evidence of high use in auto supply is
consistent with findings that temporary help in manufacturing has grown rapidly since the 1980s (Segal
1996; Segal and Sullivan 1997).
Statistical studies have shown that, on average, agency temporaries earn lower wages than
comparable workers in regular positions (Segal and Sullivan 1997, 1998). They also tend to receive fewer
benefits (Houseman 1999; Kalleberg, Reskin, and Hudson 2000). Based on these facts, it is often
assumed that temporary jobs are “bad” for workers. Yet, as our case study evidence shows, the
implications for workers are often more complicated. First, differences in compensation between agency
temporaries and regular employees in statistical studies are only averages; sometimes agency temporaries
earn more than their counterparts in regular jobs and often their compensation is similar. Moreover, the
12
effects of temporary employment on workers depend on why organizations are utilizing temporaries and
what the alternatives available to the worker are. For instance, if the organization is using temporary staff
to screen workers for permanent positions, temporary employment may have no long-term adverse effects
on these workers and may allow low-skilled workers to audition for jobs for which they would otherwise
not be considered. Finally, an organization’s utilization of temporary staff may impact not only the
temporary workers, but also the regular staff, and the effects on these two groups may be quite different.
Use of Temporary Help among Auto Suppliers
Four out of the five auto supply plants utilized temporary agency help during the period of our
interviews (Table 1). In two of the plants, the utilization of temporary employment was very high,
accounting for 20 percent or more of production employment. In one unionized plant, the use of temporary
agency workers was strictly prohibited by union contract.
As noted above, all of the plants were operating at or near full capacity and employers faced tight
labor markets during the period of study. Human resources directors reported that, while they received
many applications, the average quality of the candidates was low. The fact that auto suppliers continued
to have a large number of job applicants, albeit of low quality, despite the tight labor markets likely reflected
the fact that the entry-level jobs offered relatively high pay and good benefits for low-skilled workers. For
the most part, the auto suppliers in our study used temporary help agencies to staff increases in production
and to screen workers for permanent positions. Across the cases in our study, however, there were subtle
but important differences in employers’ motivations for using temporary staff, the terms by which temporary
staff were hired, and the implications for temporary and regular staff. We distinguish three cases.
13
Temporary Help in Unionized Plants
Auto suppliers are often faced with sudden and temporary increases in production levels and thus
have an inherent, periodic need for temporary workers. Auto Supplier D, the unionized plant prohibited
by contract from using agency temporaries, instead hired temporary workers directly when production
volumes increased significantly. The short-term workers received entry-level wages and could stay no
longer than 120 days. They gained no seniority for their work effort, nor did they gain any advantage from
working at the plant if they later sought a permanent position.
Auto Supplier E, the other union plant in our sample, was strictly limited in its use of agency
temporaries. The labor contract stipulated that individual temporary agency workers could be used for no
more than 30 days, and then only after the unionized workers turned down the opportunity of greater
overtime hours. At the end of the 30 days, the employer could either hire the temporary worker on its
payroll or terminate the temporary’s contract. In addition, by collective agreement, the agency temporaries
earned the same wage as the employer offered new production workers. Thus, in practice, the treatment
of direct-hire and agency temporary staff differed little between the two union plants. For the companies,
the one using the agency temporaries paid overhead to the agency, while the other incurred higher internal
costs in recruiting and screening its own temporary workers. Although temporary help agencies likely enjoy
economies of scale in recruitment and screening, the union company using direct-hire temporaries paid
higher wages than the union company using agency temporaries, and thus it is likely that it was easier for
the former to recruit temporary workers on its own.
It is important to note that the human resources directors at both plants expressed a desire to use
or to increase the use of agency temporaries. Even paying similar wages and agency overhead, human
14
resources directors felt that they could save money by using agency temporaries because they incurred no
benefit costs on the temporaries and because they reduced overtime paid to regular staff. Thus, the
provisions in the union contracts constrained management staffing decisions.
Use of Agency Temporaries by Non-Union “High-Road” Auto Supplier
The nonunion Auto Supplier C offered higher wages than one of the union plants in the sample and
its wages were substantially higher than those at the other nonunion auto suppliers. The company was
regarded as one of the best area employers for low-skilled workers seeking manufacturing employment.
In addition to good wages, the company had a no-layoff policy for its regular workers.
An integral complement to the company’s high-wage, strong job security policies for regular
workers, however, was the extensive use of agency temporaries who earned substantially lower wages than
entry-level regular workers, and had no benefits or job security. The company’s rationale for using agency
temporaries was two-fold. First, the company carefully screened all new hires through a temporary help
agency, which had an on-site representative. The company wanted to ensure that workers were high
performers before offering them job security at relatively high compensation. The company also wanted
to remain nonunion. According to the on-site temporary agency representative, the agency used this
probationary period to screen out individuals it deemed likely to sympathize with union causes. Temporary
agency workers had to complete a minimum of 6 months of work before becoming eligible to apply for a
regular position. Sometimes workers remained as temporaries for a year or more before being offered a
regular position or leaving. This probationary period as a temporary worker was considerably longer than
that at any other auto supplier we studied.
15
A second reason the company used the agency temporaries was to buffer regular workers in the
event of a downturn in production. According to the human resources director, the company believed it
had to maintain about 10 percent of its workforce as temporary to provide an adequate buffer for regular
workers in the event of a downturn.
At the time of our interviews, however, the fraction of production workers who were agency
temporaries had climbed to over 20 percent. The high fraction of temporaries resulted from the fact that
production had been expanding, all new hiring was done through the temporary agency, and temporaries
were converted to permanent hires at a slow rate.
Whether the fraction in temporary employment should be substantially lowered by offering more
temporaries permanent positions at a faster pace was the subject of intense debate at the time of our
interviews. On the one hand, the accounting department had calculated that a five percentage point drop
in temporary employment cost the company one million dollars per year due to higher wages and benefits
earned by permanent workers. On the other hand, human resources expressed concern that the
temporaries were not as committed to the organization as regular staff. Moreover, reportedly many good
temporaries tired of the long wait for permanent employment and quit, resulting in increased training costs
and quality problems associated with high turnover among temporaries—factors not taken into
consideration in the accounting department’s cost-savings calculations. Human resources had feedback
from other companies reporting that they hired many well-trained workers who had worked as temporaries
at Auto Supplier C, but who had quit or been passed over for promotion. The plant’s human resources
director summarized the internal debate as follows:
16
…as we were growing the business and … we’re trying to get product out the door, it’s
like ‘get some more temporaries,’ ‘get some more temporaries,’ and one morning we
woke up and we were at like 25 percent. And … quality is starting to have problems …
and now it’s like ‘we’ve got to get this temporary ratio back down.’
…we’ll start edging back down to 20 and … then the goal becomes 15 percent. We
haven’t hit it … and now, there’s always this discussion, ‘Well, it’s more cost effective to
have the temporaries.’ So it doesn’t seem to be an initiative with the executives to get that
ratio down. So even though they talk about it, we are never going to get this high rate
down, we run at around 20 percent. So, what I’m trying to say is ‘Is there a cost
advantage?’ If there is … let’s decide this and we’re going to operate within 20 to 25
percent.… But … we are in this constant state of denial, yet that number still stays up there
and yet the Vice President of Human Resources is like [saying] ‘we’ve got to get it down.’
We conducted two focus groups with production workers at this plant—one with regular
employees and the other with agency temporaries—to better understand the workers’ perceptions of the
costs and benefits of the company’s use of temporaries. Interestingly, the regular workers, all of whom had
been hired through the temporary agency, saw no problems with the system and perceived it as fair. In
marked contrast, the temporaries viewed the system as unfair. They understood that the company would
want to screen candidates through the temporary agency, but believed that a two or three month time
period—as was typical at the other nonunion plants we studied—was more than adequate to assess a job
candidate. They believed that temporaries were not treated as well by supervisors and deeply resented
the low wages they earned as temporaries. However, this resentment did not affect their productivity, for
they knew that only the best temporaries would be offered permanent positions.
In sum, at the time of our interview, company executives de facto had chosen to keep high levels
of agency temporaries. The recent dramatic growth in temporaries at this company could be viewed as
a backing off, or at least a qualification, of their human resources commitment to high wages and job
security. Their original “high road” philosophy was premised on the belief that high wages and strong
17
commitments to job security would be paid back through high quality and low turnover. However, by
having a very long probationary period at much lower wages, the company, in essence, was recapturing
some of the economic rents workers earned subsequently. At least the prevailing group of executives had
come to believe that wage and benefits savings outweighed other costs associated with this strategy.
Temporary Help Use among Other Nonunion Auto Suppliers
Auto Suppliers A and B were nonunion plants whose compensation levels were considerably lower
than those at the other three facilities studied. They each made extensive use of agency temporaries. At
the time of our interview, agency temporaries accounted for over 20 percent of production employment
at Auto Supplier B. This high use was because of a temporary increase in production. The human
resources director at this relatively small plant felt it would have been extremely difficult to recruit and
screen large numbers of workers quickly for a large increase in production, especially in the tight labor
market prevailing at the time of the interviews, without the assistance of a temporary agency.
Both plants also used temporary help agencies to screen workers for permanent positions.
Temporary agency workers were typically screened over a two to three month period. If the company
chose to offer them a position at the end of their temporary contract, they would then be placed on a
probationary status as a direct hire of the company for another 90 days.
As was the case in the other auto supply plants, human resources directors complained about the
low quality of the applicants in the tight labor markets prevailing at the time. Many applicants had little or
no experience in a manufacturing setting. Moreover, some applicants had criminal records and many had
spotty work histories. Managers emphasized that, in these tight labor markets, they were hiring job
applicants that they would have never taken a second look at in the past.
7
According to data supplied by the companies, the hourly wage earned by agency temporaries was 75
percent, 83 percent, and 56 percent of that earned by new hires at Auto Suppliers A, B, and C, respectively. The
agency bill rate was 79 percent, 85 percent, and 60 percent of the total compensation of a new hire at Auto Suppliers
A, B, and C, respectively.
18
Although the entry-level tasks that workers were expected to perform required little or no previous
skills, the jobs required “soft” skills, like being punctual for one’s shift and being able to get along with
coworkers. Production supervisors stressed the need to hire workers that “fit” into their production teams,
and the use of temporary workers provided the opportunity to see if the new workers would fit.
I have the opportunity to observe this person over a period of time. I don’t have just a
half-an-hour snap shot to go by as in an interview. I can get to know the person, get to
know their background a little bit, get to know what their behaviors are, what motivates
them, what possible problem areas I might have with the individual.
Historically, these auto suppliers would have hired new workers directly on probation. Temporary
help agencies offered a couple of advantages to companies: First, while the workers were with the
temporary agency, they were less costly to the company than if they were hired directly. Although the
wage differentials between temporaries and direct-hire production workers were not as great as at Auto
Supplier C, temporary workers assigned to Auto Suppliers A and B earned lower wages than new hires
at these companies. Moreover, the rate at which the temporary agency billed the workers was lower than
the total compensation costs of new hires.
7
Second, it was easier for a manager to terminate and replace
a worker hired through a temporary agency. The manager could simply inform the temporary agency that
the assigned worker was not satisfactory and request a replacement.
By lowering the compensation and firing costs, temporary help agencies made it more attractive
for companies to try out workers with criminal records, poor work histories, or otherwise “risky”
characteristics. In the absence of temporary help agencies, companies may have hired these risky workers
19
anyway. However, they may have chosen to offer higher wages and attract more qualified candidates
instead, bidding them away from other companies. The human resources director at Auto Supplier B
specifically discussed this alternative strategy, saying that his company had tried it, but opted instead to use
agency help.
The use of temporary help agencies as a mechanism for screening risky workers was even more
transparent at Auto Supplier A. At that company, job applicants were channeled into one of two tracks.
Applicants with good qualifications were hired directly with a 90-day probation. Applicants deemed more
risky were referred to the temporary help agency, which, in turn, could place the candidate with the
company on a temp-to-perm contract. At the end of the contract, successful candidates would be hired
directly and begin the company’s 90-day probationary period. According to the human resources director,
treatment of its agency temporaries was fair:
If you are out on the line, you would not be able to distinguish a contract employee from
one of our own full-time employees. There is no differentiation on the training perspective,
on the assignments they are given … Our intention in bringing [temporary employment
workers] in is to bring them on full time. It’s just that their backgrounds [make it uncertain
if they will be able] to meet our pretty rigorous hiring criteria. Yet, if we sense that there
is an opportunity for a good match, we will bring them in through contract. So, there is
never an intent to just bring them in for a short-term blip in production and then let them
go.
Approximately 70 percent of the company’s new hires came through the temporary help agency.
A key element allowing these auto suppliers to tryout more “risky” workers through temporary
employment agencies was the use of small production teams. Instead of the assembly line approach to
production, work at these plants was organized in small production areas, where the more experienced
workers were able to perform many, if not all, of the tasks required. This setting allowed the more
8
We develop the argument that temporary help agencies may reduce pressure on companies to increase
wages in Houseman, Kalleberg, and Erickcek 2001.
20
experienced workers to monitor the performance of the temporary worker, catch errors before the product
left the assembly area, and adjust quickly to possible bottlenecks that arose as the new workers learned
their tasks.
Although agency temporaries earned lower wages than those hired directly, many of these workers
potentially benefitted from the exposure to jobs that temporary help agencies offer. By lowering the cost
to companies of hiring them, temporary help agencies gave workers with risky backgrounds opportunities
to try out for higher-paid, full-time manufacturing jobs with good benefits, opportunities that they otherwise
might not have had. Another advantage to these workers is that if they did not succeed in the position or
did not like the job, the assignment simply ended; it was not recorded as a dismissal or a quit, which would
tarnish their employment records.
For regular workers, any costs of using agency temporaries are indirect. By lowering the cost to
companies of trying out riskier workers, temporary help agencies effectively expand the supply of potential
labor to a company. In this way, companies may avoid or minimize the need to raise wages in order to
attract more qualified candidates; because increases in wages for new workers are almost always
accompanied by increase in wages for existing workers, existing regular workers would benefit from higher
wage levels.
8
Use of Temporary Help among Hospitals
The hospitals in our study made use of agency temporaries in a number of low-skilled occupations.
In clinical areas, the use of agency temporaries in such occupations as nurse assistants and patient sitters
9
Because the agency temporaries are somewhat more likely than other workers to work part-time hours, the
fraction of hours worked by agency temporaries is probably lower than the fraction of employment accounted for by
agency temporaries.
21
was often cited in interviews. Patient sitters watch extremely ill or confused patients and call for help when
needed. Because the need for patient sitters is so variable, hospitals tend to staff this position exclusively
with agency temporaries. In non-clinical areas, all hospitals cited use of agency temporaries for clerical
functions. Also, three hospitals indicated that they used agency temporaries in food and environmental
services, although only one maintained records of this use.
Table 1 reports the fraction of total hours worked in a particular occupational area that were
accounted for by agency staff in hospitals with data that permitted such a calculation. Economy-wide,
temporary agency workers represent between 2 and 3 percent of paid employment.
9
Thus, the figures
suggest that agency use in low-skilled hospital functions is often moderately high.
One reason hospital managers commonly cited for using agency temporaries was to circumvent the
human resources department. Managers complained that human resources took too long to process new
hires, in part because new hires needed to be carefully screened to work in a hospital environment.
Managers felt that they could not perform the work with existing staff if there were several vacancies, and
so would often bring workers in as agency temporaries while they waited to have their paperwork
processed by human resources. Managers cited the ability to quickly dismiss a worker as another
advantage of using temporary agencies for new hires. Human resources departments in
hospitals—particularly the larger hospitals—tended to have elaborate and lengthy procedures that
managers had to follow in order to dismiss a worker. Therefore, some managers preferred to screen new
workers through temporary help agencies. In these situations, workers were simply being screened for a
22
few weeks with a temporary agency, and such use of temporary agencies would appear to have little effect
on these workers.
The second major reason managers cited for using temporary help agencies was difficulty in
recruiting new workers. Hospitals had come under intense pressure to reduce costs. At the same time,
given tight labor markets, their wage levels for low-skilled clerical, food, housekeeping, and clinical
positions were becoming uncompetitive with wages for low-skilled workers in other sectors. In the words
of one manager, “We had a problem attracting recruits. Frankly, McDonalds and Burger King were paying
what we were and [workers there] don’t have to go through the stress or the hassle or the hours or the
customer service that they are required to perform here.” In contrast to the situation among auto suppliers,
where managers complained that the quality of the applicants had declined, some hospital managers
complained that they had almost no applicants, good or bad, for low-skilled positions. When hospital
managers had difficulty recruiting and retaining staff in low-skilled positions, they often turned to temporary
help agencies, on the grounds that the agencies were better at recruiting workers.
However, temporary agencies could not always recruit workers at these low wages either. For
instance, one hospital manager reported: “We wanted a housekeeper … and only wanted to pay the person
$7 an hour … you can’t get a housekeeper for $7 an hour and so we’ve made calls to four agencies and
they only laugh.” Similarly, one hospital reported difficulty retaining workers in patient billing, who received
low wages but nevertheless required extensive training. When the hospital was unable to staff the position
on its own, the hospital brought in a temporary agency to staff it. Yet, the agency could not retain workers
in the position either. A more common complaint was that while temporary agencies might be able to
staff positions at low wage levels, the average quality of the workers in these positions was very low.
10
We discuss the theoretical framework of such wage discrimination and cite numerous examples in which
temporary help workers earn more than regular staff in high-skilled clinical positions in Houseman, Kalleberg, and
Erickcek (2001).
23
In several cases, hospitals reported that workers hired through the temporary help agency earned
more than regular workers on staff. In these situations, it is not so much the agency temporaries who are
potentially harmed by the arrangement, but the regular workers. By hiring through a third party like a
temporary agency, employers may effectively raise the wages for new hires without raising wages for
existing staff.
10
Such wage differentials can cause deep resentment and morale problems among regular staff,
however, if they become aware of them. One temporary agency placing nursing assistants reported poor
treatment of their assignees because regular staff resented the high pay they received. Perhaps because
of such morale issues, managers more typically reported that agency temporaries received about the same
hourly rate as regular entry-level staff, though in many instances temporaries made less in total
compensation, taking into account differences in benefits.
In general, hospitals had less financial incentive to use agency temporaries than auto suppliers. In
auto supply, temporary workers earned less than entry-level regular workers, and the agency’s bill rate to
the company was lower than the total compensation costs of hiring a new regular worker or paying
overtime to an existing worker. Thus, even though managers in auto supply plants complained about the
quality of temporary workers, most felt it was a cost-effective way of screening job candidates and staffing
for temporary increases in production. In hospitals, however, temporary workers typically earned about
the same wage as regular workers. Managers reported paying a 40 to 50 percent overhead rate to the
temporary agency. In the absence of agency temporaries, hospital managers would typically pay regular
24
workers overtime to cover understaffed positions. Therefore, with a 50 percent overtime premium, many
managers reported that the marginal hourly cost of paying overtime was about the same as hiring a
temporary.
Given that temporary workers were regarded as less reliable than their own staff, a couple of
managers in the food and environmental services areas reported movement toward reducing the amount
of temporary help and offering high hours positions for regular staff. For instance, one hospital reported
recently advertising 60 hour per week positions. The rationale was that these low-wage workers had to
work two or three jobs to make ends meet. Managers saw the chance to consolidate these hours at the
hospital, effectively at a higher wage factoring in overtime, as benefitting both workers and the hospital.
Except to the extent that their use enabled hospitals to postpone wage increases for low-skilled
workers, temporary help agencies appeared to have little adverse impact on hospital employees, and
potentially benefitted agency workers. Although the hospitals generally paid low wages, full-time workers
did receive benefits, and thus, as in auto supply, very risky workers might obtain an opportunity that they
would not have otherwise received to try out for jobs with benefits through a temporary help agency.
However, because of the low wage levels in hospitals, temporary help was relatively expensive for hospitals
to use and most managers expressed a desire to reduce, not increase, their reliance on it.
One interesting fact to emerge from the comparison of temporary workers assigned to hospitals
and auto suppliers in the same labor markets was that there seemed to be a “going” wage for low-skilled
entry-level workers in temporary services. Unless specified in a union contract, temporary workers
assigned to entry-level production positions seemed to earn little, if any, more than workers assigned to
low-skilled positions in hospitals. Thus, with the exception of the unionized firm, the wage differentials
25
between temporary and regular workers were much greater in auto supply than in hospitals. Yet, provided
a worker could tolerate a manufacturing working environment, the assignment to auto supply offered better
future prospects in terms of training and wage advancement. This example underscores the problems of
trying to draw conclusions about the effects of temporary employment through simple wage comparisons
of workers in temporary and regular positions. The implications of temporary employment for workers
depends upon a company’s reason for using agency temporaries and the workers’ job alternatives.
CONTRACTING IN LOW-SKILLED OCCUPATIONS IN HOSPITALS AND PUBLIC
SCHOOLS
No systematic national data are collected by the government on the extent to which business and
government organizations contract out services. However, some data—including data from surveys
conducted by private associations—are available on outsourcing in selected industries. For hospitals, the
best data on national trends come from the annual Contract Management Surveys conducted by Hospital
and Health Networks, a publication of the American Hospital Association. In 1999, 32 percent of
hospitals reported that they outsourced their food services, with another 6 percent reporting that they
planned to do so in the next two years. In addition, 27 percent outsourced housekeeping, while 3 percent
reported that they planned to do so in the next two years. The two most commonly outsourced functions
were pest control and laundry services, with 86 percent and 62 percent of respondents indicating that they
outsourced these functions, respectively (Sunseri 1999).
Limited government statistics are available on contracting out of transportation and food
services in primary and secondary public schools. The Federal Highway Administration estimates that
11
U.S. Department of Transportation, Federal Highway Administration, Highway Statistics 2000,
http://www.fhwa.dot.gov/ohim/hs00/mv10.htm.
26
about 30 percent of school buses were privately or commercially owned in 2000.
11
In some instances,
these statistics reflect the outsourcing of transportation services to private companies. However, in
other instances, they capture the fact that some school districts lease their buses from private companies
but continue to manage and employ all workers in this area. Among school districts nationwide that
participated in federally-funded school lunch programs, about 8 percent used food service management
companies in 1994–1995, compared to only about 4 percent in 1987–1988 (U.S. General Accounting
Office, 1996).
Table 2 shows the patterns of contracting out low-skill occupations in the hospitals and public
schools in our sample. Interestingly, the nationwide surveys previously cited on contracting out in
hospitals and public schools do not distinguish whether an organization contracts out the entire
operation or just its management. As is evident from Table 2, it is quite common for hospitals and
public schools to contract out only management functions and to keep workers on their payroll.
Whereas all of the hospitals we studied outsourced the management of food services and/or
environmental services, none outsourced the entire operation. Among public schools, five of the six
districts with a food service operation contracted it out, but only two contracted out the entire function.
The one district in our sample that contracted out custodial services contracted out the entire function.
Two of the districts in our study contracted out the entire student transportation function.
In situations where only management was outsourced, managers, who were employees of the
contract company, directed the hospital’s or school’s employees. They also took on primary
27
responsibility for hiring and firing workers, under the guidelines and with the approval of the client’s
human resources department. However, pay and benefits were established by the hospital or public
school. As is discussed below, whether or not an organization outsources its management functions or
its entire operation potentially has important implications for workers.
Why Organizations Outsource Management Functions and its Implications for Workers
Hospital and education administrators typically view areas such as food service, cleaning, and
transportation as outside their areas of “core competency.” Choices to contract out are made with a view
toward improving services and reducing costs in these areas, while freeing administrators’ time to focus on
the business of health care or education. One common reason hospital and public school administrators
gave for outsourcing was difficulty recruiting good managers in non-core areas. Because contract
companies specialize in a particular service and have many clients, they can offer managers an internal
career ladder, and therefore can attract managers more easily than an individual hospital or school can.
In addition, contractors are often part of large national companies that offer well-tested systems
of quality- and cost-control and can provide their on-site managers with technical support. For instance,
in the area of food service, contract companies have systems to estimate food usage and reduce wastage.
They have dieticians at their corporate headquarters that can consult with hospitals and public schools on
specific issues. Because they purchase in bulk, they can often get better prices on food than hospitals or
school systems can operating on their own. In public schools, a complex set of regulations govern federal
government reimbursement of school lunches for children eligible for free and reduced-price lunches.
Contract companies offer expertise in packaging and marketing meals to comply with government
28
regulations, thereby maximizing federal reimbursement. Transportation companies often offer more
systematic safety training for bus drivers than that provided by individual public school districts.
Thus far, the ways listed in which a contractor may improve quality and reduce costs would have
few consequences for workers. Indeed, in our case studies, the outsourcing of food or custodial service
management by itself often appeared to have little impact on the workers in these areas. When workers
are significantly affected by the contracting out of management, the issues generally revolve around
productivity, work loads, and staff reductions. With regard to these issues, the contractors with whom we
spoke also reported that when they took over the management of food or custodial services, these
operations frequently were overstaffed and the quality of the work was poor, requiring them to reduce staff
and discipline workers. For instance, the contract environmental services manager at Hospital C reported
greatly increasing productivity upon taking over the operation from the hospital and firing workers who had
previously been “goofing off.”
Efforts to increase productivity and reduce staff may cause deep resentment and backlash among
workers. For example, upon taking over the management of food services in Public School A, the
contractor completely reorganized the structure of jobs, increasing work loads and leading to many
complaints and quits among workers. At the time of our interview, over half of the approved slots for food
service workers were unfilled, leaving those who remained to work harder and work longer hours. These
food service workers, who were unionized, felt they were in an awkward position being school district
employees with a supervisor who was not. The union complained that, in this situation, contract company
management often ignored union contract provisions.
29
Contract managers emphasized that improvements in worker productivity do not always or even
usually lead to layoffs, however. When staff reductions were desirable, contract managers typically worked
with the human resources department at the hospital or school to reduce employment through attrition.
Additionally, in food service in public schools, managers usually tried to improve quality, and hence sales,
which could lead to a net increase in jobs in spite of any productivity improvements.
Contracting Out Entire Functions and the Implications for Workers
The contracting out of an entire function potentially has larger impacts on affected workers. In our
case studies, apart from any impacts on work structure and work loads, low-skilled workers in functions
that had been entirely contracted out received lower compensation, most importantly in the form of lower
benefits, and sometimes they lost union status.
Under the Employee Retirement Income and Security Act (ERISA) and IRS tax laws, it is difficult
for an employer to discriminate in its offerings of retirement and health insurance benefits among its full-time
employees. Therefore, organizations with high-skilled and low-skilled workers will tend to offer the same
benefits packages to workers across skill levels, and low-skilled workers benefit from generous benefits
packages offered at organizations with high-skilled workers. However, contractors with predominantly
low-skilled employees typically offer less generous benefits. Hospital E, which outsourced all of its security
function, mentioned benefits savings as a major motivation for contracting out this function. Within public
schools, food service workers primarily work part-time and receive no health insurance benefits. Under
Michigan law, however, all school employees are part of a retirement system financed by a 12 to 13
percent levy paid by the school district on these workers’ wages. Although wage and benefit cost savings
30
did not appear to be a primary factor in Public School B’s and D’s decisions to outsource their entire food
service operations, administrators acknowledged these savings.
Public Schools C and F recently outsourced their entire transportation functions. Although the
workers remained unionized in both cases, the bus drivers received lower benefits with the contract
company, primarily because of the loss of public pension. One district went to great lengths to protect bus
drivers during the transition. It hired bus drivers near retirement in other district jobs to protect their public
pension, and guaranteed jobs and wages (but not benefits) with the contract company for other drivers.
Nevertheless, the contract company was allowed to hire new workers at lower wages, and, over time,
workers in general would likely be worse off. According to the results of a national survey published in
School Bus Fleet, the average wage for bus drivers hired by contractors was $10.76 in 2000, compared
to an average wage of $12.23 among bus drivers who were public school employees.
Public School G was the clearest case of a school motivated to cut wage and benefit costs by
contracting out. This urban district was losing students largely as a result of competition with charter
schools and, under Michigan’s new school financing laws, suffered large revenue losses. Its non-
instructional support staff was all unionized, and with legal changes prohibiting bargaining over contracting
out of noninstructional services, this district immediately moved to contract out grounds services.
According to school administrators, these employees were receiving wages well above private sector
levels. Decisions to outsource custodial services were made by principals, and about half of the district’s
schools opted to contract out this function. These principals realized cost savings because the contractor,
who was not unionized, paid lower wages and benefits.
31
According to union representatives and school administrators, another key factor in principals’
decisions to contract out was quality. The quality of the cleaning service was low at many buildings and
administrators either failed to discipline poor performers or had difficulty disciplining them, in part because
of union grievance procedures.
Although union employees still worked in half of the district’s buildings, the internal competition with
the contractor greatly reduced the union’s bargaining power. The union did not sacrifice benefits, but it
made large wage concessions on higher paid custodians. Perhaps most interesting, the union dramatically
altered its position on the disciplining of workers. The union believed that its members could only retain
schools and receive higher compensation than the non-union contractor employees if its productivity was
greater than that in the contract company. Therefore, union representatives met with school officials to
encourage them to discipline or dismiss poor performers.
Why the Outsourcing of Low-Skilled Functions is Limited
Although we came across examples in our case studies of hospitals and schools contracting out
entire services, with adverse consequences for workers’ wages and benefits, contracting out is more
typically limited to management. There are several reasons for this. First, schools and hospitals are
reluctant to lose control of the quality of their staff. Both serve vulnerable populations—children and very
ill patients—and as such are particularly sensitive to the types of individuals coming in contact with them.
Several school officials mentioned the direct contact custodians, food service workers, and bus drivers have
with children as a reason for not contracting out a particular service, or for only contracting out
12
This sensitivity to contact with children was also a reason offered by one public school official for why
contracting out was more prevalent in other government sectors than in public schools.
13
Institutional arguments (e.g., Dobbin et al. 1988) hold that exposure to the public sphere places
organizations under greater pressure to conform to evolving norms about legitimate employment practices.
32
management.
12
One contractor cited the extensive contact food service workers and housekeepers have
with patients as a reason why so few hospitals fully contract out these functions.
In addition, the hospitals and schools in our study were all either public or private, nonprofit
organizations, which historically have adopted a protective attitude toward employees.
13
Public school
administrators expressed a particular reluctance to outsource jobs in order to save money at the expense
of employees. Moreover, some administrators who chose not to outsource jobs expressed a fear of
backlash from the community were they to do so. Those who chose to outsource jobs described these
decisions as among the most controversial they had made.
The main potential benefit an organization would derive from contracting out the entire function,
rather than simply contracting out the management of the function, is savings on wages and benefits. Yet,
with a few notable exceptions, the wages of low-skilled workers in hospitals and public schools in our case
studies were relatively low. Although low-skilled workers typically earned more as employees of hospitals
and public schools than they did as employees of contractors—primarily because of better benefits—the
cost savings an organization could realize from contracting out its low-skilled workforce was often
perceived as limited. Instead, the organizations in our study often felt the primary benefit of contracting out
was better management, which would realize cost savings through increased productivity and reduced
expenditures on non-labor inputs. In these instances, the impact on affected workers is ambiguous.
33
Workers may face increased workloads or job reductions, but they may also reap benefits from more
knowledgeable management.
Although the contracting out of entire functions in hospitals and public schools is currently limited,
it is likely to expand if these organizations continue to face financial pressures to cut costs. As they realize
cost savings from outsourcing management functions, organizations may seek further savings in the form
of wage and benefit reductions by contracting out low-skilled staff. In support of this notion, a
representative from a major company supplying contract food and cleaning services to hospitals and
schools noted that his company had initially focused on the provision of management services, but in recent
years had begun to offer “full service accounts”—which involves putting all workers in the function on its
payroll—on a region-by-region basis.
Finally, it is important to note that growth in competition from contractors affects wage and benefits
even at hospitals and schools that do not contract out low-skilled staff. As was illustrated in the case of
Public School G, which utilized both in-house union and contract custodial staff, the option of switching to
a contractor greatly circumscribes workers’ bargaining power.
CONCLUSION
Evidence from our case studies suggests that the effects of temporary services and contracting out
on low-skilled workers vary considerably. In some situations, the use of temporary agencies or contract
companies appears to have clear adverse effects on workers’ wages, benefits, or other employment
conditions. In other situations, any effects on workers appear minimal or even beneficial.
14
The possibility that lower compensation may reduce unemployment is an important caveat to this
conclusion.
34
Adverse effects are clearest when agency temporaries or contract company workers are substituted
for regular employees on a long-term basis and receive lower compensation than they would as regular
employees.
14
This situation arises when 1) low-skilled workers in an organization receive relatively high
compensation, and 2) employers are not blocked from substituting these employees with agency
temporaries or contract workers by employment and labor law or by workers and their unions.
With respect to the first factor, we observed several situations in our case studies where low-skilled
workers were earning compensation well above that prevailing for workers in their positions. In two of the
auto supply companies, this higher compensation was associated with unionization, and in a third, at least
in part, to the company’s perceived threat of unionization. In Michigan public schools, unions were able
to raise the wages of workers in some non-instructional services. Moreover, state law mandated that all
public school employees be covered by an expensive pension plan. We also observed differences in the
benefits received by low-skilled employees in hospitals and auto supply, on the one hand, and their
counterparts in temporary help agencies and contract companies, on the other. Federal regulations
governing private sector benefits are designed to ensure that these tax-free or tax-deferred forms of in-kind
compensation do not primarily benefit highly compensated employees; consequently, companies generally
offer the same benefits to low-wage and high-wage employees. A result of this regulation is that low-wage
workers in organizations with high-wage workers tend to receive more generous benefits than do low-wage
workers in companies, such as many temporary help agencies or contract companies, that predominantly
15
See Houseman 2001a for a discussion of benefits regulations.
35
employ low-wage workers. Thus, unions or federal and state benefit laws may raise compensation levels
of certain workers above prevailing levels, giving employers an incentive to outsource these jobs.
With respect to the second factor, unions in the auto sector were able to block the use of low-wage
temporaries through collective bargaining. In contrast, the high-wage, non-union auto supplier substantially
lowered labor costs by utilizing agency temporaries for extended time periods. In Michigan public schools,
a change to collective bargaining law greatly weakened union power and opened the way to outsourcing
of non-instructional services to lower-paid contract company workers.
These examples from our case studies underscore the important interaction between unions and
employment and labor laws in enabling low-skilled labor to raise their compensation and to maintain these
higher compensation levels by preventing outsourcing. Evolving labor and employment law, similarly, will
impact both employers’ incentives to use temporary agencies and contract labor, and the implications of
these employment forms for workers. For instance, evolving labor law governing whether and how
temporary help workers are covered by collective agreements at a client company, and the rules governing
conditions under which employees of a temporary agency may form a union, potentially will have important
implications for the future impacts of temporary help employment in low-skilled labor markets. Benefit
regulations, as noted above, provide an incentive to companies to outsource jobs, particularly low-skilled
ones. Although Congress passed a law in 1982 requiring that a client provide benefits to agency
temporaries assigned to it on a long-term basis and the IRS has cracked down on the misclassification of
employees as independent contractors,
15
some have called for further action to curb the use of alternative
36
forms of employment for the purposes of circumventing benefits and other employment regulations (e.g.,
Commission on the Future of Worker-Management Relations 1996).
Many, if not most, of the instances of outsourcing in our case studies, however, did not entail the
long-term substitution of low-paid agency temporaries or contract company workers for employees, and,
we argue, the effects on workers were largely benign or beneficial in these situations. Most of the
temporary agency help we found in our case studies was, indeed, short term. Organizations typically used
agency help to cover for employee absences, staff for temporary increases in workloads, or screen workers
for permanent positions. What is perhaps most striking about our case study evidence on the contracting
out of low-skilled workers is the fact that it did not happen very much. More often than not, when hospitals
and public schools contracted out food services and when hospitals contracted out housekeeping services,
they only contracted out the management function. The employees in these occupations earned relatively
low wages, and the labor cost savings that could be realized from outsourcing the entire function—including
savings on benefits costs—were small relative to other cost savings these organizations felt they could
achieve from outsourcing management. In the few instances where low-skilled employees’ wages had
become much higher than those at a contractor, hospitals and public schools outsourced the entire function
when they came under financial pressure to cut costs.
There is an important caveat to the conclusion that because much contracting out in low-skilled
areas only involves management, the impacts on workers are relatively minor. Even if the practice of
contracting out low-skilled work is limited, the very existence of such contractors may have a powerful
effect on wages and benefits in low-skilled labor markets. As was illustrated by the example of unionized
custodians in one of our public school case studies, it will be difficult for workers and their unions to
37
significantly increase wages and benefits of low-skilled workers if organizations have the option of
outsourcing this work to low-cost contractors.
Finally, workers may even benefit from a temporary agency or contract arrangement. Some
workers, of course, prefer temporary positions. Workers may benefit from superior management skills
provided by a contractor. Moreover, organizations often screen workers with poor work histories or
otherwise risky characteristics through temporary help agencies. Particularly with auto suppliers, where
the cost of hiring through a temporary agency was substantially lower than hiring a worker directly, our case
study evidence suggests that, often, employers would have been unwilling to try out the more risky workers
in the absence of temporary agencies. Thus, temporary agencies potentially provide an important linkage
for these workers to full-time jobs with benefits.
38
REFERENCES
Abraham, Katharine G., and Susan K. Taylor. 1996. “Firms’ Use of Outside Contractors: Theory and
Evidence.” Journal of Labor Economics 14(3): 394–424.
Burmahl, Beth. 2001. “Making the Choice: The Pros and Cons of Outsourcing.” Health Facilities
Management 14(6):16–22.
Commission on the Future of Worker-Management Relations. 1996. Report and Recommendations of
the Commission on the Future of Worker-Management Relations. Issued January 9, 1995.
Report DLR, No. 6, Special Supplement. Washington, D.C.: Bureau of National Affairs, Inc.
Dobbin, Frank, Lauren Edelman, John W. Meyer, W. Richard Scott, and Ann Swidler. 1988. “The
Expansion of Due Process in Organizations.” In Institutional Patterns and Organizations:
Culture and Environment, Lynne G. Zucker, ed. Cambridge, Massachusetts: Ballinger, pp.
71–98.
Egger, Ed. 1999. “Old Ways of Planning, Thinking Won’t Work in Today’s Volatile Health Care
Industry.” Health Care Strategic Management (September): 18–19.
Houseman, Susan. 2001a. “The Benefits Implication of Recent Trends in Flexible Staffing Arrangements.”
Working paper 2001–19, The Pension Research Council, The Wharton School, University of
Pennsylvania..
———. 2001b. “Why Employers Use Flexible Staffing Arrangements: Evidence from an Employer
Survey.” Industrial and Labor Relations Review 55(1): 149–170.
———. 1999. Flexible Staffing Arrangements: A Report on Temporary Help, On-Call, Direct-Hire
Temporary, Leased, Contract Company, and Independent Contractor Employment in the
United States. Report prepared for the U.S. Department of Labor, Office of the Assistant
Secretary for Policy, August.
Houseman, Susan N., Arne L. Kalleberg, and George A. Erickcek. 2001. “The Role of Temporary Help
Employment in Tight Labor Markets.” Working paper no. 01-73, W.E. Upjohn Institute for
Employment Research, Kalamazoo, Michigan.
Kalleberg, Arne L., Barbara F. Reskin, and Ken Hudson. 2000. “Bad Jobs in America: Standard and
Nonstandard Empoyment Relations and Job Quality in the United States.” American Sociological
Review 65: 256–278.
39
Kalleberg, Arne L., Jeremy Reynolds, and Peter V. Marsden. 1999. “Externalizing Employment: Flexible
Staffing Arrangements in U.S. Organizations.” Unpublished paper, University of North Carolina
at Chapel Hill.
Segal, Louis M.. 1996. “Flexible Employment: Composition and Trends.” Journal of Labor Research
17(4): 525–542.
Segal, Louis M.. and Daniel G. Sullivan. 1997. “The Growth of Temporary Services Work.” Journal of
Economic Perspectives 11(2): 117–136.
———. 1998. “Wage Differentials for Temporary Services Work: Evidence from Administrative Data.”
Working paper WP-98-23, Federal Reserve Bank of Chicago, Chicago, Illinois.
Sunseri, Reid. 1999. “Outsourcing on the Outs.” Hospitals and Health Networks 73(10): 46, 48, 50–52.
U.S. Department of Labor. 1998. Employment and Earnings (January), Table 39.
U.S. Department of Transportation, Federal Highway Administration, Highway Statistics 2000,
http://www.fhwa.dot.gov/ohim/hs00/mv10.htm.
U.S. General Accounting Office. 1996. School Lunch Program: Role and Impacts of Private Food
Service Companies. GAO/RCED-96-217.
40
Table 1 Use of Temporary Help among Auto Suppliers and Hospitals
Auto Suppliers
(As a percent of production employment)
Auto Supplier A 3–6
Auto Supplier B 22
Auto Supplier C 22
Auto Supplier D 0
Auto Supplier E
a
0–9
Hospitals
b
(As a percent of hours worked within occupation)
Food Service
Hospital A
Not available
Hospital E 11
Housekeeping
Hospital A Not available
Hospital E 5
Nursing and Medical Assistants
Hospital A
4
Hospital B
7
Hospital C
2
Hospital D 5
Hospital E 0
Clerical: Unit Admin. Support
Hospital A 8
Hospital B 0.4
Hospital C
0.3
Hospital D
7
Hospital E
11
a
At the time of our interview, Auto Supplier E had recently terminated all of its agency help.
b
The contract manager of food and environmental services of Hospital A reported using agency temporaries, but data
on these hires were not available. Hospital F did not maintain any central records of temporary agency use, and so is
excluded from this table.
41
Table 2 Outsourcing in Hospitals and Public Schools
Food service
Custodial/environmental
services Transportation Other
Hospital A Management Management All of security
Laboratory (some)
Landscaping
/Groundskeeping
Hospital B Management Laboratory (some)
Hospital C Management Landscaping
Hospital D Management
Hospital E Management Management All of security
Hospital F Management
Public School A Management
Public School B Entire Service
Public School C Entire Service
Public School D Entire Service
Public School E Management
Public School F Entire Service
Public School G Management Entire Service in Half of
Buildings
42
Table A-1 Wages and Benefits in Low-Skilled Occupations, Auto Suppliers, Hospitals, and Public
Schools
Occupation Starting wage Health plan offered Union
Auto Suppliers
Production
Auto Supplier A $10.60 Yes No
Auto Supplier B
$9.62 Yes No
Auto Supplier C
$13.28 Yes No
Auto Supplier D
$15.51 Yes Yes
Auto Supplier E
$12.35 Yes Yes
Public Schools
Food Service
Public School A $6.82 No Yes
Public School C $7.13 No No
Public School D
a
$6.00 No No
Public School E
$7.40 Yes
Public School G
$7.22 No Yes
Custodial
Public School A
$8.40 Yes
Public School C $9.57 Yes No
Public School D $11.49 Yes Yes
Public School E $12.83 Yes
Public School G $10.16 Yes Yes
Bus Drivers
Public School A
$11.31 No Yes
Public School C
a
Public School D
$11.42 Yes Yes
Public School E $7.42 No Yes
Public School G $11.25 Yes Yes
Table A-1 (Continued)
Occupation Starting wage Health plan offered Union
43
Hospitals
Food Service
Hospital A $6.40 PT prorated, 14% PT No
Hospital B
$7.00 16% with partial benefits; 26% without benefits No
Hospital C
$8.03 9% without benefits No
Hospital D
$7.05 20% with partial benefits; 22% without benefits No
Hospital E
$6.48 PT prorated, 22% PT No
Hospital F $6.60 PT prorated, 65% PT Yes
Housekeeping
Hospital A $6.40 PT prorated, 8% PT No
Hospital B $7.00 14% with partial benefits; 1% without benefits No
Hospital C
$7.95 5% without benefits No
Hospital D
$7.05 10% with partial benefits; 5% without benefits No
Hospital E
$6.48 PT prorated, 18% PT No
Hospital F
$6.60 PT prorated, 27% PT Yes
Nurse Assistant
Hospital A $7.20 PT prorated, 22% PT No
Hospital B $7.40 10% with partial benefits; 12% without benefits No
Hospital C $7.90 27% without benefits No
Hospital D $8.67 23% with partial benefits; 4% without benefits No
Hospital E
$6.74 PT prorated, 5% PT No
Hospital F
$8.96 PT prorated, 46% PT Yes
Clerical
Hospital A
$7.35 PT prorated, 3% PT No
Hospital B $8.00 13% with partial benefits; 23% without benefits No
Hospital C $8.98 17% without benefits No
Hospital D $8.23 33% with partial benefits; 5% without benefits No
Hospital E $7.29 PT prorated, 22% PT No
Hospital F
$7.50 PT prorated, 48% PT No
Note: Data reported are for lowest-skilled occupation within a functional area.
a
Occupation outsourced.
