WORK /AND

10.1177/0730888404271901
Handel PERCEIVED
OCCUPATIONS
JOB QUALITY
/ February 2005

Trends in Perceived Job Quality,

1989 to 1998

MICHAEL J. HANDEL
University of Wisconsin–Madison

There is significant controversy over recent trends in the material and intrinsic quality of jobs.
Neo-Fordist theories argue that material conditions such as pay, job security, promotion oppor-
tunities, and effort requirements have deteriorated for most of the workforce. Post-Fordist theo-
ries argue that new work systems are raising levels of intrinsic rewards such as job challenge,
autonomy, and cooperation and are also offering higher wages. This article tests both theories
using repeated cross-sectional data from the General Social Survey. Results suggest workers’
perceptions of the quality of their jobs remained remarkably stable on most dimensions. It is pos-
sible that this reflects various obstacles to recognizing objective changes such as flexible stan-
dards of evaluation. But stability may indicate that both schools of thought overestimate the
extent of recent changes.

Keywords: job quality; job satisfaction; neo-Fordism; post-Fordism

T here is general agreement that work in the United States has changed
significantly in the past 15 years in response to economic shocks, but
there is great controversy over whether the quality of jobs in the United States
has generally improved or declined. The first half of the postwar period,
sometimes referred to as the golden age of American capitalism, was one of
American economic dominance, rising productivity and earnings, stable or
declining inequality, and robust employment growth in the context of a gen-
erally stable industrial system, which is now often called Fordism for its reli-
ance on mass production, Tayloristic jobs, and bureaucratic management.
The second half of the postwar period was one of crisis and change as the
United States faced greatly increased international competition, declining

Author’s Note: I would like to thank Charles Halaby, Larry W. Hunter, Erik Olin
Wright, and participants at the University of Wisconsin Economic Sociology Seminar
for helpful comments and suggestions. Correspondence concerning this article
should be addressed to Michael J. Handel, Department of Sociology, 1180 Observa-
tory Drive, Madison, WI 53706; e-mail: mhandel@ssc.wisc.edu.
WORK AND OCCUPATIONS, Vol. 32 No. 1, February 2005 66-94
DOI: 10.1177/0730888404271901
© 2005 Sage Publications

66
 Handel / PERCEIVED JOB QUALITY 67

productivity growth, and pressure on both labor costs and existing manage-
ment practices. Two broad schools of thought have characterized the changes
in management and the implications for workers in very different terms.
One school of thought, which may be termed neo-Fordist theory, argues
that job quality declined as business responded to crises in the 1980s and
early 1990s by rolling back many of labor’s postwar gains and by institution-
alizing a “lean and mean” philosophy of employment relations (Harrison,
1994).
Another school of thought, which may be termed post-Fordist theory,
argues that job quality likely improved due to changing product markets, the
diffusion of information technology, and the adoption of participative man-
agement practices, though the extent of change is argued to be contingent on
management strategy. The theory recognizes that managers have the power
to choose whether to respond to altered conditions by implementing changes
that are mutually beneficial for employees and firms (Appelbaum, Bailey,
Berg, & Kalleberg, 2000; Osterman, 1994; Piore & Sabel, 1984; Zuboff,
1988).
These theories contain clear and contrasting expectations for trends in
employee well-being, but existing research generally neglects systematic
study of workers’ own perceptions of the quality of their jobs. Most studies
use standard representative data with information on changes in objective
stratification outcomes such as pay and job tenure (e.g., Neumark, 1999) or
employ qualitative case studies of the objective and subjective consequences
of change for employees in a particular workplace (e.g., Graham, 1993).
How a large sample of workers that is representative of the workforce itself
understands changes in the quality of its jobs is important for understanding
overall trends in well-being, but it is a neglected dimension of the debate
between neo- and post-Fordist perspectives.
This article uses recently available data to examine trends in subjective
perceptions of job quality for a nationally representative sample to shed light
on this question. The next section describes the neo-Fordist and post-Fordist
accounts of recent systemic change in the American workplace and notes
some issues in the use of subjective indicators. The next two sections discuss
the data, methods, and results. The final sections summarize the results and
draw conclusions for further research.
Despite various theoretical expectations that different dimensions of job
quality improved or deteriorated significantly, the findings indicate surpris-
ingly little change in workers’ views of the quality of their jobs during the
1990s both in terms of overall job satisfaction and in terms of 12 specific fac-
ets of job quality. Possible reasons for stability of perceptions and its signifi-
cance are explored in the discussion and conclusion.
68 WORK AND OCCUPATIONS / February 2005

THEORY AND PREVIOUS RESEARCH

Neo-Fordist theory claims that the quality of jobs for most workers, par-
ticularly in terms of material rewards and work pace, declined in the last 20
years. Advocates of this view argue that mean earnings stagnated or fell for
most categories of workers and that earnings inequality grew. Job security
and internal labor markets deteriorated due to downsizing, outsourcing, and
the increased use of contingent employment. Work loads increased as
employers sought to extract more effort from workers without a commensu-
rate rise in pay. These trends were facilitated, in part, by a rollback in union
influence. Although employers were initially motivated to make these
changes by economic difficulties from the late 1970s to the early 1990s, they
found these techniques so lucrative that they continued to pursue them even
after profits recovered (Graham, 1993; Harrison, 1994; Harrison & Blue-
stone, 1988; Mishel, Bernstein, & Schmitt, 2001; Schor, 1991; Taplin, 1995;
for relevant reviews, see Kalleberg, 2000; Morris & Western, 1999).
This strategy has been called neo-Fordist because the basic principles of
the traditional firm remain intact and some principles, such as inequality
between management and labor, are accentuated. Benefits such as relatively
high pay, job security, and career mobility that were extended to workers dur-
ing the growth years of the postwar period are pared back and apply to a
smaller proportion of workers as corporations seek lower costs and leaner
workforces. Postwar labor markets were always segmented into more and
less privileged groups because high wage employers used low wage subcon-
tractors to buffer fluctuations in demand, but the greater reliance on subcon-
tractors under neo-Fordism exacerbates this dualism and increases the pro-
portion of workers in the low-wage segment (Harrison, 1994; cf. Berger &
Piore, 1980; Edwards, 1979). The new practices also represent an intensifica-
tion of Taylorism and a return to prewar managerial prerogatives such as
tighter work standards, speed ups, and greater management power to flexibly
deploy personnel within firms (Dohse, Jürgens, & Malsch, 1985; Graham,
1993; Taplin, 1995). These developments are assumed to apply to most
employees but not to the most privileged, so tests of this theory will include
analyses of trends in inequality of job rewards as well as of changing means.
Aspects of the neo-Fordist position receive varied levels of support from
other research. Most researchers, regardless of theoretical position, agree
that except for a modest increase after 1995, mean earnings stagnated or fell
for most categories of workers since the early 1970s in stark contrast to the
previous 25 years. Figure 1 shows the dramatic shift in the pattern of mean
earnings growth as the steady progress of the first half of the postwar period,
an average annual growth rate of 2.26%, gave way to the current extended
 Handel / PERCEIVED JOB QUALITY 69

Trends in Real Hourly Wages, 1947-2000 ($2000)

15

14.5

14

13.5

13

12.5

12
Real Wage

11.5

11

10.5

10

9.5

8.5

7.5
1947
1949

1951
1953
1955

1957
1959

1961

1963

1965
1967

1969

1971

1973

1975

1977

1979
1981
1983
1985

1987
1989

1991

1993
1995

1997
1999
Year

Figure 1: Trends in Real Hourly Wages, 1947 to 2000 ($2,000)

SOURCE: Economic Report of the President (1989, 2001).

period of general stagnation (author’s calculations; Economic Report of the

President, 2001).
Researchers also agree that the recession of the 1980s was deeper than any
since the Great Depression and had dramatic effects on the U.S. employment
system such as acceleration in the rate of union decline. Earnings inequality
also grew rapidly in the 1980s, reaching levels unprecedented in the postwar
period. The pattern in the 1990s was somewhat more complex. Overall mea-
sures of inequality grew modestly because the more robust economy gener-
ated wage gains for those in lower decile relative to the median, which had an
equalizing effect on the earnings distribution, but a growing gap between the
top decile and the median in the same period had an offsetting disequalizing
effect (Bernstein & Mishel, 1997; Katz & Autor, 1999; Katz & Murphy,
1992).
There is now general agreement that job security and internal labor mar-
kets also eroded in the 1990s, but the effect fell disproportionately on white-
collar and more educated workers, narrowing some of the insecurity gap
between more and less privileged workers (Aaronson & Sullivan, 1998;
Farber, 1997; Heckscher, 1995; Hirsch, 1993; Mishel et al., 2001; Neumark,
1999; Osterman, 1996).
70 WORK AND OCCUPATIONS / February 2005

Claims regarding increased work loads and stress are more disputed,
although there is agreement that new quality control techniques, often criti-
cized as work intensification devices, diffused widely in the 1980s and 1990s
(Bluestone & Rose, 1998; Cappelli & Neumark, 2001; Graham, 1993;
Mishel et al., 2001; Osterman, 2000; Robinson & Bostrom, 1994; Schor,
1991; Taplin, 1995; cf. Green, 2001).
In short, research using objective indicators suggests significant work-
place changes in the direction hypothesized by neo-Fordist theory. However,
there are almost no studies that seek to understand how employees as a whole
understand or evaluate any of these trends (see Schmidt, 1999, for an
exception).
Post-Fordist theory describes an alternative model of employment rela-
tions that benefits both firms and workers and is claimed to have emerged in
response to the same economic crisis described by neo-Fordist theory. The
post-Fordist account implies rising job quality both with respect to certain
material dimensions (e.g., pay, job security) and, more strongly, with respect
to intrinsic rewards (e.g., job challenge, autonomy, workplace cooperation)
and working conditions (e.g., decreased physical workloads). According to
this view, product market shifts favoring higher quality and more customized
goods and the spread of information technology have led to a rollback of tra-
ditional Taylorist job design and to a greater use of employee involvement
practices such as self-directed work teams. These new arrangements increase
job skill requirements, task variety, and job autonomy while decreasing gross
physical effort demands. The result is greater workplace cooperation and job
enrichment and, in some accounts, higher wages (Appelbaum & Batt, 1994;
Appelbaum et al., 2000; Osterman, 1994, 2000; Piore & Sabel, 1984; Zuboff,
1988; for a review, see Smith, 1997).
Some of these ideas build on earlier theories of postindustrialism and
postmaterialism. These theories predicted that jobs would become more
skilled, remunerative, and enriched as a result of the rising education levels of
workers, the increasing proportion of professional and managerial jobs, and
the perceived value shift among workers toward more meaningful or intrinsi-
cally rewarding work rather than the traditional focus on higher pay (Bell,
1976; United States Department of Health, Education, and Welfare, 1973).
The perspective has been called postindustrial or post-Fordist because it
views the recent economic crisis as demonstrating that Taylorism and
adversarial employment relations inherited from the industrial era are obso-
lete. Changing markets and information technology exert pressures to break
with Fordism and to institute genuine policies of job enrichment, decentral-
ized decision making, employee teamwork, and management-labor
cooperation.
 Handel / PERCEIVED JOB QUALITY 71

Many proponents are careful to note that broad diffusion of this system is
contingent on management’s willingness to relinquish traditional preroga-
tives. The pressures to restructure employment relations do not necessarily
translate into real changes in practice (Piore & Sabel, 1984; Zuboff, 1988).
Management has discretion or agency in deciding whether or how to adapt to
these pressures. Nevertheless, there is some expectation and hope of a
positive trend.
Evidence from a number of studies supports aspects of post-Fordist the-
ory. The traditional postindustrial indicators of improved job quality, namely,
the education levels of workers, the proportion of professionals and manag-
ers in the workforce, and the use of computer technology at work, have all
increased substantially in recent years (Autor, Katz, & Krueger, 1998).
Research indicates that employees want to be more involved in workplace
decisions and that those who participate in employee involvement programs
generally endorse them and report greater job satisfaction, autonomy, and
intrinsic rewards on average (Appelbaum et al., 2000; Cotton, 1993; Free-
man & Rogers, 1999; Godard, 2001; Hodson, 1996, 2001). The number of
workplaces using such practices has also grown over time. According to one
time-series estimate, the percentage of workplaces in which more than 50%
of employees participated in quality circles and job rotation doubled from
27% in 1992 to 56% in 1997, and many or most of the programs present in
1992 were less than 5 years old (Osterman, 1994, 2000). Other studies sug-
gest much less dramatic but still significant levels and growth of employee
involvement practices during the 1990s (Cappelli & Neumark, 2001).
At the same time, some of the same studies suggest that many workers
doubt that management is committed to serious change and believe that many
programs are token and ineffectual and have offsetting negative implications,
such as increased stress, for workers (Barker, 1993; Cotton, 1993; Freeman
& Rogers, 1999; Godard, 2001). Neoinstitutional theory suggests that the
apparently wide diffusion of employee involvement practices may mask a
paucity of substantive change either positive or negative if adoption is ritual-
istic or if organizational commitment is weak (DiMaggio & Powell, 1983;
Meyer & Rowan, 1977). Different studies identify bandwagon effects, mana-
gerial resistance, and poor implementation as reasons for limited impact
(Abrahamson & Fairchild, 1999; Vallas, 1999, 2003; Zuboff, 1988). Find-
ings regarding the impact of employee involvement on wages are quite
mixed, and the debate remains unsettled (see Handel & Gittleman, 2004, for
a review). As with the debates over neo-Fordism, there are no studies of the
workforce as a whole that examine trends in how employees view the quality
of their jobs as it relates to job enrichment, decreased physical effort, and
increased cooperation implied by the post-Fordist account.
72 WORK AND OCCUPATIONS / February 2005

In sum, a great deal of research has made claims that the general quality of
jobs has either declined or improved on various dimensions, but little
research has actually examined changes over time in how workers them-
selves perceive the quality of their jobs.
This gap is significant for several reasons. On the most basic level, it is
important to know how the workers in these debates actually understand their
situation. All research on the changing objective characteristics of jobs
implicitly assumes that improved or deteriorating conditions translate
readily into subjective feelings of gain or loss in well-being on the part of
workers, but this assumption is rarely tested and is now widely recognized
within psychology as problematic (Kahneman, Diener, & Schwarz, 1999).
For anyone seeking to draw conclusions regarding actors’ well-being or util-
ity, which are subjective concepts, this must be considered a significant omis-
sion. Indeed, an increasing number of economists, motivated by the recogni-
tion that the ultimate standard of economic outcomes is their contribution to
subjective well-being, have begun to explore what they call the economics of
happiness. As one economist said simply, “Economic things matter only in
so far as they make people happier” (Oswald, 1997; see also Easterlin, 1996;
Frey & Stutzer, 2002). Sociologists have also begun to examine trends in sub-
jective well-being in relation to income inequality (Hout, 2003). However,
the research within economics has been conducted in complete isolation
from the research on inequality growth, and sociological studies also do not
focus on the relationship between trends in subjective well-being and
workplace-related changes specifically.
Also speaking to the importance of understanding subjective perceptions
of job quality is the rather puzzling scarcity of open popular protest regarding
the economic hardships identified by neo-Fordist theory and confirmed by
subsequent research, coinciding with the nearly complete dominance of free-
market doctrines in the political and economic discourse. Indeed, examining
Figure 1, one would be hard pressed to find a more vivid illustration of
Davies’s (1962) famous J-curve model relating relative economic depriva-
tion to rising discontent. Some contend that the main explanation for this
political quiescence is the decline in support for activist government among
the White working class, who continue to support interventionist economic
policies in principle but believe they are ineffective in practice or have bene-
fited other groups (Teixeira & Rogers, 2000). Understanding trends in
perceptions of job quality is another important piece of this puzzle.
Finally, on a methodological level, if one is interested in tracking well-
being over time using alternative indicators, some understanding of the pos-
sibilities of subjective measures of well-being is necessary.
 Handel / PERCEIVED JOB QUALITY 73

Nevertheless, subjective measures also have clear limitations. There are

issues of biases toward socially desirable responses in measures of job satis-
faction, for example. Issues surrounding the relationship between subjec-
tively perceived and objective reality become more difficult when subjective
items are used as proxies for difficult-to-measure objective characteristics
such as a job’s level of task variety. But little is known about whether such
issues are important with respect to the debate over job quality, and the new
availability of trend data, in which some biases might difference out, presents
an opportunity to examine a new source of information on an important
subject.

DATA AND METHOD

This study uses trend data from the General Social Survey (GSS) to exam-
ine whether changes in employees’ perceptions of job quality correspond
to neo-Fordist or post-Fordist predictions. The GSS has a full time series
using a single item on overall job satisfaction that extends back to 1972. The
GSS also administered 12 more detailed items in two modules on work orien-
tations administered in 1989 and 1998. These items can be grouped into
four dimensions of job quality that bear on the debates of neo-Fordism and
post-Fordism. They measure employees’ perceptions of material rewards
(pay, job security, promotion opportunities), intrinsic rewards (interesting
job, job autonomy), other working conditions (stress, work load, physical
effort, danger), and the quality of workplace interpersonal relations
(management-employee relations, coworker relations). These categories are
well established in the research literature as key dimensions of job quality
and as different facets of the more general construct of overall job satisfaction
(Clark, in press; Frey & Stutzer, 2002, p. 103; Kalleberg, 1977; Kalleberg &
Reynolds, 2001; Quinn & Staines, 1979; Rosenthal, 1989). Human resource
professionals specializing in job analysis and job evaluation also use most of
these concepts and similar indicators to rate jobs in determining compensa-
tion levels (Milkovich & Newman, 1993) or to understand determinants of
turnover, organizational climate, and job satisfaction.
More importantly, the items in these categories directly test for the differ-
ent predictions of neo-Fordist and post-Fordist theories. Neo-Fordist theory
emphasizes deterioration and increased inequality in material rewards but
also predicts increased routinization and stress and heavier work loads result-
ing from the spread of new management techniques such as just-in-time pro-
duction, peer monitoring, and continuous quality improvement programs
74 WORK AND OCCUPATIONS / February 2005

(Barker, 1993; Dohse et al., 1985; Graham, 1993; Taplin, 1995). Post-Fordist
theory emphasizes improvements in the quality of work life such as job chal-
lenge and autonomy, decreased physical effort, and improved workplace
interpersonal relations. It also focuses on potential improvements in pay and
job security that result from the higher levels of training, commitment, and
discretionary effort required from employees (Appelbaum et al., 2000; cf.
Akerlof, 1986).
Because the two accounts present generally positive or negative overall
characterizations of trends in job quality as well as more specific predictions,
one would expect employees’ responses to both the overall job satisfaction
and more specific items to change over time depending on which account
was more accurate, allowing for the possibility of offsetting effects, which
are investigated below. The 2 years of data on these questions (1989, 1998)
cover a period of interest to the job quality debate and are sufficiently far
apart that one would expect that significant changes in objective job condi-
tions would register in employees’ evaluations. These changes include those
highlighted by both neo-Fordist theory such as downsizing and increased
employment insecurity and post-Fordist theory such as the spread of
computers and employee involvement practices.
The analyses test for simple time trends in both the means and standard
deviations of all items because of the neo-Fordist implication that inequality
in job quality rose over time and because of the need to test for the possibility
of an increasing bifurcation of employment systems as a result of different
employers choosing to adopt neo-Fordist or post-Fordist strategies. Regres-
sion models with a year dummy for 1998 also control for compositional
changes in the sample that may suppress trend effects (education, age, gen-
der, race, marital status, region, industry, and occupation). The analyses also
include models that interact the year dummy with earnings and occupation,
which are reported where significant, to test for whether inequality in job
rewards by strata increased over time. The regressions are ordinary least
squares regressions for ease of interpretation, but comparison with identical
models using an ordinal logit specification indicates few substantive differ-
ences, though t-statistics are sometimes moderately higher in the logit mod-
els.1 For all analyses, the sample is all wage and salary workers responding to
the items in the years indicated.
 Handel / PERCEIVED JOB QUALITY 75

TABLE 1: Correlations Between Job Satisfaction and Other Dimensions of

Perceived Job Quality

1989 1998

Material rewards
My income is high .21 .21
My job is secure .19 .20
My opportunities for advancement are high .26 .29
Nonmaterial rewards
My job is interesting .49 .47
I can work independently .21 .22
Working conditions
How often work is stressful –.08 –.09
How often come home from work exhausted –.12 –.11
How often do hard physical work –.19 –.13
How often work in dangerous conditions –.11 –.13
How hard do you work .15 .14
Positive interpersonal relations
Management and employees .29 .27
Among coworkers .23 .19

NOTE: Samples sizes are roughly 725 for 1989 and 680 for 1998. All correlations are
significant at the .05 level.

RESULTS

JOB SATISFACTION

The most general measure of perceived job quality is overall job satis-
faction. Most researchers consider it a weighted sum of the various specific
criteria actors use to evaluate their jobs (Clark, in press; Kalleberg, 1977).
Correlations among overall satisfaction and the more specific job quality
variables in the GSS generally support this view, although not all job facets
affect overall job satisfaction equally (see Table 1). Job satisfaction is most
strongly associated with interesting work, followed by positive management-
employee relations and promotion opportunities and then by several job char-
acteristics including subjective pay evaluation, job security, independent
work, and positive relations among coworkers. To different degrees, then,
overall job satisfaction is a summary measure that incorporates workers’
evaluations of both the material and nonmaterial rewards discussed by neo-
Fordist and post-Fordist theories.
Turning to consideration of trends, if neo-Fordist theory is correct, one
would expect to observe declining levels of job satisfaction reflecting the pre-
dominantly negative view of developments in that account. By contrast, if
76 WORK AND OCCUPATIONS / February 2005

Trends in Job Satisfaction, 1989-1998 (GSS)

50 3.5

45 3.45

40 3.4

35 3.35

30 3.3
Percentage

Mean
25 3.25

20 3.2
Very Sat.
Mod. Sat.
15 3.15
A little Dis.
Very Dis.
10 3.1 Mean

5 3.05

0 3
1989 1990 1991 1992 1993 1994 1995 1996 1997 1998
Year
Note: Figures for 1992, 1995, and 1997 are interpolations

Figure 2: Trends in Job Satisfaction, 1989 to 1998 (General Social Survey)

NOTE: Figures for 1992, 1995, and 1997 are interpolations. The trend line for mean
uses a right-hand scale; all other series use a left-hand scale.

post-Fordist theory is correct, one would expect to observe job satisfaction

rise over time.
The GSS has used a consistent job satisfaction question for all survey
years since 1972, and Figure 2 plots the mean and precentage of specific
responses for the period of 1989 to 1998, the period covered by the more spe-
cific items in the work orientations modules. This item elicits the most posi-
tive responses of any examined here and shows no evidence of a time trend.
An average of 86% of workers report being very or moderately satisfied with
their jobs over the full period, and there is little systematic variation across
survey years. Neither the mean nor the standard deviation for 1989 and 1998
are significantly different from one another (see Table 2).2
One possibility is that the apparent constancy in the overall numbers
masks offsetting changes resulting from compositional shifts. The workforce
has become more white collar and female over time, for example, and these
characteristics are associated with higher levels of job satisfaction (Phelan,
1994), which may suppress a negative time trend within groups resulting
from the kinds of changes cited by neo-Fordist theory. The regression results
in the first column of Table 3, which control for a range of characteristics,
indicate that this is not the case. There appears to be no trend whatsoever in
reported job satisfaction, either positive or negative, for 1989 to 1998.
 Handel / PERCEIVED JOB QUALITY 77

TABLE 2: Trends in Perceived Job Quality, 1989 to 1998

1989 1998 1989 to 1998

a
1. Overall job satisfaction 3.30 3.28 –0.02
(0.78) (0.80) 0.02
b
Material rewards
2. My income is high 2.74 2.71 –0.03
(0.99) (1.01) 0.02
3. My job is secure 3.90 3.77 –0.13*
(0.98) (1.01) 0.03
4. My opportunities for advancement are high 2.93 2.88 –0.05
†
(1.10) (1.05) –0.05
b
Nonmaterial rewards
5. My job is interesting 3.84 3.77 –0.08
(0.97) (0.96) –0.01
6. I can work independently 3.86 3.88 0.03
(0.97) (0.93) –0.04
c
Working conditions
7. How often find work stressful 3.29 3.28 –0.01
(0.96) (0.97) 0.01
†
8. How often come home from work exhausted 3.28 3.36 0.08
(0.81) (0.84) 0.03
9. How often have to do hard physical work 2.50 2.48 –0.02
(1.19) (1.26) 0.07*
10. How often work in dangerous conditions 2.04 2.07 0.03
(1.18) (1.19) 0.01
d
11. How hard do you work 2.53 2.44 –0.08*
(0.63) (0.65) 0.02
e
Interpersonal relations
12. Management-employee relations 3.76 3.85 0.08†
(0.88) (0.95) 0.07*
13. Coworker relations 4.01 4.12 0.11**
(0.75) (0.76) 0.01

NOTE: Standard deviations are in parentheses. Data are from the General Social Sur-
vey Modules on Work Orientation (1989, 1998). The income question refers to income
from work, not total income. Figures in the final column are computed as (1989-1998) ÷
SD1989. Sample sizes for items vary from 713 to 734 for 1989 and from 681 to 687 for
1998.
a. Response choices are 1 = very dissatisfied, 2 = a little dissatisfied, 3 = moderately sat-
isfied, and 4 = very satisfied.
b. Response choices are 1 = strongly disagree, 2 = disagree, 3 = neither agree nor dis-
agree, 4 = agree, and 5 = strongly agree.
c. Response choices are 1 = never, 2 = hardly ever, 3 = sometimes, 4 = often, and 5 =
always.
d. Response choices are 1 = I work only as hard as I have to, 2 = I work hard, but not so
as to interfere with the rest of my life, and 3 = I make a point of doing the best work I can
even if it sometimes does interfere with the rest of my life.
e. Response choices are 1 = very bad, 2 = quite bad, 3 = neither good nor bad, 4 = quite
good, and 5 = very good.
†p < .10. *p < .05. **p < .01.
78 WORK AND OCCUPATIONS / February 2005

TABLE 3: Effects of Year and Covariates on Material and Nonmaterial

Dimensions of Job Quality (1989, 1998)

Job High Secure Promotion Interesting Independent

Satisfaction Pay Job Chances Work Work

Year (1 = 1998) –0.0041 –0.0071 1.5042* –0.0111 –0.1045* –0.0161

(.0440) (.0534) (.6052) (.0590) (.0510) (.0510)
Earnings 0.0853 ** 0.2972 ** 0.1813 ** 0.0618
(natural log) (.0302) (.0367) (.0472) (.0403)
Year earnings* –0.1608**
(.0602)
Technical 0.0293 –0.0471 0.1418 –0.1238 0.0931 –0.3398**
(.1094) (.1328) (.1373) (.1457) (.1235) (.1232)
Sales –0.1033 –0.0916 –0.1761 –0.1061 –0.1186 –0.1201
(.0875) (.1063) (.1106) (.1173) (.1010) (.1008)
†
Clerical –0.0118 –0.1554 0.0864 –0.0591 –0.1972* –0.0090
(.0750) (.0911) (.0946) (.1005) (.0852) (.0849)
Service –0.0175 –0.1446 0.0267 –0.2523* –0.4453** –0.2973**
(.0907) (.1102) (.1144) (.1213) (.1032) (.1032)
Protective 0.0926 –0.6601** 0.1955 –0.2533 –0.1827 –0.3107†
service (.1473) (.1789) (.1860) (.1974) (.1627) (.1647)
†
Farm 0.1706 0.0256 0.4514 0.1300 0.6957** 0.6154*
(.2085) (.2532) (.2633) (.2794) (.2466) (.2456)
Craft 0.1315 –0.2384* –0.0919 0.0478 –0.0957 –0.2775*
(.0955) (.1160) (.1206) (.1285) (.1105) (.1104)
†
Blue collar –0.1601 –0.3281** –0.1442 –0.3153** –0.3756** –0.2533*
(.0878) (.1065) (.1108) (.1175) (.1015) (.1012)
N 1,271 1,272 1,273 1,274 1,370 1,368
2
Adjusted R .047 .127 .054 .078 .113 .082

NOTE: Standard errors are in parentheses. All models control for age, education, gen-
der, race, weekly hours worked, occupation (managers and professionals omitted), in-
dustry, marital status and the interaction of marital status with female, and region. Earn-
ings for both 1989 and 1998 are expressed in 1998 dollars. All dependent variables have
been recoded so that higher values indicate more positive perceptions. Models in the
last two columns also include controls for job values (see Table 4).
†
p < .10. *p < .05. **p < .01.

As with all models in this article, a further sensitivity analysis was per-
formed by alternatively interacting year with occupation and earnings to
determine whether inequality in job quality across strata increased over time
and whether overall stability in the mean of the dependent variable masks off-
setting subgroup changes. The interactions were not significant and therefore
are not reported.
Although temporal stability in reported job satisfaction has been found in
previous decades (Barbash, 1976, p. 22), an examination of the full GSS time
series indicates that relative to the 1970s, reported job satisfaction did decline
 Handel / PERCEIVED JOB QUALITY 79

somewhat in 1980 during the deepest recession since the Great Depression,
but it has since remained stable at that lower level through the 2000 survey.
This pattern remains even when controlling for the same variables used in the
model in Table 3 (author’s calculations). Other surveys also show a signifi-
cant decline in job satisfaction between 1972 and 1977, when the American
economy first entered the period of slow growth (Quinn & Staines, 1979, pp.
210-211).
This raises the question as to whether the more recent stability of overall
job satisfaction masks more facet-specific positive and negative effects
claimed by post-Fordist and neo-Fordist theories that offset one another in
respondents’ summary evaluations. Unlike most previous work on job satis-
faction, the GSS work orientations modules have comparable sets of items on
specific job facets for two points in time that can be used to test this possibil-
ity. However, most facet-specific results confirm the impression of stability
in perceptions of job quality.

MATERIAL REWARDS

Neo-Fordist theory posits declining levels of material rewards resulting

from employer efforts to restrain labor costs and increasing inequality of
rewards, especially earnings. Post-Fordist theory posits higher levels of
skills, training, and workplace cooperation that may raise earnings because
they require higher levels of human capital and discretionary effort (cf.
Akerlof, 1986). How have workers’ own evaluations of their jobs’ material
rewards changed over time?

Earnings. Satisfaction with earnings is an obvious place to begin an exam-

ination of specific job facets. In the GSS, when workers were asked whether
they think their earnings are high, 45% disagreed with that characterization,
the least positive responses of all the items examined here. This finding is
similar to results in other studies (Alwin, 1987; Gruenberg, 1980; Mirowsky,
1987).
However, Table 2 indicates that there was no significant change in work-
ers’ average evaluations of their pay levels or in the variance of responses
between 1989 and 1998. The regression model in the second column of Table
3 also shows no significant time trend. Not surprisingly, log actual earnings
(constant 1998 dollars) have a strong effect on perceptions of high earnings,
but the interaction of year with a linear earnings term is not significant (β =
.003, SE = .058). Although measured earnings inequality increased during
the 1990s, at least in terms of the gap between the top and middle of the
distribution, interaction terms were not individually or jointly significant
80 WORK AND OCCUPATIONS / February 2005

TABLE 4: Trends in Job Values (1989, 1998)

Not Neither Important Very

How Important in a Job is: Important nor Unimportant Important Important

Material
High income
1989 2.8 13.7 60.2 23.4
1998 4.3 16.1 56.3 23.3
Job security
1989 1.1 4.9 42.2 51.9
1998 1.0 3.1 35.7 60.2
Promotion
1989 1.8 5.9 53.2 39.0
1998 2.6 8.6 51.9 36.9
Nonmaterial
Interesting work
1989 1.5 3.8 51.3 43.4
1998 1.4 3.0 44.1 51.5
Independent work
1989 5.3 18.9 49.0 26.8
1998 2.9 15.4 53.7 28.0

NOTE: The category “not important” also includes those responding “not at all impor-
tant,” who account for less than 1% of all responses to each question in both years. Sam-
ple sizes are about 760 for 1989 and 700 for 1998.

when the year dummy was interacted with dummies for income quintiles
(F = .85, p = .49).
In general, then, workers did not evaluate their earnings differently over
time, and this result does not vary by earnings strata despite some increased
inequality in the earnings distribution. This picture is reinforced by GSS data
from the same surveys on respondents’ job values. Table 4 indicates that the
percentage of workers saying they consider high income an important job
characteristic actually declined by about 4% between 1989 and 1998, though
the difference across years is not statistically significant in an ordinal logit
model. The data do not suggest any increase or decrease in workers’ concern
with earnings over the 1990s.

Job security. In contrast to earnings adequacy, Table 2 indicates there was

a significant decline in mean perceptions of job security between 1989 and
1998. The percentage strongly agreeing with the statement that their job is
secure declined from 27% to 22%. This decline is consistent with the
research on declining job security in the 1990s cited earlier and is also nota-
ble because the unemployment rate actually declined from 5.3% in 1989 to
 Handel / PERCEIVED JOB QUALITY 81

4.5% in 1998, which would be expected to move perceptions in the opposite

direction.
The regression model indicates that the change is entirely a function of a
nearly complete erosion of the job security advantage previously held by
those with higher earnings, supporting the impression that corporate down-
sizing made the 1990s a more insecure time for upper income workers than
previously. Although the main effect for the dummy for 1998 is now positive,
the negative interaction with earnings is so strong that perceived job secu-
rity is lower for the top four earnings quintiles in 1998. This is consistent
with analyses of a longer GSS time series on individuals’ perceived probabil-
ity of job loss and with studies of actual job loss in the 1980s and 1990s,
which found both job insecurity and perceptions of insecurity rising for
white-collar and more educated workers but not for less skilled workers in
the 1990s (Aaronson & Sullivan, 1998; Farber, 1997; Schmidt, 1999). The
impression of greater concern with job security is reinforced by the trend data
on job values in Table 4 that shows that the percentage of workers saying they
consider job security to be very important rose by about 8% between 1989
and 1998, an increase that is significant in an ordinal logit model (p < .01).

Promotion opportunities. Neo-Fordist theory also argues that downsizing

and delayering and a lean approach to staffing have led to the decline of inter-
nal labor markets. Table 2 indicates no significant change in mean percep-
tions of promotion opportunities and a slight decline in the variance of
responses to this question. Table 3 shows that the absence of a time trend
remains after controlling for compositional changes. There are no significant
interactions of year with either income or occupation even though percep-
tions of job security for those with higher income do seem to have declined
over time. Table 4 indicates respondents shifted modestly away from consid-
ering promotion opportunities an important job characteristic, but this
change is not significant. If there was a decline in internal labor markets in the
1990s, it did not register with respondents as significant or large enough to
warrant increased concern.
In sum, among indicators of material rewards, the trend toward increased
job insecurity, particularly for those at the upper end of the income distribu-
tion, seems to support theories of declining job quality and is consistent with
results for objective measures of employment insecurity. However, respon-
dents do not perceive either the adequacy of earnings levels or their promo-
tion prospects any differently in 1998 compared to 1989, supporting neither
neo-Fordist or post-Fordist theories on these dimensions of job quality.
For earnings, one possibility suggested by Figure 1 is that workers
respond more to changes in personal earnings levels or trajectories than to
82 WORK AND OCCUPATIONS / February 2005

changes in inequality. This leads one to expect large changes in pay percep-
tions in the early to mid-1970s, when the U.S. economy made the transition
from a regime of steady earnings growth to one of stagnation, rather than in
later decades, even though inequality grew most rapidly then. Although the
GSS did not ask detailed questions about specific facets of job quality in the
1970s, the Quality of Employment Surveys (QES) collected such data in
1969, 1972, and 1977 (Quinn & Staines, 1979, pp. 217ff.). This series covers
both years immediately preceding and following the slowdown in U.S. earn-
ings growth. The data show that the percentage agreeing strongly that their
pay is good dropped from 40% in both 1969 and 1972 to 27% in 1977. There
was a similar decline in reported job security (54% vs. 42%) and a smaller
decline in reported promotion potential (22% vs. 16%). This suggests that
workers’ perceptions of all forms of material rewards shifted downward dur-
ing the 1970s in response to declining economic conditions, but only percep-
tions of job security declined during the 1990s, when neo-Fordist and post-
Fordist theories of economic change crystallized.
However, there is a danger in viewing subjective perceptions as simple
reflections of changing objective conditions. The percentage of QES respon-
dents agreeing strongly that their fringe benefits are good declined from 42%
to 33% even though the actual average level of benefits reported in the same
surveys improved (Quinn & Staines, 1979; Staines & Quinn, 1979). This
suggests perceptions are affected by expectations as well as objective condi-
tions, a point that will be discussed further below.

INTRINSIC REWARDS

If perceptions of job security have declined somewhat and if perceptions

of other material rewards are little changed, perhaps the main trends in the
1990s were felt by workers in the area of nonmaterial or intrinsic rewards.
The GSS contains two measures of intrinsic job quality: the extent to which
respondents find their work interesting and the extent to which respondents
can work independently. Despite competing claims of increased control and
job enrichment, the raw means and standard deviations also show no signifi-
cant change over a nearly 10-year period, although the direction of change is
negative in the case of whether respondents consider their work interesting
(Table 2). Indeed, the percentage of workers agreeing strongly with the state-
ment that their work is interesting declined from 23.1% in 1989 to 19.4% in
1998, with the neutral response (neither agree nor disagree) growing by
roughly the same magnitude.
Controlling for background characteristics, the regression models in
Table 3 indicate that the decline in interesting work is significant once
 Handel / PERCEIVED JOB QUALITY 83

compositional changes in the workforce are controlled. There is no evidence

of an increase in the perceived scope for working independently in either the
raw figures or the regression model. Both findings are contrary to notions that
the intrinsic quality of jobs improved as the result of the greater use of teams
and other employee involvement practices.
One question this raises is whether expectations for more interesting and
autonomous work have risen faster than have changes on the job and masked
genuine changes in working conditions. If objective conditions change in a
positive direction but subjective standards of evaluation rise faster, the result
might be perceptions of declining conditions even when actual conditions
improve.
Table 4 shows that workers’ desires for intrinsic or nonmaterial job re-
wards have increased over time for both measures, and ordinal logit models
confirm these increases are significant (interesting work: p < .01; indepen-
dent work: p < .05). Except for the growing desire for job security, the overall
pattern of results in this table supports the view that workers increasingly
seek intrinsic rather than material rewards despite the long-run stagnation in
earnings and growth in earnings inequality (Inglehart & Abrahamson, 1994;
cf. Easterlin & Crimmins, 1991). An exogenous rise in expectations would
presumably cause perceptions of job quality to decline even if actual job con-
ditions did not change or improved somewhat.
However, the models for intrinsic rewards in Table 3 include the work
values variables as controls. Consequently, the results cannot be attributed to
compositional changes in the proportion of workers desiring more interest-
ing and autonomous jobs. The decline or stability in intrinsic work rewards is
found within categories of workers who express different desires for such job
characteristics.
Consistent with post-Fordist interpretations of information technology,
about two thirds of respondents in the 1998 GSS expressed the belief that
new technology would make work more interesting. However, occupation-
level fixed effects models indicate no relationship between changes in the
percentage of computer users in an occupation and mean perceptions of how
interesting respondents found their work (not shown).3 Likewise, there is no
cross-sectional relationship between computer use and overall job satisfac-
tion, with or without controls for background characteristics, using the GSS
2000 survey (not shown).
There is, then, no evidence that workers perceived their jobs to be more
interesting or autonomous in 1998 compared to 1989, which contradicts the
view that employers are providing more meaningful work either in response
to workers’ needs or because they find that changing product markets or
information technology impels them to offer such work. Overall, controlling
84 WORK AND OCCUPATIONS / February 2005

for the rising value workers placed on this job characteristic, it appears that
employees found their work less interesting over the course of the 1990s.

OTHER WORKING CONDITIONS

Neo-Fordist theory argues that employers have intensified the effort that
they require from workers in response to an increasingly competitive eco-
nomic environment and that employees work more hours to offset declining
pay rates. However, the postindustrial aspect of post-Fordist theory argues
that automation is making work less physically arduous. The GSS has a num-
ber of items on working conditions and effort levels that relate to this aspect
of the job quality debate.
The descriptive data in Table 2 suggest little change in the number of
workers who find work stressful and some increase in the number of workers
who say they come home from work exhausted, but the full model for the lat-
ter in Table 5 suggests that the effect is restricted mostly to clerical workers.
Clerical workers also reported substantially greater stress and physical
demands in 1998 compared to 1989 (Table 5).4 It is unclear what might be
responsible for deterioration in job quality among clerical workers, particu-
larly the reported increase in hard physical work, but it is possible that leaner
staffing and increasingly computer-paced work have increased effort
demands for clerical workers over time. None of these results supports post-
Fordist expectations, though only the results for clerical workers actually
contradict it.
Also contrary to post-Fordist theory, less skilled blue-collar and service
work involved more physical effort than other occupations in 1989. And the
reported frequency of hard physical work actually increased among these
groups by 1998 (Table 5). Similarly, the frequency of dangerous work did not
decline significantly overall between 1989 and 1998 either in raw terms
(Table 2) or after controlling for compositional variables (Table 5), and there
is no significant interaction of year with income or occupation.
Finally, contrary to neo-Fordist theory, the relative number of workers
reporting they worked so hard that it interfered with their nonwork life
declined significantly between 1989 and 1998, and there is no significant
interaction of year with income or occupation (Table 2 and Table 5). It is not
clear how to reconcile this finding with the absence of a similar downward
trend in stress and exhaustion from work. However, it seems fair to say that
there is little evidence of the overworked American in these self-reports (cf.
Schor, 1991), though there does not appear to have been any improvement in
most of these working conditions either, and physical demands seem to have
increased for a number of occupations.
 TABLE 5: Effects of Year and Covariates on Working Conditions and Interpersonal Relations at Work (1989, 1998)

Working Conditions Interpersonal Relations

Stress Exhausted Physical Work Danger Work Hard Management-Labor Coworkers

Year (1 = 1998) –0.0880 0.0478 –0.2106* 0.0897 –0.0790* 0.0966† 0.1246**

Technical –0.4166* –0.1070 0.1232 0.3296* –0.0388 –0.0220 –0.0071
Sales –0.1481 –0.0598 0.0456 –0.0940 –0.0551 0.0947 0.0344
Clerical –0.4218** –0.2014* –0.3900** –0.2062* –0.0592 0.1192 0.0745
Service 0.0181 0.0555 0.6966** 0.4367** 0.0166 0.2369* –0.0059
Protective service –0.1988 –0.1778 –0.3098 0.9008** 0.0073 0.2637 0.1503
† †
Farm –0.3851 0.1251 0.6534 0.5251 –0.0889 –0.0576 –0.0258
† †
Craft –0.2495 –0.0247 0.8671** 0.9024** 0.1241 –0.0144 0.1069
†
Blue collar –0.1956 0.2153 1.0151** 0.7964** –0.0468 –0.0414 0.0085
Interaction
Year by occupation
Technical 0.2086 0.1070 0.1331
Sales 0.0818 0.0386 0.2237
Clerical 0.5814** 0.2339† 0.3880*
Service –0.1465 0.1675 0.4167*
†
Protective service –0.5193 –0.6182* 0.3225
†
Farm –0.7789 –0.3173 0.6121
Craft 0.2746 0.1069 0.1834
Blue collar –0.0041 –0.1794 0.3695*
N 1,393 1,397 1,392 1,391 1,370 1,387 1,365
2
Adjusted R .076 .065 .300 .255 .094 .032 .015

NOTE: All models control for age, education, gender, race, weekly hours worked, occupation (managers and professionals are omitted), industry,
marital status and the interaction of marital status with female, and region. All dependent variables have been recoded so that higher values indicate
more positive perceptions.

85
†
p < .10. *p < .05. **p < .01.
86 WORK AND OCCUPATIONS / February 2005

The dominant picture that emerges from these analyses is the lack of great
change in working conditions generally. Some groups, such as clerical and
blue-collar workers, experienced deteriorating conditions on some dimen-
sions, as neo-Fordist theory predicts, but there is also a broadly shared
perception among workers that their work is not as hard now compared to
previously, as post-Fordist theory implies. These mixed and somewhat in-
consistent results do not give a strong indication as to the trends in these
dimensions of job quality.

QUALITY OF INTERPERSONAL RELATIONS AT WORK

The final dimension of job quality is the quality of interpersonal relations

at work. If teamwork has diffused as widely as post-Fordist theory claims
(Osterman, 1994) and has been successful, one would expect improved re-
lations between management and workers and among coworkers. Indeed,
one of the traditional motivations for introducing teams has been to improve
workplace cooperation and organizational climate. By contrast, if neo-Fordist
theory is correct, one would expect a deterioration in labor-management rela-
tions resulting from continued stagnation and inequality in material condi-
tions and the intensification of work.
Both descriptive results (Table 2) and the regression models (Table 5)
indicate some improvements in relations between workers and employers
and even more so among coworkers between 1989 and 1998.5 These results
do not suggest growing friction between management and workers as neo-
Fordist theory implies and may reflect either positive efforts at work reform,
such as high performance work practices, or simply the improved labor mar-
ket of the late 1990s, which may have done much to improve all social
relations at work.

DISCUSSION

Competing neo-Fordist and post-Fordist theories imply significant and

generally opposing trends in job quality along several dimensions, but work-
ers’perceptions, with some exceptions, were surprisingly stable in the 1990s.
Different theories suggest workers should feel better or worse off than previ-
ously, but workers themselves mostly did not perceive the quality of their
jobs as better or worse than in the past, nor does the stable average mask a
widening inequality in reported job quality. This general stability is also not
explained by changes in the composition of the workforce, nor was the
 Handel / PERCEIVED JOB QUALITY 87

stability in overall job satisfaction explained by offsetting changes in mea-

sures of more specific job dimensions, most of which show a similar stability.
Only in the area of job security is there clear evidence of change in subjec-
tive perceptions corresponding to observed declines in objective measures of
job security for upper-level workers. There seemed to be some tendency to
perceive work as less interesting in the 1990s despite the growth in informa-
tion technology and respondents’ own reported belief that technology would
make work more interesting. There also seemed to be an improvement in
interpersonal relations at work, particularly among coworkers, but also
between management and employees. Whether this is due to more coopera-
tive management practices or to the positive effects of the tighter labor
market is difficult to determine.

CONCLUSION

Although there are many studies on the changing objective qualities of

work and employment, there are few studies that examine trends in how
workers themselves feel about their jobs. The question naturally arises as to
whether the observed stability implies existing theories overstate objective
labor market trends or whether complexities in the ways that objective condi-
tions affect subjective perceptions may contribute to stable perceptions
despite objective changes. The following can only sketch some of these pos-
sibilities for future research to investigate.

COGNITIVE AND AFFECTIVE BARRIERS TO PERCEPTIONS

Social indicators research has often found relatively weak relationships

between objective and subjective measures of quality of life (Schuessler &
Fisher, 1985, p. 145). Coarseness of response categories may introduce mea-
surement error such that only large changes in objective conditions alter the
distribution of responses.
Job satisfaction items may be subject to social desirability response biases
because they force respondents to confront issues of whether they live up
to dominant social standards or their own individual aspirations, as some crit-
ics of John Goldthorpe’s Affluent Worker study and others have argued
(Barbash, 1976; Fantasia, 1988; Freeman & Rogers, 1999, p. 23-24, 43ff.;
Mann, 1973).
In the case of earnings, inflation may dampen dissatisfaction with stag-
nant personal earnings by making it hard to distinguish real from nominal
pay increases, an example of so-called money illusion. Earnings growth
88 WORK AND OCCUPATIONS / February 2005

over the life cycle may also generate a sense of personal advancement that
offsets negative perceptions or inhibits awareness of a flattening trend in age-
earnings profiles across birth cohorts (Osberg, 2000).
Cognitive barriers to perceiving changes in one’s own income relative to
those of others may also dampen reactions to changes in inequality. For many
years, mainstream economists disputed Bluestone and Harrison’s (1982)
claim that inequality grew in the 1980s (e.g., Kosters & Ross, 1987, 1988;
Lawrence, 1983, 1984; Lerman & Salzman, 1988). It would not be surprising
if the general population, with limited access to official statistics and analy-
ses, experienced similar difficulties.
Arguing against the view that reported perceptions are poor reflections of
objective conditions are the significant correlations among perceived earn-
ings levels and actual log earnings (.32) and raw earnings (.40) in the pooled
GSS sample. There is an even larger correlation between general perceptions
of job security and more specific perceptions of the likelihood of job loss in
the next 12 months (.50), a measure whose sample mean itself correlates .85
with the national unemployment rate across the full GSS time series for the
years 1977 to 1996 (author’s calculations; Schmidt & Thompson, 1997, p. 7,
Figure 2). The QES data also showed that as economic conditions deterio-
rated in the 1970s perceptions of material rewards fell as well.
Evidence reviewed earlier suggests that objective changes in job tasks
such as employee involvement practices are also associated in the expected
manner with subjective perceptions such as greater job satisfaction and
autonomy.

ADAPTABLE STANDARDS

Richard Easterlin (1996; Easterlin & Crimmins, 1991) and others have
found that people often adjust their subjective standards in response to
changing objective conditions. Thus, income and self-reported happiness of
Americans were positively correlated in cross-sectional surveys, but happi-
ness did not rise over time despite dramatic gains in average incomes, at least
in the first half of the postwar period. People simply readjusted their stan-
dards to reflect the rising norms of what constituted an adequate income and
evaluated their life circumstances relative to the new standard (Blanchflower
& Oswald, 2000; Easterlin, 1996; Frey & Stutzer, 2002; Oswald, 1997). This
upward readjustment of standards, known as the Easterlin paradox or the
hedonic treadmill, produced stable levels of reported well-being that masked
genuine improvements in objective living conditions. Research indicates that
people also adjust their standards in response to worsening objective condi-
tions. After experiencing misfortune, there is a transition period after which
 Handel / PERCEIVED JOB QUALITY 89

people report generally similar levels of happiness as they had previously, as

if they were returning to a fixed set-point (Kahneman et al., 1999).
Objective changes in life circumstances have a genuine impact on welfare
insofar as individuals would judge themselves better or worse off relative to a
previous norm. It is simply that most people usually compare their situations
to current norms, imparting a stability bias to subjective perceptions. Thus, it
is possible that the growth of upper white-collar jobs, the diffusion of com-
puters, and the growth of employee involvement have made work more inter-
esting and satisfying, but once they are taken for granted, their effects on per-
ceptions diminish even though they represent genuine welfare gains relative
to previous working conditions.
Likewise, adaptation to negative change might also explain the stability in
other perceptions and the broader political quiescence of the last 25 years.
Reflecting on several years of earnings stagnation in the 1970s, Joanne Mar-
tin (1981) predicted workers would experience growing feelings of relative
deprivation that would lead to greater class conflict. However, after another
decade of restructuring, high unemployment, wage stagnation, and inequal-
ity growth, Paul Krugman (1990, p. xi) concluded that the prevailing mood
was one of diminished expectations rather than of frustrated expectations and
consequently greater political passivity.
This view was echoed by Federal Reserve Bank chairman Alan
Greenspan (1998), among others, who concluded that downsizing during the
1990s lowered expectations and led to voluntary wage restraint even when
unemployment fell below levels expected to generate inflationary wage pres-
sures (p. 81; Stiglitz, 1997, p. 7; Uchitelle, 1997). However, attempts to for-
mally test whether increased job insecurity led to wage restraint in the 1990s
have produced weak results (Aaronson & Sullivan, 1998, 2000; Schmidt &
Thompson, 1997).
By way of conclusion, one might say that although the competing expla-
nations cannot be fully adjudicated here, it seems more likely that stability in
objective conditions, rather than cognitive and affective constraints, accounts
for the stability of subjective perceptions. On balance, individuals’ percep-
tions of both material and intrinsic job rewards seem reasonably well-
grounded in objective conditions.
Thus, it seems most plausible to conclude that earnings perceptions did
not change greatly during the 1990s because absolute earnings were also sta-
ble and because changes in inequality were either sufficiently subtle to
remain relatively unnoticed or not something that workers cared deeply
about when evaluating their personal situation. The decline in perceived job
security among more privileged workers also accurately reflected their
increased job displacement rates.
90 WORK AND OCCUPATIONS / February 2005

The absence of either negative or positive trends in perceptions of intrinsic

rewards suggests that neither more controlling nor genuinely empowering
work practices were more common during the 1990s. Despite superficially
high adoption rates, many employer-reported programs are likely rather
nominal efforts reflecting current management fashion rather than funda-
mental change in the structure of work, as neoinstitutional theory suggests
(Abrahamson & Fairchild, 1999; DiMaggio & Powell, 1983).
Although one cannot rule out the possibility that stability in perceived job
quality reflects the readjustment of evaluative standards in response to
changes that have genuine effects on well-being, the preceding suggests that
the main reason for stable perceptions in the 1990s is a general stability in
objective conditions. Both neo-Fordist and post-Fordist theory seem to have
overstated changes in the workplace and effects on employee welfare, but
this conclusion must be viewed as tentative. Adjudicating this question more
conclusively will require national data collected at different points in time
with consistent and strong measures of workplace organization and workers’
perceptions to decide whether or in which respects the stability in workers’
subjective well-being reflects stability in actual conditions rather than
complexities in workers’ subjective reactions to them.

NOTES

1. Results are available upon request. The General Social Survey (GSS) asks union status for
only two thirds of the GSS sample. Because models not reported here indicated union status was
generally not significant, and because its omission did not alter the coefficients of interest sub-
stantively, it was excluded from the final models.
2. Tests for differences in standard deviations between groups were conducted using Stata’s
sdtest procedure.
3. These models merged data on computer use from the Current Population Survey (1989,
1997) onto the General Social Survey files (1989, 1998) and aggregated both measures of inci-
dence of computer use and interesting work to the occupation level.
4. The omitted occupational group in Table 5 in managerial and professional.
5. In analyses not shown, ordinal logit models show that the year dummy variable is signifi-
cant at the .05 level rather than at the .10 level for the model in which the quality of management-
employee relations is the dependent variable.

REFERENCES

Aaronson, D., & Sullivan, D. G. (1998). The decline of job security in the 1990s: Displacement,
anxiety, and their effect on wage growth. Economic Perspectives, 22(1), 17-43.
Aaronson, D., & Sullivan, D. G. (2000). Unemployment and wage growth: Recent cross-state ev-
idence. Economic Perspectives, 24(2), 54-71.
 Handel / PERCEIVED JOB QUALITY 91

Abrahamson, E., & Fairchild, G. (1999). Management fashion: Lifecycles, triggers, and collec-
tive learning processes. Administrative Science Quarterly, 44, 708-740.
Akerlof, G. A. (1986). Labor contracts as partial gift exchange. In G. A. Akerlof & J. L. Yellen
(Eds.), Efficiency wage models of the labor market (pp. 66-92). New York: Cambridge Uni-
versity Press.
Alwin, D. F. (1987). Distributive justice and satisfaction with material well-being. American So-
ciological Review, 52, 83-95.
Appelbaum, E., Bailey, T. R., Berg, P., & Kalleberg, A. L. (2000). Manufacturing advantage:
Why high-performance work systems pay off. Ithaca, NY: Cornell University Press.
Appelbaum, E., & Batt, R. (1994). The New American Workplace. Ithaca, NY: ILR Press.
Autor, D. H., Katz, L. F., & Krueger, A. B. (1998). Computing inequality: Have computers
changed the labor market? Quarterly Journal of Economics, 113, 1169-1213.
Barbash, J. (1976). Job satisfaction attitudes surveys. Paris: Organisation for Economic Co-
operation and Development.
Barker, J. R. (1993). Tightening the iron cage: Concertive control in self-managing teams. Ad-
ministrative Science Quarterly, 38, 408-437.
Bell, D. (1976). The coming of post-industrial society: A venture in social forecasting. New
York: Basic Books.
Berger, S., & Piore, M. J. (1980). Dualism and discontinuity in industrial societies. New York:
Cambridge University Press.
Bernstein, J., & Mishel, L. (1997). Has wage inequality stopped growing? Monthly Labor Re-
view, 120, 3-16.
Blanchflower, D. G., & Oswald, A. J. (2000). Well-being over time in Britain and the USA. Cam-
bridge, MA: National Bureau of Economic Research.
Bluestone, B., & Harrison, B. (1982). The deindustrialization of America. New York: Basic
Books.
Bluestone, B., & Rose, S. (1998). The unmeasured labor force: The growth in work hours.
Annandale-on-Hudson, NY: Jerome Levy Economics Institute.
Cappelli, P., & Neumark, D. (2001). Do “high-performance” work practices improve establishment-
level outcomes? Industrial and Labor Relations Review, 54, 737-775.
Clark, A. E. (in press). What makes a good job? Evidence from OECD countries. In S. Bazen &
C. Lucifora (Eds.), Job quality and employer behaviour. Basingstoke, UK: Palgrave.
Cotton, J. L. (1993). Employee involvement: Methods for improving performance and work atti-
tudes. Newbury Park, CA: Sage.
Davies, J. C. (1962). Toward a theory of revolution. American Sociological Review, 27, 5-19.
DiMaggio, P. J., & Powell, W. W. (1983). The iron cage revisited: Institutional isomorphism and
collective rationality in organizational fields. American Sociological Review, 48, 147-160.
Dohse, K., Jürgens, U., & Malsch, T. (1985). From “Fordism” to “Toyotism”? The social organi-
zation of the labor process in the Japanese automobile industry. Politics and Society, 14, 115-
146.
Easterlin, R. A. (1996). Does satisfying material needs increase human happiness? In R. A.
Easterlin (Ed.), Growth triumphant (pp. 131-144). Ann Arbor: University of Michigan Press.
Easterlin, R. A., & Crimmins, E. M. (1991). Private materialism, personal self-fulfillment, fam-
ily life and public interest. Public Opinion Quarterly, 55, 499-533.
Economic report of the President. (1989). Washington, DC: Government Printing Office.
Economic report of the President. (2001). Washington, DC: Government Printing Office.
Edwards, R. (1979). Contested terrain. New York: Basic Books.
Fantasia, R. (1988). Cultures of solidarity: Consciousness, action, and contemporary American
workers. Berkeley, CA: University of California Press.
92 WORK AND OCCUPATIONS / February 2005

Farber, H. S. (1997). The changing face of job loss in the United States, 1981-1995. Brookings
Papers on Economic Activity, 1997, 55-128.
Freeman, R. B., & Rogers, J. (1999). What workers want. Ithaca, NY: ILR Press.
Frey, B. S., & Stutzer, A. (2002). Happiness and economics: How the economy and institutions
affect human well-being. Princeton, NJ: Princeton University Press.
Godard, J. (2001). High performance and the transformation of work? The implication of alter-
native work practices for the experience and outcomes of work. Industrial and Labor Rela-
tions Review, 54, 776-805.
Graham, L. (1993). Inside a Japanese transplant: A critical view. Work and Occupations, 20, 147-
173.
Green, F. (2001). It’s been a hard day’s night: The concentration and intensification of work in
late twentieth-century Britain. British Journal of Industrial Relations, 39, 53-80.
Greenspan, A. (1998). Is there a new economy? California Management Review, 41(1), 74-85.
Gruenberg, B. (1980). The happy worker: An analysis of educational and occupational differ-
ences in determinants of job satisfaction. American Journal of Sociology, 86, 247-271.
Handel, M. J., & Gittleman, M. (2004). Is there a wage payoff to innovative work practices? In-
dustrial Relations, 43, 67-97.
Harrison, B. (1994). Lean and mean: The changing landscape of corporate power in the age of
flexibility. New York: Basic Books.
Harrison, B., & Bluestone, B. (1988). The great u-turn: Corporate restructuring and the polariz-
ing of America. New York: Basic Books.
Heckscher, C. (1995). White collar blues: Management loyalties in an age of corporate restruc-
turing. New York: Basic Books.
Hirsch, P. M. (1993). Undoing the managerial revolution? Needed research on the decline of mid-
dle management and internal labor markets. In R. Swedberg (Ed.), Explorations in economic
sociology (pp. 145-157). New York: Russell Sage.
Hodson, R. (1996). Dignity in the workplace under participative management: Alienation and
freedom revisited. American Sociological Review, 61, 719-738.
Hodson, R. (2001). Dignity at work. New York: Cambridge University Press.
Hout, M. (2003). Money and morale: What growing inequality is doing to Americans’ views of
themselves and others (Working paper). Berkeley: University of California, Berkeley.
Inglehart, R., & Abrahamson, P. R. (1994). Economic security and value change. American Polit-
ical Science Review, 88, 336-354.
Kahneman, D., Diener, E., & Schwarz, N. (1999). Well-being: The foundations of hedonic psy-
chology. New York: Russell Sage.
Kalleberg, A. L. (1977). Work values and job rewards: A theory of job satisfaction. American So-
ciological Review, 42, 124-143.
Kalleberg, A. L. (2000). Nonstandard employment relations: Part-time, temporary, and contract
work. Annual Review of Sociology, 26, 341-365.
Kalleberg, A. L., & Reynolds, J. (2001). Work attitudes and nonstandard work arrangements in
the United States, Japan, and Europe. In S. N. Houseman & M. Osawa (Eds.), Growth of non-
standard work arrangements. Kalamazoo, MI: W. E. Upjohn Institute.
Katz, L. F., & Autor, D. H. (1999). Changes in the wage structure and earnings inequality. In
O. Ashenfelter & D. Card (Eds.), Handbook of labor economics (pp. 1463-1555). New York:
Elsevier.
Katz, L. F., & Murphy, K. M. (1992). Changes in relative wages, 1963-1987: Supply and demand
factors. Quarterly Journal of Economics, 107, 35-79.
Kosters, M., & Ross, M. N. (1987). The influence of employment shifts and new job opportuni-
ties on the growth of and distribution of wages. In P. Cagan (Ed.), Contemporary economic
 Handel / PERCEIVED JOB QUALITY 93

problems: Deficits, taxes, and economic adjustments (pp. 209-242). Washington, DC: Amer-
ican Enterprise Institute.
Kosters, M., & Ross, M. N. (1988). A shrinking middle class? Public Interest, 90, 3-27.
Krugman, P. (1990). The age of diminished expectations. Cambridge, MA: MIT Press.
Lawrence, R. Z. (1983). The myth of deindustrialization. Challenge, 26, 12-21.
Lawrence, R. Z. (1984). Sectoral shifts and the size of the middle class. Brookings Review, 3, 3-11.
Lerman, R. L., & Salzman, H. (1988). Deskilling and declassing: Whither the middle stratum?
Society, 25, 60-66.
Mann, M. (1973). Consciousness and action among the Western working class. London:
Macmillan.
Martin, J. (1981). Relative deprivation: A theory of distributive injustice for an era of shrinking
resources. Research in Organizational Behavior, 3, 53-107.
Meyer, J. W., & Rowan, B. (1977). Institutionalized organizations: Formal structure as myth and
ceremony. American Journal of Sociology, 83, 340-363.
Milkovich, G. T., & Newman, J. M. (1993). Compensation. Homewood, IL: BPI/Irwin.
Mirowski, J. (1987). The psycho-economics of feeling underpaid: Distributive justice and the
earnings of husbands and wives. American Journal of Sociology, 92, 1404-1434.
Mishel, L., Bernstein, J., & Schmitt, J. (2001). The state of working America. Ithaca, NY: Cornell
University Press.
Morris, M., & Western, B. (1999). Inequality in earnings at the close of the twentieth century. An-
nual Review of Sociology, 25, 623-657.
Neumark, D. (Ed.). (1999). Changes in job stability and job security [Special issue]. Journal of
Labor Economics, 17(4).
Osberg, L. (2000). Long-run trends in economic inequality in five countries: A birth cohort view
(Working Paper No. 222). Differdange, Luxembourg: Luxembourg Income Study.
Osterman, P. (1994). How common is workplace transformation and who adopts it? Industrial
and Labor Relations Review, 47, 173-188.
Osterman, P. (Ed.). (1996). Broken ladders: Managerial careers in the new economy. New York:
Oxford University Press.
Osterman, P. (2000). Work reorganization in an era of restructuring: Trends in diffusion and ef-
fects on employee welfare. Industrial and Labor Relations Review, 53, 179-196.
Oswald, A. J. (1997). Happiness and economic performance. Economic Journal, 107, 1815-
1831.
Phelan, J. (1994). The paradox of the contented female worker: An assessment of alternative ex-
planations. Social Psychology Quarterly, 57, 95-107.
Piore, M. J., & Sabel, C. (1984). The second industrial divide. New York: Basic Books.
Quinn, R. P., & Staines, G. L. (1979). The 1977 quality of employment survey. Ann Arbor: Uni-
versity of Michigan, Institute for Social Research, Survey Research Center.
Robinson, J. P., & Bostrom, A. (1994). The overestimated workweek? What time diary measures
suggest. Monthly Labor Review, 117, 11-23.
Rosenthal, N. H. (1989). More than wages at issue in job quality debate. Monthly Labor Review,
112, 4-8.
Schmidt, S. R. (1999). Long-run trends in workers’beliefs about their own job security: Evidence
from the general social survey. Journal of Labor Economics, 17, S127-S141.
Schmidt, S. R., & Thompson, C. (1997). Have workers’ beliefs about job security been keeping
wage inflation low? Evidence from public opinion data (Working Paper No. 97-4). Santa
Monica, CA: Milken Institute.
Schor, J. B. (1991). The overworked American. New York: Basic Books.
94 WORK AND OCCUPATIONS / February 2005

Schuessler, K. A., & Fisher, G. A. (1985). Quality of life research and sociology. Annual Review
of Sociology, 11, 129-149.
Smith, V. (1997). New forms of work organization. Annual Review of Sociology, 23, 315-339.
Staines, G. L., & Quinn, R. P. (1979). American workers evaluate the quality of their jobs.
Monthly Labor Review, 102, 3-12.
Stiglitz, J. (1997). Reflections on the natural rate hypothesis. Journal of Economic Perspectives,
11, 3-10.
Taplin, I. M. (1995). Flexible production, rigid jobs: Lessons from the clothing industry. Work
and Occupations, 22, 412-438.
Teixiera, R., & Rogers, J. (2000). America’s forgotten majority: Why the White working class
still matters. New York: Basic Books.
Uchitelle, L. (1997, February 27). Job insecurity of workers is a big factor in fed policy. The New
York Times, p. D6.
United States Department of Health, Education, and Welfare. (1973). Work in America: Report
of a special task force to the Secretary of Health, Education, and Welfare. Cambridge, MA:
MIT Press.
Vallas, S. P. (1999). Rethinking post-Fordism: The meaning of workplace flexibility. Sociologi-
cal Theory, 17, 68-101.
Vallas, S. P. (2003). Why teamwork fails: Obstacles to workplace change in four manufacturing
plants. American Sociological Review, 68, 223-250.
Zuboff, S. (1988). In the age of the smart machine: The future of work and power. New York: Ba-
sic Books.

Michael J. Handel is an assistant professor of sociology at the University of

Wisconsin–Madison. He researches the growth in inequality in the United
States and its relationship to the changing nature of work and organizations.
He is currently conducting a national panel survey of job skills, technology use,
and employee involvement and their relationships to labor market outcomes.