HD28
DBWEf
.M414
f3
ALFRED
P.
WORKING PAPER
SLOAN SCHOOL OF MANAGEMENT
The Imperfect Legitimation of Inequality
in Internal
Labor Markets
Maureen Scully
Assistant Professor of Management
MTT Sloan
School, E52-568
Cambridge,
MA
02139
617-253-5070
Working Paper 3520-93-BPS
January. 1993
MASSACHUSETTS
INSTITUTE OF TECHNOLOGY
50 MEMORIAL DRIVE
CAMBRIDGE, MASSACHUSETTS 02139
i
The Imperfect Legitimation of Inequality
in Internal
Labor Markets
Maureen Scully
Assistant Professor of Management
MTT Sloan
School. E52-568
Cambridge.
MA
02139
617-253-5070
Working Paper 3520-93-BPS
January. 1993
1
C (r
!
f
>.
M.I.T.
LIBRARIES
The Imperfect Legitimation
of Inequality in Internal
This study addresses whether employees believe
Labor Markets
advancement
that
is
based on merit
in
two
non-unioni/.ed, high technology companies that have formal merit-based performance evaluation and
One
promotion procedures.
reason companies use the language and logic of meritocracy
encourage employees to work harder
in the
expectation of greater rewards.
based procedures -indeed their very intent according to
That
that inequality is legitimated.
that inequality
is
based
fairly
is,
in merit.
to
implication of merit-
Edwards, 1979) -
critical theorists (e.g.,
employees may believe
on differences
An
is
that merit counts for
Employees' shared belief
is
advancement and
in the rationality
of
"impersonal," merit-based governance procedures been invoked by institutional economists to explain
reduced turnover in internal labor markets
(e.g.,
Doeringer
&
Piore,
explain employee cooptation and the general lack of employee dissent
1971) and by sociologists to
(e.g.,
Edwards, 1979).
paper theoretically challenges and empirically investigates this often-invoked notion.
that beliefs about merit vary
by how employees
fare in
1
This
argue, instead,
an organization's advancement contest.
Those
in lower positions, or those with lower mobility rates, or both are less likely to believe that merit
counts in their firm, whether from a savvy born of personal experience or from a need to make
enhancing attributions.
In
some
fashion, they cope with the
judgment implicitly
pertbrmers in a putative meritocracy and deny the claims that merit counts.
argument
that the less successful
doubt the role of merit
is
Upon
hardly surprising
self-
upon lower
cast
brief reflection, an
However,
this
simple
alternative
view has been theoretically overshadowed by overdetermined accounts of socialization and
cooptation.
This view has not been empirically investigated, specifically in a workplace setting where
questions can be addressed about whether
dimensions of
stratification, that correlates
it
is
position or
upward mobility, quite different
with the extent of belief in merit.
Finally, this
view has not
been incorporated into new kinds of economic and sociological theories about the employment
relationship,,
which might look
different they
if
employees do not believe the legitimating claims
employment
relationship.
took seriously the possibility that a
that are
fair
number of
supposed to be the ideological glue of the
The term "meritCKracy" was
a satirical invention of
negative consequences of a rigidly ment-based society.
soberly, to late capitalist systems of status
from class-based or aristocratic systems,
and reward
in
this study,
human
capital variables that
argue that
it is
has since been applied,
somewhat more
allocation, usually to distinguish
There
is
them favorably
a long sociological tradition,
of examining whether variance in occupations and incomes
merits or to class background.
of the hidden
in his fable
which family origin and unearned advantages determine
occupations and incomes (Bell, 1972, 1976).
backdrop to
It
Young (1958)
Merit variables usually include education,
economists would use in wage equations.
difficult to find "pure" merit variables, since
test scores,
is
which
is
the
attnbutable to
and the kinds of
Jencks and colleagues (1972)
most of the measurable ones are already
influenced by privileged family backgrounds and variables like marginal productivity are too hard to
measure.
In the sociological equations, the merit variables are contrasted to class variables, like
parents' occupations
liberal
and family income, and also
agenda of demonstrating
policies are needed.
Alongside
that class, race,
show.
like hard
&
sex), often with the
conducted surveys to see
to be important in determining occupation
Beliefs are of interest, in addition to, and even irrespective of,
In general, national studies of beliefs (e.g.,
Schlozman
and
and sex continue to count too much and correctives
this descriptive research, sociologists also
what variables individuals believed
United States.
to ascriptive variables (race
Huber
&
and income
in the
what the equations
Form, 973; Kluegel
&
Smith, 1986;
Verba, 1978) found strong endorsement (about two-thirds of respondents) that merits
work and
ability determine individual outcomes. Theoretical attention
was directed
to the
formidable effectiveness of meritocratic ideology as a legitimating ideology.
Researchers studying status and reward allocation have turned their attention to the features of
an organization that may help explain why individuals with similar
traits realize different returns to
For example, the structure of job openings affects whether individuals are promoted
those
traits.
(e.g.,
Mittman, 1986; Stewman
&
Konda, 1984; White, 1970). The study of what individuals believe
to be the causes of inequality should similarly shift
from the national
level to the organizational level.
National level studies of beliefs about inequality, despite giving the overall impression of consensus,
do find variance
in beliefs
about the role of merit, as
Mann
(1970) suggests. Moreover, they find
variance within the upper-class, within the middle class, and within the working class.
variance
may
arise because individuals have dilTercnt local experiences in the
their organizations.
to technician, but
Some
of this
advancement contest
in
For example, some members of the working class are promoted from assembler
some
are not; the former
may
be stronger believers in the role of merit.
Including
information about the individual's experience of being promoted or not, by conducting a study of
beliefs within
an organizational context, might help explain variance
be unaccounted for in national surveys.
This study begins
this
in beliefs that
important
would otherwise
move from
the national to
the organizational level in the area of beliefs about meritocracy.
Certainly there have been numerous studies of individual attitudes that have been conducted
Studies of satisfaction and
within organizations.
satisfaction with promotions
and willingness to work
commitment often include questions about
hard.
However, these studies
treat individuals
views about hard work and rewards as neutral, atomistic calculations, relevant within the firm as a
motivation problem.
They do
not connect with the broader stream of research on meritocracy.
The
broader social and political implications of employees' beliefs about whether merits are rewarded
must be taken into account and give a much greater significance to findings about employees'
These findings reveal not just the likelihood of employees' exerting
about meritocratic claims.
but
beliefs
more fundamentally, they
reveal the extent to
which a firm derives some normative legitimacy
from practices rooted in the widespread cultural appeals to meritocracy
in the society at large.
how
This
As
such, findings that
they are doing in the firm suggest not only that
new procedures may
study examines employees' beliefs about merit from this standpoint.
employees' beliefs vary by
effort,
have to be explored by the firm to bring people normatively on board, as argued from the procedural
justice perspective.
They
also suggest that the sense-making schemes that individuals
employ
in
committing to a firm and coping with inequality either leave the firm vulnerable to legitimacy
challenges or must be understood as involving more complexity and ambivalence than binary
accounts of legitimation
One
/
delegitimation have tended to allow.
possible finding
is
that there will be very
little
variance in employees' beliefs about merit.
People in the higher positions in organizations should believe that merit counts, since they
may
interpret their
own
experience as one of meritocratic ascent and since
bolsters their position; that
it
the people in the highest positions should promulgate self-reflective and even self-serving ideologies
has been argued for
The
some time
sociological literature
of class" (Sennett
&
in social theory (e.g., Mar.x
on the legitimation of
Cobb, 1972)
is
&
Engels, 1978: 64; Weber, 1971:956).
inequality^ suggests that
As Mann
those
whose energies and
The
literature
argues, inequality
abilities
injuries
the tendency of people in lower positions to go along with this
dominant view, accept meritocratic ideology, and blame themselves and
lower position.
one of the "hidden
deserve
on organizational
it,
is
legitimated
failures
when people
their inferior merits for their
believe that "success
have only themselves to blame"
comes
to
(1970:427).
and commitment suggests that employees may be
culture
socialized to believe the frequent articulations of the meritocratic ideal in organizations that use
Organizational culture research documents corporate versions of the rags-
merit-based procedures.
to-riches story (e.g., Martin, Sitkin,
for the choice of
who
is
Feldman
&
Hatch, 1983).
promoted or not emphasize individual
structural constraints identified
by researchers mentioned above,
the rationality of the promotion system, sustain
its
o,
,
Mt^YPr
^
Rowan, 1978)
Alternatively,
work on
in turn.
little
al.,
is
by management
and merits, rather than the
in order to
1982).
Taken
maintain the "myth" of
Theories of institutionalization
who would
take for granted
together, organizational theories can
variance in beliefs about merit should be found.
recent sociological writing on the experience of
attributions point to the likeliness of finding variance.
There
traits
also add to a portrait of employees
practices like merit-based performance evaluations.
easily build a case that
rationales given
motivational potential, and bolster the authority of
those chosen by this system (Salancik, 1977; Pondy et
(f
The
work and
These two
social psychological
literatures are considered
a growing body of research that describes a working class whose
members
are aware
iDella Fave (1980:955) defines legitimation as follows, drawing on definitions employed in previous
inequality (Alves
Rossi, 1978; Jasso
Rossi, 1980; Rainwater, 1974): "Legitimation refers
to a belief on the part of a large majority of the populace that institutionalized inequality in the
distribution of primary resources (Rawls, 1971) - such as power, wealth, and prestige - is essentially
right and reasonable." The claim that inequality is meritocratic, and particularly that hard work and
ability lead to success, is one specific contemporary form of the legitimation of inequality (e.g.,
Althusser, 1969; Giddens, 1973; Huber
Engels, 1846; Miliband,
Form, 1973; Mann, 1971; Marx
1969; Mills, 1969; Schlozman
Verba. 1978).
work on
&
&
&
&
&
of their
own
own
interests
and not easily coopted by a dominant ideology
Some examples
experiences.
on experience-based differences
in
Mann
include
that
is
not corroborated by their
(1970) on dissensus, Larkwood and others (1975)
working class images of society, Willis (1981) on the dissident
values of British working class youths, Scully (1982) on the intact self-esteem of high schoolers
how
sorted into the lowest tracks, Sabel (1982) on
working class
the
struggles over the division of labor in the workplace,
is
well aware of
role in
its
MacLeod (1984) on how Boston
power
area youths do
not necessarily believe claims that hard work leads to success, Scott (1985) on everyday forms of
ideological struggles by the peasantry against the powerful,
upward progressions does not apply
careers as orderly
Thomas (1989) on how
research on
to the lived experience of blue collar workers,
and Gamson (forthcoming) on the complexity and nuance, which should not be surprising,
belief systems of
The
members of
social psychological literature
make self-enhancing
successful people
successful people
(e.g.,
the working class
Seligman
make
et al.,
on issues such
attributions
1972).
al.,
common
as ability or hard
pattern
work) and
whereby individuals
1991).
&
traits),
The "fundamental
own
attribution error" (Jones
(Murphy
&
successes to internal factors (such
their failures to external causes (such as structural constraints or luck)
&
Sims, 1985; Mitchell, Green
&
this long-accepted social
Wood, 1981,
&
Cleveland,
psychological style of sense-making
understanding the legitimation of inequality have not been drawn by social psychologists.
implication
is
that meritocratic ideology
successful inasmuch as
cognitions.
it
&
Cleveland, 1991),
Ross, 1975) can be applied to the performance evaluation process (Murphy
The implication of
while less
appeal to external constraints and biases)
attribute their
Research on self-serving biases in attribution (Gioia
Miller
action.
failure predicts that
appeal to internal
(e.g.,
Nisbett, 1971), though no longer regarded as universal or fundamental
describes a
and affirmative
on attributions about success and
self- protecting attributions (e.g.,
1979; Weiner et
as inequality
in the
depends upon
is
their
likely to fail
for
The
in legitimating inequality to the less
blaming themselves and not employing self-protecting
Hypotheses
Sociologists have pointed to multiple and competing systems of stratification within and
outside organizations (Granovetter
&
Organizations have a distribution of positions, of
Tilly, 1986)
performance evaluations, of degree of return on human
considers whether
how an
relates to their beliefs
Beliefs about
much
individual
about
is
how much
how much
doing in any or
all
rates.
This study
of these various local mobility contests
merit counts.
how
merit counts are operational i zed by by having employees rate
(on a 7-point scale) they believe each of five items counts for advancement: hard work, ability,
performance, privilege, and luck.
These items
what individuals believe does and ought to
Coleman
&
and of mobility
capital,
&
Rainwater, 1978; Huber
&
An employee who
Verba, 1978).
are considered in
&
believes that the firm
work, and performance as very important determinants of
coming from a privileged background and luck
relationships
is
the
same
work may be
United States
of
(e.g.,
Smith, 1986; Mann, 1970; Schlozman
is
would
meritocratic
who
rate ability, hard
gets ahead in the firm,
Though
as not important.
The items
are different
a variable input, and performance
form a conceptually nor empirically sound
makes sense not
in the
level studies
and
rate
the expected pattern of
for the five variables in the hypotheses below, they are treated as five
separate dependent variables in this study.
fixed input, hard
and income
affect occupation
Form, 1973; Kluegel
numerous national
scale.
is
enough
(e.g., ability
may
be a
an output) that they might not
In addition, as a first study of these variables,
more "open" look
to aggregate variables, but to take a
at the
it
data and uncover
unexpected differences (Bailyn, 1977).
Hypotheses
human
capital,
1
to
4
relate success in
and mobility to
terms of position, performance evaluation, return on
beliefs about merit.
Hypotheses 5 to 8 address other aspects of
employee's experience in the advancement contest: their recent
lateral mobility,
whether they have
crossed a "class boundary" from hourly to salaried, whether they are a manager, and their tenure.
Hypotheses 9 and 10 consider employees perceptions of
a possible sex difference.
their
advancement, and Hypothesis
1 1
posits
Success in the firm's advancement contests
1.
Position in
tlie
organizational hierarchy. Position
of success in the firm, even
if
may
represent the ultimate attamment
The
best
to the highest positions
A
individuals socially construct alternative and local indicators.
rewards and the largest allocation of scarce societal resources attach
person exports the rewards of organizational position - from a paycheck to social esteem - into the
Locally constructed indicators of success
larger society.
(e.g., best
assembler) do not export as well.-
Job ladders and bureaucratic hierarchies have long implied a ranking of employees by merit, such
that the
most meritorious employees are
connotation,
it
is
at the top
self-enhancing for people in
role of merit
luck).
and
Similarly,
cite the role
it
is
In a world
where rank has
this
higher positions to believe in meritocracy and to
attribute their position to their merit (and also, to
background or
(Weber, 1946a).
deny the
role of non-merit factors,
such as class
self-enhancing for people in lower positions to downplay the
of non-merit factors.
Hypothesis 1. The higher the employee's position in the firm, the more strongly he or she believes:
o that hard work counts for getting ahead,
o that abihty counts for getting ahead,
o that performance counts for getting ahead.
o that coming from a privileged background does not count for getting ahead,
o that luck does not count for getting ahead.
If beliefs
about meritocracy do
the organization and a
map
demand from people
onto position, one might expect a crisis of legitimacy in
in
lower positions that people in higher positions turn
over some of their "unearned" rewards. Variance in beliefs by position looks like a class-stratified
belief system.
ladders,
may
mobility, and
However,
internal labor markets, particularly narrowly defined jobs arranged in
constrain people's social comparisons, focus people's aspirations on local upward
mask
vast differences in position and attendant rewards, thus preventing (intentionally
or incidentally) a class -stratified belief system and the discontent
it
might engender.
^Discussions of inequality take many forms, but at their basis, the interesting issue is the gap between
those in the highest and lowest positions. It is good not to lose sight of the fact that the highest
position is the highest reward, even in the process of discussing how participants in the social system
may be satisfied by other, compensatory successes, like climbing lower rungs of the ladder or
receivina good performance evaluation within their job grade.
8
2.
Upward
Absolute position may not matter as
mobility.
much
to individuals as their
success or failure in improving their position in the hierarchy via upward mobility.
organizations, their mobility rate
their absolute position.
and
may
The words of one employee
are but
where you're going
The data
success in the competition for increased rewards
an internal labor market where insiders compete to
levels.
Methods
First, there is
and controls for differences
higher rate than a slow, steady incumbent).
employee has advanced.
Rates of
company and may depend
in part
a
comparison referents
Hypothesis 2.
o that
o that
o that
o that
o that
3.
in their
to be
more
is
higher
ways
rate
which
in
(number of
fast-moving newcomer has a
it
meanings
may
of the
in different areas
be worth looking at the absolute
be more attuned to their mobility rate
immediate occupational
mean and
three
simply the absolute number of levels the
so
may
show
which captures the speed with which
different
arise,
Third, people
(to correct for the
normed measure proves
Second, there
movement may have
rate.
section,
in tenure (so that a
on when openings
ascent without the correction for
can be standardized
at
not where you
"It's
an employee's mobility
levels of the hierarchy transcended divided by years of tenure),
rate
openings
that counts."
upward mobility might be operationalized.
relative to social
fill
interviewed for this study typify this view:
I
for this study, discussed below in the
the person advances
in
be the more salient reward in the status attainment contest than
Movement upward connotes
status, particularly in
For individuals
area.
Each person's mobility
standard deviation of their group) to see
if
such
useful for understanding variance in beliefs about merit.
The greater an employee's upward mobility, the more strongly he or she believes:
hard work counts for getting ahead,
ability counts for getting ahead,
performance counts for getting ahead.
coming from a privileged background does not count for getting ahead,
luck does not count for getting ahead.
Relative return on education and tenure. This variable
failure in terms of specific social
education, their starting
human
comparison
referents.
capital, to the firm.^
is
another indicator of success or
Individuals bring different amounts of
Individuals
may
not expect to do as well as those
^Other studies have addressed whether education itself is distributed meritocratically or not. The firm
inherits already educated individuals, and personnel managers sometimes take pains to argue that
they and the firm are not in a position to correct for past inequalities of opportunity.
who
but
more
bring
do expect
to
capital, in the
do
as well as similar others,
become Vice
not expect to
form of education or years of experience,
President (despite popular Horatio Alger stories about such
he or she should expect to become Lead Technician
have done
so.
on education and tenure
believe less strongly that
Hypothesis
advancement
contest,
may
lor example, a person with a high school diploma
rises), but
a return
to the
if
meteoric
others with a high school degree
that is relatively too
low should make an individual
ment guides advancement.
The greater an employee's return on education and tenure, the more strongly he or she
3.
believes:
o
o
a
o
o
hard work counts for getting ahead.
that
that ability counts for getting ahead.
performance counts for getting ahead.
coming from a privileged background does not count for getting ahead.
luck does not count for getting ahead.
that
that
tliat
how
This measure comes closest to capturing individual's "actual" merits, but also reveals just
difficult
it
for lack of
4.
is
to
answer the question of whether the companies
good measures of
in this study "really are" meritocracies,
merit.
Performance evaluation. The performance evaluation process
individuals in the
same job area
like privilege
in a
may
and
way they may
latter
relate to
luck.
not
two variables
whether they attribute success
Employees know
know
what most directly
pits
against one another and assigns relative winners and losers.
Employees' sense of how well they are doing might be tightly tied
evaluation and
is
their
in the
to their recent
performance
firm to merit or to non-merit factors
performance evaluation (often a number from
their exact mobility rate or their relative return
are constructed by researchers), so
it
may prove
on human
to be the best
1
to 5)
capital (these
measure
for
understanding beliefs about merit.
Hypothesis 4.
she believes:
a that
o that
o that
o that
o that
The higher an employee's
last
performance evaluation rating, the more strongly he or
hard work counts for getting ahead.
ability counts for getting ahead.
performance counts for getting ahead.
coming from a privileged background does not count for getting ahead.
luck does not count for getting ahead.
10
other advancement experiences
in the
firm
Lateral moves (recent). Lateral moves can create a sense of movement, perhaps whether
5.
or not the individual
is
making
real
headway up the
vertical ladder of the organization.
(1987) found that job ladders represent idealized routes of movement, but in practice,
between ladders are as frequent and can improve career prospects.
the mobility variables discussed so
scale.
Recent
lateral
moves
far,
moves
Lateral
still
lateral
moves
are not captured in
which chart only movement up hierarchical
(within the past three years) are the ones that
DiPrete
levels of the
pay
hold the promise of
converting into upward mobility opportunities (whereas this promise for lateral moves of several
may have
years ago
expired).
5. The greater the number of lateral moves an employee has made(in
more strongly he or she should believe:
o that hard work counts for getting ahead.
Hypothesis
the past three
years), the
o tfiat ability counts for getting ahead.
o that performance counts for getting ahead.
o that coming from a privileged background does not count for getting ahead.
o that luck does not count for getting ahead.
Crossing a class boundary.
6.
Internal labor markets often have multiple ports of entry.
Hourly workers and salaried workers enter
bottom of the engineering
start at the
for hourly production
to salaried worker.
at different starting
ladder,
and technical workers.
which
It is
is
pay grades. For example, engineers
already higher than the top rung of the ladder
difficult for hourly
Essentially, they face a ceiling
on
their mobility prospects. DiPrete
(1988) found that crossing a boundary from the lower to upper
moment
significant
who have
in a career history
tiers
For those
crossed this particularly salient boundary, this single event, irrespective of other indicators
may
condition strongly their sense of success.
Hypothesis
6. If
an individual has crossed the boundary from hourly
more strongly:
o that hard work counts for
o
o
o
o
and Soule
of the civil service was a
and the greatest source of disadvantage for women.
of mobility,
believe
workers to cross the boundary
to salaried,
he or she should
getting ahead.
that ability counts for getting ahead.
that performance counts for getting ahead.
that coming from a privileged background does not count for getting ahead.
that luck does not count for getting ahead.
11
definitive aspect of being a
make
manager
a person defend the practice,
Following
Pfeffer's (1981)
organizational symbols,
more
Whether the employee
Being a manager.
7.
likely
a
manager might inlTuence
having to conduct performance evaluations, which might
mostly for the authority
it
confers (Dornbusch
argument, people in managerial positions
some of which symbolize
&
Scott, 19'
manage
in the organization
that the firm is meritocratic;
the
managers become
abilitv counts for getting ahead.
performance counts for getting ahead.
coming from a privileged background does not count for getting ahead,
luck does not count for getting ahead.
Predictions about the role of tenure in understanding beliefs
in the firm.
tinged by researchers' prior on whether firms tend to be meritocratic.
If
basically meritocratic, despite occasional, local deviations, then one
would predict
to see this pattern
tenure and belief that the firm
emerge
is
after longer tenure
meritocratic.
meritocratic, predictions about tenure
Conversely,
would stem from
one believes
if
one believes
a chronicle of
how
becomes disillusioned
of such deviations.
relationship between tenure and belief that the firm
is
that the firm is
that
that the firm
when seeing deviations from merit
at the repetition
may be
employees
and predict a positive relationship between
gives the firm the benefit of the doubt
this
A
Managers should believe more strongly:
hard work counts for getting ahead,
Tenure
would begin
beliefs
themselves to be persuaded by these symbols.
Hypothesis 7.
o that
o that
a that
o that
o that
8.
if
is
is
The
latter
second view, since reviews of performance evaluation practices suggest
Of
basically not
employee
criteria, but
initially
eventually
account suggests a negative
Hypothesis 8
meritocratic emerges.
firms to find and consistently use unbiased measures of merit.
the
is
it
is
is
based on
extremely difficult for
course, a third possibility
is
that
individual employees have different experiences of gradual confirmation or disconfirmation that
merit applies as they remain with a firm, and thus no significant effect in one direction or the other
emerges
for tenure.
Hypothesis 8. The longer an employee's tenure in the firm, the weaker his or her belief:
o that hard work counts for getting ahead.
a that ability counts for getting ahead,
o that performance counts for getting ahead.
o that coming from a privileged background does not count for getting ahead,
o that luck does not count for getting ahead.
12
Perceptions of advancement
9.
Perceived relative mobility.
mentioned, they
may
not
know
Individuals' actual mobility
their mobility rate or believe
it
is
measured above. However, as
something
to be
Their
different.
perceived advancement may, therefore, relate more strongly to their beliefs about merit.
Hypothesis
9.
The better an employee perceived his or her mobility
to be, the
more strongly he or she
believes:
o that hard work counts for getting ahead.
o that ability counts for getting ahead.
o that performance counts for getting ahead.
o that coming from a privileged background does not count for getting ahead.
o that luck does not count for getting ahead.
10.
Disappointment aiM>ut performance evaluation. Some employees who receive
performance evaluation
a meritocracy,
even
may
if it
accept that their
is
one
in
it
reflects their lesser merits
which they are doing
less well.
and believe
a
low
that the firm
is
Other employees may be
disappointed that their performance evaluation should have been higher.
This discrepancy
may
relate negatively to the belief that merit counts.
Hypothesis 10. The greater the discrepancy between the performance evaluation an employee felt he
or she deserved and the actual performance evaluation received {i.e., the greater the
disappointment), the weaker his or her belief:
o that hard work counts for getting ahead.
o that ability counts for getting ahead.
o that performance counts for getting ahead.
o that coming from a privileged background does not count for getting ahead.
o that luck does not count for getting ahead.
Demographics and controls
11. Sex.
The tendency
the socialization of girls and
to
make
women
self-enhancing attributions
suggests that
evaluations of them and to accept blame
Nelson,
&
Enna, 1978;
Dweck
&
when they
Goetz, 1978).
women
are
may
more
differ
by
women,
Research on
likely to internalize others'
receive poor ratings (e.g.,
Therefore,
sex.
Dweck, Davidson,
irrespective of position,
may
more likely to believe in meritocratic ideology, an exploratory prediction that this study examines.
Hypothesis 11. Female employees will believe more strongly:
o that hard work counts for getting ahead.
o that ability counts for getting ahead.
o that performance counts for getting ahead.
o that coming from a privileged background does not count for getting ahead.
o that luck does not count for getting ahead.
be
13
Occupational and firm controls.
individual le%
el
Table
1
This study examines mechanisms that work
and makes no a priori predictions about firm or occupation
(all
at
the
effects.
working paper) summarizes the hypotheses.
tables appear at the end of this
Method
Survey design
I
developed and administered a questionnaire to measure the preceding variables. The survey
includes attitudinal variables measured using 7-point Likert-type scales.'^ For the five dependent
variables, respondents
answered questions about how much the thought each item (hard work,
performance, privilege, and luck) counted for advancement
A
not count, 7=counts very much.
pretest, involving eight
four from one of the companies, indicated that wording
some of
the variance, since "does not count at all"
advancement was also measured using a 7-point
very much).
1
in the firm.
ability,
(The scale used was: l=does
respondents not from the companies and
as "does not count very
was perceived
much" truncated
to be the other endpoint.)
scale (l=have not
advanced
at all,
Perceived
7=have advanced
For the other "perception" variable, employees were asked what performance evaluation
they thought they deserved.
The other
employment
at the
variables were calculated from information each
history, including their
company,
the current year.
from
to 4).
most recent performance evaluation (l=low, 5=high),
starting position in the
mobility can be calculated.
They
The number of
From
their
start date
company, and current position, from which tenure and
also supplied the job
recent lateral
their starting
employee provided about
titles
they had for the three years preceding
moves was constructed from
and current positions,
I
this
information
(it
ranges
could calculated whether they had "crossed
'^Caution is certainly warranted in using ordinal variables as dependent variables, even though this
usage has a long history in the social sciences. I was concerned that coefficients might understate the
relationship, particularly of issue in interpreting whether position is not significant. Subsequent to the
analyses reported herein, I recoded the dependent variables (0= 1,23,4 and 1 = 5,6,7) and performed
logistic regressions. The pattern of results was the same, and position was not significant (nor was it
significant when position itself was recoded into fewer categories).
14
a class boundary" from hourly (non-exempt) to salaned (exempt);
passage was particularly prized
at these
Though employees supplied
etc.) are
my
interviews revealed that this
companies
the input data, the actual measures (of mobility, lateral
moves,
calculated by the researcher, and as such, these can be considered "quasi-objective" measures.
The inclusion of such measures
coefficients likely
mitigates concern about
when independent and dependent
variance
(i.e.,
inflated
variables are both attitudinal measures obtained
Common
from the same instrument) (Spector, 1987)
common method
method variance must be kept
in
mind,
however, for the relationship between perceived advancement and beliefs about merit (even though
these questions were asked several pages apart on the survey).
Selection of two companies
The study of employees'
beliefs about merit
is
most relevant
for
companies
that
have two
features of internal labor markets: merit-based promotion systems that are part of the governance
structure
and promotion from within. Promotion contests
most organizations" (Baker, Jensen
markets can use different
will
move up
criteria
&
"are used as the primary incentive device in
Murphy, 1988:600)
- such as
the organizational ladder^
for a
number of
merit, need, or seniority
-
reasons.
Internal labor
for the determination of
who
Lawler (1971:158) documents "many companies' very
frequent claims that their pay systems are based on merit," despite, he continues, evidence from
several studies of a
low correlation between pay and performance.
Cleveland (1991) suggest that
evaluation.
90%
of private organizations use
&
recently.
Murphy and
some form of formal performance
Promotions are ideally supposed to allow firms to match individuals with jobs for which
they are well-suited, although this matching
Jensen
More
Murphy, 1988; Sorenson
The two firms
that
&
may
involve occasional or even systematic errors (Baker,
Kalleberg, 1981).
were selected
for this study (of twelve firms
approached) matched on a
in the periphery of the economy may make no pretense of
offering merit-based or other criteria; they simply offer work and pay, particularly to unskilled
workers who may have no other choices.
^Of course, sweatshops and establishments
15
number of important
charactenstics, including: both are in the
500 employees and
at least
(which makes
it
more
same industry (high technology), have
are at least 15 years old (so the job paths are stable), are not unionized
likely that merit rather than seniority
the formally
is
promotion (Freeman (1982)), conduct regular performance evaluations
for
espoused basis
for
in part to identify candidates
promotion from within, and have multiple levels of blue- and white-collar job grades.
As
I
learned more about the two firms,
of potential interest for this study.
Because of
ratings.
this practice,
I
found that they did not match on one characteristic
I
Company A
uses a forced normal distribution of performance
expect a steeper relationship
evaluation and belief that merit counts
Where
at
Company A between performance
a forced normal distribution
value high ratings more strongly (because they have more value where there
is
is
used, people might
no rating inflation and
high ratings are scarce), but not take seriously lower ratings (because they are artifactual of the
constraint
on managers
more reason
attribute
to
to attribute
them
fill
the bottom categories). Thus, those
them
to lack of merit.
equations (Firm
A
to merit, and those
dynamic
If this
who
is at
who
get the best ratings have
the
more reason
get the lowest ratings have
work, the interaction term that
all
I
will
add
to
to the
x performance evaluation received) should be positively related to beliefs about
merit.
Survey administration and response issues
The survey was
(called
distributed to
Company A and Company
mailed survey, but
still
low enough
845 employees
The response
B).
in
two firms
rate to the
that
have internal labor markets
survey was 51.8%, not unusual for a
that sources of response bias warranted examination.
a logistic regression to predict non-response, following the
method
in
1
performed
Berk (1983), and found
that
none of the available variables significantly differentiated between respondents and non-respondents.
1
was limited
to variables for
which
I
had information on non-respondents: detailed work group,
location code, and sex (the companies were not able to provide
employment
history,
and of course, data on
beliefs
is
me
with additional information about
always missing for non-respondents).
had information on the distribution of performance evaluations
for the entire population.
1
A
also
Chi-
16
squared
showed
test
purpose of
that
this study,
it
my
is
sample was not significantly different from the population.
particularly
good
that neither
For the
winners nor losers in the performance
evaluation contest are over- or under-represented.
Results
This section:
1)
presents the descriptive findings about beliefs and compares respondents'
beliefs about merit in the firm versus in the U.S.,
the preliminary concern that
2) addresses
respondents think merit ought to count, 3) creates the mobility rates normed for occupational group,
4) computes the relative return on education and tenure by regressing them on position, 5) presents
the results of the examination of Hypotheses
which proves to be an important variable
1
to 11, 6)
examines whether perceived advancement,
for understanding beliefs in merit,
is
correlated with actual
advancement, and 7) considers whether ambivalence about merit better characterizes some
respondents.
Pattern of
t)eliefs
about merit
Tables 2 and 3 show the means, standard deviations, and correlations for variables
study.
Responses to the questions about merit show there
variables, although the
mean
and luck (on the same
meritocratic.
variation in beliefs about the five merit
level of belief in the merit factors (about
not count" to "counts very much") compared to the
privilege
is
scale), suggests
mean
in this
5 on a 7-point scale from "does
level of belief in the "non-merit" items,
an overall tendency toward belief that the firm
is
Table 4 shows the frequency distribution of responses to the merit questions.
This study was motivated by national level studies of beliefs about inequality.
studies, the links
between
beliefs about national
This study has some preliminary data on
this issue.
In future
and organizational mobility can be tied together.
Individuals
meritocratic and that their workplace, about which they have
may
more
believe that the U.S.
sjjecific
information,
is
generally
is less so.
I
expected this pattern, particularly since questions about opportunity worded more specifically
17
generate lower levels of belief (eg
perceive there
is
.
Schlozman
&
Verba, 1978)
inequality in the U.S. ("out there"), but
my own
Alternatively,
situation in
which would be consistent with Lemer's (1980) theory of people's views of a
latter pattern is
It
firm
is
what
1
"just world."
is fair,
In fact, this
find for this sample.
meritocratic than that the United States
in the
be that people
my own company
appears, looking at the means in Table 5, that employees believe
and privilege
may
it
is
meritocratic
more strongly
Statements about
ability,
that the
hard work,
United States generate a lower mean and about the same variance as statements
about ability, hard work, and pnvilege in the more specific context of the workplace.
In addition, the beliefs about the national and the organizational opportunity structure appear
to be only moderately correlated.
Particularly since both questions
one might e.xpect a higher correlation simply as an
artifact
were asked in the same survey,
of the method.
It
appears instead that
employees' beliefs about merit in one context do not strongly relate to or inform their beliefs about
merit in the other context.
Whether respondents think merit ought
Before proceeding,
it
is
to count
necessary to check that the merit items
(hard
work,
ability,
performance) are what individuals generally think ought to count for advancement and that the nonmerit items (privilege, luck) ought not to count. This difference
results, for
example,
count in the
What
I
first
if
is
important for interpreting the
employees do not think hard work counts, but they do not think
place, then the interpretation should not suggest
some kind of
it
ought to
crisis of legitimacy.
have called "non-merit" items should be undesirable deviations from meritocracy, not
normatively desirable bases for deciding advancement.
item ought to count, on a 7-point scale.
distinction.
I
asked respondents to
rate
how much each
Their responses confirm the posited merit
/
non-merit
Merit items are rated high in the normative questions, and non-merit items are related
low. Table 6 shows the
means and standard
deviations.
18
Construction of the measure of upward mobility relative to job category
This variable
a
is
used as one of the three measures of mobility for Hypothesis
measure of each individual's mobility
7.
Because
this variable
relative to others in the
same job category,
specifications of mobility should be used
However, future work using
this data
if
not
is
it
the
made
a
human
used in testing Hypothesis
regressed position on education and tenure.
on education and tenure
A
shown
8.
in
Table
Table
The simpler
value.
might give more emphasis to organizational social groupings
Construction of the measure of relative return on
Harder (1992)).
in
more nuanced ones do not add explanatory
by Baron and Pfeffer (1989)
return
constructed
shown
focus of this study
that give rise to local social comparisons, as urged
is
as
I
proves to be only marginally significant in the equations in this study, and
other measures of mobility are stronger,
This variable
2.
3.
To
capital
calculate relative return
on human
capital
I
saved the residual to measure each individual's relative
I
(similar to the procedure used in Pfeffer
higher residual represents relative "over-attainment."
Separate regression equations for
and Lawler (1980) and
Results of this procedure are
Company A and Company B
are
shown (and
the
second of the two equations for each company just shows other exploratory measures of human
capital that
were not retained
for conceptual or empirical reasons).
Theoretically,
it
makes sense
to
run separate equations for the two companies, because individuals have within-firm information about
how
their position
compares with
relatively high or low.
for
Company A
is
that of others of similar education
Therefore, their return
much
higher than the
education and tenure explain
much more
is
R^
calculated relative to others in the
for
Company B
is
the log of
of the variance in position in
Because of the better
to
fit
not great in this study (studies where highly skewed income
income to improve
linear
fit
for
same
(.669 versus .175).
education and tenure on the log of position does not improve the
position
and tenure and whether
for
is
firm.
it
is
The R^
Differences in
Company
A.
Company
B; the skew in
Regressing
the dependent variable use
fit).
Company
A, relative return on education and tenure
have a stronger relationship to perceived advancement for employees in
Company
A.
may
That
prove
is,
in
19
Company
A, an employee whose return on education and tenure
too low relative to his or her
is
colleagues might be more aware of this under-altainment in a context where education and tenure
align
more closely with
contrast, in
in the
Company
company, so
position,
differently in the
it
may diminish
his or her perceived
advancement
In
B, other unmeasured variables appear to contribute to the variance in ptisition
relative return
To examine
advancement
and and
on education and tenure may be of
I
include an interaction term - Firm
the equations that estimate perceived advancement.
This term
importance for perceived
on education and tenure operates
the possibility that relative return
two companies,
less
A
x the attainment residual - in
not significant and
is
is
not
shown
in
the final results in Table 9.
Examination of Hypotheses
to 11
1
Results for Hypotheses
performance
1
to 7 are
shown
This study finds not only that perceived
in the firm's mobility contest is significant for
also that absolute position in the hierarchy
equation for
in Table 9.
how much
privilege counts,
suggests that the organization
is
is
generally not significant
which
significant relationships
Position in
tlie
is
(it is
only significant
in the
discussed further below). Again, this pattern
not vertically divided between believers in meritocracy nearer the top
and disbelievers nearer the bottom. Rather, there
In general, the equations
understanding beliefs about merit, but
are relative believers
do not explain much of
must be viewed against
is
in
every
level.
the variance (less than ten percent) in beliefs, so
this finding.
organizational hierarcliy. Hypothesis
positively to believing that advancement
and disbelievers
meritocratic.
to variance in beliefs about the three merit items
The
1
results
- hard work,
argues that position should relate
show
ability,
that position
does not
relate
and performance - nor about
the role of luck.
However, position does matter
is
for beliefs about
how much
privilege counts.
The
relationship
in the predicted negative direction. This finding says that people in higher positions believe less
strongly that privilege contributes to success, while people in lower positions believe
that privilege counts.
The
belief that privilege
does count
is
a potentially
more strongly
more radicalizing
belief
20
than the belief that hard work, abihty, or privilege does not count, so
found but not the
Because privilege
latter.
is
interesting that the former
counts, either people might be neutral about
it
it
because
it
is
is
ment
the counter-normative alternative to the idea that
is
language, or they have strong opinions about
it
unfamiliar and not part of corporate
(strongly positive or negative, depending
upon
their
position).
Performance evaluation.
Performance evaluation relates only to luck The better the
performance evaluation someone receives, the
do worse
in the contest to get the limited
the attribution that luck determines
less strongly they believe that luck counts.
Those who
good performance evaluations may find some comfort
who does
In interviews,
best.
in
some of those who received
excellent performance evaluations graciously acknowledged that there was certainly an element of
luck, particularly since performance (on the shop floor or at a desk)
anonymity of a survey, they appear
is difficult
to measure.
In the
to have been less likely to attribute importance to luck.
Perceived mobility. As predicted
in
Hypothesis
9,
perceived mobility has a significant
positive relationship to hard work, ability, and performance and a significant negative relationship to
luck and privilege.
five equations.
One
It is
the only variable that
shows
significant effects in the predicted direction in all
of the main findings of this study
is
that perceived mobility is significant for
understanding beliefs about merit.
Performance evaluation discrepancy. Higher discrepancies (deserved minus
actual
performance evaluation) indicate greater disappointment with the performance evaluation received.
This variable
is
significant in the predicted direction in
two equations (hard work and luck) and
marginally significant in two equations (performance and privilege).
Employees who
are
more
disappointed feel less strongly that hard work and performance count, and feel more strongly that
privilege
and luck count.
Upward
mobility, relative return on tenure
boundary, being a manager.
and education,
lateral
Contrary to the predictions in Hypotheses
these variables have a significant relationship to beliefs about merit.
moves, crossing a class
2, 3, 5, 6,
and
7,
none of
All these variables have the
advantage, discussed above, of being "quasi-objective" measures of individuals'
employment
21
experience and history.
While
this is a mcthcxlological
advantage,
it
may
also
mean
inasmuch
that,
as
these variables are constructed by the researcher and not reported by employees, these variables
do
An employee may
not
not reflect
some understanding
the respondent has of his or her experience.
need to "make sense of" these experiences by believing to a greater or lesser extent
merit
if
these experiences are not salient in the
An
alternative interpretation
perhaps with some noise.
may
is
first place.
that these experiences
This possibility
is
in the role ol
do contribute
advancement,
to perceived
The perceived advancement
explored below.
already capture the role of these variables.
Tenure.
Tenure has a significant negative relationship both to the belief
employees may
work
that hard
counts and to the belief that ability counts, in the direction predicted in Hypothesis
just
variable
Newer
8.
with a belief that hard work and ability are rewarded, perhaps because they have
start
come from school
or start with
new
priors that their
new company
is
a meritocratic place.
longer tenure, employees believe less strongly that hard work and ability count.
Their beliefs
decrease with tenure, perhaps because as options close and careers settle to a certain pace, they
feel that their
own
hard work and ability
is
not,
is
Sex. Sex
not simply true for people
is
This result
performance
who
when mobility
is
controlled
for.
The
are frustrated about not being mobile.
not related to beliefs about merit, contrary to Hypothesis 11.
Occupation.
counts.
may
on the margin, delivering better advancement. Tenure
per se must be driving negative relationship, which remains even
effect of tenure
With
is
Employees who
are in Production are less likely to believe that
surprising, since the stereotypical view of Production
measures
on
which
to
base
evaluations
and
is
that
performance
has more usable
it
advancement
do
than
the
business/administrative or engineering occupations, which are thought to include more projects that
are
ambiguous, long-term, or accessible only by similarly skilled members of what Williamson and
Ouchi (1981)
call a "clan."
In fact, what
precisely because performance
arbiter of
who
is
may
be going on
is
that Production
employees can
measurable and not socially constructed, that performance
is
see,
not the
gets ahead.
Firm. Only one main
effect of firm is found.
Employees
in
Company A
are less likely to
22
believe that hard
work
counts.
The
may
structure of the performance evaluation system
contribute to
this belief.
Interaction of
is
steeper for
Firm A and performance evaluation. The
Company A
(the interaction term for
Perhaps
many employees
evaluations limits the
number who can
evaluations to some.
Those
significant).
strongly that hard
in
determining
who
gets
A
and performance evaluation
is
and
positive
they work hard, but the forced normal distribution of
feel
get the highest evaluation and forces managers to give low
Company A who do
work counts than those who
get stuck with the poor ratings in
Firm
slope for performance evaluation
Company A
Company
get the top evaluations in
B.
are even less likely to believe that hard
(Dummy
which evaluations.
performance evaluation can be included
well in this competition believe even
in future
work
variables
for
to estimate
And
those
more
who
work counts
for
four of the five levels of
more precisely the return
to
performance evaluation.)
Overall, perceived advancement
explained in these five equations.
advancement
relates to actual
important for the small amount of variance in beliefs
is
The next
section considers the extent to
advancement, particularly in
which perceived
light of the surprising finding that
none of
the "quasi-objective" measures of employees' experiences in the advancement contests in the firm
is
significant.
The
relationship of actual advancement experiences to perceived advancement
This study proposes that individuals might cope with meritocracy by embracing the belief in
Another way in which they might cope
merit to differing degrees.
While
true success in
advancement contests
is
limited, individuals
is
to believe they are successful.
may
perceive themselves to be
successful, whether because of inflated impressions or because other kinds of success are possible in a
variety of local contests.
person
who
may
assist the individual.
For example, a
has not experienced any upward mobility in an organization might nonetheless feel
successful if he or she has
done.
Local constructions of success
made
Indeed, the current
frequent lateral
move toward
moves
flatter
that
add interesting change to the work to be
organizations and more job rotation requires
23
corporations to encourage employees to regard such moves as
real,
not illusory or consolatory,
representations of success, while a critical sociological approach might regard these "satisfying" lateral
moves
as
mere
(eg, Baron
illusions of mobility
&
Bielby, 1986); these normative views are not
adjudicated herein.
It
worse to the extent
better or
answer
(e.g..
seems very straightforward
is
not as trivial as
Kinder
&
it
that their
may
to predict that individuals perceive their
advancement has actually been
at first appear.
Sears, 1985), as
is
The
the relationship
links
advancement
to be
relatively better or worse. This
between experiences and beliefs
is
complex
between behaviors and attitudes (eg., Ajzen
&
Fishbein, 1977). Psychologists have identified factors that might attenuate the relationship between
experiences and perceptions. Motivations
and nurture "positive
may
might naturally want to deny
intervene: people
illusions" (Taylor, 1988) about their success.
These mechanisms notwithstanding,
I
expect that, at the workplace, people's perceptions of
their mobility will be fairly well in line with their actual mobility
may
be hard to sustain.
The evidence of how well one
is
doing
At the workplace, positive illusions
is
constantly present, whether in the
form of a weekly paycheck, of promotion announcements of peers, or of having
a superior vested with the greater authority of a higher position.
to take orders
one to adduce to predict simply
the appropriate
is
that actual relative mobility experience relates to perceived relative
mobility experience, this view of the particularities of the workplace suggests a significant
Perceived advancement
relative return
is
on human
capital,
performance evaluation,
The
results are
moves, crossing a
shown
in Table
Advancement experiences other than upward mobility, experiences
be considered ancillary to upward mobility, are included alone
the second equation, the addition of
for
lateral
and
1 1
Overall, actual experiences explain about twenty percent of the variance in
perceived advancement.
method
link.
regressed on the variables from the preceding analysis - position,
boundary, being a manager, tenure, sex, occupation, and firm.
discussed below.
from
These reminders may make denial
or positive illusions a tenuous coping strategy. While no particular body of theory
upward mobility,
failure
upward mobility
comparing nested models
in
in the first
rate significantly
that
may
of the three equations.
improves
R^
In
(following the
Wonnacott and Wonnacott (1977:434-436).
The other
24
factors alone
do not explain perceived advancement
as well as the equation including the
most
direct
measure of actual advancement, although they can explain about sixteen percent of the variance
in
In the third equation, demographics and controls are included, but none
perceived advancement.
contribute significantly.
Position in
tlie
organizational hierarchy. The findings show that position does not relate to
As above,
perceived advancement.
the lack of a significant role for position suggests that perceptions
of success are not vertically stratified and clustered
at the top.
It
is
possible that the nature of the
dependent variable prompts respondents to think in terms of how they have advanced, rather than
where they have
arrived.
A
broader question about satisfaction with one's status in the firm might
have produced a significant relationship with position. In terms of the mobility contest specifically, a
higher position does not apjsear to relate to a greater sense of success in the mobility contest.
Performance evaluation. Performance evaluation
has a significant positive relationship to
Those who
get higher evaluations perceive that their
perceived advancement, in
all
three equations.
Many
advancement overall has been good.
performance evaluations, and in
my
interviews,
I
studies
found
find
that
that
employees
process
(e.g.,
Ilgen
&
to this
&
Feldman, 1983; Murphy
and perceived procedural fairness of what
is
cynical
about
employees who had gotten high and low
ratings insisted they did not take performance evaluations very seriously.
improving performance evaluations responds
are
Much
of the literature on
cynicism and focuses on how to improve the
Cleveland, 1991) in order to increase the credibility
taken to be an unfortunate aspect of the organizational
governance structure. The strong positive relationship between performance evaluation and perceived
advancement
in this study suggests that people
entirely cynical about the signal
so entirely mollified by
fair
it
who
sends and people
receive high performance evaluations
who
may
not be
receive low performance evaluations are not
procedures that they perceive themselves to be doing just as well.
Relative return on education and tenure.
Relative return on education and tenure (also
labeled "attainment residual") had only a marginally significant positive relationship to perceived
advancement and was not significant with
all
the controls added.
As discussed above,
the advantage of being a quasi-objective measure constructed by the researcher.
At
this variable has
the
same
time.
25
because
survey,
it
constructed by the researcher rather than obtained as an attitudinal measure from the
is
it
has the disadvantage that
measure from which
it
may
be,
however, that employees
college or
who
in
are
commonly used
in social science
who
if
survey research.
these courses were pitched as
ways
may compare
to
their attainment to
improve one's
research can pursue multiple and more detailed measures of this variable, which
significantly related both to perceived
Upward
movement
variables,
advancement and
to beliefs
Future
career.
may prove
to be
about merit.
Actual upward mobility rate had a significant positive relationship to
mobility.
perceived advancement.
It
get a degree in a particular subject at a nearby
take a particular on-site course to upgrade their skills
one another, particularly
for the
an organization make more detailed comparisons about more
For example, people
specific types of training.
The education
was derived asked only about terminal degrees (except
this variable
"some college" category); these categories
may
not tap the social reality of the res[X)ndent
(The equations were also run including,
which was significant
others in one's job category,"
which
is
in the equations,
in separate turns, the
absolute
and the variable "mobility relative
only marginally significant.) This seemingly simple result
to
is
a
contribution of this study, because this relationship has been assumed or overlooked, but not
empirically demonstrated.
While appealingly simple,
Lateral moves (recent).
That
advancement.
lateral
moves can
for example,
was not a foregone conclusion.
this result
This variable has a significant, positive relationship to perceived
this variable is significant in addition to actual
upward mobility suggests
that
contribute independently of actual mobility to the perception of advancement.
two employees both have zero
overall mobility rates, but
one has made two
lateral
If,
moves
recently, that person should perceive slightly better advancement, despite the fact that the stark zero
mobility rate
is true for
both.
(49.7%) report no recent
4 moves.
That
This lateral
is,
move
In this sample, of the 177
lateral
employees who have zero mobility
moves,, while 66 (37.3%) report
half the people with zero
upward movement
1
rates,
88
move, and 23 (13.0%) report 2
to
report at least
contributes positively to their perceived advancement and
them the lower perceived advancement they would have on the
Crossing a class boundary.
1
may
recent lateral move.
partly ameliorate for
basis of their zero mobility rate alone.
The experience of crossing a
class
boundary contributes
26
positively to the perception of advancement
This result
is
consistent with interviewees' reports that
number of
positive experience, even controlling for actual
this is a particularly salient
levels
advanced.
Tenure.
included.
First,
Two
Tenure
is
not significantly related to perceived mobility once
may
increase an employee's sense of general satisfaction, including feelings about
advancement. Chinoy (1956) found
workers and attributed
it
this positive relationship
of tenure to satisfaction for automobile
to longer tenured employees' efforts to reduce their cognitive dissonance
over staying in a boring job with low mobility for a long time.
tenure,
employees who have not
And
Second, in contrast,
I
favor arguments
have a negative relationship specifically to perceived advancement.
had a spurt of mobility may
yet
particularly since significant mobility appears to
1979).
the controls are
explanations pose competing predictions about the direction of the effect of tenure.
longer tenure
that tenure will
all
with longer tenure, employees
happen
who have been mobile and
may
chances diminishing,
earlier in a career in a firm
&
Konda, 1984; White,
experience their mobility as too slow.
effects are in operation for different people, they
(Rosenbaum,
are near the top of their ladder
(Stewman
or the top of the firm face fewer available openings above them
1970), and, despite their position, they
feel their
With longer
would cancel any straightforward
If
both these
significant effect
of tenure.
It
is
unexplained.
worth bearing
A
in
simpler look
mind
at the
that
much
of the variance in perceived advancement
mobility helps to explain
left
data shows an asymmetric tendency for people to "over-perceive"
themselves to be advancing well, relative to their actual mobility
lateral
is
why
rate.
The
there might be a general tendency for
they are doing somewhat better than they actually
are.
The two
role of a variable like
employees
variables, perceived
to perceive
advancement and
mobility rate, are cross-tabulated in Table 12, demonstrating this asymmetry.
Ambivalence about meritocracy
The discussion of asymmetry above suggests
sense-making about workplace experiences.
that people
have some cognitive agility
This cognitive agility
is
downplayed
in the
in their
explanations
27
for
why
sound
may
people believe that hard work counts and
like
who
people
believe in hard
work do
why
people believe that luck counts, which
not believe in luck and vice versa.
not have such internally consistent perspectives.
However, what may characterize people's
ambivalence
may
be crucial for understanding
does not seem to result
work counts; indeed,
"ambivalent," but do not
this
difficult
On
may
This ambivalence
the margin, at least, hard
ambivalence to resolve
background.
is
that ability counts
other in national level status attainment studies
answers to the questions about
ability
(e.g.,
am
I
fact,
calling these people
two items
I
"ambivalents," and "agnostics" (who believe in neither).
that
label
if
label people's
assessment of
coming from
too.
how
A more
a privileged
have been pitted against each
et al., 1979).
and privilege to see
who
But luck can count
but so does
Jencks
representing four overall ideological views, which
is
In
also believe that hard
reflect a reasonable
work does count.
Ability and privileged background are the
possible groups
may
term pejoratively, as have some social scientists
beliefs inconsistent or incoherent.
things work.
that luck counts
balanced view would be quite a sensible one
mean
some
a particular items counts
ambivalence across items
is
is
the dissensus sprinkled throughout the system
The people who believe
in dissent.
this
why
beliefs
how much
it
In tact, people
This study has iu^gucd that there
dissensus about meritocratic ideology, looking at dissensus over
make
1
cross-tabulated people's
there were people in all four cells,
"true
believers,"
"true disbelievers,"
Table 13 shows that each of the conceptually
empirically represented in this sample.
Future research on ambivalence
is
warranted.
Conclusions and Implications
One
implication of these findings
legitimating ideology precisely because
is their
own
fault, as a legitimating
it
is
that meritocratic ideology
invites those
that disbelief
is
inherently flawed as a
doing poorly, not to believe that their position
ideology would do, but to protect their esteem and question the
ideology, as psychological theories suggest people would do.
is
is
What
1
call the "irony
generated precisely by the nature of meritocratic claims.
The
and differentially rewarding employees necessarily designates some employees as
of meritocracy"
practice of ranking
relative "losers" in
28
the meritocratic contest and places
them
employees may accept meritocratic ideology and blame themselves and
failures in the merit contest (only a small
variables in this study). This argument
employees who do
is
amount of
the variance in beliefs
in this study
is
explained by the
is
consistent with
advance the alternative possibility that
less well reject meritocratic claims.
This argument appeals to a different view of
They
are not too easily fooled, they are cognitively
employees emerging
in the sociological literature:
agile in their search for self-enhancing attributions about their position
particular experiences in
may
of these
their inferior merits for their
based on one view of employees and
The hypotheses
theories of perfect legitimation
Some
into lower positions or slower mobility paths.
an
ILM
inform their particular
beliefs.
By
and mobility
rate,
and their
touting meritocratic claims, firms
inadvertently but systematically demoralize a segment of their workers, precisely in the attempt
to build their
The
commitment.
findings in this study suggest that firms are not characterized by perfect legitimation.
Since there are some disbelievers in meritocracy in the firm, the interesting question that follows
whether there
is
a crisis of legitimacy for the firm.
retain a stable status
and mobility
disbelievers will be dispersed.
the mobility contest.
This study suggests two ways in which firms might
quo despite some degree of de-legitimation.
beliefs about meritocracy
is
of).
mobility rates relate to
who won and some
is,
after all,
- may reduce
Second, employees' dependence on the firm for
their willingness to dissent, despite dissatisfaction
preserve a stable status quo.
The
many
I
things
- from
affiliation,
and
relationship of beliefs about merit, versus other forms of attachment
from
this survey.
idea of "imperfect legitimation" prompts a different view of legitimation and a different
view of ILMs. At
However,
lost
with inequality, and
to the firm, to dissent is a subject that will be treated in another analysis of the data
The
who
what theories of legitimation were
income, job security, deferred compensation, and benefits to friendships, emotional
familiarity
people
another form of the balkanization of the workforce and would
probably tend to prevent collective dissent (which
invented to explain the lack
First, if
rates are distributed through all levels of the firm, then the
Every level will have some people
This pattern
is
first,
"imperfect legitimation"
may
appear to be less stable than perfect legitimation.
suggest that the coexistence of belief and disbelief in an
ILM may
be easier to sustain than
29
complete belief
collapse
If
A
in legitimating claims.
when employees
such questioning
is
state of perfect legitimation
may
be precarious,
it
see disconfirming evidence that causes questioning of meritocratic claims.
already going on, however, then
employees to moral outrage and a push
for changes.
it
is
harder to see what
ILMs may
it
would take
to spur
provide stability to firms that use
them, but not simply or largely because they include governance structures that promise
advancement and
more complicated
infer
that thereby gain the normative
Some employees may
from the seeming
may
stability
commitment of employees.
endorse the firm's meritocratic claims.
and success of any given firm
that uses
ILM
The
picture
fair
may
be
But we should not
practices that
achieved legitimation, "false consciousness," bureaucratic control, or normative commitment.
it
has
30
WORKING BIBLIOGRAPHY
Abercrombie, N.,
Unwin.
Althauser, R.P.,
I.
Berg
&
Hill, S.,
&
Turner, B.S.
(1980).
W. M.
&
&
London: Allen
Firms, occupations, and the structure of labor markets.
on labor markets, 119-149. New York: Academic Press.
Althusser, L. (1969) Ideological state apparatuses.
Alves,
thesis.
Kalleberg, A.L. (1981)
(Ed.), Sociological perspectives
earnings.
The dominant ideology
Rossi,
P.
Who
H. (1978)
For
Lenin.
should get what? Fairness judgements of the distribution of
American Journal of Sociology,
Baker, G.P., Jensen, M.C., & Murphy,
The Journal of Finance, 43, 593-616.
In
In
K.J.
84, 541-564.
(1988).
Compensation and
incentives: Practice vs. theory.
Bailyn, L. (1977). Research as a cognitive process: Implications for data analysis. Quality
and
Quantity, 11, 97-117.
Baker, G.P., Jensen, M.C.,
&
Murphy,
K.J. (1988).
Compensation and
incentives: Practice vs. theory.
The Journal of Finance, 42, 593-616.
Organizational perspectives on stratification.
Baron, J.N. (1984).
37-69.
Baron, J.N.,
&
Bielby,
W.T
(1986).
The
Annual Review of Sociology,
proliferation of job titles in organizations
10,
Administrative
Science Quarterly, 31, 561-586.
Baron, J.N, Davis-Blake, A.,
ladders vary within and
Becker, G. (1964).
Bell, D. (1972).
Human
On
&
among
Bielby, W.T. (1986).
organizations.
capital.
New
The
structure of opportunity:
promotion
York: Columbia University Press.
meritocracy and equality.
Public Interest, 29, 29-68.
The coming of post-industrial society: A venture
Bell, D. (1976).
How
Administrative Science Quarterly, 31, 248-273.
in social forecasting.
New
York:
Basic Books.
Berger, PL., & Luckman, T. (1966). The social construction of reality:
knowledge. New York: Doubleday.
Berk, R. A. (1983)
An
A
treatise in the sociology
introduction to sample bias in sociological data.
of
American Sociological
Review, 48, 386-398.
Bielby, W.T.,
discrimination.
&
(1986).
Men
and
women
at
work: Sex segregation and
statistical
American Journal of Sociology, 91, 759-799.
Bohmstedt G.W.,
Publishers.
Baron, J.N.
&
Knoke, D. (1982).
Statistics
for social data analysis. Itasca,
IL: F.E.
Peacock
31
Bourdieu,
(Eds.),
(1977).
P.
Cultural reproduction and social reproduction
Power and ideology
Bourdieu,
&
P.,
in education.
& AH.
Karabel
In J
Halsey
Cambridge: Cambridge University Press.
Passeron. J.C. (1977). Reproduction in education, society, and culture. London: Sage
Publications
Bowles,
&
S.,
Gintis, H. (1973).
I.Q. in the U.S. class structure.
Social Policy, January/ Februarv
1973, 65-96.
Bowles, S,
& Gintis,
&
H.
Schooling
(1976).
in capitalist
America.
(1989) The diverted dream: Community colleges and the promise of
educational opportunity in America, 1900-1985. New York: Oxford University Press,
Brint, S.
Karabel,
J.
Bulmer, M. (Ed.) (1975).
Working-class images of society.
Boston: Routledge and
Kegan
Burawoy, M. (1981). Manufacturing consent. Chicago: University of Chicago
Press.
The pattern of human concerns. Trenton: Rutgers University
Press.
Cantril, H. (1965)
Chinoy, E. (1955). Automobile workers and the American Dream.
Coleman, R.P.
&
Rainwater, L. (1978) Social standing
in
MA: Beacon
Boston,
America:
New
Paul.
Press.
dimensions of class.
New
York: Basic Books.
Daniels, N. (1978) Merit and meritocracy. Philosophy
&
Moore, W. (1966). Some principles of
Class, status, and power. New York: Free Press.
Davis, K.,
Deal, T.E.,
Reading,
&
MA:
and Public
stratification
Kennedy, A. A. (1982). Corporate cultures: The
206-223.
Affairs, 3,
In R.
rites
Bendix
and
& S.M.
Lipset (Eds),
rituals of corporate
life.
Addison-Wesley.
The meek shall not
American Sociological Review,
Delia Fave, L.R. (1980)
inherit the earth: self-evaluation
stratification.
45, 955-971.
Deming, W.E. (1986). Out of the
crisis.
Cambndge, MA; MIT Center
for
and the legitimacy of
Advanced Engineering
Study.
Deutsch, M.,
& Steil,
J.
DiPrete, T.A. (1987).
(1988).
Awakening
the sense of injustice.
Social Justice Research,
Horizontal and vertical mobility in organizations.
2,
3-23.
Administrative Science
Quarterly, 32, 422-444.
DiPrete, T.A., & Soule, W.T. (1988) Gender and promotions
American Sociological Review, 53: 26-40.
in
segmented job ladder systems.
.
32
&
Doeringer, P.B.,
MA: DC.
Piore, M.J.
(1971). Internal labor markets
and manpower
analysis.
Lexington,
Heath.
Dombusch,
S.
&
Scott,
W.R.
(1975). Evaluation
and
the exercise of authority.
San Francisco: Jossey-
Bass.
Dweck, C, Davidson, S., Nelson, W., & Enna,
The contingencies of evaluative feedback in
Developmental Psychology,
Sex differences in learned helplessness: II.
and III An experimental analysis.
B. (1978).
the classroom,
14, 268-276.
W
Dweck, C, & Goetz, T. (1978) Attributions and learned helplessness. In J Harvey,
Kidd (Eds.), New directions in attribution research. Vol. 2. Hillsdale, NJ: Elbaum.
Edwards, R. (1979). Contested terrain.
New
Elworth,
J.
The
(in progress).
legitimation of
11: Miscellanies.
P.
woman
in
in the iron
and
steel industry.
Press.
organizations.
Unpublished doctoral
Department of Sociology, Stanford University.
Emerson, R.W. (1968). The fortune of the
England,
R.
York: Basic Books.
Elbaum, B. (1984). The making and shaping of job and pay structures
In P. Osterman, Internal labor markets, 71-108. Cambridge, MA: MIT
dissertation.
&
Ickes,
New
(1981).
York:
AMS
republic.
In
E.W. Emerson (Ed), Works of Emerson,
Vol.
Press.
Assessing trends in occupational sex segregation, 1900-1976.
Sociological perspectives on labor markets, 11-\\1.
New
In
I.
Berg (Ed),
York: Academic Press.
Freeman, R. B. (1982) Union wage practices and wage dispersion within establishments. Industrial
and Labor Relations Review, 36, 3-2 1
&
NoUen, S.D. (1989).
presented at
and normative commitment.
the aimual meeting of the Academy of Management, Washington, D.C.
Gamson, W.
(in press).
Gaertner, K.N.,
Internal labor markets
Paper
Talking politics.
Gellerman, S.W., & Hodgson, W.G. (1988).
Harvard Business Review, May-June, 36-41.
Cyanamid's new take on performance appraisal.
Giddens, A. (1973) The class structure of the advanced societies.
New
Giddens, A. (1979). Communication and the evolution of society.
York: Harper
New
&
Row.
York: Cambridge University
Press.
Gioia, D.A.,
An
&
Sims, H.P.
empirical analysis.
Goffman,
E. (1952).
(1985).
Self-serving bias and actor-observer differences in organizations:
Journal of Applied Social Psychology, 15, 547-563.
On
'cooling out the mark.'
33
V
Goldberg.
Bridges over contested
(1980).
eniploNment relationship.
&
Gordon, D.M., Edwards, R.,
Cambridge Lniversity Press.
M & Tilly,
Granovetter,
Habermas,
,
Reich, M. (1982).
lixploring
account of the
249-274
radical
the
I,
Segmented work, divided workers.
London:
Inequality and labor processes.
C. (1986)
Legitimation
(1975).
J.
terrain:
Journal of Economic Behavior and Organization.
crisis.
(T.
McCarthy, Trans). Boston,
Administrative Science Quarterly, 23, 466-484.
Halaby, C. (1978)
Bureaucratic promotion
Halaby, C. (1986).
Worker attachment and workplace
criteria.
MA.
authority.
American Sociological Review, 51,
634-649.
&
Hanushek
Jackson.
Harder, J.W.
Play for pay: effects of inequity in a pay-for-performance
(1992).
context.
Administrative Science Quarterly, 37, 321-335.
Hirschman, A.O. (1970). Exit, voice, and loyalty: Responses
Cambridge, MA: Harvard University Press.
to decline in firms, organizations,
and
states.
What's fair? American beliefs about distributive justice.
Hochschild, J.L. (1981).
Harvard University
&
Holmstrom, B.R.,
Ilgen,
J.
&
J.
(1989).
Form, W. (1973). Income and ideology.
DR. &
B.M. Staw
The theory of the firm. In R. Schmalensee &
organization. Volume I. Amsterdam: North-Holland Press.
Tirole,
Handbook of industrial
Huber,
Cambridge,
MA:
Press.
New
R. Willig (Eds.),
York: Free Press.
Feldman, J.M. (1983). Performance appraisal: A process focus. In L.L. Cummings
Research in organizational behavior. Vol. 5. Greenwich, CT: JAI Press.
&
(Eds.),
Jackall, R. (1988)
Moral mazes: The world of corporate managers. New York: Oxford University
Press.
&
Jasso, G.
Rossi, P. H. (1977) Distributive justice and earned income. American Sociological
Review. 42, 639-651.
Jencks,
P.,
C,
Olneck,
Bartlett, S.,
M,
R,
J.,
Schwartz, J, Ward, S, Williams,
economic success
Jones,
Corcoran, M., Crouse,
in
& Nisbett,
America.
New
R. (1971).
E^glesfield, D., Jackson, G., McClelland, K., Mueser,
J.
(1979).
Who
gets
ahead? The determinants of
York: Basic Books.
The
actor and the observer: Divergent perceptions of the causes of
behavior. In E.E. Jones, D.E. Kanouse, H.H.Kelley,
Attribution: Perceiving the causes of behavior.
RE.
Nisbett, S. Valines,
&
B.
Weiner (Eds),
Morristown, NJ: General Learning Press.
34
Kerr,
&
J.,
Managing corporate
Slocum, J.W. (1987).
Management
Journal,
Kluegel, J.R.,
&
Smith, E.R.
&
Smith, E.R.
Kluegel, J.R.,
New
ought to be.
I,
culture through reward systems.
Academy of
99-108.
(1984).
Annual Review of Sociology.
(1986). Beliefs about inequality: Americans' views of
what
and what
is
York: Aldine de Gruyter.
How
Kochan, T.A. (1979).
American workers view labor unions. Monthly Labor Review,
April: 23-
31.
Latham, G.P.,
&
Wexley, K.N. (1980).
Increasing productivity through performance appraisal.
MA: Addison-Wesley.
Reading,
Lawler, E.E.
Pay and organizational
(1971).
effectiveness:
A psychological
New
view.
York:
McGraw-Hill.
Lawler, E.E. (1973). Motivation in work organizations.
Lawler, E.E.
(1990).
Monterey, CA: Brooks/Cole.
Strategic pay: Aligning organizational strategies
and pay systems. San
Francisco: Jossey-Bass.
Lazear, E.
Political
P.
&
Rosen,
Economy,
S.
(1981) Rank-order tournaments as optimum labor contracts
Journal of
89, 841-864.
Lenski, G.E. (1954).
A
Status crystallization:
nonvertical dimension of social status.
American
Sociological Review 19, 405-413.
Lemer, M.J. (1980). The belief in a just world: A fundamental delusion.
Lincoln, JR.,
plants
&
in the U.S.
Work organization and workforce commitment:
and Japan. American Sociological Review, 50, 738-760.
Lockwood, D. (1975). Sources of
Working-class images of society.
Mann, M.
J.
York: Plenum Press.
Kalleberg, A.L. (1985).
and employees
MacLeod,
New
variation in working-class images of society.
In
A
study of
M. Bulmer (Ed),
Boston: Routledge and Kegan Paul.
(1984). Ain't no making
it.
Boulder, CO:
Westwood
Press.
(1970).
The
Mann, M.
(1973).
Consciousness and action among the Western working class.
Macmillan
Press.
Martin,
(1981).
social cohesion of liberal democracy.
American Sociological Review, 35,
423-439.
J.
resources.
In L.L.
Relative deprivation:
Cummings
Greenwich, CT: JAI Press.
&
A
London: The
theory of distributive injustice for an era of shrinking
B.M. Staw
(Eds.),
Research
in organizational behavior,
Vol. 3.
35
&
&
DM.
A (1983) Distributive injustice and unfair exchange In K S C(X)k
New
Psychological
sociological
Yoric:
perspectives.
Praeger.
Theories
and
Messick (Eds.),
of equity:
Martin, J
Marx,
,
K.,
&
Murray,
Engels,
F.
The German ideology.
(1978)
C.J.
Arthur (Ed).
New
York: International
Publishers.
McCarthy, J D & Zald, M.N. (1977).
theory. American Journal of Sociology,
Resource mobilization and social movements:
,
Medoff, J.L., & Abraham, K.G.
of Economics, 95, 703-736.
A
partial
82, 1212-1241.
(1980).
Experience, performance and earnings.
Quarterly Journal
Meyer, J.W., & Rowan, B. (1977). Institutionalized organizations: Formal structure as myth and
ceremony. American Journal of Sociology, 83, 340-363.
Miliband, R. (1969).
&
Miller, D.T.,
Ross, M.
Self-serving attribution of causality: Fact or fiction? Psychological
(1975).
Bulletin, 82, 213-225.
Mills,
C.W.
The power
(1956).
Miner, A. (1987).
elite.
London: Oxford University
Idiosyncratic jobs in formalized organizations.
Press.
Administrative Science Quarterly,
32, 327-351.
Mitchell,
T.D Green, SO.,
,
&
Wood, RE. (1981)
An
performing subordinate: Development and validation.
in
organizational behavior. Vol.
3.
model of leadership and the poor
B. Staw and L. Cummings (Ekls), Research
attributional
In
Greenwich, CT: JAI Press.
Mittman, B.S. (1986). Mobility and inequality in the civil service: Job ladders and vacancy chains in
bureaucratic labor markets. Unpublished doctoral dissertation, Graduate School of Business,
Stanford University.
Mobley, W.H., Homer, S.O., & Hollingsworth, A. (1978). An evaluation of precursors of hospital
employee turnover. Journal of Applied Psychology, 63: 408-414.
Mowday,
&
R.M. (1982). Employee-organization
psycfiology of commitment, absenteeism, and turnover. New York: Academic Press.
R.T., Porter, L.W.,
& Cleveland,
& Bacon.
Murphy, K. R.
Boston: Allyn
Offe, C. (1976).
J.
Steers,
N. (1991) Performance appraisal:
linkages:
An organizational
Industry and inequality: The achievement principle in work
and
The
perspective.
social status.
(J.
Wickham, Trans). London: Edward Arnold.
O'Keefe, M., Viscusi,
W.
K.
&
Zeckhauser, R.
schemes. Journal of Labor Economics,
2,
27-56.
J.
(1984) Economic contests; Comparative reward
36
Osterman, P. (1984). Introduction: The nature and importance of internal labor markets.
Osterman (Ed.), Internal labor markets. Cambridge, MA: MIT Press.
W.
Ouchi,
Theory Z organizations.
(1980).
T
Parsons,
An
(1949).
in sociological theory.
The
Parsons, T. (1959).
analytic approach to the theory of stratification.
New
Pfeffer,
social class as a social system.
J.
(1981).
T
Parsons (Ed), Essays
Harvard Educational Review, 297-31
& Organ, D.W. (1986). Self-reports
Journal of Management, 12, 531-544.
paradigms.
In
York: Free Press.
Podsakoff, P.M.
prospects.
In P.
in organizational
research:
1.
Problems and
Management as symbolic action: The creation and maintenance of organizational
Cummings & B. Staw (Eds), Research in organizational behavior, vol. 3, 1-52.
In L.L.
Greenwich, CT: JAI Press.
Pfeffer,
&
J.,
Cohen, Y. (1984).
Determinants of internal labor markets in organizations.
Administrative Science Quarterly, 29, 550-573.
Pfeffer,
J.,
&
Lawler,
commitment on
paradigm.
J.
attitude
(1980).
Effects of job alternatives, extrinsic rewards,
toward the organization:
A
field test of the
and behavioral
insufficient justification
Administrative Science Quarterly, 25, 38-56.
&
The 1977 Quality of Employment Survey: Descriptive statistics
with comparison data from the 1969-70 and 1972-73 surveys. Ann Arbor, MI: Institute for Social
Quinn, R.P.,
Staines, G.L. (1979).
Research, University of Michigan.
Rae, D. (1981). Equalities. Cambridge,
MA:
Harvard University Press.
Rainwater, L. (1974).
Rawls,
J.
A
(1971).
theory of justice. Cambridge,
Rescher, N.P. (1966). Distributive justice.
Rosenbaum,
J.E. (1979).
New
MA: Harvard
University Press.
York: Bobbs-Merrill.
Tournament mobility: Career
patterns in a corporation. Administrative
Science Quarterly, 24, 220-221.
Runciman, W.G.
(1966).
Relative deprivation
inequality in twentieth century England.
Ryan, W. (1971). Blaming the victim.
Salancik, G.R. (1977).
Commitment
Schlozman, K.L. (1976).
society. Politics
and
is
Coping with
New
attitudes to social
York: Vintage Books.
too easy!
the
Society, 6, 213-240.
and social justice: A study of
CA: Univ. of California Press.
Berkeley,
Organizational Dynamics,
American Dream: Maintaining
6,
62-80.
self-respect in an achieving
37
&
Schlozman, K.,
\'erba. S. (1978).
Unemployment,
class,
and
Cambridge, MA:
political response.
Harvard I'niversity Press.
J. C.
Scott,
Weapons of
(1985).
the weak: everyday
forms of peasant resistance.
New
Haven: Yale
University Press.
The Scribner-Bantam EngHsh Dictionary (1979). New York: Bantam Books.
M.
Scully,
Unpublished undergraduate
inequality
R
Sennett,
Coping with meritocracy: Education, self-esteem, and the legitimation of
(1982).
.v.
,
&
Cobb,
(1972).
J.
Harvard-Radcliffe Colleges.
thesis,
The hidden injuries of class.
New
Sher, G. (1979) Effort, ability and personal desert. Philosophy
Sorensen,
jobs.
In
I.
.A.B.,
Berg
&
Kalleberg, A.L.
(Ed.), Sociological perspectives
Method
Spector, P.E. (1987)
M>th
An
(1981).
or significant problem?
Staw, B. (1984).
and Public
Affairs, 8,
361-376.
outline of a theory of the matching of persons to
on labor markets, 49-74.
New
York: Academic Press.
variance as an artifact in self-reported affect and perceptions at work:
Journal of Applied Psychology, 72: 438-443.
Organizational behavior:
variables.
Annual Review of Psychology,
Stewman,
S.
A
review and reformulation of the
field's
outcome
35, 627-666.
Demographic models of
(1986).
York: Vintage Books.
Administrative Science
internal labor markets.
Quarterly, 31, 212-247.
Stewman,
S.
&
Konda,
Stone, K. (1972).
The
Careers and organizational labor markets: Demographic models
American Journal of Sociology, 88, 637-85.
S.L. (1983).
of organizational behavior.
origins of the job structure in the steel industry.
market segmentation. Lexington,
Taylor, S.E.,
Taylor, S.
health.
&
&
Fiske, S.T.
Brown,
J.
MA: DC.
(1982).
Gordon
In D.
et al..
Labor
Heath.
Social cognition.
New
(1988). Illusion and well-being:
A
York:
Random House.
social psychological perspective
on mental
Psychological Bulletin, 103: 193-210.
Thomas,
R.J. (1989).
Arthur, D.T. Hall,
&
Meaning and choice in a world of constraints In MB.
Lawrence (Eds), Handbook of career theory. New York: Cambridge Univ.
Blue-collar careers:
B.S.
Press.
Walzer, M. (1983). Spheres of justice:
Weber, M. (1946). Class,
in sociology.
New
status, party.
A
defense of pluralism and equality.
In H.H. Gerth
& C.W.
Mills (Eds),
New
York: Basic Books.
From Max Weber: Essays
York: Oxford University Press.
Weber, M. (1978). Economy and society: An outline of interpretive sociology.
G Roth & C. Wittich (Eds). Berkeley, CA: University of California Press.
Vol. 2.
38
Weiner, B., Frieze,
I,
Kukla, A., Reed,
L., Rest, S.,
&
Rosenbaum, R.M. (1972)
of success and failure. In E.E. Jones, D.E. Kanouse, H.H.Keliey,
(Eds.), Attribution: Perceiving the causes
White, H.
MA:
RE.
of behavior. Morristown,
Nisbett, S. Valines,
N.J.:
Chains of opportunity: System models of mobility
(1970).
Perceiving the causes
&B
Weiner
General Learning Press,
in organizations.
Cambridge,
Harvard University Press.
Williamson, O. (1975). Markets and hierarchies.
Williamson, O.,
&
New
York: Free Press.
Ouchi, W. (1981). Markets, hierarchies, and clans.
Administrative Science
Quarterly.
Willis, P. (1977).
Learning
Columbia University
to labor:
How
working class kids get working class jobs.
J.V. (1989).
York:
Press.
Wonnacott, T.H., & Wonnacott, R.J. (1977). Introductory
second edition. New York: John Wiley & Sons.
Wood,
New
Theory and research concerning
statistics
for business and economics,
social comparisons of personal
attributes.
Psychological Bulletin, 106, 231-248.
Young, M. (1958). The
rise
Zenger, T.R.
Why
relationships
(1992).
among
of the meritocracy, 1870-2033.
York: Penguin Books.
do employers only reward extreme performance?
performance, pay, and turnover.
Zucker, L.G. (1981).
New
Examining
the
Administrative Science Quarterly, 37, 198-219.
Organizations as institutions
In
SB. Bacharach (Ed), Perspectives
organizational sociology: Theory and research. Greenwich, CT: JAI Press.
in
39
Table
1
Summary
of hypotheses.
40
Table
2.
Descriptive statistics.
41
Table
3.
Correlation Matrix
42
TaB^^^^Correlation Matrix (cont^iue^^
43
Tal^^^^Co^^^^Ton
Matrix (continued)
15.
-0.214***
17.
16.
18.
15.
Education
16.
Manager
17.
Prod
-0.379***
-0.147**
IB.
Business
-0.052
-0.038
-0.347***
19.
Engineer
0.409***
-0.115**
-0.719***
-0.403*** 1.000
20.
Firm A
-0.241***
-0.106**
-0.238***
-0.113**
21.
Hard work
0.008
22. Ability
23.
Perform
24.
Privilege
25.
Luck
+ p <
.10,
1.000
1.000
1.000
-0.316***
0.057
-0.081
-0.001
0.080
0.119**
-0.002
-0.087
0.039
0.056
-0.144***
0.061
0.107**
0.024
-0.122**
0.015
0.014
-0.026
-0.235***
0.038
21.
1.000
0.103**
0.040
*
20.
19.
0.011
-0.037
-0.012
p < .05, ** p < .01,
*** p < .001
•
*
0.095*
1.000***
-0.133**
1.000
-0.016
0.561***
-0.032
0.467***
0.109**
-0.002
-0.194***
-0.154**
44
TaE^TTTorre^^on
Matrix (cont^iuS)
22.
23.
22. Ability
1.000
23. Perform
0.488***
24.
Privilege
25. Luck
+ p <
.10,
*
p
<
24.
1.000
-0.261***
-0.121***
1.000
-0.229***
-0.273***
0.230***
.05,
25.
** p < .01,
*** p < .001
1.000
45
Table
4.
)
.
46
Table
5
Comparison of respondents rating of the importance of merit items
in the United States and in their company.
.
'
Ability counts
Hard work counts
Privilege counts
*** p <
Table
In the U.S.
In the company
Mean
Mean
(
s.d.
4.17 (1.45)
4.54 (1.31)
3.68 (1.40)
(
3 .d.
)
5.01 (1.33)
5.15 (1.23)
3.08 (1.99)
Correlation
.2828 ***
.2371 ***
.3320 ***
.001
6. Employees' normative ratings of how much hard work,
ability, performance, privilege, and luck ought to count
Item
Mean
s.d.
Items that ought to count very much according to meritocratic ideology
7-point scale in this study)
a
Hard work
(7
on
47
Table
7.
Relative mobility, by company, by job category.
Job category
representative
titles)
(
n
(
%
of
CO.
)
Percent
female
Mean levels
advanced
Range
s.d.
(
)
Skew
Kurtosis
Company A (n=147)
Production
assembler,
material handler,
technician)
1
(
2.
Administration
(office support,
programmer,
scheduler,
other non-exempt)
Business
(financial analyst,
marketer,
supervisor, manager,
other exempt)
3.
4
Engineering
(R&D engineer,
47
48.9
0.341
(1.160)
0-7
4.899
26.552
71.4
1.618
(1.792)
0-8
1.825
3.935
0-10
0.116
-1.602
(32.0)
35
(24.5)
34
50.0
4.382
(3.210]
(25.8)
26
(17.7)
11.5
1.348
(2.124)
0-8
1.900
3.490
53
(19.1)
26.4
3.176
(2.840)
0-11
0.517
-0.607
45.9
1.787
(2.332)
0-12
1.718
4.648
26.8
1.976
(2.128)
0-9
0.806
-0.285
manufacturing
engineer,
quality analyst)
Company B (n=278)
Production:
(assembler,
material purchaser,
technician,
inspector)
1.
Business/Admin
(administrator,
systems analyst,
accountant,
technical writer,
trainer,
other exempt
2.
61
(21.9)
Engineering
164
(software engineer,
(59.0)
hardware engineer,
manufacturing engineer,
scientist)
3.
.
48
Table
8.
Estimates of the return on education, tenure, and other
measures of human capital for Companies A and B
Dependent variable: Position
b value (t statistic shown in parentheses)
Company B
Company A
(1)
Education
(15.227)
Tenure
in firm
2.195***
(16.747)
0.216***
(4.057)
(2)
(1)
2 .080***
(6.463)
(6.124)
0.232***
(5.146)
0.714***
0.231***
(4.564)
4.517***
(8.032)
Mobility rate
(2)
0.682***
0.254***
(4.883)
0.192
(0.468)
Performance
evaluation
Firm- specific
skills
(self rating)
-0.174
(-0.879)
-0.075
(-0.463)
0.358***
(1.469)
0.357
Productivity
relative to
similar others
(self rating)
-0.427*
(-2.107)
0.059
(0.387)
Constant
-4.763
(-6.448]
Adjusted R^
.669
-2.382
(-1.681)
.776
(2.854)
7.351
(10.292)
4.817
(3.352)
.175
Descriptive statistics for the saved residual
(referred to as "Relative return on education and tenure")
mean
0.002
2.614
-6.517
8.120
.205
4')
Table
9.
OLS e3timate3 of
50
Tiible 10.
Suitmary of significant
Table 11. OLS regression estimates of
:
52
Table 12. Examination of the incidence of "mismatch" between
perceived and actual mobility.
Shown for each cell:
Frequency
% of column
% of total
Perceived mobility
Not advanced
1,2
Mobility
rate
44
,01
to .50
.50
Column
total
Notes
(7
point scale)
Advanced
somewhat
3,4,5
Advanced
very much
6,7
53
Table 13.
Ccmbinat
9
C
0':^'^
-J
O
^j
c: t;
i
'-^
J u
MIT LIBRARIES DUPl
3
TOaO D0a5b317
M fl V
'
',
.iJate Due