You are on page 1of 20

International Review of Public Administration, 2015

Vol. 20, No. 1, 51–69, http://dx.doi.org/10.1080/12294659.2014.983216

The effect of gender discrimination in organization


Sunhee Kim*

Department of Political Science and Public Administration, Seowon University, Cheongju, Korea
(Received 28 July 2014; final version received 28 October 2014)

Since discrimination against women in the workplace is generally regarded as


having a negative impact on quality of life, systematic managerial efforts to prevent
such discrimination have been enforced. However, there are very few studies on
whether or how this discrimination affects the subjective well-being in Asia where
the unequal treatment of women has been traditionally accepted. Our study exam-
ines whether perceived/experienced discrimination against women consistently influ-
ences two subjective kinds of well-being: job satisfaction and work engagement. It
will consider how the impact varies according to different dimensions of discrimina-
tion, what kinds of moderators intervene to influence the effect on subjective well-
being and how they do so. Empirical findings from three-wave panel data with a
sample of 5987 female workers show the negative effects of discrimination on these
two forms of subjective well-being.
Keywords: discrimination; subjective well-being; gender

Introduction
The increase in the number of women in the labor force has presented both social
opportunity and challenge. In terms of social change, the growing influx of productive
women into the labor market has a positive impact on organizational productivity and
economic growth. Even if there are some evidences that there are no differences
between females and males (Jeong, 2001), discriminatory employment practices and
cultures that affect women employees’ subjective well-being still exist. Workplace
discrimination is defined as unfair and negative treatment of workers or job applicants
based on personal attributes that are irrelevant to job performance (Chung, 2001).
According to the 2009 Gender Employment Equality Index in Korea (Ministry of
Employment and Labor [MOEL], 2009), which examined discrimination amongst sala-
ried workers, there was a great difference between male and female salaried workers in
terms of employment status and position held. Amongst men, 3.44% held managerial
status and 66.7% were permanent rather than temporary workers; whereas 0.33% of
women were managers and 49.9% of women were workers with long-term tenure. The
average earnings per hour also showed inequality: 12,911 won ($11.90) for men versus
8,926 won ($8.25) for women.
Why does workplace discrimination matter? At the organizational level,
discriminatory practices have a negative impact on organizational performance and
productivity (Ensher, Grant-Vallone, & Donaldson, 2001; Foley, Hang-Yue, & Wong,
2005; Mays, Coleman, & Jackson, 1996; Sanchez & Brock, 1996; Shim, Kwon, Park,

*Email: shkim7675@hanmail.net

© 2014 The Korean Association for Public Administration


52 S. Kim

& Hwang, 2011). Moreover, perceived discrimination can affect the degree to which an
organization’s members create an integrated culture or share common values, which in
turn can affect its policies, procedures and day-to-day life (Ensher et al., 2001). At the
individual level, discrimination perceived by female employees is associated with nega-
tive outcomes such as work conflict and reduced productivity (Gutek, Cohen, & Tsui,
1996).
However, there are some limits to existing research. First, since existing empirical
research has focused heavily on the gender effect, in particular comparing the effects of
discrimination between men and women (Foley et al., 2005; Foley, Ngo, & Loi, 2006;
Gutek et al., 1996; Sweeney & McFarlin, 1997), there are few studies considering how
or to what extent discrimination has an unwanted implication for the quality of life of
women workers. Secondly, existing studies have not specified which conditions facili-
tate these effects and dismiss the factors that moderate the positive/negative relation-
ships between discrimination and quality of life. Thirdly, even if the unequal treatment
of women has been traditionally accepted in Oriental society, little research has been
done in this area.
Our study will empirically test whether discrimination really influences the quality
of life of women workers (casual effect), how different types of discrimination influ-
ence subjective well-being (dimensional effect) and what conditional moderators change
the causal relationship between discrimination and quality of life (conditional effect).

Theoretical background
The effect of discrimination on quality of life
To examine the effects of discrimination on quality of life at the organizational level,
we established a research model which consists of two types of discrimination –
perceived and experienced discrimination – as independent variables, and the two sub-
jective kinds of well-being of women workers – job satisfaction and work engagement
– as a dependent variable.

Perceived discrimination vs. experienced discrimination


We set perceived and experienced discrimination as independent variables. Perceived
discrimination is the perception of an individual that he or she is treated differently or
unfairly because of his or her membership in a particular group (Foley et al., 2005;
Sanchez & Brock, 1996).
How does the perceived discrimination cause the negative outcome? According to
social identity theory (Tajfel & Turner, 1985), people put themselves and others into
social categories which have been characterized using prototypes drawn from the cate-
gory member. The sense of belonging gives two contrasting effects; it gives people the
comfortable feeling of belonging to a category because they have a kind of human
resource to depend on, whereas it also leads to personal conflict because they received
the differential treatment based on such categories. Hence, social identity may provide
the foundation from which discrimination could be interpreted. Perceived discrimination
is an individual’s perception that he or she is treated differently or unfairly because of
his or her group membership (Sanchez & Brock, 1996).
Why is subjective discrimination more important than objective discrimination?
According to relative deprivation theory, satisfaction is the opposite of relative depriva-
tion. Such relative deprivation may come from comparing one’s status with that of a
International Review of Public Administration 53

referent group (Dion, 1986). Although people may have more resources than did the
referent group, the former felt they had less than the latter did. The more important
thing in judging fairness is not the matter of the objective state but rather the subjective
judgment.
Such perceived discrimination may be as important as actual inequality. Since previ-
ous research has focused on only one dimension of discrimination, usually perceived
discrimination (Ensher et al., 2001; Sanchez & Brock, 1996), few studies have investi-
gated the effect of experienced discrimination (Koomen & Fränkel, 1992).
We adopted not only perceived discrimination but also experienced discrimination
in the research model. First, we include the latter because one’s personal experiences
and the experiences of others with whom one identifies are an influential factor that
may affect the perception of discrimination. Second, since discrimination, by its nature,
has multiple meanings, it is difficult to cover them fully by using only one concept.
The distinction between perceived and experienced concepts will contribute to covering
the conceptually heterogeneous components of discrimination. The former is concerned
with attitude, whereas the latter is related to life events.

Job satisfaction vs. work engagement


In the dependent variables, we measure subjective well-being using job satisfaction and
work engagement. Satisfaction represents the cognitive dimension of subjective well-
being, whereas engagement represents the affective dimension. According to Pavot and
Diener (1993, p. 164), subjective well-being usually consists of two contrasting concep-
tual components: the cognitive and affective dimensions. The affective component
represents pleasant or unpleasant mood, whereas the cognitive component is described
as life satisfaction. While the affective and cognitive components of subjective well-
being are not completely independent, they are somewhat distinctive and can provide
complementary information when assessed separately.
Based on Russell (2003)’s core affect, Bakker and Oerleman (2010) divide the sub-
jective well-being into engagement, satisfaction, workaholism and burnout. As shown
in Figure 1, engagement represents the highly activated and pleasant state of one’s
mind, whereas satisfaction keeps a pleasant but low-activated mood.
Our research assumed that job satisfaction is related to the cognitive evaluation of
one’s own life in the organization, whereas work engagement is associated with the
positive/negative emotional aspects.

Job satisfaction
Job satisfaction can be defined as a pleasurable or positive emotional state resulting
from one’s own appraisal of one’s job or job experience (Locke, 1976). According to
Ensher et al. (2001), job satisfaction is a critical component of employee attitudes that
is likely to be affected by perceived discrimination.
There are many empirical studies on the effect of discrimination on job satisfaction
amongst minority groups. After analyzing Hispanic workers, Sanchez and Brock (1996)
reported that perceived discrimination negatively influences job outcomes more than
other stressors (e.g., role ambiguity and role conflict). With regard to female workers,
Sanchez and Brock (1996) demonstrated that perceived discrimination amongst female
employees creates more work tension, therefore decreasing job satisfaction.
54 S. Kim

Figure 1. A two-dimensional view of work-related subjective well-being.


Note: adapted from Bakker and Oerlemans (2010).

Why does discrimination have a negative effect on job satisfaction? It is a matter of


both justice and self-esteem or perceived control. McFarlin and Sweeney (1992)
explain that justice influences personal satisfaction. Discrimination is critical for evalu-
ating justice, not only at the individual but also at the organizational level.

Work engagement
Bakker and Leiter (2010) state that ‘work engagement is a positive, work-related state
of well-being or fulfillment characterized by a high level of energy and strong identifi-
cation with one’s work.’
Welle and Heilman (2005) explain that many women who believe they have experi-
enced discrimination, or have seen colleagues affected by it, show less engagement in
their work. Moreover, they argue that discriminatory work environments can encourage
the psychological disengagement of women from their work. The discrimination is
related to work engagement in terms of justice. In broader views, discrimination is seen
as a matter of justice. Inoue et al. (2010) show that procedural and interactional justice
is positively related to work engagement. Moreover, we consider the effect of discrimi-
nation on work engagement in terms of fairness. Perceived/experienced discrimination
is closely related to fairness: ‘Fairness is the extent to which a decision at work is per-
ceived as being fair and people are treated with respect’ (Leiter & Maslach, 2004,
p. 98). Fairness has a negative impact on burnout, opposite to work engagement. After
examining the role of fairness in supervisory support, Leiter and Harvie (1997) reported
that workers who perceived that their managers treated them fairly were less sensitive
to burnout, i.e., more work engagement.
International Review of Public Administration 55

H1: There is a negative impact of perceived/experienced discrimination on employees’ sub-


jective well-being. i.e., job satisfaction and engagement.

Dimensional effect of discrimination


The dimensional effect of discrimination is concerned with whether or not different kinds
of discrimination have the same impact on subjective well-being. Even if discrimination
negatively affects psychological well-being in organizations, multi- or sub-dimensions of
discrimination do not bring out the same effect on subjective well-being. In the other
words, various attributes of discrimination are likely to result in different effects in terms
of their content and degree. Mays et al. (1996) argue that although various kinds of dis-
crimination have different impacts on individual or organizational results, it is hard to
know which type of discrimination is comparatively more important.
A few studies have examined how different dimensions of discrimination have
unequal effects on outcomes. For example, Ensher et al. (2001) revealed that two dif-
ferent types of discrimination, in this case relational (by supervisor and co-worker) and
organizational, have distinctive effects.
Based on these studies, we suggest a second research hypothesis. To examine the
dimensional effect of discrimination, we first divide discrimination into two contrasting
dimensions: perceived versus experienced discrimination. We then divide each of them
again into six sub-domains: selection, promotion, pay and benefits, deployment, training
and lay-off.
H2: Each dimension of discrimination has different impacts on quality of life or subjective
well-being.

Conditional effect of discrimination


‘Conditional effect’ refers to the impact of contextual moderating variables on the rela-
tionship between discrimination and subjective well-being. Discrimination produces
negative psychological stressors; however, this relationship is not always direct because
buffering or facilitating moderators intervene in the process. Discrimination usually
occurs under specific contextual conditions that, to a large extent, moderate the direc-
tive relationships between discrimination and subjective well-being. Noh and Kaspar
(2003) argue that the effects of social context are involved in dealing with perceived
racial discrimination.
Personal or organizational conditions usually play a role in buffering the negative
impact of discrimination on subjective well-being within an organization. In an empiri-
cal analysis, Jackson, Williams, and Torres (2003) suggest that psychological coping
resources, such as emotional support and self-efficacy, religious belief and action
responses, reduce the impact of perceived discrimination on health. In our analysis,
such moderating conditions are concerned not only with individuals’ personal
resources, such as income and level or status within an organization, but also with
organizational characteristics, such as organizational support. The former is attributed
to individual characteristics, whilst the latter is concerned with the organization.

Personal resources
Many researchers have proved the positive function of personal resources between
discrimination and negative psychological outcome in organizations. For example,
56 S. Kim

Sanchez and Brock (1996) reported, using a sample of Hispanic employees, that a
coping resource that may interact with perceived discrimination reduces the effect of
discrimination on employees’ psychological outcomes.
Our study focused on income and higher status in organizations. Gutek et al.
(1996) explain the income effect, in which income is positively associated with a feel-
ing of power and prestige. Hence, if women had a high income, this might temper neg-
ative feelings and outcomes associated with sex discrimination.
Status or level within an organization is power. Gutek et al. (1996) argue that per-
ceived discrimination tends to be associated with lower power within an organization.
In an organizational context, Sanchez and Brock (1996) explain that the advantages
associated with high rank should facilitate the ability to cope with discrimination.
Those who take higher positions utilize resources such as authority, power and direc-
tion, all of which help them cope positively with the negative effects of discrimination.

Organizational support
Organizational support can be defined as psychological or institutional resources that
decrease the negative effects of discrimination. The indirect role of organizational support
often moderates the effect of discrimination. Many empirical studies have proven the buf-
fering effect of organizational support on the relationship between discrimination and
organizational output. Noh and Kaspar (2003) prove that moderating effects of coping,
acculturation and ethnic support exist between perceived discrimination and personal out-
comes such as depression. They explain that people who are discriminated against but
are also empowered with sufficient social resources are more likely to confront than
accept racial bias. Such buffering effects should be most effective when the source of
support is relevant to the stressor being experienced (Sanchez & Brock, 1996).
In short, the evidence regarding the roles of personal resources and organizational
support suggests that these factors function as direct or indirect buffers, thus attenuating
the negative effects of discrimination; for example, increasing job satisfaction and work
engagement.
H3: Personal resources and organizational support not only have a significant direct
impact on subjective outcomes, but also serve as conditional moderators between discrimi-
nation and subjective well-being in an organization.

Method
Our analysis is based on survey data from the Korean Women’s Development Institute
(KWDI)’s Korean longitudinal survey of women and families (KLoWF), a large longi-
tudinal panel study consisting of three waves: the first wave in 2007, the second in
2009 and the third in 2011 (KWDI, 2007, 2009, 2011). Data was collected by way of
computer-assisted personal interviewing by professional interviewers. The survey pri-
marily adopted random systemic stratified sampling methods. We analyzed the data
from 5987 of 30,359 samples: 1996 samples in the first wave, 1838 in the second wave
and 2153 in the third wave. We analyzed only female ‘salaried’ workers, excluding
other respondents such as non-workers, non-salaried workers and those employed in
special fields.
As an example shown in Table 1, perceived or experienced discrimination was mea-
sured using six items, relating respectively to selection, promotion, pay and benefits,
International Review of Public Administration 57

Table 1. Theoretical concepts and measures.

No. of
items/
Theoretical composite
concept Measure Response scale method
Perceived In the process of selection for jobs, Four-point scale (1 = Six/mean
discrimination a male is preferred over a female in disagree strongly, 4 =
spite of women having the same agree strongly)
capability.
Experienced After last survey, have you Dichotomous scale (0 Six/sum
discrimination experienced or are you experiencing = no, I have not
the several discriminations experienced, 1 = I
mentioned below? have experienced)
Job satisfaction To what extent are you satisfied Four-point scale (1 = Eight/mean
with each of the next eight fields strongly dissatisfied to
(salary/income, employment 4 = strongly satisfied)
stability, work content, work
environment, work hours, personal
development, communication and
relationship and welfare)?
Work engagement To work gives me worth and a vital Four-point scale, (1 = One/NA
life. disagree strongly to 4
= agree strongly)
Organizational The next 13 items, i.e., retirement Dichotomous scale (0 Thirteen/
support allowance, paid family leave, = no, not provided, 1 sum
childcare facilities in the workplace, = yes, provided)
etc., provide you with a sense of
increasing well-being.

deployment, training and lay-off. In order to combine the six items into a single com-
posite scale, we calculated the mean of the scores of perceived discrimination and the
sum of those of experienced discrimination.
Because we measured the work engagement by one item, we need check its the valid-
ity and reliability. First, to check the face validity, after comparing each item used with
the Utrecht Work Engagement Scale (UWES) (Schaufeli, Bakker, & Salanova, 2006), we
assessed the commonality between them. Our measure ‘To work gives me worth and a
vital life’ is similar to the measures of vigor (VI) in Schaufeli’s ‘I am bursting with energy
in my work’ (VI 1) and ‘At my job, I feel strong and vigorous’ (VI 2). Such work contrib-
utes to confirming the face validity of items we used. Second, to check the reliability,
since we used the 3-year waves of panel data rather than a single year, we computed Alpa
= .542 (test-retest reliability) by using the same items across 3 years.

Results: determinants, dimensionality and conditionality


The effect of perceived/experienced discrimination
To estimate the effect of perceived/experienced discrimination on subjective well-being
within the organization, we proceed to regression analysis with pooled data. To control
the heterogeneity, we used the panel-corrected standard error (PCSE) method, as shown
in Table 2.
To estimate the relative explanatory power of discrimination, we controlled for
related variables such age, education, type of employing organization (private or
Table 2. Regression of two types of discrimination on subjective well-being within an organization.
58

Model 1: Job satisfaction (N = 3254; R² = 0.1957; Wald Model 2: Work engagement (N = 3253; R² = 0.0508; Wald
chi²(12) = 719.15, Prob > chi² = 0.000) chi²(12) = 181.67, Prob > chi² = 0.000)
Coef. SE z P > z [95% Conf. Interval] (Beta) Coef. SE z P > z [95% Conf. Interval] (Beta)
Perceived discrimination as independent variable
Constant 2.761 0.095 29.01 0.00 2.57 2.95 – 2.911 0.084 34.68 0.000 2.747 3.076 –
Age squared .000 .000 0.50 0.618 0.000 0.000 0.008 0.000 0.000 1.98 0.048 2.46E-07 0.000 0.036
Education 1: middle −0.015 0.030 −0.49 0.625 −0.073 0.044 −0.009 −0.028 0.028 -1.01 0.314 −0.083 0.027 -0.019
Education 2: univ. 0.038 0.024 1.55 0.121 −0.010 0.086 0.029 −0.005 0.023 -0.21 0.837 −0.050 0.041 -0.004
Private company −0.030 0.025 −1.20 0.231 −0.078 0.019 −0.021 0.031 0.024 1.31 0.189 −0.015 0.077 0.025
Permanent worker −0.021 0.023 −0.92 0.360 −0.066 0.024 −0.016 −0.024 0.022 -1.11 0.268 −0.067 0.019 -0.021
Organization size −0.001 0.007 −0.13 0.898 −0.015 0.013 −0.002 0.007 0.007 1.02 0.310 −0.006 0.020 0.019
Year of employment −0.002 0.002 −1.04 0.299 −0.006 0.002 −0.017 −0.001 0.002 -0.73 0.467 −0.005 0.002 -0.013
(Log) income 0.115 0.017 6.83 0.000 0.082 0.147 0.141 0.072 0.014 5.12 0.000 0.045 0.100 0.102
Status 0.167 0.023 7.18 0.000 0.122 0.213 0.123 0.123 0.022 5.60 0.000 0.080 0.165 0.104
Union membership −0.016 0.032 −0.51 0.609 −0.078 0.046 −0.009 −0.004 0.031 -0.11 0.910 −0.064 0.057 -0.002
Organizational 0.047 0.003 14.67 0.000 0.041 0.053 0.283 0.009 0.003 3.30 0.001 0.004 0.015 0.064
S. Kim

support
Perceived −0.106 0.016 −6.67 0.000 −0.137 −0.075 −0.113 −0.071 0.015 -4.65 0.000 −0.101 −0.041 -0.087
discrimination
Experienced discrimination as independent variable
Model 3: Job satisfaction (N = 3145; R² = .1901; Wald chi²(12) Model 4: Work engagement (N = 3144; R² = .0427; Wald
= 676.39, Prob > chi² = .000) chi²(12) = 145.53, Prob > chi² = .000)

Coef. SE z P > z [95% Conf. Interval] (Beta) Coef. SE z P > z [95% Conf. Interval] (Beta)
Constant 2.551 0.089 28.55 0.000 2.375 2.726 – 2.754 0.080 34.64 0.000 2.598 2.910 –
Age squared 0.000 0.000 0.82 0.413 −1.52E-05 0.000 0.013 0.000 0.000 2.41 0.016 0.000 0.000 0.045
Education 1: middle −0.005 0.030 −0.15 0.881 −0.064 0.055 −0.003 −0.030 0.028 −1.05 0.295 −0.085 0.026 −0.020
Education 2: univ. 0.043 0.025 1.74 0.081 −0.005 0.091 0.033 −0.006 0.024 −0.23 0.817 −0.052 0.041 −0.005
Private company −0.022 0.025 −0.89 0.374 −0.071 0.027 −0.016 0.032 0.024 1.32 0.186 −0.015 0.079 0.026
Permanent worker −0.027 0.023 −1.18 0.237 −0.073 0.018 −0.021 −0.024 0.022 −1.10 0.273 −0.068 0.019 −0.022
Organizational size 0.002 0.007 0.27 0.790 −0.012 0.016 0.005 0.006 0.007 0.87 0.383 −0.007 0.019 0.017
Year of employment −0.002 0.002 −0.98 0.328 −0.006 0.002 −0.017 −0.001 0.002 −0.58 0.564 −0.005 0.003 −0.011
(Log) income 0.114 0.017 6.72 0.000 0.081 0.147 0.142 0.074 0.014 5.10 0.000 0.045 0.102 0.104
Status 0.174 0.023 7.50 0.000 0.129 0.220 0.130 0.135 0.022 6.17 0.000 0.092 0.178 0.114
Union membership −0.016 0.032 −0.51 0.611 −0.080 0.047 −0.009 −0.006 0.032 −0.18 0.856 −0.068 0.056 −0.004
Organizational 0.045 0.003 14.05 0.000 0.039 0.052 0.275 0.008 0.003 2.73 0.006 0.002 0.014 0.055
support
Experienced −0.067 0.012 −5.84 0.000 −0.090 −0.045 −0.099 −0.007 0.011 −0.67 0.502 −0.029 0.014 −0.012
discrimination
Note: Education, private company, permanent worker and union are dummy variables. Referent groups are high school graduates, non-profit and other organizations, temporary
workers and non-union members. Organizational size is the number of employees.
*P < 0.1; **P < 0.05; ***P < 0.01.
International Review of Public Administration
59
60 S. Kim

non-profit/other organization, type of employment (permanent or temporary worker),


organizational size, the tenure (year of employment), income and union membership
(union or non-union member). Nominal variables were recoded into dummy variables
in which the target group was recorded as 1 and other statuses were recorded for refer-
ence as 0.
Based on the Wald Chi² test, all models are statistically significant. Based on
R-squared, four models explain the lowest, 4.27%, to the highest, 19.57%, variance of
satisfaction and engagement. However, there are, to a large extent, gaps between job
satisfaction and work engagement. The former is explained very well – 19.57% in
model 1 and 19.01% in model 2; whereas the latter revealed a low level of explanation,
i.e., 5.08% in model 2 and 4.27% in model 4. This finding implies that the present
model is more appropriate for explaining job satisfaction than work engagement.
Next, as predicted in hypothesis H1, perceived discrimination is negatively related
to job satisfaction and work engagement. Experienced discrimination significantly
decreases job satisfaction. These results confirm Ensher et al.’s (2001) findings that,
after controlling for gender, income and other variables, perceived discrimination
explains a significant amount of variance in job satisfaction.
Perceived discrimination did not significantly explain work engagement, whereas
experienced discrimination does not have an effect on work engagement. One possible
explanation is that since job satisfaction and work engagement have different attributes
in their conceptual components, they will be influenced by different causes. There are
low coefficients between job satisfaction and work engagement: .005 in the case of par-
tial correlation in which age, income, and status controlled; and .001 in the case of
Pearson correlation. Such weak relationships appear in the empirical testing by
Rothmann (2008). Second, the low variance in experienced discrimination has contrib-
uted to weakening the statistical significance or reversing the logically expected direc-
tion of the relationship between variables. Only 14.8% of 5695 respondents
experienced discrimination. Also, the mean (.32) and standard deviation (.90) revealed
low values, which induced narrow variance and then lower covariance.
Viewing the findings from standardized Beta coefficients, the relative explanatory
power of the two types of discrimination is weaker in comparison with the Beta coeffi-
cients of other controlled variables. Among control variables, three variables (higher
income, status and organizational support) are consistently connected with higher satis-
faction and engagement. This implies that personal resources and organizational support
directly influence subjective well-being within an organization.

Dimensionality of discrimination
In relation to the dimensional effects of discrimination, Table 2 shows that two kinds
of discrimination contribute to explaining subjective well-being within an organization.
Both perceived and experienced discrimination decrease job satisfaction, whereas only
perceived discrimination decreases work engagement. Moreover, the two types of dis-
crimination have different explanatory power and significance. These differences imply
that they might comprise different dimensions.
To establish a further dimensional effect of discrimination, we compared the effect
of the two types of discrimination by inputting them into a single regression model. As
shown in Table 3, when controlling for each other, the two kinds of discrimination
have their own distinct impacts on job satisfaction and work engagement.
Table 3. Perceived discrimination versus experienced discrimination.

Model 1: Job satisfaction (N = 3141) Model 2: Work engagement (N = 3140)


[95% [95%
Coef. SE z P > z Conf. Interval] (Beta) Coef. SE z P>z Conf. Interval] (Beta)
Constant 2.738 0.096 28.58 0.000 2.551 2.926 – 2.915 0.086 34.020 0.00 2.747 3.083 –
Age squared 0.000 0.000 0.63 0.527 0.000 0.000 0.010 0.000 0.000 2.230 0.03 0.000 0.000 0.041
Education 1: middle −0.006 0.030 −0.20 0.841 −0.065 0.053 −0.004 −0.031 0.028 −1.100 0.27 −0.087 0.024 −0.021
Education 2: univ. 0.038 0.024 1.57 0.116 −0.009 0.086 0.030 −0.009 0.024 −0.380 0.70 −0.056 0.037 −0.008
Private company −0.022 0.025 −0.90 0.366 −0.071 0.026 −0.016 0.031 0.024 1.300 0.19 −0.016 0.078 0.025
Permanent worker −0.027 0.023 −1.19 0.236 −0.072 0.018 −0.021 −0.023 0.022 −1.020 0.31 −0.066 0.021 −0.020
Organization size 0.002 0.007 0.25 0.804 −0.012 0.016 0.005 0.005 0.007 0.820 0.41 −0.008 0.019 0.016
Year of employment −0.002 0.002 −1.01 0.314 −0.006 0.002 −0.017 −0.001 0.002 −0.640 0.52 −0.005 0.002 −0.012
Income 0.114 0.017 6.76 0.000 0.081 0.147 0.142 0.075 0.014 5.180 0.00 0.046 0.103 0.105
Status 0.154 0.023 6.57 0.000 0.108 0.199 0.115 0.116 0.022 5.220 0.00 0.073 0.160 0.098
Union membership −0.016 0.032 −0.49 0.622 −0.078 0.047 −0.009 −0.006 0.031 −0.180 0.85 −0.067 0.056 −0.004
Organizational 0.046 0.003 14.41 0.000 0.040 0.053 0.280 0.009 0.003 3.000 0.00 0.003 0.014 0.059
support
Perceived −0.089 0.017 −5.35 0.000 −0.122 −0.057 −0.096 −0.077 0.016 −4.740 0.00 −0.109 −0.045 −0.094
discrimination
Experienced −0.050 0.012 −4.14 0.000 −0.074 −0.026 −0.074 0.009 0.011 0.750 0.45 −0.014 0.031 0.014
International Review of Public Administration

discrimination
R² .1963 .0495
Wald chi² 699.37*** 169/007***
Note: *P < 0.1; **P < 0.05; ***P < 0.01.
61
62 S. Kim

In model 1, both types of discrimination significantly explain job satisfaction,


whereas in model 2, only perceived discrimination influences work engagement.
In model 1, 2, if comparing the Beta coefficients, perceived discrimination shows a
little more explanatory power than experienced discrimination. This implies that the
two types of discrimination have their own dimension which each type of discrimina-
tion can explain well.
Next, to see the impact of the six different dimensions of discrimination (selection,
promotion and advancement, pay and benefits, deployment, education and lay-off) on
subjective well-being within an organization, we calculated the mean value of the
dependent variables (job satisfaction and work engagement) according to each catego-
rized discrimination group (the ‘not perceived’ groups versus ‘perceived’ groups in per-
ceived discrimination, as shown in Figure 2; and the ‘not experienced’ groups versus
‘experienced’ groups in experienced discrimination, as shown in Figure 3).
When comparing the mean values of well-being between ‘discriminated’ and ‘not
discriminated’ groups, several dimensions of discrimination revealed statistical differ-
ences in the analysis of variance (ANOVA) test. In Figure 2, in the case of job satisfac-
tion, perceived discrimination in selection, pay and benefits, deployment, training and
lay-off brought out significant gaps in subjective well-being between the ‘perceived’
and ‘not perceived’ groups. However, the difference with regard to promotion appears
not to be significant. Moreover, in the case of work engagement, only two out of six
perceived discrimination dimensions reveal significant differences.
The degree of the gap between the ‘perceived’ and ‘not perceived’ groups has a
variation across the types of discrimination. The gap score between the ‘perceived’ and
‘not perceived’ group in job satisfaction is highest in the case of the pay and benefits

Figure 2. Means of job satisfaction (JS)/work engagement (WE) between ‘perceived’ and ‘not
perceived’ groups (ANOVA-test).
Note: ANOVA: analysis of variance; *P < 0.1; **P < 0.05; ***P < 0.01.
International Review of Public Administration 63

Figure 3. Means of job satisfaction (JS)/work engagement (WE) between ‘experienced’ and
‘not experienced’ groups (ANOVA-test).
Note: ANOVA: analysis of variance; *P < 0.1; **P < 0.05; ***P < 0.01.

field (3.213 minus 3.08), whereas it is lowest in the case of promotion (3.198 minus
3.151). Such variations in the gaps suggest the difference in degree of impact made by
each type of discrimination.
Figure 3 shows different the degrees of two types of organizational well-being
according to whether respondents experienced discrimination. In the case of job satis-
faction, four out of six forms of experienced discrimination show statistical signifi-
cance; whereas, in the case of work engagement, only one (pay and benefits cases) out
of six has a significant effect.
In short, the above evidence demonstrates the dimensional effect of discrimination.
All forms of discrimination do not have the same effect on subjective well-being.
Moreover, the degree of impact also appears to differ, varying with the type of discrim-
ination.

Conditionality of discrimination
Not all types of discrimination always have a direct link to subjective well-being within
an organization. The effects of discrimination are buffered or facilitated by personal or
organizational conditions, which usually play moderating roles. We hypothesized that
personal resources and social support would moderate the relationship between discrim-
ination and employees’ subjective well-being.
To see the effects of interaction, we examined the role of three different moderators:
income, status and organizational support. To regress the subjective well-being on inde-
pendent variables including interaction terms, we used the parsimony model shown in
Table 4, which used second-wave data and only seven out of 11 independent variables
64 S. Kim

adopted in Table 2, since the heteroskedasticity in pooled data and a lot of independent
variables prevent the inferring of exact estimation parameters of interaction coefficients
and constants which are key parameters for a simple slope test.
For perceived discrimination, Table 4 displays the results of the hierarchical regres-
sion analyses in which three moderators – income in model 1, status in model 2 and
organizational support in model 3 – significantly moderate the impact of perceived dis-
crimination on job satisfaction. In the case of work engagement, perceived discrimina-
tion is moderated only by status in model 6. However, the effect of experienced
discrimination on subjective well-being was not moderated by income, status and
organizational support.
Table 4. Regression analysis of interaction effect.

Model 1 Model 2 Model 3


Perceived
Discrimination B SE Beta B SE Beta B SE Beta
Job satisfaction as dependent variable
(Constant) 2.787*** .117 2.601*** .116 2.706*** .117
Age .001 .002 .019 .002 .002 .025 .001 .002 .023
Education .093*** .028 .108 .096*** .028 .111 .096*** .028 .112
Income .111*** .016 .175 .001*** .000 .166 .00***1 .000 .165
Status .163*** .037 .111 .169*** .037 .115 .164*** .037 .111
Welfare .043*** .004 .240 .043*** .004 .238 .157*** .016 .247
PD –.062*** .014 –.090 –.035** .017 –.050 –.062*** .014 –.090
Interaction term –.043*** .015 –.061 –.105*** .033 –.076 –.025* .014 –.037
PD × income PD × position PD × organizational
support
F-Value 75.571*** 75.879*** 74.581***
Adjusted R- .224 .224 .221
square
F-value change 8.436*** 10.113*** 3.050*
R-square change .004 .004 .001
Work engagement as dependent variable
Model 5 Model 6 Model 7
 
B SE Beta B SE Beta B SE Beta
(Constant) 3.063*** .111 2.957*** .110 2.973*** .111
Age .001 .002 .018 .001 .002 .022 .001 .002 .020
Education .036 .027 .049 .037 .027 .050 .037 .027 .051
Income .063*** .015 .116 .001*** .000 .112 .001*** .000 .111
Status .065* .035 .052 .068* .035 .054 .064* .035 .051
Welfare .003*** .004 .019 .003 .004 .018 .014 .015 .025
PD −.050 .014 −.085 −.031* .016 −.052 −.050*** .014 −.085
Interaction term −.019 .014 −.031 −.075** .031 −.064 −.019 .014 −.033
PD × income PD × position PD × organizational
support
F-value 10.678*** 11.267*** 10.705***
Adjusted .036 .038 .036
R-square
F-value change 1.759 5.724** 1.940
R-square change .001 .003 .001
Note: F-value change and R-square change mean refer to the changed variance against models that appeared
in Table 2.
*P < .1; **P < 0.05; ***P < 0.01.
International Review of Public Administration 65

Since it is difficult to interpret the four significant interaction terms in the regres-
sion, we illustrate the simple slope in Figures 4–7, which depict the moderating effect
of personal resources and organizational support. To depict the figures, except for the
organizational position variable, we divide, based on mean score, the respondents into
two groups. For example, in the case of organizational support, the respondents in
higher groups had scores higher than the mean in support measures whereas those in
lower groups had scores below the mean. Organizational position is a dichotomous var-
iable that consists of non-managerial or non-professional position (= 0) and managerial
or professional position (= 1).
As is evident from Figure 4, perceived discrimination has less negative impact
when income is high; thus income plays a role in buffering the negative results of
discrimination. The simple slope at the high level of income presents as statistically

Figure 4. The moderating effect of income (MV) between PD (IV) and job satisfaction (D).
Note: D: dependent variable; IV: independent variable; MV: moderating variable;
PD: perceived discrimination.

Figure 5. The moderating effect of status (MV) between PD (IV) and job satisfaction (D).
Note: D: dependent variable; IV: independent variable; MV: moderating variable;
PD: perceived discrimination.
66 S. Kim

Figure 6. The moderating effect of support (MV) between PD (IV) and job satisfaction (D).
Note: D: dependent variable; IV: independent variable; MV: moderating variable;
PD: perceived discrimination.

significant (b = –.105, standard error [SE] = .036, t-value = –.949, p < .01), but that at
the low level does not (b = –.019, SE = .020, t-value = –.948, p-value = .343).
Such a buffering effect is also observed in both organizational status and support.
The simple slope for high and low management level shows as statistically significant
(b = –.545, SE = .029, t-value = –19.102, p-value < 0.01 in the case of high
management level, and b = –.035, SE = .017, t-value = –2.099, p-value < .05 in the
case of low management level). The slope for high organizational support is greater
(b = –.087, SE = .021, t-value = –4.185, p < .01) than that for low organizational sup-
port (b = –.036, SE = .020, t-value = –1.774, p-value < .1).
Regarding job satisfaction, in Figures 4 to 6, those with greater personal resources
(income and status) and organizational resources (organizational support) show more
job satisfaction than those with fewer resources, regardless of whether perceived
discrimination has increased or decreased.

Figure 7. The moderating effect of status (MV) between PD (IV) and work engagement (D).
Note: D: dependent variable; IV: independent variable; MV: moderating variable;
PD: perceived discrimination.
International Review of Public Administration 67

In Figure 7, even where perceived discrimination exists, there is a large gap in


work engagement between respondents holding high-level status and those at low-level
status. This suggests that even if employees are experiencing the same discrimination,
those holding a low status in the organization may feel more negative effects from
perceived discrimination. The simple slope is statistically significant (b = –.106,
SE = .027, t-value = –3.907, p-value < 0.01 for high management level, and b = –.031,
SE = .016, t-value = –1.953, p-value < 0.1 for low management level).
In short, we can confirm that personal resources and organizational support serve as
conditional moderators which reduce the negative effects of perceived discrimination,
thereby acting as a protective instrument.

Discussion and implications


As more women have entered the labor market, the issue of discrimination has received
more attention in terms of both practical management and theoretical research. Discrimi-
nation is known to predict not only employees’ personal distress, but also aspects of sub-
jective well-being within organizations, such as job satisfaction and work engagement
(Ensher et al., 2001; Foley et al., 2005; Mays et al., 1996; Sanchez & Brock, 1996). Our
study examines not only whether perceived/experienced discrimination consistently influ-
ences subjective well-being within an organization, but also what dimensions and condi-
tions of discrimination affect it. We provide key findings as follows.
Firstly, empirical results show that perceived/experienced discrimination is nega-
tively associated with two subjective types of well-being within an organization: job
satisfaction and work engagement. Perceived discrimination has a negative relationship
with job satisfaction and engagement. Furthermore, experienced discrimination nega-
tively predicts job satisfaction but not work engagement, after controlling for other
related variables.
Secondly, we demonstrate the dimensional effects of discrimination on subjective
well-being within an organization, and confirm that each dimension of perceived and
experienced discrimination shows a different effect on subjective well-being within an
organization. Perceived discrimination significantly decreases work engagement, whereas
experienced discrimination does not show such an effect. Moreover, the two dimensions
of discrimination in the full regression model have their own significant variance in the
standardized Beta coefficient. Also, regarding the effect of six different kinds of discrimi-
nation (selection, promotion, pay and benefits, deployment, training and lay-off) on sub-
jective well-being, each shows different statistical significance and degree of influence.
Thirdly, to prove the effect of personal or organizational conditions as buffers or
facilitators, we investigated moderators such as personal resources, income or status
within an organization and organizational support. By analyzing the effect of interaction
in regression, we confirm that personal resources and organizational support play a
moderating role that reduces the negative effect of perceived discrimination. Such con-
ditional effects imply that perceived/experienced discrimination alone may directly link
with negative subjective well-being in an organization, as the effect of such forms of
discrimination is moderated by personal resources or organizational support. Such
results confirm the moderating effect of salary as well as of organizational position
(Sanchez & Brock, 1996).
Our research suggests that workers’ perceived discrimination, to a large extent,
influences subjective well-being within an organization. These empirical findings
suggest several theoretical and practical implications, as outlined below.
68 S. Kim

Firstly, personnel management should monitor discriminatory practices with an eye


toward prevention as they have a negative effect on subjective well-being within in the
organization. Secondly, based on the significant dimensional effect of discrimination,
management should consider examining different kinds of discrimination. Different
types of discrimination have different effects on subjective well-being within an organi-
zation. Thirdly, in practical terms, to reduce the negative effects of discrimination, orga-
nizations need to pay attention to conditional moderators.
Our study has several limits. Firstly, the low variance in experienced discrimination
has contributed to weakening the statistical significance or bringing out a logically
unexpected direction in the relationship between variables. Secondly, because we used
existing survey data rather than a self-designed survey, we cannot fully explore the the-
oretical concepts which have extensive dimensions of meaning, in particular work
engagement.

Notes on contributor
Sunhee Kim holds a PhD in public administration from Korea University. She is an
assistant professor at Seowon University, South Korea. Her research interests include
gender study, public management and welfare policy. She has published several papers
on social welfare and gender policies in journals such as the Korean Journal of Public
Administration (‘Analysis of determinants of women’s WFC [work-family conflict] in
public organizations’) and the Asian Journal of Women’s Studies (‘Exploring the effect
of four factors on affirmative action programs for women’).

References
Bakker, A. B., & Leiter, M. P. (2010). Where to go from here: Integration and future research on
work engagement. In A. B. Bakker, & M. P. Leiter (Eds.), Work engagement: A handbook of
essential theory and research (pp. 181–196). New York: Psychology Press.
Bakker, A. B., & Oerlemans, W. G. M. (2010). Subjective well-being in organizations. In
K. Cameron, & G. Spreitzer (Eds.), Handbook of positive organizational scholarship
(pp. 178–189). Oxford: Oxford University Press.
Chung, Y. B. (2001). Work discrimination and copying strategies: conceptual frameworks for coun-
seling lesbians, gay and bisexual clients. The Career Development Quarterly, 50(1), 33–44.
Dion, K. L. (1986). Responses to perceived discrimination and relative deprivation. In
J. M. Olson, C. P. Herman, & M. P. Zanna (Eds.), Relative deprivation and social
comparison: The Ontario symposium, (Vol. 4, pp. 159–179). Hillsdale, NJ: Erlbaum.
Ensher, E. A., Grant-Vallone, E. J., & Donaldson, S. (2001). Effects of perceived discrimination
on job satisfaction, organizational commitment, organizational citizenship behavior, and
grievances. Human Resource Development Quarterly, 12(1), 53–72.
Foley, S., Hang-Yue, N., & Wong, A. (2005). Perception of discrimination and justice: Are there
gender difference in outcomes? Group & Organization Management, 30(4), 421–450.
Foley, S., Ngo, H., & Loi, R. (2006). Antecedents and consequences of perceived personal gen-
der discrimination: A study of solicitors in Hong Kong. Sex Roles, 55(3/4), 197–208.
Gutek, B. A., Cohen, A. G., & Tsui, A. (1996). Reactions to perceived discrimination. Human
Relations, 49, 791–814.
Inoue, A., Kawakami, N., Ishizaki, M., Shimazu, A., Tsuchiya, M., Tabata, M., … & Kuroda, M.
(2010). Organizational justice, psychological distress, and work engagement in Japanese
workers. International Archives of Occupational and Environmental Health, 83(1), 29–38.
Jackson, J. S., Williams, D. R., & Torres, M. (2003). Perceptions of discrimination, health and
mental health: the social stress process. In A. Maney (Ed.), Social stressors, personal and
social resources and their mental health consequences (pp. 86–146). Rockville, MD:
National Institute of Mental Health.
International Review of Public Administration 69

Jeong, C. H. (2001). Gender differences in computer attitudes: New evidence from Korea.
International Review of Public Administration, 6(2), 115–123.
Koomen, W., & Fränkel, E. G. (1992). Effects of experienced discrimination and different forms
of relative deprivation among Surinamese, a Dutch ethnic minority group. Journal of Com-
munity & Applied Social Psychology, 2, 63–71.
KWDI (Korea Women’s Development Institute). (2007). KLoWF (The Korean longitudinal survey
of women and families). Seoul: KWDI.
KWDI (Korea Women’s Development Institute). (2009). KLoWF (The Korean longitudinal survey
of women and families). Seoul: KWDI.
KWDI (Korea Women’s Development Institute). (2011). KLoWF (The Korean longitudinal survey
of women and families). Seoul: KWDI.
Leiter, M. P., & Harvie, P. (1997). The correspondence of supervisor and subordinate perspectives
on major organizational change. Journal of Occupational Health Psychology, 2, 1–10.
Leiter, M. P., & Maslach, C. (2004). Areas of work life: A structured approach to organizational
predictors of job burnout. In P. Perrewé, & D. C. Ganster (Eds.), Research in occupational
stress and well being, (Vol. 3., pp. 91–134). Oxford, UK: JAI Press/Elsevier.
Locke, E. A. (1976). The nature and causes of job satisfaction. In M. D. Dunnette (Ed.), Handbook
of industrial and organizational psychology (pp. 1297–1343). Skokie, IL: Rand McNally.
Mays, V. M., Coleman, L. M., & Jackson, J. S. (1996). Perceived race-based discrimination,
employment status, and job stress in a national sample of black women: Implication for
health outcomes. Journal of Occupational Health Psychology, 1(3), 319–329.
McFarlin, D. B., & Sweeney, P. D. (1992). Distributive and procedural justice as predictors of
satisfaction with personal and organizational outcome. The Academy of Management Journal,
35(3), 626–637.
MOEL (Ministry of Employment and Labor). (2009). Gender employment equality index. Seoul:
MOEL.
Noh, S., & Kaspar, V. (2003). Perceived discrimination and depression: moderating effects of
coping, acculturation, and ethnic support. American Journal of Public Health, 93(2),
232–238.
Pavot, W., & Diener, E. (1993). Review of the satisfaction with life scale. Psychological Assess-
ment, 5(2), 164–172.
Rothmann, S. (2008). Job satisfaction, occupational stress, burnout and work engagement as com-
ponents of work-related wellbeing. Empirical Research, 34(3), 11–16.
Russell, J. A. (2003). Core affect and the psychological construction of emotion. Psychological
Review, 110, 145–172.
Sanchez, J. I., & Brock, P. (1996). Outcomes of perceived discrimination among hispanic
employees: Is diversity management a luxury or a necessity? Acaemy of Management
Journal, 39, 704–719.
Schaufeli, W. B., Bakker, A. B., & Salanova, M. (2006). The measurement of work engagement
with a short questionnaire. Educational and Psychological Measurement, 66, 701–716.
Shim, D. C., Kwon, Y. S., Park, H. H., & Hwang, S. W. (2011). Linking pay-for-performance
system with performance improvement: The role of fairness, organizational resources, and
leader/managerial engagement. International Review of Public Administration, 16(2), 49–69.
Sweeney, P. D., & McFarlin, D. B. (1997). Process and outcome: Gender differences in the
assessment of justice. Journal of Organizational Behavior, 18, 83–98.
Tajfel, H., & Turner, J. C. (1985). The social identity theory of intergroup behavior. In
S. Worchel, & W. G. Austin (Eds.), Psychology of intergroup relations (pp. 7–24). Chicago,
IL: Nelson-Hall.
Welle, B., & Heilman, M.E. (2005). Formal and informal discrimination against women at
work: The role of gender stereotypes [online]. Centre for Public Leadership, Working Paper
No. 2, pp. 24–40, Retrieved from http://dspace.mit.edu/bitstream/handle/1721.1/55933/
CPLWP0502HeilmanWelle
Reproduced with permission of the copyright owner. Further reproduction prohibited without
permission.

You might also like