Tài liệu tiếng anh tham khảo business society

/>Business & Society
/>The online version of this article can be found at:

DOI: 10.1177/0007650302041003003
2002 41: 292Business Society
Kristin B. Backhaus, Brett A. Stone and Karl Heiner
Exploringthe Relationship Between Corporate Social Performance and Employer Attractiveness


Published by:

On behalf of:



International Association for Business and Society
can be found at:Business & SocietyAdditional services and information for





Alerts:

What is This?

- Sep 1, 2002Version of Record >>
at Harvard Libraries on August 17, 2012bas.sagepub.comDownloaded from
BUSINESS & SOCIETY / September 2002Backhaus et al. / CSP AND EMPLOYER ATTRACTIVENESS
Exploring the Relationship Between
Corporate Social Performance
and Employer Attractiveness
KRISTIN B. BACKHAUS
BRETT A. STONE
KARL HEINER
State University of New York at New Paltz
Building on existing studies suggesting that corporate social performance (CSP) is
important in the job choice process, the authors investigate job seekers’ percep
-
tions of importance of CSP and explore effects of CSP dimensions on organiza-
tional attractiveness. Job seekers consider CSP important to assessment of firms
and rate five specific CSP dimensions (environment, community relations, em-
ployee relations, diversity, and product issues) as more important than six other
CSP dimensions. Using signaling theory and social identity theory, the authors hy-
pothesize differences in effects of CSP data on ratings of employer attractiveness
and find that environment, community relations, and diversity dimensions have
the largest affect on attractiveness ratings.
The demographics of our workforce are changing, highlighted by the exit
of the baby boom generation and the entrance of Generation Y workers.
The evolving needs and values of today’s workers make employee recruit
-
ment more challenging. “How-to” articles about recruiting and attracting
new employees abound in the popular business press, but scholars are
challenged to explain clearly the science of organizational attraction.

What makes an organization attractive to a potential applicant?
One stream of research suggests that job seekers prefer organizations
with whom they perceive congruence between their and the organization’s
primary values (Cable & Judge, 1994; Chatman, 1989, 1991; Judge &
Bretz, 1992; Judge & Cable, 1997; Schneider, 1987). In the job search
process, individuals tend to be more attracted to organizations with which
they perceive a match (Judge & Bretz, 1992).
292
BUSINESS & SOCIETY, Vol. 41 No. 3, September 2002 292-318
© 2002 Sage Publications
at Harvard Libraries on August 17, 2012bas.sagepub.comDownloaded from
A second stream of research focuses on attributes of the organization.
Studies suggest that organization structure (Turban & Keon, 1993) and
reward systems (Bretz, Ash, & Dreher, 1989) influence initial organiza
-
tional attraction. Organization image has also been shown to influence the
firm’s ability to attract applicants (Belt & Paolillo, 1982; Gatewood,
Gowan, & Lautenschlager, 1993; Rynes, 1991; Tom, 1971). By organiza
-
tional image, we refer to the general impressions held by those outside the
organization (Barber, 1998). Tom (1971) described image as the collec
-
tion of knowledge, beliefs, and feelings about an organization. Image may
derive from snippets of information or from in-depth involvement with the
organization. Image resides at the level of perception and can change over
time. Organizational image has been shown to influence early job choice
decisions (Fombrun & Shanley, 1990). Gatewood and colleagues (1993)
found image to be related to familiarity with the firm and personal interac
-
tion with it. Because image is constructed from a broad range of impres

-
sions created by various organizational activities, the firm’s corporate
social performance helps to inform perceptions of image (Greening &
Turban, 2000).
Corporate social performance (CSP) has been defined as “a business
organization’s configuration of principles of social responsibility, pro-
cesses of social responsiveness and policies, programs, and observable
outcomes as they relate to the firm’s societal relationships” (Wood, 1991,
p. 693). CSP is a multidimensional construct, encompassing organiza-
tional activities related to treatment of employees, the natural environ-
ment, workplace diversity, customers, product, and other issues (Berman,
Wicks, Kotha, & Jones, 1999). Research suggests that a firm’s corporate
social performance may influence perceptions of organizational attrac
-
tiveness (Greening & Turban, 2000; Turban & Greening, 1997). Our study
aims to add to this literature by investigating job seekers’ perceptions of
the importance of CSP and exploring the differential effects of CSP
dimensions on organizational attractiveness. This study adds to previous
research in a few key respects. First, we examine both the attitudes of job
seekers toward CSP and the way in which CSP information affects job
seekers as they evaluate the attractiveness of employers. Second, of 11
CSP dimensions typically considered under the rubric of CSP, most
research focuses on only 5 (employee relations, the natural environment,
product quality, treatment of women and minorities, and community rela
-
tions). Although these 5 dimensions are the ones most commonly used in
CSP research, no empirical evidence has been presented to support the
inclusion of these 5 dimensions over any others. Given the growing list of
CSP areas of concern in our society today (Stone, 2001), it is becoming
increasingly important to demonstrate empirically that these particular

Backhaus et al. / CSP AND EMPLOYER ATTRACTIVENESS 293
at Harvard Libraries on August 17, 2012bas.sagepub.comDownloaded from
CSP dimensions are considered most important to job seekers. Third, pre
-
vious research results (Greening & Turban, 2000; Turban & Greening,
1997) support the notion that there is a relationship between CSP and
organizational attractiveness, but still we cannot discern the degree to
which variations in levels of corporate social performance across various
CSP areas will affect this relationship. We address this issue. Finally, we
explore patterns of effects of combinations of CSP ratings on organiza
-
tional attractiveness.
CONCEPTUAL FRAMEWORK AND HYPOTHESES
Corporate Social Performance
Corporate Social Performance (CSP) has become an important issue to
the public, even to the extent that it influences financial investments.
Investment in socially and environmentally responsible firms increased
82% in 2 years, whereas investment in the broad market was up only 42%
(Social Investment Forum, 1999). In addition, considerable scrutiny has
been placed on the role of business in society over the past 10 years (Harri-
son & Freeman, 1999). Marketing research suggests that consumers are
paying attention to the CSP records of the companies from whom they are
making purchases (Gildea, 1994; Maignan, Ferrell, & Hult, 1999; Owen
& Scherer, 1993; Paul, Zalka, Downes, Perry, & Friday, 1997). A recent
survey revealed that 88% of consumers are more likely to buy from a com-
pany that is considered socially responsible (Smith, 1996).
Most CSP research has been conducted to measure the relationship
between financial performance and social performance. Although these
studies have been largely inconclusive, there have been positive links
found between the two (Cochran & Wood, 1984; Coffey & Fryxell, 1991;

McGuire, Sundgren, & Schneeweis, 1988; Waddock & Graves, 1994;
Wokutch & Spencer, 1987), but others have revealed unclear or negative
links (Alexander & Buchholz, 1982; Aupperle, Carroll, & Hatfield, 1985;
Cochran & Wood, 1984; Coffey & Fryxell, 1991; McGuire et al., 1988;
Shane & Spicer, 1983). However, in the process of exploring the financial
implications of CSP, researchers have uncovered other important aspects
of CSP. Waddock and Graves (1997) found that good CSP contributes to
improved financial performance. From this, they developed the good
management theory, suggesting that good management of relationships
with various stakeholders results in stronger corporate performance
(Waddock & Graves, 1997).
294 BUSINESS & SOCIETY / September 2002
at Harvard Libraries on August 17, 2012bas.sagepub.comDownloaded from
Stakeholder theory provides the underpinning for the good manage
-
ment theory and also provides conceptual support for the study reported
here. Stakeholder theory posits that companies have a responsibility to
those who have vested interests in firm performance and those who are
directly affected by the firm’s actions (Evan & Freeman, 1983; Freeman,
1984). Stakeholder theory provides a logical explanation for why CSP
matters. Treating stakeholders well, making decisions that affect employ
-
ees, communities, and the environment positively leads to positive out
-
comes. Indeed, studies support the notion that effective management of
key stakeholder relationships has positive financial implications for orga
-
nizations (cf. Berman et al., 1999).
Stakeholder theory suggests that organizations must be responsive to
the competing demands of those who hold a stake in the organization.

Employees are among the important stakeholders identified in CSP
research. Effective management of the relationship with employees was
found to affect firm financial performance (Berman et al., 1999). Mitchell,
Agle, and Wood (1997) suggested that employees have sufficient power,
legitimacy, and urgency to become salient stakeholders to top manage-
ment. As Greening and Turban (2000) argued, prospective employees
may also have this power, legitimacy, and urgency, especially when jobs
are plentiful. Moreover, based on stakeholder theory and public interest in
CSP, there is reason to argue that CSP may be an important tool in attract-
ing employees.
Signaling Theory
In his seminal work on recruitment and selection, Wanous (1992)
pointed out that job seekers require complete and accurate organizational
information to match their needs properly with organizational offerings.
However, job seekers usually have limited information about organiza
-
tions and must use bits and pieces of data to construct a view of what it
would be like to work for an organization (Barber, 1998). Organizational
characteristics have been shown to be indicative of personnel practices
(Jackson, Schuler, & Rivero, 1989), and job seekers tend to use these char
-
acteristics as clues. This use of organizational attributes as predictors of
working conditions is captured under the rubric of signaling theory. Signaling
theory suggests that individuals use various clues, dropped by the firm, to
draw conclusions about the firm’s intentions or actions (Srivastava &
Lurie, 2001). In the organizational choice process, prospective employees
use any available information to improve their efforts to make a rational
decision (Wanous, 1992). Information about certain CSP dimensions may
Backhaus et al. / CSP AND EMPLOYER ATTRACTIVENESS 295
at Harvard Libraries on August 17, 2012bas.sagepub.comDownloaded from

provide the data a job seeker needs to assess the appropriateness of the
employer.
Social Identity Theory
Social identity theory suggests that individuals derive their self-concept
in part from their membership in certain social groups (Tajfel, 1982). In
other words, we define our identity in terms of our group membership, and
enhance our self-esteem by comparing our group to lesser quality groups
(Stets & Burke, 2000). The successes and reputation of our group contrib
-
ute to our self-concept (Underwood, Bond, & Baer, 2001). Our employer
is an important source of self-concept (Tajfel, 1982). We enjoy the bene
-
fits of our employer’s positive reputation but also suffer detrimental
effects of our firm’s negative reputation (Dutton, Dukerich, & Harquail,
1994). This provides another foundation on which to argue that a firm’s
CSP will be important to job seekers as a way of selecting a self-enhancing
employment setting.
Organizational Attraction
Although the study of organizational attraction has revealed some
insights, there remains much to be learned (Barber, 1998). One stream of
extant research investigates organizational characteristics and their effects
on attraction to the organization. Structural attributes, such as decentral-
ized decision making (Turban & Keon, 1993) and reward systems (Bretz
et al., 1989), are shown to influence perceptions of attractiveness.
Gatewood and associates (1993) examined corporate image, another
organizational characteristic, finding that perception of an organization’s
image is a significant predictor of decisions to pursue employment with
that company.
Specifically in the area of CSP, Bauer and Aiman-Smith (1996) find
that individuals are attracted to employers with a proenvironmental

stance, whereas Wright and associates (Wright, Ferris, Hiller, & Kroll,
1995) concluded that organizations with positive affirmative action pro
-
grams are more successful in attracting high-quality applicants. Turban
and Greening’s (1997) study found a positive relationship between pub
-
lished ratings of firms’ CSP and participants’ ratings of firms’ attractive
-
ness. Because they find a correlation between CSP and attractiveness, the
study concluded that organizational attractiveness perceptions may be
influenced by CSP. Greening and Turban’s (2000) study expanded on their
previous study, measuring the relationship between CSP and attractive
-
ness more directly by supplying CSP data to participants and testing the
296 BUSINESS & SOCIETY / September 2002
at Harvard Libraries on August 17, 2012bas.sagepub.comDownloaded from
effects of that information on their perceptions. Again, they found a posi
-
tive relationship between CSP ratings and attractiveness ratings. Albinger
and Freeman (2000) also found that CSP influences attractiveness ratings
but only for those job seekers with high levels of job choice.
Thus, previous research suggests that attractiveness and CSP are
related. Furthermore, signaling theory and social identity theory suggest
that CSP may inform the applicant’s job choice process. On that basis, we
begin our two-phase study by determining the extent to which our sample
considered CSP relevant in the job search process.
Hypothesis 1: Job seekers rate CSP as an important organizational attribute
when considering prospective employers.
Dimensions of CSP
Methodological problems have made the study of CSP difficult

(Waddock & Graves, 1997). The primary problem relates to the
multidimensionality of CSP. Various measures have been used as proxies
for CSP, but the variation among them has led to difficulty in generalizing
findings or even drawing practical conclusions from studies. Many of
these problems have been addressed effectively by the use of data from
Kinder, Lydenberg, and Domini (cf. Graves & Waddock, 1994.) Kinder,
Lydenberg, Domini (KLD) ratings measure social performance of major
corporations, assessing firms in 11 (previously 8) categories of CSP,
which include community, diversity, employees, product, (ecological)
environment, non-U.S. operations, nuclear power, involvement in alco
-
hol, gambling, tobacco, and military contracting, as well as an “other” cat
-
egory (Waddock & Graves, 1997). Of these 11 dimensions, 5 (environ
-
ment, community relations, diversity, product issues, employee relations)
have become widely used in management research on CSP (Berman et al.,
1999; Greening & Turban, 2000; Johnson & Greening, 1999; Turban &
Greening, 1997).
These 5 dimensions have proved to be most relevant to research related
to stakeholder impressions of CSP (Berman et al., 1999; Waddock &
Graves, 1997) and have been the most studied of 11 dimensions in the aca
-
demic literature. However, with regard to recruitment, no empirical evi
-
dence has been compiled to indicate that these 5 dimensions are the most
relevant to job seekers. In addition to the weight of evidence presented in
other stakeholder/CSP studies, we suggest that signaling theory and social
identity theory support the importance of these 5 dimensions.
First, as previously indicated, signaling theory suggests that job seek

-
ers look for clues to indicate what it would be like to work for a company.
Backhaus et al. / CSP AND EMPLOYER ATTRACTIVENESS 297
at Harvard Libraries on August 17, 2012bas.sagepub.comDownloaded from
The concept of issue intensity, a term borrowed from ethical decision-
making theory (Jones, 1991), suggests that the greater the importance of
an issue to the decision maker, the more salient that issue becomes to the
decision process. Therefore, we reason that important issues about the
workplace that have direct personal impact on the individual will carry
greater issue intensity than those that have a less direct, less personal
impact. For that reason, employee relations and management of diversity,
having a greater daily effect on the worker, would signal more salient mes
-
sages about life in the firm than other, more distant dimensions. Albinger
and Freeman’s (2000) work on CSP dimensions supported this sugges
-
tion. Their study revealed that firms’ support for diversity and employee
relations was related to positive perceptions of the firm as an employer.
Second, social identity theory supports the importance of treatment of
the environment, the community, and product issues to the job seeker.
Workers derive a sense of self-identity and self-esteem from their associa
-
tion with their employer (Crocker & Luhtanen, 1990). Furthermore,
Dutton and colleagues’ (1991) study of the Port Authority of New York
and New Jersey revealed that individual employees were personally
affected by the negative image of their employer with regard to a particular
issue, namely, the treatment of the homeless community. The poor image
of the organization was transferred to the individual organization mem-
bers (Dutton et al., 1991). This work suggests that behaviors of the firm
that are visible to the public affect self-identity of the worker. Thus, we

assert that they would prefer to be associated with firms that contribute
positively to the environment and the community and produce quality
goods. Based on this, we propose:
Hypothesis 2: Job seekers consider the CSP dimensions of environment, com
-
munity relations, diversity, product issues, and employee relations more
important than the other dimensions, which include non-U.S. operations,
alcohol, tobacco, gambling, military contracting, and nuclear power.
Having argued that CSP is important to job seekers, and also having
proposed that certain aspects of CSP are more important than others, we
now turn to the area of individual differences. The following hypotheses
propose that certain aspects of CSP are more important to some groups of
job seekers than to others.
Studies indicate that women and minorities place greater value on
organizational diversity efforts than do White men (Kossek & Zonia,
1993; Thomas & Wise, 1999; Williams & Bauer, 1994). Furthermore,
Greening and Turban (2000) revealed that gender moderates the relation
-
ship between treatment of women and minorities and organizational
298 BUSINESS & SOCIETY / September 2002
at Harvard Libraries on August 17, 2012bas.sagepub.comDownloaded from
attraction under certain circumstances. On this basis, we hypothesize that
female and minority job seekers may use data about organizational treat
-
ment of women and minorities as a signal of potential working conditions
for themselves. A poor record of treatment of minority or female employ
-
ees may discourage such candidates from goingforward to apply or accept
a position. Thus, we propose the following:
Hypothesis 3: Female job seekers will rate the dimension of diversity as more

important than male job seekers.
Hypothesis 4: Minority job seekers will rate the dimension of diversity as more
important than nonminority job seekers.
Knowledge of CSP
As we have discussed, we believe that CSP may be an important ele
-
ment in recruitment. In terms of attracting potential employees, it is
important for organizations to convey specific information (Barber, 1998)
as well as information that is personally relevant to the job seeker
(Chaiken & Stangor, 1987). For example, studies show that decentralized
organization structure, allowing participative decision making, has a posi-
tive effect on applicant attraction (Turban & Keon, 1993). Reward struc-
ture has been shown to signal career growth opportunities and impact
attraction (Thompson, 1967).
Similarly, there has been increasing focus on organizational image and
its effects on applicant attraction. The concept of employer branding sug-
gests that firms must create a desirable employment image and convey
that to prospective candidates (Engaging Employees Through Your Brand,
2001). Organizational values are an integral part of the employment
image, as they convey signals to candidates indicating what it will be like
to work for the company. Second, organizational values and behaviors
create an organizational identity, and that identity is central to the individ
-
ual’s identity formation after they join the organization (Hatch & Schultz,
1997). CSP, which is a value-based construct, communicates a wealth of
information to the job seeker about the organization, and that information
can be used in the job-choice process. Thus, we argue that knowledge of
CSP has an affect on job choice.
Hypothesis 5: Firms’ CSP ratings will influence job seekers’ assessment of
employer attractiveness.

CSP can be conceptualized as a singular attribute, as we did for
Hypothesis 5, or could be seen as multiple attributes. Waddock and Graves
Backhaus et al. / CSP AND EMPLOYER ATTRACTIVENESS 299
at Harvard Libraries on August 17, 2012bas.sagepub.comDownloaded from
(1997) suggested that we can understand CSP’s functions better when we
examine them on a dimension-by-dimension basis. In our next set of
hypotheses, we build on previous findings indicating that employee rela
-
tions, product quality, concern for the natural environment, and treatment
of women and minorities appear to be more important than community
relations (Greening & Turban, 2000). We begin with this hypothesis and
continue with more specific propositions.
Hypothesis 6: Individual dimensions of CSP will have differential effects on
job seekers’ assessment of employer attractiveness.
Of the five dimensions under consideration, employee relations hits
closest to home when it comes to effects on workers. Job seekers, espe
-
cially college students, demonstrate concern for growth opportunities,
part of the employee relations dimension (Jeffords, Scheidt, &
Thibadoux, 2000). A company’s positive or negative record in terms of
treatment of employees sends a clear signal to potential employees about
the desirability of that employer (Albinger & Freeman, 2000). Therefore,
we hypothesize the following:
Hypothesis 7: Employee relations will be most influential of the five dimen-
sions on assessment of employer attractiveness.
Product issues, although seemingly less central to an employee’s life,
has been shown to be important to prospective employees. In a study of the
fast-food industry, Highhouse, Zickar, Thorsteinson, Stierwalt, &
Slaugher (1999) found that product image was one of the most important
indicators of employer image. Social identity theory suggests that an indi

-
vidual would derive a more positive sense of self-identity from association
with high-quality or prestigious products. Generalizing from this, we
hypothesize that product issues will be the second most influential of the
five dimensions, after employee relations.
Hypothesis 8: Product issues will rank second among the dimensions in its
effect on assessments of employer attractiveness.
METHOD
Procedure
This was a two-part study. In Part 1, data were collected from 297
undergraduate business students as an in-class exercise. The survey was
300 BUSINESS & SOCIETY / September 2002
at Harvard Libraries on August 17, 2012bas.sagepub.comDownloaded from
administered in upper-division accounting and management courses. Dis
-
cussion of CSP was not part of the curriculum of these classes. Almost all
students (93%) indicated that they were involved in a job search or were
beginning to gather information about employers. Most respondents were
juniors/seniors (86%), with a median age of 21. Of respondents, 50%
reported their ethnicity as White, 37% as minority, and 13% declined to
respond to the ethnicity item, and 52% were female.
Participants completed a questionnaire that first asked them to indicate
the importance of CSP in their job choice process. They reported the
importance of CSP at four points in the job-choice process, when consid
-
ering the attractiveness of a firm as a potential employer in general; when
determining positions for which to submit a résumé; when considering
whether to interview with a company; and when considering whether to
accept a job offer. Participants rated each of these using a 5-point Likert-
type scale, from 5 (very important to1(unimportant) (Cable & Judge,

1996). We were interested in determining the stages at which job seekers
consider CSP and also wanted to determine an overall score indicating
their general concern for CSP, which would be called importance of CSP.
The second section of the questionnaire asked them about the relative
importance of 11 dimensions of CSP. Participants were provided with def-
initions of each of the dimensions of CSP,drawn from the KLD definitions
to ensure that each student was rating the same construct. Again, partici-
pants used the same 5-point Likert-type scale. This phase of the study was
completed 2 months prior to beginning the next phase of the study.
Part 2 of the study examined the effects of CSP knowledge on attrac-
tiveness ratings of firms. The same sample was used for this stage. We
used a quasi-experimental design, in which participants were first asked to
rate the attractiveness of an organization using only what they already
knew about the company. After 2 weeks, students were presented with the
same list of companies but were provided with CSP data about each orga
-
nization and were told to add this information to what they already knew
about the company and rate its attractiveness. The data described the
firms’ records in 5 areas of CSP—environment, community relations,
diversity, product issues, and employee relations. We selected these 5
areas based on previous studies (Berman et al., 1999; Waddock & Graves,
1997) and from our findings from our Part 1, which supported the hypoth
-
esis that these 5 areas are the more important elements of a group of 11
dimensions. In both parts of the experiment, participants used a 6-point
Likert-type scale, ranging from 5 (very good)to1(very poor)to0(cannot
judge).
We used policy capturing (Slovic & Lichtenstein, 1971) to examine the
extent to which each CSP dimension affected the participants’ perception
Backhaus et al. / CSP AND EMPLOYER ATTRACTIVENESS 301

at Harvard Libraries on August 17, 2012bas.sagepub.comDownloaded from
of the organization. This technique has been frequently used to examine
individual differences in decision processes (Cable & Judge, 1994;
Graves & Karren, 1992; Klaas & Wheeler, 1990). Using this method, we
can systematically vary the CSP conditions to produce a large set of com
-
binations. Because we were using five dimensions of CSP and we pro
-
vided three conditions for each dimension, there were a possible 243 com
-
binations. We asked participants to make judgments on 50 of these
possible combinations. The 50 profiles were randomly selected from the
complete set.
To test for bias in the questionnaire design, we conducted a test-retest.
A control group of 80 students, members of another section of an upper-
division business course completed questionnaire one, which asked them
to rate the attractiveness of the same organizations, using the same scale as
the experimental group. Two weeks later, the students filled out the ques
-
tionnaire again but did not receive any additional information. In the
retest, part of the group received a questionnaire that listed the firms in a
different order to test for order effects. The overall multivariate tests of
significance are not statistically significant, indicating that there is no sig-
nificant difference between pretest and posttest responses when no new
information is presented (F = 1.47, p = .26). Furthermore, it also elimi-
nates the question of order effects within the questionnaire.
Measures
CSP data. For each organization they were asked to assess, participants
received ratings for each of the five CSP dimensions. As we indicated, we
used the KLD dimensions and definitions to follow mainstream CSP

research, but for methodological reasons, we modified their rating scale.
KLD assigns a value on a 5-step rating scale running from –2 (major
weakness)to+2(major strength) for each dimension. Because we
believed that the distinctions between points on a 5-point scale would be
difficult for a participant to interpret, we broadened the scale to three
points (1 = poor performance, 2 = neutral performance, and 3 = good per
-
formance). This simplified the construction of the combinations and
reduced the effect of bias in respondent interpretation. Furthermore, in
contrast to Greening and Turban (2000), we chose to use numerical rather
than verbal descriptions to avoid semantic biases that may be introduced
with narratives.
The sample of 50 organizations was drawn from the Fortune 500 using
the following two criteria: the organization had to be generally familiar to
the researchers and the group of 50 had to represent a broad range of
302 BUSINESS & SOCIETY / September 2002
at Harvard Libraries on August 17, 2012bas.sagepub.comDownloaded from
industries. The companies were listed on the survey instrument in alpha
-
betical order.
Control variables. Gender and student status were also measured, as
these have been used as control variables in previous studies (Greening &
Turban, 2000). We included these variables in our analysis of the pretest
and posttest data in Part 2 of the study.
RESULTS
Importance of CSP
In Part 1 of the study, we determined that a mean response that fell sta
-
tistically significantly above the midpoint of the scale, which is labeled
“important,” would indicate support for the hypothesis. To test this

hypothesis, we conducted one-sample t tests, using 3.5 as a test value.
Although all four items garnered a mean response of 3.5 or greater, Item 1
(when considering attractiveness in general), and Item 4 (when consider-
ing whether to accept a job offer) were significantly above 3.5. Item 1 had
a mean of 3.66 (t = 2.53, df = 293, p = .012), and Item 4 had a mean of 3.82
(t = 4.9, df = 291, p = .00).
We also assessed the importance of CSP in the job-search process by
creating a composite score, which is the sum of the four items, importance
of CSP in general, when determining positions for which to submit a
résumé, when considering whether to interview, and when considering
whether to accept a job offer. This new item called “importance of CSP,”
has a strong scale reliability of Cronbach’s alpha = .86. The high correla
-
tion among the four items suggests that they are tapping a single job-
search criteria, thus we combined them into a single measure.
Importance of CSP had a mean response of 14.59 (SD = 3.62), which
was significantly higher than the midpoint of the scale, again signifying a
level of importance that exceeds “important” (t = 2.776, df = 291, p =
.006). Thus, Hypothesis 1 is supported (see Table 1).
Hypothesis 2 proposes that job seekers consider the CSP dimensions of
environment, community relations, diversity, product issues, and
employee relations more important than other commonly used dimen
-
sions of CSP when assessing employer attractiveness. Our findings indi
-
cate support for this hypothesis. Analysis of variance revealed that the
means of the 11 dimensions differed significantly (F = 84.05, p <.000).
The 5 dimensions named in the hypothesis are indeed the 5 top-rated
dimensions. Post hoc analysis confirms that these 5 dimensions are
Backhaus et al. / CSP AND EMPLOYER ATTRACTIVENESS 303

at Harvard Libraries on August 17, 2012bas.sagepub.comDownloaded from
304
Table 1
Correlations and Descriptive Statistics
Item MSD1234 5 67 89101112131415161718
1. General 3.66 1.08
2. Résumé 3.60 1.06 .64**
3. Interview 3.50 1.05 .54** .75**
4. Job offer 3.82 1.10 .62** .57** .55**
5. Overall 14.59 3.62 .83** .88** .84** .82**
6. Age 22.59 5.10 .01 –.06 –.03 –.06 –.04
7. Gender (1 = male) 1.54 .50 .16** .15** .21** .19** .21** .09
8. Ethnicity
(1 = nonminority) 1.57 .50 –.13* –.10* –.05 –.08 –.11 .13* .00
9. Employee relations 4.49 .70 .28** .32** .27** .31** .34** .06 .14* .04
10.Product issues 3.84 .95 .37** .37** .44** .44** .48** .04 .24** .04 .40**
11. Environment 3.82 1.04 .35** .38** .38** .31** .43** .12 .15** –.07 .26** .30**
12. Diversity 3.81 1.30 .35** .33** .36** .31** .40** .02 .23** –.32** .26** .34** .24**
13. Community relations 3.73 1.05 .37** .40** .42** .34** .46** .05 .15* –.25** .30** .47** .55** .44**
14. Non-U.S. operations 3.18 1.19 .37** .37** .29** .32** .40** .05 .22** .09 .28** .49** .29** .49** .34**
15. Nuclear power 2.90 1.56 .16** .28** .20** .15** .22** –.11 .25** .00 .10 .35** .18** .17** .16** .28**
16. Tobacco 2.71 1.63 .22** .28** .29** .18** .29** .01 .19** –.09 .05 .32** .20** .22** .22** .25** .46**
17. Gambling 2.63 1.62 .23** .28** .31** .15** .29** –.01 .17** –.23** .11 .35** .24** .29** .30** .30** .46** .68**
18. Alcohol 2.48 1.59 .25** .32** .31** .19** .32** –.02 .21** –.20 .03 .31** .23** .28** .30** .29** .36** .69** .70**
19. Military 2.35 1.47 .11** .27** .31** .11** .21** –.07 .09 –.08 .10 .32** .24** .18** .25** .23** .59** .42** .52** .44**
*p < .05. **p < .01.
at Harvard Libraries on August 17, 2012bas.sagepub.comDownloaded from
significantly higher than the other 6 dimensions. Table 2 indicates the
results of t tests among each of the dimensions, demonstrating the differ
-

ences among the means. Hypothesis 2 is supported.
Hypotheses 3 and 4 predict that women and minorities place more
importance on the diversity dimension than men or nonminorities.
Women gave significantly higher ratings to diversity than men, thus sup
-
porting Hypothesis 3. Similarly, minority participants rated diversity sig
-
nificantly higher than nonminorities did. Hypothesis 4 is supported (see
Table 3).
Backhaus et al. / CSP AND EMPLOYER ATTRACTIVENESS 305
Table 2
CSP Dimensions and p values for Tests of Differences between Mean Ratings
Dimension MSD 12 34 567891011
1. Employee relations4.49 .70
2. Product issues 3.84 .95 .00
3. Environment 3.82 1.04 .00 .73
4. Diversity 3.81 1.30 .00 .63 .87
5. Community
relations 3.73 1.05 .00 .06 .086 .22
6. Non-U.S.
operations 3.18 1.19 .00 .00 .00 .00 .00
7. Nuclear power 2.90 1.56 .00 .00 .00 .00 .00 .00
8. Tobacco 2.71 1.63 .00 .00 .00 .00 .00 .00 .06
9. Gambling 2.63 1.62 .00 .00 .00 .00 .00 .00 .00 .26
10. Alcohol 2.48 1.59 .00 .00 .00 .00 .00 .00 .00 .00 .03
11. Military 2.35 1.47 .00 .00 .00 .00 .00 .00 .00 .00 .00 .22
Note: Within-groups df = 3,187;F= 84.05; p <.000. Values on the diagonals are p values for t
tests between mean ratings of each dimension.
Table 3
ANOVA, Diversity Dimension by Sex and Ethnicity

Description MSDdf FSignificance
Minority 4.27 .95 255 27.88 .00**
Nonminority 3.46 1.38
Male 3.50 1.39 266 14.89 .00**
Female 4.10 1.12
Minority males 4.15 .95 117 23.35 .00**
Nonminority males 3.02 1.45
Minority females 4.37 .95 137 8.06 .00**
Nonminority females 3.84 1.21
**p < .00.
at Harvard Libraries on August 17, 2012bas.sagepub.comDownloaded from
Effects of Knowledge of CSP
In Part 1, we confirmed previous findings suggesting that CSP is
important to job seekers. Now we move to the effect of knowledge of CSP
records on assessments of organizational attractiveness. This portion of
the study yielded 110 usable, matched pretest and posttest surveys. To
analyze our data, we fit a mixed-effects linear model (Laird & Ware, 1982;
Pinheiro & Bates, 2000). The model is fixed because we are concerned
with inferences about the specific means in the study (the company rat
-
ings), and random, as we are also concerned about the means of the ratings
for a random sample of students (Dielman, 1996). We use the mixed
effects to account for the covariance structure introduced by the repeated
measurements (the pre-post test design). The results are displayed in
Table 4, the analysis of variance table, and in Table 5, attractiveness rat
-
ings at Times 1 and 2.
In addressing the next set of hypotheses, we first classified each partici-
pant in the experiment by two between-participant factors, the partici-
pant’s gender and the participant’s student status (freshmen/sophomore or

junior/senior), to identify any differences related to these characteristics.
In our analysis, we would expect to find differences on between-partici-
pant factors, if any exist, because of the reasonably large number of partic-
ipants. In spite of the large number of denominator degrees of freedom,
neither of the between-participant factors, gender, nor student status had
significantly different levels, which indicates that these personal charac-
teristics had little effect on the ratings of the corporations at either Time 1
or Time 2.
Within-participant differences are almost surely to be judged signifi
-
cant because of the extremely large number of pre-post correlated pair-
wise comparisons. Because each test of within-participant differences has
8,730 denominator degrees of freedom, we needed to examine the magni
-
tude of the estimates of these effects to shed light on these two hypotheses.
We assume that the effects of knowledge of CSP dimensions would
become evident in the time-dimension interactions (Time 1 is pretest,
Time 2 is posttest); for example, we should recognize the effect of a time-
environment or a time-community relations interaction.
Due to the extremely large number of denominator degrees of freedom
associated with the within-participant main effects, all effects are judged
to be significantly different from zero, but the only main effect with mean
-
ingful magnitude is time. As participants were told of the CSP records of
each company, their rather favorable opinions based on name recognition
declined on the average of 1.67 points on the 6-point scale. This finding
supports Hypothesis 5, that knowledge of firms’ CSP records will
306 BUSINESS & SOCIETY / September 2002
at Harvard Libraries on August 17, 2012bas.sagepub.comDownloaded from
influence assessment of the firms. It is important to note that participants’

Time 1 scores for many companies were high, very often in the highest
category. In these instances, only two options remained at Time 2—score
the same or score lower. Thus, we see the average company score decline
from Time 1 to Time 2. Our next question, however, was, How did the CSP
scores affect ratings from Time 1 to Time 2?
After participants received the CSP ratings, the estimates of the inter
-
action parameters became considerably larger. As hypothesized, the 5
dimensions of CSP did have differential effects on participants’ assess
-
ment of the firms. The largest changes in ratings are associated with the
environment, diversity, and community relations dimensions. The affect
Backhaus et al. / CSP AND EMPLOYER ATTRACTIVENESS 307
Table 4
Analysis of Variance Table, Effects of Corporate Social Performance Knowledge
Numerator Denominator Fp
Predictor df df Value Value
Intercept 1 8,730 7,174.15 < .001
Main effects
Gender 1 95 1.66 0.20
Student status 1 95 0.05 0.83
Time 1 8,730 167.89 < .001
Environment 2 8,730 93.04 < .001
Community relations 2 8,730 2.13 0.12
Diversity 2 8,730 99.73 < .001
Product issues 2 8,730 84.90 < .001
Employee relations 2 8,730 27.58 < .001
Interactions
Time: environment 2 8,730 40.96 < .001
Time: diversity 2 8,730 35.70 < .001

Time: community relations 2 8,730 28.41 < .001
Time: product issues 2 8,730 24.98 < .001
Time: employee relations 2 8,730 15.74 < .001
Environment: community relations 4 4,859 10.88 < .001
Environment: employee relations 2 4,859 2.48 .08
Environment: diversity 3 4,859 .75 .52
Environment: product issues 3 4,859 3.10 .02
Community relations: employee
relations 4 4,859 3.69 .005
Community relations: diversity 4 4,859 8.25 < .001
Community relations: product issues 3 4,859 5.03 .001
Employee relations: diversity 2 4,859 .40 .67
Employee relations: product issues 2 4,859 5.83 .003
Diversity: product issues 3 4.859 1.28 .28
at Harvard Libraries on August 17, 2012bas.sagepub.comDownloaded from
308 BUSINESS & SOCIETY / September 2002
Table 5
Employer Attractiveness Ratings at Time 1 and Time 2
Time 1 Time 1 Time 2 Time 2
Company Name Mean Rating SD Mean Rating SD
American Express 3.85 1.20 3.57 .95
Anheuser-Busch 2.47 1.82 2.64 1.12
AT&T 3.73 1.04 2.05 1.37
Bell Atlantic 3.59 1.32 2.81 1.05
Boeing 1.87 1.89 2.86 1.07
Cisco Systems 2.15 2.10 3.65 1.21
Coca-Cola 4.12 .97 3.37 1.12
Compaq Computer 3.64 1.21 3.30 1.03
CVS 3.37 1.31 2.22 .97
Dell Computer 3.90 1.39 3.00 1.06

Delta Air Lines 3.21 1.24 2.64 1.08
Dow Chemical 1.71 1.78 2.38 1.04
DuPont 2.14 1.92 2.60 1.10
Eastman Kodak 3.42 1.63 2.89 .97
Exxon 2.96 1.35 2.97 1.05
FedEx 4.11 .90 2.93 .96
Ford Motor Company 3.45 1.38 3.33 1.42
Gap 3.83 1.25 2.66 1.11
General Electric 3.82 1.23 2.95 1.03
General Motors 3.71 1.42 2.92 1.22
Goodyear 3.28 1.64 3.52 .97
Hewlett-Packard 3.52 1.60 2.79 .95
Home Depot 3.73 1.30 3.40 1.18
Intel 3.90 1.52 3.69 .94
IBM 3.97 1.17 3.34 1.05
International Paper 1.45 1.81 2.83 1.31
JC Penney 3.32 1.23 2.82 .93
Johnson & Johnson 3.81 1.40 3.64 1.29
Kmart 3.46 .97 3.00 1.06
Lucent 2.32 2.00 3.24 1.04
McDonalds 3.91 1.14 3.39 .95
MCI WorldCom 3.36 1.40 2.80 1.01
Merrill Lynch 3.03 1.83 3.54 1.06
Metropolitan Insurance 2.54 1.78 3.12 1.01
Mobil 3.66 1.22 2.78 1.04
Motorola 3.79 1.27 2.79 1.07
PepsiCo 4.12 1.14 2.97 1.04
Philip Morris 2.31 1.71 2.80 1.06
Procter & Gamble 2.47 1.87 2.67 1.04
RJR Nabisco 2.85 1.75 3.20 1.073

Sara Lee 2.66 1.88 2.55 .96
Sears Roebuck 3.46 1.38 2.73 1.05
Sprint 3.38 1.25 2.75 .96
Sun Microsystems 1.63 1.94 2.62 1.13
at Harvard Libraries on August 17, 2012bas.sagepub.comDownloaded from
of product issues is less, whereas that of employee relations the least.
Therefore, Hypotheses 7 and 8 are not supported.
The analysis also revealed interesting patterns in the way CSP dimen
-
sion ratings affected firm ratings. As might be anticipated, participants
gave more favorable ratings to firms that had higher CSP ratings. CSP rat-
ings of 1, defined as poor, resulted in the lowest firm ratings. For all
dimensions, a CSP rating of 2, defined as neutral, resulted in significantly
higher firm ratings. This difference was quite marked, especially in the
area of environment. The difference in effect between a CSP rating of 2
and 3, good, again was significant and is particularly dramatic in the area
of community relations and diversity. The difference between a neutral
and good rating in the areas of environment and employee relations had
less impact. In the product issues dimension, a rating of neutral resulted in
a better overall firm rating than the good rating. This was the only dimen-
sion that had an outcome that deviated from the expected order. We next
describe our interpretations of these findings in the following Discussion
section.
DISCUSSION
This study continues the exploration of the role of CSP in organiza
-
tional attractiveness that was started by Turban and Greening (1997) and
Greening and Turban (2000). Reporting the results of two studies, we first
confirm that potential job seekers consider CSP important to the overall
assessment of a company. We find that job seekers consider CSP records

important at all stages of the job search, but most important when deter
-
mining whether to take a job offer. Also, as anticipated, job seekers find
some aspects of CSP more relevant than others. Namely, environment,
community relations, employee relations, diversity, and product issues
outrank 6 other categories of CSP. The study also reveals that women
Backhaus et al. / CSP AND EMPLOYER ATTRACTIVENESS 309
Texaco 2.79 1.39 2.36 1.04
Toy R Us 3.80 1.07 3.05 1.03
UPS 3.51 1.69 3.57 .95
Wal-Mart 3.71 1.21 3.62 .88
Walt Disney 4.25 1.28 3.23 1.15
Xerox 3.59 1.45 2.75 1.11
Table 5 continued
Time 1 Time 1 Time 2 Time 2
Company Name Mean Rating SD Mean Rating SD
at Harvard Libraries on August 17, 2012bas.sagepub.comDownloaded from
express greater interest than men in a firm’s diversity record, and simi
-
larly, minorities express greater interest than nonminorities. As we
explored the data, we were also interested to see how gender and minority
status affected overall concern for CSP. We found that women report a sig
-
nificantly greater concern for CSP than men. Minorities report a slightly
greater concern than nonminorities, but not significantly so. We suggest
that signaling theory may help to explain these findings. The presence of
good social performance may indicate to job seekers that a firm has posi
-
tive values. Because historically, both women and minorities have faced
barriers to entry and career growth in corporations, they may be particu

-
larly sensitive to any attributes that suggest concern for the betterment of
society or consciousness of societal needs. The presence of a positive CSP
record may suggest a better and more welcoming environment. Research
in the area of corporate social orientation, a construct that measures the
perceived importance of a firm’s responsibility to society, also suggests
that women and minorities are more philanthropically, and less economi-
cally oriented than nonminority men (Edmondson & Carroll, 1999;
Ibrahim & Angelidis, 1994). With the results of the first study as a founda-
tion, our second study investigates how knowledge of CSP affects opin-
ions about organizational reputation. Before they were presented with
CSP data, participants’ initial ratings of the organizations were high—the
top 10 mean ratings were greater than 4.0 on a 5-point scale. We investi-
gated a number of factors to try to explain this high initial rating effect.
First, we looked for a relationship between ratings and firm size but
found no apparent relationship. Using Turban and Greening (1997) as a
foundation, we also did post hoc tests to determine whether initial ratings
were related to actual CSP. Using data from Fortune magazine’s ratings of
CSP, there was no significant correlation between CSP and our partici
-
pants’ initial ratings of the firms (r = .06, p = .66). We also grouped the 50
firms by industry, and found only two noteworthy rating trends. Computer/
electronics/telecommunications tended to be rated higher overall (m =
4.00) than any other major category, whereas the petroleum industry was
rated lower (m = 3.39). The difference between these two ratings was sig
-
nificant (p = .01).
Familiarity with the firms presents a second possible explanation for
the high ratings. We selected firms for inclusion in the study on the basis of
their general familiarity to us. In their study of organizational attractive

-
ness, Gatewood and colleagues (1993) found a high correlation between
familiarity with the firm and ratings of firm image. Their findings may
help explain the unusually high ratings given by our participants to a group
of well-known organizations.
310 BUSINESS & SOCIETY / September 2002
at Harvard Libraries on August 17, 2012bas.sagepub.comDownloaded from
As discussed earlier, after receiving information about CSP, participant
ratings of organizational attractiveness dropped significantly. Part of this
drop is explained by the fact that initial ratings were so high (i.e., that there
was no place to go but down). But a significant portion of the change in rat
-
ings related to the CSP data students’ received. Participants gave higher
ratings to organizations with higher CSP scores and lower ratings to those
with poorer CSP scores. To illustrate this in another way, we conducted a
between-participants analysis. We summed the ratings of the five dimen
-
sions to calculate the overall CSP score assigned to each firm. This overall
score was highly correlated (–.77, significant at .01 level) with the net
change in ratings from Time 1 to Time 2. The lower the CSP score, the
greater the reduction in ratings. Therefore, it is clear that there was a
change in firm ratings at Time 2, and that change relates to knowledge of
CSP data.
We expected to find a change in ratings, but the more interesting ques
-
tion is, How do the specific CSP dimensions work to create this change?
Or, which dimensions seem to impress job seekers the most? Variations in
CSP ratings, from poor to good, resulted in different reactions among the
participants. For example, in the environment dimension, there was a
notable difference between the effects of a poor rating and a neutral rating.

When presented with an organization with a poor environmental rating,
reactions were quite strongly negative. However, there was little differ-
ence between a neutral and a good environmental rating. The employee
relations dimension displayed the same effect, although to a lesser extent.
Poor employee relations result in a serious downgrading of an organiza-
tion, whereas neutral and good ratings are perceived in much the same
way as one another. For these areas, it appears that as long as the record is
not poor, it is considered acceptable. A poor rating may provide a warning
to the job seeker that the organization has a negative image.
The areas of diversity and community relations work in a somewhat
different way. There was a large gap in ratings between organizations with
neutral CSP data and good CSP data. In the case of these dimensions, it
seems as if an organization has to be able to demonstrate positive accom
-
plishments to prove itself. It appears from these findings that job seekers
set a higher benchmark for employer performance in the areas of diversity
and community relations than they do for environment and employee
relations.
The interactions among pairs of dimensions also revealed some inter
-
esting patterns. Using analysis of variance, we examined interactions
among each pair of dimensions. First, firms with a low employee relations
score garnered low attractiveness ratings even when other factors were
Backhaus et al. / CSP AND EMPLOYER ATTRACTIVENESS 311
at Harvard Libraries on August 17, 2012bas.sagepub.comDownloaded from
high. For example, the mean rating for all firms with the lowest employee
relations score was 2.9. However, when paired with a high community
relations score or a high product issues score, the mean remains at 2.9, (p <
.00 for both sets of interactions. Interactions between employee relations
and the other dimensions were not statistically significant). When com

-
munity relations is scored low and employee relations is scored high, there
is a large jump in attractiveness ratings (from m = 2.9 to m = 3.2, p < .00).
However, that same effect does not take place with other dimensions.
When community relations is low and environment is high, the mean
attractiveness rating does not change.
When we look at combinations of low scores, however, we see some
different trends. When community relations and environment are both
scored low, the mean attractiveness rating drops to 2.5. However, when
environment and product issues have low scores, we reach the lowest of all
interaction means, 2.4. Although we can offer no clear explanation of this,
it suggests that there is a level of tolerance for organizational shortcom-
ings among job seekers. Although a low rating in one area might have only
a small effect on overall impressions of the firm, combinations of low rat-
ings lead to a sharp drop-off in organizational attractiveness. It appears
that a combination of low ratings, even among dimensions that are not
foremost in the job seekers’priorities, contribute to the development of an
overall negative image. Hence, even a positive record in one area, such as
environment or community relations, may not be enough to offset the neg-
ative image conveyed by other, multiple weaknesses.
Analysis of the two-way interactions presents data that stand in con-
trast to the main effects. When we examine main effects, we see that diver
-
sity and environment most strongly affect participants’ ratings of
employer attractiveness. Yet, when we view interactions among dimen
-
sions, we see that employee relations plays a large role in raising ratings
when the firm is performing poorly in another dimension. It is also inter
-
esting to compare this to Part 1 of the study, in which participants reported

that employee relations is most important to them in assessing an organi
-
zation’s attractiveness. Additional research should be conducted to con
-
tinue to understand how the dimensions interact and how job seekers value
the dimensions differently.
Overall, our results indicate that potential job seekers are interested
and concerned about a firm’s record of corporate social performance
when considering firms as prospective employers. They also indicate that
knowledge of CSP data has an effect on the way in which these prospec
-
tive employees view the reputation and attractiveness of firms. This study
also indicates that certain dimensions of CSP—environment, community
relations, employee relations, diversity, and product issues—are more
312 BUSINESS & SOCIETY / September 2002
at Harvard Libraries on August 17, 2012bas.sagepub.comDownloaded from
important than other dimensions, and have differing effects on prospec
-
tive employees’assessments of firms. Specifically, firms with a poor envi
-
ronmental record are likely to encounter hurdles in recruitment, as job
seekers are critical of weakness in that dimension. On the flip side, having
a good environmental record adds little to the attractiveness of the organi
-
zation. Similarly, demonstrating positive employee relations adds little,
but being guilty of poor employee relations creates a negative recruitment
image. Finally, the presence of multiple weaknesses creates a measurable
image problem for the firm.
Our results are also somewhat surprising, in that issues that would
seem to be most salient to job seekers, those issues that have a direct rela

-
tionship to their daily lives, were less important than broader issues.
Again, this may be answered by social identity theory. Job seekers may be
concerned about being associated with firms with poor environmental
records or poor treatment of the community. Conversely, workers may
derive a positive sense of identity from association with a firm that does
good things for the neighborhood and the world.
Our findings support the need for image management among recruit-
ing organizations. Image management refers to the attempts of an organi-
zation to construct positive perceptions of itself to stakeholders (Griffin,
1999). First, our study supports the growing movement that recognizes the
importance of corporate social responsibility to the stakeholder group that
includes prospective employees. This is a group that has not been given
sufficient attention in the CSP stakeholder literature (see Harrison & Free-
man, 1999). Second, our study provides preliminary data indicating that
these stakeholders use information about the various dimensions of CSP
in different ways. Environment, community relations, and diversity rat
-
ings create the largest change in opinions about the firm, with employee
relations and product issues having a lesser effect. There are also differing
levels of tolerance for poor, neutral, and good performance among the
dimensions. Although it appears sufficient for a firm to have a neutral
environmental record, job seekers have higher expectations when it comes
to community relations and employee relations. Good ratings in those
areas provide an edge for firms in the eyes of job seekers. Moreover,
women and minorities are more critical of a firm’s CSP record than
nonminority men, and express greater concern about these records.
This study may be limited in generalizability because of the nature of
the sample. Although upper-division students are beginning to engage in
the job-search process, another study utilizing a sample of active job seek

-
ers is necessary to test the limits of our findings. Future studies might
broaden the sample of job seekers to include individuals at different career
stages, with different skills, education and experience.
Backhaus et al. / CSP AND EMPLOYER ATTRACTIVENESS 313
at Harvard Libraries on August 17, 2012bas.sagepub.comDownloaded from
Additional research is needed to explore further the effects of CSP.
This study analyzed the response to CSP data in a hypothetical situation.
Researchers might investigate the way in which actual job candidates
react to CSP data presented to them during the job search process and how
this data influences their job choices. It would also be interesting to
explore the relative importance of CSP data compared to other organiza
-
tional attributes. In other words, does CSP matter in the job-choice pro
-
cess after we introduce other job-choice factors such as salary, working
conditions, and opportunities for promotion? The results of our study
prompt more questions about the importance of the various dimensions of
CSP. Our study suggests that positive employee relations is valued highly
and can offset a poor record in community relations. We also found that a
neutral or positive environmental record has little effect, but a negative
environmental record can result in a negative assessment of the firm.
Future research might continue to explore these variable effects and the
interaction of positive and negative ratings. Furthermore, it would also be
helpful to explore how individual differences mediate the relationship
between a firms’ CSP record and organizational attractiveness.
In our study, we provided participants with CSP data and asked them to
use this to assess the attractiveness of employers. We did not investigate
how job seekers search out and screen CSP data or how firms market CSP
data as part of the recruitment process. Furthermore, we focused on famil-

iar, Fortune 500 firms. Researchers might explore how job seekers gather
CSP data on smaller, less familiar firms, and how that information is used
in the job-choice process. Studies of actual behavior, rather than inten-
tions, may also avoid any potential for social desirability bias, which is
introduced when we ask people to report their reactions to scenarios that
might be perceived as socially undesirable.
CONCLUSIONS
Our study supports and advances the findings of Greening and Turban
(2000), suggesting that CSP is an important attribute to job seekers.
Accordingly, firms may find it effective to incorporate CSP information
into their recruitment efforts. In addition, although there are a multitude of
themes ordinarily considered under the CSP umbrella, our research sug
-
gests that firms may be well advised to concentrate their efforts on certain
dimensions of CSP and focus on the differing values of diverse job candi
-
dates as they present their corporate image.
314 BUSINESS & SOCIETY / September 2002
at Harvard Libraries on August 17, 2012bas.sagepub.comDownloaded from
REFERENCES
Albinger, H. S., & Freeman, S. J. (2000). Corporate social performance and attractiveness as
an employer to different job seeking populations. Journal of Business Ethics, 28, 243-
253.
Alexander, G., & Buchholz, A. (1982). Corporate social responsibility and stock market per
-
formance. Academy of Management Journal, 21, 479-486.
Aupperle, K. E., Carroll, A. B., & Hatfield, J. D. (1985). An empirical examination of the
relationship between corporate social responsibility and profitability. Academy of Man
-
agement Journal, 28, 446-463.

Barber, A. (1998). Recruiting employees. Thousand Oaks, CA: Sage.
Bauer, T. N., & Aiman-Smith, L. (1996). Green career choices, the influences of ecological
stance on recruiting. Journal of Business and Psychology, 10, 445-458.
Belt, J. A., & Paolillo, J. G. P. (1982). The influence of corporate image and specificity of
candidate qualifications on response to recruitment advertisement. Journal of Manage
-
ment, 8, 105-112.
Berman, S., Wicks, A. C., Kotha, S., & Jones, T. (1999). Does stakeholder orientation mat
-
ter? The relationship between stakeholder management models and firm financial per
-
formance. Academy of Management Journal, 42, 488-505.
Bretz, R. D., Ash, R. A., & Dreher, G. F. (1989). Do people make the place? An examination
of the attraction-selection-attrition hypothesis. Personnel Psychology, 42, 561-581.
Cable, D. M., & Judge, T. A. (1994). Pay preferences and job search decisions, a person-
organization fit perspective. Personnel Psychology, 47, 317-349.
Cable, D. M., & Judge, T. A. (1996). Person-organization fit, job choice decisions and orga-
nizational entry. Organizational Behavior and Human Decision Processes, 67,294-311.
Chaiken, S., & Stangor, C. (1987). Attitudes and attitude change. In M. Rosenzweig &
L. W. Porter (Eds.), Annual review of psychology (vol. 38, 575-630). Palo Alto, CA:
Annual Reviews.
Chatman, J. A. (1989). Improving interactional organizational research: A model of person-
organization fit. Academy of Management Review, 14, 333-349.
Chatman, J. A. (1991). Matching people and organizations: Selection and socialization in
public accounting firms. Administrative Science Quarterly, 36, 459-484.
Cochran, P. L., & Wood, R. A. (1984). Corporate social responsibility and financial perfor
-
mance. Academy of Management Journal, 27, 42-56.
Coffey, B. S., & Fryxell, G. (1991). Institutional ownership of stock and dimensions of cor
-

porate social performance: An empirical examination. Journal of Business Ethics, 10,
437-444.
Crocker, J., & Luhtanen, S. R. (1990). Collective self-esteem and in-group bias. Journal of
Personality and Social Psychology, 58, 60-67
Dielman, T. E. (1996). Applied regression analysis for business and economics. Belmont,
CA: Duxbury.
Dutton, J. E., Dukerich, J. M., & Harquail, C. V. (1991). Organizational images and member
identification. Administrative Science Quarterly, 39, 239-263.
Edmondson, V. C., & Carroll, A. B. (1999). Giving back: An examination of the philan
-
thropic motivations, orientation and activities of large Black owned businesses. Journal
of Business Ethics, 19, 171-179.
Engaging employees through your brand. (2001). New York: Conference Board.
Backhaus et al. / CSP AND EMPLOYER ATTRACTIVENESS 315
at Harvard Libraries on August 17, 2012bas.sagepub.comDownloaded from