Académique Documents
Professionnel Documents
Culture Documents
TH
1
Institute of International Business, Stockholm
School of Economics, Stockholm, Sweden;
2
Stockholm School of Economics Russia,
St Petersburg, Russia; 3J. Mack Robinson College
of Business, Georgia State University, Atlanta,
USA; 4Swedish School of Economics, Helsinki,
Finland
Abstract
This paper investigates the extent to which different human resource
management (HRM) practices work better in different countries. We also try
to open the black box between HRM and multinational enterprise (MNE)
subsidiary performance by considering mechanisms through which HRM
practices affect MNE subsidiary performance. The study utilizes a unique data
set consisting of subsidiaries of 241 MNEs operating in Russia, USA, and
Finland. In the partial least-square analysis used to examine our hypotheses, we
demonstrate that different HRM practices are preferable in different countries,
and that motivation and ability are important mediating variables in the HRM
MNE subsidiary performance relationship.
Journal of International Business Studies (2009) 40, 690712.
doi:10.1057/jibs.2008.83
Carl F Fey1,2,
Sergey Morgulis-Yakushev1,2,
Hyeon Jeong Park3 and
Ingmar Bjorkman4
PY
Correspondence:
CF Fey, Institute of International Business,
Stockholm School of Economics, Box 6501,
Stockholm S-11383, Sweden.
Tel: 46 8 736 9450;
Fax: 46 8 31 9927;
E-mail: Carl.Fey@hhs.se
INTRODUCTION
A large body of research has documented that the way in which a
firms human resources are managed is important for its competitiveness. This is especially important, as a firms employees can
provide one of the most important sources of competitive
advantage (e.g., Barney & Wright, 1998; Guest, 1997; Lado &
Wilson, 1994). However, while the importance of choice of human
resource management (HRM) practices is now well accepted, there
is insufficient understanding of how the contribution of these
practices to multinational enterprise (MNE) subsidiary performance may differ across countries. As Newman and Nollen
(1996: 753) stated, Until recently, the dominance of American
management theory led to a belief that one size fits all y that
effective US management practices will be effective elsewhere.
This view is now being seriously questioned. Increasing evidence
suggests that management theory and practice in the US may not
be optimal in other settings (Boyacigiller, Kleinberg, Phillips, &
Sackmann, 2004; Hofstede, 1993) because of cultural (Boyacigiller
et al., 2004; Hofstede, 1993) and institutional (DiMaggio & Powell,
Carl F Fey et al
691
present the studys results, and end with a discussion of the results and conclusions.
PY
THEORETICAL BACKGROUND
A belief that employees performance has critical
implications for a firms performance has been
common among academics and practitioners for
many years (Huselid, 1995). However, interest in
which HRM practices facilitate maximum employee
performance, and thus in turn organization performance, has intensified more recently as scholars
have suggested that collectively a firms employees
can provide a unique source of competitive advantage. The realization that human resources can be a
source of competitive advantage and firm performance has led to expanded research in the field of
strategic HRM. Wright and McMahan (1992: 298)
define strategic HRM as the pattern of planned
human resource deployments and activities
intended to enable an organization to achieve its
goals. Because firm performance stands out as a
major organizational goal, developing a better
understanding of the relationship between HRM
practices and MNE subsidiary performance is an
important goal for research in this area.
Many studies have been conducted that examine
the linkage between investments in human
resources and firm performance. The majority
of this research reports a positive relationship
between so-called high performance work practices (Huselid, 1995) and different measures of
firm financial performance. Most of the studies in
this stream of research have been based on US data
(e.g., Arthur, 1994; Delery & Doty, 1996; Huselid,
1995; MacDuffie, 1995), but several studies also have
been conducted using data from other countries
rkman, 2001;
(e.g., Bae & Lawler, 1998; Fey & Bjo
Guest, Michie, Conway, & Sheehan, 2003; Ngo,
Turban, Lau, & Lui, 1998).
Recently this line of research has received some
criticism (e.g., Boselie, Dietz, & Boon, 2005;
Gerhart, 2005; Guest, 2001; Truss, 2001) concerning the relationships posited and the notion of a
universal set of high-performance work practices
suitable for all situations and national contexts
(Delery & Doty, 1996; Paauwe & Farndale, 2006;
Schneider & Barsoux, 2003). Although there may
be some areas of universality in HRM, it may well
be that geographical settings affect which HRM
practices are suitable. Likewise, there is growing
agreement in the HRM literature that positing a
direct relationship between HRM practices and firm
performance is too simplistic, and that mediating
TH
Carl F Fey et al
692
Table 1
TH
PY
Model Development
We identified 18 articles that have focused on the
HRMperformance relationship and were published
in major journals over the last decade as a first step
in determining which practices to include in our
model, as there is no standard set of HRM practices
studied in the HRM literature. For each study, we
identified the HRM practices that were included
(see Table 1). Of the 19 practices identified, we
focused on the top six most often studied HR
practices. We then decided to exclude recruitment
from our study, as we viewed recruitment as an
early or initial stage of the employment relationship that does not directly affect a firms efforts to
make employees an effective asset for the organization. Therefore in this study we examine the effect
of five different HRM practices (or sub-bundles of
HRM practices) on MNE subsidiary performance:
competence/performance
appraisal,
employee
training, performance-based compensation, meritbased promotion, and internal communication.1
Many past studies have analyzed the relationship
between the use of specific individual HRM practices and firm performance. However, as Becker and
Practice name
Training
Recruitment
Compensation
Communication
Appraisal
Promotion
Employee participation
Teamwork
Job security
Rewards
Job descriptions
Career development
Decision-making practices
Leader development
Rotation
Quality improvement
Organizational culture
Job analysis
Other (various different practices)
1
a
4
a
8
a
a
a
10
11
12
a
a
a
a
13
14
15
a
a
a
a
a
a
a
a
a
a
a
18
a
a
17
a
a
16
Total
a
a
a
Practice included in previous studies.
Shaded practices are those that were included in our study.
a
a
13
10
8
7
6
6
5
5
4
4
4
3
3
2
2
2
2
2
5
Carl F Fey et al
693
PY
TH
Carl F Fey et al
694
TH
HRM Practices
As emphasized by Huselid (1995), HRM practices
influence employee skills and competencies
through the development of a firms human
capital. Performance appraisal systems help
employees obtain feedback on their performance
and identify ways to enhance their ability that is
useful to the company. Most performance appraisal
systems provide feedback and establish objectives
for employees by establishing targets for the selfdevelopment and training of each employee. There
is a substantial amount of literature that speaks to
the benefits of employee competence/performance
appraisal systems (Brutus et al., 2006). Indeed,
feedback has been shown to enhance an individuals work performance (Kopelman, 1986; Locke,
Cartledge, & Koeppel, 1968). Feedback is useful
because it can provide information about the
correctness, accuracy, and adequacy of work behaviors. Further, feedback provided by competence/
performance appraisal systems is helpful in creating
a sense of competence, accomplishment, and
control in workers (Bandura, 1977).
Investments in employee training are also beneficial in enhancing the human capital of the firm
(Becker, 1975). As Becker (1975: 19) writes: Most
on-the-job training presumably increases the future
marginal productivity of workers in the firms
providing it. Wright et al. (1994) recognize that
the characteristics of individuals do not provide
PY
Carl F Fey et al
695
PY
TH
Carl F Fey et al
696
TH
PY
we needed to decide on which cultural dimension(s) to focus this study. In the absence of a
consensus in the international HRM literature
concerning the dimension on which to focus, we
chose power distance, for the following reasons.
First, of the four dimensions in Hofstedes study,
the biggest differences between Russia, Finland, and
the USA were found in terms of power distance.
Clear differences in power distance among these
countries were also reported in the GLOBE study.
The considerable differences in power distance
between Russia, Finland, and the USA increase the
likelihood that we will see differences across
the countries also in the effects of HRM practices
on MNE subsidiary performance. Second, of
Hofstedes four national cultural dimensions, HR
professionals have perceived power distance to
have the most significant influence on HRM
practices (Budhwar & Sparrow, 2002). Third, power
distance is clearly one of the most established
national cultural dimensions that has not only
been part of several conceptualizations of culture
but has also been used in several studies conceptualizations and in empirical studies of how
national cultural differences impact on HRM
practices (e.g., Milliman, Nason, Von Glinow, Lowe,
& Gallagher, 1998; Schuler & Rogovsky, 1998).
The study by Hofstede (1980) has served as a
useful aid in studying the impact of national
culture on many different issues, including management practices, and will be used as a point of
reference for this study. Although Russia was not
part of Hofstedes original study, Elenkov (1997) has
replicated Hofstedes study in Russia. Hofstede
(1980) reports power distance scores of 33 for
Finland and 40 for the USA, and Elenkov (1997)
reports a score of 89 for Russia. Hofstedes study
does have weaknesses, such as being based on data
from one firm, being fairly old, and having some
methodological concerns. It is therefore important
to corroborate the findings of Hofstede with that
of the GLOBE study conducted in the 1990s. In
the latter study, the researchers studied two manifestations of power distance: as is, which reveals
perceptions concerning current organizational
practices; and should be, which reveals values
with respect to what the respondents believed the
organizational practices should be (House et al.,
2004). In the GLOBE study Russia had the highest
scores for both manifestations of power distance;
the USA was clearly higher than Finland concerning should be, while the USA and Finland had
almost identical scores concerning as is power
Carl F Fey et al
697
TH
PY
Institutions.
Along
with
national
culture,
institutions are likely to have an important effect
on which HRM practices are most effective in a
particular country. Institutions are shared collective
understandings or accepted rules of conduct
that are reflected in laws, rules, governance
mechanisms, and capital markets (North, 1990;
Scott, 2001). Institutional theory argues that
traditional values and practices are embedded in a
countrys social and economic institutions. It has
been argued that it is not desirable to examine
separate aspects of a system without locating it in
its specific societal context (Morishima, 1995).
DiMaggio and Powell (1983) suggest that there
are three major types of isomorphism that affect
organizations: coercive isomorphism, where a
powerful constituency such as the government
imposes certain patterns on the organization;
mimetic isomorphism, where organizations in
situations of uncertainty adopt the pattern exhibited by organizations in their environment that are
viewed as successful; and normative isomorphism,
where professional organizations act as the
disseminators of appropriate organizational patterns, which are then adopted by organizations
under the influence of the professional organizations. More recently, Scott (2001) has suggested
that there exist three pillars of institutional
processes: regulatory (corresponding to DiMaggio
and Powells coercive), cognitive (cf. mimetic), and
normative processes.
Institutional prescriptions play an important role
in influencing economic activity (DiMaggio &
Powell, 1983; North, 1990) and in motivating and
regulating the behavior of actors in a given
environment (Scott, 2001). Institutional theory
suggests that firms experience pressure to conform
to the norms for that environment to gain and
maintain legitimacy in relation to the environment
(e.g., Powell & DiMaggio, 1991), and thus different
institutions (present in different countries) are
likely to produce different pressures. The way
foreign and local businesses need to manage
their operations to be effective depends on the
constraints imposed by the powerful institutions
present in the country where the firm is operating.
These institutions include both formal organizations social, economic, and political bodies and
the social norms and rules that those organizations
Carl F Fey et al
698
PY
TH
Carl F Fey et al
699
PY
TH
Carl F Fey et al
700
PY
TH
METHODOLOGY
Sample and Data Selection
This research examines 100% foreign-owned subsidiaries located in three host countries: Russia, the
USA, and Finland. These countries are different in
terms of national culture, market institutions, and
competitive context, making them a good sample
Carl F Fey et al
701
Training
Employees
ability
H5, H9
H1
Competence and
performance
appraisal
Controls:
Home country,
Size, Age,
% Manufacturing,
& Int. experience
H4, H10
Merit-based
promotion
H3
Firm
performance
H7, H12
Performancebased
compensation
H2
Employees
motivation
Controls:
Home country,
Size, Age,
% Manufacturing,
& Int. experience
Internal
communication
Conceptual model.
Host country:
Russia, USA,
and Finland
H8, H13
Figure 1
PY
H6, H11
TH
Carl F Fey et al
702
TH
Measures
The following are brief descriptions of the independent and dependent variables used in this study
(internal consistency of the scales is discussed in
the results section under the heading Reliability
and validity).
Independent variables
Training: The extent to which subsidiaries used
training is measured through two items assessing
the number of days of formal training annually
received by managerial and non-managerial
PY
Carl F Fey et al
703
Mediating variables
PY
Control variables
TH
Employee ability: This construct captures employees ability. It is not a measure of individual
ability, but a measure of the overall ability of a
subsidiarys employees. This construct was measured by asking respondents to assess the quality
of the firms employees relative to that of its
competitors in overall ability, job related skills,
and educational level. Respondents indicated this
on seven-point Likert-type scales (1far below
average and 7far above average).
Employee motivation: This construct consists of
five items. In the same vein as employee ability,
this is a measure of the overall motivation of a
firms employees, not individual motivation. Two
items asked respondents to assess the quality of
the firms employees relative to that of its
competitors on motivation and work effort, using
a seven-point Likert-type scale (1far below
average and 7far above average). An additional
three items were measured using a five-point
scale (1strongly disagree and 5strongly agree):
(1) the extent to which employees behave in
ways that help firm performance; (2) the extent
to which employees contribute in a positive way
to firm performance; and (3) the extent to which
the subsidiary, compared with the parent company, has a highly motivated group of employees.
industry-based subjective measurement of financial results allows managers to take into account
short-term firm objectives in the assessment of
the financial results. While there are potential
reporting biases in such measures, research has
shown that self-reported subjective performance
data are generally reliable (e.g., Dess & Robinson,
1984). Also, similar research has shown positive
experiences with the use of subjective performance assessments (Delaney & Huselid, 1996;
Youndt, Snell, Dean, & Lepack, 1996). Moreover,
it has been shown that subjective measurements
of firm performance correlate highly with objective measurements of firm performance (Geringer
& Hebert, 1991). Our performance construct is
composed of five items using five-point Likerttype scales measuring profitability, sales, service,
operating efficiency, and quality relative to other
firms in the same industry.
Dependent variable
Carl F Fey et al
704
Employees motivation
TH
Merit-based promotion
Performance-based compensation
Internal communication
Error
MNE subsidiary performance
PY
Employees ability
Training
Statistical Method
Our hypotheses may be summarized in three basic
equations:
RESULTS
It is traditional to present PLS results in two steps.
First, the researcher ensures that the measures
to operationalize the underlying constructs
(measurement model) are valid and reliable. Once
satisfied with the measurement model, the resulting inner model coefficients can be interpreted
(the path coefficients between latent variables or
constructs).
Validity and Reliability of Measures
We assess the measurement model used here by
looking at the reliability of individual items,
the internal consistency between items expected
to measure the same construct, and the discriminant validity between constructs. Individual item
Carl F Fey et al
705
PY
Measurement model
Number of items
Training
Appraisal
Promotion
Compensation
Communication
Ability
Motivation
Performance
Table 3
2
3
3
3
3
3
3
5
Internal consistency
Construct
Table 2
TH
0.940
0.810
0.860
0.790
0.850
0.910
0.860
0.930
Hypothesis Testing
We used the total database, including country
dummies and country interaction terms, to estimate the differences in the effect of HRM practices
across countries. Since the US was considered as the
base case, we include country dummies and
Training
Appraisal
Promotion
Compensation
Communication
Training
Appraisal
Promotion
Compensation
Communication
0.884
0.205
0.188
0.242
0.207
0.591
0.254
0.374
0.237
0.683
0.287
0.313
0.567
0.182
0.650
The diagonal elements in this matrix show the square root of the average variance extracted.
Carl F Fey et al
706
Number of
alternative model
1
2
3
4
5
pvalue
Training-Motivation
Appraisal-Motivation
Compensation-Ability
Promotion-Ability
Communication-Ability
0.26
0.19
0.54
0.37
0.41
No
No
No
No
No
Employees ability
0.001
0.000
0.023
0.013
0.000
0.000
0.020
0.214
0.416*
0.063
Training
RUS Training
FIN Training
Appraisal
RUS Appraisal
FIN Appraisal
0.101*
0.110**
0.013
0.194**
0.087w
0.102*
0.001
0.000
0.103
0.016
0.003
0.014*
0.002
0.113
0.064
0.045
O
TH
U
0.061**
0.039
0.128**
0.074**
0.025w
0.007
0.430***
0.124**
0.064*
Promotion
RUS Promotion
FIN Promotion
Compensation
RUS Compensation
FIN Compensation
Communication
RUS Communication
FIN Communication
0.000
0.001
0.074
0.001
0.000
0.000
0.057
0.007
0.110
0.065
Firm performance
Employees motivation
PY
Variable
Table 5
Added path
Table 4
0.087w
0.068*
0.201***
Employees ability
Employees motivation and behavior
Interaction: Motivation Ability
R2
Observations
0.17
226
0.29
226
0.16
226
Carl F Fey et al
707
Table 6
Support received
at
at
at
at
at
at
po0.01
po0.05
po0.01
po0.01
po0.001
po0.01
TH
PY
Hypothesis
Carl F Fey et al
708
PY
TH
Carl F Fey et al
709
PY
TH
ACKNOWLEDGEMENTS
The authors gratefully acknowledge financial support
from the Swedish Research Council. They are also most
thankful to JIBS Departmental Editor Mary Ann Von
Glinow, and to the anonymous JIBS reviewers who
helped us in improving our manuscript.
NOTES
In another review of empirical studies on HRM and
firm performance, Boselie et al. (2005) found that
employee training, involvement in decision-making,
reward and compensation, and communication and
information sharing were the most commonly studied
HRM practices.
1
Carl F Fey et al
710
REFERENCES
PY
TH
Carl F Fey et al
711
PY
TH
Carl F Fey et al
712
PY
Sergey
Morgulis-Yakushev
(Sergey.MorgulisYakushev@hhs.se) is a doctoral candidate at the
Institute of International Business (IIB) at Stockholm School of Economics in Sweden and a senior
lecturer at the Stockholm School of Economics in
Russia. His research focuses on international strategic HRM and organizational culture and effectiveness in the transforming economies of Russia
and China. Sergey was born in St Petersburg, Russia,
and is a Russian citizen.
TH
Accepted by Mary Ann Von Glinow, Departmental Editor, 10 January 2008. This paper has been with the authors for two revisions.