UOTC Influences of Performance Management in Effective Teams Discussion

  

Description

Discuss what performance management is and how it influences effective teams.

Review table 11.1, define leadership behaviors (in your own words) and note which behaviors are beneficial at specific organizational activities (example: project planning, leading coworkers, etc…). Please note at least five organizational activities and be specific when responding.

Note at least two organizational capabilities and compare and contrast each

Journal of General Management
Vol. 26 No.2 Winter 2000
–
A Reappraisal ofHRM
Models in Britain
by
Pawan
s.
Budhwar
Human Resource Management is still struggling to find a strategic role.
For a better understanding ofthe subject, both management practitioners
and scholars need to study human resource management (HRM) in
context [1]. The dynamics of both the local/regional and international/
global business context in which the firm operates should be given a
serious consideration. Similarly, there is a need to use multiple levels of
analysis when studying HRM: the external social, political, cultural, and
economic environment; and the industry. Examining HRM out-of-context
could be misleading and fail to advance understanding. A key question is
how to examine HRM in context? One way is by examining the main
models of HRM in different settings. However, there is no existing
framework that can enable such an evaluation to take place. An attempt
has been made in this paper to provide such a framework and empirically
examine it in the British context.
This paper is divided into three parts. Initially, it summarises the
main developments in the field of HRM. Then, it highlights the key
emphasis of five models of HRM (namely, the ‘Matching model’; the
‘Harvard model’; the ‘Contextual model’; the ‘5-P model’; and the
‘European model’ ofHRM). Lastly, we will address the operationalisation
of the key issues and emphases of the aforementioned models by
examining their applicability in six industries ofthe British manufacturing
sector. The evaluation highlights the context specific nature of British
HRM.
This introduction looks at the need to identify the core emphasis of
the main HRM models that could be used to examine their applicability in
different national contexts. Developments in the field of HRM are now
well documented in the literature [2, 3]. The debate relating to the nature
ofHRM continues today, although the focus of the debate has changed
over a period of time. At present, the contribution ofHRM in improving
Pawan S. Budhwar is Lecturer in Organizational Behaviour and
HRM at CardiffBusiness School, UK.
Journal of General Management
Vol. 26 No.2 Winter 2000
the firm’s performance and the overall success of any organization
(alongside other factors) is being highlighted in the literature [4, 5].
Alongside these debates, a number of important theoretical
developments have taken place in the field of HRM. For example, a
number ofmodels ofHRM have been developed over the last 15 years or
so. Some of the main models are: the ‘Matching model’; the ‘Harvard
model’; the ‘Contextual model’; the ‘5-P model’; and the ‘European
model’ ofHRM [6, 7]. All these models have been developed in the US
and the UK. These models ofHRM are proj ected to be useful for analysis
both between and within nations. However, the developers of these
models do not provide clear guidelines regarding their operationalisation
in different contexts. Moreover, it is interesting to note that, although a
large number of scholars refer to these models, very few have tested their
practical applicability (exceptions being Benkhoff [8]; Monks [9]; Truss
et al. [10]). For the development ofrelevant management practices there
is then a clear need not only to highlight the main emphasis of the HRM
models but also to show their operationalisation. Such an analysis will help
to examine the applicability of these models in other parts of the world.
With the increasing levels ofglobalisation ofbusiness such investigations
have become an imperative.
Moreover, although the present literature shows an emphasis on
themes such as ‘strategic HRM’ (SHRM), the majority of researchers
persist in examining only the traditional ‘hard’ and’ soft’ models ofHRM
[11]. For the growth and development of SHRM, there is a strong need
to examine the applicability of those models ofHRM which can help to
assess the extent to which it has really become strategic in different parts
of the world, and the main factors and variables which determine HRM
in different settings. This will not only test the applicability of HRM
approaches in different regions, but will also help to highlight the context
specific nature of HRM practices.
The aims of this paper are twofold. First, to identify the core
emphasis offive main models ofHRM which can be used to examine their
applicability in different national contexts. Second, to test empirically the
applicability of these models of HRM in the British context. Before
answering why this investigation is being conducted in the UK, the main
models of HRM are briefly analysed.
Models of HRM
Five models ofHRM, which are widely documented in the literature are
chosen for analysis. They are: the ‘Matching model’; the ‘Harvard
model’; the ‘Contextual model’; the ‘5-P model’; and the ‘European
model’ ofHRM [12,13, 14]. The reason for the selection and analysis of
thesemodelsis two-fold.First, it willhelptohighlighttheirmain contribution
–
Journal of General Management
Vol. 26 No.2 Winter 2000
..
to the development of SHRM as a distinct discipline. Second, it will help
to identify the main research questions suitable for examining these
models in different national settings. The analysis begins with one ofthe
traditional models ofHRM.
The strategic fit of HRM
The main contributors to the ‘Matching model’ ofHRM come from the
Michigan and New York schools. Fombrun et al. ‘s [15] model highlights
the ‘resource’ aspect ofHRM and emphasises the efficient utilisation of
human resources (like otherresources) to meet organizational objectives.
The matching model is mainly based on Chandler’s [16] argument that an
organization’s structure is an outcome of its strategy. Fombrun et al.
expanded this premise and developed the matching model of strategic
RRM, which emphasises a ‘tight fit’ between organizational strategy,
organizational structure and HRM system, where both structure and
HRM are dependent on the organization strategy. The main aim of the
matching model is therefore to develop an appropriate ‘Human Resource
System’ that will characterise those HRM strategies that contribute to the
most efficient implementation ofbusiness strategies. The Schuler group
made further developments to the matching model and its core theme of
‘strategic fit’ in the late 19?Os[17]. The core issues emerging from the
matching models are:
1. Do organizations show a ‘tight fit’ between their HRM and
organization strategy where the former is dependent on the
latter? Do personnellHR managers believe they should
develop HRM systems only for the effective implementation
of their organization strategies?
.2. Do organizations consider their HRs as a cost and use them
sparingly? Or, do they devote resources to the training of
their HRs to make the best use of them?
3. Do HRM strategies vary across different levels of
employees?
The soft variant of HRM
Beer et al. [18] articulated the ‘Harvard Model’ of HRM. It is also
denoted as the ‘Soft’ variant ofHRM [19], mainly because it stresses the
‘human’ aspect of HRM and is more concerned with the employeremployee relationship. The model highlights the interests of different
stakeholders in the organization (such as shareholders, management,
employee groups, government, community and unions) and how their
interests are related to the objectives of management. It also recognises
the influence ofsituational factors (such as the market situation) on HRM
policy choices. According to this model, the actual content of HRM is
described in relation to four policy areas i.e. human resource flows,
Journal of General Management
Vol. 26 No.2 Winter 2000
reward systems, employees’ influence and work systems. Each of the
four policy areas is characterised by a series of tasks to which managers
must attend. The outcomes that these four HR policies need to achieve
are commitment, competence, congruence, and cost effectiveness. The
model allows for analysis of these outcomes at both organizational and
societal levels. As this model acknowledges the role ofsocietal outcomes,
it can provide a useful basis for comparative analysis of HRM [20]. The
key issues emerging from this model which can be used for examining its
applicability in different contexts are:
1. What is the influence ofdifferent stakeholders and situational
and contingent variables on HRM policies?
2. To what extent is communication with employees used as a
means to maximise commitment?
3. What level of emphasis is given to employee development
through involvement, empowerment and devolution?
The contextual model of HRM
Researchers at the Centre for Corporate Strategy and Change at the
Warwick Business School developed this model. They examined strategy
making in complex organizations and related this to the ability to transform
HRM practices [21,22]. Hendry and associates argue that HRM should
not be labelled as a single form of activity. Organizations may follow a
number of different pathways in order to achieve the same results. This
is mainly due to the existence of a number of linkages between the outer
environmental context (socio-economic, technological, political-legal and
competitive)and inner organizationalcontext (culture, structure, leadership,
task-technology and business output). These linkages directly contribute
to forming the content of an organization’s HRM. The core issues
emerging from this model are:
1. What is the influence of economic (competitive conditions,
ownership and control, organization size and structure,
organizational growth path or stage in the life cycle and the
structure of the industry), technological (type ofproduction
systems) and socio-political (national education and training
set-up) factors on HRM strategies?
2. What are the linkages between organizational contingencies
(such as size, nature, positioning ofHR, and HR strategies)
and HRM strategies?
Strategic integration of HRM
The existing literature reveals a trend in which HRM is becoming an
integral part of business strategy – hence, the emergence of the term
SHRM. It is largely concerned with ‘integration’ and ‘adaptation’. The
purpose of SHRM is to ensure that [23]:
–
Journal of General Management
VoL 26 No.2 Winter2000
..
1. HRM is fully integrated with the strategy and strategic needs
of the firm;
2. HR policies are coherent both across policy areas and across
hierarchies; and
3. HR practices are adjusted, accepted, and used by line
managers and employees as part of their every day work.
Based on such premises, Schuler [24] developed a 5-P model of
SHRM that melds five HR activities (philosophies, policies, programs,
practices and processes) with strategic needs. This model, to a great
extent, explains the significance ofthese five SHRM activities in achieving
the organization’s strategic needs, and shows the inter-relatedness of
activities that are often treated separately in the literature. This is helpful
in understanding the complex interaction between organizational strategy
and SHRM activities.
The model raises two important issues (also suggested by many
other authors in the field) for SHRM comparisons. These are:
1. What is the level of integration of HRM into the business
strategy?
2. What is the level ofresponsibility for HRM devolved to line
managers?
European model of HRM
Based on the growing importance of HRM and its contribution towards
economic success and the drive towards Europeanisation, Brewster [25]
proposes a ‘European model ofHRM’. His model is based on the premise
that European organizations operate with restricted autonomy. They are
constrained at both the international (European Union) and national levels
by national culture and legislation, at the organization level by patterns of
ownership, and at the HRM level by trade union involvement and
consultative arrangements [26, p. 3]. Brewster suggests the need to
accommodate such constraints when forming a model ofHRM. He also
talks about ‘outer’ (legalistic framework, vocational training programs,
social security provisions and the ownership patterns) and ‘internal’ (such
as union influence and employee involvement in decisionmaking) constraints
on HRM. Based on such constraints, Brewster’s model highlights the
influence of factors such as national culture, ownership structures, the
role ofthe state and trade unions on HRM, in different national settings.
The European model shows an interaction between HR strategies,
business strategy and HR practice and their interaction with an external
environment constituting national culture, power systems, legislation,
education, employee representation and the constraints previously
mentioned. It places HR strategies in close interaction with the relevant
Journal of General Management
Vol. 26 No.2 Winter 2000
organizational strategy and external environment. One important aim of
this model is to show factors external to the organization as a part of the
HRM model, rather than as a set of external influences upon it.
From the above analyses, it can be seen that there is an element of
both the contextual and 5-P models of HRM present in Brewster’s
European model. Apart from the emphasis on ‘strategic HRM’, one main
issue important for cross-national HRM comparisons emerges from
Brewster’s model. This is:
•
What is the influence of international institutions, national
factors (such as culture, legal set up, economic environment
and ownership patterns), and national institutions (such as the
educational and vocational set-up, labour markets and trade
unions) on HRM strategies and HRM practices?
Recently, Budhwar and associates [27, 28,29,30] have proposed
a framework for examining cross-national HRM. They have identified
three levels of factors and variables that are known to influence HRM
policies and practices and which are worth considering for cross-national
HRM examinations. These are national factors (such as national culture,
national institutions, business sectors and dynamic of the business
environment), contingentvariables (such as the age, size, nature, ownership,
and life cycle stage of the organization, the presence of trade unions and
HR strategies, and the interests of different stakeholders) and
organizational strategies and policies (related to primary HR functions,
internal labour markets, levels ofintegration and devolvement, and nature
ofwork). This framework is used to examine the applicability ofthe issues
arising from the five HRM models in British organizations. But why
conduct this form of investigation, and in the British context?
As mentioned already, there is a scarcity of this type of research.
So far, only Truss et al. [31] have examined the applicability of some of
the models of HRM in a few UK case companies. Apart from their
research, there is scarcely any study that conducts the type ofinvestigation
described here. There are, then, two main reasons for conducting this
investigation in British companies. First, a UK sample possesses the
characteristics suitable to test the operationalisation ofthe main emphases
and critical issues ofthe five models ofHRM. Second, the HRM function
in the UK is under intense pressure due to competitive conditions, and the
restructuring and rightsizing programmes going on in British organizations,
as well as the pressure on British firms from EU and other international
players to stay competitive and meet the EU regulation regarding the
management ofhuman resources. In such dynamic business conditions it
is worth examining the HRM function in context. Moreover, since the five
models have been developed among Anglo-Saxon nations, it is sensible to
test them initially in these countries before recommending their testing in
others parts of the world.
–
Journal of GeneralManagement
Vol. 26 No.2 Winter 2000
..
The Research Methodology
Sample and data collection
A mixed methodology, using a questionnaire survey and in-depth interviews,
was adopted. During the first phase of the research, a questionnaire
survey was conducted between August 1994 and December 1994 in
British firms having 200 or more employees in six industries in the
manufacturing sector (food processing, plastics, steel, textiles,
pharmaceuticals and footwear). The respondents were the top personnel
specialist (one each) from each firm. The response rate ofthe questionnaire
survey was approximately 19 per cent (93 out of500 questionnaires). The
items for the questionnaire were constructed from existing sources, such
as those developed by Cranfield researchers in their study ofcomparative
European HRM [32] and other studies (see for example [33, 34]). The
questionnaire consisted of 13 sections. These were: HR department
structure, role of the HR function in corporate strategy, recruitment and
selection, pay and benefits, training and development, performance
appraisal, employee relations, HRM strategy, influence ofnational culture,
national institutions, competitive pressures and business sector on HRM,
organizational details. Public limited companies represented approximately
one-third of the sample, with the remainder from the private sector. The
industry-wide distribution of respondents is shown in Table 1.
Table 1: Sample Industry Distribution
Indtitry
Food Processing
Plastics
Steel
Textiles
Pharmaceuticals
Footwear
Percentage .
17.2
17.2
16.1
17.2
21.5
10.8
Analysis of the demographic features of the sample suggests that
the sample was representative ofthe total population. Sixty-two per cent
of sample organizations were medium-sized and employed 200-499
employees, 14 per cent employed 500-999 employees, 15 per cent 10004999 employees, and 8 per cent employed 5000 or more employees.
In the second phase of the research, 24 in-depth interviews were
conducted with personnel specialists representative of those firms which
participated in the first phase of the research. The interviews examined
six themes, viz. the nature ofthe personnel function, integration ofHRM
into the corporate strategy, devolvement ofHRM to line managers, and
the influences of national culture, national institutions and business
environment dynamic on HRM.
Journal of General Management
Vol. 26 No.2 Winter 2000
Measures
Multiple regression analysis and descriptive statistics are used to analyse
questionnaire data. Table 1 in the Appendix shows the main dependent
and independent variables used for multiple regression analysis. Table 2
in the Appendix presents the mean scores of respondents regarding the
influence of different aspects of national factors (culture, institutions,
business environment dynamic and business sector) and HR strategies on
HRM policies and practices. The qualitative data is content analysed. In
the discussion, survey results are complemented by key messages coming
from the qualitative interviews.
Findings of the Study
The matching models suggest a strong dependence ofHRM on organization
strategy, i.e, HRM is mainly developed for the effective implementation
of organization strategies. The results show that in 34.6 per cent of the
organizations under study personnel is involved from the outset in the
formation ofcorporate strategy, and 42 per cent oforganizations actively
involve HRM during the implementation stage of their organizational
strategies. Such a trend of ‘active’ personnel management is further
evident from 55 per cent of sample organizations having personnel
representation at board level. Moreover, 81.1 per cent ofthe respondents
believe that their HRM has become proactive over the last five years (i.e.
more involved in decision making).
Such results reflect the growing strategic and proactive nature of
the British personnel function. There is support for such findings in the
existing literature [35, 36].
The second reason to examine the matching models in a crossnational context is to assess whether human resources are considered as
a cost (‘use them sparingly’) or as an asset (spend on training to ‘make
their best use ‘). The results suggest that British organizations claim to be
spending variable though reasonable proportions oftheir annual salaries
on human resource development (HRD) related activities (see Table 2).
Table 2: Proportion of Annual Salaries and Wages Currently
Spent on Training and Development
Value(%)
Percentage of Sample
Nil
0.1- 2.00
2.01-4.00
4.01- 6.00
6.01 or more
Don’t know
41.3
7.6
3.3
1.1
46.7
–
–
Journal of General Management
Vol. 26 No.2 Winter2000
..
A similarpattern characterizes the number ofdays training provided
to different levels ofemployees (see Table 3). The substantial majority of
British firms have increased (rather than maintained or reduced) their
training spend across all categories of staff over the last five years (see
Table 4). There is evidence that this investment has been directed
particularly in the areas of performance appraisal, communication,
delegation, motivation and team building.
Table 3: Average Number of Days Training and Development
Given to Staff Categories Per Year
Number ofDays
Nil
0.1-3.00
3.01-5.00
5.01-10.00
10.1 and above
Don’t know
Mana}!erial(%)
1.2
24.4
20.9
7.0
5.8
40.7
Different Cat~ories of Staff
Prof,/Technical(%)
Clerical(%)
1.1
22.8
21.7
14.7
4.6
40.9
2.3
35.6
13.8
4.6
3.5
40.2
Manual(%)
1.2
24.7
11.7
11.8
9.4
41.2
These developments in the British HRD scene appear to be
consistent with the increased realisation by both business and government
that the development ofhuman resources has been neglected for too long
[37].
Table 4: Nature of Change in Amount of Money Spent on
Training Per Employee
Nature ofChange Mana}!erial(“/o)
Increased
Same
Decreased
Don’t know
59.8
21.7
7.6
10.9
Different Categories of Staff
Clerical(%)
Prof,/Technical(“/o)
63.0
18.5
8.7
9.8
53.3
28.3
7.6
10.9
Manual(%)
60.9
20.7
7.6
10.9
Another key emphasis of the matching model suggests a variation
in HRM strategies across different levels of employees. This is clearly
evident from the results as the nature and type of approach to the
management of different levels of employees vary significantly (see for
example, Tables 3 and 4). This aspect is further highlighted later in this
paper. Based on the above evidence, it seems that the British personnel
function still plays an implementationist role rather than being actively
involved in strategy formulation. On the other hand, there is a strong
emphasis on training and development.
Important Situational Determinants
One of the basic assumptions of the Harvard model of HRM is the
influence of a number of situational factors (such as work force
Journalof General Management
Vol. 26 No.2 Winter 2000
characteristics, unions, labour legislation and business strategy) and
different stakeholders (such as unions, government and community) on
HRM policies. The impact of a few of the situational factors and
stakeholders (proposed by Beer et al. [38D was examined during the
multiple regressions, analysis of means scores and the analysis of
interview results.
Taking the number of employees as a characteristic of the work
force [39, 40], the regression results show that small British organizations
(those having less than 499 employees) are likely to recruit their managerial
staffby advertising externally. Medium size organizations (those having
500 to 999 employees) are likely to recruittheirclerical staffas apprentices.
Large organizations (those having 1000 to 4999 employees) are more
likely to use assessment centres to train their human resources. Lastly,
very large firms (having over 5000 employees) are less likely to recruit
their managerial staff by advertising internally and their manual staff
through the use of word of mouth method. These firms are likely,
however, to recruit their professional staff with the help of consultants.
Moreover, large UK firms are more likely to adopt formal career plans,
succession plans and planned job rotation to develop their human resources
(for details see Table 1 in Appendix).
Support for these findings can be found in the literature (see for
example, [41D. The size ofan organization has a positive relation with the
formalism of their HRM policies [42]. Therefore, as the size of the firm
becomes large, logically, the degree offormalism ofits personnel function
increases and the organization obtains the help ofrecruitment agencies to
recruit its professional employees.
The results show a strong impact of labour laws, educational and
vocational training set up (highlighting government policy) and unions on
British HRM policies (see Table 2 in Appendix). Unions in the UK are
now playing a more supportive role [43]. The implementation of labour
legislation is also having significant influence on UK HRM policies.
Various pressures groups also contribute in this regard (for example,
against age discrimination). Over the last decade or so, the education and
vocational set-up in the UK has initiated a number of programmes and
qualifications such as the national vocational qualifications (NVQs),
investorsinpeople (IIP) and’opportunity2000’ .These are now significantly
influencing HRMin British organizations [44].
The results also show a number ofsignificant regressions regarding
the impact of HR strategies on British HRM. Results in Table 1 in the
Appendix show that organizations pursuing a cost reduction strategy are
more likely to recruit their clerical and manual staffas apprentices. These
organizations are likely to adopt an effective resource allocation HR
strategy. Organizations pursuing a talent improvement HR strategy are
–
Journal of General Management
Vol. 26 No.2 Winter 2000
less likely to recruit their manual staff by word of mouth method.
However, sample firms pursuing a talent acquisition HR strategy are
likely to use consultants to recruit their managerial staff and recruitment
agencies for manual staff. These organizations are also likely to adopt
assessment centres to train their staff.
Most of the above results seem to be logical. For example, by
recruiting employees as apprentices organizations not only pay them less
but also train and prepare them for working in the long run in their
organizations. Hence, it helps to reduce the costs. Similarly, by recruiting
employees externally, organizations increase the opportunity to improve
their talent base.
The second key emphasis of the Harvard model of HRM suggests
extensive use of communication with employees as a mechanism to
maximise commitment [45, p. 63]. Ninety-one per cent of British
organizations share information related to both strategy and financial
performance with their managerial staff. However, this percentage is
significantly lower for other categories of employees (see Table 5).
Table 5: Employees Formally Briefed about Strategy or
Financial Performance
Tvoe ofInformation
Strategy
Financial Performance
Both
Neither
Managerial(%)
–
6.5
91.3
2.2
Different Categmes of Staff
Prof/Technical(%) Clerical(%)
8.0
14.8
65.7
11.6
8.6
39.5
42.0
9.9
Manual(%)
6.4
38.5
23.6
31.5
There can be a number of explanations for the difference in the
sharing of strategic and financial information with different levels of
employees in British organizations. Whilst noting that top personnel
specialists are now more and more involved in strategy making, it seems
that top management continue to be reluctant to devolve responsibility to
line managers for the dissemination offinancial and strategic information.
These issues are further examined when discussing the 5-P model.
The above discussion suggests applicability of the Harvard model
ofHRM in British organizations. The results showed an impact oflabour
laws, education vocational set-up, unions, work force characteristics and
HR strategies on HRM policy choices. There are encouraging results on
the communication of information with different levels of employees
regarding sharing strategic and financial performance and on employee
development through their involvement and training.
Contextual Factors
The main issue against which the relevance of the contextual model can
be evaluated is the impact on HRM policies and practices of economic
Journal of General Management
Vol. 26 No.2 Winter 2000
(characterized by competitive pressures, ownership and life cycle stage),
technological (type ofproduction system)and socio-political (characterised
by national education and training set-up) factors and organizational
contingencies (such as size, age and nature of organization).
The results show a strong influence of competitive pressures on
British HRM policies and practices (see Table 2 in Appendix). To achieve
a competitive edge in such situations, they are focusing particularly on
total customer satisfaction and the restructuring oftheir organizations. As
competitive pressures are also forcing British organizations to enter into
new business arrangements (such as alliances), so these are having direct
influence on HRM policies and practices.
The results also show the impact of increasingly sophisticated
informationand communicationstechnologyon HRM policies and practices
(see Table 2 in the Appendix). Further evidence indicates that the
majority of respondents suggest these technologies mainly influence
training, appraisal and transfer functions. Why? Because with the change
in technology, employees need to be trained to handle it. To see if they
have achieved the required competence they are appraised and if
required, transferred to suitable positions.
Finally, we summarise the relevance of the contextual model of
HRM in terms ofthe impact oforganizational contingencies. Contingent
variables such as size of the organization, presence of HR strategy and
presence of unions were examined above, as were the impacts of
ownership and organizational life cycle stage. These variables do not
seem significantly to impact HRM in British organizations.
Nevertheless, there is significant evidence overall regarding the
applicability of the contextual model ofHRM in British organizations.
Strategic Integration and Devolvement of HRM in Britain
Our discussion now focuses on the relevance of the ‘5 P’ model ofHRM
in British organizations. To achieve this, results regarding the integration
of HRM into corporate strategy and the devolution of responsibility for
HRM to line managers are examined. The detailed results are presented
elsewhere [46], but are summarized below.
In brief, the level of integration is measured on the basis of the
following four scales:
a) representation of Personnel on the board;
b) presence of a written Personnel strategy;
c) consultation ofPersonnel (from the outset) in the development
of corporate strategy; and
d) translation ofPersonnel/HR strategy into a clear set ofwork
programmes.
–
Journal of General Management
Vol. 26 No. 2 Winter 2000
The level ofdevolvement is measured on the basis ofthe following
three scales:
a) primary responsibility with line managers for HRM decision
making (regarding pay and benefits, recruitment and selection,
training and development, industrial relations, health and
safety, and workforce expansion and reduction);
b) change in the responsibility of line managers for HRM
(regarding pay and benefits, recruitment and selection, training
and development, industrial relations, health and safety, and
workforce expansion and reduction); and
c) percentage ofline managers trained in performance appraisal,
communication, delegation, motivation, team building and
foreign language.
High integration is the result of personnel representation at board
level, the personnel function being consulted about corporate strategy
from the outset, the presence of a written personnel strategy, and the
translation of such a strategy into a clear set of work programmes. As
mentioned earlier, the personnel function is represented at board level in
the majority (55 per cent of organizations). For our sample companies,
87.4 per cent have corporate strategies. Of these, 34.6 per cent consult
the personnel function at the outset, 42 per cent involve personnel in early
consultation, and only 13.6 per cent involve personnel during the
implementation stage. Over a quarter (26.4 per cent) of sample
organizations did not have a personnel strategy, 29.9 per cent had an
unwritten strategy and 43.7 per cent had a written personnel strategy. A
clear majority (57.4 per cent) of organizations felt that their personnel
strategy was translated into clear work programmes.
High devolvement is the result of: primary responsibility for pay,
recruitment, training, industrial relations, health and safety and expansion/
reduction decisions lying with the line (see Table 6); line responsibility for
these six areas on an increasing trend (see Table 7); and, evidence of
devolved competency with at least 33 per cent of the workforce being
trainedin appraisals, communications,delegation, motivation,teambuilding
and foreign languages.
Budhwar’s [47] analysis shows that when the four measures of
integration are summated and divided into a single scale ofhigh and low
type, 50.5 per cent of the sample organizations would be categorised as
having high integration and 49.5 per cent fall into the low integration
category. The average score of the summated integration scale for a1193
organizations is .50. These results show a moderate level of integration
being practised in the sample industries. On the other hand, the summated
scales demonstrate a low level of devolvement. Sixty-one per cent of the
sample practise low levels of devolvement of HRM to line managers.
Journal of General Management
Vol. 26 No.2 Winter 2000
Table 6: Primary Responsibility for Major Decisions on
Personnel Issues
Personnel Issues
Related to:
PayandBenefits
Recruitment and Selection
Training andDevelopment
Performance Aonraisal
Industrial Relations
Health and Safety
Workforce
Expansion/Reduction
WorkSystem/Job Design
HRDept.
Line
Mgt.
48.3
17.2
15.1
17.5
36.3
18.5
14.3
12.9
18.3
6.9
13.2
32.6
Line Mgt in
Consultation
wi!il1lB.l)llUt.
11.0
34.4
22.5
30.4
25.3
19.6
19.4
19.4
44.1
IIR Dilpt. in
COllSuJtation
withLineMat.
26.4
35.5
44.1
45.2
25.2
29.3
17.1
40.2
7.6
33.7
18.5
Figures in the above cells represent validpercentage, calculated after excluding the missing
values.
Table 7: Change in Responsibility of Line
Management for Different Personnel Issues
PellSonnelIssues
Increased (%)
Same(%)
Decreased (%)
Pay andBenefits
Recruitment and Selection
Training and Development
Performance Appraisal
Industrial Relations
Healthand Safety
Workforce
Expansion/Reduction
WorkSystem/Job Design
27.2
43.5
69.6
28.9
61.5
65.2
48.9
23.9
37.8
63.3
35.2
7.6
7.6
6.5
2.2
7.8
3.3
38.9
54.4
6.7
43.3
53.3
3.3
60.0
The results confirm the relevance of the 5-P model of HRM in
British organizations. They also help to examine the main emphasis of
Brewster’s [48] European model of HRM, i.e, the linkages between
corporate strategy and HRM strategy.
Conclusion
Overall, the results show a mixed picture, i.e. from strong to moderate
applicability of the mentioned HRM models in Britain. The study aimed to
examine HRM in context, and the findings should be useful for relevant
policy makers. In particular, it seems that the sample firms are practising
a relatively low level of devolvement in comparison to the integration
function. Ifthe HRM function is to become more strategic, then the level
of practice of both these concepts has to increase. Such demands are
likely to increase in future as more and more firms restructure and
become lean in order to respond to competitive and other pressures [49].
The study has two main limitations. First, it is restricted to six
industries ofthe UK manufacturing sector. Second, the views ofonly top
..
Journal of General Management
Vol. 26 No. 2 Winter 2000
personnel specialists were examined. In order, therefore, to obtain a more
comprehensive picture, research needs to be extended to other business
sectors and to the views of other key actors (such as line managers).
Future research could also build upon this study by investigating other
models ofHRM and their applicability in different national contexts.
Appendix
Table 1: Factors Determining HRM Practices in
British Organizations
Independent.
Varin/J/es
Introductory
lifecycle stage
Turnaround
lifecycle stage
Less than 499
employees
Between 500599 employees
Between 10004999
employees
More than
5000
employees
Public Limited
Company
State-owned
organization
Organizations
incorporated
between 18691899
Organizations
incorporated
between 19001947
lJependentVariables
Training and development
through planned iob rotation
Communication through
immediate superior
Recruiting managerial staff by
advertising externally
Recruiting managerial staff by
advertising externally
Recruiting clerical staff from
recruitment agencies
Recruiting clerical staff as
apprentices
Training and development
through assessment centres
If
BiJta .
t·value
0.2102
0.2984*
2.3790
0.1629
-0.2663*
-2.0720
0.3695
-0.3038*
-2.6170
0.3695
0.3658**
3.0590
0.1014
-0.3184*
-2.4220
0.3337
0.2891*
2.4600
0.2607
0.3547**
2.8530
0.1563
-0.2835*
-2.1800
0.1039
0.3223*
2.4550
0.3698
-0.4529**
-3.9340
Recruiting managerial staff by
advertising internally
Recruiting
professionals/technical staff by
use of search/selection
consultants
Recruiting manual staffby
word of mouth
Training and development
through formal career plans
Training and development
through succession plans
Training and development
though planned job rotation
Recruiting managerial staff by
advertising externally
Recruiting managerial staff
from current employees
Recruiting clerical staff from
current emnlovees
Recruiting manual staff by
word of mouth
0.1406
0.375**
2.9170
0.1685
0.4105**
3.2460
0.2102
0.3873**
3.0880
0.3695
0.4436**
3.8050
0.0830
-0.2881*
-2.1700
0.2842
-0.2583*
-2.0650
0.3698
-0.3342**
-2.9100
Commnnication through trade
unions or work councils
0.7445
-0.216**
-3.0370
Recruiting manual staff from
current employees
0.1557
0.2609*
2.0240
Continued …
Journal of General Management
Vol. 26 No.2 Winter 2000
Table 1 Continued:
Independent
Variable
Organizations
incorporated
between19481980
Cost reduction
HRstrategy
Talent
improvement
HRstrategy
lJepen4ent Variables
Recruitingclericalstaffby
advertising externally
Recruitingmanualstaffby
advertising externally
Trainingand development
throughassessment centres
Communication through
immediatesuperior
No formalcommunication
methods
Communication through
suggestion box(es)
Recruitingclericalstaff from
currentemployees
Recruitingclericalstaff as
apprentices
Recruiting manualstaff as
apprentices
Recruitingmanualstaffby
word of mouth
.Jf
Beta
tvalue
0.2465
-0.3931**
-3.2110
0.1974
-0.2767*
-2.1550
0.2607
0.4364**
3.3780
0.1629
-0.3255*
-2.5320
0.3517
0.3265**
2.7370
0.0858
0.2929*
2.2090
0.2842
-0.3019*
-2.4240
0.3337
0.4182**
2.9450
0.1330
0.3646**
2.8240
0.3698
-0.3655**
-3.2440
0.2787*
2.0930
0.3024*
2.2880
0.2857*
2.2090
0.2882*
2.0300
0.3593**
2.9750
0.3502**
2.6960
-0.255*
-2.1820
0.5656**
6.4000
Recruiting managerial staffby
use of search/selection
0.0777
consultants
Talent
acquisition HR Recruitingmanualstaff from
0.0914
strategy
recruitmentagencies
Trainingand development
0.2607
throughassessment centres
Effective
Recruitingclericalstaff as
resourceHR
0.3337
apprentices
strategy
Recruitingmanagerial staffby
0.3695
advertising externally
Recruitingmanualstaffby
0.1226
Unionised
advertising internally
firms
Communication through
0.3517
attitudesurvey
Communication throughtrade
0.7445
unions or work councils
* Significance at .05 level; **Significance at .01 level
–
Journal of General Management
Vol. 26 No.2 Winter 2000
..
Table 2: Influence of Different Aspects of
National Factors on HRM
Aspectsoff”lational (;ultttre
1
2
3
4
5
Way in which managers are socialised
Common values, norms of behaviour and customs
The influence of pressure groups
Assumptions that shape the way managers perceive and
think: about the organization
The match to the organization’s culture and ‘the way we
do things around here’
N(ltif.}1Inl
1
2
3
4
5
1
2
3
4
5
6
T-
o
‘011.6
National Labour Laws
Trade Unions
Professional Bodies
Educational and Vocational training set-up
International Institutions
A~l1ects QflIusinessEnvironment
Increased national/international competition Globalisation of corporate business structure
Growth of new business arrangements, e.g. business
alliances, joint ventures and foreign direct investment
through mergers and acquisitions
More sophisticated information/communication
technology or increased reliance on automation
Changing composition of the workforce with respect to
gender, age, ethnicity and changing employee values
Downsizing of the workforce and business reengineering
Heightened focus on total management or customer
satisfaction
No. ofCases
Mean
84
81
58
18.07
20.28
10.47
84
25.98
86
35.58
82
61
56
84
54
40.91
21.72
15.11
27.62
20.07
72
27.56
66
19.01
70
19.62
48
12.39
69
23.13
78
26.92
71
22.95
79
20.35
82
28.96
56
15.35
61
16.39
37
10.54
49
14.40
39
13.10
Aspects qfBusinessSector
1
2
3
4
5
6
7
8
Common strategies, business logic and goals being
pursued by firms across the sector
Regulations and standards (e.g. payments, training,
health and safety) specific to your industrial sector
Specific requirement/needs of customers or suppliers
that characterise your sector (i.e. supply chain
management)
The need for sector-specific knowledge in order to
provide similar goods/services in the sector
Informal or formal benchmarking across competitors in
the sector (e.g, best practices of market leaders)
Cross-sector co-operative arrangements, e.g, common
technological innovations followed by all firms in the
sector
Common developments in business operations and work
practices dictated by the nature of the business
A labour market or skill requirement that tends to be
used by your business sector only
Respondentswere askedto allocate a totalof100points to the different aspects oftheabove
nationalfactors.
Journal of General Management
Vol. 26 No.2 Winter 2000
References
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
Jackson, S. E. and Schuler, R. S., ‘Understanding Human Resource
Management in the Context ofOrganizations and their Environment’ ,
Annual Review ofPsychology, Vol. 46, 1995, pp. 237-264.
Legge, K., Human Resource Management: Rhetorics and
Realities, Chippenham: MacMillan Business, 1995.
Sisson, K. and Storey, J., The Realities of Human Resource
Management, Buckingham: Open University Press, 2000.
Guest, D. E., ‘Human Resource Management and Performance: A
Review and Research Agenda’, International Journal ofHuman
Resource Management, Vol. 8, No.3, 1997, pp. 263-276.
Schuler. R. S. and Jackson, S. E., Strategic Human Resource
Management, London: Blackwell, 1999.
Brewster, C., ‘Towards a European Model of Human Resource
Management’, Journal of International Business Studies, Vol.
26, No.1, 1995, pp. 1-22.
Legge, K., 1995, op. cit.
Benkhoff, B., ‘A Test ofthe HRM Model: Good For Employers and
Employees’, Human Resource Management Journal, Vol. 7,
No.4, 1997,pp.44-60.
Monks, K., ‘Global or Local? HRM in the Multinational Company:
The Irish Experience’, The International Journal of Human
Resource Management, Vol. 7, No.3, 1996, pp. 721-735.
Truss, C., Gratton, L., Hailey, H., McGovern, P. and Stiles, P., ‘Soft
and Hard Models ofHuman Resource management: A Reappraisal’ ,
Journal ofManagement Studies, Vol. 34, No.1, 1997, pp. 53-73.
Legge, K., 1995. op. cit.
Brewster, C., 1995, op. cit.
Legge, K., 1995. op. cit.
Poole, M., ‘Editorial: Human Resource Management in An
International Perspective’, International Journal of Human
Resource Management, Vol. 1, No.1, 1990, pp. 1-15.
Fombrun, C. J., Tichy, N. M. and Devanna, M. A., Strategic
Human Resource Management, New York: Wiley, 1984.
Chandler, A., Strategy and Structure, Cambridge, MA: MIT
Press, 1962.
Schuler, R. S. and Jackson, S. E., ‘Organizational Strategy and
Organizational Level as Determinants of Human Resource
Management Practices’, Human Resource Planning, Vol. 10,
No.3, 1987, pp. 125-141.
Beer, M., Spector, B., Lawrence, P. R., Quinn Mills, D. and
Walton, R. E., Human Resource Management, New York: Free
Press, 1984.
Legge, K., 1995. op. cit.
Poole, M., 1990. op. cit.
..
Journal ofGeneral Management
Vol. 26 No. 2 Winter 2000
..
[21] Hendry, C and Pettigrew, A.M., ‘Patterns of Strategic Change in
the Development of Human Resource Management’, British
Journal ofManagement, Vol. 3, 1992, pp. 137-156.
[22] Hendry, C., Pettigrew, A. M. and Sparrow, P. R., ‘Changing
Patterns of Human Resource Management,’ Personnel
Management, Vol. 20, No. 11, 1988, pp. 37-47.
[23] Schuler, R. S., ‘Linking the People with the Strategic Needs ofthe
Business’, Organizational Dynamics, Summer, 1992, pp. 18-32.
[24] Ibid.
[25] Brewster, C., 1995, op. cit.
[26] Ibid.
[27] Budhwar, P., ‘Taking Human Resource Management Research To
The Next Millennium: Need For An Integrated Framework’,
Annual Academy of Management Conference, Chicago, 1999.
[28] Budhwar, P. and Debrah, Y., ‘Rethinking Comparative and Cross
National Human Resource Management Research,’ The
International Journal of Human Resource Management, 2001
(forthcoming).
[29] Budhwar, P. and Sparrow, P., ‘An Integrative Framework For
Determining Cross National Human Resource Management
Practices’, Human Resource Management Review, 2001
(forthcoming) .
[30] Budhwar, P. and Sparrow, P., ‘National Factors Determining
Indian and British HRM Practices: An Empirical Study’,
Management International Review, Vol. 38, Special Issue 2,
1998, pp. 105-121.
[31] Truss, C., Gratton, L., Hailey, H., McGovern, P. and Stiles, P.,
1997, op. cit.
[32] Brewster, C. and Hegewisch, A., (eds.) Policy and Practice in
European Human Resource Management, London and New
York: Routledge, 1994.
[33] Baird, L. and Meshoulam, r., ‘Managing Two Fits of Strategic
Human Resource Management’, Academy of Management
Review, Vol. 13, No.1, 1988, pp. 116-128.
[34] Jackson, S. E., Schuler, R. S. and Rivero, J. C., ‘Organizational
Characteristics as Predictors of Personnel Practice’, Personnel
Psychology, Vol. 42, No.4, 1989, pp. 727-786.
[35] Legge, K., 1995, op. cit.
[36] Hendry, C., Human Resource Management: A Strategic
Approach to Employment, Bath: Butterworth-Heinemann, 1998.
[37] Townley, B. ‘Communicating with Employees’ , in Sisson, K. (ed.),
Personnel Management: A Comprehensive Guide to Theory
and Practice inBritain, Blackball Business: London, 1996, pp.595633.
[38] Beer, M., Spector, B., Lawrence, P. R., Quinn Mills, D. and
Walton, R. E., 1984, op. cit.
[39] Jackson, S. E. and Schuler, R. E., 1995, op. cit.
[40] Tayeb, M., Organizations and National Culture, London: Sage,
1988.
Journal of General Management
Vol. 26 No.2 Winter 2000
[41] Sisson, K. and Storey, J., 2000, op. cit.
[42] Brewster, C. and Hegewisch, A., 1994, op. cit.
[43] Heery, E., ‘Annual Review Article 1996’. British Journal of
Industrial Relations, Vol. 35, 1997, pp. 87-109.
[44] Collin, A. and Holden, L., ‘The National Framework for Vocational
Education and Training’, in. Beardwell, 1. and Holden, L., (eds.),
Human Resource Management. London: Pitman Publishing, 1997,
pp.345-377.
[45] Truss, C., Gratton, L., Hailey, H., McGovern, P. and Stiles, P.,
1997,op. cit.
[46] Budhwar, P., ‘Strategic Integration and Devolvement of Human
Resource Management in the British Manufacturing Sector’ , British
Journal of Management, Vol. 11, No.4, 2000, (in Press).
[47] Ibid.
[48] Brewster, C., 1995, op. cit.
[49] Terry, M. and Purcell, J., ‘Return to Slender’ , People Management,
23 October, 1997,46-51.
–
510853
research-article2013
ROP34210.1177/0734371X13510853Review of Public Personnel AdministrationVermeeren et al.
Article
Does Leadership Style
Make a Difference? Linking
HRM, Job Satisfaction, and
Organizational Performance
Review of Public Personnel Administration
2014, Vol. 34(2) 174­–195
© 2013 SAGE Publications
Reprints and permissions:
sagepub.com/journalsPermissions.nav
DOI: 10.1177/0734371X13510853
rop.sagepub.com
Brenda Vermeeren1, Ben Kuipers1,
and Bram Steijn1
Abstract
With the rise of New Public Management, public organizations are confronted
with a growing need to demonstrate efficiency and cost-effectiveness. In this
study, we examine the relationship between public organizational performance and
human resource management (HRM). Specifically, we focus on job satisfaction as a
possible mediating variable between organizational performance and HRM, and on
the influence of a supervisor’s leadership style on the implementation of Human
Resource (HR) practices. Drawing on a secondary analysis of data from a national
survey incorporating the views of 6,253 employees of Dutch municipalities, we tested
our hypotheses using structural equation modeling. The findings indicate that (a)
job satisfaction acts as a mediating variable in the relationship between HRM and
organizational performance and (b) a stimulating leadership style has a positive effect
on the amount of HR practices used, whereas (c) a correcting leadership style has no
effect on the amount of HR practices used.
Keywords
HRM, leadership style, job satisfaction, organizational performance, public sector,
Dutch municipalities
Introduction
During the last three decades, public sector performance has become an increasingly
important issue. With the rise of New Public Management, targets, performance, and
1Erasmus
University Rotterdam, The Netherlands
Corresponding Author:
Brenda Vermeeren, Erasmus University Rotterdam, Room M7-13, P.O. Box 1738, 3000 DR, Rotterdam,
The Netherlands.
Email: vermeeren@fsw.eur.nl
Vermeeren et al.
175
a more business-oriented management approach have come to play central roles within
the public sector (Boyne, Meier, O’Toole, & Walker, 2006; Osborne & Gaebler, 1992;
Pollitt & Bouckaert, 2004). Several innovations in the field promised to increase the
quality of public service while reducing its costs. However, research into human
resource management’s (HRM) contributions to these developments in the public sector has been scarce (Boyne, Poole, & Jenkins, 1999; Gould-Williams, 2003). This
neglect persists despite the fact that employees (those who deliver public services) are
crucial to achieving superior public performance. High-quality services require highly
qualified and motivated personnel (Batt, 2002).
Based on numerous studies in the private sector, we can conclude that human
resource (HR) practices and organizational performance are at least weakly related
(Boselie, Dietz, & Boon, 2005; Guest, 2011; Paauwe, 2009). However, research comparing HRM in the public and private sectors suggests that the HR policies and practices in these sectors differ in many important areas (Boyne et al., 1999). In particular,
public organizations are more likely than private organizations to engage in activities
associated with the role of model employer. Such activities imply commitment to staff
training, trade union, and workforce participation in decision making, promotion of
equal opportunities, and a concern for the welfare of employees to meet their personal
and family needs. Given these empirical findings, we cannot simply assume that the
relationship between HRM and performance will be the same in the public sector.
In private sector–based research on HRM and performance, the assumption is that
an underlying causal link that runs through employee outcomes (in the form of
employee attitudes and behavior) connects HR practices with organizational performance (Boselie et al., 2005; Guest, 2002; Paauwe & Richardson, 1997). In other
words, HR practices are implemented to influence employees, with the ultimate aim to
positively influence the organization’s performance. Job satisfaction is conceptualized
as one of the key indicators of employee outcomes in HRM and performance research
(Guest, 2002; Purcell & Hutchinson, 2007). Previous research has demonstrated a
positive relationship between HRM and job satisfaction (e.g., Guest, 2002; Steijn,
2004) and between job satisfaction and performance (e.g., Hackman & Oldham, 1975;
Judge, Thoresen, Bono, & Patton, 2001; Taris & Schreurs, 2009). These findings support the idea that job satisfaction acts as a mediating variable in the relationship
between HRM and performance. At this time, only a few studies have examined that
mediating relationship (e.g., Ahmad & Schroeder, 2003; Gelade & Ivery, 2003), but
more research is needed to understand how HRM and organizational performance are
related. Such research is even more important in the context of the public sector, as
previous research showed differences in job satisfaction between public and private
sector employees (DeSantis & Durst, 1996).
In general, in the HRM literature is stated that the HR practices perceived or experienced by employees will be those enacted by their supervisors (Bowen & Ostroff,
2004; Paauwe, 2009; Purcell & Hutchinson, 2007; Wright, Gardner, Moynihan, &
Allen, 2005). To influence employee outcomes positively, supervisors require welldesigned HR practices for use in their management activities. Den Hartog, Boselie,
and Paauwe (2004) stressed the important role that supervisors play in implementing
176
Review of Public Personnel Administration 34(2)
an intended HRM policy, as differences in implementation at this level may be attributable to supervisors’ different leadership styles. Such differences in implementation
and communication may lead to variation in employees’ HR perceptions. However,
scholars have uncovered little empirical evidence that bears on the role of supervisors’
leadership styles in HRM implementation. Focusing on leadership style can provide
additional insight into how supervisors influence the implementation of HR
practices.
This study adds to prior research in three ways. First, we focus specifically on the
relationship between HRM and organizational performance in the public sector.
Second, we test whether job satisfaction acts within a public context as a mediator
between HRM and organizational performance. Third, we focus on the influence of a
supervisor’s leadership style on the implementation of HR practices. Thus, our main
research question is as follows:
Research Question: To what extent is the relationship between HRM and the performance of public organizations mediated by job satisfaction and what is the influence of a supervisor’s leadership style on the implementation of HR practices?
After a theoretical exploration of the literature on HRM, job satisfaction, organizational performance, and leadership, we will formulate several hypotheses and test
them using survey data from 6,253 employees of Dutch municipalities. We perform
these tests using structural equation modeling (SEM). We will then discuss our findings. Finally, we conclude by describing suggestions for future research and implications for theory and practice.
Literature Review
The increased focus on performance in the public sector has encouraged a large amount
of research (Boyne et al., 2006; Halachmi & Bouckaert, 1996). In particular, the
impact of management on performance in public organizations has been frequently
studied (Meier, O’Toole, Boyne, & Walker, 2007; Nicholson-Crotty & O’Toole, 2004).
The O’Toole and Meier (1999) model of management is well known and has often
been used to test the impact managers may have on the performance of public organizations. In one of their articles, O’Toole and Meier (2008) focused on the internal side
of management and, in particular, on the contribution of “the human side” of public
organizations to organizational performance in public education. Their results indicate
that the power of HRM in attracting and developing an organization’s human capital
is important to organizational performance. Gould-Williams (2003), in turn, examined
the relationship between HRM and performance in local government in the United
Kingdom. He found, the more HR practices are used within an organization, the
greater the impact on organizational performance. In both articles, the authors stated
that more research is needed to explore the relationship between HRM and organizational performance in the public sector.
Vermeeren et al.
177
As the existing literature has paid little attention to the relationship between HRM
and performance in a public context, we must turn to the general HRM literature to get
more insight. However, that literature contains a very diverse array of theoretical perspectives, definitions, measurements, methodologies, and research fields (Boselie et
al., 2005). Nevertheless, following Paauwe (2009), we can conclude that there is at
least a weak relationship between HR practices and organizational performance. Yet,
despite the fact that several studies indicate a link between HRM and performance,
significant challenges to a full understanding of this relationship still exist (Boselie
et al., 2005; Bowen & Ostroff, 2004; Guest, 2011; Paauwe, 2009).
In this study, we adopt a micro approach to HRM. This approach reflects a more
operational view of HRM by focusing specifically on the effect of multiple HR practices on individuals (Wright & Boswell, 2002). By using this micro approach, we
attempt to acquire more insight into the impact of multiple HR practices on individuals
(measured through job satisfaction) and, subsequently, on organizational performance.
By focusing on job satisfaction as a mediating factor, our aim is to generate a better
understanding of what takes place between HRM and performance. Furthermore,
scholars frequently identify the leadership style of supervisors (who are increasingly
charged with implementing HR practices) as a variable essential to a better understanding of the relationship between HRM and performance (Bowen & Ostroff, 2004;
Paauwe, 2009; Purcell & Hutchinson, 2007; Wright et al., 2005). In this respect,
Purcell and Hutchinson (2007) used the term “people management” to mark the distinction between a supervisor’s leadership style and the application of HR practices.
This distinction is based on the assumption that supervisors require well-designed HR
practices to use in their people management activities and that their leadership style
will influence the way they enact these practices.
The Mediating Role of Job Satisfaction
Guest stated in 1999 that, given the growing interest in research on the relationship
between HRM and performance, a focus on workers’ viewpoints has become increasingly important. An analysis of 104 articles by Boselie et al. (2005) confirms Guest’s
impression that the linking mechanisms between HRM and performance have largely
been disregarded. To understand how HR practices influence employees and improve
worker performance in ways that are beneficial to the organization, research is required
that concentrates on employee perceptions of HR practices and establishes relationships between their job satisfaction and organizational performance, to take one example (Purcell & Hutchinson, 2007). One model that takes this focus is the Paauwe and
Richardson (1997) model on HRM, HRM outcomes and organizational performance.
In this model, the first element consists of HR practices such as recruitment, rewards,
and employee participation. This element influences the so-called HRM outcomes,
such as job satisfaction and motivation. Both of these elements affect the third element, organizational performance, which involves performance indicators related to
the effectiveness, quality, and efficiency of the organization.
178
Review of Public Personnel Administration 34(2)
A variety of studies have examined separate parts of this model. Focusing specifically on the public sector, a number of studies have explored the relationship between
HRM (Element 1) and HRM outcomes (Element 2; for example, Gould-Williams,
2004; Steijn, 2004) and between HRM outcomes (Element 2) and organizational performance (Element 3; for example, Kim, 2005; Ostroff, 1992). The model by Paauwe
and Richardson (1997) adds to this research through its explicit focus on the mediating
effect of HRM outcomes on the relationship between HRM and organizational performance. Moreover, the Paauwe and Richardson model adds to existing public sector
research by promoting an explicit concentration on the concept of HRM itself. This
concentration marks an important difference with the aforementioned management
model by O’Toole and Meier (2008). Therefore, we use the Paauwe and Richardson
model as the starting point for our research. However, while that model offers an
exhaustive range of options to consider for each element, we limit ourselves to job
satisfaction as the only included HRM outcome.
The introduction of job satisfaction enables us to refine the relationship between
HRM and organizational performance. To a large extent, positive employee outcomes
depend on employees’ perceptions of how much the organization cares about their
well-being and values their contributions (Gould-Williams, 2007; Vermeeren, Kuipers,
& Steijn, 2011). In this respect, the degree of job satisfaction will depend on the fulfillment of employee’s needs and values (Hackman & Oldham, 1975). To increase organizational performance, it is likely important that the organization must not only meet
the needs of customers, but also meet those of employees (Schneider & Bowen, 1993).
This assertion is based on the assumption that if organizations care for their employees, these employees will care for the organization (and their customers). In other
words, this argument is based on the assumption that a happy worker is a productive
worker (Taris & Schreurs, 2009). In this respect, the degree to which HR practices are
introduced can be conceptualized as a marker of the extent to which an organization
values and cares for employees. As noted above, previous research has demonstrated
a positive relationship between HRM and job satisfaction (e.g., Guest, 2002; Steijn,
2004) and between job satisfaction and performance (Hackman & Oldham, 1975;
Judge et al., 2001; Taris & Schreurs, 2009).1 These findings support the idea that job
satisfaction acts as a mediating variable in the relationship between HRM and performance. However, this relationship is mostly studied in separate parts and seldom
examined within one design. We will therefore study the relationships among HRM,
job satisfaction, and organizational performance in one model. Following this plan,
our first hypothesis is as follows:
Hypothesis 1: Job satisfaction acts as a mediating variable in the relationship
between HRM and organizational performance.
The Role of Leadership Style
For many years, HRM and leadership were separate research areas. Gradually, interest
in combining these two areas has grown. The connection between these areas is based
Vermeeren et al.
179
on the proposition that employees are likely to be influenced by the HR practices they
experience and their supervisor’s leadership style (Purcell & Hutchinson, 2007).
Supervisors need HR practices to support their management activities, and the way
supervisors enact these practices is influenced by their leadership style. However, previous research on the relationship between HRM and performance paid little attention
to supervisors’ leadership styles. One of the few studies that did attend to leadership
style demonstrated that leadership and employee satisfaction with HR practices have
a strong and independent impact on such employee attitudes as job satisfaction and
commitment (Purcell & Hutchinson, 2007).
However, this demonstration does not allow us to say much about the influence of
different leadership styles on the use of HR practices within an organization. It is
appropriate to assume a relationship exists between different leadership styles and
HRM, because the choice of which HR practices to use appears to be linked to leadership style. For example, Zhu, Chew, and Spangler (2005) have shown that transformational leaders influence organizational outcomes by their use of “human-capital-enhancing
HRM.” Human-capital-enhancing HRM is defined as an approach to managing people
that achieves competitive advantage through the strategic development of a highly
committed and capable workforce (Zhu et al., 2005). Their assumption is that transformational leaders possess a clear vision of what the organization will be, and what it
will do, in the future. HRM plays a critical role in the communication process between
leaders and employees, because without such HRM activities as staffing and training
the leader’s vision will not be transmitted effectively.
Today, scholars in the field of leadership research use many and varied conceptualizations of leadership. Despite differences among these conceptualizations, we can
detect a certain commonality. This commonality is not of jargon, but of the ideas that
underpin the language used. Many conceptualizations are based on a distinction
between an internally and intrinsically directed, people-oriented, and stimulating leadership style versus an externally and extrinsically directed, task-oriented and correcting leadership style (Howell & Avolio, 1993). For example, this distinction underpins
the differentiation made between transformational versus transactional leadership
(Bass & Avolio, 1994) and participative versus authoritive leadership (Likert, 1961).
With respect to the relationship between leadership style and HRM, Guest (1987) has
argued that a more correcting leadership style could be linked to hard HRM and that a
more stimulating leadership style could be linked to soft HRM. In his research, he
refers to the classic distinction in McGregor (1960) between theory X and theory Y.
The “hard” version of HRM is widely acknowledged to place little emphasis on workers’ concerns. In contrast, “soft” HRM would be more likely to pay attention to workers’ outcomes (Guest, 1987).
We will also use McGregor’s distinction between theory X and theory Y. This distinction, despite frequent criticism (Bobic & Davis, 2003), still remains useful for
distinguishing between the different leadership styles a supervisor can adopt. Theory
X assumes that employees are not self-motivated and will avoid work if possible.
Employees, therefore, must be closely supervised and corrected when necessary.
Employees are seen as factors in the production process. Theory Y, in contrast, assumes
180
Review of Public Personnel Administration 34(2)
Stimulating
Leadership
1
2A
HRM
Correcting
Leadership
1
Job
Satisfaction
1
Organizational
Performance
2B
Figure 1. Conceptual model.
that employees are ambitious and self-motivated and can play a crucial role within the
organization. Supervisors must ensure that their employees are properly stimulated by
paying attention to their values and needs. It is in this context that Guest (1999) stated
that if more HR practices are used, the impact on workers will be larger. Based on the
idea that an HRM system should be designed to meet employees’ needs for skills and
motivation and provide them with the opportunity to profile themselves to improve
their performance (Appelbaum, Bailey, Berg, & Kalleberg, 2000), we would expect
that a stimulating leadership style (theory Y) would be accompanied by the use of a
greater number of HR practices tailored to invest in employees and meet their needs
than would be the case for a correcting leadership style (theory X), in which employees are seen as factors in the production process. This leads us to our second hypothesis, which consists of two separate parts:
Hypothesis 2a: A stimulating leadership style has a positive effect on the amount
of HR practices used within an organization.
Hypothesis 2b: A correcting leadership style has a negative effect on the amount of
HR practices used within an organization.
Figure 1 shows the overall theoretical model representing the hypotheses thus
developed above. In the following sections, we present the methodology for testing
this model and our empirical results.
Research Methods
A quantitative study was carried out to address our research question. This section
describes the data and the measurement procedure, including the results of a confirmatory factor analysis using AMOS version 16.
Vermeeren et al.
181
Data
To test our hypotheses about the direct and indirect relationships between the variables
we apply a quantitative research design. For our analysis, we used data from a Dutch
national survey on well-being among municipal employees. In 2005, a public sector
organization representing municipalities approached 29,626 employees of Dutch
municipalities in all functional areas (e.g., administrative, sociocultural, legal and
information and communication technology functions), asking them to fill out a questionnaire about employee well-being via Internet or mail. Of these employees, 7,918
respondents participated in the research. The respondents with missing data for the
analyzed variables were removed from the sample, which resulted in a file with 6,253
respondents. The data for the resulting sample are as follows: 58% are male, the predominant age is 45 to 54 years (37.5%), and the predominant educational level is
secondary (vocational) education (43.1%). When compared with general population
data (A+O fonds Gemeenten, 2005), the sample’s deviation from the general population is small (2%-6%). Despite the response rate of 26.7%, the respondents are generally representative of the population with respect to gender, age, and educational level.
The respondents also worked in different municipalities spread across the Netherlands
and in organizations of various sizes.
Measures
HRM. HRM and performance research exhibits little consistency in the selection of
HR practices by which to measure HRM. Boselie et al. (2005) analyzed 104 important
HRM and performance studies and identified as many as 26 different HR practices
that are used in different studies. No single agreed, or fixed, list of HR practices or
systems of practices exists by which to measure HRM (Guest, 2011; Paauwe, 2009).
Nevertheless, a certain consensus regarding the measurement of HRM has emerged in
the scientific literature on HRM and performance during the past decade. More than
half of the articles published after 2000 made use of Ability, Motivation, and Opportunity (AMO) theory (Paauwe, 2009). AMO theory proposes that an HRM system
should be designed to meet employees’ needs for skills and motivation and, after
meeting those needs, provide them with opportunities to use their abilities in various
roles (Appelbaum et al., 2000). The underlying idea is that employees will perform
well if they have the requisite abilities, when they are motivated and when they obtain
the opportunity to profile themselves (Appelbaum et al., 2000).
In our study, an existing data set is used for secondary data analysis. Although this
data set can be employed to search for the presence of HR practices within organizations, it was not developed for this specific purpose. The survey only measures 10
different HR practices used to a limited extent, and it is not able to measure all the
aspects of HRM proposed by AMO theory. In particular, the survey does not allow us
to determine whether an HR system provides employees with opportunities to use
their abilities in various job roles. Despite this limitation, we use this list of practices
as an indicator of the extent to which HR practices were used in public organizations.
182
Review of Public Personnel Administration 34(2)
Researchers often advocate the study of an HRM system instead of individual HR
practices (Wright & Boswell, 2002). Organizations rarely use HR practices in isolation;
they more typically use them in combination. This system approach adheres to the principle “the whole is more than the sum of its parts” and examines a bundle of HR practices. In this study, we have followed the system approach. In the survey, employees
were asked about the use of 10 different HR practices within their organization (job
evaluation conversations, assessment interviews, personal development plans, training
plans, career plans, competency management, population aging HRM policy, mobility
management, job rotation, and individual coaching). This particular list has been used
in previous research (Steijn, 2004). In accordance with Guest’s suggestion, we counted
how many of these practices were present in the organization according to its employees. Cronbach’s alpha is widely used to demonstrate consistency among a set of items
and, based on the score, it might be argued that a bundle of HR practices can be observed
(Guest, Conway, & Dewe, 2004). The Cronbach’s alpha of the HR bundle is .70. This
is within the range for acceptable internal consistency. The assumption is that the use of
more HR practices suggests the existence of a better developed HRM policy within an
organization. In making this assumption, we can only say something about the surplus
value of HRM in general terms. However, we do not know whether some individual
practices have stronger effects than others, how each of the individual practices affects
performance and whether complementarities or synergistic interdependent relationships among such practices can further enhance organizational performance (Delaney
& Huselid, 1996; Guest et al., 2004; Sels et al., 2006).
Job satisfaction. Job satisfaction is measured using one item: “All things considered,
how satisfied are you with your job?” The answers were given using a 5-point Likerttype scale ranging from very dissatisfied (1) to very satisfied (5). Although there is
some disagreement regarding how to measure job satisfaction, previous research
shows that job satisfaction can reliably be measured using only one item (Nagy, 2002;
Wanous, Reichers, & Hudy, 1997).
Organizational performance. To measure organizational performance, perceptions of
performance and objective performance indicators can be studied (Delaney & Huselid,
1996; Kim, 2005). In this article, the focus is on employee perceptions of organizational performance because objective performance data are not available in the database. When objective performance data are not available, subjective (perceptual)
performance measures may be a reasonable alternative (Delaney & Huselid, 1996;
Kim, 2005). There is evidence of a strong correlation between perceptual and objective measures at the organizational level, although there is always some doubt regarding perceptual measures of performance (Kim, 2005). In this study, we used one item
to measure performance, “the perception that the organization is doing good work,”
utilizing a 5-point Likert-type scale, ranging from totally disagree (1) to totally agree
(5). The use of only one indicator is clearly an important limitation, but at least we are
able to characterize how employees assess their organization’s performance.
Vermeeren et al.
183
Leadership style. To measure the influence of leadership style, we used two latent variables that correspond to the distinction between stimulating and correcting leadership
(cf. Bass & Avolio, 1994; Likert, 1961; McGregor, 1960). The specific items can be
found in the appendix. All answers were given on a 5-point Likert-type scale ranging
from totally disagree (1) to totally agree (5).
Descriptive and reliability statistics were computed for the individual items and the
two scales (see Table 1). To show the strength of the associations between the items,
Table 1 displays the correlations matrix. The correlations are all significant at the 1%
level.
To test whether the distinction between the two leadership styles is supported by the
data, we performed confirmatory factor analysis using AMOS version 16. Unlike
exploratory factor analysis, in which only the number of factors and observed variables are specified, confirmatory factor analysis permits specification and testing of a
more complete measurement model (Byrne, 2001). The simultaneous estimation of
the measurement models allows us to examine the relationships between the items and
their latent constructs as well as the relationships among the constructs themselves.
Furthermore, one also receives information on whether the items load only on their
target variable, or whether they load on the other dimension as well (unidimensionality
of factors). Based on the results of the confirmatory factor analysis, the measurement
model was modified where necessary. The modifications made to enhance the model
included the introduction of error correlations.2 Reasons for error correlation include
respondents’ inability to answer questions, a lack of effort on the part of the respondents to provide the correct answers or other psychological factors, or inadequately
worded questions on the survey questionnaire (Byrne, 2001).
For evaluating the convergent validity of the measurement model, Anderson and
Gerbing (1988) suggested examining the construct loading and determining whether
each estimator’s coefficient is significant. For this model, the regression weights range
from .69 to .89 and all are significant (see Table 1). These coefficients may be interpreted as indicators of the validity of the observed variables, that is, how well they
measure the latent dimension or factor. For this model, convergent validity has been
achieved. With regard to discriminant validity, we note that the items related to the
same construct are always more closely correlated with one another than with the
items for the other construct. In addition, Bagozzi and Philips (1982) suggested that
discriminant validity in SEM is achieved if the unconstrained model has a significantly lower chi-square value than the constrained model. In this study, the chi-square
value for the unconstrained model (CMIN 1711.061/df 62) appears to be significantly
lower than that for the constrained model (CMIN 2722.621/df 63). Thus, for this
model, discriminant validity has been achieved. Finally, the R2 in Table 1 is a measure
of reliability, which indicates how consistently the observed variable measures the
latent dimension. The explained variance corresponding to the observed variables
indicates that the respective factor explains an adequate portion of the variance
(between 47% and 78%; Perry, 1996).
The overall fit of the measurement model was tested using absolute and relative fit
indices, which indicated a good fit. In general, a chi-square test is used to assess the
184
SD
Correcting leadership
X1
3.38 1.032
X2
3.71
.980
X3
3.33 1.060
Stimulating leadership
Y1
3.56 1.091
Y2
3.50 1.116
Y3
3.19 1.091
Y4
3.64 1.103
Y5
3.54 1.067
Y6
3.51 1.123
Y7
3.13 1.165
Y8
3.76 1.070
Y9
3.77 1.056
Y10 2.94 1.092
M
—
.513
.385
.438
.454
.502
.463
.473
.463
.365
.377
.393
.406
.456
.462
.477
.474
.486
.471
.382
.376
.398
X2
—
.569
.560
X1
.429
.486
.504
.457
.481
.499
.484
.426
.434
.511
—
X3
—
.782
.736
.635
.698
.728
.654
.680
.642
.592
Y1
—
.772
.678
.718
.794
.697
.664
.659
.624
Y2
—
.651
.735
.748
.715
.656
.631
.613
Y3
—
.685
.682
.619
.558
.598
.543
Y4
—
.770
.656
.643
.661
.578
Y5
—
.681
.656
.650
.623
Y6
—
.611
.642
.602
Y7
—
.712
.544
Y8
—
.533
Y9
—
Y10
Table 1. Measurement Model: Means, Standard Deviations, Correlations, Standardized Estimates, R2 (N = 6,253).
.841 (.011)
.885 (.011)
.865 (.011)
.774 (.012)
.844 (.011)
.878 (.011)
.800 (.012)
.757 (.012)
.755 (.011)
.711(.012)
.726 (.013)
.687 (.013)
.758 (.013)
Standardized ML
Estimates (SE)
.708
.783
.748
.599
.712
.772
.639
.573
.570
.506
.528
.472
.575
R2
Vermeeren et al.
185
sample data in relation to the implied population data. However, there are concerns
about using the chi-square test because its probability is sensitive to sample size
(Jöreskog, 1993). In larger samples (as in this research), the chi-square test almost
always leads to the rejection of the model because the difference between the sample
covariances and implied population covariances will lead to a higher chi-square value
if the sample size increases.3 As a result, a number of alternative fit measures have
been developed (Hu & Bentler, 1999), including the goodness-of-fit index (GFI), the
adjusted goodness-of-fit index (AGFI), the normed fit index (NFI), and the comparative fit index (CFI). The values for this model were .959 (GFI), .940 (AGFI), .972
(NFI), and .973 (CFI). In the social sciences, a cutoff value of .95 is the prescribed
norm (Hu & Bentler, 1999). Based on these fit indices, one can conclude that the
model is a good fit. In addition, the root mean square error of approximation (RMSEA)
value of .065 indicates that the model is a reasonable fit (Byrne, 2001).
Finally, a traditional measure of scale reliability is Cronbach’s alpha, which measures internal consistency among items on a scale. The Cronbach’s alpha for the stimulating leadership scale is .95 and for the correcting leadership scale is .78. Based on
these results, one may conclude that the reliability coefficients provide independent
corroboration for the results obtained from the use of confirmatory factor analysis. The
results show that the distinction between the two leadership styles is supported by the
data.
Control variables. Of course, several other variables can affect HRM, job satisfaction,
and organizational performance. Therefore, Guest (1999) emphasized that several
controls must be in place to take account of individual and organizational factors. Following Guest, our control variables are divided into two groups. In the first group, we
controlled for individual characteristics (gender, age, and educational level). These
controls are based on the assumption that different groups within organizations may be
managed differently with the result that their perceptions will be different. Then, we
controlled for one important organizational characteristic: organizational size. This
control is based on the assumption that large organizations pursuing improved performance have more resources with which to provide their employees a large HRM
policy.
We coded gender as a dummy variable (1 = female). The category of age was subdivided into five categories (1 = 15-24 years; 2 = 25-34 years; 3 = 35-44 years; 4 =
45-54 years; and 5 = 55 years and older). Educational level was also subdivided into
five categories (1 = primary education; 2 = lower vocational education; 3 = higher
general secondary education, preparatory academic education; 4 = higher vocational
education, candidate exam; and 5 = scientific education). Finally, the category of organizational size was subdivided into seven categories (1 = fewer than 100 employees; 2
= 101-500 employees; 3 = 501-1,000 employees; 4 = 1,001-5,000 employees; 5 =
5,001-10,000 employees; 6 = 10,001-20,000 employees; 7 = more than 20,000
employees). Because we used secondary data analysis, we were restricted to these
categories in measuring the control variables.
186
Review of Public Personnel Administration 34(2)
Table 2. Means, Standard Deviations, and Correlations (N = 6,253).
M
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
Gender
Age
Educational level
Organizational size
HRM
Job satisfaction
Organizational performance
Stimulating leadership
Correcting leadership
SD
1
2
3
4
.42 .493
—
3.57 .958 −.223**
—
3.18 1.169 .071** −.116**
—
2.76 1.269 −.009
.007
.159**
—
3.73 2.04
.004
.045** .093** ,175**
3.78 .933 .037** −.014
.008 −.016
3.48 .956 −.011
.005
.040** .043**
3.46 .914 .008
−.002 −.008
.000
3.47 .854 −.007
.014 −.045** .016
5
6
7
8
9
—
.150**
—
.206** .319**
—
.251** .416** .443** —
.188** .240** .325** .649** —
Note. HRM = human resource management.
**p < .01. Results The hypothesized relationships among the variables were analyzed using SEM. This statistical methodology allows us to test the full conceptual model in a simultaneous analysis. In addition, SEM enables us to analyze simultaneously the direct and indirect relationships among the dependent and independent variables. Finally, SEM also enables us to compare different models (Byrne, 2001). We built our SEM model using AMOS version 16. To examine whether the data were normally distributed, the index of multivariate kurtosis was considered. Bentler (2005) has suggested that, in practice, values above 5.00 are indicative of nonnormality. Our data have a score of 4.94, which indicates that it is normally distributed. In Table 2, the means, standard deviations, and correlations of the study variables are presented.The results show that, of the 10 HR practices, employees observed, on average, the use of 4 HR practices within their organizations. The most frequently observed HR practice was job evaluation conversations, and the least frequently observed practice was job rotation. Employees were generally satisfied with their jobs. The average score for this variable on a 5-point scale was 3.78. Moreover, employees perceive the organization to be doing good work, with the average score on a 5-point scale being 3.48. Finally, the average score for the stimulating leadership style was 3.46 on a 5-point scale; the average score for the correcting leadership style was 3.47. To test the proposed relationships, a causal structure was posited that resulted in a structural equation model. First, we tested the hypothesis that job satisfaction acts as a mediating variable in the relationship between HRM and organizational performance. A distinction can be made between fully mediated and partially mediated models (Wood, Goodman, Beckman, & Cook, 2008). Therefore, in SEM, two different models must be created. In the first model, the direct relationship between HRM and organizational performance was fixed at zero. In the second model, the direct relationship and indirect relationship between HRM and organizational performance were estimated. By using the chi-square difference test and other global-fit measures, one can test the models against each other. In Table 3, the fit indices are presented. The chisquare difference test implies that the relationship between HRM and organizational 187 Vermeeren et al. Table 3. Fit Indices for the Fully and Partially Mediated Models. Model Fully mediated model Partially mediated model χ2 df GFI AGFI NFI CFI RMSEA 189.389 8.670 7 6 .990 .999 .970 .998 .874 .994 .877 .998 .065 .008 Note. GFI = goodness-of-fit index; AGFI = adjusted goodness-of-fit index; NFI = normed fit index; CFI = comparative fit index; RMSEA = root mean square error of approximation. .162 Age -.118 .158 Educational Level Organizational Size .053 .074 .163 .158 HRM (.038) -.044 Job Satisfaction (.024) .294 Organizational Performance (.127) .020 Figure 2. Result of structural equation modeling. performance is partially mediated by job satisfaction. Furthermore, the partially mediated model shows a better model fit than the fully mediated model.In Figure 2, the partially mediated model is shown. Only the statistically significant relationships are described (with a significance level of .01). The numerical scores on all lines indicate standardized regression coefficients (β), and the scores in brackets are the explained variances. Second, we analyzed the effect of leadership style on HRM. We assumed that the amount of HR practices perceived by employees would be influenced by their supervisors’ leadership styles. We distinguished between stimulating and correcting leadership to test our hypotheses that (a) a stimulating leadership style has a positive effect on the amount of HR practices used within an organization and (b) a correcting leadership style has a negative effect on the amount of HR practices used within an organization. The overall model fit was tested using several fit indices. The model fit values were .999 (GFI), .997 (AGFI), .996 (NFI), and .998 (CFI), implying that the model was a very good fit. In addition, the RMSEA, with a value of .015, also indicated that the model is a good fit.The model in Figure 3 is the result. Only the statistically significant relationships are shown (with a significance level of .01). The numerical scores on all lines indicate standardized regression coefficients (β), and the scores in brackets are the explained variances. The results show that a stimulating leadership style has a significantly positive effect on the implementation of HR practices, supporting Hypothesis 2a, whereas a correcting leadership style appears to have no effect on the amount of HR practices used, rejecting Hypothesis 2b. 188 Review of Public Personnel Administration 34(2) .086 Age -.116 .159 -.045 .649 Educational Level Organizational Size .054 HRM (.102) .076 .050 .163 Job Satisfaction (.177) -.025 .252 .161 Organizational Performance (.229) .029 .310 .439 Stimulating Leadership -.054 .069 Correcting Leadership Figure 3. Result of structural equation modeling. Table 4. Results of Cross-Model Validation Showing R2 for the Three Samples. Predicted variable HRM Job satisfaction Organizational performance Full sample 20% sample 80% sample Difference in R2 for 20%-80% sample .102 .177 .229 .109 .197 .240 .100 .173 .231 .009 .024 .009 Note. HRM = human resource management. When we compare the model in Figure 2 with the model in Figure 3, we see that the first model shows a statistically significant and positive relation between HRM and organizational performance. However, the model in Figure 3 shows that this relation becomes weaker when the variables related to leadership style are included. Therefore, we also examined whether supervisors’ leadership style influences the relationship between HRM and performance (moderating effect). However, these effects do not appear to be significant. These results imply that leadership style has its own, independent, effect. Finally, model validity was achieved through cross-model validation. Camilleri (2006) suggested attaining cross-validation in three phases. In the first phase, data are divided into two data sets. One data set consists of a random selection of 20% of the data collected from respondents; the second data set consists of a random selection of 80% of the data collected. In the second phase, SEM by means of a path analysis that calculates the structural fit index (measured by R2) is conducted for both the data sets. The third phase consists of examining the differences between the calculated structural fit indices obtained for each data set. The extent of model validity is determined by the similarity in the variance accounted for by each data set. The results of the cross-model validation are presented in Table 4. Given the fact that the differences in the explained variances are small, the cross-model validation provided satisfactory results. Vermeeren et al. 189 Discussion Looking at the main independent and dependent variables, we expected that a supervisors’ leadership style has an influence on the implementation of HR practices. Our research provides empirical evidence that a supervisor’s leadership style, and specifically a stimulating leadership style, is important to the HRM–performance relationship within an organization. When we compare Figure 2 with Figure 3, we see that adding “leadership” importantly increases explained variance. As such, the results of this study emphasize the important role of supervisors in the HRM and performance model, as was previously suggested by Wright et al. (2005) and Paauwe (2009), among others. When we look at the results in greater detail, we find evidence of the positive relationship between a supervisor’s leadership style and the HR practices conducted within the organization, as previously shown by Purcell and Hutchinson (2007) and Zhu et al. (2005). More specifically, a stimulating leadership style is demonstrated to have an important effect on the implementation of HR practices. In contrast, a correcting leadership style appears to have no effect on the amount of HR practices used. Thus, our hypothesis that a stimulating leadership style has a positive effect on the amount of HR practices used within an organization is confirmed, whereas our hypothesis that a correcting leadership style has a negative effect on the amount of HR practices used within an organization must be rejected. Nevertheless, the results are in line with the research discussed by Guest (1987), which argued that a stimulating leadership style (theory Y) could be linked to soft HRM (HRM focusing on the development, motivation, and commitment of employees). Furthermore, it would be interesting in future research to test Guest’s (1987) idea that theory X (with a correcting role for the supervisor) is linked to hard HRM (a focus on rewards and determinations of whether employees do what the organization requires). To study this relationship, data must include such elements of HRM as performance-related pay. An additional interesting result is that a stimulating leadership style appears to be very important to employees’ degree of satisfaction, while the correcting leadership style has a negative influence on job satisfaction. Finally, a stimulating leadership style and a correcting leadership style have a positive effect on organizational performance, although the effect of the stimulating leadership style is much larger. Our research also provides empirical evidence for the mediating relationship between HRM and organizational performance. The results indicate a direct effect and an indirect effect of HR practices on organizational performance, as is already assumed in the Paauwe and Richardson (1997) model. Our analysis shows that when employees perceive a more elaborate use of HR practices, organizations do achieve a better score for their performance. Moreover, when more HR practices are used, employees experience greater satisfaction, which positively influences organizational performance. This study adds to previous research by confirming the hypothesis that job satisfaction acts as a mediating variable in the relationship between HRM and organizational performance. This important finding provides more insight into employees’ reactions to HRM and its effect on organization performance. These reactions have been largely disregarded in previous research (Boselie et al., 2005). 190 Review of Public Personnel Administration 34(2) Looking at the results in greater detail, we see that older employees and employees with higher education levels perceive a greater use of HR practices. This suggests that different groups within organizations (e.g., younger and older employees) are managed differently. In addition, organizational size has a relatively large effect on HRM, as can be concluded from its high beta weight. In line with Guest’s (1999) assumption, this finding indicates that the HRM policy of organizations is influenced by such contextual variables as the size of the organization. Finally, our study supports the idea that a focus on HRM as a method of increasing organizational performance is also relevant in the public sector. Based on this study, conclusions regarding the relationship between HR practices and organizational performance in private organizations (cf. Paauwe, 2009) also appear applicable to public sector organizations. In line with the results of previous research (e.g., Gould-Williams, 2003; Kim, 2005; O’Toole & Meier, 2008), public organizations appear to be more successful if they value their employees and if they utilize a more extended set of HR practices. In addition, this study illustrates the important role supervisors play in this relationship in the public sector. Conclusion In the introduction, we stated that public sector performance has become an increasingly important issue over the past three decades. Several innovations in the field have promised to increase the quality of public service while reducing its costs. However, research into the contributions of HRM to these developments has been scarce. Our main research question, therefore, was “To what extent is the relationship between HRM and the performance of public organizations mediated by job satisfaction, and what is the influence of a supervisor’s leadership style on the implementation of HR practices?” Based on the data and arguments presented in this study, one can conclude that a positive relationship exists between HRM and organizational performance in the public sector. Specifically, by studying the relationships among HRM, job satisfaction, and organizational performance in a single model, this research showed that job satisfaction partly mediates the relationship between HRM and organizational performance. Moreover, this study showed that the choice to use HR practices is influenced by a supervisor’s leadership style. Despite these findings, the limits of this article suggest lines of further research. This study used a cross-sectional data set restricted to Dutch municipalities. Its findings, therefore, have limitations with respect to internal and external validity. A longitudinal data set would increase internal validity, as such data enable researchers to make stronger causal claims. HRM–performance research is dominated by crosssectional research, which generates considerable discussion of questions regarding “what came first?” (Guest, 2011). Are public organizations more successful if they value their employees, or do public organizations value their employees if they are more successful? Or are both propositions true? A similar problem can be observed with respect to the relationship between job satisfaction and performance (Judge et al., 2001; Taris & Schreurs, 2009). For this reason, a longitudinal research design would Vermeeren et al. 191 be preferable in further research. With respect to external validity, we have examined the HRM and performance relationship in the public sector by focusing on Dutch municipalities. More research is needed to determine whether the HRM–performance relationship holds for different kinds of public sector organizations and different countries. Finally, the selection of the data source (survey) may have influenced some of the results. The use of only one survey instrument may create distortions in the data, in particular regarding common method bias (Podsakoff & Organ, 1986). This is specifically a question with respect to the connection between job satisfaction and organizational performance. The strong relationship between these two variables may be attributable to the fact that employees were asked to rate their job satisfaction and their perceptions of organizational performance. This potential problem highlights the importance of replicating our research, ideally by using objective performance indicators. This study not only generates recommendations to further enhance HRM and performance research in the public sector. Based on its observations, this study also provides possible starting points for improving the performance of public organizations through their employees. To increase organizational performance, it appears important that organizations invest in employees’ needs by implementing HR practices. Moreover, this study suggests that the stimulating leadership style is very important to employee satisfaction, while the correcting leadership style negatively influences job satisfaction. This suggestion further implies that when a public sector organization wishes to acquire an involved and motivated staff, its supervisors must as... Purchase answer to see full attachment