Samenvatting artikelen
One HRM fits all? A meta-analysis of the effect of HRM practices in
the public, semipublic, and private sector. (Blom, Kruyen, Van der
Heijden & Van Thiel)
Abstract
For a long time, public and semipublic organizations have borrowed HRM practices from the private
sector to enhance employee performance (especially during NPM). There are plenty of arguments,
however, that business-like practices are less effective outside the private sector context because of
sector-specific conditions. Based on the ability-motivation-opportunity model, we performed a
three-level meta-analysis to investigate differences in effects of HRM practices on individual
performance across sectors. Significant differences exist, but the expectation that the effects of HRM
practices are largest in the private sector and smallest in the public sector is not supported. More
specifically, the differences between public, semipublic, and private sector are not straightforward.
Theory
AMO Model and Individual Performance
The AMO model (Appelbaum) focuses on the effects of HRM practices on performance at the
individual level of analysis.
Previous literature has defined individual performance in terms of behaviours and actions that have
an impact on the organization’s goals and are under the control of the individual. These behaviours
can be positive or negative and are often differentiated;
- In-role performance (task performance/ job-specific task proficiency) is defined as doing
what one is hired to do
- Extra-role performance (contextual performance/ organizational citizenship behaviour) is
defined as performance that goes beyond the call of duty for the good of the organization
Building on social exchange theory, the AMO model posits that if employees have the ability,
motivation, and opportunity to do their job, they will demonstrate increased effort, which, in turn,
will result in a higher performance. HRM practices that are aimed to enhance employees’ abilities,
motivation, and opportunities are thought to be viewed as beneficial by these employees and
provide them the incentives to perform.
- The ability dimension is defined as employees having the skills, knowledge, and abilities to
perform
- The motivation dimension is defined as employees’ willingness and drive to perform
- The opportunity dimension refers to employees having the responsibility, authority, and
opportunity to solve problems and make decisions
Following the AMO model, HRM practices can be classified into:
- Ability-enhancing HRM practices focus on increasing employee knowledge, skills, and
abilities
- Motivation-enhancing practices aim to increase employee motivation and include practices
such as contingent rewards, performance management, and internal promotion
opportunities
- Opportunity-enhancing practices focus on employee participation and empowerment and
typical examples are direct participation, job design, and team working
Although previous research has demonstrated positive effects of all three types of HRM practices in
the light of employee performance, differences in effects depending upon the type of HRM practices
are to be expected.
,Public and Semipublic Sector Characteristics Affecting the Impact of HRM practices
Based on the extant literature in this scholarly field, we argue that differences in effects of HRM
practices on individual performance across sectors stem from the variety in organizational goal
ambiguity, personnel constraints, and employee motivation.
- Organizational goal ambiguity is defined as ‘the extent to which an organizational goal or set
of goals allows leeway for interpretation, when the organizational goal represents the
desired future state of the organization’. Ambiguous goal lower the effect of HRM practices
on individual performance.
In the public sector organizational goals are considered harder to measure, more abstract,
more diverse and often more conflicting compared with ones in the private sector.
Consequently, it is harder to design effective training programs in public organizations.
We argue that organizational goal ambiguity is lower in semipublic organizations in
comparison with public sector ones. Semipublic organizations are often single purpose
organizations, so they are expected to have less organizational goal ambiguity.
- Personnel constraints have been found to be more prevalent in the public sector, which, in
turn, are expected to attenuate the effects of HRM practices. The limited discretion of public
managers to hire and discharge employees affects their ability to adopt certain HRM
practices. They experience difficulties in implementing HRM practices to empower
employees due to higher levels of formalization.
There is little research on semipublic organizations, but they appear to be more similar to
private organizations with respect to flexibility in hiring practices and perceptions of the
presence of unnecessary rules.
- Employee motivation. Employees have distinct values, motives, and attitudes that may
influence the effectiveness of HRM practices. In general, employees in the public sector are
supposed to have a higher level of intrinsic and altruistic motivation than employees in the
private sector.
Semipublic organizations carry out public tasks but may operate under private sector
conditions. Therefore, it could be argued that both intrinsic and extrinsic motivational
aspects are present in the semipublic sector.
Hypotheses
Hypothesis 1: The effect of ability-enhancing HRM practices on individual performance is larger in
the semipublic sector than in the public sector (a) and smaller than in the private sector (b).
Hypothesis 2: The effect of motivation-enhancing HRM practices on individual performance is larger
in the semipublic sector than in the public sector (a) and smaller than in the private sector (b).
Hypothesis 3: The effect of opportunity-enhancing HRM practices on individual performance is
larger in the semipublic sector than in the public sector (a) and smaller than in the private sector (b).
Method
Search Strategy
,Inclusion Criteria
Only studies that met the following six criteria were included in our meta-analysis:
1) Studies had to provide correlations for the relationship between individual HRM practices or
AMO-based HRM practices, on one hand, and individual performance, on the other hand.
2) Only studies that examined, for example, preferences of HRM practices.
3) Only studies that provided organization-specific information needed to test our research
hypotheses were included.
4) Only studies that provided the necessary statistical information to perform our meta-
analysis were included.
5) In case a sample was used in multiple studies, only the study that provided the most
information was included.
6) Only studies that were published in English were included.
In the end, 65 articles were selected and coded.
Coding Procedure
In addition to the main variables of interest, study characteristics were coded using a coding scheme
developed by the first author.
‘Method’ and ‘Results’ sections of the different included studies were the primary sources of
information during the coding process. In case crucial information was missing as indicated in the
inclusion criteria, the corresponding author of the empirical work was contacted to obtain the
required data.
Operationalization of Variables
HRM practices according to AMO model à using a two-step procedure, HRM practices were coded as
either ability-, motivation- or opportunity-enhancing practices. In the first step, the practices were
coded as one of the 26 distinguished types of individual HRM practices identified by Boselie et al. In
the second step, these individual practices were coded accordingly to the AMO model.
Individual performance à as explained in the theoretical section, we adopted the widely used
categorization of in-role and extra-role performance. We added a third type of performance, which
we coded as ‘general individual performance’.
Sector à to determine the appropriate sector code, we, first, coded organization type into one of
eight categories. Second, these types were coded into one of the three sectors. The public sector
consisted of central government and state/regional/local government bodies. The semipublic sector
consisted of education and health organizations, because no studies were found that examined
other types of semipublic organizations. The private sector consisted of manufacturing and service
businesses.
Geographical area à findings from previous studies suggest that the effects of HRM practices in
different sectors vary across countries, party due to institutional and cultural differences. To create
, this variable, we coded whether the study was conducted in an Anglo-Saxon, European, Asian, or
other geographical area.
Meta-Analytic Procedure
Our final dataset contained many interdependent effect sizes, as most studies investigated the
effects of various overlapping HRM practices on individual performance. In this study, a three-level
meta-analytical approach was used to account for these dependencies without overestimating
results.
To investigate potential differences between sectors, the following steps were taken:
1) Meta-analytic correlations were calculated using a three-level meta-analysis approach, in
which we controlled the geographical area. Next, we checked several heterogeneity
statistics.
2) Meta-analytic correlations between HRM practices and individual performance outcomes for
each sector were calculated by fitting three-level random effects models.
3) To test the formulated hypotheses, multigroup analyses were conducted using the results
from the WLS estimation. In these analyses, parameters are constrained to be equal across
groups to test for differences between sectors.
Publication bias à due to a possibility that nonsignificant findings go unreported, meta-analyses
could present a too optimistic view of the state of the literature. To assess this impact of publication
bias on the effects of HRM practices on performance in general, we used Egger’s test of the intercept
and Duval and Tweedie’s trim and fill method. None of the intercepts estimated were significant,
whereas the trim and fill analyses indicated that only the relationship between motivation-
enhancing practices and general performance may be influenced by publication bias.
Results
Effect of HRM Practices on Performance Outcomes
In Table 2, meta-analytic correlations are presented between the HRM practices and individual
performance outcomes. For all combinations, the results show a positively significant correlation.
Therefore, no additional analyses were conducted for these relationships.
Effects of HRM Practices on Performance Outcomes per Sector
To investigate the effects of HRM practices on individual performance in each sector, multiple
MASEM models were tested. Figure 2 presents the results for the model with in-role performance as
dependent variable and shows that ability-enhancing practices have a significant effect in both the
semipublic and private sector. In contrast, no significant effect is found for motivation-enhancing
practices in both the semipublic and the private sector. Finally, opportunity-enhancing practices
appear to have a significant in the semipublic and the private sector, while no significant effect is
found in the public sector.
