FACTORS INFLUENCING THE
IMPLEMENTABILITY AND IMPLEMENTATION
SUCCESS OF HRM CHANGE PROGRAMS:
CONTENT-, PROCESS- AND CONTEXT-RELATED
ANTECEDENTS
Master Thesis
MSc Business Administration, Specialization Change Management
MSc Human Resource Management
University of Groningen, Faculty of Economics and Business
Ulrike D. Mathies Student number: s2093456 G. Bakkerstraat 93a 9713 HE Groningen Tel: +31 (0)628 569646 u.d.s.mathies@student.rug.nl Date: 26 June 2013 Supervisors Drs. Metha Fennis-Bregman Prof. Dr. Onne Janssen
FACTORS INFLUENCING THE IMPLEMENTABILITY AND IMPLEMENTATION SUCCESS OF HRM CHANGE PROGRAMS:
CONTENT-, PROCESS- AND CONTEXT-RELATED ANTECEDENTS
ABSTRACT
HRM instruments often are not implemented successfully or not the way they were intended. This study examined four factors that came forth from HRM literature as possible antecedents of successful HRM program implementation and can be controlled by the designers of HRM change programs: content clarity, solicitation of intra-organizational commitment (or: politics), participation, and facilitative support provided by the HR department. A newly defined construct termed implementability was hypothesized to mediate the relationship between predictors and implementation success. Results of a cross-sectional multi-case study of 87 different HRM implementation activities indicated that politics was the strongest predictor of implementability and implementation success, followed by content clarity; participation and facilitative support were non-significant in the presence of the other factors. A mediating function of implementability was not supported. Limitations include sample size, sample composition, cross-sectional design, and the operationalization of the variable implementability. This thesis makes contributions to the field of change management by simultaneously assessing factors related to the content, process, and context of change measures. It adds value to the field of HRM by applying concepts and insights of change management research to issues of implementability and implementation of HRM programs. The study’s empirical outcomes have practical implications for HR managers as regards HRM program design: A combination of “hard” power-coercive elements (engaging in intra-organizational politics) and rational-empirical strategies (clear structure, user-friendliness) was the most advantageous for the implementation of planned change interventions; “soft” measures (participative program development and support during implementation) did not play a role in combination.
MANAGEMENT SUMMARY
Human Resource departments regularly face the challenge that the instruments or programs they design, such as a new appraisal system or a changed recruitment policy, are implemented by line managers or supervisors elsewhere in the organization. In this process of passing operational HRM activities on to others information can be lost in translation. As a result, HRM instruments often are not implemented successfully or not the way they were intended. In change management literature the failure rate of planned organizational change is even estimated to be as high as 70 percent (Burnes, 2004). Therefore, it is crucial for HR managers to know, which factors within the realm of their control will contribute to a successful implementation outcome and can help bridge the described gap between intended HRM and actual outcome.
In this thesis, HRM programs were treated as change interventions and, accordingly, looked at through the lens of change management theory, methodology, and practices. The focus was on four factors that in HRM literature were named as possible antecedents of successful HRM program implementation: content clarity, solicitation of intra-organizational commitment (aka ‘politics’), participation, and facilitative support provided by the HR department; a newly defined construct termed implementability was hypothesized to mediate the relationship between predicting factors and implementation success.
Empirical research by means of a survey indicated that merely two factors played a significant role for goal achievement and the reduction of resistance in line managers: solicitation of intra-organizational support and content clarity. Participation and the facilitative support of the HR department only were relevant predictors of successful HRM implementation if looked at by themselves; in the presence of the other factors they became non-significant. Thus, despite of the overwhelming amount of literature advocating participation as the essential means to assure support for a change measurement, results indicated that considering political sensitivities and paying attention to the user friendliness of the program contents were the strongest unique drivers of implementation success.
INTRODUCTION
Organizational change can take a multitude of different forms and sizes on a continuum from emergent to planned and incremental to radical (Weik & Quinn, 1999; Burnes, 2004; van de Ven & Sun, 2011). It can be perceived as large or small with respect to its impact on change recipients and/or with respect to the material and financial resources involved (Lines, 2005). The introduction of a new Human Resource Management (HRM) program in an organization, or a division thereof, most often belongs to the category of deliberately planned, yet incremental change. Nevertheless, the modification of HRM policies and procedures can bring about significant change processes in an organization by affecting the motivation and cooperation of employees, transforming organizational culture e.g. regarding quality or client services, and ultimately impacting organizational performance (Pichault and Schoenaers, 2003). Examples for such HRM change programs are measures, practices or tools related to recruitment/selection/hiring practices, performance appraisal procedures, reward and promotion systems, staff development and training, and job design issues, just to name a few key areas of HRM. Even though HRM in content, they are managed change in nature and, just like any other organizational change, have a certain likelihood of not being implemented successfully. In change management literature the failure rate of planned organizational change initiatives is estimated to be as high as 70 percent or even exceeding that (Beer and Nohria, 2000; Burnes, 2004). Consequently, a lot of theoretical and practical research has been dedicated to the question how to increase the success likelihood of planned change initiatives (Burnes, 2009). The aim of this thesis is to look at the introduction and enactment of new HRM programs in the organization through the lens of change management and to explore some of the variables impacting the ease or difficulty with which they are implemented. I am conceptualizing this “ease or difficulty of implementation” as implementability, a construct that has not been discussed as such in the published literature to date. In my definition, it is understood as a property of the new HRM program, which can be manipulated by its designers in the HR department. It reduces resistance in change recipients and promotes the program’s successful implementation. Implementability is hypothesized to have a mediating function between its antecedents and the organizational outcome variable of implementation success.
validating a measurement instrument. Informed by HRM literature, a discussion of the specific context of this research will lead to a guiding research question; for the development of a conceptual model change management theory is applied and a set of hypotheses will be presented; this is followed by a detailed description of the quantitative cross-sectional research set-up and its results; finally, theoretical and practical implications will be considered. Since HRM change programs comprise tools, practices, policies, and procedures, all of these terms can be used to signify a HRM change program.
HR Departments and Line Managers
HRM programs are regularly designed by Human Resource (HR) managers with the goal of “securing the availability, employability, motivatedness, and vitality of employees” (Emans, 2009, p. 90). Other organization members, namely line managers and supervisors, are subsequently in charge of enacting – or implementing – these tools in their departments, with or without the help of the HR managers. This principle of devolvement – or devolution – means that HR managers hand over their conceptions and delegate operational HRM activities to line managers. Line managers, in turn, by taking ownership, play a crucial role in translating HRM policies into reality and bridging the “gap between prescription and practice” (Caldwell, 2004, p. 197). If they are not successful in implementing a HRM policy as intended, even the best-designed and most strategic HRM program will be ineffective or, worse, damaging (Gratton and Truss, 2003).
present study in a broader framework, a few and rather recent exceptions to this rule will be discussed next.
Research on HRM Implementation
The discrepancy between “intended” and “actual” (or “implemented”) HRM is subject of a theoretical exposé by Wright and Nishii (2007), who explore the multi-level nature of the HRM-performance link from the perspective of Strategic Human Resource Management. In their model, shown below in Figure 1, the relationship between HRM practices and organizational performance is mediated by a number of processes and variables that span both group and individual levels of analysis. “Intended HRM practices” represent the result of HRM strategy, an ideally designed HRM system that HR managers believe will lead to positive organizational outcomes; “actual HRM practices” show that not all intended practices are in fact implemented fully, consistently, and the way they are conceived, which may be detrimental to organizational performance; connecting the two variables, the process of implementation entails the adjustment or transformation of existing systems and practices, leading to organizational change (Wright and Nishii, 2007). It is this very intersection of intended HRM, actual HRM, and the process of implementation that the present thesis will focus on (framed in Figure 1).
FIGURE 1
Mediated Model of the Relationship between HRM Practice and Organizational Performance
An empirical multi-methods study, testing the hypothesis of a gap between “intended” versus “implemented” HRM, sheds some light on the mechanisms underlying the implementation process (Khilji and Wang, 2006). By means of cross-sectional, single-industry analysis of qualitative and quantitative data pulsing HR managers, line managers, and non-managerial employees alike (195 interviews, 508 questionnaire responses, distributed over 12 organizations), three main constructs were measured: HRM practices (intended and implemented), HRM satisfaction, and organizational outcomes. Findings confirmed implemented HRM may be fundamentally different from intended HRM, and indicate that consistent implementation increases employee motivation and satisfaction, which in turn is positively related to organizational performance. Thus, minimizing factors responsible for gaps between intended and implemented HRM is paramount. Organizations rated high on implementation distinguish themselves by at least four significant factors: “incorporating the use of cultural and structural changes in developing effective HRM systems, ensuring employee involvement, developing employee-friendly policies and making HR departments accessible, and providing management support and commitment in implementing changes throughout the organization” (Khilji and Wang, 2006, p. 1186). The authors conclude that one cannot make the link between HRM and performance without understanding how exactly HRM programs are translated into practice; even though companies throughout the world are adopting similar practices, it is the implementation of these practices that can lead organizations to a competitive advantage.
“actual HRM” is a combination of line managers’ leadership behavior, HRM programs, and organizational context.
Finally, a recent Dutch dissertation is dedicated entirely to the role of line managers in the implementation of HRM programs, specifically to the possible constraints experienced by them and the effect of these constraints on their HRM implementation effectiveness (Bos-Nehles, 2010; partially published in Nehles and Boon, 2006; (Bos-Nehles, van Riemsdijk, Kok, and Looise, 2006; and several conference papers). Devolution literature suggests five factors that inhibit line managers in the successful implementation of HRM programs on the work floor: (1) Line managers may not have the desire to perform HRM responsibilities; (2) they lack the capacity and time to do so; (3) they have insufficient HR-competences; (4) there is a shortage of support and advice from the HR department; and (5) they are not provided with clear HRM policies & procedures (Bos-Nehles, 2010). The author investigated and measured the actual effect of devolution constraints on implementing HRM by means of case study research, using quantitative and qualitative cross-sectional data collected in the Netherlands; for the quantitative part, a new instrument for the HRM domain was developed and validated. Consistent with the idea of “perceived HR practices” in Wright’s and Nishii’s model, the dependent variable HR implementation effectiveness was determined by surveying employees about their satisfaction with the HR activities of their line managers. Results indicate that line managers consider the mentioned factors (with the exception of desire) to be highly relevant yet not as constraining as hypothesized; that the intensity of the constraining factors varies by organizational context; and that line managers are perceived by their subordinates as effective – as opposed to ineffective – in implementing HRM practices (Bos-Nehles, 2010). Practical advice given to organizations includes improving support services offered by HR managers and providing specific HR-training to line managers.
and participation of employees and line managers, and (4) facilitative HR departments offering advice and support; in addition, factors in the line managers, such as individual skills level, and in the general organizational context and contingencies are named as affecting HRM outcomes. Still, the current research on implementation lacks a theoretically grounded and empirically validated measurement instrument of these factors.
Antecedents of Implementability
Following Pettigrew’s (1987) model of organizational change, antecedents of implementability are seen as related to the content, process and context of the change program. A number of antecedents, covering all three categories, have empirically been identified as stimulating the implementation of HRM programs (Emans, 2009; Emans, Postema, Weering, Peelen, and Boeve, 2011). These factors, or implementation levers, share the characteristic that they can be influenced and manipulated by the program-designing HR managers. Emans’ study (Emans, 2009; Emans et al., 2011) offers a first overview of the impact of the implementation levers on the implementability of HRM tools. Starting off with an exploratory multiple-case study of HRM programs in eight different organizations, mainly based on interviews, a number of levers were identified; in a quantitative follow-up study, an instrument for measuring these factors was developed and tested. Results indicate that an overall of seventeen levers can be discerned and by factor analysis be positioned with one of five groups of levers such as the smoothness of the implementation process or the degree of collaboration in the process. Due to the small sample (n = 20) and the complexity of the instrument, however, the conducted research requires further refinement and validation. This thesis will build on the work of Emans and colleagues (2011) by focusing on four managerially controllable content-, process-, and context-related implementation levers and their respective effects on the implementability – and eventually implementation success – of the HRM change program. The four chosen antecedents of implementability – content clarity, solicitation of intra-organizational commitment, participation, and facilitation by the HR department – have been selected due to their relevance as described in the literature above. In addition, based on Emans’ et al. (2011), all four variables are expected to have a clearly positive effect on implementability.
What are the effects of content clarity, solicitation of intra-organizational commitment, participation, and facilitative support provided by the HR department on the implementability and eventual implementation success of a HRM change program?
The originality and value of this research are multifold: Firstly, it defines the concept of implementability, which as such has not been described yet in either change management or HRM research. Secondly, this thesis makes a contribution to the field of change management by simultaneously assessing factors related to the content, process, and context of change measures. Within the change management literature, to date, only few studies have integrated these three categories, which are common to all change efforts (Armenakis & Bedeian, 1999; Walker, Armenakis, and Bernerth, 2007; Self, Armenakis, & Schraeder, 2007). Thirdly, it adds value to the field of HRM by applying concepts and insights of change management research to the issue of the implementation of HRM programs, and by offering practical recommendations to HR managers as to how to bridge the well-documented gap between intended HRM practices and implemented HRM practices (Khilji and Wang, 2006).
CONCEPTUAL MODEL
In order to approach the research question, each of the four independent variables will be explored in more detail; an overview of their definitions and dimensions is provided in Table 1. The variables will be linked by a set of hypotheses to the concept of implementability, which is hypothesized to have a partially mediating function to the organizational outcome variable of implementation success. Figure 2 displays the relations as proposed in this section.
Content Clarity
Concluding from implementation literature, clarity of the HRM change program is an important content-related antecedent of implementability (Khilji and Wang, 2006; Purcell and Hutchinson, 2007; Bos-Nehles, 2010; Emans et al., 2011). Along with Bowen and Ostroff’s (2004, p. 209) “understandability” feature of a strong HRM system, content clarity means “a lack of ambiguity and ease of comprehension of HRM practice content”. As such, it combines the aspects of simplicity and user friendliness, and can be defined as the degree to which a program is easy to explain, easy to understand, and easy to use.
In psychology and social sciences simplicity serves as a general principle in cognition and information processing, helping us cope with an immensely complex world (Chater, 1999); and simplifying heuristics play a major role in complex decision making by providing time saving shortcuts (Rubinson, 2009). In change management case studies, simplicity of the interventions and their easy comprehension by potential adopters are named as vital success factors (Atun, Menabdeb, Saluvere, Jesse, Habicht, 2006). Following Emans et al. (2011, p.8), simplicity can be operationalized as “the degree to which a program is devoid of elements that are hard to grasp for the actors involved” and be measured with a scale validated in their research.
User friendliness relates to the presence of unambiguous, pragmatic policies and procedures with clear definitions of how to use them and how to put them into practice. These are necessary to keep line managers from interpreting HRM programs according to their own idiosyncratic understandings and to prevent multiple categorizations (Bowen & Ostroff, 2004). Bos-Nehles (2010) developed items to measure how understandable HRM programs are in reality and whether they are sufficiently concrete to be used in practice.
Given the many positive effects associated with the discussed dimensions of content clarity, it is hypothesized that content clarity will have a positive impact on both the implementability and on implementation success of the HRM program.
Hypothesis 1a: Content clarity of the HRM program is significantly positively related to the implementability of the HRM program.
Solicitation of Intra-organizational Commitment (Politics)
Solicitation of intra-organizational commitment to the HRM change program is a process-related antecedent of implementability. The unconditional commitment of upper management and other intra-organizational stakeholders is considered to be vital for the advantageous completion of the HRM change program (Khilji and Wang, 2006; Purcell and Hutchinson, 2007; Emans et al., 2011). During both the development and the introduction of a new HRM program, HR specialists need to pay attention to organizational politics and power relations and to actively assure the endorsement of influential groups and individuals – aspects that can be measured by using Emans’ validated items for the lever attention to organizational politics.
In change management literature, this factor is mirrored by Kotter’s (2007, reprint of 1995, p. 98) requirement of “creating a powerful enough guiding coalition” for any transformation effort in the organization; here, under the leadership of a key line manager, senior managers as well as members from within and outside of the formal hierarchy come together as a team with the shared mission of making the envisioned change a success. Similarly, but from a practitioners’ point of view, Sirkin, Keenan and Jackson (2005) emphasize the importance of enthusiastic commitment and visible backing of the organization’s most influential executives for the vitality of the transformation project. Based on an initial 225-company study they found a consistent correlation between the success (or failure) of change initiatives and four “hard”, i.e. measurable, factors that organizations are able to both communicate and influence. Without paying attention to these hard factors first, they argue, some change initiatives would never get to the “soft” issues of culture, leadership, and motivation. Senior management commitment represents the “C1” of their DICE approach to change, which
comprises the elements of project duration, performance integrity, top-level and local-level commitment, and coping effort. Top-level commitment is seen as essential for communicating the need for change to lower managers and employees and, thus, for engendering their commitment in turn.
Hypothesis 1b: Solicitation of intra-organizational commitment to the HRM change program is significantly positively related to the implementability of the HRM program.
Hypothesis 2b: Solicitation of intra-organizational commitment to the HRM change program is significantly positively related to the implementation success of the HRM program.
Participation
Participative program development is another process-related antecedent of implementability mentioned in implementability research (Khilji and Wang, 2006; Purcell and Hutchinson, 2007; Emans et al., 2011). It refers to the degree to which the enactors and receivers of a HRM tool were able to contribute to its development and can be measured with the instrument developed by Emans et al. (2011). HRM research has increasingly acknowledged the value of “business partnerships” between HR specialists and line managers – a fact that was reemphasized by case study research of 76 UK-based hotels of Hilton International concluding that perceptual convergence between the partners can lead to increased business performance (Maxwell and Watson, 2006).
found to be positively associated with three hypothesized dimensions of implementation success: organizational commitment, change goal achievement, and minimization of resistance (Lines, 2004, 2007). Despite of a growing body of literature explicating the benefits of participatory practices in change implementation efforts, however, Lewis and Russ (2012) document in a qualitative study that participative communication approaches are underused and that input solicitation and use vary greatly.
Participative program development in the context of HRM can apply to both line managers and affected employees. Consistent with Brown and Cregan’s (2008) large-scale survey research on organizational change cynicism (OCC) two levels of involvement can be distinguished: information sharing and involvement in decision making.
Information sharing is a passive, top-down form of employee involvement, which does not include an opportunity for employees to influence decisions and thus does not lower management authority (Brown and Cregan, 2008). The HRM department just gives updates about the development of the program and its planned implementation. Case studies on organizational change have reinforced the importance of communication processes in implementation situations; especially line managers play a significant role in affecting the attitudes of employees toward change initiatives (Lewis, 2000). Brown and Cregan’s (2008) survey results indicate that information provided by management is associated with greater employee understanding of management decisions and lower levels of OCC, such as frustration, disillusionment and general negative feelings that can result in resistance to change.
Involvement in decision making implies getting line managers and/or affected employees to participate actively in the development of the program and its planned implementation. Sharing responsibility for decision making means that employees are able to bring their workplace experiences to the decision-making table, and is thought to be more durable in reducing OCC than is information sharing alone (Brown and Cregan, 2008).
Hypothesis 1c: Participative program development is significantly positively related to the implementability of the HRM program.
Hypothesis 2c: Participative program development is significantly positively related to the implementation success of the HRM program.
HRM Support
Facilitative support provided by the HRM department is a context-related variable of the implementability of a HRM change program and regularly mentioned by implementation researchers (Khilji and Wang, 2006; Purcell and Hutchinson, 2007; Bos-Nehles, 2010; Emans et al., 2011). While HR specialists have no influence on their external context, they do have limited impact on some aspects of the internal organizational context, of which they are part. First and foremost, they can strive to play a supportive and facilitating role during the implementation phase by e.g. optimizing their accessibility for line managers, being ready to give advice as needed, and/or even carrying some of the implementation workload themselves. This need has already been recognized in Cunningham and Hyman’s (1999) early case studies on the effects of devolution, where line managers complained about a lack of direction, availability or even visibility of personnel specialists; the authors conclude that these tensions could have detrimental effects on the successful implementation of HRM operations. Bos-Nehles’ (2010) research reemphasizes the demand for – and importance of – HRM support: Operationalized as a two-factor model comprising HRM support services and HRM support behavior, the support construct was positively related to perceived HRM implementation effectiveness, when not controlled for organizational context. In addition, analysis of variance showed significant differences between organizations in terms of the support factor, which could be attributed to the service orientation of the HR manager, the structure of the HR function, and the form of support demanded (Bos-Nehles, 2010).
collaboration, and power-coercive strategies, involving the use of politics and authority. Together with participation, facilitative support can be understood as part of the normative-reeducative approach of Chin and Benne’s (1985) taxonomy. Szabla’s (2007) survey research study of cognitive, emotional, and intentional responses to planned change found that the group of participants perceiving a normative-reeducative change implementation process – even though just marginally ahead of those perceiving a rational-empirical approach – held the most positive beliefs, experienced the most positive emotions, and had the highest intentions to support the implementation of the change.
Therefore, I hypothesize that clearly demonstrated support by HR managers will positively affect the implementability as well as the implementation success of the newly devised HRM program.
Hypothesis 1d: Facilitative support by the HRM department is significantly positively related to the implementability of the HRM program.
Hypothesis 2d: Facilitative support by the HRM department is significantly positively related to the implementation success of the HRM program.
p. 139), aspects covered by the variables soliciting intra-organizational commitment and participation in the present model.
Implementability and Implementation Success
In this research implementability is conceptualized as the ease or difficulty with which a new HRM program is introduced and put into practice in an organization. Emans et al. (2011) describe the task to take care of a program’s implementability as “optimizing the whole of content, process and context factors” (p. 5) so that “those who are supposed to use it feel inclined and enabled to do so” (p. 3). Parallel to the concept of readiness for change for change recipients, implementability is understood as a property of the HRM change program that reduces resistance in change recipients and promotes the program’s successful implementation. Holt, Armenakis, Field and Harris (2007a, 2007b) define readiness for change as a multifaceted construct with four dimensions concerning the individuals’ belief in their own change-related self-efficacy, in the appropriateness of the change, in management support of the change, and in personal benefits of the change. Using aspects of these definitions, the dimensions of implementability of a HRM program are determined as follows: The program is designed in a way that line managers are (1) capable of using the tool and (2) willing to do so. Lines 2007 considers the willingness aspect – and its opposite resistance among those affected by the change – part of the organizational outcome variable implementation success, but stresses that it could quite as well be seen as a variable mediating the relationship between a change approach and its outcome. In this research, it is hypothesized that implementability is an indispensable, but not necessarily sufficient, condition for the successful implementation of the HRM change program, and that as such it mediates the relationship between the implementation levers and implementation success. This is consistent with Jones et al. (2005) who found a mediating role of their readiness for change construct between change management strategies and components of change implementation success.
Hypothesis 3: Implementability mediates the relationship between the independent variables and the organizational outcome variable of implementation success.
TABLE1
Overview of the Dimensions and Operationalization of the Variables
Factor Definition Sub-Dimensions Operationalization # Items
Content clarity Degree to which a program is easy to explain, easy to understand, and easy to use. • Simplicity • User friendliness • Program transparency; lack of complexity (Emans et al., 2011) • HR instruments being
clear, concrete, and easy to use (Bos-Nehles, 2010) 3 3 Solicitation of intra-organizational commitment Degree to which power relations in the organization have been considered and support solicited Solicitation of endorsement
of main stakeholders and influential parties during development/ introduction of the program (Emans et al., 2011)
4
Participation Degree to which implementers and receivers of the HRM tool were able to contribute to its development
HR department keeping line managers and
employees informed about program development and allowing varying degrees of influence (Emans et al., 2011)
6
HRM support Degree to which HR department plays a supportive and facilitating role during program implementation phase • HRM support orientation • HRM support behavior • HRM-accessibility: readiness to help (Emans et al., 2011) • HRM-coworker: Active role of HRM department, alleviation of workload (Emans et al., 2011) 6 3 Implementa-bility Ease or difficulty with which a new HRM program is introduced and put into practice in an organization • Design that promotes line managers’ willingness
• Design that enables line managers to use program • Line managers’ willingness to use HRM program; absence of resistance (Lines, 2004) • Line managers’ ability to
use HRM program 3 removed Implementation success Successful introduction of a new HRM program • Goal achievement • Goal congruence
RESEARCH METHOD
In order to test the hypotheses I conducted a cross-sectional multi-case study of different HRM implementation activities. An “implementation activity” is defined as the enactment of a new HRM program in a certain organization or department thereof.
Data Collection
Sample Size. Sample size determines the amount of sampling error inherent in a test result. A
priori sample size calculation for multiple regressions was conducted to estimate the minimum number of responses needed for this research. Given the conventional probability level (p ≤ .05), an anticipated medium effect size (f2 = 0.15), a desired statistical power level (1-β ≥ 0.8), and the number of four predictors in the model, the calculated minimum required sample size is 84 valid responses.
Sample. For each activity, one Dutch HR manager was asked to fill in a self-administered,
organization (range = 1-39 years), and held positions ranging from support tasks to director and board functions with a focus on HR advisor rank. Participants represented diverse industries including (in the order of frequency) government, healthcare, professional services, industry and agriculture, education, retail, food and beverage, other services, and transport. More than half of the participants’ organizations had a size of more than 500 employees.
Instrument. The independent variables were measured based on the instrument developed and validated by Emans et al. (2011; unpublished original data provided by the authors). This Dutch-language questionnaire with originally 81 items translates seventeen antecedents, or levers, of implementability into concrete statements about their nature. In order to refine the scales, all items eliminated in Emans’ et al. (2011) second or third analysis were removed. For the purpose of this study, the Emans et al. scales named simplicity, politics, participation, HRM-coworker and HRM-accessibility were employed; for modifications and additions to these scales see below, “Measures”. I requested that the respondents recall a recently implemented HRM program and give some basic information on its nature. Then they were asked to indicate to what degree each of the statements applied to the case in question. Measures were obtained on 7-point Likert scale ranging from 1 (totally disagree) to 7 (totally agree) and presented to the participants in thematic order. Questions and associated instructions were provided in Dutch. Scales adopted or adapted from questionnaires other than Emans et al. (2011) were either originally phrased in Dutch (Bos-Nehles, 2010) or translated from English employing back-translation techniques as described by Usunier’s (1998). This method achieves translation equivalence by using two independent translators, one of whom translates from source language into target language and the other one the result vice versa. Then the resulting source-language texts are compared and controlled for translation errors. Following Saunders (2009) the final survey was successfully pilot tested (n = 10) to minimize the likelihood of problems with data collection and to allow some assessment of the questions’ validity and reliability.
Measures
By means of literature study I double-checked whether the scales developed by Emans et al. (2011) reflected the latest insights from change management and HRM research. In the case of content clarity I added a validated user friendliness scale as second dimension next to simplicity; for the dependent measures implementability and implementation success all scales had to be adopted and/or adapted from other sources. A number of additional questions were designed to collect biographical and organization-related data. In the following the operationalization of each variable will be explained in detail; an overview of all constructs is presented in Table 1.
Content clarity. The variable content clarity (α = 0.83) is composed of two sub-dimensions:
simplicity and user friendliness. The simplicity dimension was measured by a 3-item scale, testing statements such as “The HRM program is transparent” and “It is easy to explain how the program works” (Emans’ et al., 2011; α = 0.61). Reliability analysis required the removal of one of the originally four items: “The program consists of a multitude of rules and procedures”. For the user friendliness component Bos-Nehles’ (2010; α = 0.88) 3-item scale was employed; example items are “The HR instruments the line managers are provided with are clear and understandable” and “The HR instruments are easy to use”.
Solicitation of intra-organizational commitment. This variable was tested with the 4-item
attention to politics scale developed by Emans et al. (2011; α = 0.85), using items such as “During the introduction of the HRM program support of the main stakeholders was ascertained” and “During the development of the HRM program support of important and powerful parties in the organization was ascertained”.
Participation. Participation of line manager and participation of employees were assessed
Support by the HRM department. For the variable HRM support Emans’ et al. (2011; α =
0.87) 4-item scale HRM-accessibility and 5-item scale HRM-coworker were applied. Factor analysis required a different splitting of the sub-dimensions than suggested by Emans et al. (2011): Six items, all four of HRM-accessibility and the first two of HRM-coworker, belonged to our sub-dimension of HRM support orientation (α = 0.91). Example items are “If there are problems with the implementation of the program, the HRM department is always immediately available for help”; “One person in the HRM department, who is easy to approach, is available for advice and support when there are problems with the implementation of the program”; and “During the implementation of the HRM program, the HRM department customizes their services for the line managers”. The remaining three items of Emans’ et al. HRM-coworker scale constitute our sub-dimension of HRM support behavior (α = 0.77), which is represented by items such as “During the implementation of the HRM program, the HRM department eases the line managers’ burden” and “The contribution of the HRM department to the implementation of the HRM program reduces the workload of the affected line managers”.
Implementation success. The ultimate outcome variable implementation success was
operationalized by two dimensions, goal achievement and goal congruence. The sub-dimension goal achievement was tested by Lines’ (2004, 2007; α = 0.84) two-item scale, consisting of the items “The objectives of this change were achieved” and “The change has contributed to the achievement of strategic goals”. For the sub-dimension goal congruence, HR managers were asked to indicate the degree to which the achieved goals are congruent with the intended outcomes of the HRM program. Factor analysis confirmed that the two sub-dimensions belong to the same overall variable of implementation success (α = 0.71).
Background variables. The background variables used in the questionnaire were age, gender,
work experience, industry branch, organization size, and type of HRM program.
Factor analysis. Factor analysis indicated six major factors that for the most part are
consistent with the six variables of this research. There were two deviations: The variable HRM support consisted of two factors that split along the lines of the sub-dimensions of HRM support orientation and HRM support behavior. Due to the high Cronbach’s alpha (α = 0.87) and the content-relatedness of the sub-dimensions as discussed in the literature section, the variable was still treated as one entity in this research. Contrary, factor analysis suggested that the variables of solicitation of intra-organizational commitment and participation were very close to each other, basically representing different sides of the same factor. This can be explained by the fact that both components of participation (information sharing and involvement in decision making) also can be seen as acts of political behavior and may increase intra-organizational commitment. Nevertheless, based on the argumentation of the literature research, both aspects were operationalized as two separate variables.
Data analysis
can be applied to small samples with more confidence and is advocated as supplement to the causal steps (Hayes, 2009; Hayes, 2013). Therefore, as a second step in data analysis, I have applied bootstrapping to confirm or disconfirm the results of the regression analyses and to get a more refined view on the testing of Hypothesis 3.
RESULTS
Preliminary Data Analysis
Means and Correlations. Descriptive data (means, standard deviations) and intercorrelations
among the different predictor and criterion variables are displayed in Table 2. Correlations demonstrated that the four independent cluster variables content clarity, politics, participation, and HRM support were significantly positively related to the ultimate outcome variable implementation success. The highest correlation was obtained between politics and implementation success (r = 0.54, p < 0.01). The independent variables, with the exception of politics (r = 0.32, p < 0.01), however, were not significantly correlated to the supposed mediator implementability. Strong inter-variable correlations existed between politics and all other independent variables, the one with participation (r = 0.61, p < 0.01) and HRM support (r = 0.49, p < 0.01) being the strongest. In addition, HRM support was significantly positively correlated to both content clarity (r = 0.50, p < 0.01) and participation (r = 0.29, p < 0.01)
Control variables. As shown in Table 2, the control variables age, gender, work experience,
organization size, and type of HRM program were not significantly correlated with either of the outcome variables in this study, whereas industry branch showed a moderately significant correlation with implementability (r = -0.22, p < 0.05). As regards correlations with the predictor variables, gender was positively related to HRM support (r = 0.30, p < 0.05); industry branch to politics (r = -0.29, p < 0.05); and type of HRM program to participation (r = 0.24, p < 0.05).
TABLE 2
Descriptive statistics and Pearson correlations of the variables
n M SD 1 2 3 4 5 6 7 8 1 Age 87 41.29 10.23 - 2 Experience 87 8.45 8.84 .58** - 3 Gender 87 1.68 0.47 -.10 -.23* - 4 Industry branch 87 5.47 2.80 .20 .07 -.16 - 5 Organization size 87 4.01 1.27 .18 .22* .18 .14 - 6 Type HRM program 87 2.20 1.96 -.14 -.10 .08 -.02 -.01 - 7 Content Clarity 75 4.94 1.06 -.01 -.05 .15 -.15 -.00 .10 (.83) 8 Politics 75 5.17 1.25 .17 .04 .17 -.29* .04 .23 .43** (.85) 9 Participation 73 4.13 1.27 .13 -.07 .15 -.19 -.02 .24* .17 .65** 10 HRM Support 69 4.73 1.05 -.04 -.16 .30* -.18 .12 .18 .50** .49** 11 Implementability 81 4.53 1.09 .04 .02 .11 -.22* -.04 .08 -.06 .32** 12 Implementation Success 85 4.89 1.30 .11 -.02 .10 -.16 .01 .08 .49** .54**
Note. Crohnbach alpha reliability estimates appear on the diagonal. *p < .05. **p < .01
n M SD 9 10 11 12 1 Age 87 41.29 10.23 2 Experience 87 8.45 8.84 3 Gender 87 1.68 0.47 4 Industry branch 87 5.47 2.80 5 Organization size 87 4.01 1.27 6 Type HRM program 87 2.20 1.96 7 Content Clarity 75 4.94 1.06 8 Politics 75 5.17 1.25 9 Participation 73 4.13 1.27 (.87) 10 HRM Support 69 4.73 1.05 .29* (.87) 11 Implementability 81 4.53 1.09 .20 .16 (.69) 12 Implementation Success 85 4.89 1.30 .32** .45** .17 (.71)
industry branch, organization size, and type of HRM program was deemed necessary for the study, and regression analyses were performed on the complete sample regardless of the control variables.
Model Testing
In order to test the study’s conceptual model for a mediating effect of implementability a series of simple and hierarchical regression analyses was employed. Following Baron and Kenny’s (1986) four-step approach three conditions must be fulfilled to detect mediation between an independent variable X and a dependent variable Y. Variable M is considered a mediator if (1) X significantly predicts Y, (2) X significantly predicts M, and (3) M significantly predicts Y controlling for X. In step (4) is determined whether the effect of X on Y decreases to zero when controlling for M, indicating full mediation, or not, indicating partial mediation. Results and detailed data of all four steps are presented in Tables 3-6. The regressions were performed on the standardized scores of the scales of content clarity, politics, participation, and HRM support (Aiken & West, 1991).
Step 1: Direct effect of independent variables on outcome variable. Standard multiple
overlap between participation and politics, a standard concern that has been observed in linear mediation models involving highly correlated predictors (Hayes, 2013). Even though each of the two independent variables by themselves were significantly positively associated with implementation success, they compete with each other when both included as predictors in a mediation model, with the result that politics cancels out the effect of participation.
TABLE 3
Standard Multiple Regression Analyses Predicting Implementation Success
b SE β p 1. Content Clarity .31 .14 .26 .03* 2. Politics .41 .18 .34 .02* 3. Participation .04 .16 .04 .79 4. HRM support .17 .15 .14 .26 Overall multiple R .61 R² .38 Adjusted R² .34 F 9.9*** Note. N = 69; * p < 0.05; **p < 0.01; ***p < 0.001
Step 2: Effect of independent variables on mediator. Standard multiple regression analysis
TABLE 4
Standard Multiple Regression Analyses Predicting Implementability
b SE β p 1. Content Clarity -.33 .14 -.31 .02* 2. Politics .45 .18 .44 .01* 3. Participation -.07 .16 -.06 .70 4. HRM support .13 .15 .12 .38 Overall multiple R .42 R² .18 Adjusted R² .12 F 3.39* Note. N = 69; * p < 0.05; **p < 0.01; ***p < 0.001
Step 3: Effect of mediator on outcome variable when controlling for independent variables.
Hierarchical multiple regression analysis were performed to investigate the effect of implementability on implementation success when controlling for the predictor variables content clarity, politics, participation, and HRM support (see Table 5). Results yielded a non-significant association (b = -0.07, n.s), indicating that implementability was not related to implementation success. Hypothesis 3, proposing that implementability serves as mediator between the independent variables and implementation success, is therefore not supported.
TABLE 5
Hierarchical Multiple Regression Predicting Implementation Success (Controlling for Predictor Variables)
Model 1 Model 2 b SE β p b SE β p 1. Content Clarity .31 .14 .26 .03* .29 .15 .24 .06* 2. Politics .41 .18 .34 .02* .44 .19 .37 .02* 3. Participation .04 .16 .04 .79 .04 .16 .03 .81 4. HRM support .17 .15 .14 .26 .18 .15 .15 .24 Implementability -.07 .14 -.06 .60 R .62 .62 R² .38 .39 Adjusted R² .34 .34 F 9.9*** 7.89*** Note. N = 69; * p < 0.05; **p < 0.01; ***p < 0.001
Step 4: Effect of independent variables on outcome variable when controlling for mediator.
conducted. In order to check for partial or full mediation, the results of Step 1 have to be compared to those of Step 4. A mediation effect can be confirmed if (1) the effect of the mediator on the dependent variable is significant, and (2) the effect size of the independent variables as shown by their b-values are considerably lower in Step 4 than in Step 1. Hierarchical multiple regression analysis were performed to investigate the indirect effect of the predictor variables content clarity, politics, participation, and HRM support on implementation success when controlling for implementability (see Table 6). In comparison to Step 1, results of Step 4 displayed only very slight changes in the effect of the predictor variables on the dependent variable implementation success. Of all independent variables only content clarity showed changes in both b- and p-values in the right directions (change b = 0.31 to b = 0.29, change p = 0.03 to p = 0.06); these changes were too small to have a significant statistical effect. For all other independent variables there were either no changes at all or very slight ones in the opposite direction. Therefore, neither a full nor a partial mediation between implementability and implementation success can be confirmed and Hypothesis 3, as already concluded from Step 3, is not supported.
TABLE 6
Hierarchical Multiple Regression Predicting Implementation Success (Controlling for Implementability)
Model 1 Model 2 b SE β p b SE β p Implementability .09 .16 .07 .55 -.07 .14 -.06 .60 1. Content Clarity .29 .15 .24 .06 2. Politics .44 .19 .37 .02* 3. Participation .04 .16 .03 .81 4. HRM support .18 .15 .15 .24 R .73 .62 R² .01 .39 Adjusted R² -.01 .34 F .36 7.89*** Note. N = 69; * p < 0.05; **p < 0.01; ***p < 0.001 Bootstrapping
samples) four times, content clarity, politics, participation and HRM suport were entered one at a time as the independent variable while the remaining three served as covariates; implementability was added as mediator and implementation success as dependent variable. This procedure does not result in a single overview of the total indirect effect across all independent variables, but rather in estimates for each predictor. The coefficients of these four bootstrap tests are summarized in Table 7.
TABLE 7
Model Coefficients for Mediation Analysis Using Bootstrapping
Implementability Implementation Success
Coeff. SE p Coeff. SE p Implementability -.07 .14 .60 Content Clarity -.30 .13 .02* .29 .15 .05 Politics .41 .16 .01* .44 .19 .02* Participation -.06 .14 .70 .04 .16 .80 HRM support .12 .14 .38 .18 .15 .24 Constant -.02 .11 .89 5.00 .12 .000*** R² .17 .38 F 3.34* 7.89*** Note. N = 69; * p < 0.05; **p < 0.01; ***p < 0.001
For politics both the total effect (point estimate = 0.41, with a 95% bias-corrected bootstrap confidence interval of 0.06 to 0.77; p < 0.05) and the direct effect (point estimate = 0.44, with a 95% bias-corrected bootstrap interval of 0.07 to 0.82, p < 0.05) on implementation success were very strong and significant, while the indirect effect was negative and statistically not significant (point estimate = -0.03, with a 95% bias-corrected bootstrap interval of -0.17 to 0.08). There was no evidence that implementability served as a mediator to implementation success, leading to the rejection of Hypothesis 3 for politics.
The direct effect sizes for both participation (point estimate = 0.04, with a 95% bias-corrected bootstrap interval of -0.28 to 0.36; p = 0.81) and HRM support (point estimate = 0.18, with a 95% bias-corrected bootstrap interval of -0.12 to 0.48; p = 0.24) were small and not significant. In both cases, the indirect effects were statistically not different from zero and insignificant, thus also not supporting Hypothesis 3.
In conclusion, the bootstrapping tests were in line with the regression analyses data and confirmed the outcomes of Baron and Kenny’s causal steps. All conclusions made about the support or rejection of Hypotheses 1-3 were kept intact.
DISCUSSION
Findings
between the predicting factors and implementation success as outcome. Broadly speaking, the results showed that politics in the form of solicitation of intra-organizational commitment and, to a lower degree, content clarity played an overwhelming role in influencing both implementability and implementation success. Taken all together, the significant factors accounted for more than a third of variance in implementation success, and for about a fifth of variance in implementability. Rather surprisingly, neither participation nor HRM support appeared as significant factors and can therefore not be considered to be predictors in this study. In-depth analyses yielded no indication for a mediating effect of implementability for any of the study variables.
Politics. As a first result, politics was the strongest predictor of both implementability and
political action does not necessarily have to be self-serving but, used the right way, can majorly influence the success of planned change. Nevertheless, the use of politics epitomizes Beer and Nohria’s (2000) “hard”, power-coercive approach to change management.
Content clarity. Content clarity was the second strongest predictor of implementability and as
such confirmed previous research: Less unnecessary complexity in the content of the HRM program and more focus on customer concerns triggered higher levels of goal achievement and overall more positive implementation results. This phenomenon has mainly been observed in HRM research, where it was sub-categorized as “simplicity” (Emans et al., 2011), “understandability” (Bowen and Ostroff, 2004) or “user friendliness” (Bos-Nehle, 2010). Contrary to this expected outcome, content clarity was negatively associated with implementability. Instead of promoting the willingness of line managers to implement a new HRM program as predicted, an increase in simplicity appeared to have a revered effect and lead to more resistance in line managers. User friendliness proved to be not significant in this relationship. A possible explanation for this seemingly paradoxical observation can be found in recent research on resistance to change. Pieterse & Caniëls (2012), discussing the role of discourse in their case study of the implementation of a new IT system for an airline, attribute resistance to differences in professional cultures in cross-functional teams. They argue that change programs contain subjective, informal and linguistic dimensions that may not be understood correctly by people from different professional backgrounds, such as from a different department; this can lead to confusion, resentment, withdrawal from cooperation, and eventually the failure of the change project. Applied to the feature of content simplicity in my research, it could be that a HRM program that, in the opinion of their creators, is designed to be “transparent, “easy to explain” and “void of complexity” is perceived differently by the receivers of the program. They may understand it as either not simple at all or, on the other extreme, as too simplistic. Either way such “unarticulated non-alignments of professional discourses” (Pieterse & Caniëls, 2012, p. 811) may lead to a breakdown in communication and result in resistance from the side of line managers.
Participation. Even though participation in change management literature often is advocated
implementation outcome. As regards creating willingness among line managers, participation did not play a role at all, even when the politics scale was removed from the regression. These findings are diametrically opposed to Lines’ (2004, 2007) research, in which the use of participation was positively associated with both change goal achievement and minimization of resistance. A comprehensive meta-analysis of the older literature on participation and performance, however, found participation to have a minor impact on performance and questioned the practical significance of participation as a means of influencing performance (Wagner, 1994). Furthermore, Rafferty and Simons (2005) encountered a phenomenon similar to the current finding in their examination of antecedents of readiness for fine-tuning and corporate transformation changes. In their research participation was only significantly associated with readiness for fine-tuning changes, but not with readiness for corporate transformation change in the presence of a range of other measures; mainly trust in organizational leadership cancelled out the effect of participation on readiness for corporate transformation changes. From all this it can be concluded that participation – the centerpiece of Beer and Nohria’s (2000) “soft” approach – is not a panacea for all change situations, but that it depends on the presence of other factors whether or not participation will be a strong driver of change outcome variables.
HRM Support. Contrary to the prediction, the facilitative support provided by the HR
department during and after the implementation of a new HRM program was neither a significant predictor of implementation success nor of implementability when in the presence of the other factors. Entered by itself it did predict implementation success, but not implementability. In Chin and Benne’s (1985) taxonomy facilitative support, together with participation, belongs to the normative-reeducative approach to change management, which is associated with positive cognitive, emotional, and intentional responses in change recipients (Szabla, 2007). The current findings suggest that even though support by the HR department may trigger positive reactions and intentions in those affected by the change, it does not help reduce resistance in change recipient, nor does it contribute to better goal achievement and increased implementation success when accompanied by other measures.
a limit on the relationships the construct has with the predictor variables on the one hand and implementation success on the other. Only politics and, in a negative way, content clarity have shown to have a direct effect on implementability, whereas participation and HRM support do not. Implementability was not significantly positively associated with implementation success and did not have a mediating function between the independent and outcome variables. Taking the limitations of this study into account, one can cautiously conclude that absence of resistance in those affected by the change program does not necessarily lead to positive outcomes as regards change implementation success and goal achievement. Interestingly and in line with this finding, recent literature on resistance to change and/or change readiness has put much emphasis on the affective and emotional elements of this attitude; however, while they identify change-supportive behaviors, positive job attitudes, job satisfaction, and organizational commitment as key outcomes of individual change readiness, they do not claim that these would automatically lead to change implementation success (Bouckenooghe, 2010; Rafferty, Jimmieson, & Armenakis, 2013).
Conclusion. Evaluating this study’s finding in the light of Chin and Benne’s (1985) three-change strategy classification it is remarkable that with soliciting intra-organizational commitment, which can be interpreted as using political activities to access power, the “hard” power-coercive approach came out as the most effective strategy to acquire implementation success. With content clarity, which appeals to people’s rational self-interest, the second place was taken by the rational-empirical change approach. The “soft” strategy of the normative-reeducative approach as represented by participation and facilitative support, however, which in literature and empirical studies has been associated with increased readiness for organizational change (Szabla, 2007; Choi and Ruona, 2011), did not play a role for either reduced resistance or implementation success. Contrary to Beer and Nohria’s (2000) explicitly recommendation to combine “hard” and “soft” strategies to change, in this study the combination of power-coercive and rational-empirical approaches proved to be the most advantageous.
Theoretical and Practical Contribution
this research can lead to the establishment of a refined instrument to measure the implementability of HRM change programs. For the time being the survey questions can be used as a checklist for HR managers when designing new HRM tools; additional modifications and testing are needed to develop them into a validated tool that can serve both practitioners and researchers.
Limitations and Suggestions for Future Research
the program, as intended, but a property of the line managers. Furthermore, this dysfunctional scale, even though reliable by itself, could not be combined with the lack of resistance scale; this was indicated by a low combined Cronbach’s alpha and by scores on different factors in factor analysis. After removal of the two ability questions, the remaining rudimentary construct was limited to absence of resistance in line managers, which gave the variable an inadvertent connotation of readiness for change. Implementability as a concept, however, comprises features related to the change program, not characteristics or attitudes of the change recipients. This dysfunctional operationalization of the variable may also be the reason why no mediation effect could be detected. The precise conceptualization of implementability needs reconsideration, based on which valid scales can be adopted, adapted or developed for further testing.
More research is also needed to gain a better understanding of the full process leading to implementation success. A larger number of predicting factors should be included into the model and be analyzed for possible interactions; especially contextual variables should be tested for moderating effects. In addition, the direct link between politics and implementation success deserves further inquiry. I see value in drawing from literature and empirical studies on power and change, and in investigating the mechanism for how this antecedent contributes to implementation success. In this context, more insight is needed into the inter-relationship between politics and participation, and how the former can cancel out the effect of the latter. Finally, the present study has been conducted in the Netherlands with Dutch HR professionals in Dutch companies. In Dutch national culture, discourse, agreement and mutual understanding form integral parts of professional and private socialization. Results regarding the predominant role played by politics may be different for different cultural environments; this question could be probed by repeating the research in other countries.
CONCLUSIONS
intra-organizational support and content clarity. Participation and the facilitative support of the HR department only were relevant predictors of successful HRM implementation when looked at by themselves; in the presence of the other factors, however, they became non-significant. Thus, despite of the overwhelming amount of literature advocating participation as the essential means to assure support for a change measurement, results indicated that considering political sensitivities and paying attention to the user friendliness of the program contents were the strongest unique drivers of HRM implementation success. Contrary to popular theories, an approach dominated by “power” elements, such as having strategic informal chats with the right people or engaging in goal-oriented networking, turned out to be the best guarantee for a positive change implementation result. Following the results of this study, it should be combined with content-related elements that trigger people’s rational self-interest; examples for such elements are a clear structure, lack of complexity and other features, which make the program easy to use. In general it can be concluded that a combination of power-coercive and rational-empirical strategies proved to be the most advantageous for the implementation of a planned change intervention.
REFERENCES
Aiken, L.S. & West, S.G. 1991. Multiple regressions: testing and interpreting interactions. Newbury Park, CA: Sage.
Armenakis, A.A., & Bedeian, A.G. 1999. Organizational change: a review of theory and research in the 1990s, Journal of Management, 25: 293-315.
Atun, R.A., Menabdeb, N., Saluvere, K., Jesse, M., & Habicht, J. 2006. Introducing a complex health innovation—Primary health care reforms in Estonia (multimethods evaluation). Health Policy, 79(1): 79-91.
Baron, R. M., & Kenny, D. A. 1986. The moderator–mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality & Social Psychology, 51:1173-1182.
