Abstract
Across organizations, most leadership development programs share a similar formalized and centralized structure. This similarity reflects a neglect of the specific contexts in which leadership development programs are implemented. The objective of this article is to understand the insensitivity to organizational contexts that is inherent in the structure of such programs. To do so, we monitored the implementation of a leadership development program in a multinational company from the construction industry. Over a period of 18 months, we collected material from four countries across four continents, with 41 semi-structured interviews, participant observations, and multiple documents. Our analysis builds on Pettigrew’s contextualist framework, combining power and contexts to understand the program structure. We suggest the insensitivity of leadership development programs to contexts lies in power dynamics: the legitimization strategies of the program designers are mostly decisive in shaping the structures of these programs. Contexts are not absent, however. They are present and valued in the legitimization strategies but only as long as they justify the structure of the program as designed by these actors; if contexts call for alternative structures, legitimization strategies neglect or bias contexts. In other words, power strategies instrumentalize contexts in the implementation of a leadership development program. Overall, our research calls for re-embedding the study and practice of leadership development programs into organizations, with a recommendation to those involved in their implementation to broaden the scope of their programs in terms of structures, contexts and power for enhanced program effectiveness.
Introduction
Most leadership development programs (LDPs) look alike across organizations (Pinnington, 2011) . No matter the industry or the sector, chances are that they will have been designed by the corporate services of the company, rely on a panoply of texts, material mediums, etc. to target the training of so-called individual-based leadership skills (Kelly et al., 2006). While rarely discussed, this similarity between mainstream LDPs is a concern: not only is the scope narrowed to the same repetitive leadership refrain (Bolden and Gosling, 2006), but the contextual fit and related program effectiveness are likely to be reduced (Cohn et al., 2005). The thousands of millions invested in leadership development every year run the risk of not generating the expected returns on investment.
How can we understand this lack of program sensitivity to contexts? To date, the few existing research studies examine this issue through the lens of the theories and practices underpinning LDPs. For example, Pinnington (2011: 347) ascribes this similarity and associated insensitivity to contexts to an “absence of everyday discussions on different approaches to leadership and its development, combined with a lack of awareness of the fundamentals of mainstream theories of leadership.” This hypothesis, at the cognitive level, explains that the stakeholders involved in LDP implementation fail to see the assumptions and implications behind their program and end up replicating the same “leadership concept” (Probert and Turnbull James, 2011: 138) of the dominant practices without any specific fit to contexts.
Our research develops a complementary perspective based on empirical grounds. First, rather than capturing this similarity between programs through underlying theories and practices, we look at their structures. Second, building on Pettigrew’s contextualist framework (1985), we do not offer a cognitive explanation but rather look at the interplay of power and contexts to understand the program structure. Our research question is therefore: How do power dynamics contribute to the insensitivity of the LDP structure to organizational contexts?
To answer this question, we studied the implementation of a coaching program aimed at developing leadership in a global company from the equipment industry: CONTRUCT. 1 We collected material from four countries across four continents over 18 months (June 2015–December 2016), with combined analysis of 41 semi-structured interviews, 5 days of participant observation, and multiple documents, including 14 PowerPoint presentations made by the program designers.
The article is structured into four parts. First, we present some of the reasons suggested in the literature to explain why context is overlooked in leadership development, and we then introduce our analytical framework: Pettigrew’s contextualist framework and the conceptual complements we are proposing. Second, we describe our research method: a case study-based with consideration of the temporal perspective. Third, we present our findings, especially the prevalence of legitimization and delegitimization strategies deployed by the actors in charge of program implementation, and the limited influence of contexts. Fourth, we explain the insensitivity of program structure to contexts in light of the instrumentalization of contexts, which attests to the political dimension of the LDP implementation in a global company; this opens to our recommendation, i.e. a broadened scope of programs in terms of structures, contexts and power.
Re-contextualizing leadership development in organizations
Why is context overlooked in leadership development?
LDPs are increasingly being used as a way to ensure the human capacities expected to sustain the future of companies. Despite being on the rise, they look mostly similar. First, in terms of practices: they take the form of 360-degree feedback, coaching, and/or mentoring (Pinnington, 2011). Second, in terms of their underlying theories: they mostly convey a similar perspective of leadership, building on charismatic and transformational approaches to leadership. Leadership is narrowed to an individual rather than collective skill (Day, 2001), with a focus on intrapersonal competence. Third, these programs are likely to have been implemented by a central corporate team (we will talk about centralization) (Conger, 2010), and rely on a panoply of formal mediums (texts, flyers, intranet, etc.) (we will talk about formalization) (Gagnon and Collinson, 2014).
When examined, this similarity in leadership development leading to a lack of sensitivity to contexts is captured at the first two of these levels – practices and theories – and is attributed to either an absence of discussion on the underlying leadership theories (Pinnington, 2011) and/or over-reliance on competency thinking in the programs (Bolden and Gosling, 2006; Carroll et al., 2008; Probert and Turnbull James, 2011). In other words, a cognitive hypothesis is usually advanced in the study of the similarity between LDPs. We empirically explore an alternative proposition that distances itself from this literature in at least two ways. First, we study programs not in terms of their practices or underlying theories, but rather their structure, detailed below as centralized and formalized. Second, our conceptual framework, introduced in the next section, echoes calls to better explore power dynamics and contexts in leadership development (Collinson, 2014; Edwards et al., 2013; Gagnon and Collinson, 2014).
Pettigrew’s contextualist framework: Contents, processes and contexts
We chose the contextualist approach (Pettigrew, 2012) which “challeng[es] … the largely acontextual, aprocessual and ahistorical research on organizational change” (1037). This framework, which considers the temporal dimension of the phenomena under study, appeared particularly suited to the study over time of the implementation of an LDP that represented organizational change for the multiple stakeholders involved.
Pettigrew’s analytical framework combines three interrelated concepts – contents, processes and contexts – that the researcher endeavors to understand. Contents, for the purposes of our study, refer to the structure of an LDP. Processes are the power dynamics that develop among stakeholders. Finally, contexts are the organizational (internal) or environmental (external) features (see Figure 1, terms in bold capital letters).
Analytical framework: three interrelated key concepts and theoretical complements.
While contextualism offers a “meta-framework” (Leavy, 1996: 442), it calls for more specific theories to analyze each of its three components – structures, power dynamics and contexts – and their interrelationships.
In our study (see italics in Figure 1), we capture structures through the degree of centralization and formalization of the LDP. Formalization refers to the extent to which a program relies on written text (rules, requirements, specifications, contracts, codes of conduct, etc.) or materials (access to equipment, computer software, etc.) (Nizet and Pichault, 2012). Centralization can be measured by the power of stakeholders over decisions: we will talk about a centralized program in which decisions are controlled by a small number of people, if not just one person, situated on top of the organizational hierarchy, as well as a decentralized program in which power is distributed among numerous stakeholders, situated at lower levels of the hierarchy (Mintzberg, 1983).
To analyze power dynamics, we refer to an approach to power inspired by Weber emphasizing agency (1978), and we differentiate between pressure strategies and legitimization strategies (Bourgeois and Nizet, 1993; Pfeffer, 1981). Pressure strategies involve threats used to obtain what people want. Legitimization strategies describe the ability for actors to make connections between what is perceived as legitimate in a given context and what they are trying to obtain (legitimization), or what they are trying to distance themselves from and disqualify (delegitimization).
To analyze contexts, we refer to the distinction made by Pettigrew (1985) between internal and external contexts. Internal contexts 2 are represented by different dimensions such as the size of the organization, degree of centralization of decisions, positions held by different people (C-suite, analysts, operators, etc.). External contexts correspond to the different segments of the organizational environment – economic, political, cultural, etc. – (Hatch and Cunliffe, 2012) – and to their characteristics, i.e. their degree of complexity, stability, etc. (Mintzberg, 1978).
Research method: Temporality and triangulation in a case study
Research design and data collection
Our approach to LD as an organizational phenomenon supported the decision to conduct a single case study (Hlady Rispal, 2002). Given the prevalence of highly centralized and formalized LDPs depicted in both the analytical and normative literatures, it was only logical for us to analyze a program with such a structure. Such a program had recently been introduced in the global company CONSTRUCT where our access to information was facilitated by the support of the person – let’s call him Joe – in charge of implementing the program; he was an alumnus from the same university as one of the researchers. CONSTRUCT is a civil engineering company. It designs and builds large-scale infrastructure such as bridges, tunnels, hotels, shopping centers, etc. For the past 10 years, it has experienced significant growth, notably due to the takeover of foreign companies as part of a diversification process. The company is located in 30 countries across the world and has about 40,000 employees. Each site tends to specialize in the construction of particular types of infrastructure. Facing an increasingly globalized market, the executive team aimed to evolve CONSTRUCT into a “global company” during the period under study, enabling it to move from building various infrastructure projects to the manufacture of components that fit the products of the multiple sites where the company is active. This shift, which requires improvements in communication and coordination between the different locations and services, represents a significant challenge, especially given the firm’s linguistic and cultural diversity. To reflect its importance, a “director of diversity” was recently appointed.
In 2014, the executive team of CONSTRUCT created a Corporate University, the CONSTRUCT University (CU), responsible for designing different corporate programs, including a coaching program. This coaching program called “Coaching for Leaders” (CFL) was designed to “support the [CONSTRUCT]’s Group strategy by reinforcing the identified leadership competencies connected to the organisations strategy, individual development, and career development.” According to internal documents, the program was further portrayed as “a very efficient way to leverage leaders’ performance development” to help “deal with the complex situations [leaders] face in their roles”, when the “tough economic climate invites us to find cost-efficient coaching solutions” and a “coaching culture can strengthen our attractiveness as an employer who supports individual development.” Typical competences included: “dealing with ambiguity”, “standing alone”, “listening”, “courage”, “strategic agility”, etc.
Under the internal process, a pilot project that Joe supervised was launched to study the viability and parameters of the program. It is this pilot phase for program implementation that provided our research material. It began in June of 2015 when the issue emerged and took place in four countries located in as many continents: United States, Netherlands, Brazil, and China. It proved successful, and the decision to extend the program to six additional countries was made in late 2016, when the data collection period ended.
The 18-month data collection period was beneficial in several ways for our study. First, as the program was being implemented, new events occurred, which meant we had to keep a timeline and update obsolete information (e.g. initially there were six coaching candidates in China, and three withdrew). Keeping track of these events was an initial way to integrate temporality into the study. Second, our analytical framework was progressively refined, making relevant events that previously seemed insignificant, in particular those that occurred before the start of the data collection period, ranging from a few months before to several years; these reconstructions operated by stakeholders, or “retrospective” were a second way to approach temporality, with a longer timeframe in mind, on top of the “real-time analysis” of the material (Pettigrew, 2012: 1308). Third, new data information was systematically compared against the information already obtained, either through another interview or document, as a way to validate, complement, or nuance it.
Several complementary types of data were collected. First, we conducted, recorded, and transcribed 41 interviews with 24 people – some, like Joe, were interviewed more than once – totaling 32 hours. There were five categories of interviewees: (a) the C-suite, especially executives from the HR department (seven people); (b) members of the Corporate University, including Joe, in charge of program design and implementation (6 people, 19 interviews); (c) local HR Managers from the four countries selected for the pilot study (four people); (d) members of Global Management Consulting (GMC), the consultancy firm in charge of coordinating the program under the control of the CU, as well as a local coach, who was not a member of GMC (three people); and (e) several participants of the pilot phase of the program (four people).
Another point of note is that we had access to plenty of printed and digital documents used in the context of the program, including 14 PowerPoint presentations, totaling some 325 slides, used by CU members in their interactions with the steering committee or with local HR managers as well as future coaches and coachees. Finally, the on-site researcher attended six work meetings with CU members, local HR, and/or program participants, where Joe either presented the program (using the PowerPoint slides) or fixed issues related to program implementation.
These research data are highly complementary, and the relevance of the different elements varies according to the moments and objectives of the analysis. For example, some interviews – in particular with local HR managers – accurately describe the structure of the LD practices formerly implemented in their countries. The PPT presentations are also very pertinent when it comes to describing legitimization strategies insofar as they were designed to inform future program users, and to convince them of its soundness. And given that one of the analytical variables is the degree of formalization, it is only natural to pay attention to texts.
We conducted our data analysis over two phases building on Pettigrew’s framework as introduced above. First, meetings between the researchers allowed our propositions to be refined: we discussed the different modalities of the processes at play (pressure and/or legitimization, legitimization strategies referring to one theme or another), and the specific dimensions of the internal and external contexts likely to shed light on the low sensitivity of program structure to organizational contexts. Once these analytical elements were in place, the whole body of material was analyzed, which allowed us to confirmation, illustrate, and nuance our conceptual propositions. More precisely, a thematic content analysis was carried out for the 325 slides, documents, and interviews, to weigh up the importance – or highlight the absence – of the different themes (such as the various legitimization strategies or contextual elements).
Coaching for Leaders as a centralized and formalized program
CONSTRUCT’s CFL program did not start from scratch. Since the early 2000s, several countries had already been offering different types of leadership development practices on demand. In most cases, their existence resulted from HR managers, on their own initiative, requesting coaching training. They then answered requests from managers who wanted to use the services of an internal coach or seek guidance from an external coach. These initiatives are thus mostly decentralized and largely informal.
In two countries, the United States, a pilot country, and France, who was not part of the pilot countries, such practices within the firm have evolved recently toward greater formalization. In France, a coaching network has been formed together with a “coaching office” where members of the network receive coaching requests, discuss demand, establish a contract, and then direct applicants toward the most suitable coach, whether internal or external. As of December 2016, this network was still alive, but had lost momentum. In the United States, a similar network exists, but most coaching services go through external coaches.
The CFL program both builds on and differs from existing programs, but appears largely centralized and formalized, something Joe encapsulated as a “structured coaching approach led by the CU.” First, the program is highly centralized. Most decisions are made and implemented by a relatively small number of people – the C-suite (especially the HR VP), the related project steering committee, and the CU. Some are major decisions and actions, such as launching the project, conducting a pilot study in four different countries, supervising the recruitment of coaches in each of the four countries of the pilot study (with the support of local HR managers and GMC), and identifying a few executive managers – about six per country – to participate in the pilot phase (again with the support of the local managers and GMC). Others are smaller in scope, such as the process for requesting a coach, the number of coaching sessions, their duration, etc. Centralization can also be seen in the CU’s ability to follow the progress of each coachee through a chart that Joe has in his office. This chart leads us to the issue of formalization.
The program is indeed highly formalized. The CU registers all of the individual or group contacts (by phone or Skype) involved in the CFL program in written documents. Some of these documents are then directed to the coaching users. This is accomplished through a virtual object, the “Operator software that represents the interface through which coaches will find all the relevant information to begin and conduct their coaching sessions, designated by the term “toolkit.” This toolkit provides coachees with general expectations regarding coaching, registration steps, and evaluation forms. From reading the different documents that the applicant must fill out throughout the process – about 10 in total – we learn that a standard coaching program comprises 4–6 sessions each lasting 2 hours and is preceded and followed by a 3-hour third-party meeting with the coach, coachee, and the manager or HR manager. It is also through this platform that future recipients enroll in the coaching program and go through the different required steps, up to the final evaluation, which is completed with the manager three months later and relates to behavioral changes.
In contrast with the high degree of centralization and formalization of the new LDP, CONSTRUCT’s previous practices – some of which are still in place – have significantly different structural features. Some practices, such as coaching on demand implemented in a few countries, are both decentralized (decisions are made by multiple actors, such as coaching recipients, internal coaches, local HR Managers, etc.) and informal (with very few written materials to guide the practice). The French and American experiences showcase a certain degree of formalization (see the existence of a “coaching office” in France) and are relatively decentralized: most decisions – such as the choice of the coach – are made by employees with lower hierarchical status.
In sum, in contrast to previous practices, the recently launched CFL program presents a high degree of centralization and formalization, typical of mainstream LDPs.
Findings: The prevalence of power and the limited influence of contexts in shaping the CFL program structure
Power: A prevalent influence mostly through legitimization strategies
Legitimization: A wide variety of (de)legitimization strategies. The CFL program has given rise to an abundant number of legitimization and delegitimization strategies adopted by the people in charge of its implementation, mostly demonstrated by CU members. These strategies are directed at both “the top” of the organization chart, i.e. the steering committee responsible for granting the extension of the program to other countries, and “the bottom,” i.e. the company’s managers who are the program’s potential future clients.
The first
3
set of legitimization strategies involve associating the formalization and/or centralization of the program with positive connotations, such as “maturity,” “quality,” “clarity,” “visibility,” or even the anticipated “success” of the program as displayed on slides. The Senior VP for Talent Development mentions that one of the “success factors” or “drivers” of this LDP is its “structured” nature, the goal being to “structure the coaching program as much as possible.” Several PowerPoint slides note that “in many countries, coaching already exists with different levels of maturity,” maturity here referring to the fact that in some countries, coaching “is not offered in a professional way,” to quote the Senior HR VP. However, here is one noticeable counter-example: the VP of the Leadership Academy distanced herself from this overall negative image of past programs by mentioning that in France participants had good visibility of the program, which is a desirable quality. Otherwise, interviewees explicitly associated “coaching quality” with its dimensions of formalization and centralization: In some countries, there were ad hoc requests for coaching, and people had to find their way to a coach. [The rationale behind the present program is] to control the quality of coaching that is offered through a more structured offer that can be monitored centrally. (emphasis added)
Among the positively connoted terms, those that refer to “leadership” – such as “leader,” “leadership,” “leading,” etc. – deserve specific attention. This topic has been prevalent – and valued – over the years in the company. For example, it is used to distinguish between hierarchical levels, with terms such as “leading business,” “leading leaders,” “leading people,” and even sometimes “emerging leaders” as displayed on slides. The leadership theme is also present and valued in the external context (see fourth strategy below). These internal and external valorizations of “leadership” lay the groundwork for a second set of legitimization strategies: the reliance on “leadership” to promote the program within the company. Indeed, “leadership” is closely associated with the coaching program to be implemented and seems to support its promotion. The program is called “coaching for leaders” and the associated material (PPT presentations and internal documents designed for its beneficiaries) makes abundant references to the leadership theme. For example, the PPT slides indicate that the program is supposed to “accelerate leaders’ performance,” that CONSTRUCT’s partner consulting firm works “for all leaders all around the world,” etc. Similar to what we found for the first set of strategies, these references to leadership are not used to characterize the company’s previously implemented programs.
The program benefits from positive comments, as do the actors in charge of its implementation and smooth functioning. Promoting these people in this way falls under the third set of strategies. The program’s designers and supporters make abundant use of words with connotations of expertise such as “University,” “expertise,” “professional,” etc. GMC’s website presents itself as being a “world expert” in its domain. The word “University” appears in the header of each written – printed and online – document connected to the program. Slides inform future program participants that they will “work with a professional senior coach.” Several of the PowerPoint slides mention the research collaboration established with one of the authors of this article, outlining that CONSTRUCT thus benefits from the “expert analysis” of this “professor,” etc. As in the other two sets of strategies, such positive phrasing is not used to characterize the actors in charge of the alternative program, who are instead described using more neutral terms, such as “ways of approaching coaching.” The only exception is the reference to “certified coaches” from the French network. We suggest that overall such positively connoted terms reinforce the legitimacy of the centralized actors in charge of the program, which in turn contributes to the program’s centralized structure.
The fourth set of strategies sustaining the program’s centralization and formalization relates to its finances. Several arguments can be made here. First, references are made to similar centralized and formalized programs in comparably large companies that, according to CU members, generate significant returns on investment. By referring to several studies conducted by consulting firms, the ROI is estimated to range from 529 to 788%, though one slide mentions: “Other studies show that the ROI calculation is often unreliable and irrelevant.” This preoccupation with ROI goes hand-in-hand with close attention paid to costs: numerous slides display the cost of operations already implemented and/or of future expenses that need to be considered in the decision whether or not to extend the program to other countries, with several options such as using GMC. This concern for costs also emerges in interviews, as in the case of the VP for Learning who points out the “hidden costs” of more informal and decentralized practices: “For each coaching request, a different person is solicited to provide information, which means that potentially we have 10 people, full time, here to provide support. Even if costs are not always visible, they can rapidly spiral.”
The fifth set of strategies connects the formalization of the program to the recent strategic orientation that CONSTRUCT has been pursuing with the objective of becoming a “global firm.” The US Director for Talent presents this new orientation as a “big change,” as if their “world had become global suddenly”, accompanied by “structural changes.” From the perspective of the VP Technology Platforms, this requirement to “work globally” “increases the complexity of product development” since “now you need to take into account greater complexity of production, i.e. the same components need to work for more than one product.” This also implies improving lateral communication between the different entities of the company. However, the CFL project team interprets this objective of globalization in a very specific way, seeing it as an argument in favor of unifying and standardizing processes. Indeed, the CFL program intends to offer a standardized and global program that fits a “one company approach” likely to “provide unified quality and global processes” that contrast with “ways of working with coaching [that] have been carried out in different countries without any global guidelines or unified group strategy”, as Joe explains. The Dutch HR Manager compares the CFL program, which offers a “uniform way of approaching coaching,” to former experiences in which “the program used certified coaches but no structure was given.”
The sixth set of legitimization strategies refers not to language, as above, but to symbol manipulation (Pfeffer, 1981). Budget accounts and estimates, aside from the information they provide, also function as symbols that bestow legitimacy on this specific program and demonstrate the commitment and engagement of those in charge of it. In the same vein, all the material (slides and documents) dedicated to action or to the evaluation of the coaching process also contains symbols likely to further legitimize the program as it inherently matches the expectations, values, and interests of the steering committee (who will see in these figures/symbols signs of rigor and compliance with internal rules and processes demonstrated by the CU), as well as those of the program participants (who will interpret these as proof of appropriate preparations in operational terms and as a signal of the CU’s competence and trustworthiness, etc.).
We have identified a total of six strategies – five using language and one relying on the manipulation of the symbols – which, depending on the case, legitimize the centralization and/or formalization of the program, i.e. make it mainstream. At the same time, several of these strategies delegitimize previous LD practices, either by not associating the same positive connotations with them, by doing so but to a lesser extent, or by attributing opposing characteristics to them, i.e. with negative connotations.
Several authors (Bourgeois and Nizet, 1995; Pettigrew, 1985; Pfeffer, 1981) have explored the specificity of legitimization strategies compared to other modes of influence. Here, we follow Pfeffer’s hypothesis (1981) that legitimization has the effect of mobilizing support and quieting opposition, thus making the relationship not very conflictual. This hypothesis seems relevant in characterizing the attitudes of the managers involved in the pilot phase of the program. Some obviously supported the program. According to the Brazilian HR manager, managers were “very excited and very eager to get the program underway”; according to Dutch HR manager, the number of applicants outweighed the number of available spots for the pilot phase. Other managers solicited for the pilot phase were less forthcoming, albeit without showing signs of opposition, i.e. without exercising counter-power. The Chinese HR manager mentioned a financial director who was supposed to enroll in the pilot phase but in the end expressed other priorities and was unable to allocate time to participate. In the same country, several designated candidates were late enrolling and would likely have to give up their spots to colleagues. In fact, we identified only one clear case of counter-power: one manager in the Netherlands expressed dissatisfaction with the two coaches presented to him and requested – successfully – that he be allowed to work with his own coach.
While overall the implementation of this centralized and formalized program was not very conflictual, some targeted issues were more conflictual and gave rise to another type of power strategy: pressure.
Pressure strategies for targeted issues
When legitimization strategies proved insufficient to justify program implementation, we observed the use of pressure. This concerned targeted issues, essentially those relating to costs and the CU’s control over the program.
First, the steering committee members pressured the CU to reduce program costs, suggesting that they might not grant the authorization to convert the pilot into a program should it prove too expensive. This is one of the “gate stopper[s]” that would “stop the project” that Joe had identified in several of his PPT presentations (even displayed with a stop sign). He spoke on several occasions of the threat that the Steering Committee might not give its “green light” to the continuation of the program in the four pilot countries or to its expansion to other countries.
Second, pressure strategies also appear in connection with the more complex issue of the CU’s control over the program (respondents used the word “grip”). The main way in which the CU maintains this control is – a bit paradoxically – to use an external provider, i.e. the services of the international firm GMC. GMC monitors the entire process on the CU’s behalf, and this outsourcing was contested by certain local HR managers who did not understand the rationale for changing the previous formula as it required them to end long-standing supplier relationships. Local HR departments, especially in the US and China, would rather use their own suppliers, but need authorization from the CU to do so. This leads to pressure strategies, with threats coming from multiple stakeholders: local HR managers threatening to resort to local suppliers; the CU threatening not to accept local suppliers chosen on a more decentralized basis, etc.
In sum, our analysis shows that the centralized and formalized structure of the CFL program is largely shaped by power dynamics, in the form of two sets of strategies. First, a wide variety of legitimization and delegitimization strategies, which go hand-in-hand with relationships that involve little conflict; and second, a more limited set of pressure strategies that correspond to more conflictual relationships.
Contexts: Small direct influences but strong presence in legitimization strategies
If processes (power dynamics) largely shape the CFL program’s centralized and formalized structure, what about contexts? Two main results emerge from our data. First, the direct influence of contexts appears limited. Second, these same contexts are largely used in legitimization strategies.
Limited direct influences of contexts
Contexts are identified as having two direct influences in the shape of the program structure, both attributed to internal contexts.
The first relates to the characteristics of the decision-making process that led to the adoption of the program. This process originated in the C-suite, which mandated the CU to design and implement the CLF pilot phase. Right from the beginning, we see that these central actors – the C-suite and the CU – are in control. This centralized decision-making process, typical of CONSTRUCT’s internal context, thus exerts a direct influence on the centralized dimension of the CFL program.
The second direct influence relates to the specific position held by the CU in the internal context of the company, which, in fact, is a dual position. First, it appears that the CU fundamentally occupies, in Mintzberg’s terminology (1978), an analyst position. And Mintzberg (1983) shows how the objectives usually pursued by analysts, as well as the specific competencies they hold, generally encourage and reinforce a type of formalization that the author encapsulates as a “standardization of processes,” i.e. a type of formalization that tends to control and standardize behaviors. But at the same time the CU holds a position as service vendor: it is just one among many other potential program providers. The clients – in our case, local HR Managers – can choose alternative local suppliers outside of CONSTRUCT. It is even to their advantage to do so when the internal offer is more expensive than the equivalent service obtained outside of the company. This latter position tends to reinforce the power of local HR Managers and encourage a more decentralized type of program. In sum, the dual position of the CU has ambiguous effects on the program structure: the CU’s position as an analyst tends to reinforce its centralized and formalized structure, while its position as a vendor works in favor of other structures.
Contexts highly present in legitimization strategies
While contexts have a limited direct influence, they are omnipresent in the (de)legitimization strategies detailed above. References are made to internal contexts when respondents talk about existing coaching practices already in place in the organization (strategy 1), the financial side of the program, and especially its expected ROI and the hidden costs generated by the previous initiatives (strategy 4), and CONSTRUCT’s shift towards becoming a “global firm” (strategy 5). References are also made to external contexts, such as the collaboration between CONSTRUCT and external partners like GMC or the University (strategy 3), similar programs implemented in other large firms (strategy 4), and the globalized state of the current economic context (strategy 5). Next, we discuss these findings.
Discussion: From the instrumentalization to the consideration of contexts in LDP implementation
First, we discuss how, in the case of CONSTRUCT, program designers instrumentalize contexts in their power strategies to support a formalized and centralized program. Then, from a more prescriptive perspective, we call on all stakeholders involved in LDP implementation to broaden the scope of the program structures they take into account, and beyond, the scope of contexts and power, for enhanced contextual fit and increased effectiveness.
The instrumentalization of contexts to support a centralized and formalized program structure
Our study of the design and implementation of CONSTRUCT’s coaching program shows that contexts have a limited direct influence while power dynamics are prevalent, essentially in the form of (de)legitimization strategies. But in these (de)legitimization strategies, contexts are very present. Indeed, contexts are valued by those who rely on legitimization strategies as long as they serve their interests, which in this case is to implement a program with a specific structure, i.e. centralized and formalized. For example, the expected ROI that this centralized program is supposed to generate refers to the internal context of a focus on cost-saving, while the external context of programs with a similar structure implemented in large firms is cited as a way to legitimize centralization and formalization.
And when contexts do not support the program structure that actors want to implement, they are either neglected or biased. A case of bias concerns the context of the “global company,” interpreted by those in charge of the coaching program as a call to uniformize LD practices, and thereby formalize the program, rather than allowing for more customization in each department, division and country. Another case of bias relates to the financial context, and particularly the “hidden costs” which those in charge of CFL associate with more decentralized and less formalized practices already in place in the company. They suggest that these costs are “significant” without quantifying them, while proven techniques are available to do so (Savall and Zardet, 2008). Nor do they provide budgets for alternative program structures.
An example of neglect is the (quasi) absence of references by interviewees in charge of program design and implementation to the internal context of linguistic and cultural diversity. According to several of them, in particular local HR managers from the countries involved in the pilot phase, these aspects of diversity are an important dimension of the company’s context. However, they are not taken into account in the program design, and consequently do not appear in the legitimization strategies. In fact, we only came across two documents from the coaching program that mentioned this issue: one slide (out of the 325 reviewed) and the front cover of a written document. These portray people from culturally diverse backgrounds who are busy cooperating. No text accompanies these representations, as if the actors in charge of these programs were not quite able to clearly articulate this cultural diversity.
In both cases, we see this as the instrumentalization of contexts by power: positive instrumentalization that involves selecting and valuing contexts that obviously justify CFL’s centralized and formalized structure; and negative instrumentalization when contexts are neglected or biased because they do not support this structure.
Our results reveal that power strategies encompass both legitimization and delegitimization, and the same applies to the instrumentalization of contexts. Some instrumentalization strategies are based on legitimization, such as the ROI example, while others involve delegitimization, such as the example of the hidden costs generated by previous practices. From a temporal perspective, we can say that delegitimization strategies mark a break, or at least some distance, from the past, while legitimization strategies focus on the future as imagined by those in charge of the program.
Table 1 summarizes the main results in response to the question of how power dynamics contribute to the insensitivity of programs to contexts. It shows that two types of power strategies – pressure and legitimization – are used in the implementation of the coaching program at CONSTRUCT. While pressure strategies and some legitimization strategies do not refer to contexts, most legitimization strategies use them as a way to secure acceptance of the coaching program’s formalized and centralized structure.
Power strategies developed by the main actors in charge of CFL program implementation.
CFL: Coaching for Leaders; CU: CONSTRUCT University.
Broadening the scope of LDP structures, contexts and power for enhanced effectiveness
Our analysis suggests an alternative way of thinking about and conducting the implementation of LDPs. It not only recommends broadening the scope of available programs to consider alternative program structures, but also to the scope of contexts as well as of power dynamics.
Broadening the scope of structures
We have seen that the LDPs previously implemented at CONSTRUCT had structures that were different from those of the newly implemented CFL program. Various researchers have also shown that such programs also exist in other organizations. For example, Peel's (2008) study about coaching and mentoring programs implemented in small and medium-sized enterprises in Wales reveals a centralized but informal type of program. Centralized since executives largely control the decisions, in particular those related to the appointment of program beneficiaries; but informal given that these decisions are based on criteria that are not very explicit and are rarely written down, but rather have to do with the quality of the relationship with the beneficiary. An opposite structure – decentralized and formalized – was identified by Fatien Diochon et al. (2018) in a company from the automobile industry. In that case, on their own initiative, a group of middle managers obtained authorization from the C-suite to create a space dedicated to learning how to lead and collaborate more creatively. This decentralized space appears rather formalized, however, since it is promoted through posters and flyers and is equipped with numerous artifacts – furniture, clothes, etc. – likely to boost creativity. Finally, a study by Ivanaj and Persson (2012) on mentoring practices in large French companies offers an illustration of a program that is mostly informal and decentralized. The program in question is implemented at the initiative of the interested party (the mentor or mentee) and meets informal needs of a professional nature (advice, idea-sharing, recommendations by superiors, etc.) or a psychosocial nature (role-modeling, managing emotions, etc.). Our research thus invites LDP stakeholders to look beyond the dominant centralized and formalized programs and to consider a broader range of structures over an LDP implementation.
Broadening the scope of contexts
Broadening the scope of structures that we just mentioned should allow stakeholders involved in LDP implementation to come up with designs that better fit internal and external organizational contexts. These organizational contexts are indeed very diverse. External contexts, first, encompass multiple sectors: legal, economic, technological, political, cultural, etc. (Hatch and Cunliffe, 2012) – likely to each vary across several dimensions: stability (stable vs. dynamic), simplicity (simple vs. complex), hostility (hostile vs. friendly), etc. (Mintzberg, 1978). Then, internal contexts are equally diversified, notably in terms of the size of the organization, the technology used in production, the degree of complexity of the tasks handled by operators, etc. (Donaldson, 1999). Organizations, in particular large ones, can in addition have multiple departments or divisions and markets in several countries; other said, they can be very heterogeneous. This research thus suggests to all stakeholders involved in the implementation of an LDP that they should pay careful attention to the characteristics of their organization’s internal and external contexts to determine which ones are essential, and thereby adapt the program structure. It also suggests taking into account the heterogeneity in the organization, which will require the implementation of a program with different structures in sync with the different contexts of the organization.
Broadening the scope of power dynamics
Taking into account these multiple structures – including decentralized structures – and multiple external and internal contexts will extend the power dynamics to additional actors beyond those observed in the case of CONSTRUCT, where they were mostly manifested by the CU. Depending on whether they are external (e.g. consultants) or internal (e.g. the executive team, or HR managers), and for these internal actors, whether they hold a central or more peripheral position, stakeholders are likely to (de)legitimize a given structure depending on their interests. Power dynamics broadened in this way do not contradict the requirement of contextual fit. Indeed, as indicated by Mintzberg (1983), political games usually generate intense debate covering all aspects of an issue. In short, it is very likely that a broader range of different contexts will be carefully examined, which will reduce partial selections and biases, and such adjustments in program structure(s) are likely to increase their effectiveness (Donaldson, 1999).
Conclusion
The article starts with a question on the lack of sensitivity of LDPs to the organizational contexts in which they are implemented. The answer lies in the prevalence of power dynamics and in particular in the (de)legitimization strategies that instrumentalize contexts. Those responsible for implementing LDPs often limit their scope to only centralized and formalized structures. Furthermore, they value the contexts that justify these same structures (positive instrumentalization) and neglect or bias those that support alternative structures (negative instrumentalization). This means the structure lacks fit with the organizational contexts, which consequently results in a loss of effectiveness. These main analytical results can be framed within Pettigrew’s framework (see Figure 2).
LDP Limited scope of LDPs as depicted in our analyses.
We can also use Pettigrew’s framework to represent the practical recommendations of this article in terms of LDP implementation. First of all, the article suggests that those involved in implementation should broaden their scope by considering different program structures. We further recommend conducting thorough analysis of the different internal and external contexts for enhanced program effectiveness. This fit to contexts and associated choice among different program structures is likely to be debated by a plurality of actors (see Figure 3).
Broadened scope of LDPs as recommended.
In sum, our research calls those involved in the implementation of an LDP to broaden the scope of each of the three constituents of Pettigrew’s model: structures, contexts and power. It therefore evidences the need to think of LDPs as embedded organizational phenomena, and to re-embed their study into organizational dynamics.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.