The value-rational organizational form as the context for distributed ambidexterity By

Authority

The key organizational paradox of DA in the authority dimension is in assuring the local autonomy required for creative exploration at the same time as the centralized control required for efficient exploitation. The former calls for “distributed” leadership (Cullen & Yammarino, 2014) and the latter for hierarchical leadership (Von Krogh, Nonaka, & Rechsteiner, 2012). The value-rational organizational form synthesizes these demands by enacting the principle that authority—whether local and distributed or centralized and hierarchical— should be based on widely-acknowledged ability to contribute to the organization’s shared purposes (as suggested by Aime, Humphrey, DeRue, & Paul, 2013). This principle can be operationalized by ensuring that activities are chartered appropriately and that structures are appropriately matrixed. Under these conditions, the value-rational organization’s authority structure can take a form we call “participative contingent centralization”: here authority in exploratory decision-making is participative by being distributed, while authority in exploitation decision-making is participative by ensuring that hierarchy operates participatively, and the locus and extent of centralization are contingent on how best to serve the shared purposes. If members share the higher purposes of the organization (with the help of the Latency/values subsystem), the resulting form of hierarchy (needed for the efficiency and exploitation) will not be experienced as alienating and will not undermine creative exploration, but instead will be experienced as “autonomy-supporting” (Deci & Ryan, 1987) and will enable DA.

Much of the scholarship in our field would be skeptical of the concept of participative centralization, assuming that centralization and participation are polar opposites (e.g. McCaffrey, Faerman, & Hart, 1995). However, as these constructs have been defined more precisely in organizational research, they are not opposites. Centralization is assessed by ascertaining the lowest hierarchical level at which a decision can be made without prior consultation with a superior (Pugh & Hickson, 1976). Participation is assessed by ascertaining the lowest hierarchical level at which real influence on the decision is exerted (Hage & Aiken, 1970). Centralization and participation are thus better conceptualized as independent, orthogonal dimension of the authority structure. The value-rational organization is, we argue, high on both dimensions.

Adler et al. (1997) describe various management techniques supporting participative contingent centralization at the NUMMI auto assembly plant, and these techniques’ impact on the firm’s ambidexterity. This study documents the coexistence and mutually supportive relationship between (a) a distributed form of leadership operative in the shop-floor and engineering teams aiming for radical innovation (in that setting, radical innovation took the form of major model changeovers), (b) a participative form of hierarchical authority where shop-floor personnel, both directly and through their union, had strong upward influence over centralized, organization-wide policies, and (c) a process of negotiating the charters of the various exploration- or exploitation-oriented teams in light of their possible contributions to the shared purpose of “producing high-quality, low-cost cars.” This participative contingent centralization supported a high degree of ambidexterity—both distributed and other types—evidenced in the organization’s remarkable combination of efficiency and flexibility.

The matrix type of authority structure is a second key innovation in management technique that has operationalized contribution-base authority and thereby facilitated participative contingent centralization and DA. DA requires that contributors attend simultaneously to the efficient exploitation of established functional expertise and to creative exploration of new capabilities and new products. When organizations attempt to orchestrate such synthesis efforts by relying on the familiar monocratic hierarchy of authority, the result is typically an overemphasis on just one of these priorities, usually exploitation. DA, therefore, often requires a matrix structure with multiple dimensions of accountability, where the locus of authority on one dimension versus others is contingent on the nature of the decisions that need to be taken (Galbraith, 1994). Matrix structures, however, are notorious for the challenging “organizational politics” engendered by their multiple reporting relationships, and as a result, these structures are difficult to sustain and organizations have suffered many implementation failures (Burns, 1989; Larson & Gobeli, 1987). Nevertheless, competitive pressures have pushed firms to persist in trying to master these challenges, and in many firms today matrix is a taken-for-granted condition; indeed, there has been an evolution over time toward matrix structures with more than two dimensions (Galbraith, 2008; Heckscher, 2007; Strikwerda & Stoelhorst, 2009). The key to mastering these implementation challenges lies in a value-rational commitment to shared purpose and the ethic of contribution: it is this shared purpose that enables individual contributors, functional managers, and project managers to search together for synthetic resolutions of the competing priorities of exploration and exploitation.

Heracleous and Wirtz (2014) offer a nice description of this kind of matrixed participative centralization at Singapore Airlines. Here, the organization’s ambidexterity depended in part on the efforts of a centralized product innovation department. While this department centralized the firm’s larger exploration investments, its centralization had a strongly participative flavor because the department’s teams were staffed primarily by people from the line organization, some of whom spent two or three years on rotation here and some of whom worked on shorter-term or part-time assignments. The work of these teams was organized in a matrix fashion, linking the product innovation staff and the various line functions.

Participative centralization may also underlie the results of several studies showing that transformational leadership contributes importantly to DA within top-management teams (Jansen, George, Van den Bosch, & Volberda, 2008; Nemanich & Vera, 2009). Measured as a combination of intellectual stimulation, inspirational motivation, individualized consideration, and idealized influence, transformational leadership appears to overlap to a considerable degree with our construct of value-rational authority. The exception is the “inspirational motivation” component, which, from our Weberian viewpoint, seems underspecified because it could refer either to an affectually-based charismatic inspiration or to a value-rational type of inspiration. Future research might usefully seek to clarify which aspects of transformational leadership support DA and whether value-rationality is a better predictor.

Kaiser illustration. To facilitate the pursuit of DA and the achievement of its diverse goals, Kaiser has created a sizable “Performance Improvement” staff, which functions in a matrix relationship with their line management client groups (Schilling et al., 2010a). Indeed, Kaiser’s authority structure is extensively matrixed. Managers and non-managerial personnel have learned to accommodate themselves to the unusual degree of organizational complexity required to ensure that both exploitation and exploration tasks are kept constantly in view.

The labor-management partnership helps Kaiser meet its DA challenges by supporting participative contingent centralization across the wider organization. First, the unit-based teams afford important opportunities for distributed leadership and engage aggressive innovation efforts: leadership within these teams is determined by contribution, not by hierarchical position. Second, these teams are chartered through an explicit negotiation process that is anchored in the Value Compass: the teams negotiate with goals with management and union sponsor as a function of their ability to advance one or more of the Value Compass goals (Kaiser Permanente Labor Management Partnership, 2015). And third, to the extent that decisions need to be centralized, a hierarchy of joint labor/management committees governs decision-making, from the national, to the regional, and down to the facility level. While this centralized structure has undoubtedly added organizational overhead, it ensures that the decisions made centrally by higher-level decision-makers are seen as legitimate by lower levels in the organization. The union coalition’s partnership with management in joint pursuit of the Value Compass goals reassures members that that these centralized decisions are oriented towards the organization’s ultimate purposes. This legitimacy has enabled Kaiser to undertake several potentially highly controversial moves without major internal conflict, including several major cost-cutting initiatives, some of which involved large-scale redundancies (which, thanks to advance planning and partnership, were handled by retraining and reemployment).