Organization Design

John Joseph, Oliver Baumann, Richard M. Burton, Kannan Srikanth, John Joseph, Oliver Baumann, Richard M. Burton, Kannan Srikanth, John Joseph, Oliver Baumann, Richard Burton, Kannan Srikanth

• Publisher: Emerald Publishing Limited
• Year: 2018
• Print ISBN: 9781787563308
• eBook ISBN: 9781787563315
• Topic: Business
• Subtopic: Business Strategy
• Index: Business

This volume of Advances in Strategic Management explores emerging trends and contemporary research in the field of organization design. It reflects a renewed focus on the universal problems of organizing—the division of labor and the integration of effort. At the same time, it incorporates new ideas on designs for cooperation in organizations.

The studies published here employ a wide variety of theories and research designs to contribute to this renewal of organization design research, considering collaborative ways of working, organizational learning, and strategic innovation.

PART I
FIT AND COORDINATION

DESIGNING A CULTURE OF COLLABORATION: WHEN CHANGING BELIEFS IS (NOT) ENOUGH

Özgecan Koçak and Phanish Puranam

ABSTRACT

Organizational cultures that facilitate collaboration are valuable, but little is known about how to create them. The authors investigate the microfoundations of this problem using computational models of dyadic coupled learning. The authors find that merely altering initial beliefs about the consequence of actions (without altering the consequences themselves) can under some conditions create cultures that promote collaboration. The results of this study show why the right initial “framing” of a situation – established for instance through persuasive rhetoric, an inspiring vision, or careful recruitment choices – may under the right conditions be self-reinforcing, instead of becoming empty symbolism.

Keywords: Culture; collaboration; coupled learning; computational models; organization design; socialization; sense-making; sorting

INTRODUCTION

Organizational culture, defined as a system of shared assumptions, values, and norms, can facilitate collaboration and thus improve performance (Chatman & O’Reilly, 2016; O’Reilly & Chatman, 1996; Schein, 1983, 2004). Failures of collaboration such as those observed after mergers, across internal organizational units, or in the aftermath of too rapid expansion through hiring, are often attributed to cultural divergence (e.g., Van den Steen, 2010a). Conversely, cultures are often credited with the success of organizations (e.g., Pfeffer, 1995). Since collaborative actions are often difficult to measure and verify, there are inevitable limits to the use of contracts and authority (Gibbons & Henderson, 2012). Culture can fill in the gaps in formal administrative systems to promote collaboration (O’Reilly & Chatman, 1996).

This leads naturally to the question of whether and how a culture that supports collaborative efforts can be designed. One cannot simply mandate a particular set of assumptions, values, or norms on the basis of authority, and expect them to persist. In this chapter, we focus on a particular, precisely demarcated “object of intervention” in the process of setting up collaborative cultures, and then offer some initial results on how the effectiveness of interventions can be traced back to the properties of the “context of intervention.” With respect to the object of intervention, we build on a conceptual framework that distinguishes between design interventions that alter individuals’ beliefs about the consequences of their actions and interventions that alter the actual consequences of those actions.

While recognizing that designers can and do intervene in shaping outcomes of behavior, in this chapter we take the structure of interdependence between actors (i.e., how their payoffs depend on each other’s actions) as given. We then examine the conditions under which organizational designers can elicit collaborative behavior by merely shaping beliefs about payoffs – which we conceptualize to stem from the assumptions, mental models, and representations that actors hold (Cannon-Bowers & Salas, 2001; Rouse & Morris, 1986; Thagard, 2005; Van den Steen, 2010b). These beliefs are the “object of intervention” in our analysis.

It is a well-established view in the literature that organizations constitute the psychological environment in which choice occurs (Simon, 1945) and exert a powerful influence on the beliefs of their constituent individuals (March & Simon, 1958, chapter 3). Our analysis helps to see how far one can go toward building a culture of collaboration merely by shaping beliefs; it responds to the recent rethinking of organizational culture presented by Chatman and O’Reilly (2016) and in particular to their call for renewed attention to the underlying psychological mechanisms that shape cultures (p. 2). As they note, beliefs and assumptions form a bedrock layer of culture, harder to observe than norms or artifacts, but nonetheless foundational (p. 16). The analysis in this chapter provides a rigorous framework to describe the conditions under which individual beliefs aggregate into stable patterns of collaborative actions.

We build on prior influential work to identify the kinds of cultural interventions that designers can use to influence beliefs. Schein summarizes decades of his own and others’ work when he writes:

Cultures basically spring from three sources: (1) the beliefs, values, and assumptions of founders of organizations; (2) the learning experiences of group members as their organization evolves; and (3) new beliefs, values, and assumptions brought in by new members and leaders. (Schein, 2004, p. 255)

Thus, organizational designers can try to shape beliefs in an organization: (1) by training new entrants in a top-down fashion (socialization), (2) by influencing the process of interaction and learning among the organization’s members (sense-making), and (3) by selecting organizational members on the basis of cultural fit (sorting).¹

Clearly, these mechanisms do not operate in isolation from each other. In fact, they must be mutually reinforcing – if the beliefs that individuals bring to an organization or imbibe from its current members at the time of entry do not align with what they experience as they live and interact in the organization, the beliefs are unlikely to remain either stable or shared (Schein, 1983, p. 8). Even brainwashing has its limits, when confronted by the constant intrusion of reality (Schein, Schneier, & Barker, 1961). The explicit focus on the processes that reinforce or weaken initial beliefs is the first distinctive feature of our analysis. We assume that the initial beliefs held by individuals are modified through feedback to their actions. To formalize this, we build on models of coupled learning that assume bounded but adaptive rationality (e.g., Knudsen & Srikanth, 2014; Lave & March, 1975).

The second distinctive feature of our analysis concerns the “context of intervention.” We assume that the key element of context is the feedback that individuals get on their actions. Specifically, we assume that the structure of interdependence between individuals in organizations, which determines the feedback to individuals’ joint actions, can exhibit a lot of diversity. We draw on game theoretic representations of common social dilemmas to formalize this variation (e.g., Camerer, 2003; Camerer & Knez, 1996).

Third, while our approach builds on a tradition of using computational models to understand culture (e.g., Harrison & Carroll, 2006), it uniquely takes a micro-foundational approach to understand the links between individual beliefs and how they are shaped through interaction in a context. We model processes of interaction rather than incorporating their outcomes, such as influence of managers or peers, as exogenous parameters that affect the degree of enculturation.

Our analysis gives sharp predictions about the conditions under which efforts to shape beliefs through cultural design efforts are likely to be effective in creating collaboration, and why. The results suggest that cultural interventions aimed primarily at endowing actors with an initial set of beliefs (for instance through socialization and sorting), as well as interventions that affect the process through which actors update their beliefs through interaction (i.e., sense-making), when combined, promote the emergence of improved collaboration.

Put differently, cultural interventions that only shape beliefs and how they are revised (without altering the actual consequences of actions) may nonetheless contribute to the creation of a culture of collaboration across a range of contexts. Our results imply that the right initial “framing” of a situation – established through persuasive rhetoric and an inspiring vision – may be self-reinforcing under the right conditions, rather than merely empty symbolism. This has implications for practice in situations involving inter-organizational collaboration (e.g., between departments or in post-merger integration) as well as the process of hiring and enculturation of new employees.

USING GAMES TO REPRESENT INTERDEPENDENCE STRUCTURES

Organizations exist to solve problems of collective action that prevent individuals from reaching socially optimal outcomes (Chatman & O’Reilly, 2016; Miller, 1992). The precise nature of the problem of collective action that individuals face depends on the nature of interdependence between them. Interdependence between agents in an organization implies that the returns to A of A’s actions depend on B’s action (and vice versa) (Puranam, Raveendran, & Knudsen, 2012). It is a consequence of design choices about the division of labor (i.e., how to partition and allocate tasks), as well as the integration of effort (i.e., how to motivate and coordinate them) (Raveendran, Puranam, & Warglien, 2015).

We take it as a premise that a fruitful understanding of how cultural design interventions operate cannot be achieved without reference to the underlying interdependence structure between agents. Game theory provides a useful and intuitive set of representations of interdependence structures that highlight unique challenges to collaboration (Camerer & Knez, 1997; Knez & Camerer, 1994). The utility of representations of interdependence in the form of games has been recognized by scholars across a variety of fields broadly related to organizations (e.g., Dawes, 1980; Kelley & Thibaut, 1978; Kollock, 1998; Ostrom, 1990).²

A general partitioning of game structures is into those that feature challenges to cooperation, challenges to coordination, and both (Camerer & Knez, 1997). Note that it makes little sense to talk of pure cooperation or pure coordination success – the success of each presumes the other. Success at both cooperation and coordination is necessary for successful collaboration. However, failure in either cooperation or coordination is individually sufficient to generate collaboration failures.

The defining attribute of a game that features a cooperation challenge is that (at least some) players have to choose an action that produces a sub-optimal outcome for themselves, for the group to attain the best outcome (Camerer & Knez, 1997). The most famous of these games is the Prisoner’s Dilemma, or the Public Goods game, which serves as the model for why groups may fail to build a public good or maintain a common pool resource, and why free-riding may occur in team tasks.

In Table 1, the collaborative outcome is achieved when both A and B cooperate and yields an individual payoff of δ (where 0 < δ < 1), but the Nash equilibrium involves both defecting (and yields an individual payoff of 0 < ε < δ). The challenge is to get agents to converge to the collaborative outcome, avoiding the temptations to free ride (1 − δ) and the opportunity cost of being exploited (0 − ε) that drag them to the Nash equilibrium that yields payoffs of ε to both.

Table 1. Public Goods (Prisoner’s Dilemma) Payoff Matrix.

	B Chooses Action 1	B Chooses Action 2
A Chooses action 1	δ, δ	0, 1
A Chooses action 2	1, 0	ε, ε

Notes: In all cases 1 > δ>ε>0, where δ lies in (0,1) and ε in (0, δ).

Games that feature a pure coordination challenge lack barriers to cooperation because each player’s payoff is maximized when the group’s payoff is maximized. Matching (Schelling, 1960) and Stag-Hunt (Skyrms, 2004) are two important instances of pure coordination games. The Matching game popularized by Schelling (1960) is one where both players are better off as long as their actions match (both pick action 1 or action 2), but it does not matter which particular actions they choose (Table 2). Standardizing on one among multiple equivalent choices is an instance – what matters are that we reach agreement, but what we agree on is not important. There are thus two collaborative outcomes possible in this case and yield the same average payoff (1.0). There is no incentive to deviate if each player knows for sure what action the other player will take (i.e., each collaborative outcome is a Nash equilibrium). The difficulty lies in finding and converging to one, to avoid the mis-coordinated choices that lead to payoffs of 0 < δ < 1.

Table 2. Matching Payoff Matrix.

	B Chooses Action 1	B Chooses Action 2
A Chooses action 1	1, 1	δ, δ
A Chooses action 2	δ, δ	1, 1

No...

Table of contents

• Cover
• Title Page
• Reviewing, Revisiting, and Renewing the Foundations of Organization Design
• Part I Fit and Coordination
• Part II Configuration and Control
• Part III DIVISION Of Labor and Organizational Learning
• Part IV Structure and Strategy
• Part V New Organizational Forms and Problem Solving
• Index