Application paper

Organizational Routines as Patterns of Action: Implications for Organizational Behavior Brian T. Pentland1 and Thorvald Hærem2

1Eli Broad College of Business, Michigan State University, East Lansing, Michigan 48824; email: [email protected] 2Department of Leadership and Organizational Behaviour, BI Norwegian Business School, NO-0442 Oslo, Norway; email: [email protected]

Annu. Rev. Organ. Psychol. Organ. Behav. 2015. 2:465–87

The Annual Review of Organizational Psychology and Organizational Behavior is online at orgpsych.annualreviews.org

This article’s doi: 10.1146/annurev-orgpsych-032414-111412

Copyright © 2015 by Annual Reviews. All rights reserved

Keywords

patterns of action, narrative network, practice theory, behavioral decision making

Abstract

In recent years, organizational routines have been studied in a wide variety of settings, including law,medicine, accounting, and engineer- ing. This fieldwork has led to a broader understanding of organiza- tional routines as repetitive, recognizable patterns of interdependent action, carried out by multiple actors. Routines are seen as practices that are situated in a social/material context.Within an organizational routine, individual actions are situated in a broader pattern of actions that can be represented as a network. Recognizing patterns of interde- pendent action as a unit of analysis entails a research paradigm that has implications for a range of topics in organizational behavior.

465

A nn

u. R

ev . O

rg an

. P sy

ch ol

. O rg

an . B

eh av

. 2 01

5. 2:

46 5-

48 7.

D ow

nl oa

de d

fr om

w w

w .a

nn ua

lr ev

ie w

s. or

g A

cc es

s pr

ov id

ed b

y U

ni ve

rs ity

o f

C al

if or

ni a

- R

iv er

si de

o n

08 /1

3/ 22

. F or

p er

so na

l u se

o nl

y.

INTRODUCTION

The term routine is often associated with work or behavior that is automatic, low skill, repetitive, or even habitual (Becker 2004, Reynaud 2005). Ashforth & Fried (1988) used the term mind- lessness to capture the core idea. This view of organizational routine builds on ideas about cognitive efficiency, habit, and standard operating procedures (e.g., March & Simon 1958). Economists who refer to routine work (Autor & Dorn 2013) use the term in this narrow sense as well.

Although this class of phenomenon will always be an important part of the economic and organizational landscape, field research has identified a wide variety of settings in which repetitive patterns of action are prevalent but cannot be described as automatic, low skill, ormindless. These include medical work (Edmondson et al. 2001, Goh et al. 2011, Greenhalgh 2008), legal work (Brown & Lewis 2011), accounting work (Burns & Quinn 2011, Burns & Scapens 2008, Quinn 2014), social work (Foldy & Buckley 2010), engineering and design work (D’Adderio 2008, Gaskin et al. 2010, Leonardi 2011, Rerup & Feldman 2011), technical work (D’Adderio 2003, Pentland&Rueter 1994, Reynaud 2005), research and development (Bresman 2013, Bruns 2009, Howard-Grenville 2005), and many others (Becker 2004, Feldman 2000, Rice & Cooper 2010, Turner & Rindova 2012).

For better or worse, the definition of organizational routine has expanded from the narrow, stereotypical examples of mindless assembly line work to include a much broader range of phe- nomena. Therefore, for the purposes of this article, we adopt the definition from Feldman & Pentland (2003, p. 95): “repetitive, recognizable patterns of interdependent actions, carried out bymultiple actors.” In their editorial introduction to the Journal ofManagement Studies special issue on the microfoundations of routines and capabilities, Felin et al. (2012, p. 1355) note that this definition has become “widely accepted.”Beyond the definition itself, the research literature on organizational routines has begun to focus on patterns of action as the unit of analysis (Pentland & Feldman 2005). We discuss the implications of these developments throughout this article.

We begin this article by clarifying the definition of routines as pattern of actions and how this emphasis relates to traditional research paradigms in organizational behavior. Our review of the literature confirms the growing consensus that patterns of interdependent action can be a fruitful unit of analysis and that such patterns can be understood as a ubiquitous aspect of organizing in general (Weick 1969). However, we also find that explanations for how such patterns form, persist, and change could be improved. To help fill this gap, we argue that concepts from heuristic information processing (Evans 2008, Gigerenzer 2000, Gigerenzer& Todd 1999, Hogarth 2001, Kahneman & Klein 2009, Simon 1987) may provide more precise microfoundations for orga- nizational routines. At the same time, the focus onpatterns of actionmay lead to novel perspectives on topics such as decision making, transactive memory systems, and organizational change.

BEYOND MINDLESSNESS: A BROADER VIEW OF ORGANIZATIONAL ROUTINES

Repetitive, recognizable patterns of action are not necessarily fixed, automatic, or mindless. The confusion over the labeling and definition of this phenomenon stems, in part, from the fact that the word routine can be used as both an adjective and a noun.As an adjective, it is amatter of degree— a process or pattern of activity can be more or less routine (Reynaud 2005). There are objective measures for capturing the degree of routinization or variety in a pattern of action (Cohen & Bacdayan 1994, Pentland 2003, Su et al. 2013).

466 Pentland � Hærem

A nn

u. R

ev . O

rg an

. P sy

ch ol

. O rg

an . B

eh av

. 2 01

5. 2:

46 5-

48 7.

D ow

nl oa

de d

fr om

w w

w .a

nn ua

lr ev

ie w

s. or

g A

cc es

s pr

ov id

ed b

y U

ni ve

rs ity

o f

C al

if or

ni a

- R

iv er

si de

o n

08 /1

3/ 22

. F or

p er

so na

l u se

o nl

y.

Definitional Components

As a noun, the definition of routine offered by Feldman & Pentland (2003) refers to a category of phenomenon that is very broad but has definite limits. For a phenomenon to qualify as an or- ganizational routine, all four parts of the definition need to be satisfied; they are individually necessary and jointly sufficient conditions. It is worth considering each of them in turn.

Repetitive. The role of repetition can be thought of as a continuum. At one extreme, we have assembly lines, call centers, and other operations in which patterns of actions are repeated fre- quently (Nyberg 2009). At the other extreme, however, one can imagine patterns that do not occur (e.g., disaster recovery plans). The existence of a plan, a guideline, awritten procedure, or any such artifact is neither necessary nor sufficient to define an organizational routine. Even if repetition is rare, it is a necessary part of the definition.

Recognizable pattern. The concept of repetition is closely related to the more subtle question of whether the pattern of action is recognizable. If a pattern cannot be recognized as an instance of a prior pattern, it is impossible to say whether it has repeated or not. Actions in the pattern may be widely distributed in time and space. The pattern may contain interruptions, exceptions, errors, and improvisation, and it can change over time (Feldman & Pentland 2003). As a result, it is not straightforward to recognize and count instances of a routine (Pentland&Rueter 1994). Given an appropriate representation of the routine, there are a variety of mathematical techniques that can be used to recognize, categorize, and cluster patterns of action according to similarity. As a matter of definition, however, if a pattern of action cannot be recognized, there is no organizational routine.

Interdependent actions. The actions within an organizational routine must be interdependent. Interdependence can take many forms, such as sequential flow of materials or information from one action to the next (Puranam et al. 2012, Thompson 1967). If a set of actions is not in- terdependent, it cannot be anorganizational routine, nomatter howoften it repeats or howeasily it can be recognized.

Multiple actors. Finally, we distinguish between individual routines and organizational routines. Individual-level tasks and habits will always be important, but organizational routines require the involvement ofmultiple actors (Becker 2005, Reynaud 2005). The actors need not be copresent or even aware of each other’s role in carrying out the overall pattern (Rice & Cooper 2010). A consistent or stable set of actors is not necessary, and recent research on routines has incorporated the possibility of nonhuman actors, such as computerized workflow systems (e.g., Pentland et al. 2010).

Defining organizational routines as patterns of action expands the potential applicability of the concept tremendously because it is no longer restricted to the narrow, stereotypical example of mindless work.

Patterns of Action

The focus on patterns of action is one of the distinctive features of this emerging body of research. It builds on Weick’s (1969) assertion that it is actions that are organized, not people, as he ad- vocated a shift from organization to organizing. Rousseau (1997, p. 536) argued that the “evolution from organization to organizing changes both the phenomena traditionally studied by

467www.annualreviews.org � Organizational Routines

A nn

u. R

ev . O

rg an

. P sy

ch ol

. O rg

an . B

eh av

. 2 01

5. 2:

46 5-

48 7.

D ow

nl oa

de d

fr om

w w

w .a

nn ua

lr ev

ie w

s. or

g A

cc es

s pr

ov id

ed b

y U

ni ve

rs ity

o f

C al

if or

ni a

- R

iv er

si de

o n

08 /1

3/ 22

. F or

p er

so na

l u se

o nl

y.

organizational research and meaning of some traditional concepts.” Rousseau’s comments are exemplified by the organizational routines research paradigm.

The essential move is to treat the pattern of action as the unit of analysis (Pentland & Feldman 2005). This breaks from research paradigms in which the focus is on specific kinds of decisions or on the actors making the decisions. For example, in traditional decision-making research (e.g., Gigerenzer 2000, Kahneman&Klein 2009, Tversky&Kahneman 1974,Whyte 1989), decisions or actions are predicted based on the properties of individuals (attitudes, expertise, biases, etc.) or groups (cohesion, conflict, diversity, etc.).With some exceptions, the paradigmatic research design considers a single type of action or decision (making a purchase, awarding a raise, etc.), as shown on the left side of Figure 1. Even in research on escalating commitment (e.g., Sleesman et al. 2012), the same actor is prompted to make a similar decision.

By contrast, research on organizational routines examines sequences of different actions, in which actions at one point in time influence or trigger subsequent actions, as shown on the right side ofFigure 1. This paradigm has been applied to such diverse activities as performing a surgical procedure (Edmondson et al. 2001) or designing a building (Gaskin et al. 2010). Each action involves an actor, but not necessarily the same actor. The set of actors involved in carrying out a routine may be widely distributed in time and space, they may not be aware of each other, and they are not necessarily human (Pentland & Feldman 2005). This stands in sharp contrast to traditional group research, in which the paradigmatic research design includes copresent human individuals engaged in a single decision (e.g., Kahneman 2011, Plous 1993).

The explicit focus on patterns of action distinguishes organizational routines from related concepts, such as values, norms, and culture. Whereas culture involves webs of meaning (Geertz 1973), routines involve webs of action. Concepts like values, norms, and culture are distinctive aspects of organizations, and they do not necessarily entail repetitive, recognizable patterns of interdependent action. Furthermore, action patterns are often inconsistent with organizational values and norms (Rice&Cooper 2010).One contribution of this research stream is the recognition that organizational routines are conceptually and empirically distinct from these other phenomena.

Routines as Practice

The definition of routines offered by Feldman & Pentland (2003) has been described as the practice-based view (Parmigiani & Howard-Grenville 2011, Radwan & Kinder 2013) because it conforms to ontological assumptions of theories of practice (Bourdieu 1990, Feldman &

Single-action paradigm Action-pattern paradigm

Action

Action/ actor

Action/ actor

Action/ actor

Action/ actor

Action/ actor

Action/ actor

Action/ actor

Action/ actor

Actor • Demographics

• Personality

• Motivation

• Attitudes

• Training

Figure 1

Single actions versus patterns of action.

468 Pentland � Hærem

A nn

u. R

ev . O

rg an

. P sy

ch ol

. O rg

an . B

eh av

. 2 01

5. 2:

46 5-

48 7.

D ow

nl oa

de d

fr om

w w

w .a

nn ua

lr ev

ie w

s. or

g A

cc es

s pr

ov id

ed b

y U

ni ve

rs ity

o f

C al

if or

ni a

- R

iv er

si de

o n

08 /1

3/ 22

. F or

p er

so na

l u se

o nl

y.

Orlikowski 2011, Gherardi 2000, Nicolini 2011) and structuration (Feldman 2003, Giddens 1984). Feldman & Pentland (2003) argue that the properties of routines are consistent with an underlying duality of structure and agency that has become central to contemporary theories of organizational stability and change (Farjoun 2010, Tsoukas & Chia 2002).

This ontological perspective is important because it recognizes the essential duality of this phenomenon. This duality is sometimes discussed in terms of levels (e.g., Becker 2005, Vromen 2011) or aspects (Feldman & Pentland 2003). The sidebar Dual Aspects of Organizational Routines provides some typical terminology that has been used to distinguish different aspects of routines.

The tendency to focus on one side of the duality or the other has led to some debate in the theoretical literature on routines. On one hand, some have argued that routines should be un- derstood as latent phenomena (Hodgson 2008), because routines may be dormant for extended periods of time (Birnholtz et al. 2007) and observable performances contain variations thatmay be due to extraneous factors, such as interruptions, that are not part of the routine. This perspective aligns with the metaphor of routines as genes (Nelson &Winter 1982). On the other hand, some have argued that the performative aspect is key, because empirical research on routines is best served by analysis of the observable performances (Pentland et al. 2010). From the practice perspective, these two aspects form a duality (Feldman& Pentland 2003, Feldman&Orlikowski 2011): They are not just mutually causal; they are mutually constitutive.

The practice perspective also helps to clarify the distinction between rules and routines, a topic over which there is frequent confusion. Organizations have rules (Blau 1955), but rules are not routines (Grote 2012, Reynaud 2005, Grote et al. 2009). Reynaud (2005) offers a thorough discussion of the difference between rules and routines, noting that rules must be interpreted and translated into practice (routine) before they have any practical effect. Even when organizations encode rules into computerized artifacts, the results are not always as intended, because the encoded rules must be appropriated into practice (DeSanctis & Poole 1994, Pentland & Feldman 2008). Like any practice, routines are generative in the sense that they continue to produce new patterns of action over time (Bourdieu 1990, Feldman 2000, Feldman & Pentland 2003).

Actors, Artifacts, and the Material Context

The practice-based view of routines also emphasizes that action is situated in some material context. In this respect, it builds on work by Suchman (1987), Lave (1988), Hutchins (1995), Wenger (1998),Orlikowski (2007) andotherswhohave shownhowmaterial circumstances shape the conduct of work, action, and interaction in fundamental ways. In the literature on routines, many scholars have commented that routines are adapted and connected to the particular or- ganizational context in which they are performed (Cohen et al. 1996, Cohendet & Llerena 2003,

DUAL ASPECTS OF ORGANIZATIONAL ROUTINES

Latent Expressed Potential Actual Ostensive Performative Deep structure Surface structure

469www.annualreviews.org � Organizational Routines

A nn

u. R

ev . O

rg an

. P sy

ch ol

. O rg

an . B

eh av

. 2 01

5. 2:

46 5-

48 7.

D ow

nl oa

de d

fr om

w w

w .a

nn ua

lr ev

ie w

s. or

g A

cc es

s pr

ov id

ed b

y U

ni ve

rs ity

o f

C al

if or

ni a

- R

iv er

si de

o n

08 /1

3/ 22

. F or

p er

so na

l u se

o nl

y.

Turner&Fern 2012). In field research on organizational routines,material artifacts are verymuch present. D’Adderio (2011, p. 197) argues that artifacts and materiality should be “at the very center” of routines theory. Nyberg’s (2009) fieldwork vividly illustrates the close interaction of people andmachines in call centers. Pentland et al. (2011, p. 1369) describe the collection of actors carrying out an invoice processing routine as a “sociomaterial ensemble” because it includes both humans and a computerized workflow system. Leonardi (2011) argues that when we observe a routine in practice, we are inevitably observing some form of technological artifact as well.

Artifacts appear in at least three kinds of roles in the research literature on organizational routines. First, material artifacts serve as affordances and constraints (Norman 2002, Robey et al. 2012). In this role, technical artifacts are tools; they operate at the discretion of their current user.

Second, artifacts can encode the intentions of managers or designers (D’Adderio 2008, Pentland & Feldman 2008). In many cases, this can include formal controls (Burns & Quinn 2011). In this role, they are more than just tools, because they shape the course of action. Of course, people may engage in workarounds or creative reinvention (Boudreau & Robey 2005, DeSanctis & Poole 1994).

Finally, artifacts participate in routines as actors that take actions.More than just shaping action, theyactivelyparticipate.For example, ina studyof invoiceprocessing routines,Pentland et al. (2011) observed that 65%of the actions were carried out by people, but the remainder were carried out by the computerized workflow system. This included routing information to the human decision makers, but also making approval decisions according to predefined rules. The role of nonhuman actors has become an explicit topic in recent research on routines (D’Adderio 2011, Leonardi 2011, Nyberg 2009, Pentland& Feldman 2008, Pentland et al. 2011). Therefore, throughout this article, the term actor is meant to refer to social situation participants that are not necessarily humans.

The emphasis on material context and nonhuman actors is consistent with actor-network theory (Latour 2005, Law 1992) but stands in contrast to the mainstream organizational liter- ature. Zammuto and colleagues (2007, p. 750) noted that less than 3% of the articles published in last 10 years in the top four organizational journals feature technology and organization. Technology is a ubiquitous and transformational presence in contemporary organizations (Brynjolfsson & McAfee 2011), but it is absent from most research on organizational behavior. Thus, research on organizational routinesmay provide an opportunity to incorporate thematerial side of organizing into our research.

Levels of Analysis

Focusing on actions rather than actors has significant implications for conventional concepts of levels of analysis (individual, group, organization) as well. Levels of analysis are usually nested, in the sense that an individual belongs to a group or subunit, and that subunit belongs to an overall organization (Chan 1998). This kind of hierarchical structure conforms to managerial norms and facilitates the use of familiar statistical techniques concerning the analysis of variance across levels (Kozlowski et al. 2013). A similar approach can be applied to organizational routines, except that the logic applies to actions rather than actors.

When we switch from actors to actions, what we traditionally call level of analysis becomes granularity (Pentland et al. 2010). Actions can be nested within other actions (e.g., hiring includes interviewing), but unlike traditional levels of analysis, the decomposition can be carried out infinitely far (Abell 1987). This is because we are nesting actions instead of objects. Actions occur at a moment in time, and those moments can be sliced into very small increments (Su et al. 2013), whereas objects such as individual humans cannot. The broadened concept of routine has been applied to a wide range of activities, ranging from long-term, distributed processes such as

470 Pentland � Hærem

A nn

u. R

ev . O

rg an

. P sy

ch ol

. O rg

an . B

eh av

. 2 01

5. 2:

46 5-

48 7.

D ow

nl oa

de d

fr om

w w

w .a

nn ua

lr ev

ie w

s. or

g A

cc es

s pr

ov id

ed b

y U

ni ve

rs ity

o f

C al

if or

ni a

- R

iv er

si de

o n

08 /1

3/ 22

. F or

p er

so na

l u se

o nl

y.

technology roadmapping for the semiconductor industry (Howard-Grenville 2005) to very short- term, localized processes such as office work (Su et al. 2013).

Theconceptofmeta-routines (Lewin et al. 2011,Winter 2003) provides a superficial analogy to traditional levels of analysis, but it does not necessarily conform to the concept of nesting. This is because a meta-routine is a routine for changing other routines (Winter 2003). Total quality management (Hackman&Wageman 1995) is a classic example of ameta-routine; it can be applied to the refinement of nearly any kind of routine that has measurable results. However, it does not makes sense to include eachof those other routines as a part of a total qualitymanagement routine in the sense that a particular individual is a member or employee of a particular subunit. Likewise, the levels referred to by Becker (2005) and Vromen (2011) are not levels of nesting or aggregation; they are levels of structure, as in the linguistic metaphor of surface versus deep structure.

REPRESENTING ROUTINES AS NETWORKS

The research literature has often relied on qualitative, textual descriptions of routines. Although qualitative description may always be the best way to express context and meaning, we also need a way to represent what Cohen (2007, p. 781) referred to as “pattern in variety.” The patterns of action in a routine generally have a strong central tendency; this is what makes them recognizable. However, they may also have many variations. For example, using archival records from a computerized workflow processing system, Pentland et al. (2011) observed more than 1,000 distinct patterns of action in the performance of an invoice processing routine. This variability demonstrates the deficiency of the traditional concept of routines as a fixed sequence: Even within a basic accounting procedure, variations are common.

To address this fundamental aspect of the phenomenon, Pentland& Feldman (2007) suggested representing routines as networks of functional events. The concept of a functional event was adapted from the structural analysis of narrative (Bal 1985; Hendricks 1973); it refers to a distinct event in a narrative that advances the plot. It is a piece of a story, such as “the princess kisses the frog.” Thus, in a narrative network, the nodes include both the action and the actor(s) involved in carrying out the action. Pentland & Feldman (2007) refer to these nodes as narrative fragments and the overall structure as a narrative network. It is also possible to construct a simplified version in which the network nodes are simply actions, rather than functional events (e.g., Pentland et al. 2010). In either case, this stands in contrast to a traditional social network, in which the nodes are usually individual actors.

The nodes in a narrative network are related by sequence, so the ties are directional. At each node, the network shows what just happened (prior events) and what could happen next (sub- sequent events). In this way, a narrative network maps out the possible paths (or story lines) that are a natural part of every organizational routine.

A narrative network can be represented as a directed graph or a matrix, as shown in Figure 2. Matrix elements, aij, express the probability of a tie from event ai to aj. Unlike those in a social network, the nodes in a narrative network can be self-connected because events can repeat. Figure 2 shows two routines, each of which consists of five actions. The top row depicts each routine as a directed graph; the bottom row depicts each routine as a network. The diagrams (and the corresponding networks) could be expanded to include a “start” and “end,” if so desired. On the left is a very simple, straight-line routine, where each action triggers the next, without exception. This is the stereotypical fixed sequence. In the corresponding matrix, the transition probability from one event to the next is 1.0 for each path and zero elsewhere. On the right is a routine that includes some possibilities for variation. The probabilities in the matrix are for purposes of illustration.

471www.annualreviews.org � Organizational Routines

A nn

u. R

ev . O

rg an

. P sy

ch ol

. O rg

an . B

eh av

. 2 01

5. 2:

46 5-

48 7.

D ow

nl oa

de d

fr om

w w

w .a

nn ua

lr ev

ie w

s. or

g A

cc es

s pr

ov id

ed b

y U

ni ve

rs ity

o f

C al

if or

ni a

- R

iv er

si de

o n

08 /1

3/ 22

. F or

p er

so na

l u se

o nl

y.

The narrative network provides a way to characterize the central tendency of a pattern, as well as the variability. It can also be used to characterize task complexity by counting the pathways in the network (Hærem et al. 2015). Narrative networks have been applied in field research in medical settings (Goh et al. 2011, Hayes et al. 2011), accounting information systems (Yeow & Faraj 2011), and coordination in emergency response (Constantinides & Barrett 2012). The network representation solves the central problem of conceptualizing organizational routines as patterns with variations, as suggested byCohen (2007). Further, it can be used in empirical research to objectively identify and compare routines and detect change over time (Pentland et al. 2010). At the same time, it highlights a central question about themicrofoundationsof organizational routines: How are the actions within a routine connected? To address this question, we turn to concepts from behavioral decision making (Kahneman 2011, Plous 1993, Simon 1959, Winter 2013).

ORGANIZATIONAL ROUTINES AND BEHAVIORAL DECISION MAKING

Within an organizational routine, one action cues subsequent actions (Schulz 2008,Winter 2013). Sequential cueing of actions formed the basis for the influential article by Cohen & Bacdayan (1994), who studied the formation of routines within a two-player card game called Target the Two. They argued that “the routines of a group can be viewed as the concatenation of such procedurally stored actions, each primed by and priming the actions of others” (Cohen & Bacdayan 1994, p. 557). However, that study focused on how knowledge was stored (e.g., procedural versus declarativememory), rather than thenature of the cueingmechanism. Following Cohen&Bacdayan’s landmark study, there has been ongoing interest in the cognitivemechanisms

Very simple routine More complex routine

1 2 3 4 5

0 1.0 0 0 0

1 2 3 4 5

0 0 1.0 0 0

0 0 0 1.0 0

0 0 0 0 1.0

0

1

2

3

4

5 0 0 0 0

0 0.4 0.6 0 0

1 2 3 4 5

0 0.1 0 0 0.9

0 0 0 1.0 0

0 0 0 0.2 0

0

1

2

3

4

5 0 0 0 0

1

5

3 4

2

Figure 2

Routines as networks of action.

472 Pentland � Hærem

A nn

u. R

ev . O

rg an

. P sy

ch ol

. O rg

an . B

eh av

. 2 01

5. 2:

46 5-

48 7.

D ow

nl oa

de d

fr om

w w

w .a

nn ua

lr ev

ie w

s. or

g A

cc es

s pr

ov id

ed b

y U

ni ve

rs ity

o f

C al

if or

ni a

- R

iv er

si de

o n

08 /1

3/ 22

. F or

p er

so na

l u se

o nl

y.

for storing routines (Lazaric 2011, Lazaric&Denis 2005,Winter 2013). However, the processing of information cues per se has not been a substantial research focuswithin studies of organizational routines. Recent agent-based simulation models assume that one action triggers the next, but the triggering mechanism is not explicitly modeled (Cohen et al. 2014, Miller et al. 2012). Likewise, empirical studies have not focused on information cues within routines.

This lack of research attention is unfortunate because the connection between actions is one of the defining features of the phenomenon. Organized patterns of action can be very complex. Within a narrative network, additional nodes and ties can have an exponential effect on the number of pathways (Hærem et al. 2015). Each performance of a routine involves a single pathway, but there may be thousands of possible pathways. The question is, how is a particular pathway selected and enacted by the participants?

Situational strength provides a starting point for examining the role of information cues in shaping a course of action. In a recent review of the concept, Meyer et al. (2010, p. 122) define situational strength as

. . .implicit or explicit cues provided by external entities regarding the desirability of potential

behaviors. Situational strength is posited to result in psychological pressure on the individual to

engage in and/or refrain from particular courses of action. . . .

Organizational routines embody courses of action, so the concept of situational strength is clearly relevant. Meyer et al. (2010, pp. 126–27) review the literature to derive a model of situational strength with four dimensions: clarity, consistency, constraints, and consequences. Because they are recognizable and repetitive, organizational routines should provide participants with cues that are relatively clear and consistent. Interdependence of actions within a routine introduces im- portant constraints. Finally, there are often significant, clearly defined rewards and penalties for compliance (or lack of compliance) with the pattern.

In a naturalistic setting, however, the set of cues can quickly become very complex (Hærem et al. 2015). Real organizational routines are not confined to pairs of participants, as in Cohen& Bacdayan’s (1994) experimental turn-taking card game.Rather, the routinesmaybe carried out by apotentially open-ended set of actors, distributedover several locations. Inmany settings, it is common to have multiple performances of the same routine at the same time (e.g., cooking multiple meals in a kitchen), and it is common to have interruptions (Su et al. 2013). As action patterns often contain variations, itmaybedifficult to recognizewhatprioractionsare relevant, and todecidewhat todonext. Given this level of complexity, it would not be surprising if simplifying heuristics (Kahneman&Klein 2009) were involved in processing the information cues. However, our review of the literature shows that the processing of information cues within organizational routines has not been examined.

Intuitive and Analytical Processing of Information Cues

To help address this gap, we draw on the literature on intuitive and analytical decision making, which builds upon dual-processing theories (Evans 2008). Although there are skeptics (e.g., Allison & Hayes 2012, Kruglanski & Gigerenzer 2011, Keren & Schul 2009), dual-process theories have become dominant. Dual-process models hold that intuition and analysis are parallel and interactive modes of information processing that are served by separate cognitive systems, System 1 and System 2 (Kahneman 2011, Stanovich & West 2000).

System 1, which handles intuitive processing, is context dependent, associative, heuristic, tacit, intuitive, implicit, automatic, and parallel. Furthermore, it is often accompanied by affect (Dane& Pratt 2007, Hodgkinson et al. 2009). Although this kind of processing is rapid, requires very few

473www.annualreviews.org � Organizational Routines

A nn

u. R

ev . O

rg an

. P sy

ch ol

. O rg

an . B

eh av

. 2 01

5. 2:

46 5-

48 7.

D ow

nl oa

de d

fr om

w w

w .a

nn ua

lr ev

ie w

s. or

g A

cc es

s pr

ov id

ed b

y U

ni ve

rs ity

o f

C al

if or

ni a

- R

iv er

si de

o n

08 /1

3/ 22

. F or

p er

so na

l u se

o nl

y.

cognitive resources, and is robust to stressors, it is slow to learn and unlearn (Schneider & Chein 2003). This is so-called mindless processing. Mindless in this sense does not refer to lack of in- telligence or smartness; it refers to efficient andautomatic utilization of past experience byheuristic processing. Often it refers to the absence of careful and elaborate search and analyses, although a parallel system takes care of this.

System 2 handles analytical processing, which is context independent, rule based, analytic, explicit, and serial. System2 is relatively affect-free (Dane& Pratt 2007, Hodgkinson et al. 2009). It is slow, cognitively demanding, and sensitive to stress. However, it is associatedwithmore rapid learning and is therefore of particular value in novel or exceptional situations (Schneider &Chein 2003). This is so-called mindful processing.

Figure 3 shows one possible way to theorize the role of heuristics in shaping the course of action within an organizational routine. Each action may generate information cues that may (or may not) trigger subsequent actions. The key is to recognize that the trigger is not automatic; it is a phenomenon that can be investigated. Winter (2013) speculates about the role of System 1 and System 2 processing in his discussion of routine microfoundations but conflates the processing of the cues with subsequent processing required to carry out the action.

Heuristics in Patterns of Action

Behavioral decision making treats human decision makers as heuristic information pro- cessors. Based on task-related cues, a heuristic is selected and action follows from the heuristic (Kahneman & Tversky 1974, Kruglanski & Gigerenzer 2011). It is important to note that in this article, we are using the term heuristic to refer to a basic class of cognitive process that operates at a neurological level (Evans 2008). This is distinct from the colloquial usage that lumps heuristics together with routines, recipes, and rules of thumb.

Although heuristics are at the mindless end of the spectrum, Gigerenzer (2000) still sees heuristics as what makes humans smart. Heuristics are fast and frugal and usually lead to very good decisions. By contrast, Kahneman and colleagues see heuristics as (mindless) triggers of decisions that often lead decisionmakers to biased decisionmaking. Kahneman&Tversky (1974) show that heuristics can lead to predictable traps,whereasGigerenzer (2000) shows that heuristics can lead to good decisions, as long as the heuristics are ecologically valid. Kahneman & Klein (2009) try to bridge the two views by discussing the concept of ecological validity of heuristics, arguing that sometimes the environment shifts and renders otherwise efficient heuristics invalid. The quality of the outcomes hinges on the ecological validity of the heuristic.

Action

Functional eventAction1

Action2

Action3

Actor Heuristics

• Recognition

• Simplification

• • •

Information cue

Information cue

Information cue

Figure 3

Heuristics mediate the processing of information cues.

474 Pentland � Hærem

A nn

u. R

ev . O

rg an

. P sy

ch ol

. O rg

an . B

eh av

. 2 01

5. 2:

46 5-

48 7.

D ow

nl oa

de d

fr om

w w

w .a

nn ua

lr ev

ie w

s. or

g A

cc es

s pr

ov id

ed b

y U

ni ve

rs ity

o f

C al

if or

ni a

- R

iv er

si de

o n

08 /1

3/ 22

. F or

p er

so na

l u se

o nl

y.

Kahneman & Klein (2009) discuss the value of ecologically valid cues and skilled intuition. They distinguish between twomain categories of heuristics: simplifying heuristics and recognition heuristics. Simplifying heuristics are of main interest for the heuristics and bias research paradigm (Kahneman & Klein 2009). Recognition heuristics are related to expertise and learning. Simon (1987, 1992)was interested in heuristic decisionmaking as recognition of cues,which“provided the expert access to information stored in memory and the information provides the answer” (Simon 1992, p. 154). Thus, heuristic decision making is a matching process in which cues trigger actions based on the heuristics that are associatedwith that cue. Learning the appropriate cues and heuristics depends on the stability of the environment; wicked or turbulent envi- ronments with unexpected changes may make old learning obsolete and new learning difficult (Hogarth 2001).

Behavioral decision theory suggests that heuristics based on cues generated by prior actions may act as triggers for future actions. In the context of a repetitive, recognizable pattern of actions, however, it seems possible that heuristics that exploit cues about one’s prior action might bias the decision-making process. In particular, one’s own past actions may be both more available and more representative (Tversky & Kahneman 1973, 1974). Escalation of commitment provides a well-known illustration of how past actions may bias future actions (e.g., Sleesman et al. 2012). However, if the heuristics are ecologically valid, this fast, frugal information processing makes organizational routines efficient.

Recent simulation models on the formation and dynamics of organizational routines have begun to incorporate some of these ideas. For example, Cohen et al. (2014) use a multiagent simulation to show how habitual, heuristic decision processes (matching cues and actions) lead to the formation of routines. Miller et al. (2012) also use a multiagent simulation to investigate routine formation. These simulations begin to demonstrate how individual psychological pro- cesses can lead to collective performances, but they have not explicitly modeled simplification and recognition heuristics and have not considered the issue of ecological validity.

In normative decision models, ecological validity has been interpreted in terms of a relatively static background in which there is a clear, normatively optimal outcome. A heuristic is valid if it produces an optimal decision in that context. Within an organizational routine, however, the ecology includes the network of other specific actions that have just occurred or may be expected to occur. Simplifying heuristics like availability and representativeness will tend to favor the immediate context of action. Recognition heuristics may tend to favor recent per- formances of the same routine (taking the same action in the same situation). These heuristics may seem ecologically valid in the context of the particular pattern of action because they enable another efficient performance of the routine. However, the overall pattern of action may still be dysfunctional.

To the extent that human actors are involved, path dependency and lock-ins (e.g., Sydow et al. 2009) may be the product of heuristics that are ecologically valid in the context of the current action pattern. Nonhuman actors may (or may not) increase the information processing capacity but do not necessarily improve the efficiency or effectiveness of the overall action pattern.Wicked environments and changes in technology may make old routines dysfunctional and new ones difficult to acquire.

Can Organizational Routines Be Mindful?

Since Cyert & March (1963) first articulated the behavioral theory of the firm, there has been a canonical tension between routine and rational decisionmaking.March&Simon (1958) argued that routines were needed to deal with the cognitive constraints posed by bounded rationality. The

475www.annualreviews.org � Organizational Routines

A nn

u. R

ev . O

rg an

. P sy

ch ol

. O rg

an . B

eh av

. 2 01

5. 2:

46 5-

48 7.

D ow

nl oa

de d

fr om

w w

w .a

nn ua

lr ev

ie w

s. or

g A

cc es

s pr

ov id

ed b

y U

ni ve

rs ity

o f

C al

if or

ni a

- R

iv er

si de

o n

08 /1

3/ 22

. F or

p er

so na

l u se

o nl

y.

key insight of behavioral theory was that managers do not seem to follow optimal, economic decision models. Instead, they follow routines.

This opposition between routine (mindless) and rational (mindful) persists in both individual- and organization-level theory. The behavioral theory of the firm (Cyert & March 1963) and theories of behavioral decision making (Kahneman 2011) are, in several ways, parallel, although they concern different entities and levels of analysis. Both theories are concerned with how the actor, the firm, or the individual relies on procedures to compensate for bounded rationality. Both theories suggest that decision makers operate across a range of possibilities, from more or less mindless behavior to extensive search and analysis.

Table 1 provides a simple classification of the literature on decision making and routines that builds on Reynaud (2005). As in most such tables, the categories are ideal types. At the individual level, intuitive decision making occupies the “mindless” quadrant, whereas rational, analytical decision making occupies the “mindful” quadrant.

The lower-left quadrant represents mindful individual action (Langer 1989). In normative treatments of decisionmaking, System2 remains the preferablemodeof processing (Milkman et al. 2009). There are many studies of individual decision making that describe problem-solving techniques designed to help individuals from falling into the traps of intuitive processing (e.g., Russo & Schoemaker 2002).

At the level of the organizational routine, there is a parallel set of categories. The “mindless” quadrant (upper right) is occupied by routines that are effectively locked-in; they are relatively fixed and unchanging (Cohen 2007). March& Simon (1958) portrayed routines in this quadrant as efficient tools to process information and perform in well-known environments. This is how Gigerenzer (2000) sees the role of heuristics in individuals’ decisionmaking: as simple, frugal rules that pick up the critical cues in a complex information environment and allow the decision maker to act upon them fast, reliably, and efficiently. At the same time, it is also possible to portray routines as potential sources of error and inertia known as competency traps (March 1991). The argument is that organizations become proficient (competent) at particular patterns of action and become trapped in those patterns even after the organizational environment has changed and rendered them less functional. This argument resonates with Kahneman&Tversky’s (1974) view of heuristics as sources of bias and error. There is considerable support for the idea that orga- nizational routines can be dysfunctional and difficult to change. Rice & Cooper (2010) offer an extensive catalog of apparently mindless patterns of action.

The lower-right quadrant of Table 1, however, suggests the possibility of mindful organiza- tional routines (Levinthal & Rerup 2006). Drawing on ideas of mindful attention (Langer 1989) and heedfulness (Ryle 1949), Weick & Roberts (1993, p. 357) argue that more information processing should improve reliability: “. . .intensified efforts enable people to understand more of the complexity they face, which then enables them to respond with fewer errors.”

Table 1 Decision making and routines

Individual Organizational

Mindless Intuitive System 1

Standard operating procedures Locked-in Path following

Mindful Analytical System 2

Organizational routine Generative Path making

476 Pentland � Hærem

A nn

u. R

ev . O

rg an

. P sy

ch ol

. O rg

an . B

eh av

. 2 01

5. 2:

46 5-

48 7.

D ow

nl oa

de d

fr om

w w

w .a

nn ua

lr ev

ie w

s. or

g A

cc es

s pr

ov id

ed b

y U

ni ve

rs ity

o f

C al

if or

ni a

- R

iv er

si de

o n

08 /1

3/ 22

. F or

p er

so na

l u se

o nl

y.

This argument reinforces the traditional assumption that the difference between mindless and mindful routines concerns the ability to process information. Winter (1985, p. 109) argues that routinized decision processes at the organization level often use a lot of information:

There is little basis in reality for the view that routinized decision processes are necessarily or

typically information poor. The real situation is very nearly the opposite. Because rationality is

bounded, it is not possible for individuals or organizations to improvise the effective use of large

amounts of information. Such use is possible only as a result of prior investment in some sort of

system (or skill, habit, routine, or program) that “mechanistically” picks out, from the enormous

range of possible processing activities, the actual processing that the information is to receive.

It is worth noting that this was written nearly 30 years ago. For decades, decisions have been guided/assisted by computer technology (Silver 1991). Now, computerized decision making is becoming even more sophisticated and pervasive (Brynjolfsson & McAfee 2011). Although routine decision making at the individual level may be characterized by limited information processing and bounded rationality, that is not necessarily true at the organization level, where multiple actors are involved and some of them are able to integrate terabytes of data. With the aid of automated decision making, the traditional opposition of routine versus rational is blurring. Organizational decisions can be routine and rational. The application of heuristics on the individual level may also involve processing of huge amounts of information—much more than typically can be processed within a given time interval in a mindful mode. Rather than information process- ing per se, a defining difference between themodes on both levels of analysismay be the flexibility and openness to alternative interpretations and alternative patterns of actions. Flexibility implies a greater repertoire of actions and a denser, more elaborate set of pathways within the narrative network.

This is not to suggest that every organizational routine is mindful; any such claim would ob- viously be nonsense. Rather, the point is that the new, broader definition of routines as patterns of action breaks down the traditional dichotomy of routine versus rational and has opened up this quadrant to investigation. Furthermore, to view routines as patterns of actions suggests that we can look at decisions as embedded in a networkof other decisions and information cues. This clearly sets some new challenges for how we study routines and decisions about which actions to make.

METHODOLOGICAL ISSUES AND OPPORTUNITIES

When we shift our inquiry to the domain of actions, we confront some unique methodological concerns, and we need a new set of tools. In this section, we discuss some of the methodological issues and opportunities that result from a focus on action patterns in organizational behavior.

Although some studies analyze large numbers of events and apply statistical tests, event data are fundamentally qualitative data (Abbott 1992). There are some established techniques for identifying processes and sequence analysis. For example, event structure analysis (Corsaro & Heise 1990) provides a method for inductively deriving patterns of action from ethnographic observations. Optimal string matching (Abbott 1995) offers a method for comparing sequences. However, all of these methods depend on collecting data about sequences of events or actions that may be distributed in time and space and difficult to observe.

Availability and Quality of Data

Fieldwork on organizational routines can be expensive and time-consuming. Rigorous studies based on videotaping and coding large numbers of performances in field settings (e.g., Su et al.

477www.annualreviews.org � Organizational Routines

A nn

u. R

ev . O

rg an

. P sy

ch ol

. O rg

an . B

eh av

. 2 01

5. 2:

46 5-

48 7.

D ow

nl oa

de d

fr om

w w

w .a

nn ua

lr ev

ie w

s. or

g A

cc es

s pr

ov id

ed b

y U

ni ve

rs ity

o f

C al

if or

ni a

- R

iv er

si de

o n

08 /1

3/ 22

. F or

p er

so na

l u se

o nl

y.

2013) are rare. The increasing availability of digitized data provides an enticing alternative. For example, Pentland et al. (2010) were able to analyze thousands of performances of the same routine in four different organizations by extracting archival data from a workflow system. However, digitized data collected from computer systems are biased toward whatever the designers and managers felt was needed for an audit trail. Archival data accurately record some actions while systematically obscuring others. Although digitized archival data can contain thousands of performances, they generally contain little or no information about the context of the routine; the attitudes, identity, or relationships of the participants; or other factors that may be of interest to researchers.

Field research on organizational routines has tended to emphasize the deeper contextual insights that are possible through observation, interviews, or participant observation. These methods provide researchers with much richer information but naturally tend to limit the number of organizations, routines, and performances that can be studied. Because of the emphasis on traditional fieldwork methods, most empirical studies have been limited to very small numbers of organizations and routines.

Identification and Comparison

To generalize across larger numbers of routines and organizations, we need to identify and compare routines as they occur in different settings (Ragin & Becker 1992). Techniques for comparing patterns of actions are available (e.g., Pentland et al. 2010), but they require a common lexicon to make valid and reliable comparisons. At present, there is no agreed upon lexicon for coding and comparing that can be transferred across situations. This difficulty is exacerbated by the fact that patterns of action can be decomposed into smaller and smaller component actions (Abell 1987). The granularity of the description has a direct effect on the apparent complexity of a routine (Pentland 2003). In current empirical research, decisions about lexicon and granularity are generally a matter of researcher judgment or convenience.

Methodological Innovation

There is a tremendous opportunity to increase the sophistication of the methodology in this do- main. At least three avenues of innovation seem particularly promising by building on well- established techniques from other disciplines, such as computer science. First, analysis of action patterns has been limited to whole sequences or pairs of actions (Pentland et al. 2010). Higher- order patterns (n-grams) have not been used, although it seems likely that higher-order de- pendencies exist in organizational processes. Second, hidden Markov models (HMMs) have rarely been applied to research in organizational behavior. For example, Sánchez and colleagues (2008) demonstrate the use of HMMs in recognizing activity patterns in hospitals. HMM techniques are potentially useful because changes in patterns of action may indicate hidden states or hidden events that would otherwise be difficult to detect. Third, the temporal dimension of routines has received very little attention. Routines operate on very different time scales (seconds, minutes, hours, weeks, months, years). Su and colleagues (2013) offer a very sophisticated analysis of action patterns and the timing of those patterns as a way to measure routineness. More careful attention to the duration and timing of actions within a routine may lead to a different un- derstanding of whether a pattern of action is stable or changing. As these methods depend on having relatively large, valid, and reliable sources of data, it remains to be seen if available archival sources will be adequate.

478 Pentland � Hærem

A nn

u. R

ev . O

rg an

. P sy

ch ol

. O rg

an . B

eh av

. 2 01

5. 2:

46 5-

48 7.

D ow

nl oa

de d

fr om

w w

w .a

nn ua

lr ev

ie w

s. or

g A

cc es

s pr

ov id

ed b

y U

ni ve

rs ity

o f

C al

if or

ni a

- R

iv er

si de

o n

08 /1

3/ 22

. F or

p er

so na

l u se

o nl

y.

NEW RESEARCH DIRECTIONS

Although these methodological issues are challenging, focusing on patterns of action opens up a range of exciting new research opportunities. Here, we briefly mention a few examples.

Antecedents and Consequences of Action Patterns

Although it has become a cliché to “open the black box” in organizational research, the focus on action patterns provides a research strategy for doing exactly that (Pentland & Feldman 2005). The first question must be whether a recognizable pattern exists (Abbott 1992): That is, is there a routine? Given a pattern, we can ask about the antecedents and consequences of the pattern. It may be possible to identify specific variations as well. Dynamics of action patterns are also an emerging area of inquiry (D’Adderio et al. 2012) because they connect to the larger question of organizational change (Farjoun 2010). How do action patterns form (Dionysiou & Tsoukas 2013)? What factors influence their tendency toward stability or change (Schulz 2008)? These questions often require the analysis of sequential data, which requires specific methods, such as optimal string matching (Abbott 1995, Gaskin et al. 2010) or Markov chains (Pentland et al. 2010), that are not widely used in mainstream organizational research.

Microfoundations of Routines

There has been considerable interest in the microfoundations of routines and capabilities (Abell et al. 2008, Cohen 2012, Felin & Foss 2005, Greve 2013, Winter 2013). For example, Abell et al. (2008, p. 496) offer a model of microfoundations in which individuals are modeled by “a simplified Cobb–Douglas function” that includes skill and motivation for each in- dividual. The model assumes that rational individuals will deploy their skills at optimal levels in response to incentives, thereby enacting routines. The Cobb–Douglas model provides a convenient basis for aggregating individual output (Greve 2013), but it does not take into account the defining feature of the phenomenon being modeled: the pattern of interdependent actions.

A more fruitful approach might be to examine the role of information cues within an orga- nizational routine from the perspective of System 1 and System 2 processing, as suggested by Winter (2013).We expect that System 1 processing may be prevalent, but this just begins to frame the potential questions that could be investigated. For example, to what extent do simplifying heuristics matter compared with recognition heuristics? Are there particular kinds of heuristics that are unique to the formation and persistence of organizational routines? Research on this topic would respond to Schulz’s (2008) general question of what keeps routines on track. Individuals differ with respect to preferences for intuitive and analytical thinking (Dane&Pratt 2007, Pacini& Epstein 1999). Investigating how such individual differences predict variation, change, and stability in routines might be a productive contribution to the study of the microfoundations of routines.

One might also investigate the possibility of mismatches (or misfits) at different levels of analysis. For example, can a mechanistic, mindless routine include some amount of individual-level System 2 (analytical) processing? Can a mindful routine include individual- level System 1 (intuitive) processing? What would be the implications of these kind of mis- fits within a routine? This kind of inquiry might investigate the ecological validity of various kinds of heuristics, which is another way to describe the fitness of the heuristic for a given context.

479www.annualreviews.org � Organizational Routines

A nn

u. R

ev . O

rg an

. P sy

ch ol

. O rg

an . B

eh av

. 2 01

5. 2:

46 5-

48 7.

D ow

nl oa

de d

fr om

w w

w .a

nn ua

lr ev

ie w

s. or

g A

cc es

s pr

ov id

ed b

y U

ni ve

rs ity

o f

C al

if or

ni a

- R

iv er

si de

o n

08 /1

3/ 22

. F or

p er

so na

l u se

o nl

y.

Transactive Memory Systems

Research on transactivememory systems (TMS) usually focuses on how individuals become aware of the unique content of knowledge of others. TMS are sometimes described as knowing what others know (Lewis 2003). Much research has been devoted to understanding the structure and content of TMS (e.g., Gupta & Hollingshead 2010, Hollingshead 2001). Agent-based models of routines have demonstrated that TMS formation is a critical aspect of routine formation (Miller et al. 2012). To carry out a routine efficiently, actors need to remember which other actors can perform necessary tasks. Thus, TMS are recognized as an essential aspect of routines.

If we look at TMS in the context of a network of actions, it suggests a novel perspective. Instead of focusing on the knowledge of others, which may be difficult to observe, we might focus on the reactions of others. For example, we may have no idea about the knowledge or training of airport security officers, but we have a very good idea how they would react if we tried to board an air- plane with a fake ID or a concealed weapon. The general question becomes, do we know how otherswill process particular information cues? The information cues and reactions are visible and directly relevant to the course of action within the routine, but cue processing has not been ex- amined as part of the TMS.

Networks of Decisions

Theory and research on behavioral decision making are generally at the individual level, mainly studied in individual settings using laboratory experiments. Other decision-making research (e.g., at the group and organization levels) takes a broader view, but the behavioral decision research paradigm focuses on single decisions taken in isolation from other actions or decisions (Bromiley 2010, Kahneman 2011, Kahneman& Klein 2009, Sleesman et al. 2012). Most studies have been controlled experiments that hold theorganizational context constant, or have considered decisions as independent and isolated events (Bromiley 2010).

Rather than theorizing about particular decisions by particular decision makers, researchers should be able to locate the decision within a network of interdependent decisions. This approach would go beyond traditional concerns about framing, heuristics, and biases because decisions are treated as part of a network of other decisions. Locating decisions in a network makes sense be- cause field research suggests that decisions in one part of an organization can be strongly in- terdependent with decisions in other parts. For example, Zbaracki & Bergen (2010) analyze the classic example of setting prices. In principle, this task can be treated as an individual decision, but their analysis demonstrates that pricing decisions are not located in one person, one location, or one point in time. Price setting involves ongoing negotiations between sales, marketing, and other parts of an organization, each of which has competing interests and objectives. When the truce breaks down, the network of action is disrupted, and decisions about pricing are disrupted.

Dynamics: Stability and Change

The dynamics of organizational routines presents another potentially fruitful area for research. Organizational routines have often been associated with path dependence (Schulz 2008, Sydow et al. 2009) and inertia (Aldrich 1999). Path dependence refers to the tendency for past actions to influence future actions (Sydow et al. 2009). Schulz (2008) identifies a large number of mech- anisms that tend to support path dependence. Sydow et al. (2009) suggest that self-reinforcement tends to lead toward lock-in—a fixed, unchanging pattern of action. The tendency toward lock-in is associated with organizational inertia and competency traps (March 1991). The tendency

480 Pentland � Hærem

A nn

u. R

ev . O

rg an

. P sy

ch ol

. O rg

an . B

eh av

. 2 01

5. 2:

46 5-

48 7.

D ow

nl oa

de d

fr om

w w

w .a

nn ua

lr ev

ie w

s. or

g A

cc es

s pr

ov id

ed b

y U

ni ve

rs ity

o f

C al

if or

ni a

- R

iv er

si de

o n

08 /1

3/ 22

. F or

p er

so na

l u se

o nl

y.

toward self-reinforcing patterns has practical importance because, as Gersick&Hackman (1990) point out, routines can form very fast, in as little as a single performance.

Although path dependence explains inertia, routines can also undergo drift over time and endogenous change (Feldman & Pentland 2003). Garud et al. (2010) refer to this process as path creation or pathmaking. Using a computer simulation, Pentland et al. (2012) show that variations (due to exception, error, improvisation, or innovation) canbe incorporated over time, as newpaths are formed within the network of action that represents the routine. In this view, routines are dynamic systems; their tendency to stay the same or change remains an active topic of research (D’Adderio et al. 2012).

INTERNATIONAL AND CROSS-CULTURAL ISSUES

International and cross-cultural issues have not figured prominently in the research literature so far. There is an important opportunity here, however, because the idea of routinized patterns of action can be extended across any geographic scale. When people with diverse backgrounds find themselves engaged in a pattern of interdependent actions, it may exacerbate some of the problems discussed above.

PRACTICAL IMPLICATIONS

There are a number of practical implications that can be drawn from the research literature on organizational routines. These are not entirely novel, but it is reassuring that they resonatewith good managerial practice. The first implication concerns the importance of designing networks of actions and practices that flow toward desirable outcomes. If we adopt the view of managing as designing (Boland& Collopy 2004), then a good understanding of organizational routines will be crucial for good organizational design. Actions are not isolated from their social and material context or from other interdependent actions. By adopting a more inclusive view of the relevant actors and the pathways to success (e.g., a positive customer experience), the design is more likely to be robust.

Second, it is important to avoid overreliance on rules to implement the design. Rules are fraught with what Winter termed naive top-downism—the mistaken belief that once a manager makes a rule, it will be followed and will have the intended result (Pentland & Feldman 2008, Reynaud 2005). Even when encoded into computerized artifacts, rules do not necessarily have the desired effect. Rather than relying on rules, organizations should try to develop practices (Grote 2012).

Routines can be an efficient vehicle for learning because patterns of action can form fast, in as little as a single performance (Gersick &Hackman 1990). The resulting routine may (or may not) be reasonably well adapted to a particular context. If the context changes, the routine is likely to persist, unchanged. Thus, it is important to periodically review and revise established patterns. In other words, to remain adaptive and avoid competency traps, organizations need active meta- routines, routines for changing routines (Lewin et al. 2011, Winter 2003). Systematic review is important because, as Rice & Cooper (2010) point out, normal expectations about feedback and learning are often violated in organizational routines. Participants may not get usable information about outcomes, and even if they do get such information, and have some incentive to improve, they may be unable to do so because of the overall pattern of interdependence.

CONCLUSION

Organizational routines are a deceptively complex phenomenon. For many years, they were simply taken for granted—frequently mentioned, but basically ignored. Now, they have gained

481www.annualreviews.org � Organizational Routines

A nn

u. R

ev . O

rg an

. P sy

ch ol

. O rg

an . B

eh av

. 2 01

5. 2:

46 5-

48 7.

D ow

nl oa

de d

fr om

w w

w .a

nn ua

lr ev

ie w

s. or

g A

cc es

s pr

ov id

ed b

y U

ni ve

rs ity

o f

C al

if or

ni a

- R

iv er

si de

o n

08 /1

3/ 22

. F or

p er

so na

l u se

o nl

y.

attention as a legitimate topic in organizational research and an object of debate and even some controversy. The key to this renewed interest was fieldwork: taking a closer look at what people in organizations actually do. Fieldworkers spent time observing mundane, operational activities in awide variety of settings. As a result, we know that repetitive patterns of action are not necessarily mindless. Those repetitive patterns have become a viable unit of analysis and have been recognized as a central aspect of organizing, but the nuances of how they form, persist, and change have not been fully explained.

DISCLOSURE STATEMENT

The authors are not aware of any affiliations, memberships, funding, or financial holdings that might be perceived as affecting the objectivity of this review.

ACKNOWLEDGMENTS

Thanks go to Avinash Venkata Adavikolanu for his help in preparing the final document.

LITERATURE CITED

Abbott A. 1992. From causes to events: notes on narrative positivism. Sociol. Methods Res. 20(4):428–55 Abbott A. 1995. Sequence analysis: new methods for old ideas. Annu. Rev. Sociol. 21(1):93–113 Abell P. 1987. The syntax of social life: the theory and method of comparative narratives. Oxford, UK:

Clarendon Press Abell P, Felin T, Foss NJ. 2008. Building micro-foundations for the routines, capabilities, and performance

links. Manag. Decis. Econ. 29:489–502 Aldrich H. 1999. Organizations Evolving. Beverly Hills, CA: Sage Allison C, Hayes J. 2012. The Cognitive Style Index. London: Pearson Ashforth BE, Fried Y. 1988. The mindlessness of organizational behaviors. Hum. Relat. 41(4):305–29 Autor DH, Dorn D. 2013. The growth of low-skill service jobs and the polarization of the US labor market.

Am. Econ. Rev. 103(5):1553–97 Bal M. 1985. Narratology: Introduction to the Theory of Narrative. Toronto: Univ. Toronto Press Becker MC. 2004. Organizational routines: a review of the literature. Ind. Corp. Change 13:643–78 Becker MC. 2005. The concept of routines: some clarifications. Camb. J. Econ. 29:249–62 Becker MC, ed. 2008. Handbook of Organizational Routines. Cheltenham, UK: Edward Elgar Birnholtz JP, Cohen MD, Hoch SV. 2007. Organizational character: on the regeneration of Camp Poplar

Grove. Organ. Sci. 18(2):315–32 Blau P. 1955. The Dynamics of Bureaucracy. Chicago: Univ. Chicago Press Boland RJ Jr, Collopy F, eds. 2004. Managing as Designing. Stanford, CA: Stanford Univ. Press Boudreau MC, Robey D. 2005. Enacting integrated information technology: a human agency perspective.

Organ. Sci. 16(1):3–18 Bourdieu P. 1990. The Logic of Practice. Stanford, CA: Stanford Univ. Press Bresman H. 2013. Changing routines: a process model of vicarious group learning in pharmaceutical R&D.

Acad. Manag. J. 56:35–61 Bromiley P. 2010. Looking at prospect theory. Strateg. Manag. J. 31:1357–70 Brown AD, Lewis MA. 2011. Identities, discipline and routines. Organ. Stud. 32(7):871–95 Bruns HC. 2009. Leveraging functionality in safety routines: examining the divergence of rules and perfor-

mance. Hum. Relat. 62:1399–426 Brynjolfsson E, McAfee A. 2011. Race Against the Machine. Lexington, MA: Dig. Front. Burns J,QuinnM. 2011. The routinisation ofmanagement controls in software. J.Manag.Control22(1):5–24 Burns J, Scapens RW. 2008. Organizational routines in accounting. See Becker 2008, pp. 15–30

482 Pentland � Hærem

A nn

u. R

ev . O

rg an

. P sy

ch ol

. O rg

an . B

eh av

. 2 01

5. 2:

46 5-

48 7.

D ow

nl oa

de d

fr om

w w

w .a

nn ua

lr ev

ie w

s. or

g A

cc es

s pr

ov id

ed b

y U

ni ve

rs ity

o f

C al

if or

ni a

- R

iv er

si de

o n

08 /1

3/ 22

. F or

p er

so na

l u se

o nl

y.

ChanD. 1998. Functional relations among constructs in the same content domain at different levels of analysis: a typology of composition models. J. Appl. Psychol. 83(2):234–46

Cohen MD. 2007. Reading Dewey: reflections on the study of routine. Organ. Stud. 28(5):773–86 Cohen MD. 2012. Perceiving and remembering routine action: fundamental micro-level origins. J. Manag.

Stud. 49:1383–88 Cohen MD, Bacdayan P. 1994. Organizational routines are stored as procedural memory: evidence from

a laboratory study. Organ. Sci. 5(4):554–68 CohenMD,BurkhartR,DosiG, EgidiM,MarengoL, et al. 1996.Routines andother recurring actionpatterns

of organizations: contemporary research issues. Ind. Corp. Change 5(3):653–98 CohenMD,LevinthalDA,WarglienM.2014.Collective performance:modeling the interactionof habit-based

actions. Ind. Corp. Change 23:329–60 Cohendet P, Llerena P. 2003. Routines and incentives: the role of communities in the firm. Ind. Corp. Change

12:271–97 Constantinides P, BarrettM. 2012. A narrative networks approach to understanding coordination practices in

emergency response. Inf. Organ. 22(4):273–94 Corsaro WA, Heise DR. 1990. Event structure models from ethnographic data. Sociol. Methodol. 20:1–57 Cyert RM, March JG. 1963. A Behavioral Theory of the Firm. Englewood Cliffs, NJ: Prentice-Hall D’Adderio L. 2003. Configuring software, reconfiguring memories: the influence of integrated systems on the

reproduction of knowledge and routines. Ind. Corp. Change 12:321–50 D’Adderio L. 2008. The performativity of routines: theorising the influence of artefacts and distributed

agencies on routines dynamics. Res. Policy 37(5):769–89 D’Adderio L. 2011. Artifacts at the centre of routines: performing the material turn in routines theory. J. Inst.

Econ. 7(2):197–230 D’Adderio L, FeldmanMS, LazaricN, Pentland BT. 2012. Call for papers—special issue on routine dynamics:

exploring sources of stability and change in organizations. Organ. Sci. 23(6):1782–83 Dane E, Pratt MG. 2007. Exploring intuition and its role in managerial decision making. Acad. Manag. Rev.

32(1):33–54 DeSanctis G, Poole MS. 1994. Capturing the complexity in advanced technology use: adaptive structuration

theory. Organ. Sci. 5(2):121–47 Dionysiou DD, Tsoukas H. 2013. Understanding the (re)creation of routines from within: a symbolic

interactionist perspective. Acad. Manag. Rev. 38(2):181–205 Edmondson AC, Bohmer RM, Pisano GP. 2001. Disrupted routines: team learning and new technology

implementation in hospitals. Adm. Sci. Q. 46(4):685–716 Evans JSB. 2008. Dual-processing accounts of reasoning, judgment, and social cognition.Annu. Rev. Psychol.

59:255–78 Farjoun M. 2010. Beyond dualism: stability and change as a duality. Acad. Manag. Rev. 35(2):202–25 Feldman MS. 2000. Organizational routines as a source of continuous change. Organ. Sci. 11(6):611–29 FeldmanMS. 2003. A performative perspective on stability and change in organizational routines. Ind. Corp.

Change 12(4):727–52 Feldman MS, Orlikowski WJ. 2011. Theorizing practice and practicing theory. Organ. Sci. 22(5):1240–53 Feldman MS, Pentland BT. 2003. Reconceptualizing organizational routines as a source of flexibility and

change. Adm. Sci. Q. 48(1):94–118 Felin T, Foss NJ. 2005. Strategic organization: a field in search of micro-foundations. Strateg. Organ.

3(4):441–55 FelinT, FossNJ,HeimeriksKH,MadsenTL. 2012.Microfoundationsof routines and capabilities: individuals,

processes, and structure. J. Manag. Stud. 49(8):1351–74 Foldy EG, Buckley TR. 2010. Re-creating street-level practice: the role of routines, work groups, and team

learning. J. Public Adm. Res. Theory 20:23–52 Garud R, Kumaraswamy A, Karnøe P. 2010. Path dependence or path creation? J. Manag. Stud. 47:760–74 Gaskin JE, SchutzDM,BerenteN, LyytinenK. 2010. TheDNAof designwork: physical and digitalmateriality

in project-based design organizations. Acad. Manag. Proc. 2010(1):1–6 Geertz C. 1973. The Interpretation of Culture. New York: Basic Books

483www.annualreviews.org � Organizational Routines

A nn

u. R

ev . O

rg an

. P sy

ch ol

. O rg

an . B

eh av

. 2 01

5. 2:

46 5-

48 7.

D ow

nl oa

de d

fr om

w w

w .a

nn ua

lr ev

ie w

s. or

g A

cc es

s pr

ov id

ed b

y U

ni ve

rs ity

o f

C al

if or

ni a

- R

iv er

si de

o n

08 /1

3/ 22

. F or

p er

so na

l u se

o nl

y.

Gersick CJG, Hackman JR. 1990. Habitual routines in task-performing groups.Organ. Behav. Hum. Decis. Process. 47:65–97

Gherardi S. 2000. Practice-based theorizing on learning and knowing in organizations. Organization 7(2):211–23

Giddens A. 1984. The Constitution of Society. Oxford, UK: Polity Gigerenzer G. 2000. Adaptive Thinking: Rationality in the Real World. Oxford, UK: Oxford Univ. Press Gigerenzer G, Todd PM. 1999. Simple Heuristics That Make Us Smart. Oxford, UK: Oxford Univ. Press Goh JM, GaoG, Agarwal R. 2011. Evolvingwork routines: adaptive routinization of information technology

in healthcare. Inf. Syst. Res. 22(3):565–85 Greenhalgh T. 2008. Role of routines in collaborative work in healthcare organisations. BMJ 337:a2448 Greve HR. 2013. Microfoundations of management: behavioral strategies and levels of rationality in orga-

nizational action. Acad. Manag. Perspect. 27(2):103–19 Grote G. 2012. Safety management in different high-risk domains—all the same? Saf. Sci. 50(10):1983–92 Grote G, Weichbrodt JC, Günter H, Zala-Mezö E, Künzle B. 2009. Coordination in high-risk organizations:

the need for flexible routines. Cogn. Technol. Work 11(1):17–27 Gupta N, Hollingshead AB. 2010. Differentiated versus integrated transactive memory effectiveness: It

depends on the task. Group Dyn. 14(4):384–98 Hackman JR,WagemanR. 1995. Total qualitymanagement: empirical, conceptual and practical issues.Adm.

Sci. Q. 40:309–43 Hærem T, Pentland BT, Miller KD. 2015. Task complexity: extending a core concept. Acad. Manag. Rev. In

press. doi: 10.5465/amr.2013.0350 Hayes GR, Lee CP, Dourish P. 2011. Organizational routines, innovation, and flexibility: the application of

narrative networks to dynamic workflow. Int. J. Med. Inform. 80(8):e161–77 Hendricks WO. 1973. Methodology of narrative structural analysis. Semiotica 7:163–84 Hodgkinson GP, Sadler-Smith E, Burke LA, Claxton G, Sparrow PR. 2009. Intuition in organizations:

implications for strategic management. Long Range Plann. 42(3):277–97 Hodgson GM. 2008. The concept of a routine. See Becker 2008, pp. 15–30 Hogarth RM. 2001. Educating Intuition. Chicago: Univ. Chicago Press Hollingshead AB. 2001. Cognitive interdependence and convergent expectations in transactive memory.

J. Personal. Soc. Psychol. 81(6):1080–89 Howard-Grenville JA. 2005. The persistence of flexible organizational routines: the role of agency and or-

ganizational context. Organ. Sci. 16(6):618–36 Hutchins E. 1995. Cognition in the Wild. Cambridge, MA: MIT Press Kahneman D. 2011. Thinking, Fast and Slow. New York: Macmillan Kahneman D, Klein G. 2009. Conditions for intuitive expertise: a failure to disagree. Am. Psychol. 64(6):515 Kahneman D, Tversky A. 1974. Subjective probability: a judgment of representativeness. In The Concept of

Probability in Psychological Experiments, ed. CA Staël von Holstein, pp. 25–48. Dordrecht, Neth.: Springer

Keren G, Schul Y. 2009. Two is not always better than one: a critical evaluation of two-system theories. Perspect. Psychol. Sci. 4(6):533–50

Kozlowski SWJ, Chao GT, Grand JA, Braun MT, Kuljanin G. 2013. Advancing multilevel research design: capturing the dynamics of emergence. Organ. Res. Methods 16(4):581–615

Kruglanski AW, Gigerenzer G. 2011. Intuitive and deliberate judgments are based on common principles. Psychol. Rev. 118(1):97–109

Langer EJ. 1989. Mindfulness. Reading, MA: Addison-Wesley Latour B. 2005. Reassembling the Social: An Introduction to Actor-Network-Theory. Oxford, UK: Oxford

Univ. Press Lave J. 1988. Cognition in Practice: Mind, Mathematics and Culture in Everyday Life. Cambridge, UK:

Cambridge University Press Law J. 1992. Notes on the theory of the actor-network: ordering, strategy, and heterogeneity. Syst. Pract.

5(4):379–93

484 Pentland � Hærem

A nn

u. R

ev . O

rg an

. P sy

ch ol

. O rg

an . B

eh av

. 2 01

5. 2:

46 5-

48 7.

D ow

nl oa

de d

fr om

w w

w .a

nn ua

lr ev

ie w

s. or

g A

cc es

s pr

ov id

ed b

y U

ni ve

rs ity

o f

C al

if or

ni a

- R

iv er

si de

o n

08 /1

3/ 22

. F or

p er

so na

l u se

o nl

y.

Lazaric N. 2011. Organizational routines and cognition: an introduction to empirical and analytical con- tributions. J. Inst. Econ. 7(2):147–56

LazaricN,Denis B. 2005.Routinization andmemorization of tasks in aworkshop: the case of the introduction of ISO norms. Ind. Corp. Change 14:873–96

Leonardi PM. 2011. When flexible routines meet flexible technologies: affordance, constraint, and the im- brication of human and material agencies. Manag. Inf. Syst. Q. 35(1):147–67

Levinthal D, Rerup C. 2006. Crossing an apparent chasm: bridging mindful and less-mindful perspectives on organizational learning. Organ. Sci. 17(4):502–13

LewinAY,Massini S, PeetersC. 2011.Microfoundations of internal and external absorptive capacity routines. Organ. Sci. 22(1):81–98

Lewis K. 2003.Measuring transactive memory systems in the field: scale development and validation. J. Appl. Psychol. 88(4):587–604

March JG. 1991. Exploration and exploitation in organizational learning. Organ. Sci. 2(1):71–87 March JG, Simon HA. 1958. Organizations. Newark: NJ: Wiley Meyer RD, Dalal RS, Hermida R. 2010. A review and synthesis of situational strength in the organizational

sciences. J. Manag. 36(1):121–40 MilkmanKL, ChughD, BazermanMH. 2009.How can decisionmaking be improved? Perspect. Psychol. Sci.

4(4):379–83 Miller KD, Pentland BT, Choi S. 2012. Dynamics of performing and remembering organizational routines.

J. Manag. Stud. 49(8):1536–58 NelsonRR,Winter SG. 1982.AnEvolutionary Theory of EconomicChange. Cambridge,MA:HarvardUniv.

Press Nicolini D. 2011. Practice as the site of knowing: insights from the field of telemedicine. Organ. Sci.

22(3):602–20 Norman DA. 2002. The Design of Everyday Things. New York: Basic Books NybergD. 2009.Computers, customer service operatives and cyborgs: intra-actions in call centres.Organ. Sci.

30(11):1181–99 Orlikowski WJ. 2007. Sociomaterial practices: exploring technology at work. Organ. Stud. 28(9):1435–48 PaciniR, Epstein S. 1999. The relationof rational and experiential informationprocessing styles to personality,

basic beliefs, and the ratio-bias phenomenon. J. Personal. Soc. Psychol. 76(6):972–87 Parmigiani A, Howard-Grenville J. 2011. Routines revisited: exploring the capabilities and practice per-

spectives. Acad. Manag. Ann. 5(1):413–53 Pentland BT. 2003. Conceptualizing and measuring variety in organizational work processes. Manag. Sci.

49(7):857–70 Pentland BT, Feldman MS. 2005. Organizational routines as a unit of analysis. Ind. Corp. Change

14(5):793–815 Pentland BT, Feldman MS. 2007. Narrative networks: patterns of technology and organization. Organ. Sci.

18(5):781–95 Pentland BT, Feldman MS. 2008. Designing routines: on the folly of designing artifacts, while hoping for

patterns of action. Inf. Organ. 18(4):235–50 Pentland BT, FeldmanMS, BeckerMC, Liu P. 2012.Dynamics of organizational routines: a generativemodel.

J. Manag. Stud. 49(8):1484–508 Pentland BT, Hærem T, Hillison D. 2010. Comparing organizational routines as recurrent patterns of action.

Organ. Stud. 31(7):917–40 Pentland BT, Hærem T, Hillison D. 2011. The (n)ever changing world: stability and change in organizational

routines. Organ. Sci. 22(6):1369–83 Pentland BT, Rueter HH. 1994. Organizational routines as grammars of action.Adm. Sci. Q. 39(3):484–510 Plous S. 1993. The Psychology of Judgment and Decision Making. New York: McGraw-Hill PuranamP,RaveendranM,KnudsenT. 2012.Organizationdesign: the epistemic interdependenceperspective.

Acad. Manag. Rev. 37(3):419–40 Quinn M. 2014. Stability and change in management accounting over time: a century or so of evidence from

Guinness. Manag. Account. Res. 25:76–92

485www.annualreviews.org � Organizational Routines

A nn

u. R

ev . O

rg an

. P sy

ch ol

. O rg

an . B

eh av

. 2 01

5. 2:

46 5-

48 7.

D ow

nl oa

de d

fr om

w w

w .a

nn ua

lr ev

ie w

s. or

g A

cc es

s pr

ov id

ed b

y U

ni ve

rs ity

o f

C al

if or

ni a

- R

iv er

si de

o n

08 /1

3/ 22

. F or

p er

so na

l u se

o nl

y.

Radwan L, Kinder S. 2013. Practising the diffusion of organizational routines.Environ. Plann. A 45:2442–58 Ragin CC, Becker HS. 1992. What is a Case? Exploring the Foundations of Social Inquiry. Cambridge, UK:

Cambridge Univ. Press Rerup C, FeldmanMS. 2011. Routines as a source of change in organizational schemata: the role of trial-and-

error learning. Acad. Manag. J. 54:577–610 Reynaud B. 2005. The void at the heart of rules: routines in the context of rule-following. The case of the Paris

Metro Workshop. Ind. Corp. Change 14:847–71 Rice RE, Cooper SD. 2010. Organizations and Unusual Routines: A Systems Analysis of Dysfunctional

Feedback Processes. Cambridge, UK: Cambridge Univ. Press Robey D, Raymond B, Anderson C. 2012. Theorizing information technology as a material artifact in in-

formation systems research. In Materiality and Organizing, ed. PM Leonardi, BA Nardi, J Kallinikos, pp. 217–36. Oxford, UK: Oxford Univ. Press

Rousseau DM. 1997. Organizational behavior in the new organizational era. Annu. Rev. Psychol. 48(1):515–46

Russo JE, Schoemaker PJH. 2002. Winning Decisions: Getting It Right the First Time. New York: Random House

Ryle G. 1949. The Concept of Mind. London: Hutchinson House Sánchez D, Tentori M, Favela J. 2008. Activity recognition for the smart hospital. IEEE Intell. Syst.

23(2):50–57 Schneider W, Chein JM. 2003. Controlled & automatic processing: behavior, theory, and biological

mechanisms. Cogn. Sci. 27(3):525–59 Schulz M. 2008. Staying on track: a voyage to the internal mechanisms of routine reproduction. See Becker

2008, pp. 228–55 SilverMS. 1991.Decisional guidance for computer-based decision support.Manag. Inf. Syst.Q. 15(1):105–22 Simon HA. 1959. Theories of decision-making in economics and behavioral science. Am. Econ. Rev.

49(3):253–83 Simon HA. 1987. Making management decisions: the role of intuition and emotion. Acad. Manag. Exec.

1(1):57–64 Simon HA. 1992. What is an “explanation” of behavior? Psychol. Sci. 3(3):150–61 SleesmanDJ,ConlonDE,McNamaraG,Miles JE. 2012.Cleaning up the bigmuddy: ameta-analytic reviewof

the determinants of escalation of commitment. Acad. Manag. J. 55(3):541–62 Stanovich KE, West RF. 2000. Advancing the rationality debate. Behav. Brain Sci. 23(05):701–17 Su NM, Brdiczka O, Begole B. 2013. The routineness of routines: measuring rhythms of media interaction.

Hum. Comput. Interact. 28(4):287–334 Suchman LA. 1987. Plans and Situated Actions: The Problem of Human-Machine Communication.

Cambridge, UK: Cambridge Univ. Press Sydow J, Schreyögg G, Koch J. 2009. Organizational path dependence: opening the black box.Acad. Manag.

Rev. 34(4):689–709 Thompson JD. 1967. Organizations in Action: Social Science Bases of Administrative Theory. New York:

McGraw-Hill Tsoukas H, Chia R. 2002. On organizational becoming: rethinking organizational change. Organ. Sci.

13(5):567–82 Turner SF, Fern MJ. 2012. Examining the stability and variability of routine performances: the effects of

experience and context change. J. Manag. Stud. 49(8):1407–34 Turner SF, Rindova V. 2012. A balancing act: how organizations pursue consistency in routine functioning in

the face of ongoing change. Organ. Sci. 23(1):24–46 Tversky A, KahnemanD. 1973. Availability: a heuristic for judging frequency and probability.Cogn. Psychol.

5(2):207–32 Tversky A, Kahneman D. 1974. Judgment under uncertainty: heuristics and biases. Science

185(4157):1124–31 Vromen JJ. 2011. Routines as multilevel mechanisms. J. Inst. Econ. 7:175–96 Weick KE. 1969. The Social Psychology of Organizing. Reading, MA: Addison-Wesley

486 Pentland � Hærem

A nn

u. R

ev . O

rg an

. P sy

ch ol

. O rg

an . B

eh av

. 2 01

5. 2:

46 5-

48 7.

D ow

nl oa

de d

fr om

w w

w .a

nn ua

lr ev

ie w

s. or

g A

cc es

s pr

ov id

ed b

y U

ni ve

rs ity

o f

C al

if or

ni a

- R

iv er

si de

o n

08 /1

3/ 22

. F or

p er

so na

l u se

o nl

y.

Weick KE, Roberts KH. 1993. Collective mind in organizations: heedful interrelating on flight decks. Adm. Sci. Q. 38(3):357–81

Wenger E. 1998. Communities of Practice: Learning, Meaning, and Identity. Cambridge, UK: Cambridge Univ. Press

Whyte G. 1989. Groupthink reconsidered. Acad. Manag. Rev. 14(1):40–56 Winter SG. 1985. The case for “mechanistic” decision making. In Organizational Strategy and Change, ed.

JM Pennings, pp. 99–113. San Francisco: Jossey-Bass Winter SG. 2003. Understanding dynamic capabilities. Strateg. Manag. J. 24(10):991–95 Winter SG. 2013. Habit, deliberation, and action: strengthening the microfoundations of routines and

capabilities. Acad. Manag. Perspect. 27(2):120–37 Yeow A, Faraj S. 2011. Using narrative networks to study enterprise systems and organizational change. Int.

J. Account. Inf. Syst. 12(2):116–25 Zammuto RF, Griffith TL, Majchrzak A, Dougherty DJ, Faraj S. 2007. Information technology and the

changing fabric of organization. Organ. Sci. 18(5):749–62 Zbaracki MJ, Bergen M. 2010. When truces collapse: a longitudinal study of price-adjustment routines.

Organ. Sci. 21(5):955–72

487www.annualreviews.org � Organizational Routines

A nn

u. R

ev . O

rg an

. P sy

ch ol

. O rg

an . B

eh av

. 2 01

5. 2:

46 5-

48 7.

D ow

nl oa

de d

fr om

w w

w .a

nn ua

lr ev

ie w

s. or

g A

cc es

s pr

ov id

ed b

y U

ni ve

rs ity

o f

C al

if or

ni a

- R

iv er

si de

o n

08 /1

3/ 22

. F or

p er

so na

l u se

o nl

y.

Annual Review of

Organizational

Psychology and

Organizational Behavior

Volume 2, 2015 Contents

Organizational Psychology Then and Now: Some Observations Edgar H. Schein . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

Group Affect Sigal G. Barsade and Andrew P. Knight . . . . . . . . . . . . . . . . . . . . . . . . . . 21

The Modeling and Assessment of Work Performance John P. Campbell and Brenton M. Wiernik . . . . . . . . . . . . . . . . . . . . . . . . 47

Justice, Fairness, and Employee Reactions Jason A. Colquitt and Kate P. Zipay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75

Methodological and Substantive Issues in Conducting Multinational and Cross-Cultural Research Paul E. Spector, Cong Liu, and Juan I. Sanchez . . . . . . . . . . . . . . . . . . . . 101

Leadership Development: An Outcome-Oriented Review Based on Time and Levels of Analyses David V. Day and Lisa Dragoni . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133

Beyond Lewin: Toward a Temporal Approximation of Organization Development and Change Jean M. Bartunek and Richard W. Woodman . . . . . . . . . . . . . . . . . . . . . 157

Beyond the Big Five: New Directions for Personality Research and Practice in Organizations Leaetta M. Hough, Frederick L. Oswald, and Jisoo Ock . . . . . . . . . . . . . 183

Corporate Social Responsibility: Psychological, Person-Centric, and Progressing Deborah E. Rupp and Drew B. Mallory . . . . . . . . . . . . . . . . . . . . . . . . . 211

Time in Individual-Level Organizational Studies: What Is It, How Is It Used, and Why Isn’t It Exploited More Often? Abbie J. Shipp and Michael S. Cole . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237

vi

A nn

u. R

ev . O

rg an

. P sy

ch ol

. O rg

an . B

eh av

. 2 01

5. 2:

46 5-

48 7.

D ow

nl oa

de d

fr om

w w

w .a

nn ua

lr ev

ie w

s. or

g A

cc es

s pr

ov id

ed b

y U

ni ve

rs ity

o f

C al

if or

ni a

- R

iv er

si de

o n

08 /1

3/ 22

. F or

p er

so na

l u se

o nl

y.

Dynamics of Well-Being Sabine Sonnentag . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261

Low-Fidelity Simulations Jeff A. Weekley, Ben Hawkes, Nigel Guenole, and Robert E. Ployhart . . . 295

Emotional Labor at a Crossroads: Where Do We Go from Here? Alicia A. Grandey and Allison S. Gabriel . . . . . . . . . . . . . . . . . . . . . . . . . 323

Supporting the Aging Workforce: A Review and Recommendations for Workplace Intervention Research Donald M. Truxillo, David M. Cadiz, and Leslie B. Hammer . . . . . . . . . 351

ESM 2.0: State of the Art and Future Potential of Experience Sampling Methods in Organizational Research Daniel J. Beal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 383

Ethical Leadership Deanne N. Den Hartog . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 409

Differential Validity and Differential Prediction of Cognitive Ability Tests: Understanding Test Bias in the Employment Context Christopher M. Berry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 435

Organizational Routines as Patterns of Action: Implications for Organizational Behavior Brian T. Pentland and Thorvald Hærem . . . . . . . . . . . . . . . . . . . . . . . . . 465

Pay, Intrinsic Motivation, Extrinsic Motivation, Performance, and Creativity in the Workplace: Revisiting Long-Held Beliefs Barry Gerhart and Meiyu Fang . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 489

Stereotype Threat in Organizations: Implications for Equity and Performance Gregory M. Walton, Mary C. Murphy, and Ann Marie Ryan . . . . . . . . . . 523

Technology and Assessment in Selection Nancy T. Tippins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 551

Workplace Stress Management Interventions and Health Promotion Lois E. Tetrick and Carolyn J. Winslow . . . . . . . . . . . . . . . . . . . . . . . . . 583

Errata

An online log of corrections to Annual Review of Organizational Psychology and Organizational Behavior articles may be found at http://www.annualreviews.org/ errata/orgpsych.

Contents vii

A nn

u. R

ev . O

rg an

. P sy

ch ol

. O rg

an . B

eh av

. 2 01

5. 2:

46 5-

48 7.

D ow

nl oa

de d

fr om

w w

w .a

nn ua

lr ev

ie w

s. or

g A

cc es

s pr

ov id

ed b

y U

ni ve

rs ity

o f

C al

if or

ni a

- R

iv er

si de

o n

08 /1

3/ 22

. F or

p er

so na

l u se

o nl

y.

A nn

u. R

ev . O

rg an

. P sy

ch ol

. O rg

an . B

eh av

. 2 01

5. 2:

46 5-

48 7.

D ow

nl oa

de d

fr om

w w

w .a

nn ua

lr ev

ie w

s. or

g A

cc es

s pr

ov id

ed b

y U

ni ve

rs ity

o f

C al

if or

ni a

- R

iv er

si de

o n

08 /1

3/ 22

. F or

p er

so na

l u se

o nl

y.

1. ar:
1. logo: