ORGANIZATIONAL PATH DEPENDENCE: OPENING THE BLACK BOX · 2012-11-16 · ORGANIZATIONAL PATH...

ORGANIZATIONAL PATH DEPENDENCE:OPENING THE BLACK BOX

JORG SYDOWGEORG SCHREYOGG

Freie Universität Berlin

JOCHEN KOCHEuropean University Viadrina

To enable a better understanding of the underlying logic of path dependence, we setforth a theoretical framework explaining how organizations become path dependent.At its core are the dynamics of self-reinforcing mechanisms, which are likely to leadan organization into a lock-in. By drawing on studies of technological paths, weconceptualize the emergent process of path dependence along three distinct stages.We also use the model to explore breakouts from organizational path dependence anddiscuss implications for managing and researching organizational paths.

The discourse on organizational innovationand change has become more complex. On theone hand, there is an ever-increasing demandfor more flexible or even fluid “new” organiza-tional forms. On the other hand, studies stress-ing organizational inertia and the historical im-printing of decision making (“history matters”)have come to the fore in management and or-ganizational theory. There seems to be abroadly shared feeling that we need to under-stand better how organizations can lose theirflexibility and become inert or even locked in.Among the most conceptions, referred to pathdependence has recently gained prominence.Many contributions refer to path dependence toilluminate organizational rigidities, stickiness,or inflexibility.1 But what is path dependence

supposed to mean exactly? In organizational re-search the term is used mostly as a broad labelindicating all kinds of imprinting effects of thepast on organizational behavior (e.g., recently,Beckman & Burton, 2008). A closer examinationquickly reveals that the predominant usage ismore metaphorical than theoretical in nature. Aclear specification is usually missing. Thismeans, at the same time, that no indicators areavailable that allow for examining whether ornot the organizational phenomena in questionare actually path dependent. If we want “pathdependence” to provide more than a synonymfor persistence, then we need a theoreticalframework clarifying the notion and helping usbetter understand the conditions and dynamicsunder which organizations become path depen-dent. By addressing this gap in managementand organization research, we aim to offer aframework designed to explain organizationalpath dependence.The endeavor to explain organizational rigid-

ities and structural inertia is not new in man-agement and organization research. Over theyears, scholars have accumulated ample evi-dence on change-inhibiting forces. Various stud-ies have highlighted cases of persistence andirreversibility of organizational strategies, de-signs, and competences by drawing, for in-stance, on awkward routines, groupthink, orfixed cognitive maps (e.g., Beckman & Burton,2008; Burgelman, 2002; Collinson & Wilson, 2006;Gilbert, 2005; Hannan & Freeman, 1984; Helfat,1994; Huff & Huff, 2000; Stimpert, Wasserman, &

We thank the four anonymous reviewers for their helpfulcomments, and we particularly thank former associate edi-tor Pamela Tolbert for her thoughtful advice. Earlier versionsof the manuscript profited significantly from discussions inthe subgroup on path dependence and creation at the 21stEGOS Colloquium in Berlin, 2005, and at the annual meetingof the Academy of Management in Atlanta, 2006. We aregrateful to the German Research Foundation (DFG) for fund-ing the doctoral program Research on Organizational Pathsat Freie Universitat Berlin and to its members for providinga stimulating environment for the research.

1 A quick search for references to path dependence inpapers published between 1995 and 2008 in three leadingscholarly journals (Administrative Science Quarterly, Orga-nization Science and Organization Studies) showed thatmore than eighty papers referred to this concept. That isabout 4.3 per cent of the articles published in those journalsover this time span—an average of 0.3 papers per issue.

� Academy of Management Review2009, Vol. 34, No. 4, 1–000.

1

Fn1

Copyright of the Academy of Management, all rights reserved. Contents may not be copied, emailed, posted to a listserv, or otherwise transmitted without the copyrightholder’s express written permission. Users may print, download, or email articles for individual use only.

Jayaran, 1998; Tripsas & Gavetti, 2000). Path de-pendence, however, is supposed to mean morethan the mere existence of timeworn routines,cognitive rigidities, or structural inertia. It is,first of all, a process. Its distinguishing featuresneed elaboration.The starting point of any advanced path de-

pendence thought stresses the importance ofpast events for future action or, in a more fo-cused way, of foregoing decisions for currentand future decision making. Hence, decisionsare conceived of as historically conditioned—“bygones are rarely bygones” (Teece, Pisano, &Shuen, 1997: 522). In short, the basic thesis holdsthat history matters (e.g., Nooteboom, 1997;Sewell, 1996).This essential insight has certainly advanced

the understanding of emerging organizationalphenomena and has helped to overcome theahistorical and unbounded view of rationalchoice thought. We learn that history can bequite important for explaining strategic choicesand organizational failures. While we appreci-ate this insight, merely focusing on the fact ofpast dependence (Antonelli, 1999) implies takinga fairly broad view—too broad a theoretical per-spective: if we base path dependence explana-tions on the history matters argument only, thenotion is likely to become indistinct. All humanactivity and organizational processes are im-printed by their history in a way, so we wouldend up by concluding that all organizationaldecisions and actions are path dependent. Sucha ubiquitous, all-embracing understanding ofpath dependence would bring us close to a tru-ism. Path dependence relates to more specificconstellations; it includes features such as sus-tained persistency and lock-in, which are defi-nitely not a common characteristic of decisionprocesses. A theory of organizational path de-pendence therefore needs a more elaboratedframework, which takes us beyond the mere in-sight that past events influence subsequentactions.To gain a deeper understanding of the organ-

izational patterns considered to be path depen-dent, along with their underlying causal mech-anisms, it is instructive to explore the cases andconceptual suggestions provided by studies ontechnological paths. Paul David (1985, 1986) pro-vides the most prominent example of technolog-ical path dependence—the well-known stan-dard of the QWERTY keyboard and its amazing

predominance for more than 100 years. Thisstandard has spread around the world and, puz-zlingly enough, has never been seriously chal-lenged by all the newly developed, technicallymore efficient alternatives. David explains thisinefficient long-term predominance as being theresult of a path-dependent process, which wasset up owing to some initial events and ad-vanced mainly through network externalitiesleading into a technological lock-in early on.The QWERTY case and similar case studies

from technology diffusion, economic history,and evolutionary economics (e.g., Antonelli,1999; Callon, 1992; Castaldi & Dosi, 2006; Dosi,1982; Hughes, 1987) offer intriguing evidence ofsimilar persistence in national and global con-texts. Arthur (1989, 1994) was the first to model aformal theory of path dependence and to exposeincreasing returns as the major process driver.Later on, this thinking was extended to the eco-nomics of institutions (North, 1990). However, upto now, studies of path dependence (in this spec-ified sense) neither addressed the persistence oforganizations nor explored the logic and dy-namics of internal organizational processesleading to a lock-in. Here we therefore aim toelaborate a theory of organizational path depen-dence and lock-in. Building on the evidence andinsights from research on technological paths,we develop a theoretical framework to gain abetter understanding of how organizationalpath dependence comes into existence. In pur-suing this aim, we also integrate insights frominstitutional economics (North, 1990), as well asfrom political science (in particular, Mahoney,2000; Pierson, 2000, 2004; Thelen, 1999), theoriesof institutionalization (Lawrence, Winn, & Jen-nings, 2001; Powell & DiMaggio, 1991; Tolbert &Zucker, 1996), and related organization studies(Boeker, 1989; Johnson, 2007; Stinchcombe, 1965).In a subsequent section we will, however, alsoshow precisely where the differences to theseorganizational approaches can be found.In essence, we suggest a framework that dif-

ferentiates three developmental phases of pathdependence, starting with (1) singular historicalevents, (2) which may, under certain conditions,transform themselves into self-reinforcing dy-namics, and (3) possibly end up in an organiza-tional lock-in. The three phases are each as-sumed to be governed by different regimes. Thesuggested model aims at providing an explan-atory framework but also an operational scheme

2 OctoberAcademy of Management Review

for investigating claimed path dependence inand of organizations. Furthermore, we explorewhether and, how organizational path depen-dencies and lock-ins can be overcome (un-locked). We conclude by considering implica-tions for research and management.

ADVANCING A DYNAMIC FRAMEWORK OFORGANIZATIONAL PATH DEPENDENCE

Valuable insights into the dynamic nature ofentrapping or locking processes have alreadybeen provided by studies from evolutionary eco-nomics and economic history, although these, asalready indicated, focus almost exclusively ontechnological innovation at the field or marketlevel (Arthur, 1989, 1994; David, 1985, 1986; Dosi,1982, 1997). At the core, these studies identifyself-reinforcing processes as drivers that arelikely to accumulate in a specific path of action.These inherent self-reinforcing dynamics thateventually lead to an irreversible state of totalinflexibility or lock-in (David, 1985) are seen asbecoming increasingly systemic forces, beyondthe control of the individual actor. In otherwords, the individual actor becomes entrappedin the system’s dynamics.It is difficult to conceptualize the general logic

of this type of entrapping process. Arthur (1994;see also Pierson, 2000: 253) has advanced itsmost explicit characterization. In his view theprocess of becoming path dependent can becharacterized by four general properties:

1. Nonpredictability—there is an indetermi-nacy of outcome.

2. Nonergodicity—several outcomes are possi-ble (multiple equilibria), and history selectsamong the possible alternatives.

3. Inflexibility—the actors are entrapped, so ashift to another option is impossible.

4. Inefficiency—actions resulting from thepath lock the market into an inferior solu-tion.

These four properties provide a first orienta-tion for differentiating between path-dependentand non-path-dependent processes. However,the properties seem to be somewhat overgener-alized, and they do not actually apply to thewhole process of becoming locked into a path.Rather, they appear to cover specific episodes inthis process. Take, for instance, unpredictabil-ity; this trait applies only to the beginning of theprocess, when the outcome is actually unpre-

dictable. Later on, as the path is increasinglyformed, by implication, the actions becomemoreand more predictable. Having arrived at thelock-in stage, the behavior even becomes fullypredictable. The reverse is true for nonergodic-ity and inflexibility: it is only at the later stagesthat a path process rigidifies. In the beginningthe process is assumed to be flexible. And, sim-ilarly, inefficiency becomes a feature of the laterstages only; initially, before a path is shaping,the situation is open (“unpredictable”) and—asArthur (1994: 116) himself stresses—the choicesmay well be efficient. It is only at a later stagethat a more efficient option may emerge, whichactors can no longer choose because they arelocked in, thus causing inefficiency.These brief considerations advise us to differ-

entiate among explicitly different stages in theformation of a path and to specify their struc-tural properties. To elaborate on a theory of or-ganizational path dependence, we thereforesuggest subdividing the whole process of evolv-ing path dependence into three stages governedby different causal regimes and constituting dif-ferent settings for organizational action and de-cision making.

Phase I—the Preformation Phase—is charac-terized by a broad scope of action. The effect ofa choice of options cannot be predicted (see alsoMahoney, 2000: 511). Once a decision is made,this choice may, however, amount to a smallevent that unintentionally sets off a self-reinforcing process. This moment of enteringinto the dynamics of a self-reinforcing processcan be thought of as a “critical juncture” (Collier& Collier, 1991), and it indicates the end of thePreformation Phase. Drawing on complexity the-ory, this transition comes close to “bifurcation”(Kauffman, 1993).In Phase II—the Formation Phase—a new re-

gime takes the lead: the dynamics of self-reinforcing processes (Arthur, 1994). A dominantaction pattern is likely to emerge, which rendersthe whole process more and more irreversible.By implication, the range of options narrows,and it becomes progressively difficult to reversethe initial choice or the initial pattern of ac-tion—that is, a path is evolving. Decision pro-cesses in Phase II are, however, still contingent;they are “nonergodic”—not accidental, but theydo not yet fully converge to a fixed-point distri-bution (David, 1985).

2009 3Sydow, Schreyogg, and Koch

The transition from Phase II to Phase III—theLock-in Phase—is characterized by a furtherconstriction, which eventually leads to a lock-in—that is, the dominant decision pattern be-comes fixed and gains a deterministic charac-ter; eventually, the actions are fully bound to apath. One particular choice or action pattern hasbecome the predominant mode, and flexibilityhas been lost. Even new entrants into this fieldof action cannot refrain from adopting it. Whenfaced by more efficient alternatives, individuals’and organizations’ decision processes and es-tablished practices continue to reproduce thisand only this particular outcome. The occur-rence of a lock-in renders a system potentiallyinefficient, because it loses its capability toadopt better alternatives.Figure 1 illustrates the process across the

three stages. This differentiated framework isintended as a general model of path depen-dence; its functioning, however, is likely to differfrom context to context according to the prevail-ing conditions, particularly market versus hier-archy. The contextual specifics when applied toan organizational context—the target field ofthis contribution—will be outlined in subse-quent sections.

Preformation Phase

Phase I can be characterized as an open situ-ation with no significantly restricted scope ofaction. From a theoretical point of view, the

question that arises is how this initial state canbe conceptualized in more distinctive terms. Thetechnological path studies—if at all—have con-ceived of the initial situation as being unre-stricted. The search for alternatives starts fromscratch, and decisions are unconstrained.Such framing of the first stage in the rational

choice tradition, however, paradoxically ignoresthe fact that the development of a path is em-bedded and connected with other developments;it cannot be considered a completely separateprocess without any imprints from the past. Inbrief, history matters in the Preformation Phase,too. In organizations, initial choices and actionsare embedded in routines and practices, theyreflect the heritage—the rules and the culture—making up those institutions (e.g., Child, 1997;March, 1994; Tolbert & Zucker, 1996). Institutionsare “carriers of history” (David, 1994), and his-tory cannot be intermittent; it does not matteronly occasionally—it always matters! A concep-tualization of the activities in the PreformationPhase thus cannot start from scratch; it has toaccount for institutional imprints.On the other hand, history in this broad sense

is not destiny; we have to draw a clear distinc-tion between historical-institutional influencesand imperatives. The notion of path dependencedoes not refer to a state of determinacy from thebeginning; it sheds light on a tapering processthat possibly ends in a lock-in. Increasing pathdependence implies an initial scope of choice.Otherwise, the theory would lose its very point:

FIGURE 1The Constitution of an Organizational Path


F1

to make tapering processes in organizationalreality better understood. Thus, Phase I shouldbuild on a historically framed or imprinted con-tingency and, therefore, neither on the assump-tion of determinacy nor on that of completelyunrestricted choice. The shadow in Phase I inFigure 1 is intended to indicate this institutionalheritage.A related issue is the triggering of further re-

actions. The initial choice in a process that be-comes path dependent later on is not simply asingle event; it is an impetus, a trigger stimulat-ing further actions, which may accumulate in anorganizational path. Arthur (1994: 14) character-izes these initial choices as “small events.”Drawing on complexity theory (Kauffman, 1993),we conceptualize the triggering as bifurcation:small events may cause unintended, far-reach-ing consequences—as is the case, for instance,with the well-known butterfly effect.2 Becauseseveral outcomes initially are possible, the his-torical sequence of choices becomes decisive indetermining the final outcome; the first choice(or action) however, is random (David, 1985).It is doubtless appealing to conceive of trig-

gers of path dependence as small and randomevents, but for the purpose of organizationalanalysis, we need to expand the scope. Sinceorganizations are social systems and not mar-kets or natural entities, triggering events in or-ganizations are likely to prove to be not so in-nocent, random, or “small” (cf. also Bassanini &Dosi, 2001). For instance, in the case of the VHSmonopoly that has been intensively studiedfrom a path dependence perspective (Cu-sumano, Mylonadis, & Rosenbloom, 1992), thetriggering event was neither a random nor asmall one. Rather, it was Matsushita’s initialmove to secure content delivery through anagreement with major Hollywood studios thathappened to become the crucial step in defeat-ing the technologically superior Sony Beta stan-dard. Similarly, in Cowan’s (1990) study of nu-clear power plants, initial choices reflectedintentions, not randomness. These (and other)

cases invite a rethinking of the small eventframing: path dependence may be triggered by“bigger” events or even strategies as well.A less randomized modeling of these initial

activities thus seems advisable—at least for or-ganizational contexts. It is, however, importantto realize that in the nonlinear logic of pathdependence, irrespective of whether the initialactions are big or small, they can never be con-sidered causal determinants. A determined pro-cess would follow a prescribed course of eventsright from the beginning, as is the case withlinear cause-and-effect laws. Opposed to that,the very point of early path developments is thatthey are contingent in character. Moreover, theiroutcomes are unforeseeable consequences ofpurposeful action (Merton, 1936). The outcomecannot be known unless the process has beenformed.

Formation Phase

Phase II is characterized by the gradual emer-gence of an organizational path. The scope ofaction is assumed to narrow increasingly be-cause of the “pull” of the evolving path. An ini-tially unknown regime3 happens to take thelead, which favors a particular type of decisionor action pattern and reproduces them over acertain period of time. This phase commenceswith a critical juncture at the passage fromPhase I to II. A decision made or an action takenin Phase I amounts to a trigger for the furtherdevelopment of the organization or an organiza-tional subsystem. However, not all cases of com-peting solutions culminate in path dependence.It is therefore of critical importance to indicatesuch cases in which path dependence is likelyto develop.Early studies on technological path depen-

dence (David, 1985, 1986) highlighted the centralrole of self-reinforcing processes for path build-ing. Arthur (1989, 1994) elaborated on these driv-ing forces and specified “increasing returns” asthe decisive feature; this builds on the assump-tion that the decision to reproduce a particular

2 Here the flap of a butterfly’s wings represents a smallrandom change in the initial condition of the system (atmo-sphere), which sets in motion a chain of events eventuallycausing a large-scale change (tornado). Had the small eventnot occurred, the development of the whole system mighthave been vastly different (for a more detailed account, seeHilborn, 2004).

3 It should be stressed that in Arthur’s well-known polyaurn simulation, the self-reinforcing mechanisms are alreadyset up right from the beginning (Arthur, 1989); the experi-menter determines the rules. From our point of view, how-ever, this is owing to the necessities of a simulation modelrather than a theoretical statement.


Fn2

Fn3

option is suggested by a utility calculus. Whilein no way we exclude this case, it seems toorestrictive a starting point for the organizationalcontext (see also critical comments by Crouch &Farrell, 2004; Eden, 2004; Ortmann, 1995). Focus-ing on utility-driven behavior only implies a dis-regard for important insights of organizationalstudies. Self-reinforcing patterns in organiza-tions have been shown to result from other fac-tors as well, such as emotional reactions (uncer-tainty avoidance, intergroup revenge, etc.),cognitive biases (selective perception, blindspots, implicit theories, etc.), and even politicalprocesses (gaining and maintaining power, re-ciprocal negotiation). These aspects have to beincluded in a theory of organizational paths toadapt to the scope of organizational behaviorpossibly activating self-reinforcing effects. Moreprecisely, we suggest including different formsof positive feedback cycles based on specifi-cally organizational forces.A related problem of the technological path

dependence studies results from their focus onindividual decision making. This exclusive fo-cus on individuals does not account for the in-stitutional setting in which organizational posi-tive feedback processes happen to occur. It isthe broader organizational context (e.g., hiddenassumptions of the organization, organizationalculture, status and role system, and institution-alized practices) that informs decision makersand provides the basis, indirectly and inadver-tently, for the development of self-reinforcingloops.On a general level, the concept of increasing

returns highlights positive feedback processes—that is, the increase of a particular variableleads to a further increase of this very variable.More specifically, the notion of increasing re-turns indicates self-reinforcing processes withincreasing benefits; repetitive pursuits to earnthis increasing rent are likely to culminate in apatterned dynamic. Eventually, a dominant so-lution emerges in terms of recursive action pat-terns (Giddens, 1984). The flip side of these re-turns is that the whole process becomes moreand more irreversible, particularly in cases ofhigh investments and/or high fixed costs (Ghe-mawat, 1991). Decision processes in Phase II,however, are still contingent or “nonergodic”(David, 1985)—that is, while essentially con-strained, choices are still possible (the shadowof Phase II in Figure 1 is designed to indicate the

still prevalent situation of contingency). A sub-sequent section elaborates on major self-reinforcing processes in organizations.

Lock-in Phase

The transition from Phase II to Phase III ischaracterized by a further restriction of thescope. The focal action pattern is replicatedeven more, which eventually leads the wholesetting into a lock-in. Because of the circum-stances, this lock-in may be of a predominantlycognitive, normative, or resource-based nature(Giddens, 1984). Although organizational studiesmostly emphasize the role of managerial cogni-tions or beliefs or resources, organizational lock-ins are also likely to be combinations of all threedimensions.In its extreme form, the dominant pattern

gains a deterministic character, and alternativecourses of action are no longer feasible for var-ious reasons: high switching costs, sunk costs,monopoly, and so forth. By implication, furtherdecisions (owing to lack of alternatives, they areactually no longer decisions) are bound to rep-licate the path. Even newcomers are forced toadopt it. Agents continue to reproduce this andonly this particular outcome. This extreme formof lock-in has been found with technologicalsolutions (e.g., the QWERTY keyboard).Considering organizational paths, however,

the context seems to be significantly different,requiring a somewhat modified conception oflock-in. Organizational settings cannot readilybe equated with markets and monopoly. Be-cause of their social character, organizationalprocesses are more complex and ambiguous innature. They are not likely to amount to a stateof full determinacy, which excludes any alterna-tive choices. Rather, self-reinforcing dynamicsare expected to bring about a preferred actionpattern, which then gets deeply embedded inorganizational practice and replicated. Hierar-chy provides formal authority and legitimate in-fluence on members’ behavior; orders can poten-tially stop inefficient replication. On the otherhand, fixed and inflexible behavior is a widelyrecognized feature in the organizational changeliterature. It is well known that, from time totime, despite hierarchical control, it is extremelyhard to change organizational action patterns(e.g., Beer & Nohria, 2000; Kaufman, 1995); theyare quasi locked in. We should nevertheless re-


frain from reifying organizational paths and at-tributing an objective quality to social rigidify-ing processes. In organizational settings,therefore, we suggest conceptualizing the finalstage of a path-dependent process in a less re-strictive way—as a predominant social influ-ence, leaving some scope for variation (Pierson,2000; Thelen, 1999; for illustrative examples) seeBruggeman, 2002; Burgelman, in press; Eden,2004; Hollingsworth, 2006.In more detail, it seems promising to conceive

of the lock-in stage in terms of an underlyingcore pattern (invisible “deep structure”), withsome variation in practicing it (visible activitylevel). Actors in the final phase do not simplyexperience the path; rather, as “knowledgeableagents” (Giddens, 1984), they have scope in in-terpreting the organizational patterns. This in-dividual interpretation of the core (path) is likelyto bring about some variation in actual organi-zational action patterns. While the underlyingpath structure is fixed, its replicative practice issubject to some variation. In a way, this argu-ment echoes the conception of routines ad-vanced by Feldman and Pentland (2003), stress-ing, on the one hand, the ostensive side as afixed, overarching pattern and, on the otherhand, the performative side as the actual prac-ticing of a routine involving some variation. Inconclusion, for organizational settings it seemsmore adequate to conceive of the lock in statenot in terms of total rigidity but, rather, as amatter of degree, accounting for variance in theactual practicing of the organizational path. Acorridor may best serve to illustrate this reason-ing; the shadow in Phase III in Figure 1 is de-signed to indicate this adaptation. Althoughhighlighting these differentiations, the lock-inphase is nevertheless constitutive for path de-pendence. If actors were not locked in, onewould not call the process path dependent.Whatever the best conceptualization of organ-

izational lock-ins, the more controversial featureof this stage is efficiency. In David’s initialframework (1985) inefficiency was considered anecessary element, because he set out to ex-plain a puzzle: how could an inferior solutionlike the QWERTY keyboard endure in a marketeconomy? So he started with inefficiency rightfrom the beginning. Subsequent work called thiselement into question and suggested a modifiedperspective (Arthur, 1994; Pierson, 2000)—the ar-gument being that path analyses address non-

ergodic inflexible processes (as opposed to lin-ear ergodic processes). A separate question iswhether the state finally reached is efficient orinefficient, and the answer to this question is notconsidered part of the theory of path depen-dence. Although we appreciate this argument,from our point of view, it misrepresents the veryintention of path analyses. The primary interestis not in the formal logic of nonlinear noner-godic processes as such; rather, it is nourishedfrom congealing processes and puzzling persis-tencies that are likely to hamper present andfuture scopes of action. In other words, it is atleast potential inefficiency that is worrying andmakes path dependence a matter of high impor-tance. We therefore advocate including ineffi-ciency in an organizational theory of path de-pendence.It is true that a narrowing organizational pro-

cess and lock-in do not automatically mean im-mediate inefficiency or losses. Path dependenceand efficiency, however, do not refer to a certainpoint in time; instead, a longer time horizon iscovered, necessarily including the alerting riskof becoming dysfunctional. From a strategic, fu-ture-oriented point of view, rigidity therefore al-ways means potential inefficiency. If an organi-zation or a significant practice (e.g., combiningspecific R&D capabilities with marketing skills)has become locked-in, there is inherently thedanger of becoming inefficient, either in the faceof new, more efficient alternatives or changedinternal or external circumstances calling fornew solutions. Latent inefficiency becomesmanifest when an organization confronted withthese change requirements cannot adopt newmeasures because it is confined to the existingpath of action, which binds it to the historicalsolutions (Leonard-Barton, 1995; Schreyogg &Kliesch, 2007). This dysfunctional flip or ration-ality shift from initial reinforcing earnings tostrong barriers to change and losses shouldtherefore be considered a constitutive elementof organizational path analyses.In any case, calling a lock-in “inefficient” al-

ways implies a base of reference—a comparisonwith another standard. The base of referencecan differ; it is not a fact but, rather, depends onthe perspective taken (focusing on a group, adepartment, the whole organization, the field).By implication, discussing the inefficiency of anorganizational path always requires the expo-sure of the base of reference applied.


To sum up, the proposed theory conceptual-izes an organizational path as a tapering socialprocess. Starting (Phase I) with contingency, acritical event (decision, accident, etc.) favors asolution leading unpredictably to a critical junc-ture. If it triggers a regime of positive, self-reinforcing feedback, this solution progressivelygains dominance (Phase II). This pattern islikely to become persistently reproduced and tocrowd out alternative solutions to an extent thatit gets locked in (Phase III) and is accompaniedby immediate or future inefficiency. In short,organizational path dependence can be definedas a rigidified, potentially inefficient action pat-tern built up by the unintended consequences offormer decisions and positive feedback pro-cesses.

COMPARING RELATED CONCEPTIONS

The suggested framework needs discussionand refinement to further clarify the causal logicof path-building processes. In particular, theconcrete forms of self-reinforcing organizationaldynamics need elaboration. In a first step, how-ever, it seems advisable to sharpen the model’sdistinguishing features by contrasting it withrelated conceptions that also highlight the im-portance of initial conditions and events for or-ganizational development, such as imprintingor escalating commitment.

Imprinting

The concept or imprinting (Beckman & Burton,2008; Boeker, 1989; Johnson, 2007; Stinchcombe,1965) figures prominently among approachesthat seem to address a process very similar toorganizational path dependence. Basically, thisconcept postulates that either initial cognitiveschemes, competences, and so forth—of afounding entrepreneur or team, for instance—orspecific contextual conditions (organizationalstructure, postwar depression, internet boom,etc.) at the time of founding imprint organiza-tional processes at later stages and, eventually,amount to a replicated pattern. Although thereare doubtless striking similarities that leadmany authors to either simply equate imprint-ing and path dependence or to conceive theformer as a specific variant of the latter (e.g.,Beckman & Burton, 2008), the process of becom-

ing path dependent is governed by a differentlogic.First, the replicated pattern in the imprinting

approach is ready-made at the beginning; it is aspecific scheme that persists and continues toinfluence future processes. In contrast, the ge-stalt of an organizational path is not clear at allin the early stage; it is an unforeseeable productof later processes, which are initially unknown.Path dependence is an offspring of the nature ofthe process. Second, because of this, a theory oforganizational paths—as opposed to the im-printing approach—has to explain the unfold-ing process of path formation, not only the re-production of structural properties because ofeither efficiency or a lack of competition (Stinch-combe, 1965), or the presence of institutionaliza-tion processes (Johnson, 2007). Nevertheless, im-prints doubtless play an important role in manyorganizational processes. In path-dependent or-ganizational processes they can, for instance,explain the restrictions in the PreformationPhase.

Escalating Commitment

Another concept that shares striking similari-ties with organizational path dependence is es-calating commitment (Ross & Staw, 1993; Staw,1976). As happens in cases of path dependence,particularly in the inefficient Lock-in Phase, es-calating commitment prevents organizationaldecision makers from changing their course ofaction, despite continued negative feedback onthe outcome. Instead of stopping, the agents rep-licate the inefficient solution—in particular, thetendency to throw good money after bad (seeGuler, 2007)—for various reasons.There is, however, a major difference between

escalating commitment and path dependenceexplanations. The latter consider a process witha more or less accidental beginning and alonger phase of success; it is only in the finalstage that the persistent course of action shiftsinto inefficiency. In contrast, escalating commit-ment captures situations where the course ofaction fails from the very beginning. Since thereare no increasing returns or similar enhancingeffects, it highlights another problem area—namely, pathological decision behavior basedon the dynamics of self-justification and fears oflosing face. In effect, escalating commitment isproblematic from the very beginning.


Commitment/Sunk Cost

There is another related stream of thoughthighlighting persistence through resource com-mitments and subsequent exit barriers (Ghema-wat, 1991). Early investment is assumed to re-strict the future scope of action. This argumentcomes close to sunk cost. Sunk cost, however,has its own ambiguity. As we can learn frommicroeconomics, sunk costs are only a psycho-logical outcome of imagination; from a rationalchoice point of view (e.g., Pindyck & Rubinfeld,2005), they are irrelevant. Organizations couldignore them because they are, in fact, not rele-vant for their future decisions. If there are newand better projects, the capital market will pro-vide fresh money to overcome the old structure.Apart from this idealized counterargument (it

takes an efficient capital market for granted),the simple fact of past investments (in terms ofsunk costs) cannot be equated with path depen-dence, since, in consequence, all investmentswould bring about path dependence (for a moresophisticated argument in this direction, see Ar-row, 2004). This refers back to our initial state-ment that we should refrain from conceptualiza-tions that end up considering all pastdependence as path dependence.In a similar vein, Pierson (2000) raises the is-

sue of whether increasing returns or other self-reinforcing processes should be made a neces-sary element of path theory. Among others, hehighlights pure complementarities as beinglikely to bring about persistency. From our pointof view, this argument refers to a different typeof rigidity. Pure complementarities without self-reinforcing processes characterize a stable situ-ation of fitting resources, but not a process thateventually leads to a lock-in. If there is no esca-lating self-reinforcing process, switching to newand better opportunities may be difficult but notincreasingly impossible.

Structural Inertia

This is another well-known argument on or-ganizational persistency that has been ad-vanced by population ecology (Gresov, Have-man, & Olivia, 1993; Hannan & Freeman, 1984;Hannan, Plos, & Carroll, 2004; Ruef, 1997). In thisperspective structural inertia—the hyperstabil-ity of organizational arrangements in spite ofenvironmental change—is a universal organiza-

tional feature that develops in the course ofstructuring the organization. Routinizing and in-stitutionalizing organizational activities areseen as imperative in order to guarantee stake-holders reliability, accountability, and, finally,survival in competitive environments. Inertia isconsidered a precondition for effective organi-zational acting but, paradoxically enough, even-tually threatens the organization’s survival, be-cause it is likely to bring about a mismatch withchanging environmental conditions.Again, the phenomenon is somewhat similar

to organizational path dependence, but the fo-cused process and its explanations are clearlyat variance. Inertia occurs via the intended es-tablishment of reliable organizational struc-tures; there are no structural dynamics. It is auniversal requirement that all organizationshave to fulfill. And all organizations, especiallywhen growing and aging, are also expected tobecome hyperstable, with difficulties in meetingnew environmental challenges. Opposed tothat, the suggested framework of path depen-dence does not apply to all organizations (ithighlights special cases only) and requires anavalanchelike process to bring about a lock-in.The focus is on explaining the process and itsvarious stages.4

Reactive Sequences

While subscribing to the suggested type ofpath-building process, Mahoney (2000) developsa second type—namely, efficient or inefficienttrajectories built up by reactive sequences. Thisprocess is characterized by a chain of modularevents governed by singular cause-and-effectrelationships. A focal event, B, is assumed to bethe effect of a prior event, A, and at the sametime the cause of a future event, C, etc., accu-

4 Carroll and Harrison (1994) point to the importance ofpositive feedback in the ecological model, but only withrespect to density dependence; a more general consider-ation of the importance of self-reinforcing processes doesnot seem to be intended. Other evolutionary and, more re-cently, coevolutionary theoriests make explicit use of thenotion of path dependence and tend to prefer it to otherconcepts (e.g., Helfat, 1994; Nelson &Winter, 1982; Volberda &Lewin, 2003). Given its relatedness to economic evolutionarytheory, one seedbed of path, dependence research (e.g., Dosi,1982; Witt, 1997), this certainly comes as no surprise. Never-theless, the concept of path dependence has not yet beenfully utilized in this stream of research either.


Fn4

mulating in a reaction chain: A � B � C � D �E � F. Thus, initial event A is expected to affectB, but it unintentionally triggers a multistagedevelopment. The final state or, better, an inter-mediate result can—very much like pursuing alaw suit—be traced back to the releasing event.This intermediate state is also likely to shapefuture action; it is, however, not in any waylocked in or inefficient.In contrast to the path dependence model ad-

vanced above, the intermediate state of a se-quence of causal reactions is not reached byincreasingly reproducing a specific pattern, andthere is no connecting logic that explains thesuccession of the singular sequences. Althoughthe idea of reactive sequences doubtless pro-vides insights into the evolvement of historicalprocesses, it does not fit into a theory of pathdependence. Without path drivers and thecausal logic of a lock-in, a theory of organiza-tional paths loses its very point.Furthermore, the sequence argument raises

some conceptual questions. First of all, se-quences seem simply to occur. In contrast toprocesses explained by the regime of self-reinforcing mechanisms, the concept of causalreactive chains does not provide a logic thatexplains why the sequences take place this wayand not in another way. Why do reactive se-quences accumulate? Superimposing, ex post, atrajectory on reactive sequences does not pro-vide an explanation. Another problem is gener-alization. A theory of path dependence aims atexplaining a particular class of processes (Pet-tigrew, Woodman, & Cameron, 2001). Drawingon reactive sequences does not, however, tran-scend a singular case: singular reasons are sup-posed to explain singular events only.

Institutionalizing

Contextual shaping forces play a major role inneoinstitutional theory, and its concept of insti-tutionalization also seems to come close to pathdependence (e.g., Powell & DiMaggio, 1991;Scott, 2001; Tolbert & Zucker, 1996). Most impor-tant, this theory highlights the relevance of thesymbolic-normative environment of organiza-tions and how this influences the formal andinformal structuring of organizations over time.Apart from the pace of the development (Law-rence et al., 2001), it addresses institutional in-ertia and stability by revealing how a specific

organizational structure or form becomessedimented and taken for granted over time,preferably across sets of organizations.While neoinstitutional theory elucidates im-

printing and stabilizing processes and, in partic-ular, sensitizes us to the relevance of symbolic-normative contexts (e.g., Hargadon & Douglas,2001), in its present form it does not address thesystemic logic of an escalating reinforcement ofan action pattern or a path (see, however, Eden,2004, and Holm, 1995). The theoretical focus,therefore, differs significantly and explainsother constellations.

AT THE HEART OF ORGANIZATIONAL PATHDEPENDENCE: SELF-REINFORCING

MECHANISMS

So far, we have conceptualized path-buildingprocesses as processes of a diminishing scopeof action that unintentionally develop their ownpull and are driven by positive feedback. It is atime-based theoretical concept differentiatingbetween different states of flexibility/choice andstability/determinism, respectively. The dy-namic eventually flips over into rigidity. At theirheart, such processes can be explained by oneor a combination of several self-reinforcing so-cial mechanisms. In this section we elaborate onthese mechanisms in an organizational con-text.5

In the field of technology development anddiffusion, different types of self-reinforcing dy-namics have been identified (Arthur, 1994;Cowan, 1990; David, 1985; Katz & Shapiro, 1985;North, 1990). Since they have been developed atthe market level, these mechanisms cannotreadily be transferred to organizational analy-sis. In our view, four mechanisms in particularare likely to contribute to the development oforganizational path dependence: coordinationeffects, complementarity effects, learning ef-fects, and adaptive expectation effects. Belowwe aim to combine different streams of thoughtto build a framework of self-reinforcing dynam-ics at the level of single organizations and or-ganizational subunits. We discuss these fourmechanisms and show how they apply to organ-izational settings.

5 For the more general debate on social mechanisms inorganization theory, see, for instance, Pajunen (2008).


Fn5

Coordination Effects

Initially put forward in institutional econom-ics (North, 1990), these effects relate to the heartof organizational functioning. They build on thebenefits of rule-guided behavior: the more ac-tors adopt and apply a specific institution (i.e.,an organizational rule or routine), the more effi-cient the interaction among these actors is,since the behavior of the actors is rule guidedand can therefore be anticipated and reactionscan be considered in advance. Coordinationcosts can be significantly reduced. In conse-quence, it becomes more attractive to adoptthese rules when other individuals also followthem.The best-known illustrative example of this

effect at the institutional level is the decisionregarding right-hand traffic versus left-handtraffic; the institution became fixed early on be-cause of the obvious benefits of following it—uncertainties involved in human interactioncould successfully be reduced (North, 1990: 23).Another well-known example that applies di-rectly to the level of single organizations isworking time regimes, which guarantee effi-cient cooperation. There is a striking similarityto the economies of scale effect (North, 1990):increasing the number of participants results indecreasing (coordination) cost per unit.Miller and Friesen (1984) developed the con-

cept of internal consistency, which comes veryclose to that of coordination effects. This propo-sition stresses the advantages of an internal fitamong the various elements of an organization(see also Miller, 1992). Coordination effects thusresult from the benefits of following the samesingle rule or set of related rules to which othersare willing to conform. As a result, through theadvantages of continued replication, a specificpattern of practices is likely to become fixed.The fixing power of such arrangements hasbeen proved in cases where organizationalmembers have recognized new challenges andset out to change their practices but failed to doso because they could not get rid of their well-attuned activity sets and routines.A striking example of such path dependence

was provided by Tripsas and Gavetti (2000), whoportrayed the difficulties Polaroid experiencedin changing their R&D priorities into a newproduct development competence. Similarly,Gilbert (2005) described newspaper companies

who stuck to self-reinforcing rules for producinga newspaper and thereby became unable to ex-ploit new online opportunities. More recently,Koch (2008) provided evidence of similar pat-terns in German quality newspapers. Adoptingjoint rules of quality journalism brought aboutsignificant coordination advantages and thelasting constitution of a once successful busi-ness model. Nowadays, the flip side of this pathis broadly discussed.

Complementary Effects

A well-known explanation for complementari-ties are economies of scope, which exist whenthe cost of producing and selling two or moregoods or services together is lower than the costof producing and selling them separately (Pan-zar & Willig, 1981). On a more general level,complementarities mean synergy resulting fromthe interaction of two or more separate but in-terrelated resources, rules, or practices (Pierson,2000; Stieglitz & Heine, 2007). In the case ofcomplementarities, the advantages of repeat-edly combining interrelated activities do notsimply add up; they produce an additional sur-plus: K(x � y) � K(x) � K(y). Take, for instance,marketing skills and R&D capabilities, whichmay add up to a “core competence” (Prahalad &Hamel, 1990) of a company or a division. David(1994: 214) calls such combinations “institutionalclusters.” In complementary settings, self-reinforcing processes occur when routinesand/or practices are interconnected in such away that it becomes ever more attractive to ex-ploit the synergies or—when referring to the re-verse side—to save misfit costs caused by solu-tions deviating from the established cluster/organizational capability. As a result, distinctsets of activity patterns become progressivelydominant (Leonard-Barton, 1995) and, addition-ally, deeply embedded in an organization(“deep structure”)—that is, they become organi-zationally path dependent.There are many other examples that can fur-

ther illustrate this effect. Take, for instance,“Fordism,” which is characterized by comple-mentary management systems in human re-sources (hiring and firing of low skilled labor),operations (mass production), and organization(hierarchy of control), and which for quite sometime constituted, through repeated practicing, aspecific capability amounting to a competitive


organizational advantage. Ultimately, the insti-tutional cluster became path dependent (Piore &Sabel, 1984).

Learning Effects

The learning effect theory holds that the moreoften an operation is performed, the more effi-ciency will be gained with subsequent itera-tions. The operation becomes more skillfullyperformed (faster, more reliable, and with lesserrors), which, in turn, means decreasing aver-age costs per unit of output (Argote, 1999). Andthe more attractive the chosen solution becomesbecause of accumulated skills and decreasingcost, the less attractive it is to switch to newlearning sites (where the actors would have tostart from scratch). Only sticking with the oncechosen solution promises continued returns—although, as is well known, the resulting costcurve flattens after a while.Self-reinforcing learning effects can be found

at various organizational levels. A well-knownexample from organizational learning points tothe fact that a focus on the advantages of ex-ploitative learning may increasingly drive outexplorative learning (March, 1991, 2006). For var-ious reasons (e.g., the prevailing organizationalculture and reward system), the motivation toimprove everyday practices is likely to gainmore acceptance or legitimacy (and, thus, morerewards) from the organization, whereas the mo-tivation to look for fresh alternatives and to crit-ically examine well-established organizationalpractices is likely to shrink progressively. Thismyopia or preference for repetitive exploitativelearning builds on the self-reinforcing dynamicsof learning effects, eventually ending up in anorganizational path along the familiar prac-tices.A related effect has been highlighted by the

“architecture of simplicity” (Miller, 1993), inwhich an organization develops a successful setof strengths and tends to focus all learning abil-ities on refining this success; it exploits thisstrength through gaining learning effects whileneglecting other opportunities. The exploitationis easier (more efficient) the simpler the institu-tional cluster; therefore, the self-reinforcing dy-namics bring about unintended increasing sim-plicity. Ultimately, “it turns into a monolithic,narrowly focused version of its former self, con-verting a formula for success into a path toward

failure” (Miller, 1993: 116). Learning effects areoften reinforced and extended by earnings fromcoordination costs and complementarities.

Adaptive Expectation Effects

These self-reinforcing effects relate to the in-teractive building of preferences. With this con-cept, as opposed to neoclassical economics, in-dividual preferences are not considered to befixed; instead, they are assumed to vary in re-sponse to the expectations of others. Oftenquoted examples highlight the need for socialbelonging and the desire to end up on the win-ning side. The more people are expected to pre-fer a particular product or service (and not an-other), the more attractive that product or servicebecomes (Leibenstein, 1950). Since users are of-ten uncertain about the right choice, they feelrewarded by the fact that others are likely toprefer the same. Because of this self-reinforcingdynamic, a dominant solution is likely toemerge, more often than not by way of a self-fulfilling prophecy (in most cases on the basis ofmore or less random first choices and from hear-say).In the context of organizations, the informal

diffusion of best practices often follows thislogic (Szulanski, 1996). Organizational membersare willing to adopt these practices becausethey expect others to do the same and wish toend up on the side of the winners. This tendencyis reinforced by other drivers, such as legitimacyseeking or signaling; individuals or subsystemsnot subscribing to the best practices are afraidof losing legitimacy and—if associated withfailure—of becoming stigmatized as “outsiders”(Kulik, Bainbridge, & Cregan, 2008).Early on, McGregor’s (1960) Theory X nicely

illustrated the dynamics of such self-reinforcingadaptive expectations and subsequent self-fulfilling prophecies in organizations. The start-ing point of his Theory X spiral is managers’implicit assumptions, about the nature of theiremployees—as being interested only in mone-tary rewards, hating to take on responsibility,and shirking wherever they can. This implicittheory of human behavior not only defines theset of expected managerial behaviors but alsoessentially frames management’s decisions onreward systems and organizational design (inparticular: a strong emphasis on control andauthority), which, in turn, evoke corresponding


reaction patterns (especially passivity, indo-lence, and apathy). Observing those reactions islikely to confirm and reconfirm exactly thoseassumptions about behavior managers havemade, based on their implicit Theory X. Theseconfirmed expectations then reinforce the em-phasis on restrictive organizational structuresand controls, thereby unconsciously advancinga vicious circle (see also Leonard-Barton, 1995;Masuch, 1985; Repenning & Sterman, 2002). Inthis case a dominant organizational designemerges because of a self-reinforcing spiral thatis based on “expectations of expectations” (Lu-hmann, 1995).

Adding an Enhancing Context?

Some authors add contextual conditions as afurther reinforcing effect of and in institutions.Pierson (2000), for instance, highlights institu-tional density as a salient determinant likely toconverge into self-reinforcing effects in organi-zations. In his view organizations (in particular,formal political institutions) are more prone tobring about path-building forces than markets,because they act in “a far, far murkier environ-ment” (Pierson, 2000: 260) with weaker forces tocorrect inefficient courses of action over time.The complexity of organizational goals and theuncertainty of the causal links between actionsand outcomes render the organizational fieldinherently ambiguous, and organic correctionsof inefficient action are less likely to occur herethan in markets. Therefore, practices, once es-tablished, gain momentum more easily and cre-ate a fertile ground for developing increasingreturns or other types of positive self-reinforcingfeedback. More generally, Pierson considersambiguity and complexity important conditions,which amount to self-reinforcing effects andsubsequent path dependence.Although addressing doubtless significant

contextual conditions for path development, theconditions of ambiguity and complexity shouldnot be misconceived as self-reinforcing mecha-nisms in their own right. This also holds true forother factors addressed in the literature as self-reinforcing dynamics, such as “uncertain expec-tations” or “power structure” (Beyer, 2005). Theseare relevant contextual factors, but they shouldnot be equated with self-reinforcing mecha-nisms. Enhancing contexts—however importantthey may be—neither lead directly to path de-

pendence nor represent a necessary or even suf-ficient condition for the occurrence of path de-pendence (see also Arthur, 1989). A theory oforganizational path dependence has to differen-tiate properly between self-reinforcing mecha-nisms on the one hand and enabling institu-tional contexts on the other. In consequence,Pierson’s insights should encourage further re-search to explore the contextual conditions en-hancing (or hindering) the unfolding of self-reinforcing mechanisms and subsequentconstitution of organizational paths.

IMPLICATIONS FOR MANAGEMENTAND RESEARCH

Path dependence and its far-reaching conse-quences doubtlessly constitute an issue of highrelevance in strategic management and organ-izational decision making. From a managerialpoint of view, the fatal consequences of beinglocked in raise the pressing question of whetherorganizational paths can be dissolved or in anyway escaped. It is true that no path is forever,but this is no relief from the perspective of aparticular organization, since path dependencemay exist for quite some time.Path dissolution may occur through unfore-

seen exogenous forces, such as shocks, catastro-phes, or crises; these are likely to shake thesystem, thereby causing the organization tobreak away from the path (Arthur, 1994: 118).However, path dissolution may also occur be-cause of an insidious change in organizationaldemography or the “incomplete” socialization ofnew organizational members (Tolbert, 1988). Inthis vein, Castaldi and Dosi (2006) refer to thepossibility of coincidental delocking in terms ofa by-product of other organizational decisions. Anice illustration of such coincidental path disso-lution at an organizational level is provided bythe Intel case and the highlighted moves in thememory business (Burgelman, 1994, 2002;Burgelman & Grove, 1996). With this perspective,however, path dissolution amounts to an acci-dental process, be it revolutionary or evolution-ary, which—nobody knows—may or may not oc-cur. Adopting this view clearly has a fatalistic orat least a passive flavor to it. We are condemnedto wait, probably for a long time since we knowsuch events are rare. The firm may go bankruptlong before a disrupting event occurs. An activeand alert attitude thus seems imperative.


AQ:1

Facing strategic rigidity or even a threat to theorganization’s survival, decision makers wouldcertainly be eager to learn more about possibleinterventions designed to escape or unlock or-ganizational paths. But can organizations actu-ally break path dependence? This is an intricateendeavor, since the idea of deliberately break-ing a path is self-contradicting in a way. If wedefine path dependence as a situation in whichindividual actors or organizations have losttheir power to choose among alternatives, thenthe assumption that the same actors can un-lock the path is obviously inconsistent. Path-dependent behavior, strictly speaking, excludespath-breaking behavior. The idea of unlockingorganizational paths, therefore, can only work ifwe put the mechanisms of deterministic patternreproduction into perspective. In other words, itis necessary to construe and integrate an exog-enous perspective—that is, an activity that isnot under the regime of path dependence. Suchintegration of an external lens or—if you like—a“second-order observation” (von Foerster, 1991)enables knowledgeable agents to reflect prac-tices in terms of path dependence and poten-tially opens a window for path-breaking activi-ties.

Deliberately Breaking Organizational Paths

When discussing the actual possibilities ofdissolving organizational paths, we need toclarify what path breaking in an organizationalcontext means precisely. Is it the destruction of arigidified action pattern? Does it mean restoringthe situation as depicted in Phase I? Is it thebroadening of the “corridor” in Phase III? Is itthe realized switch to a superior alternative?From our point of view, each of these alterna-tives is not exclusive. Rather, path breaking canvary in intensity and complexity. Without ex-cluding more complex cases, we therefore sug-gest defining a minimum condition for a situa-tion to be categorized as “path breaking.” Sincethe process of becoming path dependent hasbeen framed as progressively eliminating thescope of decision making, this minimum condi-tion is the effective restoration of a choice situ-ation—the insertion of at least one alternativecourse of action. However, opening the windowfor an alternative is necessary but not sufficient.The new alternative has to be a superior one

(Arthur, 1994), because implanting an inferiorone would not constitute a real choice.The suggested theory of path dependence can

inform the endeavor to intentionally unlock or-ganizational paths. As already pointed out, themajor drivers rendering a process path depen-dent are self-reinforcing dynamics. By implica-tion, the possibility of escaping from or breakinga path depends very much on interrupting thelogic and the specific energy of the self-reinforcing patterns of the process in question.The first step in any path-breaking intervention,thus, requires understanding and reflecting onnot only the fact of being path dependent butalso the drivers that made this happen. At thesame time, this step brings the necessary dis-tancing from the replicating dynamics. Reflect-ing on the practice that is usually taken forgranted indicates taking a critical stance (Moon,1999; Schon, 1983) by changing the mode of ac-tivity: from doing (operational mode) to observ-ing and reflecting (observational mode), therebygaining access to the closed dynamics.Understanding self-reinforcing dynamics re-

quires theoretical knowledge as well as practi-cal skills and abilities. This is all the more truesince the organizational dynamics of path for-mation are more often than not hidden dynam-ics. Subconscious blinders, perceptual defense,and blind spots tend to inhibit reflection pro-cesses (Saffold, 1988; Sorensen, 2002). As a con-sequence, in many cases a special effort isneeded to get in touch with the hidden agenda.Tools have been developed to facilitate suchprocesses. Assumption surfacing, for instance,is a well-known technique designed to makehidden patterns in organizational settings ac-cessible, to open them up for critical reflection,and to put them on the organizational discourseagenda (Kettinger, Teng, & Guha, 1997; Mason &Mitroff, 1981).Beyond discourse, however, reflecting path-

bounded practices often requires addressing theemotional side of inertial organizational pat-terns as well. Organizations frequently resistreflecting on hidden features. For instance, intheir analysis of the Intel case, Burgelman andGrove (1996: 15) found that “emotional attach-ment on the part of the top management to thebusiness” was intertwined with inertial self-perceptions. Dissolving and/or altering suchemotional barriers often requires special ap-proaches—in particular, clinical approaches,


such as “reframing” (Bandler & Grinder, 1982), oreven psychoanalytical techniques (Kets deVries, 2006). Such interventions may help to un-lock emotional blinders and to reflect the domi-nant framing manifested in replicative prac-tices.As is well-known, such individual-centered

approaches are limited. In some cases the wholesystem is controlled by a regime of unwrittenrules that suggest refraining from self-reflection(Scott-Morgan, 1994). Communication researchhas revealed that such closing behavior is oftencaused by paradoxical interactions such as adouble bind (Hennestad, 1990; Watzlawick, 1963),basically maintained and fixed by self-reinforc-ing behavioral patterns. So-called systemic ap-proaches (Campbell, Coldicott, & Kinsella, 1994;Selvini-Palazzoli, 1986) react to exactly this well-known denial. In these approaches a totally newtype of intervention has been developed—theparadoxical intervention. This is a disguised in-tervention designed to irritate the closed self-reinforcing system from outside. The changeagent no longer tries to persuade the system toreflect and modify its hidden rules; rather, he orshe “asks” the system to produce more of thebehavior that is suggested should be changed.The underlying idea is that a prescribed resis-tance shifts the logic; it ceases to be resistanceand becomes compliance (Watzlawick, Weak-land, & Fisch, 1974). On the other hand, resistingthis prescription means changing. In self-reinforcing systems these paradoxical interven-tions bring about a kind of pattern implosion,likely to shake the system of long-standing “de-fensive routines” (Argyris, 1990) into a new con-figuration and to open the window of opportu-nity for path-breaking changes. A paradoxicalintervention comes as a surprise to the system; itattempts to stop the self-reinforcing dynamicsby advancing an implosion of the routinized ri-gidity.Such interrupting approaches potentially re-

open the scope of action. Opening the scope,however, does not automatically imply unlock-ing a path. Whether or not an existing organiza-tional path can actually be broken in the sensedefined above basically depends on the revers-ibility of the process. The path-breaking en-deavor cannot ignore the history that broughtabout path dependence. The chance of actuallyrestoring choice depends on the character of theself-reinforcing dynamics and the possibility of

creating a new advantageous situation. Majorfeatures here are resource commitment, revers-ibility, and transferability of experience (Arthur,1994; Ghemawhat, 1991; Gilbert, 2005; Karim &Mitchell, 2000). In cases where learning effectsare the major drivers of path dependence, thereversibility is at least questionable. Learningeffects are acquired in a specific field of practiceand cannot easily be transferred to a new con-text (of a new alternative). Setting up a newcourse of action takes quite some time to matchthe learning effects of the existing course (evenwhen it has been proved inferior). In those casescreating an effective alternative for the restora-tion of choice requires an extra effort—a (possi-bly costly) subsidy to help the new alternativecatch up with the existing one (Arthur, 1994).The situation is different for other self-

reinforcing mechanisms—for instance, for coor-dination effects. The logic of this driver verymuch depends on the willingness to conform torules. What is required here, therefore, is thewillingness to switch to a new regime of rulesmandated by a change agent or a project group.As Arthur (1994) points out, this is more likely tohappen the higher the certainty that the othersalso prefer the new alternative regime of rules.In cases where this certainty is missing (or ruletakers are at least uncertain about the actionpreferences of others), the situation is difficult tochange, because actors wait for more certainty.In all of these cases, additional initiatives arerequired to reduce uncertainty about others’ ac-tion.One may argue that unlocking or breaking an

organizational path should be easier than es-caping a technological or market lock-in. Orga-nizations by their very nature are characterizedby a central authority and hierarchical control,which allows (top) management to create viableorganizational alternatives in terms of incentivesystems, formal rules, or planning procedures.Organizations rely on command and control,even most new organizational forms. Neverthe-less, because of informal processes, the lack oftransparency, and unintended consequences ofactions in organizations, it is unrealistic to as-sume that the development of an organization iscompletely under management control. The ten-dency toward diminishing control, actual con-trol loss, and the illusion of control have beenlong-standing issues in organizational research(see, for instance, Downs, 1967, and Streatfield,


2001). In consequence, the scope for an organi-zation to unlock path dependence simply by or-der is clearly limited. More often than not, suchmerely formal approaches will fail because—asdepicted above—it is so difficult to stop self-reinforcing processes in organizations. Specialtechniques are required.

Researching Organizational Paths

The suggested three-stage model of organiza-tional path dependence and its focus on self-reinforcing mechanisms not only provide aplatform for considering path-breaking inter-ventions but also suggest guidelines for furtherresearch on organizational paths. According tothis framework, any attempt to prove the claimof path dependence needs to cover the followingthree features, in a way reversing the processdepicted in Figure 1.The first step of a thorough path analysis is

the identification of strategic persistence or opera-tional rigidity of or within a particular organiza-tion. Discovering such structural or institutionalinertia allows the researcher to assert path de-pendence. Because the logic of hyperstability isoften hidden, inert practices are easier to un-cover or make visible in situations of radicalchange (e.g., Tushman, Newman, & Romanelli,1986). However, organizational paths can alsosurface in incremental or even creeping changeprocesses. In any case, the contrafactual main-tenance of a specific pattern is an indicator of apath-induced, potentially inefficient organiza-tional lock-in. With regard to the efficiency as-pect, at the very least, the existence of a superioralternative that the organization is unable topursue has to be demonstrated. A major chal-lenge is presented by the case of potential orstrategic inefficiency—that is, proving that thereplicate behavioral pattern is likely to damagethe organization in the future. In these cases,where no empiri cal proofs can be provided, itbecomes a matter of good reasons to build aconvincing case.The second major element of an advanced

path analysis is the identification, exploration,and reconstruction of the self-reinforcing feed-back mechanisms possibly underlying the organ-izational rigidity in question. The identificationor detection of one or more such mechanisms atwork is a complex task that requires recognizingpatterns in the broad flow of everyday practices.

If there are no self-reinforcing mechanisms “atwork,” the presumption of an existing organiza-tional path cannot be substantiated and has tobe rejected.The third essential part of a systematic path

analysis is the search for a triggering event thatwas likely to have set the path-building processin motion. Tracer studies (e.g. Lee, 1999) mayhelp to identify these initial events mostly un-known to the actors. Also, special attention hasto be devoted to identifying the critical junctureat which a self-reinforcing process leading toorganizational path dependence was activated.It is here where the first working of one or moreself-reinforcing mechanisms can be uncovered.These three elements offer a rough guideline

for both explanatory and exploratory researchon organizational path dependence. Because ofthe process character of the framework, a longi-tudinal research design is required, whichtraces sequences of events and actions in orga-nizations (e.g., time chronologies, simple andcomplex time series). Only the examination ofthese detailed processes in time (and space) willallow us to identify and explicate the working ofthe fundamental social mechanism(s) underly-ing the constitution of organizational path de-pendence “from micro behaviors to system dy-namics, and back” (Castaldi & Dosi, 2006: 108).

CONCLUSIONS

A detailed conception of organizational pathdependence has much to offer when we are aim-ing to solve the puzzle of how organizations be-come locked in and adhere contraintuitively tohistorical solutions. To this end, we have de-fined organizational path dependence as a pro-cess that (1) is triggered by a critical event lead-ing to a critical juncture; (2) is governed by aregime of positive, self-reinforcing feedbackconstituting a specific pattern of social prac-tices, which gains more andmore predominanceagainst alternatives; and (3) leads, at least po-tentially, into an organizational lock-in, under-stood as a corridor of limited scope of action thatis strategically inefficient. The proposed frame-work not only provides a deeper understandingof the historicity of inertial phenomena beyondthe general principle that “history matters” butalso pushes the explanation beyond such well-known concepts as organizational imprinting,institutional legacy, and structural inertia.


This framework of path dependence also of-fers insights into the possibilities and limita-tions of breaking out of organizational path de-pendence. In particular, path breaking requiresa thorough understanding of the social mecha-nisms driving the path process. Understandingthese mechanisms, in turn, provides a platformfor developing path-breaking interventions.Organization research focusing on path depen-

dencies and path breaking would nicely supple-ment not only the present trend toward processstudies, with their inevitable historical component(e.g., Pettigrew et al., 2001; Van de Ven & Poole,2005), but also studies of organizational practicesand their replicate dynamics (e.g., Feldman &Pentland, 2003; Jarzabkowski, 2008; Lawrence etal., 2001). In both cases the theory of organizationalpath dependence would help to focus on particu-lar types of dynamics and, thus, would comple-ment rather than substitute for other process the-ories of organizations. However, the conditionsthat are conducive to path dependence and possi-ble ways of unlocking paths await further explo-ration.A further challenging theoretical question is

whether the emergence of paths does not simplyoccur but can also be deliberately broughtabout. The idea of (intentional) path creationwas introduced by Garud and Karnøe (2001),who referred to Schumpeter (1942) and his con-ception of a “destroying” entrepreneurship. Cre-ative agency and the power of generating mo-mentum are seen as basic ingredients for pathcreation, although entrepreneurship, as influ-enced by its own history and important institu-tions, may well also be a source of path depen-dence (Staber, 2005).Another important issue addresses the level of

analysis. Our three-stage model aims at explain-ing path-building processes of and in organiza-tions. However, other levels of analysis are also ofgreat relevance—particularly by the individuallevel, the network level, and the field level. Orga-nizational members, with their cognitive sche-mata, learning habits, response patterns, and soforth, do play a role in path-building processes inorganizations. Although a vast body of researchon individual rigidities is available (see Huff &Huff, 2000: 46–59, for a review), individual pathresearch stills awaits elaboration.The same is true of interorganizational rela-

tions. Nowadays, many organizations are embed-ded in more or less complex networks of relation-

ships. Collaborative relationships are likely tobecome path dependent too, thereby affecting thedevelopment of an organization (and vice versa).Some network-related studies of path dependen-cies are already available, highlighting, for in-stance, lock-ins and lock-outs in the course of thedevelopment of a network (Gulati, Nohria, & Za-heer, 2000) or the possibility of arising networkinertia (Kim, Oh, & Swaminathan, 2006). Othersshed light on the impact of network embedded-ness on organizational path dependencies (e.g.,Walker, Kogut, & Shan, 1997). But, as in the case ofmost studies at the organizational level of analy-sis, these tend to consider path dependence in thesense of history matters only, and not in the spec-ified way suggested by the explanatory conceptadvanced here.The third level important for organizational

path analysis is the field or industry, which hasbeen addressed by the majority of path depen-dence research so far (starting with David, 1985).Research already evidences the significance ofthe field for explaining organizational inertia,beyond the influence of technological path de-pendencies. The well-known study of the Scot-tish knitwear industry (Porac, Thomas, Wilson,Paton, & Kaufer, 1995), for example, convincinglydemonstrates that it was primarily the valuesystem of the industry and not the individualorganization that brought about inertia. Anotherexample is Hollingsworth’s (2006) study of lead-ing research organizations that were involved inmajor scientific discoveries. This study not onlypoints to the path-dependent development of theorganizations (and even of single laboratories)investigated but also to that of the institutionalfields in which they were embedded.Future research should account more explicitly

for these different levels and their interplay,which may well amount to “cross-catalytic feed-back” (Paul David, personal communication) be-tween not only these levels but also organization-al and technological path processes. Beyond this,we are convinced that more rigorous research onorganizational paths can enrich organization sci-ence significantly, aswell as our understanding ofpuzzling rigidities in organizational life.

REFERENCES

Antonelli, C. 1999. The economics of path-dependence inindustrial organization. International Journal of Indus-trial Organization, 15: 643–675.


Argote, L. 1999. Organizational learning: Creating, retainingand transferring knowledge. Boston: Kluwer Academic.

Argyris, C. 1990. Overcoming organizational defenses. Bos-ton: Prentice-Hall.

Arrow, K. J. 2004. Path dependence and competitive equilib-rium. In T. W. Guinnane, W. A. Sundstrom, & W. C.Whatley (Eds.), History matters: Essays on economicgrowth, technology, and demographic change: 23–35.Stanford, CA: Stanford University Press.

Arthur, W. B. 1989. Competing technologies, increasing re-turns, and lock-in by historical events. Economic Jour-nal, 99: 116–131.

Arthur, W. B. 1994. Increasing returns and path dependencyin the economy. Ann Arbor: University of MichiganPress.

Bandler, R. J., & Grinder, J. 1982. Reframing: Neurolinguisticprogramming and the transformation of meaning. Moab,UT: Real People Press.

Bassanini, A. P., & Dosi, G. 2001. When and how chance andhuman will can twist the arms of Clio: An essay on pathdependence in a world of irreversibilities. In R. Garud &P. Karnoe (Eds.), Path dependence and creation: 41–68.Mahwah & London: Lawrence Erlbaum Associates.

Beckman, C. M., & Burton, M. D. 2008. Founding the future:Path dependence in the evolution of top managementteams from founding to IPO. Organization Science, 19:3–24.

Beer, M., & Nohria, N. (Eds.). 2000. Breaking the code ofchange. Boston: Harvard Business School Press.

Beyer, J. 2005. Pfadabhangigkeit ist nicht gleich Pfadabhan-gigkeit! Wider den impliziten Konservatismus einesgangigen Konzepts [Path dependence does not equalpath dependence! Against the conservatism implicit inthis concept] Zeitschrift fur Soziologie, 34(1): 5–21.

Boeker, W. 1989. The development and institutionalization ofsubunit power in organizations. Administrative ScienceQuarterly, 34: 388–410.

Bruggeman, D. 2002. NASA: A path dependent organization.Technology in Society, 24: 415–431.

Burgelman, R. A. 1994. Fading memories: A process theory ofstrategic business exit in dynamic environments. Ad-ministrative Science Quarterly, 39: 24–56.

Burgelman, R. A. 2002. Strategy as vector and the inertia ofcoevolutionary lock-in. Administrative Science Quar-terly, 47: 325–357.

Burgelman, R. A. In press. Strategic consequences of co-evolutionary lock-in: Insights from a longitudinal pro-cess study. In G. Schreyogg & J. Sydow (Eds.), The hiddendynamics of path dependence. London: Palgrave-Macmillan.

Burgelman, R. A., & Grove, A. S. 1996. Strategic dissonance.California Management Review, 38(2): 8–25.

Callon, M. 1992. The dynamics of techno-economic networks.In R. Coombs, P. Saviotti, & V. Walsh (Eds.), Technolog-ical change and company strategies: 72–102. London:Academic Press.

Campbell, D., Coldicott, T., & Kinsella, K. 1994. Systemic workwith organisations. London: Karnac Books.

Carroll, G. R., & Harrison, J. R. 1994. On the historical effi-ciency of competition between organizational popula-tions. American Journal of Sociology, 100: 720–749.

Castaldi, C., & Dosi, G. 2006. The grip of history and thescope of novelty: Some results and open questions ofpath dependence in economic processes. In A. Wimmer& R. Kossler (Eds.), Understanding change: Models,methodologies, and metaphors: 99–128. Basingstoke, UK:Palgrave Macmillan.

Child, J. 1997. Strategic choice in the analysis of action,structure, organizations and environment: Retrospectand prospect. Organization Studies, 18: 43–76.

Collier, R. B., & Collier, D. 1991. Shaping the political arena:Critical junctures, the labor movement, and regime dy-namics in Latin America. Princeton, NJ: Princeton Uni-versity Press.

Collinson, S., & Wilson, D. C. 2006. Inertia in Japanese orga-nizations: Knowledge management routines and failureto innovate. Organization Studies, 27: 1359–1387.

Cowan, R. 1990. Nuclear power reactors: A study in techno-logical lock-in. Economic Journal, 106: 541–567.

Crouch, C., & Farrell, H. 2004. Breaking the path of institu-tional development? Alternatives to the new determin-ism. Rationality and Society, 16(1): 5–43.

Cusumano, M. A., Mylonadis, Y., & Rosenbloom, R. S. 1992.Strategic maneuvering and mass-market dynamics: Thetriump of VHS over Beta. Business History Review, 66:51–94.

David, P. A. 1985. Clio and the economics of QWERTY. Amer-ican Economic Review, 75: 332–337.

David, P. A. 1986. Understanding the economics of QWERTY:The necessity of history. In W. N. Parker (Ed.), Economichistory and the modern economist: 30–49. Oxford: Black-well.

David, P. A. 1994. Why are institutions the “carriers of his-tory”? Path dependence and the evolution of conven-tions, organizations and institutions. Structural Changeand Economic Dynamics, 5(2): 205–220.

Dosi, G. 1982. Technological paradigms and technologicaltrajectories: A suggested interpretation of the determi-nants and directions of technical change. Research Pol-icy, 11: 147–162.

Dosi, G. 1997. Opportunities, incentives and the collectivepatterns of technological change. Economic Journal, 107:1530–1547.

Downs, A. 1967. Inside bureaucracy. Boston: Little, Brown.

Eden, L. 2004. Whole world on fire: Organizations, knowl-edge, and nuclear weapons devastation. Ithaca, NY:Cornell University Press.

Feldman, M. S., & Pentland, B. 2003. Reconceptualizing or-ganizational routines as a source of flexibility andchange. Administrative Science Quarterly, 48: 94–118.

Garud, R., & Karnøe, P. 2001. Path creation as a process ofmindful deviation. In R. Garud & P. Karne (Eds.), Path


AQ:2

AQ:3

dependence and creation: 1–38. Mahwah & London:Lawrence Erlbaum Associates.

Ghemawat, P. 1991. Flexibility and commitment: The dynam-ics of strategy. New York: Free Press.

Giddens, A. 1984. The constitution of society. Outline of thetheory of structuration. Cambridge: Polity Press.

Gilbert, C. G. 2005. Unbundling the structure on inertia: Re-source versus routine rigidity. Academy of ManagementJournal, 48: 741–763.

Gresov, C., Havemen, H. A., & Olivia, T. A. 1993. Organiza-tional design, inertia and the dynamics of competitiveresponse. Organization Science, 4: 181–208.

Gulati, R., Nohria, N., & Zaheer, A. 2000. Strategic networks.Strategic Management Journal, 21: L203–215.

Guler, I. 2007. Throwing good money after bad? Political andinstitutional influences on sequential decision makingin the venture capital industry. Administrative ScienceQuarterly, 52: 248–285.

Hannan, M. T., & Freeman, J. 1984. Structural inertia andorganizational change. American Sociological Review,49: 149–164.

Hannan, M. T., Plos, L., & Carroll, G. R. 2004. The evolution ofinertia. Industrial and Corporate Change, 13: 213–242.

Hargadon, A. B., & Douglas, Y. 2001. When innovations meetinstitutions: Edison and the design of the electric light.Administrative Science Quarterly, 46: 476–501.

Helfat, C. E. 1994. Evolutionary trajectories in petroleum firmR&D. Management Science, 40: 1720–1747.

Hennestad, B. W. 1990. The symbolic impact of double bindleadership: Double bind and the dynamics of organiza-tional culture. Journal of Management Studies, 27: 265–280.

Hilborn, R. C. 2004. A simple model for stochastic coherenceand stochastic resonance. American Journal of Physics,72: 528–533.

Hollingsworth, R. 2006. A path-dependent perspective on in-stitutional and organization factors shaping major sci-entific discoveries. In J. Hage & M. Meeus (Eds.) Innova-tion, science, and institutional change: 432–442. NewYork: Oxford University Press.

Holm, P. 1995. The dynamics of institutionalization: Transfor-mation processes in Norwegian fisheries. Administra-tive Science Quarterly, 40: 398–422.

Huff, A. S., & Huff, J. O. 2000. When firms change direction.Oxford: Oxford University Press.

Hughes, T. P. 1987. The evolution of large technological sys-tems. In W. E. Bijker, T. P. Hughes, & T. J. Pinch (Eds.), Thesocial construction of technological systems: 51–82.Cambridge, MA: MIT Press.

Jarzabkowski, P. 2008. Shaping strategy as a structurationprocess. Academy of Management Journal, 51: 621–650.

Johnson, V. 2007. What is organizational imprinting? Cul-tural entrepreneurship in the founding of the Paris Op-era. American Journal of Sociology, 113: 97–127.

Karim, S., & Mitchell, W. 2000. Path-dependent and path-breaking change: Reconfiguring business resources fol-

lowing acquistions in the U.S. medial sector, 1978–1995.Strategic Management Journal, 21: 1061–1081.

Katz, M. L., & Shapiro, C. 1985. Network externalities, com-petition, and compatibility. American Economic Review,75: 424–440.

Kauffman, S. A. 1993. Origins of order: Self-organization andselection in evolution. Oxford: Oxford University Press.

Kauffman, H. 1995. The limits of organizational change. Edi-son, NJ: Transaction Books.

Kets de Vries, M. 2006. The leader on the couch: A clinicalapproach to changing people and organizations. WestSussex, UK: Wiley.

Kettinger, W. J., Teng, J. T. C., & Guha, S. 1997. Businessprocess change: A study of methodologies, techniques,and tools. MIS Quarterly, 21: 51–80.

Kim, T. -Y., Oh, H., & Swaminathan, A. 2006. Framing interor-ganizational network change: A network inertia per-spective. Academy of Management Review, 31: 704–720.

Koch, J. 2008. Strategic paths and media management: Apath dependency analysis of the German newspaperbranch of high quality journalism. Schmalenbach Busi-ness Review, 60: 51–74.

Kulik, C. T., Bainbridge, H. T. J., & Cregan, C. 2008. Known bythe company we keep: Stigma-by-association effects inthe workplace. Academy of Management Review, 33:216–230.

Lawrence, T. B., Winn, M. I., & Jennings, P. D. 2001. Thetemporal dynamics of institutionalization. Academy ofManagement Review, 26: 624–644.

Lee, T. W. 1999. Using qualitative methods in organizationalresearch. Thousand Oaks, CA: Sage.

Leibenstein, H. 1950. Bandwagon, snob, and Veblen effects inthe theory of consumer’s demand. Quarterly Journal ofEconomics, 64: 183–207.

Leonard-Barton, D. 1995. Wellsprings of knowledge. Boston:Harvard Business School Press.

Luhmann, N. 1995. Social systems. Stanford, CA: StanfordUniversity Press.

Mahoney, J. 2000. Path dependence in historical sociology.Theory and Society, 29: 507–548.

March, J. G. 1991. Exploration and exploitation in organiza-tional learning. Organization Science, 2: 71–87.

March, J. G. 1994. A primer on decision making. New York:Free Press.

March, J. G. 2006. Rationality, foolishness, and adaptive in-telligence. Strategic Management Journal, 27: 201–214.

Mason, R. O., & Mitroff, I. I. 1981. Challenging strategic plan-ning assumptions: Theory, cases and techniques. NewYork: Wiley.

Masuch, M. 1985. Vicious circles in organizations. Adminis-trative Science Quarterly, 29: 14–33.

McGregor, D. 1960. The human side of enterprise. New York:McGraw-Hill.

Merton, R. K. 1936. The unanticipated consequences of pur-


AQ:4

posive social action. American Sociological Review, 1:894–904.

Miller, D. 1992. Environmental fit versus internal fit. Organi-zation Science, 3: 159–178.

Miller, D. 1993. The architecture of simplicity. Academy ofManagement Review, 18: 116–138.

Miller, D., & Friesen, P. 1984. Organization: A quantum view.Englewood Cliffs, NJ: Prentice-Hall.

Moon, J. 1999. Reflection in learning & professional develop-ment. London: Kogan Page.

Nelson, R. R., & Winter, S. G. 1982. An evolutionary theory ofeconomic change. Cambridge, MA: Harvard UniversityPress.

Nooteboom, B. 1997. Path dependence of knowledge: Impli-cations for the theory of the firm. In L. Magnussion & J.Ottosson (Eds.) Evolutionary economics and path depen-dence: 57– 78. Cheltenham, UK:

North, D. C. 1990. Institutions, institutional change and eco-nomic performance. Cambridge” Cambridge UniversityPress.

Ortmann, G. 1995. Formen der Produktion: Organisation andRekusivitat [Form of production: Organization and re-cursiveness]. Opladen: Westdeutscher Verlag.

Pajunen, K. 2008. The nature of organization mechanisms.Organization Studies, 29: 1449–1468.

Panzar, J. C., & Willig, R. D. 1981. Economies of scope. Amer-ican Economic Review Proceedings, 71: 268–272.

Pettigrew, A. M., Woodman, R. W., & Cameron, K. S. 2001.Studying organizational change and development:Challenges for future research. Academy of Manage-ment Journal, 44: 697–713.

Pierson, P. 2000. Increasing returns, path dependence, andthe study of politics. American Political Science Review,94: 251–267.

Peirson, P. 2004. Politics in time. Princeton, NJ: PrincetonUniversity Press.

Pindyck, R. S., & Rubinfeld, D. L. 2005. Microeconomics (6thed.) Upper Saddle River, NJ: Prentice-Hall.

Piore, M. J., & Sabel, C. F. 1984. Second industrial divide:Possibilities for prosperity. New York: Basic Books.

Porac, J. F., Thomas, H., Wilson, F., Paton, D., & Kaufer, A.1995. Rivalry and the industry model of Scottish knit-wear manufacturers. Journal of Management Studies,26: 397–416.

Powell, W. W., & DiMaggio, P. 1991. The new institutionalismin organization analysis. Chicago: University of ChicagoPress.

Prahalad, C. K., & Hamel, G. 1990. The core competence of thecorporation. Harvard Business Review, 68(3): 79–91.

Repenning, N. P., & Sterman, J. D. 2002. Capability traps andself-confirming attribution errors in the dynamics of pro-cess improvement. Administrative Science Quarterly,47: 265–295.

Ross, J., & Staw, B. M. 1993. Organizational escalation and

exit: Lessons from the Shorehand nuclear power plant.Academy of Management Journal, 36: 701–732.

Ruef, M. 1997. Assessing organization fitness on a dynamiclandscape: An empirical test of the relative inertia the-sis. Strategic Management Journal, 18: 837–853.

Saffold, G. S. 1988. Culture traits, strength, and organization-al performance: Moving beyond “strong” culture. Acad-emy of Management Review, 13: 546–558.

Schon, D. A. 1983. The reflective practitioner. New York: BasicBooks.

Schreyogg, G., & Kliesch, M. 2007. How dynamic can capa-bilities be? Strategic Management Journal, 28: 913–933.

Schumpeter, J. A. 1942. Capitalism, socialism and democracy.New York: Harper & Row.

Scott, R. 2001. Organizations and institutions (2nd ed.). Thou-sand Oaks, CA: Sage.

Scott-Morgan, P. 1994. The unwritten rules of the game: Mas-ter them, shatter them and break through the barriers toorganizational change. New York: McGraw-Hill Profes-sional.

Selvini-Palazzoli, M. 1986. The hidden games of organiza-tions. New York: Random House.

Sewell, W. H., Jr. 1996. Three temporalities: Toward an event-ful sociology. In T. J. McDonald (Ed.), The historic turn inthe human sciences: 245–280. Ann Arbor: University ofMichigan Press.

Sorensen, J. B. 2002. The strength of corporate culture and thereliability of firm performance. Administrative ScienceQuarterly, 47: 70–91.

Staber, U. 2005. Entrepreneurship as a source of path depen-dency. In G. Fuchs & P. Shapira (Eds.), Rethinking re-gional innovation and change: Path dependency or re-gional breakthrough? 107–126. New York: Springer.

Staw, B. W. 1976. Knee deep in the big muddy: A study ofescalating commitment to a chosen course of action.Organizational Behavior and Human Performance, 16:27–44.

Stieglitz, N., & Heine, K. 2007. Innovations and the role ofcomplementarities in a strategic theory of the firm. Stra-tegic Management Journal, 28: 1–15.

Stimpert, J. L., Wasserman, M. E., & Jayaran, M. 1998. Strate-gic trajectories and patterns of innovation. In G. Hamel,C. K. Prahalad, H. Thomas, & D. O’Neal (Eds.), Strategicflexibility: Managing in turbulent environments: 51–73.New York: Wiley.

Stinchcombe, A. 1965. Social structures and organizations. InJ. G. March (Ed.), Handbook of organizations: 142–193.Chicago: Rand McNally.

Streatfield, P. J. 2001. The paradox of control in organizations.New York: Routledge.

Szulanski, G. 1996. Exploring internal stickiness: Impedi-ments to the transfer of best practices within the firm.Strategic Management Journal, 17(Winter Special Issue):27–43.

Teece, D. J., Pisano, G., & Shuen, A. 1997. Dynamic capabili-


ties and strategic management. Strategic ManagementJournal, 18: 509–533.

Thelen, K. 1999. Historical institutionalism and comparativepolitics. Annual Review of Political Science, 2: 369–404.

Tolbert, P. S. 1988. Institutional sources of organizationalculture in major law firms. In L. G. Zucker (Ed.), Institu-tional patterns and organizations: 101–113. Cambridge,MA: Ballinger.

Tolbert, P. S., & Zucker, L. 1996. The institutionalization ofinstitutional theory. In S. R. Clegg, C. Hardy, & North,C. E. (Ed.), Handbook of organizational studies: 175–190.London: Sage.

Tripsas, M., & Gavetti, G. 2000. Capabilities, cognition andinertia: Evidence from digital imaging. Strategic Man-agement Journal, 21: 1147–1161.

Tushman, M. L., Newman, W. H., & Romanelli, E. 1986. Con-vergence and upheaval: Managing the unsteady pace oforganizational evolution. California Management Re-view, 29(1): 29–44.

Van de Ven, A. H., & Poole, M. S. 2005. Alternative approaches

for studying organizational change. Organization Stud-ies, 26: 1377–1404.

Volberda, H. W., & Lewin, A. Y. 2003. Co-evolutionary dy-namics within and between firms: From evolution toco-evolution. Journal of Management Studies, 40:2111–2136.

von Foerster, H. 1991. Through the eyes of the other. In F.Steier (Ed.), Research and reflectivity: 63–75. London:Sage.

Walker, G., Kogut, B., & Shan, W. 1997. Social capital, struc-tural holes and the formation of an industry network.Organization Science, 8: 109–125.

Watzlawick, P. 1963. A review of the double bind theory.Family Process, 2: 132–153.

Watzlawick, P., Weakland, J., & Fisch, R. 1974. Change: Prin-ciples of problem formation. New York: Norton.

Witt, U. 1997. “Lock-ins” vs. “critical masses”—Industrialchange under network externalities. International Jour-nal of Industrial Organization, 15: 753–773.

Jorg Sydow ([email protected]) is a professor of management at the School ofBusiness & Economics, Freie Universitat Berlin, and a visting professor at the Univer-sity of Strathelyde’s Graduate School of Business. He received his doctorate from theFreie Universität Berlin. His research focuses on management and organization the-ory, strategic partnering and interfirm networking, technology and innovation man-agement, project management, and industrial relations.

Georg Schreyogg ([email protected]) is a professor of organization andleadership at the School of Business & Economics, Freie Universität Berlin. He re-ceived his doctorate from Universitat Erlangen-Nurnberg. His current research inter-ests include organizational change, path dependence, and organizational capabili-ties.

Jochen Koch ([email protected]) is a professor of management and organiza-tion at the European University Viadrina, Frankfurt. He received his doctorate from theFreie Universität Berlin. His research interests include modern and postmodern orga-nization theory, theories of leadership, media management, and the theory of organ-izational and strategic path dependence.


ORGANIZATIONAL PATH DEPENDENCE: OPENING THE BLACK BOX · 2012-11-16 · ORGANIZATIONAL PATH...

Documents

Transcript of ORGANIZATIONAL PATH DEPENDENCE: OPENING THE BLACK BOX · 2012-11-16 · ORGANIZATIONAL PATH...