Adaptive capacity to technological change: A microfoundational approach

• Published date: 01 June 2017
• Author: Vikas A. Aggarwal,Hart E. Posen,Maciej Workiewicz
• DOI: http://doi.org/10.1002/smj.2584
• Date: 01 June 2017

Strategic Management Journal

Strat. Mgmt. J.,38: 1212–1231 (2017)

Published online EarlyView 1 November2016 in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/smj.2584

Received 1 February 2015;Finalrevision received 1 August 2016

ADAPTIVE CAPACITY TO TECHNOLOGICAL CHANGE:

A MICROFOUNDATIONAL APPROACH

VIKAS A. AGGARWAL,1*HART E. POSEN,2and MACIEJ WORKIEWICZ3

1INSEAD, Fontainebleau, France

2University of Wisconsin-Madison, Madison, Wisconsin, U.S.A.

3ESSEC Business School, Cergy-Pontoise, France

Research summary: We take a microfoundational approach to understanding the origin of

heterogeneity in rms’ capacity to adapt to technological change. We develop a computational

model of individual-level learning in an organizational setting characterized by interdependence

and ambiguity. The model leads to organizational outcomes with the canonical properties of

routines: constancy,efcacy, and organizational memory.At the same time, the process generating

these outcomes also produces heterogeneity in rms’ adaptive capacity to different types of

technological change. An implication is that exploration policy in the formative period of routine

development can inuence a rm’s capacity to adapt to change in maturity. This points to a host

of strategic trade-offs, not only between performance and adaptive capacity, but also between

adaptive capacities to different forms of change.

Managerial summary: Why are rms differentiallyeffective at adapting to technological change?

We arguethat rms differ in the adaptive capacity of the routines that underlie their capabilities.

These differences arise well beforechange occurs, and result because rms build routines that are

differentially responsive to signals of performance decline associated with technologicalchange.

Thus, early managerial efforts to build superior productive efciency must be complemented

by efforts to build superior adaptive capacity. Our theory suggests that managers can prepare

for technological change by implementing policies, in the formative period of organizational

development, that promote individuals’ explorationof novel actions. However, there aretrade-offs

because preparation aimed at building adaptive capacity to one type of technologicalchange may

limit adaptive capacity to other types of change. Copyright © 2016 John Wiley & Sons, Ltd.

INTRODUCTION

The development of the strategy literature has been

strongly inuenced by two key themes: the drivers

of inter-rm performance heterogeneity, and the

factors underlying heterogeneity in the capacity

of rms to adapt to technological change. These

themes have evolved somewhat independently. The

Keywords: microfoundations; routines; adaptive capacity;

technological change; inter-rm heterogeneity

*Correspondence to: Vikas A. Aggarwal, INSEAD, Boulevard

de Constance, 77305 Fontainebleau Cedex, France. E-mail:

vikas.aggarwal@insead.edu

Authors contributed equally and are listed alphabetically.

Copyright © 2016 John Wiley & Sons, Ltd.

former focuses on endowments of unique resources

and capabilities (e.g., Barney, 1991; Peteraf, 1993;

Winter, 2000), while the latter focuses on fac-

tors such as complementary assets, organizational

interdependencies, and the nature of the change

itself (e.g., Henderson and Clark, 1990; Taylor

and Helfat, 2009; Tripsas, 1997; Tushman and

Anderson, 1986). In this article, we seek to link

these themes theoretically. We propose that the

same process underlying the emergence of routines,

which ultimately serve as building blocks of rm

capabilities, also endows rms with heterogeneity

in the capacity to adapt to technological change.

We suggest, moreover, that a manager’s strategic

choices during the formative period of a routine’s

Adaptive Capacity to Technological Change 1213

development (e.g., via setting the organization’s

exploration policy) may serve as a critical lever,

enabling rms to systematically develop routines

that are differentially adaptable to alternate forms

of technological change.

Organizational routines have become an increas-

ingly central theoretical construct in the strategy

literature because of their close link to rm capa-

bilities (Teece, 1982; Winter, 2000, 2003), and

consequently, the question of whether and how

these routines might change has become impor-

tant (Feldman and Pentland, 2003).1Routines may

reect managerial efforts to purposively coordinate

the actions of individuals— for example, through a

combination of standard operating procedures and

top-down hierarchy (Cyert and March, 1963; March

and Simon, 1958). At the same time, a substan-

tial stream of the literature on routines character-

izes their initial development as stemming from

a bottom-up process of learning (e.g., Cohen and

Bacdayan, 1994; Greve, 2008; Hodgson, 2008;

Hutchins, 1991; Knudsen, 2008; Levitt and March,

1988; Miner, Ciuchta, and Gong, 2008; Nelson and

Winter, 1982) that results in an organization-level

property that can be dened simply as “a repetitive

pattern of interdependent actions, involving mul-

tiple actors” (Feldman and Pentland, 2003: 96).2

These emergent routines then serve as the basis for

“continuity in the behavioral patterns” of organiza-

tions (Nelson and Winter, 1982: 96), and have been

thought to limit adaptation to external change (Han-

nan and Freeman, 1984; Leonard-Barton, 1992).3

1The literature has employed a range of terms that broadly relate

to the term routines. As Cohen, Levinthal, and Warglien (2014)

discuss, a range of “recurring organizational patterns” have been

studied with labels such as routine, practice, standard operating

procedure, etc.They propose using collective performance to

describe such patterns, with the common denominator being that

these all involve “habit-based actions.” This idea is consistent

with what we model here; for simplicity (and because a broader

discussion of terminology is beyond the present scope), we retain

the term routine, recognizing that our use of this term is in the

spirit of this notion of “collective performance.”

2See Becker (2004) for a review of the literature examining the

emergence of routines. Pentland and Feldman (2005) note that,

while standard operating procedures may indeed reect top down

coordination, they may also reect the outcome of a process of

codifying an emergent routine.

3Routines reect one of three mechanisms of organizational coor-

dination (March and Simon, 1958; Thompson, 1967), which

additionally include hierarchical authority and inter-individual

communication. While routines may emerge from bottom-up

individual-learning, the literature has also recognized the impor-

tant role that purposeful managerial action plays on them once

formed (March and Simon, 1958; Teece etal., 1997).

Yet, the literature also recognizes that the mech-

anisms of routine development do not abate after

routine emergence, and routines themselves are able

to adapt in dynamic settings (Feldman and Pent-

land, 2003). Recognizing that routines, as emergent

organizational properties, can and do adapt is a rst

step, but it is not alone sufcient. What matters to

strategy scholars is to identify sources of systematic

heterogeneity. However, the extant literature offers

only a very limited understanding of how systematic

heterogeneity in the capacity of routines to endoge-

nously adapt to technological change emerges and

evolves. Moreover, we have limited insight into the

ways in which managers might guide this process

through explicitly enacted organizational policies.4

To address this gap, we pursue a microfounda-

tional approach (Barney and Felin, 2013; Felin,

Foss, and Ployhart, 2015; Foss and Pedersen, 2014;

Winter, 2013). We propose that the value of doing

so resides not in studying individuals’ behaviors

per se, but rather from identifying how individ-

ual behaviors interact and aggregate to generate

macro-level phenomena. Our approach is anchored

in individual-level learning and the formation of

“habit” (Winter, 2013). To the degree that routines

can be conceptualized as “structures of interlocking

individual habits” (Hodgson, 2008: 25), the evolv-

ing beliefs of individuals are central not only to the

emergence of routines (Becker, 2004; Greve, 2008;

Knudsen, 2008; Levitt and March, 1988; Miner

et al., 2008), but also to their endogenous evolution

and adaptation over time (Becker et al., 2005).

We computationally model a process that could

lead to the emergence of routines in organizations.

We build on the bandit model of individual learn-

ing under uncertainty (e.g., March, 1996; Posen

and Levinthal, 2012), extending this to an orga-

nizational environment in which one individual’s

actions impact the outcomes of the others in the

organization (Levinthal and March, 1993; Thomp-

son, 1967). The individual-level learning process is

one in which individuals observe organization-level

outcomes, and then update beliefs based on these

outcomes. Individuals thus act in the context of sub-

stantial ambiguity (March and Olsen, 1975). As a

baseline result, our model of individual-level learn-

ing in an ambiguous organizational environment

produces outcomes with the canonical properties

of routines: constancy, efcacy, and organizational

4See Foss et al. (2012) for a recent discussion that highlights these

issues.

Copyright © 2016 John Wiley & Sons, Ltd. Strat. Mgmt. J.,38: 1212–1231 (2017)

DOI: 10.1002/smj