Understanding Dynamic Capabilities

Author(s): Sidney G. Winter

Source: Strategic Management Journal, Vol. 24, No. 10, Special Issue: Why Is There a Resource
-Based View? Toward a Theory of Competitive Heterogeneity (Oct., 2003), pp. 991-995
Published by: John Wiley & Sons
Stable URL: [Link]
Accessed: 27/10/2009 15:31

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
[Link] JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
[Link]

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@[Link].

John Wiley & Sons is collaborating with JSTOR to digitize, preserve and extend access to Strategic
Management Journal.

[Link]
 Strategie Management Journal
Strat. Mgmt. J., 24: 991-995 (2003)
Published online inWiley InterScience ([Link]). DOI: 10.1002/smj.318

UNDERSTANDING DYNAMIC CAPABILITIES

SIDNEY [Link]*
The Wharton School, University of Pennsylvania, Philadelphia, Pennsylvania, U.S.A.

Defining ordinary or 'zero-level' capabilities as those that permit a firm to 'make a living' in the
short term, one can define dynamic capabilities as those that operate to extend, modify or create

ordinary capabilities. Logically, one can then proceed to elaborate a hierarchy of higher-order
capabilities. However, it is argued here that the strategic substance of capabilities involves

patterning of activity, and that costly investments are typically required to create and sustain
such patterning?for in product development. Firms can accomplish without
example, ' change
reliance on dynamic capability, by means here termed 'ad hoc problem solving. Whether higher
order capabilities are created or not depends on the costs and benefits of the investments relative
'
to ad hoc problem solving, and so does the 'level of the game at which strategic competition
effectively occurs. Copyright ? 2003 JohnWiley & Sons, Ltd.

Many strategy scholars remain skeptical about the of course will really prove to be so only if it is
value of the concept of 'dynamic capabilities.' widely adopted.
While some see dynamic capabilities as the key to
competitive advantage (Teece, Pisano, and Shuen,
1997), others seem to doubt that there actually are ROUTINES AND CAPABILITIES
such things. Still others believe that they exist,
but that are 'born, not made'?i.e.,
suspect they
Following my own proposal (Winter, 2000), I
they doubt that deliberate efforts to strengthen such
are a genuine begin by founding the concept of organizational
capabilities option for managers.
capability on the broader concept of organiza
And some believe that while they are a genuine
are not necessarily tional routine: An organizational capability is a
option, they something that
high-level routine (or collection of routines) that,
confers competitive advantage. This note seeks to
together with its implementing input flows, confers
reduce the mystery surrounding both the terminol
an a set
It identifies some key upon organization's management of deci
ogy and the phenomenon.
issues and argues that clarity is served by keep
sion options for producing significant outputs of a
particular type. For present purposes, the points
ing these issues distinct. As regards terminology,
deserving emphasis here are the connotations of
it offers a proposal that seems constructive?but
'routine'?behavior that is learned, highly pat
terned, repetitious, or quasi-repetitious, founded in
Key words: dynamic capabilities; change; cost-benefit; part in tacit knowledge?and the specificity of
problem solving
objectives. Brilliant improvisation is not a rou
^Correspondence to: Sidney G. Winter, The Wharton School,
University of Pennsylvania, 2000 Steinberg Hall-Dietrich Hall, tine, and there is no such thing as a general
3620 Locust Walk, Philadelphia, PA 19104-6370, USA. purpose routine.

Copyright ? 2003 JohnWiley & Sons, Ltd.

 992 S. G. Winter

'DYNAMIC CONNOTES CHANGE that stationary process are the zero-level capabili
ties, the 'how we earn a living now' capabilities.
There is a broad consensus in the literature that Without them, the firm could not collect the rev
'dynamic capabilities' contrast with ordinary (or enue from its customers that allows it to buy more
'operational') capabilities by being concerned with inputs and do the whole thing over again. By con
change. Collis (1994) is particularly explicit and trast, capabilities that would change the product,
formal in making the point that dynamic capabil the production process, the scale, or the customers
ities govern the rate of change of ordinary capa (markets) served are not at the zero level. New
bilities. Following the example of the differential product development, as practiced in many firms,
calculus, he points to the existence of second is a prototypical example of a first-order 'dynamic
order, third-order, etc. dynamic capabilities and capability.' The capabilities that support the cre
explicitly makes the extension 'ad infinitum.' This ation of new outlets by McDonalds or Starbucks
terminological approach is adopted here, but with are another prototypical example, focused on the
an important caveat as to whetherhigher-order domain of scale and (geographic) markets rather
capabilities 'exist' in an interesting
sense. From than product attributes. These examples are proto
a logical point of view, the 'existence' of higher typical because they unquestionably involve first
order rates of change is in question only in the order change, given the definition of the zero level,
mathematical sense that some derivatives might and it is equally beyond question that they are
not exist; and from a computational point of view, highly patterned and 'routine' in many respects.
a time sequence of N + 1 values of a variable suf Given the terminological framework under con
fices to compute one value of the Nth order rate of struction here, these examples are a conclusive
are similar to answer to anyone who doubts the 'existence' of
change. But if dynamic capabilities
capabilities in that they involve patterned activity dynamic capabilities. (Of course, their doubts may
oriented to relatively specific objectives, then there relate to a different understanding of 'dynamic
is no guarantee that the organizational processes capability.')
governing high-order change are highly patterned, It is worth noting that the 'zero level' is only
and substantial reason to think otherwise. In this locally defined. For a firm that does its own R&D,
important substantive sense, high-order dynamic the producing and selling the product is zero-order
capabilities do not necessarily exist. This point is activity. For an independent R&D lab, developing
new products is zero order activity.
pursued below.

THE 'ZERO LEVEL' IN THE THERE ARE MANY WAYS TO CHANGE

CAPABILITY HIERARCHY
It is quite possible to change without having a
Constants and technical issues aside, everything is dynamic capability. To begin with, change often
the derivative of its integral and the integral of occurs by force majeure from the environment,
its derivative. To make effective use of the con predictably or not, for better or worse. Whether
cept of a hierarchy of rates of change, we need it is because such an external challenge arrives
a convention to identify the 'zero level,' the ana or because an autonomous decision to change is
logue of position for variables moving in space. made at a high level, organizations often have to
Because capabilities are complex, structured and cope with problems they are not well prepared for.
multidimensional, this question may not have an They may be pushed into 'firefighting' mode, a
answer that seems both clear and compelling in all high-paced, contingent, opportunistic and perhaps
cases. There is, however, a heuristic guide avail creative search for satisfactory alternative behav
able that conforms to common sense and existing iors. It is useful to have a name for the category
practice, at least for the capabilities of firms com of such change behaviors that do not depend on
peting in markets. Consider a hypothetical firm 'in dynamic capabilities?behaviors that are largely
equilibrium,' an organization that keeps earning non-repetitive and at least 'intendedly rational'
its living by producing and selling the same prod and not merely reactive or passive. I propose 'ad
uct, on the same scale and to the same customer hoc problem solving'. Ad hoc problem solving is
population over time. The capabilities exercised in not routine; in particular, not highly patterned and

Copyright ? 2003 JohnWiley & Sons, Ltd. Strat. Mgmt. J., 24: 991-995 (2003)
 Understanding Dynamic Capabilities 993

not repetitious. As suggested above, it typically construction, equipment and furnishings, advertis
appears as a response to novel challenges from ing campaigns,
new
employees, etc., and create

the environment or other relatively unpredictable a new outlet.

The size of that central capabil
events. Thus, ad hoc problem solving and the exer ity determines the pace at which new outlets can
cise of dynamic capabilities are two different ways be opened. For these sorts of commitments to be
to change?or two categories comprising numer economically sound, the capability must be exer
ous different ways to change. cised: to have a dynamic capability and find no
Of course, close study of a series of 'fires' may occasion for change is merely to carry a cost bur
well reveal that there is pattern even in 'firefight den. On the other hand, an aggressive search for
ing.' Some of the pattern may be learned and such occasions may also be a mistake. Attempting
contribute positively to effectiveness, and in that too much change?perhaps in a deliberate effort
sense be akin to a skill or routine. In organiza to exercise the dynamic capability?can impose
tional improvisation, as in jazz, creative achieve additional costs when the frequent disruption of
ment typically rises from a foundation of pat the underlying capability outweighs the compet
terned and practiced performance, a fund of micro itive value of the novelty achieved. There is an
patterns that are recombined and sequenced in cre ecological demand for balance between the costs
ative ways (Miner, Bassoff, and Moorman, 2001). of the capability and the use that is actually made
to highly dynamic environments may of it. (Of course, deciding whether some dynamic
Responses
also be patterned at a higher level, guided by capability is needed is only a small part of the total
adherence to relatively simple rules and structural problem of making profitable capability invest
principles (Eisenhardt and Martin, 2000). At the ments; the larger part is deciding which among the
other end of the spectrum, even the most incre many promising but uncertain investments should
mental effort at product modification can run into be undertaken?recognizing there are likely to be
snags that are beyond the scope of the trade-offs or other interactions among them.)
unexpected
dynamic capability, and require a complementary By contrast, the costs of ad hoc problem solving
dose of ad hoc problem solving. To acknowledge largely disappear if there is no problem to solve.
these points is not, however, to concede that there Many of those costs take the form of opportunity
is no difference between dynamic capabilities and costs of personnel who have alternative produc
ad hoc problem solving; to say that would be to tive roles in the (zero-level) capability. True, an
that relied on
indulge in the 'shades of gray' fallacy. For the con organization dynamic capabilities
and the contrast to be useful aids to under could conceivably have a similar cost pattern. This
cepts
would mean that people could step out of their
standing, it is not necessary that the pure forms
exist in the world, or even that we have high 'inter zero-level roles and into their dynamic capability
rater reliability' in sorting real cases into only two roles?their learned, patterned change roles?and
boxes. then step back again when change was completed.
conceptual
The plausibility of this image is undercut by the
'rustiness' problem: successful maintenance of a
skill or routine typically requires frequent exercise.
CONTRASTING COST STRUCTURES of whether that objection is decisive in
Regardless
itself, it seems that, in practice, prominent exam
Dynamic capabilities typically involve long-term ples of dynamic involve a
capabilities generally
commitments to specialized resources. The more lot of specialized who are committed
personnel
pervasive and detailed the patterning of the activity full time to their change roles, and other types of
involved, the higher the costs of the commitments investments as well. The contrast with the lighter
tend to be. The ability to sustain a particular pat cost burdens of ad hoc problem is clear.
solving
terned approach to new product development, for
example, depends to some extent on continuity in
the engineering personnel involved; there may be NO 'RULE FOR RICHES' HERE
substantial continuity in facilities and equipment
as well. Similarly, an established replicator orga It should now
be apparent that it is not neces
nization has a central staff that is the locus of its sarily advantageous for a firm to invest in (first
ability to bring together real estate, design skills, order) dynamic capabilities. Rivals who rely on ad

Copyright ? 2003 John Wiley & Sons, Ltd. Strat. Mgmt. J., 24: 991-995 (2003)
994 S. G. Winter

hoc problem solving to accomplish change when superiority at a lower level. Here, the same skepti
needed are carrying a lower cost burden. If oppor cal conclusion about advantage rests on the alter
tunities for competitively significant change are native argument that ad hoc problem solving is
sparse enough or expensive enough to realize, then always a substitute for dynamic capability and may
the added cost of dynamic capabilities will not be be economically superior. Collis also makes, how
matched by corresponding benefits on the aver ever, the related interesting suggestion that there is
age?even if an occasional notable success might a historical tendency for the locus of competitive
suggest the contrary. Also, if the change environ action to rise in the capability hierarchy. Strategic
ment does sustain dynamic capabilities relative to innovation often involves 'changing the game' in a
ordinary capabilities (plus ad hoc problem solv way that 'takes it to a higher level'?a phrase that
firms a often connotes a focus on strengthening higher
ing), competition among many pursuing
similar dynamic capabilities order change capabilities. This notion appeals at
strategy may compete
the descriptive level and there is clearly some logic
away the rents, because (for example) product mar
kets are saturated with rival innovations or because to it. Knowledge advances cumulatively, imitation
the salaries of scientists and engineers are bid up. spreads solutions around, and problems that are
A related and long-familiar example of a disadvan visible, urgent, and recur at high frequency tend
to get solved before problems with the opposite
tageous dynamic capability is innovative R&D that
does not pay off in the presence of strong rivals attributes. But these considerations do not suffice
who invest only in imitative R&D (see, for exam tomake the progression to a higher 'order' of com

ple, Nelson and Winter, 1982). The (descriptive) petition a logical necessity, since the levels differ
rule for riches is to occupy a favored and relatively in the cost-benefit balance of capability invest
ments, and exogenous change could at any time tip
uncontested place in the ecology of behaviors?for
an existing balance in favor of lower-order capabil
example, by having a strong R&D unit when others
ities supplemented by ad hoc problem solving. The
in the industry lack both innovative and imitative
an individual actor can exert argument for such upward progression is therefore
capabilities. While
missing an appropriate assumption that restricts the
some influence on that ecology, it may well have
character of exogenous change in such a way as to
difficulty in identifying the favored and uncon
assure that the investment in higher-order capabil
tested niches, or in forestalling unfavorable change
ities tends to pay off, while the cost-cutting move
after such a niche has been exploited for a while.
in the opposite direction does not. Just how such
Hence, the rule is of little prescriptive value.
an assumption might be framed is unclear, but the
logic is incomplete without it.

HIGHER-ORDER CAPABILITIES
REPRISE

If exogenous change is 'competence destroying' some of the mystery and confusion sur
Probably
at the level of first-order dynamic
capabilities, the of arises
rounding concept dynamic capability
those who invest in routinizing the response to from linking the concept too tightly to notions of
familiar types of change may find themselves dis effectiveness at dealing with change
generalized
advantaged relative to more flexible players who and generic formulas for sustainable competitive
have invested in higher-order capabilities. Delib advantage. The argument here is that clarity is
erate investments in organizational learning may, served by breaking this linkage. There is no way
for example, facilitate the creation and modifica to hedge against every contingency. There is no
tion of dynamic capabilities for the management of That
general rule for riches. investing in dynamic
acquisitions or alliances (Zoilo and Winter, 2002). (of whatever order) can be a partial
capabilities
Collis (1994) argues that the existence of higher hedge against the obsolescence of existing capa
order capabilities provides a rebuttal to any claim bility, and can sometimes yield relatively sustain
that there is generally advantage to be had from able advantage, is obvious from the nature of
strength at any particular level of dynamic capabil 'dynamic capability,' as defined here. That this
ity: there is always a higher level, and in his view cannot be uniformly or inevitably advantageous is
superiority at the higher level always 'trumps' equally obvious, from the meaning of 'investing:'

Copyright ? 2003 John Wiley & Sons, Ltd. Strat. Mgmt. J., 24: 991-995 (2003)
 Understanding Dynamic Capabilities 995

the thought experiment of raising the costs while REFERENCES

holding the gross benefits constant makes the net
benefit disappear, and certainly the world is capa Collis DJ. 1994. Research note: how valuable are

ble of turning such a thought experiment into a real organizational capabilities? Strategic Management
Journal, Winter Special Issue 15: 143-152.
experiment. The concept of dynamic capability is
Eisenhardt KM, Martin J. 2000. Dynamic capabilities:
a helpful addition to the tool kit of strategic anal what are they?' Strategic Management Journal,
ysis, but strategic analysis itself remains a matter Special Issue 21(10-11): 1105-1121.
of understanding how the idiosyncratic attributes Miner AS, Bassoff P, Moorman C. 2001. Organizational
and learning: a field Administra
of the individual firm affect its prospects in a par improvisation study.
tive Science Quarterly 46: 304-337.
ticular competitive context.
Nelson RR, Winter SG. 1982. The Schumpeterian
trade-off revisited. American Economic Review 72:
114-132.

ACKNOWLEDGMENTS Teece D, Pisano G, Shuen A. 1997. Dynamic capabilities

and strategic management. Strategic Management
Journal 18(7): 509-533.
I am indebted to David Collis, Tammy Madsen,
Winter SG. 2000. The satisficing principle in capability
Maurizio Zoilo, and participants in the Wharton Issue
learning. Strategic Management Journal, Special
Technology Mini-Conference for comments on 21(10-11): 981-996.
earlier drafts; remaining flaws are my own respon Zoilo M, Winter SG. 2002. Deliberate learning and
the evolution of
sibility. Research support from the Reginald H. dynamic capabilities. Organization
Science 13: 339-351.
Jones Center of the Wharton School is gratefully
acknowledged.

Copyright ? 2003 John Wiley & Sons, Ltd. Strat. Mgmt. J., 24: 991-995 (2003)