ArticlePDF Available

Industry Evolution and the Strategic Role of Complementarities in the Theory of the Firm

January 2002

Authors:

Nils Stieglitz

Frankfurt School of Finance & Management

Klaus Heine

Erasmus University Rotterdam

Content uploaded by Klaus Heine

Content may be subject to copyright.

· 1 ·

First draft, preliminary version

Industry Evolution and the Strategic Role of Complementarities

in the Theory of the Firm

Nils Stieglitz and Klaus Heine

DRUID Winter Conference at Aalborg, Denmark,

January 17-19, 2002

Nils Stieglitz Klaus Heine,

Organizational and Personnel Economic Policy Unit

Economics Unit Department of Economics

Department of Economics Philipps-University Marburg

Philipps-University Marburg Universitaetsstrasse 25a

Universitaetsstrasse 24 35032 Marburg, Germany

35032 Marburg, Germany heine@wiwi.uni-marburg.de

stieglitz@wiwi.uni-marburg.de

· 2 ·

1. Introduction

Recent evolutionary economic research on innovations and technological change has

stressed the importance of complementary resources and institutions explaining the innovative

activities of firms and countries (Teece/Pisano/Shuen 1997, Heine/Kerber 2002).

Complementary assets play an important role in the contract based theory of the firm, too, and

recent theorizing has stressed their importance for the existence, boundaries, and

organizational structures of the firm. Despite this common interest, there has been virtually no

cross-fertilization between contract based theories and the evolutionary theory of the firm. In

our paper we show how the theory of complementarities offers the potential to integrate the

insights of these different views within a common conceptual framework. By aligning incentives

to the creation and usage of knowledge in organizations, a deeper understanding of the

behavior of the innovative firm in the evolution of markets is gained.

In the evolutionary theory of the firm, the need to access and to coordinate diverse

complementary resources is highlighted. In this view, the firm is a collection of similar and

complementary resources which cannot be traded in markets (Richardson 1972). In other

words, the firm is not seen as a simple well-behaved production function or as a atomistic

contract combining smoothly the assets of super-rational agents. In the evolutionary theory of

the firm the specific knowledge is stressed which is necessary to combine crucial resources for

the production of unique goods. This leads to the paradigmatic idea that the firm constantly

faces the challenge to create new knowledge and to use it profitably in the market place

(Teece/Pisano/Shuen 1997). A key problem for an innovative firm is to gain access to

complementary resources which are not part of the firm's knowledge base yet and to

coordinate the use of these resources.

The need to coordinate the creation and usage of complementary activities is also a major

theme in the works of Milgrom and Roberts (1995). While not a theory of the firm proper,

they nevertheless show that strong complementarities make it necessary to coordinate their

creation and usage centrally. From the perspective of incomplete contracts (Hart 1995),

complementary and specific assets should be placed under single ownership. And more

recently, Rajan and Zingales (2001) have extended the incomplete contracting framework. In

· 3 ·

their theory, the firm is a collection of critical and complementary resources, and the degree of

uniqueness of the resources allocates power within the organization. From their perspective,

the firm is faced with the challenge to use the knowledge it created because there is, for

example, always the threat employees will expropriate the knowledge and use it themselves.

Therefore, complementary intangible assets can be a powerful device to prevent employees

from leaving the firm and ensures that new knowledge will be used where it has been created.

It is tempting to connect this recent thinking on the importance of complementarities with

the broader view of the meaning of knowledge cultivated in evolutionary thinking. This sort of

integrative reasoning leads to a conceptual framework in which the creation and protection of

knowledge plays a crucial role. From its focus our framework builts upon a framework

developed by Teece and others (Teece 1996, Teece/Pisano 1994, Teece/Pisano/Shuen

1997), in which the organization of knowledge and complementary assets plays the strategic

role to get a sustainable competitive advantage. However, the difference we will make is to

give more weight to the contractual nature of the firm.

The paper is organized as follows. In a first step we take a look on some central reasoning

in the evolutionary theory of the firm that highlights the meaning of knowledge-creation and use

of knowledge, heterogeneity and market processes. But also evolutionary theorizing is a

powerful tool to criticize neoclassical economics it lacks up to now a coherent body from

which “hard facts” can be derived as easily as in neoclassical economics. Not overcoming this

problem pretty satisfying, our next step is to figure out and to integrate evolutionary reasoning

with recent research in the (neoclassical) theory of the firm, that focuses on incentives,

complementarities, and appropriation of rents. The third step is then to summarize our findings

and to sketch out some normative conclusions concerning business policy as well as public

policy.

2. Knowledge, Heterogeneity, and Market Process

According to the later work of Schumpeter (1950) we may argue that it is nowadays not

so much the genious entrepreneur who implements most of innovations, but that this is done via

· 4 ·

large established firms (Tripsas 1997).1 These firms have market power and large resources,

and the crucial task for them is to sustain and to strengthen their position by managing and

coordinating the different capabilities and resources they have built up over time. From such a

resource-based view of the firm it becomes clear that it is worthwhile to draw closer attention

on the properties of industry evolution and the interrelatedness of these properties.

Also the following three properties of industry evolution are not clear cut definitions they

play an almost overwhelming role in evolutionary theorizing. In short, they are: (1) the respect

of the crucial role of knowledge, i.e., in an economy agents have to cope with the task of

creation (innovation) and use (imitation) of knowledge; (2) the heterogeneity of products and

factors – and hence the heterogeneity of firms; (3) the existence of cycles, which means that an

economic system does not behave as a stable mechanistic system like a clockwork, but mere

as an biological system that develops over a life-cycle. It is possible to call such an approach

“Darwinian”, in the sense that it is granted that single entities might differ from each other by

mutation but that there is to the same time a population of genetic similar entities. In the long

run such a population evolves through its capability to adapt to its environment that selects the

entities with the superior genetic program (Mayr 1984, pp. 95). In the realm of economics we

may speak instead of species of a population of activities embedded in firms that form an

industry which evolves through variation of its technological and organizational skills. In an

economy the selection of firms and activities is done via competition that is constrained by

rules and institutions like antitrust law or securities regulation.2 Because the criteria of selection

are made by man some authors (Commons 1934, Vanberg 1995) call it an artificial selection

that may be used to drive an industry in the direction that the society wants to direct it to.

From a methodological point of view we might say that the notion of variation and selection

1 This is not to say that there is nowadays no longer an entrepreneur that pilots a firm. The

opposite is right, however, the entrepreneurial function is not longer done by a single

entrepreneurial Superman but by a sophisticated management that holds the

entrepreneurial function in a large firm (Cohendet/Llerna/Marengo 2000). This

entrepreneurial function of the management may be seen as a specific asset which is

characterized by its “dynamic capabilities” to strategically manage the firm in an ever

changing environment (Teece/Pisano 1994).

2 For this point see the seminal paper of Alchian (1950).

· 5 ·

are part of the paradigmatic core of evolutionary economics that are in sharp contrast to the

“static walrasian worlds” of neoclassical economics (Vanberg 2001).3

In the following we will take a closer look on the meaning of creation and use of

knowledge, firm heterogeneity, and market processes. But also it would be tempting to discuss

every point broadly from different strands of thought we will restrict ourselves, and so we will

focus only on selected items.

(1) Creation and use of knowledge

When we analyse the problem of creation and use of knowledge, a first idea might suggest

to apply the twin of “innovation” and “imitation” to get a deeper understanding of an economy

of knowledge. Besides the fact that “innovation” and “imitation” are very broad labels which

have to be refined for theoretical and practical usage, it is a crucial question on which front of

Coase’ firm (1937) we are theorizing these labels. So, the meaning of innovation and imitation

undertakes a so called “Gestalt-Switch” (Schlicht 1998), if we take a step inside or outside the

firm. In a metaphoric sense, outside the firm-gates firms appear like islands in the sea of

markets. Firms are seen as the creators of new knowledge, or in terms of Schumpeter (1934)

they create new production-functions, which outperform the more reluctant firms. In this

outside-view the competition process is organized via markets in which price signals are the

medium through that knowledge is transmitted. This sort of transmission and organization of

knowledge in a market economy is a point intensely stressed by Hayek (1937, 1943, 1968)

and put forward by Austrians like Kirzner (1973). And in the tradition of Schumpeter and

Hayek it is not surprising that there are tries to develop an integrative framework that

combines the ideas of the young Schumpeter (1934) and of Hayek (1937, 1945).4 But the

severe question arises if this sort of thinking that was developed for the analysis of market-

coordination5 is also feasible for theorizing the coordination-problems inside the firm, or for

theorizing the myriads of organizational forms between market and hierarchy (hybrids,

networks, symbiotic arrangements etc.).

3 For a broader outline of evolutionary theorizing, see for example Dosi and Nelson (1994).

4 For such an integrative market-view see the tentative outline of Kerber (1997).

· 6 ·

Inside the firm we switch from coordination via markets to coordination via hierarchy

(Coase 1937). The coordination-problems of hierarchies in large multiproduct firms are in the

center of the reasoning about industry evolution in the work of the late Schumpeter (1950).

Because in the following sections the analysis of the coordination-problems in the innovating

large firm are put forward, we may now highlight firm heterogeneity.

(2) Firm heterogeneity

Firm heterogeneity can be seen as a consequence of two intertwined causes. The first is

that firms economize on different knowledge-bases, at last in part. E.g., firms develop different

product and process innovations. By doing so they create or use new knowledge about the

(re-)combination of factors and activities, and hence they differentiate from each other. But

such a differentiating process of “creative destruction” is not an isolated activity without

premises, because the capability to alter or to adapt the knowledge-base relies heavily on a

firm’s (dynamic) capabilities which were acquired in the past through (mutual) learning of the

agents a firm consists of. So, we may say that historical processes are at work in shaping the

firm’s knowledge-base, or put it in another way, path dependences are at work in the

unfolding of a firm’s knowledge-base.

Relying on Richardson (1972) it is possible to make a distinction between a firm’s

capabilities and a firm’s activities for which the capabilities are needed. “What concerns us

here is the fact that organisations will tend to specialise in activities for which their

capabilities offer some comparative advantage; these activities will, in other words,

generally be similar…” (Richardson 1972, p. 888). This notion of similarity leads to the

idea that firms face a co-evolution of the selected field of activities and the built up firm-

specific capabilities. But similarity of activities is only one factor which shapes a firm’s specific

characteristics leading to firm heterogeneity. The other factor is complementarity (Richardson

1972); in a broad sense, by complementarity is meant that the value of one capability or of one

activity is positively correlated to another capability or activity. From this follows directly that

increasing one capability or activity in order to generate additional value needs a simultaneous

5 See, e.g., the Lerner vs Hayek/Mises debate about the possibility of central planning.

· 7 ·

increase in the related capabilities or activities. Therefore, we may conclude that firm

heterogeneity is also a function of complex complementarities, put together in a more or less

unique organizational system which is able to manage and to exploit the existing

complementarities. Such reasoning following Richardson (1972) sheds not only light on the

question why firms differ, but also on the question where the boundaries of the firm may lie, a

question not easy to answer from a pure transaction-cost perspective.

In short, it becomes clear that capabilities, activities and complementarities are features of

heterogeneous firms that need coordination and appropriate incentives for the involved agents.

This organizational task has to be done by a central management.

(3) Market process

The notion of market process denotes the fact that organization and firm-behaviour does

not take place in a static (walrasian) world that reproduces to every time the same quality of

products and factors in the same quantity. The meaning of market process is that firms interact

in a changing world of preferences, technology, products and institutions. The central features

of market process-analysis have got recognition elsewhere (Klepper 1997, Schreiter 1994,

Heuss 1965); so, for our purposes it is enough to take into account that a firm will have to

organize its capabilities according to its selected activities in a changing environment.

Also the investigation of market processes is a manifold field there is, at least one, theme

that is hardly to underestimate. It is the cyclical evolution of entire industries and technologies,

or a little bit more de-emphasized, it is the cyclical rise and fall of products from an infant stage

of development to a mature one. Broadly seen, we suggest that a firms activities run through

from a stage of infancy to a stage of maturity; and in every stage of the life-cycle there is need

for a strategic management that centrally coordinates the interplay between activities,

capabilities and complementarities. Central coordination is needed in order to generate the

appropriate amount of new knowledge and to use the stock of knowledge generated in the

past. Also it is not easy to derive strong normative recommendations for organizing firms from

such a general assertion it seems plausible that a firm at the beginning of a life-cycle will face

some different problems than a firm engaging in a well established industry. In this regard an

· 8 ·

often mentioned organizational pattern is that in a “new” industry more “fluid” organizational

conceptions are appropriate while an “older” industry needs some more routinized

conceptions that might allow a higher degree of division of labour and knowledge. The need

for more fluid organizational concepts is also plausible if we take into account that in an early

developmental stage of an industry firms not only face a high degree of uncertainty belonging to

the accurate connection of activities, capabilities and complementarities, but also that there

may be different strategic paths for a firm which will yield an equi-efficient performance in the

future. A recent example for the existence of such different but equi-efficient organizational

paths leading to multiple-equilibria may be found in the lasting coexistence of different

governance-systems and the underlying governance-philosophies.6 In short, the respect of

market processes gives weight to the fact that the organization of knowledge is not a static

phenomenon but a dynamic one that needs the capacity of a strategic management that directs

the learning process on firm level.

Also evolutionary theorizing about industry evolution highlights central features, as e.g.,

creation and use of knowledge (learning), firm heterogeneity and market processes, these sort

of conceptual thinking appears to often only as a lucid critique of neoclassical economics

without being explicit and rigorous as neoclassical economics. Surely, from a puristic

methodological point of view this must not be a failure, but it leads to the situation that the two

schools (paradigms) of economic thinking often focus on the same problems without cross-

fertilizing each other. One field in that such a cross-fertilization might be of great value is the

modern theory of the firm. While the importance of complementarity is often stressed by

evolutionary theorists the associated coordination, incentive and internal appropriation

problems are seldom analysed in detail.7 The question is if it might be possible to align the

conceptual and empirical power of the evolutionary theory of the firm with the more narrow

neoclassical theory of the firm, that investigates the behaviour of the firm’s agents from a

rigorous contractual point of view.

6 For a deeper analysis of the path dependence of whole governance-systems from an

evolutionary point of view see Heine and Stieglitz (2001) and Heine and Kerber (2002).

7 For the fruitfulness of such an integrative approach see also the outline of Cohendet, Llerna,

and Marengo (2000, pp. 110) and of Coriat and Dosi (1999).

· 9 ·

In the next sections we will try to align neoclassical contract based theories of the firm with

evolutionary reasoning as it has been developed in the so called resource-based view of the

firm. From such a syntheses it might not only be possible to get further and deeper insights into

the working of real world firms and real world industries but also to fill the gap between high-

powered theories and the political advice that can be derived from those theories, as e.g. in the

field of merger control or the legislation of corporate law (Heine/Stieglitz 2001, Heine/Kerber

2002).

3. Technology, Coordination, and Complementarities

The seminal papers of Milgrom and Roberts (1990, 1995) mark the starting point for the

renewed interest in the role of organizational complementarities. Essentially, two key insights

form the basis of the analysis of Milgrom and Roberts. First, building on the works of Topkis,

they used mathematical lattice theory to elucidate economic problems.8 Other scholars have

picked up this method and applied it to such diverse subjects as oligopoly pricing (Vives

1999) and growth theory. Second, on a conceptual level, they highlighted the importance of

complementary activities for explaining the adoption of new organizational structures, the

development of new technologies, and for vertical integration.

The starting point for Milgrom and Roberts (1990, 1995) is the empirical observation that

the adoption of modern manufacturing technologies was followed by widespread changes in

organizational structures, human resource policies, and market positioning. This led to the

emergence of the “modern manufacturing firm”. For example, technologies like Computer-

Aided-Design and Computer-Aided-Manufacturing made the production process much more

flexible. With less specialized equipment, it was possible to offer more varieties of major

products and update the product lines more frequently. Thereby, “on-demand” production

became feasible in a range of industries. In addition, firms implemented new human resource

policies with fewer job classifications, reduced inventory stocks (“just-in-time”), increased

8 For a different modelling methodology, see especially Levinthal (1997), Levinthal (2000), and

Ghemawat and Levinthal (2000). Instead of using lattice theory, these works build on the

evolutionary nk-simulation approach developed by Kauffmann (1993). N stands for the

number of choice variables, while k for the number of variables each choice depends

upon.

· 10 ·

reliance on subcontractors, and a higher emphasis on speed in order processing, production,

and delivery. Moreover, in R&D, design, product, and process engineering was integrated to

increase the speed of the introduction of new products.

Milgrom and Roberts (1990) explain the emergence of these arrangements by arguing that

the various activities, and characteristics described, are mutually complementary and tend to

be adopted together. They give a very precise meaning to the term of complementarity: “The

defining characteristic […] of complements is that if the levels of any subset of

activities are increased, then the marginal return to increases in any or all of the

remaining activities raises. It then follows that if the marginal costs associated with

some activities fall, it will be optimal to increase the level of all of the activities.”

(Milgrom/Roberts (1990), S. 514). Thus, their concept of complementarity is more restrictive

than the one outlined above, since it focuses on symmetric or interdependent relationships

(Alchian/Woodward 1987). Accordingly, if the marginal return of adopting or increasing one

activity raises, it becomes more attractive to change other activities.

From this basic property of complementary activities, Milgrom and Roberts derive a range

of implications (for details, see Milgrom/Roberts 1995). If strong complementarities between

activities exist, the gains from increasing every component are larger than the sum of the

individual increases. While complementarities direct the search for new ventures and

improvements through their effect on the marginal return of activities, decentralized co-

ordination or mutual adjustment will often be not enough to exploit those benefits. There is a

marked tendency for coordination failures if strong complementarities exist. To see this,

consider the following coordination game where two managers decide independently over the

values for the activities x and y. The following table (Table 1) shows the total profit they want

to maximize (no corporate governance problem here).

· 11 ·

Low y High y

Low x 5 4

High x 34 + x

Table 1

Assume that x < 1, so the initial choices are (low, low). If x increases to x > 1, the new

optimum is (high, high). The problem is that uncoordinated search will not lead to the new

equilibrium (as there is no incentive to change x or y unilaterally) and the firm can be stuck in

the suboptimal initial equilibrium. Thus, organizations will often lack the motivation to change,

since the gains can only be realized if other activities are also modified. It follows that, to reap

these benefits of complementary activities, some kind of central coordination is usually

required. Individual searches and adaptations will fail to converge upon an optimal equilibrium

to begin with, and mistaken variations are less costly when they are coordinated than when

they are made independently. This underlines the importance of coherence between activities.

Imitation of coherent systems of complementary activities is more difficult since it not

enough to imitate one activity but a whole range in a highly specific manner. Dynamically,

upward or downward movement of whole system of complementary activities tend to

continue, so there is the possibility that a vicious or virtuous cycle is set in motion until a new

equilibrium is reached. Lastly, these feedbacks between complementary activities make the

existence of multiple-equilibria more likely.

Overall, strong complementarities give rise to the need of mutual adjustment and

particularly of central strategic direction by general management. Here, general management’s

task is to identify the relevant complementarity structures, and to recommend a fruitful

direction accordingly. Milgrom and Roberts (1995) stress that the informational requirements

to obtain these benefits are quite low, since it is more than enough to identify

interdependencies between activities and tasks, and communicate these to lower level

managers by setting appropriate organizational objectives and rules. The lower level manager

· 12 ·

then just has to concentrate on attaining his objectives, without having to worry about other

parts of the organization.

This conception of strategic direction has a long tradition in strategic management theory,

where the basic idea is that different competitive strategies entail corresponding sets of activity

systems (Andrews 1987, Porter 1996). Changes in competitive positioning will lead to far-

reaching adjustments in firm’s technologies, organizational structures, and corporate resources.

But in the Milgrom and Roberts framework, technological and marketing skills are common

knowledge, and firms are able to purchase all input factors on the markets. If one drops this

key premise, and takes the heterogeneity of firms’ capabilities and resources into account, then

the capability and resource base of the firm will have to change correspondingly, too. New

capabilities do not mysteriously fall from the sky, but they are the outcome of deliberate

learning processes as Winter (2000, p. 984) has stressed: “To create a significant new

capability, an organization must typically make a set of specific and highly

complementary investments in tangible assets, in process development, and in the

establishment of relationships that cross the boundaries of the organizational unit in

which the process is deemed to reside.” Consequently, there is not only a need to coordinate

well-defined complementary activities, but the interdependent learning processes as well.9

By coordinating complementary investments and corresponding activities through the

setting of organizational objectives and structures, the competitive strategy of a firm decides

what kind of knowledge and how much of it will be created and how it will be exploited in the

markets. The analysis of strategic paths by Teece, Pisano and Shuen (1997) is extended by

taking this need to coordinate complementary learning processes and activities into account.

More generally, to be successful, radical technological and strategic innovations call forth

for active, centralized direction. A radical innovation is more systemic (Rosenberg 1982) in

nature, since the whole organization has to be modified and curtailed to its needs. In the

9 According to Foss and Foss (2000), this is an important reason why firms exist: “It is in the

handling of some of the coordination problems associated with interdependence between

tasks that we find the rationale of the firm.”

· 13 ·

framework of Milgrom and Roberts, radical innovations show strong complementarities with

other activities of the firm e.g. production, marketing, and so forth. From this perspective, the

need for active, centralized direction may decrease over the industry life-cycle.

In the early stages of industry evolution, radical innovations are still quite common as a

dominant design has not yet emerged (Abernathy/Utterback 1978). Moreover, the

competitive strategies of firms are constantly evolving as firms are trying to find attractive

market positions. In this early stage of industry evolution active, centralized strategic direction

is needed to manage the learning processes in and between complementary activities

(Marengo 1992). Central direction allows to rapidly change these processes and activities in

the light of new specific knowledge about technological and market conditions. Since multiple-

equilibria exist, and activity systems are quite complex, they are themselves the object of the

organizational learning process. As a consequence, central direction is needed to experiment,

and guide the search for a consistent activity system. Mutual adjustment may not only fail to

identify relevant complementarities, but is slow to respond to new developments and reverse

past decisions.

In the later stages of industry evolution, when incremental innovations are driving

competition, a firm must have a coherent activity system tightly in place in order to survive the

selection process. If a firm still continues to look for a basic structure to deal with

complementary activities, its ability to compete is seriously restricted. While competitors have

settled for a competitive strategy and focus on perfecting their ways of doing business, they are

still defining their businesses. An imitation of competitor’s activity systems is not feasible,

because activity systems are complex in nature and most of it non-observable to outsiders.

Firms which have an consistent activity system and organizational rules may rely less on central

direction. Decentralized decision making and mutual adjustment are effective ways to

coordinate complementary activities. Active central direction plays a much smaller role now.

The danger of this arrangement is that the firm is slower to adopt to radical new technologies.

It may take some time until general management realizes the need to overhaul the activity

system.

· 14 ·

The need for consistent activity systems seems to be part of the reason why established

firms often organize their systemic innovative effort in autonomous teams or independent

divisions. This organizational arrangement gives management the freedom to devise a

competitive strategy and organizational structure which takes the complementarities of the new

product into account, and which is not restrained by existing products and their activity

systems.10 Experimentation and learning processes for the new product are conducted without

inhibiting the production and marketing of other products. This is especially relevant for

diversified firms which rely on independent divisions to coordinate the activities for different

products and markets. For firms who have to manage different generations of products, on the

other hand, autonomous teams open the possibility for focused and controlled experiments

which allow to analyze the impact of the new generation on the existing activity system en

miniature. If the product is ready to be launched, the activity system can then be modified

accordingly. Much the same is true for internal process innovations. Then again, the

management of incremental and autonomous product and process innovations can be

decentralized since the capabilities and corresponding activity system is already in place.

Broadening the basic theory, Holmstrom and Milgrom (1994) argue that complementarities

not only exist between the firm’s activities but also between their organizational incentives.

While activities form coherent activity systems, incentives form incentive systems that will be

adopted on a whole. Specially, they apply this reasoning to the make or buy decision and

maintain that high performance incentives, workers ownership of assets, and worker freedom

from direct controls are complementary instruments to prevent workers from shirking. Thus,

the decision to buy is a decision over what kind of incentive system will be put in place.

Likewise, if a firm makes the product itself, it installs a different incentive system in which the

firm owns the assets, the worker is paid a fixed wage and is supervised by the firm. This basic

idea can be easily extended. Burns and Stalker (1961) claimed that so called organic

structures are the efficient response to rapidly changing and highly complex environments.

Organic structures are characterised by large spans of control, decentralized decision-making,

10 For example, Clark and Fuijmoto (1994) argue in their study of the automobile industry that

product managers with centralized decision rights (“heavyweight team manager”) are

needed to manage product development processes which cut across organizational

· 15 ·

low levels of specialisation, and only a small amount of formal rules. The mechanized structure,

on the other hand, relies on formal rules, a high degree of internal specialization, small spans of

control, and centralized decision-making. One does not have to follow Burns and Stalker

(1961) in each and every point in order to accept the basic idea, namely that different

organizational variables form organizational systems. More recently, Teece (1996) has

identified four different archetypical “governance modes” which could be interpreted as

organizational systems. What remains unclear in his analysis is why the characteristics are

complementary and have to be adopted together.

Obviously, the Milgrom and Roberts framework is not without its flaws from an

evolutionary point of view. The analysis already indicated that Milgrom and Roberts (1990) do

not take the heterogeneity of firms into account. The activities they discussed are generic in the

sense that the activities are beneficial to all firms within an industry or even economy wide. This

does not have to be the case if firm heterogeneity and differences in competitive strategy are

taken into account. Resources and activities are, therefore, firm-specific. In addition, the

complementarity and interaction between activities may be firm-specific as well

(Porter/Siggelkow 2001). Milgrom and Roberts only analyse the case of generic activities

which interact in a generic way. Still, the Milgrom and Roberts framework is important as it

illuminates the coordination problems associated with strong complementarities and the key

role of centralized direction. While the evolutionary theory of the firm has always stressed the

need to closely coordinate systemic innovations, the Milgrom and Roberts framework takes a

broader approach by taking into consideration a wider range of activities, and by deepening

our theoretical understanding of this phenomena.

4. Complementarities, Property Rights, and Incentives in the

Incomplete Contracts Framework

Today, the theory of incomplete contracts is arguably the dominant theory to explain the

existence and boundaries of the firm (e.g. Schmitz 2001). Building on transaction cost theory

by Klein, Crawford and Alchian (1978) and Williamson (1985), Grossman and Hart (1986)

functions. On the other hand, functional teams relying on mutual adjustment fail to

coordinate these systemic innovations.

· 16 ·

and Hart and Moore (1990) have build on the basic idea that hierarchy is the efficient

governance mode if contracts are incomplete.11 Their key insight is ownership of physical

assets matters since it allocates the right to decide in events not specified by the initial contract

(Hart 1991). They equate ownership of an asset with residual rights of control, or, more

specifically, with the right to exclude others from the use of an asset (Hart/Moore 1990). This

corresponds closely with the legal definition of ownership. Ownership of assets is a source of

bargaining power that can be tapped into if the need to renegotiate an incomplete contract

arises. The allocation of ownership rights has in that way a profound influence on the

distribution of the ex post surplus generated by a contractual relationship. Assuming that

rational individuals will foresee this link between ownership and surplus distribution, the

allocation of residual rights influences the ex ante incentives to generate this surplus. For

instance, the incentive to invest into relationship-specific assets is higher if the asset is owned

by the individual who decides over the amount of investment. In this case, she receives a larger

part of the generated surplus as it is the case if a third party owns the asset.

Overall, who owns what, is determined by the structure of incentives of the contracting

parties. Integration increases the acquiring firm’s incentive to make relationship-specific

investments, since it receives the residual control rights and a larger share of the created

surplus will accrue to the firm. The cost of integration is that the incentives of the acquired firm

will be diminished because it receives a smaller fraction of the surplus created by its

investment. Thereby, who acquires whom depends on the precise shape and structure of the

relevant incentives to invest (Hart 1991). This basic thought can be applied to a range of

different situations. First, it is not only possible to discuss the incentives to invest into the

creation or enhancement of an asset, but also to discuss the effort put into activities before

surplus distribution. Potentially, there is a great deal to learn about the creation and usage of

assets by utilizing the property rights framework put forth by Grossman and Hart (1986).

Second, they explicitly model the incentive problems associated with complementary

physical assets (Hart/Moore 1990, Hart 1991). Two assets are said to be strictly

11 See Hart (1995) and Williamson (2000) for a discussion of the differences between

transaction cost economics and incomplete contract theory.

· 17 ·

complementary if both are unproductive unless they are used together (Hart/Moore 1990, p.

1135). Once again, an interdependent relationship is assumed. It follows directly that single

ownership of the two complementary assets is the efficient contractual arrangement

(Hart/Moore 1990). The acquiring firm’s incentive to invest will increase while having no effect

on the incentive of the acquired firm. In this situation integration yields only benefits but no

costs. The reason is that the assets are unproductive and yield no marginal return absent an

agreement between the two owners. Therefore, investment incentives can only increase

through integration, but not diminish.12 Schmitz and Sliwka (2001) extend the analysis by

assuming that a separate firm B not only choose the asset’s investment level but also the

degree of complementarity between B’s asset and an asset of a firm A. The more

complementary B’s asset, the more productive are the two assets. But B’s bargaining power

decreases as B becomes more dependent on A, and this reduces B’s investment incentive.

Now integration of complementary assets is no longer costless. A basic trade-off arises: While

integration leads to full complementarity but underinvestment, non-integration entails a loss in

complementarity but increases investment incentives. Integration then only will occur if the

benefits of stronger complementarity are greater than the costs of diminished investment

incentives. Therefore, in cases of complementary assets or strong “synergy” between assets,

single ownership is the appropriate ownership arrangement.

What is interesting about the above analysis is the focus on the allocation of property rights

and how this, in turn, affects the incentives of the rights’ holder. In the light of this, ownership

structures have an impact on the creation and usage of knowledge in firms. But the outlined

model is very restrictive. The implications are derived under the assumption that the productive

knowledge needed to coordinate the use of the assets is costless. If we once again drop this

assumption, and take the heterogeneity of firms into account, conclusions are less clear.

Richardson (1972) already argued that activities could be complementary but dissimilar, since

they rely on a different set of capabilities. In a similar vain, Demsetz (1995) claimed that

vertical integration forces the central contractor (e.g. the owner) to digest knowledge of

different kind. If knowledge acquisition is costly, “forsaking reliance on a specialized body of

12 But see Hart (1995, pp. 68), for special cases in which joint ownership is the efficient

arrangement.

· 18 ·

knowledge and using multiple kinds of knowledge instead raises the cost per unit of

knowledge acquired.” (Demsetz 1997, p. 428). Taking the costs (and benefits) of acquiring

new knowledge through learning and experimentation into account, the above analysis of

vertical integration has to be crucially amended.

However, the fundamental point retains its validity, namely, that how property rights are

allocated affect the incentives of the contracting parties. Of course, this insight is not new. It

formed the conceptual underpinnings of earlier works on property rights (Demsetz 1967,

Alchian/Demsetz 1972). The “new property rights school” (Foss/Foss 2001a) has the merits

to provide a more tightly reasoned, game theoretic approach to analyze contracts and to give a

very precise economic meaning to ownership. But the more rigorous foundations come at the

price of neglecting other aspects of property, especially the contractible rights to access, use,

modify, and expropriate an asset. Most of the recent discussion has concentrated on the

residual right to withdraw the asset. But if one is interested in the organizational structures, a

more finely grained framework is needed to understand the theoretical relationships between

ownership, incentives, and decision rights on the one hand, and the creation and usage of

knowledge on the other.

In an important way, residual, discretionary rights to use an asset is the prerequisite for

creating new knowledge as Foss and Foss (2000, p. 39) have stressed: “This connection is a

consequence of the fact that learning by doing usually requires the exercise of use rights

over some assets. This implies that patterns of learning by doing […] depends on the

allocation of use rights between different individuals over time, and the division of

labour may be one reason for reallocation of user rights.” The incentive to learn and

innovate will, of course, depend on the benefits of this activity, and the incentive to innovate

increases if the innovator captures part of the generated surplus. The internal allocation of user

and income rights thus has an impact on the creation of knowledge in the firm. On the other

hand, the usage of existing knowledge may be curtailed as user rights are privately consumed

(“shirking”) while the firm fails to exploit its capabilities in the market place.

· 19 ·

Once again, it is general management’s discretionary task to settle this basic trade-off13

through the firm’s competitive strategy. Hence, in a fundamental way, organizational structure

follows competitive strategy. Put differently, the level, origins, and speed of the firm’s

innovative activities decides the corresponding incentives and organizational structure

(Cohendet/Llerena/Marengo 2000). To do this, general management – whether owner or

hired gun – needs the control rights over the firm’s asset including the rights to redefine and

reallocate specific user rights. Otherwise, they would not be able to change and adapt

incentives and organizational structure.

One may ask, what all the later have to do with the role of complementary assets? Hart and

Moore (1990) set out to explain the optimal allocation of ownership rights in cases of

complementary physical assets. Their argument points to the importance of property rights and

incentives in the creation and usage of assets. But the analysis can be broadened by

understanding property as a bundle of contractible rights. This opens up the possibility for a

more detailed study of the boundaries and the organizational structure of the firm.

5. Critical Resources, Power, and Webs of Complementarities

In a stream of recent papers, Rajan and Zingales (1998, 2000, 2001a) have developed a

“critical resource theory” of the firm. Taking Grossman and Hart (1986) and the resource-

based view of the firm (Wernerfelt 1984) as their starting points, they claim that power flows

from the ownership of valuable resources that are in short supply (Rajan/Zingales 1998). The

allocation of power14 affects the distribution of the surplus, and hence, the incentives and

decisions of those involved. Thereby, power not only affects the incentive to invest but the

incentive to enter in a relationship. Above that, the allocation of power also determines the

range of feasible actions. For example, a senior manager which wields the power to allocate

organizational roles and use rights over assets has a broader range of possible actions than a

lower level manager.

13 See March (1991) for a more complete treatment of the trade-off between the creation and

usage of knowledge.

14 Rajan and Zingales (2000, p. 204) define power as “the control over valuable resources over

and above that determined through explicit contract in a competitive market.”

· 20 ·

The key research question for Rajan and Zingales is how it is possible for a firm to have

power over human resources. As it is, legal ownership of human capital rests with the

individual, and is thereby, absent slavery, non-contractible. But a firm needs bargaining power

in order to extract part of the surplus which is generated by its employees. Otherwise, the

employees receive pay equal to their marginal productivity and thereby appropriate all of the

generated surplus. This, of courses, decreases profit rates. For example, Chacar and Coff

(2000) find empirical evidence that research analysts in investment banking are able to extract

most of the generated surplus: While gross performance looks splendid, residual performance,

e.g. profits accruing to the firm’s owners, are sobering. On a different level, Bhide (2000)

reports that 71 percent of the firms included in the Inc 500 (a list of young, fast growing firms)

were founded by people who are developing and marketing product ideas created by the

previous firm that employed them. This finding highlights that today many inventive firms are

not able to exploit the growth options they generate. This adds a new dimension to the

question how firms create and use knowledge: Many firms create new knowledge, but fail to

use it themselves. A crucial problem faced by a firm is, therefore, how firms secure the option

to use the knowledge created by their employees for themselves. They paid for it, after all.

Not surprisingly, according to Rajan and Zingales (2000), we find the answer in the

existence of complementary assets which provide the economic link that glues the firm

together. “A complementarity is said to exist when the unit and the firm can together

create more value than they can going their own separate ways.” Individuals are tied to

the firm through their quasi-rents, and it wields some power over them because disobedience

would jeopardize the quasi-rent or joint value. Complementary assets are the outcome of the

specialization to the existing resource base of the firm (Rajan/Zingales 2001b). Despite making

them more dependent, individuals are ready to acquire firm-specific skills and knowledge if the

long-term rewards outweigh the costs. One important way to create these situations is to give

employees privileged access to critical resources (Rajan/Zingales 1998). A critical resource is

a valuable resource which is in very short supply. Examples include existing customer and

client lists, single individuals, brand reputation, and corporate culture. Of course, the same

holds true not only for individuals, but for firms specializing their resources to other firms. An

important example are specialized customer-supplier relationships (Rajan/Zingales 2000).

· 21 ·

Existing firms are, from this perspective, webs of past investments that created

complementarities around a critical resource (Zingales 2000). Since it takes time to build such

a web, it can not be reproduced instantly. Individuals or other firms create new firm-specific

resources only if these webs are valuable and worth to gain access to. An important task is the

coordination and direction of these investment into new complementary resources. Rajan and

Zingales (2000) argue that this general position is a source of power of its own right. While not

controlling a resource as such, a person acts as the gatekeeper by allocating the access rights

to the valuable web of resources.

Concerning innovative activities, firms are better able to exploit the generated growth

options if they not only have the necessary complementary resources in place, but are also

able to create new options faster than competitors or new firms. This way, the incentives to

leave the firm are small for employees, since they would risk failure in direct competition and

the expected rewards for being loyal are high. “Therefore, the most enduring leads are

based upon a combination of organizational capabilities and technological leadership.

When a firm innovates at a very fast pace, and it has a large specialized core of

employees to implement these innovations, its opportunities will be well protected.”

(Rajan/Zingales 2001a, p. 208). Accordingly, firms with a narrow, focused corporate strategy

are in a better position to create and exploit growth options than diversified firms.15 More

focused firms not only provide the complementary assets to successfully market innovations,

but offers the potential to retain technological leadership in its field.16

Innovations may also be hampered if incentives to exploit growth options are misaligned

internally. In that case, the trouble is that an organizational unit may have to pay the costs,

while the potential reward accrues to another unit. Zingales (2000) provides some anecdotal

evidence for this claim. With its publishing and new media divisions, German media

conglomerate Bertelsmann was well positioned to exploit the opportunities opened by the

15 In a similar vain, Rotemberg and Saloner (1994) show that the employee’s incentives to create

and communicate new knowledge to upper level manager are higher if the firm pursues a

narrow business strategy.

16 See Markides (1995) for detailed empirical evidence supporting the relative superiority of

focused corporate strategies. Of course, the reasons may be different than the one given

by Rajan and Zingales (2001a).

· 22 ·

Internet. Still, Bertelsmann got off to a very slow start. Zingales (2000, p. 1649) offers the

interpretation that the divisions would have had to invest heavily to develop a new business

while fearing that it could be reassigned to an other or a new division in the future. As a result,

the divisions were reluctant to commit themselves, and Bertelsmann did not exploit this

opportunity.

Overall, a common theme in the works of Rajan and Zingales is how firms secure part of

the surplus and the growth options they create. Put differently, the economic link between the

creation and use of knowledge cannot be taken as given. This is a very well-known

phenomenon in market evolution, and the peculiar relationship between innovator and imitator

has been debated for decades. What has often been overlooked, though, is the internal

problem of internal rent appropriation, and Rajan and Zingales move it into the spotlight.

What is interesting about their treatment is that they explicitly take account of firm

heterogeneity and Schumpeterian competition. This adds an important dimension to the

discussion, since the ability to use the internally created knowledge differs across firms and is

an outcome of past decisions. Only if a firm has crafted the appropriate web of

complementary assets and capabilities in the past, it is able to exploit internally its generated

growth opportunities.

Employees only have an incentive to leave the firm if they expect that they are able to

generate a higher surplus elsewhere and are able to capture more of the surplus in absolute

terms. A competitor will be an attractive new employer if he is ready to pay higher wages. This

is only possible if the competitor has a more valuable web of complementary assets in place.

The employee will then switch firms as long as the wages are higher, even if the share of the

captured surplus decreases. The employee shares relatively more of the generated surplus to

gain access to the more valuable web of assets. Individuals even might be ready to accept

temporarily lower wages to gain access in hope of high future wages. But when do employees

start their new businesses, then?

Evidently, they must expect to produce a larger surplus by themselves than the surplus

which they can capture by staying in their old firm or by switching to a competitor. These

· 23 ·

constellations appear if existing webs of assets and capabilities fail to add much value to

innovations, or even prevent them. In the later case, the existing firm might fail to perceive the

economic potential of these new ideas. Radical innovations are often competence-destroying

(Tushman/Anderson 1986). Because of this characteristics, incumbent firms are not in a good

position to exploit radical new ideas, since the old web of complementary assets is nearly

worthless for that venture. Accordingly, employees could be better off by leaving the firm and

starting their new venture. The more radical an innovation, the more new ventures will be

started, and the seeds of creative destruction will often be sown in the incumbent firms.

It will depend on the exact nature of the technology and market structure to pinpoint how

radical innovations are (Tripsas 1997, McGahan/Silverman 2001). Often, seemingly radical

technological innovations only destroy assets and capabilities on parts of the value chain, while

leaving other parts intact. For example, in the pharmaceutical industry, the advent of

biotechnology, changed fundamentally how research and development is organized. Hence,

many new entrants entered in downstream markets. Their primary task is the creation of new

knowledge, and they are very R&D intensive. But these ventures still rely on the vertical

integrated pharmaceutical firm to market their products. The reason is that the resources and

capabilities of the traditional pharmaceutical firms like marketing, regulatory management, and

brand names, are complementary to new ventures’ innovations (for empirical studies, see

Arora/Gambardella 1990, Shan/Walker/Kogut 1994, Rothaermel 2001). The biotechnology

firms are ready to share part of the innovative rent to gain access to these valuable

complementary assets. Vertical integrated biopharmaceutical firms have remained the

exception.

6. Conclusions: Complementarities, Innovations, and Industry

Evolution

The starting point of our discussion was the need of firms to constantly create and exploit

new knowledge. This is one of the main features of the evolutionary theory of the firm. What

has often been lacking in evolutionary theorizing has been how incentives, and organizational

structures, influence the organizational learning process. Specifically, while the importance of

complementary capabilities is often acknowledged, the associated coordination, incentive,

· 24 ·

and internal appropriation problems are seldom analyzed in detail. But these questions are

addressed in the contract based theories of the firm.

In the works of Milgrom and Roberts, the need to coordinate complementary activities is

highlighted. Central direction, organizational goals and rules, mutual adjustment, and

organizational autonomy are important instruments to cope with this coordination problem. The

theory of incomplete contracts offers the potential to discuss the link between the allocation of

property rights, and the incentive to create and use complementary assets and capabilities. But

the modern property rights literature restricts itself to analyzing residual rights of control, e.g.,

the right to withdraw an asset. This approach is probably too narrow to discuss the full range

of property rights and their influence on incentives. Finally, Rajan and Zingales identify internal

appropriation of innovative rents as an important problem. They argue that complementary

assets and capabilities are important instruments for firms to successfully exploit its growth

options. Their analysis takes explicitly account of firm heterogeneity and links up well with

Teece (1986) who shows the importance of complementary assets for innovative and imitative

activity.

To see how these three problems relate to each other, and how they are shaped by the

forces of Schumpeterian competition, it is useful to look at stylized stages of industry evolution

and analyze how successful firms manage complementary activities, capabilities and resources.

Two dimensions seem to be especially useful to characterize industry evolution. First, the

degree of cumulativeness is a measure of how radical an innovation is: “Cumulativeness

captures the properties that innovations and innovative activities of today form the

starting point for innovations of tomorrow and that innovative firms of today are more

likely to innovate in the future in specific technologies and along specific trajectories

than noninnovative firms.” (Dosi/Malerba/Orsenigo, p. 216). A low degree of

cumulativeness, therefore, indicates a more radical innovation. The second dimension is

industry growth, and is a rough measure for industry evolution. It is a stylized fact that

industries often evolve through stages of high growth into saturated stages of low growth.

· 25 ·

Taken as a whole, we can distinguish four different constellations of industry evolution which

are depicted in the following table (Table 2).17

Market growth

Degree of Cumulativeness

High Low

Low (1) Exploratory Stage (4) Stagnant Phase

High (2) Growth Stage (3) Mature Stage

Table 2

New markets and industries are usually created by radical innovations which constitute new

technological paradigms (Dosi 1988) and new value networks (Christensen/Rosenbloom

1995). In this exploratory or embryonic stage, major product innovations are taking place

which eventually evolve into a dominant design (Abernathy/Utterback 1978, Agarwal 1998)

which channels further innovations. While market volume is low, growth rates are increasing.

The establishment of a stable technological trajectory often marks the beginning of the growth

stage which is characterized by high growth and declining entry of new firms. Often, an

industry shakeout and consolidation signals the transition to the mature stage with low growth

rates. In the fourth phase, the stagnant phase, two challenges for established firms are of

interest for our discussion. First, since growth opportunities are dried up, established firms will

look to other product markets to use their capabilities and diversify accordingly

(Dosi/Teece/Winter 1992, Montgomery 1994). Second, there is a threat that incumbents are

threatened by innovative new entrants who are successfully entering the market. Both

challenges have in common that diversification and the threat of innovative entrants are often

characterized by low degrees of cumulativeness, since new capabilities, new value networks,

and new activity systems are needed. We will discuss each stage in turn, and ask

• how complementary activities are coordinated,

• how incentives are structured,

17 For a detailed discussion of the different stages, see Klepper (1997).

· 26 ·

• how the internal appropriation problem is dealt with.

(1) Exploratory stage

In the exploratory stage, the creation of knowledge is of paramount interest to early movers

in the market. Because of the low degree of cumulativeness, the technology itself and the

corresponding technological capabilities are key areas for organizational learning. As outlined

above, centralized direction is needed to coordinate the complementary investments and

activities, and to guide the search for a consistent activity system. The objectives which guide

lower level managers’ behavior are loose, and open to constant revision, since there is a need

to experiment with ‘how to do things’. Formal rules only play a minor role in coordinating

complementary activities. Thus, the organization heavily relies on active central direction.

Concerning incentives, emphasis is placed on learning and experimenting. Employee’s scope

for action is broad, but constrained by above mentioned objectives. By setting objectives, the

corresponding use and decision rights are allocated. The allocation of residual income rights,

e.g. stock options, is important for setting pecuniary incentives to learn. The internal (and

external) division of labour is low, and internal organizational structures are horizontal. Since a

web of complementary capabilities and assets is just beginning to develop, the internal

appropriation problem is a serious threat to an innovative firm. An internal instrument for

coping with that threat is the use of different schemes of surplus sharing (stock options) with

early employees. Above that, a crucial task for general management is the rapid build up of

these webs. A first step is to identify valuable complementary capabilities, and decide whether

to develop them internally or acquire them externally through mergers or networks.18

Complementary resources do not only guard the firm against internal appropriation, but is an

important barrier to imitation and entry for newcomers, too.

18 Also in the law and economics literature these questions have got attention under the label

”symbiotic arrangements“ in which the structure and working of “thick contracts” for

managing complex transactions is analysed. For an overview see Schanze (1998).

· 27 ·

(2) Growth stage

During the growth stage, the degree of cumulativeness of new knowledge increases. The

creation of knowledge is still dominant, but the areas of organizational learning shift from

technology to market conditions, especially consumer preferences. The coordination of

complementary activities relies less heavily on active centralized direction, since vital

interdependencies have already been identified. Only if the firm changes it’s competitive

strategy and corresponding activity system, the need for active centralized direction comes up

again. Formal rules and specific objectives are thus gaining in significance. The internal division

of labor deepens as the coordination of differentiated tasks and activities is more easily

accomplished. Correspondingly, organizational learning processes multiply and accelerate, and

are becoming more decentralized (Marengo 1992). Changes in the incentive structure reflects

this shift, and residual use and decision rights become more specified and restricted. In other

words, since more is known about the technology and market structure, contracts are

becoming more “complete”, and the room for employee’s discretion narrows. Overall, the

vertical hierarchy of the firm deepens. The complementary capabilities and assets of the firm

are still developing, but they already provide an important competitive advantage for early

movers. Late movers are, therefore, disadvantaged along two dimensions: They not only have

to develop their capabilities and assets but experiment with complementary activities. Since

early movers already established basic structures and capabilities, experimenting and learning

is more expensive for late movers. This appears to be one of the reasons why early movers

often outperform later entrants (Mueller 1997). Furthermore, because of this, the internal

appropriation problem is becoming less severe. While still an important challenge, the (static)

internal rent appropriation through high wages and salaries begins to be the more serious

threat, and firms are more reluctant to establish organizational schemes for surplus sharing for

later employees.

(3) Mature stage

In the mature stage, market growth slows, and price competition often intensifies. The

(efficient) usage of existing knowledge becomes increasingly more important as technological

· 28 ·

opportunities and market structure is known. Firms which are still active in the market have a

consistent activity system tightly in place, with heavy reliance on formal rules, a deep and highly

differentiated vertical structure, mutual adjustment, and inactive central direction. The internal

division of labor is high. The level of organizational learning depends on the exact nature of the

industry, but it will usually be smaller than in the earlier stages of industry evolution. The

incentive system places emphasis of static efficiency, and less on learning and innovation.

Respectively, use and decision rights are narrow and well-specified, and not often revised. The

classic principal-agent theory can be fruitfully applied here. Concerning appropriation, the

dynamic appropriation problem is not important anymore. Since a web of complementary

capabilities and assets is now in place, firms are able to expropriate parts of the surplus

generated by their employees.

(4) Stagnant Stage

In the stagnant stage, firms often face the challenge to generate new growth opportunities.

As these are dried up in the traditional market, firms look elsewhere, to foreign markets or

new products. Without analyzing the decision to diversify in detail, two observations seem to

be possible from the above analysis. A first crucial question is which of the existing capabilities

and assets the firm is able to use, and the answer determines how much emphasis is placed on

the creation of new capabilities. The second question which needs to be addressed is how to

organize the new venture. Often, diversified firms have more than one autonomous

organizational unit, and this reflects the need to have different activity systems. Both questions

are interrelated, and the organizational challenge is how to integrate the desire for joint use of

organizational capabilities and assets with the need to establish autonomous activity systems.

Much the same is valid in analyzing challenges to the incumbent firms posed by radical

innovations during “stagnant” stages of market evolution. Usually, incumbent firms have an

advantage over new entrants, since they have complementary resources in place which allows

for quick imitation (Teece 1986). But a crucial source of organizational inertia is the old

activity system of the firm which may not longer be able to accommodate the new technology,

and which cannot be changed easily. Established firms are overwhelmed by the parallel need

· 29 ·

to develop new technological capabilities and a corresponding new activity system. Often,

incumbents try to accommodate new technologies within established ways of doing things, and

they only learn after developing technological skills that they ought to refit their activity system.

This idea is supported by Christensen and Rosenbloom (1995) and their analysis of the disk

drive industry, and Tripsas (1997, p. 139) who concludes her detailed study of the typesetter

industry: “The initial products developed by established firms were consistently inferior

to those of new entrants. The need for both new technical skills and new architectural

knowledge proved difficult for incumbents to manage.”

Besides a more detailed impression of what a firm might be or what a firm ought to do our

reasoning might shed some additional light on policy questions, as, for example, the regulation

and deregulation of business forms, like private limited companies, public companies, or trusts.

Another crucial policy question might be the treatment of horizontal and vertical mergers. Why

these policy question might become interesting is easy to explain (Rajan/Zingales 2000,

2001b). The cause is that we use a functional concept of the firm in which the allocation of

property rights is the device to manage a firm successful through an evolving industry. But the

allocation of property rights is also an essential legal concept to regulate business or industry,

e.g., jurisdictions provide special business forms for doing business, which are more or less

mandatory. So, we might face a clash of a functional concept of the firm with a concept that is

based purely legally. Or, put it differently, there might be a divergence between an economic

and a juridical concept of a firm.

Concrete business forms are “legal fictions” (Jensen/Meckling 1976) that solve some

economic problems pretty well but that are not permanently (re-)constructed and adapted by

law and economics scholars in order to fit the manifold needs of managers, shareholders,

creditors, and workers.19 From that divergence between legal forms and the functional

reasoning about the appropriate allocation of property rights in a firm results a permanent need

to adapt the legal forms that are in order. This does not mean that the pressure to adapt the

legal order may not change over time; so, it is possible that the need to adapt the legal order is

stronger in times of rapid technological change than in times of gradual technological change. A

19 The process of institutional adaptation is a crucial question that is beyond the scope of this

paper. For an overview see Vanberg and Kerber (1994), for a more narrow analysis of

European corporate law see Heine and Kerber (2002).

· 30 ·

striking example how a jurisdictional legal order might have to adapt is the case of antitrust

laws against vertical mergers in the media industry (Rajan/Zingales 2000). While a few years

ago content creation, production, and distribution were typically organized as single and

independent businesses in the value chain of movie production, nowadays giants like Time

Warner, Disney, Sony or Bertelsmann control more and more parts of the value chain. From a

puristic view of antitrust this concentration of activities is a severe problem of market power

(Harvard paradigm) or a simple problem of welfare economics (Chicago paradigm). From the

here presented view the vertical mergers in the media industry are in the first line an adjustment

of the appropriate boundaries of the firm that takes place as a strategic action to cope with

complementarities, incentives and appropriation of rents in a rapidly changing technological

environment. E.g., a movie does not only entertain customers in a film-theatre. Nowadays it is

launched together with a brand name under that a multitude of entertainment articles are sold

(toys, books, videos, CDs, clothes, food etc.). Because launching a strong brand name is very

costly and the duration of a brand in the entertainment industry is short, the industry faces a

time-cost trade-off that forces the industry to exploit brands vigorously and in a concerted

manner. Since the critical resource is the brand name and not the movie or the book it makes

sense to organize the whole value chain centrally from a headquarter that controls the brand

and all activities have to be centred around the brand, so that all generated rents can be caught

(Rajan/Zingales 2000). That is not to say that all vertical mergers are no problem at all, but it is

to say that antitrust policy has carefully to look where the appropriate boundaries of the firm

lie. Surely, this is no easy task, when it is assumed that the appropriate boundaries of a firm

alter in the evolution of an industry or are highly industry specific. But in our view it is highly

recommended to use such a concept that gives more weight to an elaborated theory of the

firm. An antitrust-policy that uses only broad concepts of market-power or welfare economics

and which forgets the strategic meaning of coordinating complementarities, incentives and

appropriation of rents in the evolution of industries seems to be more an obstacle to industry

evolution than a facilitator of evolution.

To sum up, we believe that a framework, which integrates the basic ingredients of

evolutionary thinking, like the crucial role of knowledge, firm heterogeneity, and market

processes, with recent thinking in the contract based theory of the firm, provides not only a

framework for a deeper understanding of industry evolution, but gives also normative hints for

· 31 ·

an accurate business policy as well as for a public policy that is to support industry. Besides

the results we may obtain from such an integrative framework it seems to us that the

resourceful combination of findings in evolutionary thinking and in neoclassical economics is a

promising scientific strategy to discover some missing links in the theory of the firm and

industry evolution.

· 32 ·

References

Abernathy, W.J./Utterback, J.M. (1978): Patterns of Industrial Innovation, in: Technology

Review, Vol. 80, 1978, pp. 40-47.

Agarwal, R. (1998): Evolutionary Trends in Industry Variables, in: International Journal of

Industrial Organization, Vol. 16, pp. 511-525.

Alchian, A.A. (1950): Uncertainty, Evolution, and Economic Theory, in: Journal of Political

Economy, Vol. 58, pp. 211-221.

Alchian, A.A./Demsetz, H. (1972): Production, Information Cost, and Economic Organization, in:

American Economic Review, Vol. 62, pp. 777-795.

Alchian, A.A./Woodward, S. (1987): Reflections on the Theory of the Firm, in: Journal of

Institutional Economics, Vol. 143, pp. 110-136.

Andrews, K. (1987): The Concept of Corporate Strategy, 3rd ed., Homewood.

Arora, A./Gambardella, A. (1990): Complementarity and External Linkages: the Strategies of the

Large Firms in Biotechnology, in: Journal of Industrial Economics, Vol. 4, pp. 361-379.

Bhide, A. (2000): The Origin and Evolution of New Businesses, Oxford.

Christensen, C.M./Rosenbloom, R.S. (1995): Explaining the Attacker’s Advantage:

Technological Paradigms, Organizational Dynamics, and the Value Network, in: Research

Policy, Vol. 24, pp. 233-257.

Clark, K./Fujimoto, T. (1991): Product Development Performance, Strategy, Organization, and

Management in the World Auto Industry, Boston (Ma.).

Coase, R. (1937): The Nature of the Firm, in: Economica, Vol. 4, pp. 386-405.

Cohendet, P./Llerena, P./Marengo, L. (2000): Is there a Pilot in the Evolutionary Firm, in: Foss,

N./Mahnke, V. (eds.): Competence, Governance, and Entrepreneurship, Oxford, pp. 95-

115.

Commons, J.R. (1934): Institutional Economics – Its Place in Political Economy, Madison.

Coriat, B./Dosi, G. (1999): Learning how to Govern and Learning how to solve Problems: On the

Co-Evolution of Competences, Conflicts and Organizational Routines, in: Dosi, G. (ed.),

Innovation, Organization and Economic Dynamics, Cheltenham, pp. 294-324.

Demsetz, H. (1967): Toward a Theory of Property Rights, in: American Economic Review, Vol.

57, p&p, pp. 347-359.

Demsetz, H. (1995): The Economics of the Business Firm, Cambridge (UK).

Demsetz, H. (1997): The Firm in Economic Theory: A Quiet Revolution, in: American Economic

Review, Vol. 87, p&p, pp. 426-429.

Dosi, G. (1988): Sources, Procedures, and Microeconomic Effects of Innovation, in: Journal of

Economic Literature, Vol. 26, pp. 1120-1171.

Dosi, G./Teece, D./Winter, S.G. (1992): Toward a Theory of Corporate Coherence, in: Dosi,

G./Giannetti, R./Toninelli, P. (eds.): Technology and Enterprise in a Historical Perspective,

Oxford, pp. 185-211.

Dosi, G./Malerba, F./Orsenigo, F./Malerba, L. (1994): Evolutionary Regimes and Industrial

Dynamics, in: Magnusson, L. (ed.): Evolutionary and Neo-Schumpeterian Approaches to

Economics, Dordrecht, pp. 203-229.

· 33 ·

Dosi, G./Nelson, R.R. (1994): An Introduction to Evolutionary Theories in Economics, in: Journal

of Evolutionary Economics, Vol. 4, pp. 153-172.

Foss, K./Foss, N. (2000): Economic Organization in a Process Perspective, in: Krafft, J. (ed.):

The Process of Competition, Cheltenham, pp. 27-47.

Foss, K./Foss, N. (2001a): Assets, Attributes and Ownership, in: International Journal of the

Economics of Business, Vol. 8, pp. 19-37.

Ghemawat, P./Levinthal, D. (2000): Choice Strcutures, Business Strategy and Perfomance: A

Generalized NK-Simulation Approach, Working Paper WP 00-05, Wharton School

University of Pennsylvania.

Grossman, S./Hart, O. (1986): The Costs and Benefits of Ownership: A Theory of Vertical and

Lateral Integration, in: Journal of Political Economy, Vol. 94, pp. 691-719.

Hart, O. (1991): Incomplete Contracts and the Theory of the Firm, in: Williamson, O./Winter,

S.G. (eds.): The Nature of the Firm, New York, Oxford, pp. 138-158.

Hart, O. (1995): Firms, Contracts, and Financial Structure, Oxford.

Hart, O./Moore, J (1990): Property Rights and the Nature of the Firm, in: Journal of Political

Economy, Vol. 98, pp. 1119-1158.

Hayek, F.A. (1937): Economics and Knowledge, in: Economica, Vol. 4, pp. 33-54.

Hayek, F.A. (1945): The Use of Knowledge in Society, Vol. 35, pp. 519-530.

Hayek, F.A. (1968): Der Wettbewerb als Entdeckungsverfahren, Kiel.

Heine, K./Kerber, W. (2002): European Corporate Laws, Regulatory Competition and Path

Dependence, in: European Journal of Law and Economics, Vol. 13, forthcoming.

Heine, K./Stieglitz, N. (2001): Zur Konvergenz von Corporate Governance Systemen: eine

innovationsökonomische Analyse, in: Gerum, E./Krag, J./Lingenfelder, M. (eds.):

Betriebswirtschaftliche Studien, Working Paper, No. 18.

Heuss, E. (1965): Allgemeine Markttheorie, Tuebingen et al.

Holmstrom, B./Milgrom, P. (1994): The Firm as an Incentive System, in: American Economic

Review, Vol. 84, pp. 972-991.

Jensen, M./Meckling, W.H. (1976): Theory of the Firm: Managerial Behavior, Agency Costs and

Ownership Structure, in: Journal of Financial Economics, Vol. 3, pp. 305-360.

Kauffmann, S. (1993): The Origin of Order, New York.

Kerber, W. (1997): Wettbewerb als Hypothesentest: eine evolutorische Konzeption

wissenschaffenden Wettbewerbs, in: Delhaes, K./Fehl, U. (eds.), Dimensionen des

Wettbewerbs, Stuttgart, pp. 29-78.

Kirzner, I.M. (1973): Competition and Entrepreneurship, Chicago.

Klepper, S. (1997): Industry Life Cycles, in: Industrial and Corporate Change, Vol. 6, pp. 145-

181.

Levinthal, D.A. (1997): Adaptation on Rugged Landscapes, in: Management Science, Vol. 43,

pp. 934-950.

March, J.G. (1991): Exploration and Exploitation in Organizational Learning, in: Organization

Science, Vol. 10, pp. 71-87.

Markides, C. C. (1995): Diversification, Refocusing, and Economic Performance, Cambridge

(Ma.).

Mayr, E. (1984): Die Entwicklung der biologischen Gedankenwelt, Berlin et al.

McGahan, A./Silverman, B.S. (2001): How does Innovative Activity Change as Industries

Mature?, in: International Journal of Industrial Organization, Vol. 19, pp. 1141-1160.

· 34 ·

Milgrom, P./Roberts, J. (1990): The Economics of Modern Manufacturing: Technology, Strategy,

and Organization, in: American Economic Review, Vol. 80, pp. 511-528.

Milgrom, P./Roberts, J. (1995): Complementarities and Fit: Strategy, Structure, and

Organizational Change in Manufacturing, in: Journal of Accounting and Economics, Vol.

19, pp. 179-208.

Milgrom, P./Qian, Y./Roberts, J. (1991): Complementarities, Momentum, and the Evolution of

Modern Manufacturing, in: American Economic Review, Vol. 81, p&p, pp. 84-88.

Montgomery, C.A. (1994): Corporate Diversification, in: Journal of Economic Perspectives, Vol.

8, pp. 163-178.

Mueller, D.C. (1997): First-mover Advantages and Path Dependence, in: International Journal of

Industrial Organization, Vol. 15, pp. 827-850.

Porter, M.E./Siggelkow, N. (2001): Contextuality within Activity Systems. Working Paper,

Wharton School.

Rajan, R.G./Zingales, L. (1998): Power in a Theory of the Firm, in: Quarterly Journal of

Economics, Vol. 113, pp. 387-432.

Rajan, R.G./Zingales, L. (2000): The Governance of the New Enterprise, in: Vives, X. (ed.):

Corporate Governance, Cambridge (UK), pp. 201-226.

Rajan, R.G./Zingales, L. (2001a): The Influence of the Financial Revolution on the Nature of

Firms, in: American Economic Review, Vol. 91, p&p, pp. 206-211.

Rajan, R.G./Zingales, L. (2001b): The Firm as a Dedicated Hierarchy: A Theory of the Origins

and Growth of Firms, in: Quarterly Journal of Economics, Vol. 116, pp. 805-851.

Rotemberg, J.J./Saloner, G. (1994): Benefits of Narrow Business Strategies, in: American

economic Review, Vol. 84, pp. 1330-1349.

Rothaermel, F.T. (2001): Complementary Assets, Strategic Alliances, and the Incumbent’s

Advantage: an Empirical Study of Industry and Firm Effects in the Biopharmaceutical

Industry, in: Research Policy, Vol. 30, pp. 1235-1251.

Schanze, E. (1998): Symbiotic Arrangements, in: Newman, P. (ed.): The New Palgrave of

Economics and the Law, Vol. 3, London, pp.554-559.

Schlicht, E. (1998): On Custom in the Economy, Oxford.

Schmitz, P.W. (2001): The Hold-Up Problem and Incomplete Contracts: A Survey of Recent

Topics in Contract Theory, in: Bulletin of Economic Research, Vol. 53, pp. 1-17.

Schmitz, P.W./Sliwka, D. (2001): On Synergies and Vertical Integration, in: International Journal

of Industrial Organization, Vol. 19, pp. 1281-1295.

Schreiter, C. (1994): Evolution und Wettbewerb von Organisationsstrukturen, Göttingen

Schumpeter, J. (1934): The Theory of Economic Development, Cambridge (Ma.).

Schumpeter, J. (1950): Capitalism, Socialism, and Democracy, New York

Shan, W./Walker, G./Kogut, B. (1994): Interfirm Cooperation and Startup Innovation in the

Biotechnology Industry, in: Strategic Management Journal, Vol. 15, pp. 387-394.

Teece, D.J. (1996): Firm Organization, Industrial Structure, and Technological Innovation, in:

Journal of Economic Behavior and Organization, Vol. 31, pp. 193-224.

Teece, D.J./Pisano, G. (1994): The Dynamic Capabilities of Firms: An Introduction, in: Industrial

and Corporate Change, Vol. 3, pp. 537-556.

Teece, D.J./Pisano, G./Shuen (1997): Dynamic Capabilities and Strategic Management, in:

Strategic Management Journal, Vol. 18, pp. 509-533.

Tripsas, M. (1997): Unraveling the Process of Creative Destruction: Complementary Assets and

Incumbent Survival in the Typesetter Industry, in: Strategic Management Journal, Vol. 18,

Summer Special Issue, pp. 119-142.

· 35 ·

Vanberg, V. (1995): John R. Commons: Institutionelle Evolution durch absichtsvolle Selektion, in:

Samuels, W.J./Biddle, J.E. (eds.): J.R. Commons‘ ‘Legal Foundations‘ und seine Beiträge

zum Institutionalismus, Duesseldorf, pp. 69-94.

Vanberg, V. (2001): Evolutorische Ökonomik: Homo Oeconomicus, Markt und Institutionen,

Freiburg Discussion Papers on Constitutional Economics, No. 01/4, Freiburg.

Vanberg, V./Kerber, W. (1994): Institutional Competition among Jurisdictions: An Evolutionary

Approach, in: Constitutional Political Economy, Vol. 5, pp. 193-219.

Vives, X. (1999): Oligopoly Pricing, Cambridge (Ma.).

Williamson, O.E. (2000): The New Institutional Economics: Taking Stock, Looking Ahead, in:

Journal of Economic Literature, Vol. 38, pp. 585-613.

Winter, S.G. (2000): The Satisficing Principle in Capability Learning, in: Strategic Management

Journal, Vol. 21, pp. 981-996.

Zingales, L. (2000): In Search of New Foundations, in: The Journal of Finance, Vol. LV, pp.

1623-1653.

ResearchGate has not been able to resolve any citations for this publication.

The Satisficing Principle in Capability Learning

Article

Full-text available

Oct 2000
STRATEGIC MANAGE J

Sidney Winter

Whether an organization has a certain capability is often a matter of degree. Thus, in the context of initial learning of a capability, there is generally no clear-cut or automatic answer to the question of when an organization should be expected to cut back its learning efforts and affirm that the desired capability has been achieved. This paper offers a simple conceptual model for this question, based on the satisficing principle. More specifically, the question addressed is: ‘When does overt learning stop?’—where ‘overt’ learning is understood as being marked by observable allocation of attention and resources to the task of acquiring the capability. The model provides the framework for a discussion of various influences on the aspiration level in the satisficing model, and hence on the nature of the capability that has been achieved when learning stops.Overt learning efforts may be resumed at some time later if external factors operate to lift aspiration levels relevant to the capability. The paper discusses how such ‘re-ignition’ of learning may occur as a result of an organizational crisis, or of the institution of a quality management program. Copyright © 2000 John Wiley & Sons, Ltd.

Wettbewerb als Hypothesentest: Eine evolutorische Konzeption wissenschaffenden Wettbewerbs: Seine Rolle in der Entstehung und Ausgestaltung von Wirtschaftsordnungen

Chapter

Jan 1997

Wolfgang Kerber

Corporate diversification

Article

Jan 1994
J ECON PERSPECT

C.A. Montgomery

Theory of the Firm: Managerial Behavior, Agency Costs, and Capital Structure

Article

Jan 1976
J FINANC ECON

Exploration and Exploitation in Organizational Learning

Article

Jan 1991

J.G. March

Toward a Theory of Propery Rights

Article

May 1967

Harold Demsetz

Product Development In The World Auto Industry: Strategy. Organization and Performance

Article

Apr 1994
DESIGN STUD

Erik Jan Hultink

Corporate Diversification

Article

Michael L. Katz

Unraveling the Process of Creative Destruction: Complementary Assets and Incumbent Survival in the Typesetter Industry

Article

Jul 1997
STRATEGIC MANAGE J

Mary Tripsas

When radical technological change transforms an industry established firms sometimes fail drastically and are displaced by new entrants, yet other times survive and prosper. Drawing upon an unusually rich data set that covers the technological and competitive history of the typesetter industry from 1886 to 1990, this paper uses a combination of quantitative and qualitative analysis to unravel this process of creative destruction. It argues that the ultimate commercial performance of incumbents vs. new entrants is driven by the balance and interaction of three factors: investment, technical capabilities, and appropriability through specialized complementary assets. In this industry, specialized complementary assets played a crucial role in buffering incumbents from the effects of competence destruction, and an analysis that examined investment or technical capabilities in isolation would have led to misleading results. This work thus highlights the importance of considering multiple perspectives when examining the competitive implications of technological change. © 1997 by John Wiley & Sons, Ltd.

Institutional Economics: Its Place in Political Economy

Article

Jan 1961

John R. Commons

Industry Evolution and the Strategic Role of Complementarities in the Theory of the Firm

Recommended publications

Extending Methods to Express Change Requirements

The evolution of global strategy

A Study on the Historical Evolution of Advertising Industry in Korea

EVOLUÇÃO E CONFIGURAÇÃO ATUAL DAS INDÚSTRIAS MOVELEIRAS MUNDIAL E BRASILEIRA