Annals of Business Administrative Science 15 (2016) 15–27

http://doi.org/10.7880/abas.0150810a
Received August 10, 2015; accepted September 2, 2015
Published in advance on gbrc.jp: November 22, 2015

Strategy and Structure Follow

Technology: A Spinout Proposition of
J. D.Thompson’s Organizations in Action
Nobuo TAKAHASHIa)

Abstract: This study reconfigures part one of Thompson (1967) as

a theoretical restatement of Chandler’s (1962) historical evidence.
When organizations grow, their growth orientations and strategies
emerge from their technical rationality, according to Thompson’s
first criteria of technology instrumentality. Regarding
instrumentally reasonable/rational organizations, according to his
second criteria of economy, organizational structures such as
horizontal departmentalization, vertical hierarchies, and
multidivisional forms become necessary to minimize coordination
costs. In other words, when discussing growth strategies and
multidivisional forms, Chandler claimed that “structure follows
strategy,” but Thompson rightfully claimed that “strategy and
structure follow technology.”

Keywords: technical core, technical rationality, multidivisional

forms, assessment, structure follows strategy, Japan Quality
Award

a)
Graduate School of Economics, University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo, Japan,
[email protected]
A part of this paper was originally published as Takahashi (2013) in Japanese.
©2016 Global Business Research Center. This is an Open Access article distributed under the terms of
the Creative Commons Attribution License, which permits unrestricted reuse, distribution, and
reproduction in any medium, provided the original work is properly cited.
 15
Takahashi

Introduction

Thompson (1967) has taken much interest in his unique concepts,

such as the technical core (Shimizu, 2010). However, his book’s
overall assertions are difficult to understand, and not much focus
has been placed on them. The book comprises two parts, “part one”
and “part two,” without any titles. This study reconfigures “part one”
in its entirety as a theoretical study of Strategy and Structure
(Chandler, 1962). 1 Chandler advocated the proposition that
“structure follows strategy” when discussing growth strategy and
multidivisional forms, but Thompson rightfully proposed that
“strategy and structure follow technology” in essence.

Technical Core

People find organizations in non-random, planned, and

reasonable/rational actions. Organizations in Action is the
appropriate title. This pioneering idea by Thompson is also linked
with Weick’s (1979) notion of organizing (Takahashi, 2009). In other
words, Thompson’s unique notion of the organization is “an open
system subject to the criteria of rationality” (Thompson, 1967, p.
11).
The non-random and planned actions are based on technologies.
Technical rationality is evaluated according to (a) instrumental
criteria and (b) economic criteria, where (b) economic
1 Almost entirely independent of this, “part two” develops an administrative
theory based on technology as the key concept. Its motif is a
redevelopment of administrative theory as attempted by Simon (1947) and
Cyert and March (1963) in the Simon–March–Cyert stream of study
(Thompson, 1967,
p. 9), which considered technology as its foundation (Takahashi, 2013).
From a different angle, what is attempted in part one (especially
 Thompson, 1967, pp. 28–35) can also be considered a summary of the
problems involved in administrative theory (Takahashi, 2008, 2015b) as
critiqued by Simon (1947, chap. 2), particularly “organization by purpose,
process, clientele, place.”
16
Strategy and structure follow technology

criteria are considered after (a) instrumental criteria have been met.
Since (a) instrumental criteria are primarily important (Thompson,
1967, pp. 14–15), we first consider (a) instrumental rationality.
A purposive organization must have core technologies for
achieving goals at its core. The existence of these core technologies
generates a type of homeostasis or self-stabilization. Complex
organizations are formed to operate technologies that would be
impossible for each individual to operate (Thompson, 1967, p. 15).
Open systems that follow these rational criteria have three levels,
as noted by Parsons (1960, chap. 2): technical, managerial, and
institutional. At the institutional level, open systems exist as a part
of broader social systems; however, at the technical level—which is
the lowest—technical rationality is pursued as a closed system by
eliminating uncertain variables from the technical core as much as
possible (Thompson, 1967, p. 11).

Direction of Expansion of Organizations

An organization usually has only some core technologies within its
domain and never possess all core technologies. Thus, organizations
are not self-sufficient (Thompson, 1967, p. 26). Simultaneously with
these deficiencies, an organization also has excess unused
resources. As an organization grows, the direction of growth is not
random. Excesses and deficiencies orientate organization’s growth
or growth strategy.
Three dimensions of domain

The following three dimensions determine organizational domain

(Thompson, 1967, p. 40).2
2 Thompson wrote that these three dimensions appear “in examining the
17
Takahashi

1. Technology included
2. Population served
3. Services rendered
Each of these three dimensions of organizational domain
correspond to the following examples of instrumental criteria for
technology (Thompson, 1967, pp. 15–18).3
1. Serial interdependence, as conceived by scientific
management4 2. Standardization, as conceived in a bureaucracy
3. Specialization, as conceived in administrative theory
An organization grows to set organizational boundaries around
activities that would be crucial contingencies (Proposition 4.1).5 This
provides the following subpropositions.
1. When continuous reciprocal interdependence is high, a domain

concept of organizational domain” (Thompson, 1967, p. 40), although in a

corresponding section (Thompson, 1967, p. 26), he cites Levine and White
(1961) in saying that while 2 and 3 are identical, 1 is “diseases covered”
or, as Thompson stated, a “range of products” and not technology.
3 Thompson suggests three types of technologies: (1) long-linked technology,
(2) mediating technology, and (3) intensive technology and provides an
example of each (Thompson, 1967, pp. 15–18). In fact, however, one case
can fit any of these types, so it fails as a typology. For example, a general
hospital is given as an example of (3) intensive technology (Thompson,
1967, p. 17). However, each individual professional uses a broad range of
standardized technologies (2). Moreover, no hospital will administer
injections without checking the patient, examining the patient, and
making a diagnosis; thus, the actions taken there have serial dependence
(1). In fact, Thompson gives Bell’s Telephone System as an example of a
combination of (1), (2), and (3) (Thompson, 1967, p. 44), and in that
sense, the three types are not typology at all, but rather they should be
described as three factors of technology or instrumental criteria.
4 A relation in which action B is executed only after action A is successfully
 completed, action C is executed only after action B is successfully
completed, and so forth.
5 A contingency is one source of uncertainty outside an organization. An
accurate definition is given in Chapter 12 (last chapter) of the book by
Thompson (1967).
18
Strategy and structure follow technology

expands such that the technology in it expands through vertical

integration (Proposition 4.1a).
2. When standardization occurs, a domain expands by increasing
the population that is receiving a service (Proposition 4.1b). 3.
When specialization occurs, a domain expands by taking on
clients6 (Proposition 4.1c).
In other words, an organization grows along the three dimensions
of organizational domain corresponding with the instrumental
criteria of technology, thereby leading to a clear orientation in
organizational growth and emergence of growth strategy.
Balancing components

Resources are not necessarily continuously divisible and can only

be obtained in definite units (Thompson, 1967, p. 45).
Consequently, unused portions will always remain in some
components of the organization, and the organization will try to
grow (Proposition 4.2) until they almost completely use the unused
portion that is most difficult to reduce. Thompson refers to this as
the balancing of components.7 The result of this balancing is that
the organization expands its organizational domain to match its
capacity whenever they have excess capacity (Proposition 4.3).
Power

An orientation in organizational growth will result from the

aforementioned excesses or deficiencies, but when the firm8 has
6 Proposition 4.1c uses the phrase “the object worked on,” but the example
given (Thompson, 1967, pp. 43–44) is related to clients.
7 Although Thompson does not mention it, Penrose (1959, pp. 68–72) makes
the same assertion (Takahashi, 2002, 2015c).
8 Thompson’s original wording was “organization,” although it should be
more correctly “firm” in this subsection, since legally distinct firms in a
power relation will function as one organization and should therefore be
viewed as a “transfirm organization” (Takahashi, 2014).
19
Takahashi

power over the others who control the required activities, there is no
need to formally incorporate those activities by growing (Thompson,
1967, p. 48).
If a firm has the ability to take action without considering the
actions of its competitors, that firm will have power over its
competitors (Thompson, 1967, p. 31). Dependency and power are
two sides of the same coin (Emerson, 1962); thus, firms will
maintain alternative means to minimize the power of others
(Thompson, 1967, Proposition 3.1), or they will try to obtain
prestige (Thompson, 1967, Proposition 3.2), enter into contracts
(Thompson, 1967, Proposition 3.3a), co-opt executives (Thompson,
1967, Proposition 3.3b), or coalesce through combinations or joint
ventures (Thompson, 1967, Proposition 3.3c). These arguments are
becoming conventional wisdom, being almost the same as those of
“resource dependence perspective” made later by Pfeffer and
Salancik (1978).9

Decentralized Division
Next, if an instrumentally rational/reasonable organization
attempts to improve its efficiency based on economic criteria, a
horizontal departmentalization and vertical hierarchy as well as a
multidivisional form of organizational structure are required to
minimize coordination costs.
Departmentalization and hierarchy

An organization that is attempting to increase efficiency will

group its positions to minimize coordination costs through
horizontal

9 However, although Pfeffer and Salancik (1978) are often cited, they are
mentioned as metaphors. Research that expands and validates the
resource dependence perspective is limited, and the idea of resource
dependence perspective has been broadly accepted without rigorous
testing (Pfeffer, 2003).
20
Strategy and structure follow technology

departmentalization and vertical hierarchies (Thompson, 1967,

Proposition 5.1).
First, three types of internal interdependence exist (Thompson,
1967, pp. 54–55).10
(A) Pooled interdependence, where no direct relation or contact
exists, but the failure of one component puts the entire
organization at risk and threatens other components
(B) Sequential interdependence, where a one-way, direct relation
exists
(C) Reciprocal interdependence, where a bi-directional, direct
relation exists
In managing internal interdependence, coordination costs are
minimized by grouping and localizing reciprocally interdependent
positions to be tangential to each other in the case of (C) (Thompson,
1967, Proposition 5.1a), or grouping and localizing sequentially
interdependent positions autonomous under the condition that they
adhere to plans and standards in the case of (B) (Thompson, 1967,
Proposition 5.1b). In the case of (A), the organization groups
positions will be homogeneous within each group (Thompson, 1967,
Proposition 5.1c). This is departmentalization.
However, a hierarchical structure is created when
interdependence cannot be confined within a group. In other words,
a group with reciprocal interdependence will seek to form a
second-order group or cluster (Thompson, 1967, Proposition 5.2)
and will rely on a task force or project organization for coordination
(Thompson, 1967, Proposition 5.4d). A group with sequential
interdependence will form tangential clusters (Thompson, 1967,
Proposition 5.3) and rely on a committee for coordination
(Thompson, 1967, Proposition 5.4c). For

10 Thompson asserts a containment relation, where A ⊃ B ⊃ C, although

actually one can select either B or C, where they do not overlap each
other.
21
Takahashi

homogeneous positions or groups, organization will blanket them

under rules that cross divisional lines (Thompson, 1967, Proposition
5.4a), and liaisons will be assigned to manage rule-making agencies
for standardization. (Thompson, 1967, Proposition 5.4b).
Decentralized division

The technical core always should be appropriately geared to both

input and output activities (Thompson, 1967, p. 19). Therefore, an
organization will create the following type of self-control situation for
the technical core to minimize coordination costs.
1. An organization will attempt to seal off core technologies from
environmental influences (Thompson, 1967, Proposition 2.1). 2.
Thus, the organization will surround the technical core with
input and output components to buffer it from environmental
influences11 (Thompson, 1967, Proposition 2.2).
3. Furthermore, the organization influences the environment to
smoothen and level input and output transactions (Thompson,
1967, Proposition 2.3).
4. The organization will also attempt to anticipate and adapt to
environmental changes that cannot be buffered or smoothed
(Thompson, 1967, Proposition 2.4).
 5. If that also fails, the organization will resort to rationing 12
11 Buffers are created in response to quantitative changes, although
standardization in response to qualitative changes, as discussed by
March and Simon (1958), should inherently be established. Preventative
maintenance is also classified as a buffer (Thompson, 1967, p. 20). When
preventative maintenance is neglected, however, machinery will stop
working for long periods of time due to sudden malfunctions, resulting in
great losses. Therefore, categorizing it as a buffer is a mistake since it
does not take much time to smooth and level the downtime of machinery
with, for instance, weekly inspections.
12 Rationing is the securing of high-priority activities and functions in
emergencies at the expense of other activities and functions (Thompson,
1967, p. 23). An example of this is the “safe mode” of a personal
computer;
22
Strategy and structure follow technology

(Thompson, 1967, Proposition 2.5).

An organizational structure does not become complicated only
because it increases in scale (Thompson, 1967, p. 74). If the
technical core can be isolated from boundary-spanning activities
except for scheduling, then it is reasonable to have a centralized
organization comprising functional divisions (Thompson, 1967,
Proposition 6.3). In contrast, when reciprocal interdependencies
exist between major components of an organization, segmentation
will occur; to minimize coordination costs, the components will
form self-sufficient clusters, such as “decentralized divisions,” and
each cluster will have its own domain (Thompson, 1967,
Proposition 6.4). 13 Thus, Thompson uses technology to explain the
multidivisional form proposed by Chandler (1962).14

Implications of the Japan Quality Award

Thompson’s idea of technical rationality is based on rational
criteria and enables the following types of assessments:
(A) When standards of desirability are crystallized, if belief15 in a
 cause/effect relation is complete, an efficiency test is conducted
to determine the level of attainment toward a state of 100%
desirability (Thompson, 1967, p. 86).

when a computer freezes up due to a problem caused by a bug, the

minimum level of device drivers needed to diagnose the issue (such as the
keyboard driver) will operate, and other functions will be unavailable.
13 The first half of Chapter 6 of the book by Thompson (1967) also discusses
the “task environment” and its two dimensions of degree of homogeneity
and degree of stability, which were proposed by Dill (1958).
14 However, the case of GM in the study by Chandler (1962) is rather one of
centralization from a holding company to a multidivisional form, and
Thompson does not explain this.
15 Instrumental action is linked to desired outcomes and beliefs regarding
cause-effect relations (Thompson, 1967, p. 14).
23
Takahashi

(B) When belief in a cause/effect relation is incomplete, an

assessor will conduct an instrumental test (Thompson, 1967, p.
86). (C) When standards of desirability are ambiguous, a
reference group is set, and a social test is conducted and
compared with the reference group (Thompson, 1967, pp. 86–87).
The Japan Quality Award (JQA) Council, which has granted its
namesake award since 1995, also recommends conducting similar
self-assessments.16 Specifically, results-based numbers are used for
(A). In the case of (B), if desirable results are known, an assessor will
analyze the key success factors involved. (C) is known as
benchmarking. Thompson notes that instrumental test (B) is
preferable to social test (C) and that efficiency test (A) is preferable
to instrumental test (B) (Thompson, 1967, Proposition 7.1).
However, the JQA states that rather than a preference, the
assessment should be conducted as a sequence, with
benchmarking (C) done first, followed by the analysis of key success
factors (B), and finally by results-based numbers (A). This yields a
historical improvement in the results-based numbers (Thompson,
1967, Proposition 7.2a), and benchmarking is performed to identify
good and bad parts in comparison with the reference group
(Thompson, 1967, Proposition 7.2b). When an organization has no
assessment competence, it can rely on an external evaluation
(Thompson, 1967, Propositions 7.5, 7.5a, and 7.5b). However,
self-assessment is fundamental to the assessment process.

16 Organizational activation has been a topic of discussion in Japan’s

business community post the oil crisis of the 1970s (Takahashi, 1992).
JQA’s self-assessment has been proven by many award-winning
companies in the past 20 years to have an impact on business innovation
in companies that are gradually declining and need organizational
activation (Takahashi, 2015a).
24
Strategy and structure follow technology

Conclusion

As shown above, part one of Thompson’s (1967) book is

reconfigured as follows. Growth orientations or growth strategies
emerge from technical rationality evaluated by instrumental criteria
as organizations grow. For instrumentally reasonable/rational
organizations, a consideration of economic criteria reveals that
organizational structures such as horizontal departmentalization,
vertical hierarchies, and multidivisional forms are necessary to
minimize coordination costs. In other words, when discussing
growth strategy and a multidivisional structure, business historian
Chandler advocated that “structure follows strategy,” although a
logical consideration by Thompson reveals that “strategy and
structure follow technology.” Any correlation between strategy and
structure is a spurious one.17
Thompson uses the term “technology,” which leads to
misunderstanding because of its limited application, but that was
not originally the case. For example, the JQA assessment has in fact
been implemented across all industries. Thompson’s “technology”
and technical rationality is not limited to manufacturing but applies
to all industries, and the idea that “strategy and structure follow
technology” is a universal proposition.

Acknowledgements
This work was supported by JSPS KAKENHI Grant Number 26380454
for FY 2014–2018.

17 Chandler (1962, Introduction) actually stated that organization building

follows growth strategy and thus “structure follows strategy,” but this is
often given in reverse order, with organization building coming first
(Mizuno, 2013).
25
Takahashi

References
Chandler, A. D. (1962). Strategy and structure: Chapters in the history of
the industrial enterprise. Cambridge, MA: MIT Press.
Cyert, R. M., & March, J. G. (1963). A behavioral theory of the firm.
Englewood Cliffs, NJ: Prentice-Hall.
Dill, W. R. (1958). Environment as an influence on managerial
autonomy. Administrative Science Quarterly, 2, 409–443. Emerson, R. M.
(1962). Power-dependence relations. American Sociological Review, 27,
31–40.
Levine, S., & White, P. E. (1961). Exchange as a conceptual framework for
the study of interorganizational relationships. Administrative Science
Quarterly, 5, 583–601.
March, J. G., & Simon, H. A. (1958). Organizations. New York, NY: Wiley.
Mizuno, Y. (2013). Make provision for future growth under adverse
circumstances. Annals of Business Administrative Science, 12, 311–326.
doi: 10.7880/abas.12.311
Parsons, T. (1960). Structure and process in modern societies. New York,
NY: Free Press.
Penrose, E. T. (1959). The theory of the growth of the firm. Oxford, UK:
 Basil Blackwell.
Pfeffer, J. (2003). Introduction to the classic edition. In J. Pfeffer & G. R.
Salancik, The external control of organizations: A resource
dependence perspective (pp. xi–xxx). Stanford, CA: Stanford
University Press.
Pfeffer, J., & Salancik, G. R. (1978). The external control of organizations:
A resource dependence perspective. New York, NY: Harper & Row.
Shimizu, T. (2010). Soshiki wa kankyohendou ni ikani taiousurunoka?
[How does an organization cope with environmental change?: Technical
notes on Thompson (1967)]. Akamon Management Review, 9(9), 663–678
(in Japanese).
Simon, H. A. (1947). Administrative behavior: A study of decision-making
processes in administrative organization. New York, NY: Macmillan.
Takahashi, N. (1992). An evaluation of organizational activation. Omega,
20(2), 149–159.
26
Strategy and structure follow technology

Takahashi, N. (2002). Penrose Kaisha seichou no riron wo yomu [On

Penrose’s The theory of the growth of the firm]. Akamon Management
Review, 1(1), 105–124 (in Japanese).
Takahashi, N. (2008). Gentei sareta gourisei wa dokoni [Where is
“bounded rationality”?: Technical notes on Simon (1947, 1957, 1976,
1997)]. Akamon Management Review, 7(9), 687–706 (in Japanese).
Takahashi, N. (2009). Soshikika towa nanika? [What is organizing?:
Technical notes on Weick (1979)]. Akamon Management Review, 8(5),
233–262 (in Japanese).
Takahashi, N. (2013). Randamu dewanai koi nonakani soshiki o miidasu
[Finding organizations in nonrandom action: Technical notes on
Thompson (1967)]. Akamon Management Review, 12(4), 327–348 (in
Japanese).
Takahashi, N. (2014). Transfirm organization view. Annals of Business
Administrative Science, 13, 31–46. doi: 10.7880/abas.13.31 Takahashi,
N. (2015a). Behind the shell: Rigid persons clung onto it. Annals of
Business Administrative Science, 14, 1–14. doi: 10.7880/abas.14.1
Takahashi, N. (2015b). Where is bounded rationality from? Annals of
Business Administrative Science, 14, 67–82. doi: 10.7880/abas.14.67
Takahashi, N. (2015c). An essential service in Penrose’s economies of
growth. Annals of Business Administrative Science, 14, 127–135. doi:
10.7880/abas.14.127
Thompson, J. D. (1967). Organizations in action: Social science bases of
administrative theory. New York, NY: McGraw-Hill.
Weick, K. E. (1979). The social psychology of organizing (2nd ed.).
Reading, MA: Addison-Wesley.

27
Copyright of Annals of Business Administrative Science
is the property of Global Business Research Center and
its content may not be copied or emailed to multiple sites
or posted to a listserv without the copyright holder's
express written permission. However, users may print,
download, or email articles for individual use.