ArticlePDF Available

Organizational size and efficiency: An information processing view

Authors:

Abstract

The relationship between organizational size and performance has long been of interest. Size, however, has been defined and measured in global ways, resulting in few (if any) unequivocal findings. Three specific dimensions of size--as measured by the number of headquarter resources and policies, the number of products or services, and the number of resources and services of subunits -- were analysed in relation to organizational performance efficiency in a computer simulation. Analyses of simulation results suggest that some dimensions of size, particularly the number of headquarter resources and policies, are related significantly to performance efficiency.
MANUSCRIPT
How controls influence organizational information
processing: insights from a computational modeling
investigation
Chris P. Long
1
Sim B. Sitkin
2
Laura B. Cardinal
3
Richard M. Burton
2
Published online: 15 August 2015
Springer Science+Business Media New York 2015
Abstract In this study, we use a series of computational models to investigate an
information processing perspective on organizational control use. We evaluate and
compare the information processing capabilities of various formal and informal
control configurations under different information uncertainty conditions. We find
that a wide range of formal controls can be used to direct subordinates performing
interdependent tasks while a more narrow range of informal controls are most
effective for directing subordinates who perform complex tasks. Results of this
study provide a basis for formalizing an information processing perspective on
organizational control implementation that differs but is complementary to the
current emphasis on agency in organizational control research.
Keywords Organizational control Information processing Computational
modeling
&Chris P. Long
cpl32@georgetown.edu
Sim B. Sitkin
sim.sitkin@duke.edu
Laura B. Cardinal
lbcardinal@bauer.uh.edu
Richard M. Burton
rmb2@mail.duke.edu
1
McDonough School of Business, Georgetown University, Washington, DC 20007, USA
2
Fuqua School of Business, Duke University, Durham, NC 27708, USA
3
C. T. Bauer College of Business, University of Houston, Houston, TX 77204, USA
123
Comput Math Organ Theory (2015) 21:406–436
DOI 10.1007/s10588-015-9191-z
1 Introduction
From the earliest management research, implementing control has been recognized
as one of the four fundamental functions of management (e.g., Fayol 1949
identified: controlling, organizing, planning, and coordinating). This is because
organizational controls are the primary mechanisms that managers use to direct
attention, motivate, and encourage organizational members to act in ways to meet
organizational objectives (Kirsch 1996; Ouchi 1977,1979). Despite the scholarly
legacy, topical significance, and practical necessity of research on organizational
controls, in the years since Oliver (1998: Editor’s Letter) observed that ‘‘the study of
organizational control has a long history in administrative science and yet the need
to examine the processes and implications of this phenomenon has never been
greater,’’ researchers have continued to be challenged with understanding the
motivations for and components of various types of organizational controls
(Cardinal et al. 2011).
While researchers have acknowledged the role that organizational controls play
in facilitating organizational information processing (Galbraith 1973; Ouchi 1979;
Makhija and Ganesh 1997), scholars continue to maintain a relatively more limited
understanding of how managers may use controls to process production data within
organizational units. We believe, however, that comprehending these dynamics are
critical since managers currently operate in complex, interdependent environments
where they and their subordinates are charged with processing increasing amounts
of information. Managers’ capacities to effectively manage that information serves
as a primary source of an organization’s competitive advantage and largely
determines how well an organization innovates and adapts (Hagel and Brown 2007;
Turner and Makhija 2006).
To advance our understanding of the role that organizational controls play in
these activities, we build here on the principles of information processing theory
(Galbraith 1973,1977) to increase our understanding of the information processing
perspective on organizational control implementation. This perspective examines
how managers may use organizational controls to facilitate the transfer and
processing of production information. Managers specifically focus their efforts here
on ensuring that their subordinates possess information when and where they need it
to effectively perform their job tasks (Tushman and Nadler 1978).
We focus on the information processing perspective in this paper because
understanding how information processing concerns influence managers’ applica-
tions of controls has lagged well behind perspectives examining how agency
concerns influence managers’ control-related decisions. Currently, the agency
perspective on control implementation constitutes arguably the most influential of
all control research. The agency perspective builds from agency principles to argue
that managers attempt to promote production efficiencies by applying controls that
limit the relational risks that emerge from subordinate opportunism (Ouchi 1979).
While the findings presented by the agency perspective on organizational control
have been influential for both academics (e.g., Levinthal 1988) and practitioners
(e.g., Ghoshal and Moran 1996), we concur with scholars who have argued that the
How controls influence organizational information processing407
123
agency perspective provides a picture of managerial control use that is incomplete
in important ways (Long 2010; Lubatkin et al. 2007). In this paper, we specifically
focus on how the agency perspective’s emphasis regarding how managers use
controls to minimize their relational risks deemphasizes the important role that
controls play in facilitating information transfers between managers and their
employees.
Because information processing concerns comprise an increasingly salient
component of managerial decision-making, we contend that it is important for
scholars to understand how these issues affect managerial control implementation
(Leifer and Mills 1996; Simons 1995). To facilitate this, we focus our study on
investigating how managers may use organizational controls to direct superior-
subordinate information exchanges across a range of information processing
demands. Using this approach, we develop a perspective on control that is distinct
but complementary to the one presented by the agency perspective. As a result, our
study provides a basis for understanding how managers integrate agency and
information processing concerns in ways that extend previous control research and
present a more complete picture of control use in organizations.
Due to the complex confluence of control configurations and task factors that we
examine, we have chosen to employ a series of computational models. In the
tradition of Cohen et al. (1972), March (1991), and Harrison and Carroll (1991), we
use existing theory to model the information processing demands that managers
encounter. We then measure the capacity of various configurations of controls to
facilitate information processing within those conditions. By examining the patterns
of results that we obtain, we increase our understanding of how managers may use
organizational controls to facilitate effective information processing within their
organizational units.
We commence our discussion by distinguishing the core elements of the
information processing perspective that we present here from the one described by
the agency perspective on organizational control. We then discuss how the
information processing perspective provides a view of control use that is distinct
from but complementary to the findings presented by the agency perspective.
Thereafter, we describe our study, our computational model and our results. In the
final section of the paper, we examine the implications of our study for
organizational control research.
2 Theory
The information processing perspective highlights how managers may use controls
to help ensure that their subordinates get the correct production information when
and where they need it and can most effectively use it to perform their job tasks
(Makhija and Ganesh 1997; Tushman and Nadler 1978). Building from the general
premise of information processing theory (Galbraith 1973), the information
processing perspective develops the idea that controls serve as key mechanisms
that managers can use to direct task information to their subordinates. In this
capacity, managers use controls to both decrease their subordinates’ task-related
408 C. P. Long et al.
123
uncertainty and to increase the ability of those subordinates to achieve prescribed
production standards (Tushman and Nadler 1978).
Scholars’ efforts to develop this perspective on control to date have focused
primarily on examining the information processing capacities of organization-level
structures and information systems (Daft and Lengel 1986; Tushman and Nadler
1978). While research directed at this level of analysis is important, managers can
often do relatively little to alter the core components of the more overarching
infrastructures within which they work. Managers often do, however, maintain
significantly greater discretion over the controls they apply to their subordinates’ job
tasks (Simons 1995). Because of this and because these types of controls differ from
organization-level coordination and control mechanisms in both their composition
and the effects they produce (Ouchi 1979), we specifically examine here how
information processing concerns impact the ways in which managers apply controls
to their subordinates’ production efforts.
Consistent with previous research (Ouchi 1977; Perrow 1965; Thompson 1967),
we distinguish controls by the portion of the production process to which they are
applied: the inputs, processes and outputs of organizational production tasks
(Cardinal 2001). Managers use input controls to direct how material and human
elements of their production processes are qualified, chosen, and prepared through
selection, training, and/or socialization (e.g., Van Maanen and Schein 1979).
Managers use process controls such as production rules and behavioral norms to
ensure that individuals perform actions in a specific manner. Managers also use
output controls when they prescribe and monitor quality and/or quantity outcome
standards (e.g., customer satisfaction levels, production schedules) (e.g., Ouchi
1977,1979; Kirsch 1996,1997).
2.1 The agency perspective
Although we contend that information processing concerns influence managers’
applications of controls in important ways, we acknowledge that scholars’ efforts
to develop an understanding of these dynamics have lagged well behind research
that has primarily studied how issues of agency impact organizational control
implementation. The agency perspective on control described in this research
builds from agency theory principles and argues that managers attempt to promote
production efficiencies by applying controls to mitigate the relational risks that
emerge from subordinate opportunism (Kirsch 1996; Ouchi 1979). This perspec-
tive’s robust findings have been applied in a variety of contexts from traditional
manufacturing (Eisenhardt 1985) to IT system implementation (Kirsch 1996,
1997).
The agency perspective specifically focuses on how managers attempt to
minimize their relational risk exposure by aligning the controls they apply with their
perceived capacity to measure and monitor subordinate task efforts. Managers
evaluate these factors to determine how to develop input, process or output-based
contracts that effectively motivate subordinates while efficiently mitigating their
relational risk exposure.
How controls influence organizational information processing409
123
2.2 The information processing perspective
Despite the prevalence of the agency perspective in organizational control research,
we contend that the relative emphasis this perspective places only on what
information the manager needs to control tasks versus what information their
subordinates actually need to accomplish those tasks leads the agency perspective to
provide an incomplete view of control use in organizations. We argue that if
managers just follow agency prescriptions and apply controls only at points in the
production process where their relational risk exposure is minimized, they may fail
to deliver key pieces of production information when and where that information
can most effectively be used in their subordinates’ task efforts (Long et al. 2002).
The information processing perspective, however, presents a different view that
highlights how controls serve as mechanisms by which managers can exchange
important task information with their subordinates (Leifer and Mills 1996; Makhija
and Ganesh 1997). According to this perspective, managers use controls to both
gather information about their subordinates’ capacities to achieve production
standards and direct task information to their subordinates in ways that guide those
subordinates to more effectively fulfill their task responsibilities. This perspective
conceptualizes the focus of managers’ efforts to prepare, direct, and review their
subordinates’ production efforts as ways of enhancing those subordinates’ capacities
to understand task-related issues and improve their ability to achieve desired task
outcomes (Galbraith 1973; Ouchi 1979).
Thus, while the information processing perspective assumes antecedent condi-
tions similar to those outlined by the agency perspective (i.e., that individuals within
organizations often maintain different goals and priorities), the information
processing perspective describes a very different set of managerial responses to
these conditions. To summarize the differences between these two perspectives and
to help motivate our discussion going forward, we summarize the key contrasts
between the agency and the information processing perspectives on organizational
control in Table 1.
Table 1 Comparison of agency and information processing perspectives on control
Issue Agency perspective Information processing perspective
Primary motivation for control
use
Manage relational risks Reduce information uncertainty
Ongoing managerial concern Subordinate opportunism Impeded information flows
Purpose of controls Measure and monitor
subordinate activities
Facilitate effective information
sharing
Focal controls used Singular controls Singular and configurations of
multiple controls
Superior-subordinate
information exchange
Limited Extensive
Desired outcome Alignment of incentives Increase subordinates’ task
knowledge
410 C. P. Long et al.
123
According to information processing theory, managers are driven to consider
information processing issues as they become concerned about the level of
information uncertainty that their subordinates experience. Consistent with the
definition presented by Galbraith (1973,1977), information uncertainty describes
the difference between the information that subordinates require to perform their
tasks and the information that those individuals already possess.
Managers attempt to address their concerns about information uncertainty by
applying controls in ways that increase the quality and quantity of task information
that their subordinates receive. In other words, managers use controls to reduce
subordinates’ levels of information uncertainty by delivering needed task informa-
tion to their subordinates when and where they can most effectively use that
information to complete their production tasks (Galbraith 1973; Tushman and
Nadler 1978).
2.3 Investigating the information processing perspective on organizational
control
Because not all control configurations are equally effective for every situation, the
information processing perspective seeks to identify which configurations of
controls best enable managers to deliver needed task information to their
subordinates under specific information uncertainty conditions. Thus, our capacity
to develop the information processing perspective on organizational control use
rests on correctly identifying what controls managers need to apply to effectively
Task
Complexity
Hi/Lo
Task
Interdependence
Hi/Lo
Information
Uncertainty
Control Context
Formal
Informal
Information
Quality
(“Richness”)
“Fit”
Configurations of:
Input Controls
Process Controls
Output Controls
Timing of Information
Transfer
Fig. 1 Conceptual model of research question: determining the ‘‘fit’’ of organization control
configurations with various information uncertainty conditions
How controls influence organizational information processing411
123
direct information flows in their units in ways that effectively address their
subordinates’ information uncertainty problems (Tushman and Nadler 1978). To
accomplish this objective, we build on previous information processing research and
argue that it is important to evaluate which task controls provide the best ‘‘fit’’ for
key information uncertainty conditions that managers encounter (Tushman and
Nadler 1978). The components of the information processing perspective that we
describe above and we examine in this study are outlined in Fig. 1.
Figure 1displays the challenge that managers face in choosing configurations of
organizational controls that most effectively direct needed task information to
subordinates within their organizational units. In our study we attempt to identify
which configurations of controls provide the best fit across a range of information
processing demands. In our discussion below we both describe the factors that
comprise the information processing context within which managers operate and the
elements of the control-based decisions that managers must make.
2.4 The components of ‘‘fit’’ in the information processing perspective
When attempting to establish a ‘‘fit’’ between their context and the controls they
apply, we argue that managers balance three factors: the level of information
uncertainty that they and their subordinates encounter, the timing of their control
applications (i.e., of information transfers), and the quality of the information that
they seek to transfer. Building from information processing theory principles, we
argue that the controls that provide the best fit are those that most readily enable
managers to reduce the levels of information uncertainty that they and their
subordinates encounter (Tushman and Nadler 1978). Consistent with the prior work
of several scholars (e.g., Galbraith 1973; Tushman 1978; Tushman and Nadler
1978), we argue that task complexity and task interdependence jointly and
independently affect the levels of information uncertainty that subordinates
experience.
2.4.1 Task complexity
Complex tasks possess unknown or uncertain alternatives, inexact means-ends
connections, and, often, a series of subtasks which are unable to easily be integrated
with a primary task. Research on task complexity suggests that more complex tasks
both require and stimulate increased information exchange because individuals who
work on complex tasks often must rely on information that others provide them
(Tushman 1978; Van de Ven et al. 1976). While information exchanges in these
units can lead to higher performance, Galbraith (1973) argues that the capacity of
organizations to facilitate the processing of information related to complex tasks is
key to realizing these gains.
What specific forms of controls managers employ to achieve these ends is less
well understood. For example, some research suggest that, as tasks become more
complex and information processing demands increase, managers may attempt to
remedy control problems by incorporating more training and resource planning into
the work process (Galbraith 1973,1977). Others have observed that, under these
412 C. P. Long et al.
123
conditions, managers may focus on implementing process or goal-based directives
(Cardinal 2001).
2.4.2 Task interdependence
Task interdependence describes an attribute of multiple tasks where changes in one
task affects the state of and, thereby, elicits adjustments in other tasks. Increasing
levels of task interdependence can increase information uncertainty because it
forces subordinates both to rely on the potentially uncertain inputs of others to
complete their focal production tasks and to attempt to coordinate with others on the
completion of their task objectives (Thompson 1967). The more interdependent
particular tasks are, the more information that individuals must exchange to adapt,
adjust and achieve high effective levels of mutual understanding (Galbraith 1973;
Tushman 1978).
Scholars suggest that, in response to increasing levels of task interdependence
(Thompson 1967) managers will adjust the controls they implement in ways that
facilitate coordination, joint decision-making, and collective problem solving.
Similar to the current state of research on responses to task complexity, what
specific forms of controls managers employ in response to high levels of task
interdependence are less well understood. For example, Van de Ven and colleagues
(1976) and Daft and Lengel (1986) describe a wide range of formal and informal
coordination mechanisms that managers can employ to effectively respond to
information demands related to task interdependencies.
2.5 Managerial responses to information uncertainty: when to apply
controls
To accommodate information uncertainty demands, we argue that managers must
make two decisions regarding the application of organizational controls within their
organizational units. First, they must decide when they will exchange information
with their subordinates. Second, they must decide what type (i.e., quality) of
information they will exchange through those interactions.
Regarding when managers exchange information with their subordinates, the
types of controls that managers select are important because controls provide
opportunities to transfer information to subordinates at critical times in the
production process. To facilitate information processing, managers must choose to
apply configurations of input, process and output controls that enable them to
exchange task information at times when subordinates can most effectively use that
information to perform their production tasks. Generally, managers who use input
controls exchange production information with subordinates before they commence
tasks; managers who employ process controls exchange production information
with subordinates while they are working on tasks; and managers who use output
controls exchange production information with subordinates after their tasks are
completed.
While managers may choose to exchange information with their subordinates at
single points or multiple points in the production process, some preliminary research
How controls influence organizational information processing413
123
suggests that managers who apply multiple forms of controls may promote the best
fit with a range of organizational demands (Kirsch 2004). Work by Long et al.
(2002), for example, provide evidence that applying controls at multiple points in
the production process enhances information exchange. They reason that managers
who communicate with subordinates at multiple production points can more easily
adjust when they will exchange task information with their subordinates. Thus,
multiple controls enable these managers to more readily ensure that required
information is available to subordinates when they can most effectively use it to
perform their production tasks (Cardinal 2001).
2.6 Managerial responses to information uncertainty: controlling
information quality
We argue that, when evaluating how to respond to information uncertainty
demands, managers will also make decisions about the quality of the information
they will exchange with subordinates. Managers specifically select controls that
ensure their subordinates are consistently provided with data of a ‘‘richness’’ or
quality sufficient to help them effectively complete their tasks (Daft and Lengel
1986).
Daft and Lengel (1984,1986) describe how the quality of information is largely a
function of ‘‘the potential-information carrying capacity of the data’’ (i.e.,
information ‘‘richness’’) that is exchanged. Richer data allows individuals to
transmit comparatively more information within a given unit of time. The higher the
richness of the information that individuals exchange, the greater the chance that
individuals will maintain similar interpretations of that data. This happens because
actors who exchange richer data may more readily reconcile ‘‘multiple and
conflicting interpretations about an organizational situation’’ and can more easily
align their data interpretations (Daft and Lengel 1986, p. 556). More recently, the
emergence of big data has further enhanced the volume as well as richness of data
(Galbraith 2012).
The media that managers choose for the transmission of information comprises a
key component of this dynamic. Often these media choices are determined, at least
in part, by the control context within which the individual resides (Ouchi 1979).
Managers who operate within organizations that rely on formal controls tend to
employ bureaucratic elements such as job specialization and centralized decision-
making and generally exchange information of low to moderate richness. Here
subordinate tasks are directed by codified rules, standard operating procedures, and
written memoranda. While managers who use these media can more readily
transmit information to multiple receivers, the low richness of the information they
exchange tends to make it more difficult to quickly reconcile conflicting data
interpretations (Daft and Lengel 1986).
In contrast, managers who rely on informal controls tend to exchange
information of a much higher richness and use more informal means such as
face-to-face discussions and the transmission of cultural norms to direct their
subordinates’ activities (Cardinal et al. 2004; Ouchi 1980). Through these
exchanges, individuals transmit information concurrently through multiple audio
414 C. P. Long et al.
123
and visual channels (i.e., speech, facial expression, body language, etc.). While it
takes more time to transmit information to multiple receivers through these
channels, managers who rely on informal communications are able to provide
subordinates with more immediate feedback, and facilitate rapid clarifications of
divergent data interpretations.
2.7 Developing the information processing perspective on control use
We focus our investigation on increasing our understanding regarding how controls
can be used to facilitate information processing in organizational units. Building
from our descriptions of relevant factors, we use our study to attempt to determine
which configurations of organizational controls provide the best ‘‘fit’’ with a range
of information processing demands.
Due to the complex nature of the relationships we investigate, we have chosen to
develop and employ a series of computational models to vary the implementation of
control configurations, as well as levels of task complexity and task interdependence
in ways consistent with current conceptualizations of these constructs. Using our
approach, we aim to identify the most effective configurations of organizational
controls within various information uncertainty conditions.
3 Methods
3.1 Computational software
For our study, we chose the commercial software version 2.2 of the Vite’Project
(also VITE’) discrete event, agent-based computational modeling platform. Within
the parameters specified by the modeler, boundedly-rational computational agents
stochastically perform tasks and make decisions while communicating and
coordinating their work on projects containing other boundedly-rational computa-
tional agents. We selected the VITE’ program for this study because it
operationalizes the information processing view of organizations within the
decision-making and communication properties that agents in the platform possess.
Our study, thus, builds on a growing body of scholarly work that has developed and
employed this platform in organizational research (Jin and Levitt 1996; Levitt et al.
1994,1999; Long et al. 2002; Cardinal et al. 2011). The core components of VITE’
are described in the Appendix.
3.2 Study design
Consistent with recent observations on simulation research (e.g., Davis et al. 2007;
Harrison et al. 2007), we employ VITE’ as a laboratory for examining and
comparing the information processing capacities of various organizational control
configurations (Burton 2003). We developed a 2 929297 study design to
examine how 7 control configurations (single input control, single process control,
single output control, combined input/process control, combined process/output
How controls influence organizational information processing415
123
control, combined input/output control, combined input/process/output control) can
be used to facilitate efficient information processing under two task complexity
conditions (high/low), two task interdependence conditions (high/low), and two
control formalization conditions (formal/informal).
We used VITE’ to create two task complexity conditions, two task interdepen-
dence conditions, seven control configurations, and two control context conditions
where levels of formalization could be manipulated. Following the general outline
provided by Jin and Levitt (1996), we used VITE’ to model a sequence of
production tasks, assign computational agents to tasks and construct project
organizations within which work by computational agents was performed. This
approach enabled us to generate a set of organizational ‘‘projects’’ comprised of
theoretically critical variables (Harrison et al. 2007). We then systematically varied
key dimensions of those projects to compare the efficacy of various control
configurations under different information uncertainty conditions.
3.3 Project organization
Using an approach consistent with Long et al. (2002), we sought to examine our
research questions using a simple organization. This allowed us to isolate aspects of
our basic, focal phenomena while comparing and contrasting the effects of our key
variables. Specifically, we modeled an organization consisting of one project
manager agent, two team manager agents and two teams comprised of two team
member agents each (referred to hereafter as ‘‘project manager,’’ ‘‘team manager,’
and ‘‘team member’’). We differentiated these seven computational agents by the
tasks they performed, their organizational position, and the compensation they
received.
Team members devoted all their time each computational work day (9 h) to
performing work on a single generic production task while team managers applied
their efforts to three control tasks. To model managers’ organizational control
applications, we sequenced the placement of managers’ control tasks at key points
relative to subordinates’ tasks. Consistent with the approach used by Long et al.
(2002), team managers performed their input control tasks before team members
began working on their production tasks. Team managers performed process control
tasks while team members performed their production tasks. Lastly, team managers
performed output control tasks after team members completed their tasks.
The production sequence we developed enabled us to model how team managers
perform two basic control activities related to information processing: (1) to
communicate task specifications to subordinates, and (2) to monitor whether
subordinates perform work in alignment with those specifications. Specifically, we
used VITE’s failure dependency links to model managers’ monitoring of
subordinates’ tasks. As described in Appendix, failure dependency links enable
agents ‘‘downstream’’ in the production process to identify stochastically generated
problems or ‘‘failures’’ with the products generated by agents ‘‘upstream.’’ When
downstream agents detect problems, they request and work with upstream agents to
correct the identified errors. This is similar to managers monitoring subordinates and
requesting that their subordinates fix or ‘‘rework’’ problems that they observe.
416 C. P. Long et al.
123
When modeling input controls, we placed team members ‘‘downstream’’ from
team managers. As subordinates detected problems or ‘‘failures’’ with managers’
applications of input controls, they issued requests for managers to correct those
problems. In contrast, our models of process and output controls placed team
managers downstream of team members. As team managers detected problems with
the tasks that subordinates performed, those team managers issued requests and
worked with team members to correct or ‘‘rework’’ those task errors.
Using variations of this process we were able to develop and test different
organizational control configurations. When applying single forms of controls (e.g.,
a single focus on input control) both team managers in an organization distributed
their time each computational work day to three of the same type of control tasks
(e.g., three input control tasks). When applying a combination of two controls (e.g.,
input/process control), one team manager in an organization devoted 2/3 of their
time to one type of control task (e.g., input control), and 1/3 of their time to the other
control task (e.g., process control). The other manager allocated their time
differently, devoting 1/3 of their time to the first control task (e.g., input control) and
2/3 of their time to the second control task (e.g., process control). Thus, the
organization distributed team managers’ time and effort equally over two forms of
control. When applying a combination of input/process/output controls both team
managers in an organization distributed their time equally over three different
control tasks (i.e., one input control task, one process control task, and one output
control task). Our manipulations of controls are outlined in Table 2.
We use Fig. 2to outline the production process and the applications of input,
process, and output controls used in this study. Team managers apply input controls
to subordinates working on production tasks (1). Team members perform tasks
within input control parameters provided by their team managers (2a). As team
members identify problems with the input controls that team managers apply, they
send requests to managers to modify or ‘‘rework’’ those input controls (2b). While
subordinates perform tasks (2a), team manager agents also exert process controls
Table 2 Attributes for various configurations of controls
Input
controls
Process
controls
Output
controls
Input/
process
controls
Process/
output
controls
Input/
output
controls
Input/
behavior/
output
controls
Time when
controls are
applied
Before
tasks
During
tasks
After
tasks
Before and
during
tasks
During
and after
tasks
Before and
after
tasks
Before,
during, and
after tasks
Average time managers in organization focus on particular control(s) (in hours per virtual work day)
Input control 9.0 0.0 0.0 4.5 0.0 4.5 3.0
Process control 0.0 9.0 0.0 4.5 4.5 0.0 3.0
Output control 0.0 0.0 9.0 0.0 4.5 4.5 3.0
How controls influence organizational information processing417
123
(3a). As managers detect failures in team members’ task products, they send
requests to agents to correct these identified deficiencies (3b). Once team members
complete their tasks, they forward the products of their efforts to team managers for
evaluation (4a). As managers identify problems with subordinates’ work products,
they issue requests for subordinates to correct those errors (4b). Throughout, the
project manager supervises the process (5) while attending to the information and
decision requests submitted by team managers and team members.
3.4 Control context
To evaluate how variations in information quality impacts information processing
efficacy, we examined two control context conditions in our study we manipulated
VITE’s ‘‘organizational culture’’ parameters and the frequency and attendance of
agents (i.e., project manager, team managers, team members) at organizational
meetings. Please refer to the Appendix for details about described parameters.
3.4.1 Formal control context
Project managers, team managers, and team members in our formal control context
were all connected within a formal hierarchy. To model how agents in these control
contexts exchange information through formal communication channels (Ouchi
1980), multi-agent meetings were not scheduled. To model the characteristics of a
formal control context in which formal structures and procedures are key
coordination and standardization mechanisms, formalization and centralization
PM
T1 SL
ST1A ST1B
1
1
T1 Output
Control
1
1
T1 Process
Control
1
T1 Input
Control
1
Subordinate
Task Activity
Subordinate
Task Activity
11
2a
2a
2b 2b
3a 3a
3b 3b
4b 4b
4a 4a
Applied Cont rol
Actor-Manager Rework/
Control Adjustment
Unit Produced Unit Produced
5
Manager Agent
Project Manager Agent
(Manages Two Units)
Team
Member
Agent 1
Team
Member
Agent 2
Fig. 2 Diagram of model production process (Adapted from Long et al. 2002)
418 C. P. Long et al.
123
were set to high and matrix strength was set to low. In addition, we modeled
information exchanges in mature rule systems by setting team experience to high.
3.4.2 Informal control context
All agents in informal control contexts were connected within the organizational
hierarchy. To model informal communication mechanisms, we set formalization to
low. Centralization was also set to low to model how team members in these
systems make decisions on how to deal with failed task products. In addition, we set
both matrix strength and team experience to high to model the cohesiveness of a
typical informal control context. Lastly, to model how information control contexts
rely on the development of agents’ collective understanding (Ouchi 1979,1980), we
scheduled daily meetings which we required all organizational agents to attend.
3.5 Task complexity
Task complexity is a characteristic of production tasks that describes the difficulty
and variability of the work undertaken by individuals. To operationalize differing
levels of task complexity, we manipulated the ‘‘Requirement Complexity’’ of
agents’ tasks as well as the ‘‘Application Experience’’ and ‘‘Application Skill’’ that
agents possess for performing those tasks.
3.5.1 Requirement complexity
Requirement complexity within VITE’ is a characteristic of production tasks that a
team manager and team members perform. It describes ‘‘the number and difficulty
of functional requirements that need to be satisfied to complete each activity’’
(VITE’ Handbook 1998, p. 24). Tasks with higher levels of requirement complexity
possess more functional requirements, are more difficult to complete, and stimulate
increased information exchanges between actors (Tushman and Nadler 1978).
Under conditions of high task complexity, the requirement complexity of both their
tasks and the tasks their subordinates perform are set to ‘‘high.’’ Under the condition
of low task complexity, the Requirement Complexity of their tasks and the tasks
subordinates perform are set to ‘‘low.’
3.5.2 Application experience
Application experience within VITE’ describes an individual agent’s capability on a
specific set of tasks. It determines how quickly and accurately agents can process
project-relevant information. VITE’ allows modelers to toggle experience levels
between ‘‘low’’ and ‘‘high.’’ When agents possess a ‘‘high’’ level of application
experience, they understand how a particular task can and should be performed and
can effectively perform their tasks with little need to request information from other
agents in the project. Under the high task complexity condition, we set the
Application Experience levels of agents to ‘‘low’’. Under the low task complexity
condition we set each agent’s Application Experience to ‘‘high.’
How controls influence organizational information processing419
123
3.5.3 Skill level
Skill level within VITE’ describes an individual agent’s proficiency on a particular
task-relevant skill and, similar to Application Experience, determines how rapidly
and accurately agents can process project relevant information. Agents with lower
skill levels initiate more frequent information requests from other agents in the
model. They do this in order to supplement their existing skill knowledge.
Consistent with our manipulations of Application Experience we set the Skill Levels
of all agents to ‘‘low’’ under the high task complexity condition and ‘‘high’’ under
the low task complexity condition.
3.6 Task interdependence
We modeled levels of task interdependence by manipulating connections between
the tasks subordinate actors performed and both the solution complexity and task
uncertainty of the tasks all actors performed.
Connections between subordinate tasks For this study, we manipulated task
interdependence by varying connections between team members’ tasks. To generate
conditions of high task interdependence, we sequenced the timing of agents’ tasks in
ways that modeled reciprocal interdependence and linked subordinate tasks using
failure dependency connections. This increased the extent to which the effective
completion of one team member’s task was dependent on the effective completion
of the other team members’ tasks. In conditions where we modeled low task
interdependence, we did not connect team member tasks.
Solution complexity describes a characteristic of production tasks that specifically
relates to ‘‘the extent to which an activity’s requirements affect and are affected by
the requirements of other functionally interdependent activities’’ (VITE’ Handbook
1998, p. 24). Those who perform tasks with ‘‘high’’ levels of solution complexity
are dependent on the outputs of others’ tasks and initiate information exchanges to
obtain those inputs. In contrast, those who perform tasks with ‘‘low’’ levels of
solution complexity are much less dependent on the outputs produced by others.
Hence, under conditions of high interdependence we set the solution complexity of
the tasks to ‘‘high’’ and under the condition of low interdependence we set those
parameters to ‘‘low.’
Uncertainty represents a key element of task interdependence that specifically
describes ‘‘the extent to which information needed to complete an activity is
unavailable at the time the activity starts’’ (VITE’ Handbook 1998, p. 24). We
conceptualize uncertainty as an element of task interdependence because ‘‘high’
levels of uncertainty describe conditions where individuals must rely on the outputs
of others’ tasks to complete their focal tasks. Whereas when uncertainty is low, the
tasks that individuals complete are less dependent on the task outputs produced by
others. Under conditions of high interdependence, we set the task uncertainty of the
tasks that all agents performed to ‘‘high’’ and under conditions of low interdepen-
dence we set those parameters to ‘‘low.’’ Table 3summarizes the manipulations for
task complexity and task interdependence used in this study.
420 C. P. Long et al.
123
Table 3 Computational manipulations of task complexity and task interdependence
Task
complexity
Task
interdependence
Application
experience
Generic skill
level
Functional
complexity
Solution
complexity
Task
uncertainty
Subordinate
tasks
Low Low High High Low Low Low Not connected
High Low Low Low High Low Low Not connected
Low High High High Low High High Connected
High High Low Low High High High Connected
How controls influence organizational information processing421
123
3.7 Procedure
We used analysis of variance (ANOVA) procedures to test the efficacy of focusing on
specific formal and informal control configurations within four uncertainty condi-
tions. We used overall project cost (in thousands of dollars) as the performance
measure for various control combinations examined in this study. This is consistent
with traditional control research that has identified effective controls as the least
costly method for directing subordinate tasks (Eisenhardt 1989; Levinthal 1988).
In this study, we examine the costs incurred by each project organization (i.e.,
one project manager, two team managers, four subordinates) to produce 100
production units. The cost to complete a full production run (i.e., 100 units) is
calculated in thousands of ‘‘VITE’’ dollars. To develop statistical measures, the
overall production costs of 5 complete production runs were recorded and averaged
for each control/formalization/task uncertainty condition.
1
We conducted ANOVAs for each of our four task uncertainty conditions. Within
each condition, we compared the cost of producing 100 production units when
agents operate within one of 14 control conditions (seven formal and seven informal
control configurations). Relatively greater information processing efficiencies were
obtained when a particular control configuration enabled agents to produce 100
production units at a significantly lower production cost than alternative control
configurations. This is consistent with the approach used by Long et al. (2002).
We used post hoc analyses to obtain the most effective control configurations
within each information uncertainty condition. We classified ‘‘clusters’’ of the most
effective control options by identifying control configurations that were not
significantly different (p [0.05) from each other in the effects they produced on
production costs but were significantly different (p \0.05) from more expensive
(i.e., less effective) control options.
4 Results
We present the results we obtained for the four information uncertainty conditions in
Table 4. In the discussionthat follows, we describe the general and specific observations
which emerge from these analyses and evaluate how our findings refine and extend our
understanding of how controls affect organizational information processing.
4.1 Results by condition
4.1.1 Low task interdependence/low task complexity
Overall, the results we obtain suggest that a broad range of formal and informal task
control configurations provide managers with the means to effectively direct
1
This is generally consistent with the approach used by Long et al. (2002). We concluded that five
production runs would be sufficient as the standard deviations of five production runs were very small
when compared to the overall project cost for each run.
422 C. P. Long et al.
123
Table 4 Mean project cost (in thousands of dollars) when managers use different formal and informal control configurations to manage tasks of varying complexity and
interdependence in formal and informal control contexts
Low task complexity/high task interdependence High task complexity/high task interdependence
Formal control Informal control Formal control Informal control
MSD MSD MSDMSD
Input 22.4* 0.5 39.8 1.0 Input 77.4 1.3 52.8* 0.9
Process 21.2* 0.5 75.2 5.3 Process 87.5 1.4 138.4 2.9
Output 25.2 0.2 184.1 5.2 Output 80.2 1.8 211.1 4.4
Input/process 22.3* 0.4 41.0 1.9 Input/process 76.9 1.0 56.6 0.9
Process/output 21.8* 0.4 150.2 1.2 Process/output 81.1 1.4 180.3 3.6
Input/output 22.0* 0.5 118.1 6.6 Input/output 76.4 0.4 132.0 1.8
Input/process/output 22.0* 0.3 105.5 4.9 Input/process/output 77.1 1.1 116.5 2.6
F1660.0*** F2426.4***
How controls influence organizational information processing423
123
Table 4 continued
Low task complexity/low task interdependence High task complexity/low task interdependence
Formal control Informal control Formal control Informal control
M SD M SD M SD M SD
Input 13.3 0.1 12.6* 0.3 Input 47.0 0.7 26.7* 0.3
Process 12.6* 0.3 12.6* 0.2 Process 51.0 0.8 39.2 0.6
Output 12.9* 0.2 16.1 0.3 Output 43.2 0.4 35.5 0.6
Input/process 12.9* 0.3 12.5* 0.2 Input/process 48.0 0.9 26.4* 0.5
Process/output 12.9* 0.5 15.8 0.2 Process/output 47.6 0.6 42.2 1.2
Input/output 12.9* 0.4 15.1 0.2 Input/output 45.3 0.8 33.5 0.2
Input/process/output 12.9* 0.3 14.6 0.3 Input/Process/Output 46.1 0.8 32.1 0.3
F106.7*** F731.2***
* Bolded values indicate the lowest total project cost within each information uncertainty condition. ‘‘Clusters’’ of multiple, effective control configurations indicate
control options that did not produce significantly different effects on production costs (p [0.050). The results obtained by these options were significantly different than
more expensive control options (p \0.050)
*** p \0.001
424 C. P. Long et al.
123
information processing when both task complexity and task interdependence are
low. In this condition, there exists a significant main effect of control configuration
on information processing efficiency (F (1, 70) =106.7, p \0.001). We obtain
essentially equivalent results when managers use formal process, output, input/
process, process/output, input/output, and input/process/output controls and when
managers apply informal input, process and input/process controls (all represent
among the least expensive control options at p \0.050; n =70). This suggests that
when organizations encounter low levels of information uncertainty, managers
maintain a wide variety of essentially equivalent formal and informal control
options for effectively managing information processing demands within their units.
4.1.2 High task interdependence/low task complexity
When task interdependence is high and task complexity is low there exists a
significant main effect of control configuration on information processing efficiency
(F (1, 70) =1660.0, p \0.001). As task interdependence rises and task complexity
remains low, we observe that configurations of formal input, process, input/process,
process/output, input/output, and input/process/output controls provide managers
with equivalently effective mechanisms to direct their subordinates’ activities (all
represent the least expensive options at p \0.050; n =70).
4.1.3 Low task interdependence/high task complexity
There exists a significant main effect of control configuration on information
processing efficiency (F (1, 70) =731.2, p \0.001) if task complexity is high and
task interdependence is low. Here, configurations of informal input or input/process
controls appear to provide the most effective means of directing organizational
information processing efforts (i.e., these represent the least expensive options at
p\0.050; n =70).
4.1.4 High task interdependence/high task complexity
Lastly, there exists a significant main effect of control configuration on information
processing efficiency (F (1, 70) =2426.4, p \0.001) if task complexity is high and
task interdependence is high. In this condition, our results suggest that informal
input controls provide managers with the most effective means of directing
organizational information processing (this represents the least expensive option at
p\0.050; n =70).
5 Discussion
The findings we obtain from this study enable us to make several contributions to
organizational research. First, we highlight the role that controls play in helping
organizations direct the use of task-related information. Second, we explore the
capacity of various configurations of controls to accommodate a range of different
How controls influence organizational information processing425
123
information processing demands. Our research results, thus, provide the basis for
understanding how and why managers integrate and balance information processing
and agency-based concerns within their overall control portfolio. In the sections that
follow, we describe these contributions in more detail and outline how our findings
advance our understanding of organizational control implementation.
5.1 Distinguishing the information processing perspective
Although we distinguish the information processing aspect of controls that we
develop in this paper from the role that controls play in managers’ efforts to mitigate
relational risk, we contend that both categories of concerns impact managerial
control implementation and performance (Kirsch 2004; Leifer and Mills 1996). As a
result of the findings we obtain from this study, we suggest that scholars should
examine how managers use controls to both manage relational risks and ensure that
their subordinates possess the information necessary to effectively complete their
organizational tasks.
In explaining how managers integrate their responses to these concerns, we first
acknowledge that task complexity and task interdependence can jointly and
independently compromise a manager’s capacity to predict, direct, and measure
their subordinates’ production behaviors. This is because, as task complexity and
task interdependence increase, it takes more managerial effort and resources to
specify and monitor how subordinates should perform their production tasks.
If managers’ capacities to measure their subordinates’ production outputs remain
constant while managers find it more difficult to observe and program subordinate
tasks, agency prescriptions dictate that managers should increase their reliance on
formal output controls (Ouchi 1977,1979). Under these conditions, agency theory
suggests that managers should direct their employees using formal contracts that
specify the outputs that subordinates need to deliver and then simply monitor
whether subordinates deliver the outputs that they are contracted to provide
(Levinthal 1988). The agency perspective outlines how these decisions enable
managers to align subordinate work incentives and to shift primary production risks
to their subordinates (Eisenhardt 1989).
While agency theory would encourage managers to emphasize the application of
formal output controls in conditions of our study where task complexity and task
interdependence are high, we observe that a singular emphasis on formal output
controls is among the most effective information processing options only under
conditions when both task complexity and task interdependence are both at low
levels. In fact, we find that managers who rely only on formal output controls to
direct their subordinates’ work efforts as levels of information uncertainty rise
significantly decrease their organization’s capacity to process information and
achieve production efficiencies.
To fully understand the implications of these findings, we reiterate that the
information processing perspective describes how managers will choose configu-
rations of controls to direct task information to their subordinates when and where
they can most effectively use it to perform their production tasks. As levels of task
complexity and task interdependence rise, the amount of information that
426 C. P. Long et al.
123
subordinates require to effectively complete their tasks increases as well. While
managers may address their agency concerns by exchanging information with
subordinates at points in the production process where their relational risk is
minimized, managers who rely exclusively on those control choices may not be able
to deliver information when that data is most needed by subordinates in their task
efforts.
The general proposition to managers that emerges from our study is that, as
information uncertainty rises, managers should couple applications of formal output
controls that they use to manage relational risks with the control configurations that
we have identified as the most effective choices for managing particular information
uncertainty demands. By tailoring their control applications in ways consistent with
the observations we present here, we contend that managers can effectively manage
their relational risks and direct subordinates through tasks of varying complexity
and interdependence.
5.2 Effective managerial responses to complex tasks
The results we obtain suggest that managers should respond to increasing levels of
task complexity by emphasizing more informal means of control and targeting their
control mechanisms towards the input and process portions of their subordinates’
task efforts. In these conditions, managers might focus their control efforts on
emphasizing training and socializing their subordinates before they actually engage
their tasks or informally monitoring subordinates while they perform their tasks
(e.g., informal behavioral monitoring).
These findings enable us to explain accounts of managerial actions that describe
how rising levels of information uncertainty often lead managers to expand their
control portfolios past the use of just formal output controls (Kirsch 2004; Turner
and Makhija 2006). For example, managers in the information technology industry
who direct the work of complex and interdependent groups of programmers often
tend to manage information-based concerns by focusing their attention on selecting
talented programmers (input controls) and then providing programmers with
information about programming processes (process controls). They do this while, at
the same time, managing their relational risks by evaluating how well programmers’
finished products align with pre-specified objectives (formal output controls)
(Choudhury and Sabherwal 2003).
We find that how well managers accommodate information processing concerns
can profoundly impact production efficiencies. Specifically, we observe that if
managers adhere only to agency prescriptions and rely solely on formal output
controls to direct subordinates performing complex tasks (i.e., of either low or high
interdependence), they will increase information processing costs by about 50 %
over results they would obtain by using informal input controls (i.e., when task
interdependence is high) or informal input and process controls (i.e., when task
interdependence is low).
In addition to these findings, we note that the apparent importance of input
controls for managing complex tasks challenges the general proposition posed by
the agency perspective that input control mechanisms should serve as a ‘‘fall-back’’
How controls influence organizational information processing427
123
form of control (Snell and Youndt 1995, p. 716) that managers employ only when
they cannot construct effective process-based or outcome-based contracts. In fact,
our findings suggest that input controls provide managers with among the most
effective mechanisms for managing tasks that are both complex and interdependent.
While more research is necessary on this, Snell (1992) suggests one reason for this
finding is that managers who utilize input controls to exchange information with
subordinates before they commence work on their tasks are able to ‘‘prevent
performance problems,’’ (p. 297) before they occur.
5.3 Effective managerial responses to interdependent tasks
As task interdependence increases but task complexity remains low, we observe that
managers who establish high levels of formalization appear to be the most
successful at managing information process flows. Our findings suggest that, as long
as managers apply controls using formal communication media (i.e., such as memos
and SOPs), they maintain a wide variety of equally effective control alternatives.
Interestingly, we find that managers who use only formal output controls under
these circumstances will increase their production costs over the most effective
control option (i.e., formal process controls) by about 20 %. Thus, our result refines
observations by Galbraith (1973) that managers often rely heavily on formal
controls to mitigate the increases in information uncertainty that occur in the context
of task interdependencies. First, we observe that managers may effectively target
those standardization efforts across a wide array of formal control configurations.
Second, we see that formal output controls are not among managers’ most effective
control options (mean: 25.2; SD =0.2).
This finding is important because it challenges theoretical propositions presented
by Leifer and Mills (1996) that managers rely increasingly on rich information
media as information uncertainty increases. Because we distinguish the different
effects that task interdependence and task complexity have on information
uncertainty, we show how rich information media provide effective mechanisms
for managing interdependent tasks only if those tasks are also complex.
5.4 Integrating information processing and agency perspectives on control
While we observe that information processing considerations are important, we
again stress here our contention that managers should not ignore agency concerns in
control-based decision-making. Recently, Cardinal et al. (2004) reported observa-
tions of a moving company where employees performed simple but highly
interdependent tasks. While managers in this organization effectively utilized input
and process controls to exchange information with their employees about their daily
tasks, management’s failure to combine those controls with formal output controls
to address agency concerns ultimately compromised management’s capacity to
effectively motivate their employees and protect the organization from employee
deviance (Cardinal et al. 2004).
Because our research suggests that the perspectives diverge and address different
control-related problems, we argue that scholars should more closely examine how
428 C. P. Long et al.
123
managers concurrently address both agency and information processing concerns in
their applications of organizational controls. While the potential value of this
integrated approach may ultimately rest on determining how managers balance their
attention concurrently to multiple issues, it will be important for researchers to
examine when managers focus more of their attention on information processing
demands and when managers focus more of their attention on agency demands
(Choudhury and Sabherwal 2003; Leifer and Mills 1996).
How managers sequence their attempts to address these demands could prove to
be an important issue. For example, because the relational risks that managers
encounter are often determined by the information processing demands that their
units face, scholars may want to examine how managers adjust their perceptions of
relational risks based on the information processing demands they encounter. An
additional issue for scholars to investigate here may be how managers become
aware of information processing concerns after they have failed to effectively
reduce relational risks. By acknowledging and further investigating these issues,
scholars may begin to develop a much more complete understanding of how
managers use multiple controls to direct their subordinates’ work activities.
5.5 The role of formalization
While the agency perspective focuses primarily on how managers apply formal
controls, we argue that it is important for scholars to specify both the level of
formalization and the element of the production process that managers target in
their control applications. Our results support this general proposition in that we find
that managers who direct simple yet interdependent tasks are most effective when
they use configurations of formal input, process, and output controls. Here, it
appears that formalization allows managers to efficiently coordinate activities across
multiple employees.
In contrast, managers hoping to effectively direct subordinates who perform
complex tasks should use informal communication media to apply input and process
controls. While more research is necessary on this finding, one explanation for this
might be that managers and subordinates who use rich communication media can
adjust their comprehension of complex tasks in real time and can more rapidly
cultivate higher levels of mutual understanding about task complexities.
By providing more detailed prescriptions of how various media may be applied
to accommodate specific information processing demands, we refine and extend
insights about the relationship between tasks and organizational design offered by
Huber and McDaniel (1986), and Daft and Lengel (1986). Past providing general
propositions that managers may use formal rules and regulations or informal face-
to-face meetings under various conditions, our research helps identify where in the
production process those efforts should be targeted. We suggest that managers who
use formal controls and direct highly interdependent tasks should focus on
developing a wide range of formal control options: from developing and applying
standard operating procedures, to conducting formal training, to codifying output
standards. Alternatively, we argue that managers who use informal controls to direct
How controls influence organizational information processing429
123
highly complex tasks should focus primarily on implementing effective socializa-
tion and training mechanisms and enforcing process norms.
5.6 Multiple controls versus single controls
As researchers begin to examine how managers integrate their efforts to manage
relational risks with their efforts to promote information processing, scholars will
need to develop an understanding of how managers balance multiple forms of
controls. This is a potentially important conclusion that merits further investigation
since it challenges the general argument presented by the agency perspective that
managers are most efficient when they rely on singular forms of controls
(Eisenhardt 1989; Ouchi, 1979).
While control theorists who have utilized the risk management perspective (e.g.,
Ouchi 1979; Snell 1992) have occasionally suggested that input, process, and output
controls ‘‘are not mutually exclusive and are frequently used in combination,’’
(Snell and Youndt 1995, p. 712), researchers have generally avoided examining the
joint applications of these controls. However, because our research suggests that
effective managers will often use multiple controls to both manage relational risks
and promote effective information processing, we argue that managers should
devote more attention to understanding how managers balance their applications of
multiple controls (Kirsch 2004; Long et al. 2002).
We note that, in three study conditions we examined, configurations of multiple
controls were among the most effective control options. This suggests that managers
who seek to promote effective information processing will often have to integrate
multiple controls. For managers who must tangle concurrently with multiple
concerns, using multiple controls may allow them to more effectively engage
production deficiencies as they arise instead of hoping that they, a priori, have
identified the potential source of their production problems (Long et al. 2002).
While multiple control options appear to exist within many of our study
conditions, we do observe that, as information processing demands increase, the
number of viable information exchange options decrease. Our findings here suggest
that, while a range of control options might work under conditions of low
information uncertainty, the stakes for managers to choose the appropriate control
remedy rise significantly as task complexity and interdependence jointly and
independently increase.
5.7 Limitations and future research
The perspective adopted in conceptualizing this study is a strength but also
represents a potential limitation of this research. Specifically, our focus on
relationships between control and information processing captures only one
representation of these processes. The advantage of our work is that it highlights
how managers use controls to provide subordinates with information about their
task obligations and organizational roles (Kirsch 2004). However, because multiple
factors influence managers’ control-based decisions, future research should directly
430 C. P. Long et al.
123
examine how managers integrate and balance multiple concerns (agency, etc.) that
impact control implementation efforts.
In addition, while our use of computational modeling provides some key
advantages over more traditional research approaches, this design decision also
limited this research effort in some respects. Although VITE’ provided us with the
capability to assess and systematically vary key factors influencing control
implementation, future research should compare the results we obtain here with
results generated in the field. By combining approaches, scholars will be able to
provide a more complete picture of how closely managers’ decisions and actions
converge to the observations we present.
An additional issue that future research should address is how managers’ control
applications change as individuals’ capabilities change. Recent research has shown
how issues related to individual learning can be a critically important aspect of
control implementation (Kirsch 2004; Makhija and Ganesh 1997; Turner and
Makhija 2006). While we suggest that each condition in our study represents a
snapshot of individuals’ levels of task knowledge and interdependence, we
acknowledge that an individual’s level of task comprehension and extent to which
they work interdependently are dynamic factors. As such, future research should
examine how managers alter applications of organizational controls as the
complexity and interdependence of organizational tasks change.
Appendix
This appendix provides an overview of the VITE’ program
EC. 1. Overview
The commercial software version 2.2 of the Vite’Project (also VITE’) discrete event
computational model is comprised of an agent-based computational modeling
platform. Within the parameters specified by the modeler, boundedly-rational
computational agents stochastically perform tasks and make decisions while
communicating and coordinating their work on projects containing other bound-
edly-rational computational agents.
According to Levitt et al. (1999, p. 1483), VITE’ ‘‘models the total information-
processing capacity of an organization as the aggregate information-processing
capacities of its nodes, modified by the efficiency of the communication network-
comprised of vertical relationships defined by the formal structure, and emergent
lateral relationships driven by activity interdependencies—that connects the nodes.
The simulator computes the total information-processing load on the organization
from the project requirements for direct work and coordination (i.e., control) work.
Organizational performance is determined by how closely the organization’s
capacity to handle information aligns with the load that it is presented.’’ The ‘‘load’’
that Levitt et al. (1999) describe refers to the overall amount of information
processing effort that agents working on a project within an computational
organization collectively produce.
How controls influence organizational information processing431
123
While others have described the core components of the VITE’s computational
modeling platform in detail (Jin and Levitt 1996; Levitt et al. 1994,1999), below we
briefly describe the three basic components of the VITE platform: agents, tasks, and
the organizational structure.
EC. 1.1. Agents
Agents within VITE’ work on ‘‘projects’’ consisting of linked tasks. How they
perform project tasks is determined stochastically by their behavioral matrix which
is developed from information processing principles. Consistency between compo-
nents of agents’ behavioral matrices and actual human behavior have been tested
and verified using both empirical research and extant organizational practice (Levitt
et al. 1999). A computational agent’s behavioral matrix specifies their capacity to
process and exchange the information necessary to complete their assigned tasks.
EC. 1.2. Tasks
Agents perform work by stochastically transferring activities from their computa-
tionally generated ‘‘in-tray’’ (i.e., work to do) to their ‘‘out-tray’’ (i.e., completed
work). How quickly that transfer occurs depends on how efficiently agents can
process information related to their own tasks (i.e., production work) and
interdependent tasks performed by other agents (i.e., coordination work). The
priority the organization places on particular activities as well as the order in which
those activities stochastically arrive in an agent’s in-tray determines what activities
an agent addresses at any particular point in time.
2
To facilitate the effective completion of their tasks, individual agents may also
exchange two important categories of information with other agents in the project.
First, agents stochastically issue ad-hoc information requests to other agents who
perform interdependent tasks. The purpose of these information exchanges is to
coordinate and ensure that agents performing interdependent tasks make compatible
choices in their respective activities. Second, agents often request that agents
pursuing interdependent tasks ‘‘rework’’ failed outputs of task efforts. This most
often happens when agents ‘‘downstream’’ in the production process, identify
stochastically generated problems or ‘‘failures’’ with the products generated by
‘upstream’’ agents.
How efficiently individual agents exchange information with each other is a
crucial component of an overall project. This is because when an individual agent
submits a request for information or requests that other agents ‘‘rework’’ task
products, the agent making the request will suspend work on their task and ‘‘wait’
for their request to be answered. Each time that an agent waits for their information
requests to be answered, the overall project is delayed, thereby compromising how
efficiently the organization process information.
2
Agents’ decisions are set at a default level where there exists 50 % probability that they work on
activities which currently maintain the ‘‘highest priority’’ in their in-tray. In addition, they maintain a
20 % probability of working on activities that first appeared in their in-tray and a 20 % probability of
selecting activities that last appeared in their in-trays. 10 % of their activities are randomly selected.
432 C. P. Long et al.
123
EC. 1.3 Organizational structure
How efficiently agents exchange information within a project is dependent both on
agents’ behavioral matrices (described above) and the design of the organization
within which agents perform work. Agents within a project are connected to each
other within a hierarchy and assigned one of three roles in a decision-making
hierarchy: (in order of descending authority) project managers, team managers, or
team members. An agent’s hierarchical position determines the types of decisions
they make and the amount of information they are required to process (Levitt et al.
1994; Jin and Levitt 1996).
The way in which information is transferred between agents is further determined
by VITE’s four ‘‘organization’’ parameters: centralization, formalization, team
experience, and matrix strength.
Centralization determines the level of the organizational hierarchy where
decisions on ‘‘reworking’’ activity failures occur. In projects with higher central-
ization, agents higher in the organization (i.e., hierarchy) make rework decisions
while in projects with lower centralization (i.e., decentralization), agents who
perform tasks tend to make decisions on handling task failures.
Formalization describes the frequency with which agents transmit information
requests to other agents. When formalization is high, agents focus on their particular
functional duties and make fewer information requests of other agents. When
formalization is low, agents seek to collaborate more and initiate a higher number of
information requests of other agents.
Matrix strength specifies how often agents respond to the information requests of
other agents. When matrix strength is low, agents focus primarily on their individual
tasks and respond less readily to other agents’ requests for information. Agents in
organizations with higher matrix strength are more collaborative and respond more
to other agents’ information requests.
Team experience reflects the total amount of project-related experience that a
team possesses. Agents possessing higher levels of team experience work more
efficiently because they are familiar with each other and the demands of a particular
project.
EC. 1.4. Project outcomes
Each run of the simulation calculates the collective agent-based information
processing effort needed (in thousands of dollars) to generate a specified number
production units. This cost is generated based on the collective amount of time it
takes all the agents in a particular project organization to process the information
necessary to complete their task work.
References
Burton RM (2003) Computational laboratories for organization science: questions, validity, and docking.
Comput Math Organ Theory 9:91–108
How controls influence organizational information processing433
123
Cardinal LB (2001) Technological innovation in the pharmaceutical industry: managing research and
development using input, behavior, and output controls. Organ Sci 12:19–36
Cardinal L, Sitkin S, Long C (2004) Balancing and rebalancing in the creation and evolution of
organizational control. Organ Sci 15:411–431
Cardinal L, Turner S, Fern M, Burton R (2011) Organizing product development across technological
environments: performance trade-offs and priorities. Organ Sci 22:1000–1025
Choudhury V, Sabherwal R (2003) Portfolios of control in outsourced software development projects. Inf
Syst Res 14:291–314
Cohen MD, March JG, Olsen JP (1972) A garbage can model of organizational choice. Adm Sci Q
17:1–25
Daft R, Lengel RH (1984) Information richness: a new approach to manager information processing and
organization design. In: Staw B, Cummings LL (eds) Research in organizational behavior. JAI
Press, Greenwich
Daft R, Lengel RH (1986) Organizational information requirements, media richness and structural design.
Manag Sci 32:554–571
Davis JP, Eisenhardt KM, Bingham CB (2007) Developing theory through simulation methods. Acad
Manag Rev 32:480–499
Eisenhardt K (1985) Control: organizational and economic approaches. Manag Sci 31:134–149
Eisenhardt K (1989) Agency theory: an assessment and review. Acad of Manag Rev 31:57–74
Fayol H (1949) General and industrial management (Storrs C, trans). Pitman, London
Galbraith JR (1973) Designing complex organizations. Addison-Wesley, Reading
Galbraith JR (1977) Organization design. Addison-Wesley, Reading
Galbraith JR (2012) The future of organization design. J Organ Design 1:3–6
Ghoshal S, Moran P (1996) Bad for practice: a critique of the transaction cost theory. Acad Manag Rev
1:13–47
Hagel J III, Brown JS (2007) The only sustainable edge: why business strategy depends on productive
friction and dynamic specialization. Harvard Business School Press, Boston
Harrison J, Carroll G (1991) Keeping the faith: a model of cultural transmission in formal organizations.
Adm Sci Q 36:552–582
Harrison JR, Zhiang L, Carroll GR, Carley KM (2007) Simulation modeling in organizational and
management research. Acad Manag Rev 32:1229–1245
Huber G, McDaniel RR (1986) The decision-making paradigm of organization design. Manag Sci
32:572–589
Jin Y, Levitt R (1996) The virtual design team: a computational model of project organizations. Comput
Math Organ Theory 2:171–196
Kirsch L (1996) The management of complex tasks in organizations: controlling the systems development
process. Organ Sci 7:1–21
Kirsch L (1997) Portfolios of control modes and IS project management. Inf Syst Res 8:215–239
Kirsch L (2004) Deploying common systems globally: the dynamics of control. Inf Syst Res 15:374–395
Leifer R, Mills PK (1996) An information processing approach for deciding upon control strategies and
reducing control loss in emerging organizations. J of Manag 22:113–137
Levinthal D (1988) A survey of agency models of organizations. J Econ Behav Organ 9:153–155
Levitt RE, Cohen GP, Kunz JC, Nass CL, Christiansen T, Jin Y (1994) The ‘‘virtual design team’’:
simulating how organization structure and information processing tools affect team performance. In:
Carley KM, Prietula MJ (eds) Computational organization theory. Lawrence Erlbaum Associates,
Hillsdale
Levitt R, Thomsen J, Christiansen TR, Kunz JC, Yan J, Nass C (1999) Simulating project work processes
and organizations: towards a micro contingency theory of organizational design. Manag Sci
45:1479–1495
Long CP (2010) Control to cooperation: examining the role of managerial authority in portfolios of
managerial actions. In: Sitkin SB, Cardinal LB, Bijlsma-Frankema K (eds) Organizational control.
Cambridge University Press, Cambridge, pp 365–395
Long C, Burton RM, Cardinal LB (2002) Three controls are better than one: a computational model of
complex control systems. Comput Math Organ Theory 8:197–220
Lubatkin M, Lane PJ, Collin S, Very P (2007) An embeddedness framing of governance and
opportunism: towards a cross-nationally accommodating theory of agency. J Organ Behav 28:43–58
Makhija MV, Ganesh U (1997) Control and partner learning in learning-related joint ventures. Organ Sci
8:508–527
434 C. P. Long et al.
123
March JG (1991) Exploration and exploitation in organizational learning. Organ Sci 2:71–87
Oliver C (1998) Critical theory perspectives on control. Adm Sci Q 43:257–292
Ouchi W (1977) The relationship between organizational structure and organizational control. Adm Sci Q
22:95–113
Ouchi W (1979) A conceptual framework for the design of organizational control mechanisms. Manag
Sci 25:833–848
Ouchi W (1980) Markets, bureaucracies, and clans. Adm Sci Q 25:129–141
Perrow C (1965) Hospitals technology, structure and goals. In: March JG (ed) Handbook of organizations.
Rand-McNally, Chicago
Simons R (1995) Levers of control: how managers use innovative control contexts to drive strategic
renewal. Harvard Business School Press, Boston
Snell S (1992) Control theory in strategic human resource management: the mediating effect of
administrative information. Acad Manag J 35:292–327
Snell S, Youndt M (1995) Human resource management and firm performance: testing a contingency
model of executive controls. J Manag 21:711–737
Thompson JD (1967) Organization in action: social science bases of administrative theory. McGraw-Hill,
New York
Turner K, Makhija MV (2006) The role of organizational controls in managing knowledge. Acad Manag
Rev 31:198–217
Tushman ML (1978) Technical communication in R&D laboratories: the impact of project work
characteristics. Acad Manag J 21:624–645
Tushman ML, Nadler DA (1978) Information processing as an integrating concept in organization design.
Acad Manag Rev 3:613–624
Van de Ven AH, Delgecq AL, Koenig R (1976) A task contingent model of work-unit structure. Adm Sci
Q 19:322–338
Van Maanen J, Schein E (1979) Toward a theory of organizational socialization. In: Staw B, Cummings L
(eds) Research in organizational behavior. JAI Press, Greenwich
VITE’ Handbook (1998) Vite’ Corporation
Chris P. Long is an Assistant Professor in Management at the McDonough School of Business,
Georgetown University. In his research, he examines how control, trust, and fairness is enacted by leaders
and experienced by employees within both traditional organizations and new organizational forms. Chris
earned his Ph. D. in Management from Duke University.
Sim B. Sitkin is Professor of Management, Director of the Behavioral Science and Policy Center, and
Faculty Director of the Fuqua/Coach K Center on Leadership and Ethics at Duke University. He is an
internationally recognized expert on the effect of formal and informal organizational control systems and
leadership on risk taking, accountability, trust, learning, change, and innovation—all areas in which he
has published and consulted extensively. Sim is Editor of the Academy of Management Annals and
Founding Editor of Behavioral Science and Policy and a Fellow of the Academy of Management and
Society of Organizational Behavior. His previous service has included the Academy of Management
Board of Governors, Organization Science Senior Editor, and Journal of Organizational Behavior
Associate Editor. His most recent book is entitled Organizational Control.
Laura B. Cardinal is Professor of Strategic Management at the C. T. Bauer College of Business at the
University of Houston. Her expertise is principally in the area of implementation of innovation goals and
strategies and includes the effects of organizational control and coordination on innovation, R&D, and
new product development teams. Laura sits on the Board of Directors of the Strategic Management
Society, is an Associate Editor of the Academy of Management Annals, and serves on the editorial boards
of the Academy of Management Journal,Organization Science, and Strategic Management Journal. Prior
service includes roles as Director of the Burkenroad Institute for the Study of Ethics and Leadership at
Tulane University, Chair for the Competitive Strategy Interest Group of the Strategic Management
Society, and Division Chair for the Technology and Innovation Management Division of the Academy of
Management. Her most recent published book is entitled Organizational Control.
How controls influence organizational information processing435
123
Richard M. Burton is Professor Emeritus of Organization and Strategy at The Fuqua School of Business,
Duke University. He is Honorary Professor at the University of Southern Denmark and the University of
Aarhus. He holds the BS, MBA and DBA from the University of Illinois. Rich is Associate Editor for the
Journal of Organizational Design, Associate Editor for the Strategic Management Journal, and Co Editor
of a research series on Information and Organization. Previously, he was Senior Editor for Organization
Science, Department Editor for Strategy, Organizational Design and Performance for Management
Science. He is a founding member of the Organization Design Community. He has published seven books
and some seventy articles on strategy, organization and management science in Organization Science,
Management Science, Administrative Science Quarterly, Computational and Mathematical Organization
Theory, among others. His most recent book is Organizational Design: A Step-by-Step Approach, 2011,
with Professors DeSanctis and Obel.
436 C. P. Long et al.
123
... Nevertheless, the size of an organisation is a factor that has been ignored so far in the study of the relationship between the intensity of knowledge management and competitive advantage. In the literature we can find studies which show relationship between size and organisational effectiveness (Amah, Daminabo-Weje, & Dosunmu, 2013), efficiency (Burton, Minton, & Obel, 1991), performance (Manojlović, 2016) and innovation (Forés & Camisón, 2016). The practice of knowledge management varies depending on the company size. ...
Article
Full-text available
Objective: This article analyses the intensity of knowledge management processes in companies of different size. It also examines the role of these processes in the formation of competitive advantage by these businesses. Company size is defined as the number of employees. Research Design & Methods: This article is based on a quantitative study in which 1258 companies were examined. Several statistical tests were used to analyse the data, including the U Mann-Whitney test, linear regression and Pearson correlation. Findings: It was found that the intensity of knowledge management processes was higher for bigger firms. The results also show that, regardless of the size, entities with more intensive knowledge management processes were relatively more competitive. Another finding is that knowledge management processes were regarded especially important for the competitiveness of large companies, with 250 employees or more. Implications & Recommendations: The implications of this study can be valuable for managers preparing to execute knowledge management processes. Based on insights from the study managers can plan strategically and make informed decisions about what type of knowledge management initiatives to implement. Contribution & Value Added: The size of an organisation is a factor that so far has been ignored in the study of the relationship between the intensity of knowledge management and competitive advantage. Specific characteristics of a company that result from its size determine its unique approach to knowledge management – the principles and rules that apply to large organisations cannot easily be scaled down and implemented in SMEs.
... The modern startups, instead, seem to keep a low level of formalization and a precise focus on efficiency and effectiveness, on exploitation and exploration (Miles, Miles, Snow, 2006). Burton, Minton and Obel (1991) have followed the intuition of Kimberly (1976) by which the company size should consider not only the personnel, but also physical capacity, inputs and outputs, and discretionary resources (available to an organization). The size of the organization should then be defined as the "number of resources and policies, and the number of products and services". ...
Conference Paper
Full-text available
The concept of scientific management introduced by F. W. Taylor (1914) was based on the idea that the setup of internal rules on task management was “the” solution for coordinating the workforce and achieving managerial control. More than a century later, rules as coordination mechanisms are still embedded in metaphoric and visual artifacts, like organizational forms and their representation in organization charts. Theory and practice of organization design have stated and refined such rules (as organization forms), so that despite the identification of some new emerging forms (Spaghetti organization, Organizational Improvisation, etc.) the contingency approach to organization design has always been able to explain them, both at the inter-and intraorganizational level (see Burton & Obel, 1998). Nevertheless, the joint effect of globalization and economic downturn have boosted the emerging of “modern startups” that, differently from the traditional Small and Medium Sized Enterprises (SMEs) see to be designed to grow –not just to survive. In fact, while Dodge, Fullerton & Robbins (1994) found that no matter what their stage of life cycle is, the main aim of SMEs is “survival”, these “modern startups” seem to be designed to grow. Their novelty probably seek for new rules of internal design.
Chapter
In today's volatile business environment, it is more important than ever that managers, whether of a global multinational or a small team, should understand the fundamentals of organizational design. Written specifically for executives and executive MBA students, the edition of this successful book provides a step-by-step 'how to' guide for designing an organization. It features comprehensive coverage of the key aspects of organizational design, including goals, strategy, process, people, coordination, control and incentives. These aspects are explained through the use of a unique series of 2 x 2 graphs that provide an integrated, spatial way to assess and plan organizational design. The new edition features a number of important improvements, including a new framework for understanding leadership and organizational climate, the introduction of the concept of manoeuvrability and a completely new chapter examining joint ventures, mergers, partnerships and strategic alliances.
Chapter
In today's volatile business environment, it is more important than ever that managers, whether of a global multinational or a small team, should understand the fundamentals of organizational design. Written specifically for executives and executive MBA students, the edition of this successful book provides a step-by-step 'how to' guide for designing an organization. It features comprehensive coverage of the key aspects of organizational design, including goals, strategy, process, people, coordination, control and incentives. These aspects are explained through the use of a unique series of 2 x 2 graphs that provide an integrated, spatial way to assess and plan organizational design. The new edition features a number of important improvements, including a new framework for understanding leadership and organizational climate, the introduction of the concept of manoeuvrability and a completely new chapter examining joint ventures, mergers, partnerships and strategic alliances.
Chapter
This chapter, mostly based on the work by Forrest and Orvis (Kybern Int J Cybern Syst Manag Sci 45:1308–1322, 2016), introduces two important principles of efficiency—management efficiency and organizational inefficiency. The former addresses the problem of how management efficiency can be achieved, while the latter investigates the structure of employees’ efforts and devotion toward realizing the mission of the organization. These results help to enrich the managerial understanding on what can be improved and what cannot.
Book
Cambridge Core - Organisation Studies - Organizational Design - by Richard M. Burton
Article
Purpose The purpose of this paper is to introduce two important principles of efficiency, one on the management of a business entity and the other on the structure of employees’ efforts and devotion toward realizing the mission of their organization. Design/methodology/approach All discussion and reasoning are established on some of the traditional methods of microeconomics and on the basis of the systemic yoyo model. Here, the yoyo model plays the role of intuition, while the traditional methods are utilized to present the exact details underneath the systemic thinking. Findings What is discovered include how management efficiency can be achieved by being flexible in terms of allowing individual employees to have conflicting personal values and how organizational inefficiency always exists no matter how the business entity is set up. Research limitations/implications The established results are applicable in all business scenarios without foreseeable limitations. Practical implications By understanding these results, business managers could simply devote more of their time and effort on being flexible in terms of management styles and focusing on the “big” picture of the corporation instead of dwelling on how to improve employees’ efficiencies. Originality/value This paper establishes two very important, very useful results for managers. These results are expected to enrich the managerial understanding on what can be improved and what cannot.
Chapter
Decomposition methods have been intensively studied both as techniques for solving large-scale models as well as a means for analysing different control mechanisms in quantitative approaches for decentralized planning structures. The reason for this latter application lies in the straight forward interpretation of most classical decomposition methods in terms of information flow between different levels in a decentralized organization. The development of aggregation approaches for the solution of large-scale models has not followed the same pattern. In this area, the emphasis has been on the algorithmic development and the efficiency of the aggregation methods as solution techniques. The interpretation of aggregation and disaggregation techniques as models for analysing control mechanisms in decentralized planning structures are sparse although there are some exceptions [1, 2]. In this chapter, we compare different types of aggregation approaches with standard decomposition approaches focusing on their different formulations and on the informational requirements in different organizational structures, such as the functional or the multidivisional structure. We exemplify the different approaches by means of a small example and extend it partly to a well-known example for multidivisional and functional organization given by Burton/Obel [3].
Chapter
Organizational design examines how an organization should be put together, i.e. its structural form, its information system, its decision processes, and its incentives for efficiency and effectiveness. Contingency theory posits that the most efficient design will depend on these design alternatives and on the organizational situation. Mathematical models provide a precise language and clear statement of the organizational structure for a rigorous analysis and better understanding of the managerial choices for the organizational design. We provide a brief overview of organizational mathematical models. These include: mathematical programming models, team theory, game theory, and agency theory models. We examine, in more detail, the modelling of hierarchical organizations using mathematical programming models within a contingency framework. External contingencies, such as environmental uncertainty, size, and technology are related to the choice of the organizational design. Organizational design includes the organizational form (e.g. M-form, U- form, or matrix), the decision processes and the resource allocation scheme, (e.g. price or budget), and the incentive scheme (e.g. group or individual). The information system (calculation, storage and exchange) ties the organization together. The organizational characteristics of complexity, formalization, specialization and decentralization are related directly to the mathematical models of the organization. Different empirical research methods such as computer simulations and laboratory experimentation based upon mathematical programming models are presented to examine contingency hypotheses. Various contingency concepts are restated precisely and new insights are developed. Finally, we access promising future research directions.
Article
It is the objective of this study to describe three fundamentally different mechanisms through which organizations can seek to cope with this problem of evaluation and control. The three will be referred to as markets, bureaucracies, and clans. In a fundamental sense, markets deal with the control problem through their ability to precisely measure and reward individual contributions; bureaucracies rely instead upon a mixture of close evaluation with a socialized acceptance of common objectives; and clans rely upon a relatively complete socialization process which effectively eliminates goal incongruence between individuals. This study explores the organizational manifestations of these three approaches to the problem of control.
Article
The abstract for this document is available on CSA Illumina.To view the Abstract, click the Abstract button above the document title.
Article
The abstract for this document is available on CSA Illumina.To view the Abstract, click the Abstract button above the document title.
Article
The interaction network of an organized group is always patterned to some extent by the size of the group. By using the criterion of interaction possibilities, one can distinguish groups as small, medium, large, and giant groups; each has distinctive characteristics. The number of potential relationships in a group of given size may be determined by simple mathematical formulas and increases rapidly as the size of the group increases. Empirical studies suggest that the effect of size on efficiency of task performance depends upon the type of task assigned, but that increasing size is positively correlated with group stability, the proportion of resources devoted to self-maintenance, the uniformity of organizational design, and the incidence of communication problems.