The Interactive Effects of Perceived Environmental Uncertainty and Control System Type on

Managers’ Willingness to Recommend Strategic Change

Margaret Christ
The University of Georgia
Terry College of Business
mchrist@uga.edu

Anna Cianci
Wake Forest University
cianciam@wfu.edu

and

Stuart Napshin
Kennesaw State University
Coles College of Business
snapshin@kennesaw.edu

August 2016

Electronic copy available at: http://ssrn.com/abstract=2824135

 The Interactive Effects of Perceived Environmental Uncertainty and Control System Type on
Managers’ Willingness to Recommend Strategic Change

ABSTRACT

To remain competitive today, flatter organizations must rely more on middle

management in order to adapt in response to demands in the dynamic environments in which

they operate. Yet oftentimes organizations do not change even in the face of evolving markets.

Using an experiment, we examine how the type of control system used by organizations can

shifting middle manager perception of environmental stability as well as influencing their

willingness to recommend organizational change.

Consistent with our expectations, we find a positive association between middle

managers’ perceived environmental uncertainty and their willingness to recommend

organizational strategic change. This relationship is moderated by control system type,

especially for those middle managers whose perceived environmental uncertainty is low. In

particular, we find that relative to interactive control systems, diagnostic control systems are

more effective in directing manager attention toward environmental uncertainties and increasing

the willingness of those managers to recommend a strategic change. Our results are an initial step

in understanding the relationships between environmental uncertainty and control systems and

thus can help senior managers design and implement control systems for flatter organizations in

dynamic environments where middle managers are critical to initiating adaptive strategic change.

Keywords: control systems; perceived environmental uncertainty (PEU); strategic change.

Data Availability: Contact the authors.

Electronic copy available at: http://ssrn.com/abstract=2824135

 INTRODUCTION

There are numerous examples of companies that do not initiate strategic change in the

face of environmental threats until firm performance has deteriorated. Examples include

Hewlett-Packard, which only initiated a strategic shift after reporting disappointing financial

results (Worthen and Sherr 2011), and Research in Motion Ltd. (the maker of Blackberry

devices), a once dominate, now struggling smart phone company that only reacted to the

introduction of the iPhone after their market share had significantly deteriorated (Cummins

2011). Such failure to respond to the changing environment may result from the failure of

managers to recognize the need for strategic change when the firm is performing well financially

(e.g., Greve 1998, 2003). In dynamic environments identifying and acting on environmental

change prior to financial deterioration is a critical but under examined context.

We contend that organizational and individual factors contribute to managers’

willingness to recommend strategic change in the face of environmental uncertainty. At the

organizational level, we focus control systems. Diagnostic and interactive control systems are

important, yet distinct, control systems used by senior managers to direct middle manager

attention and action (Simons 1990, 1995; Garg et al. 2003; Kober et al. 2007). In a diagnostic

control system senior management initiates contact with middle management only when the

financial system indicates that a performance target has been missed. Alternatively, in an

interactive control system senior management maintains an ongoing personal involvement with

middle management to establish programs and review monthly financial performance and action

plans. These differences between diagnostic and interactive control systems are likely to affect

the extent to which middle managers focus their attention on external factors when assessing

organizational performance and evaluating current strategies.

Electronic copy available at: http://ssrn.com/abstract=2824135

 At the individual level, we examine perceived environmental uncertainty (PEU), which is

an individual’s constructed perception of uncertainty of the external environment (Fisher, 1996).

PEU is shown to influence attention (Garg et al. 2003), control system impact (Agbejule 2005)

and the identification of organizational improvements (Sorensen and Stuart 2000).

We develop a theoretical model that describes how PEU drives middle manager

willingness to recommend strategic change and how that relationship is influenced by the control

system used by the organization. We propose that the control system type (i.e., diagnostic vs.

interactive) will moderate the influence of PEU on willingness to recommend strategic change

by the way it affects middle managers’ attention and how it focuses or diffuses the ultimate

responsibility of the strategic decisions being made.

To test our model we employ 152 MBA students with management experience in a 2x2

experiment in which PEU is measured and dichotomized into two levels (high or low); and

control systems (diagnostic or interactive) are manipulated between participants, consistent with

prior research (e.g., Simons 1990, 1991, 1995; Tessier and Otley 2012).

Results of our study find that diagnostic control systems are more effective than

interactive control systems for middle managers with low PEU. Under the diagnostic control

system, low PEU managers are less likely to perceive the market as stable and attractive and are

more willing to recommend strategic change. Interactive control systems did not have these

effects on low PEU managers. Further, middle managers with high PEU were not differentially

affected by either type of control system and their willingness to recommend strategic change

remains relatively high regardless of the control system type.

Overall, these results suggest that diagnostic control systems may more effective for

organizations operating in high velocity environments where environmental change may have

2
occurred but firm performance has not yet deteriorated. Although the willingness of high PEU

managers to recommend strategic change is not differentially impacted by interactive or

diagnostic control systems, the decisions of low PEU middle managers is improved by diagnostic

control systems.

This study extends the literature by providing experimental evidence of the effect of

control systems on individual strategic behavior when individual perception of environmental

uncertainty differs. By distinguishing between managers who are lower or higher in PEU, we

show that low PEU managers, in particular, are influenced by the type of control system imposed

to more carefully scrutinize market conditions and incorporate those external factors into their

decision to recommend strategic change.

Importantly, unlike prior research which has generally focused on strategic change in

environments where firm financial performance has typically already deteriorated (e.g.,

Langfield-Smith 1997, 2006; Chenhall 2003; Henri 2006), we examine an environment where

firm performance has not deteriorated but where the industry is showing fundamental change.

This is important given the tendency of firms to not recognize or adapt to environmental change

until after firm financial performance has started to suffer (e.g., Greve 1998, 2003). Recognizing

externalities and the effect they may ultimately have on firm performance is a critical

management skill and our study shows that there are control systems that the organization can

use to facilitate this. In particular, we show that diagnostic control systems are more effective

than interactive control systems at facilitating organizational adaptation to environmental change

prior to firm performance deterioration.

In supplemental analysis, we explore how the type of control system affects managers’

assessment of the risk associated with a strategic change. Our results show that managers’ risk

3
assessment fully mediates the relationship between control system type and willingness to

recommend change, further suggesting that the control system itself changes the way that

managers view the environment in which they are operating. Although risk assessments are an

integral part of organizations’ control activities, there is very little research exploring how the

types of control systems impact these risk perceptions. Given the important role of risk

assessment in organizational strategy (e.g., COSO 2004), it is important to explore these

relationships.

The rest of the paper is organized as follows. In the next section, we review research on

the effects of PEU and control systems on strategic behavior and develop our theoretical model

and hypotheses. We then describe our research method and present our results. We conclude by

discussing the implications and limitations of our research and offering suggestions for future

research.

LITERATURE ANALYSIS AND HYPOTHESES DEVELOPMENT

We illustrate our theoretical model that describes how the individual factor (Perceived

Environmental Uncertainty) and the organizational factor (Control Systems) interact to affect

managers’ willingness to recommend strategic change in Figure 1.

<<Insert Figure 1 about here>>

Individual Factor: Perceived Environmental Uncertainty

Perceived Environmental Uncertainty (PEU) is an individual’s perception of uncertainty

that results from factors outside the organization (i.e., the external environment) (e.g., Milliken

1987; Tymon et al. 1998). Prior empirical research concludes that individuals respond, “not to

the physical environment [itself] but the environment they have constructed themselves. It is this

environment they perceive as being real” (Fisher 1996). PEU affects both the attention to, and

use of, information (Daft et al. 1988; Sawyerr 1993; Mangaliso 1995; Garg et al. 2003) and prior
4
research shows that it can influence control system design, use and impact (e.g., Gul and Chia

1994; Fisher 1996; Chenhall 2003; Agbejule 2005). PEU has also been found to be positively

associated with environmental scanning (e.g., Mangaliso 1995; Garg et al. 2003) and the

identification of organizational improvements (e.g., Sorensen and Stuart 2000; Tripsas and

Gavetti 2000; Ahuja and Lampert 2001).

Consistent with this literature, which highlights the importance of senior management

cognition (Nadkarni and Barr 2008) on strategic decisions, we expect PEU to play a role in

middle manager assessment of environmental attractiveness. Managers who perceive the

environment as low in uncertainty consider the environment to be relatively stable, and attractive

(e.g., Gul and Chia 1994; Fisher 1996; Agbejule 2005) and therefore will believe the current

market conditions to be favorable. In contrast, managers who perceive the environment as high

in uncertainty are likely to consider the environment to be unstable, unattractive, and overall

unfavorable (e.g., Gul and Chia 1994; Agbejule 2005). Thus, we hypothesize:

H1: There is a negative relationship between a manager’s PEU and that manager’s
perception of market favorability.

PEU’s link to a willingness to recommend change flows through middle manager’s

perception of the attractiveness of environment. Simply stated, managers who perceive low

uncertainty consider the environment relatively stable and attractive. Under these cognitive

conditions, managers are less likely to recommend strategic change because they do not perceive

environmental shifts (Garg et al. 2003) that could deteriorate the firm’s performance (Greve

1998, 2003) or the need for organizational improvements (Tripsas and Gavetti 2000). However,

middle managers who do perceive high uncertainty consider the environment unstable and

potentially unattractive. Under these cognitive conditions, managers are more likely to

recommend strategic change because they are sensitized to potential environmental shifts (Garg

5
et al. 2003) that could deteriorate their organization’s financial performance (Greve 1998, 2003)

and are looking for options for firm level improvements (Tripsas and Gavetti 2000). Thus, we

hypothesize:

H2: Managers who perceive market favorability as high are less likely to recommend
strategic change than managers who perceive market favorability to be low.

Organizational Factor: Control Systems

Given that middle managers operate within an organizational context, we expect that

their individual level decisions will be influenced by organizational level factors. Control

systems are mechanisms designed to direct employee behavior such that it helps the organization

achieve its objectives. Controls can take many forms such as policies and procedures,

performance measurement systems, codes of conduct and supervisory approval. However, in

addition to the specific objectives of the control system design (e.g., to require approvals, or to

set budget targets), prior research suggests that the type of control system itself conveys

information to the employees subjected to the control. For example, Christ et al. 2008 show that

relative to output controls, which focus on outcomes, behavior controls, which set limits on

employee actions, are often perceived to be signals of distrust. Similarly, incentives framed as

bonuses versus penalties convey a more trusting signal to employees (Christ et al. 2012) which

affects employees’ effort level.

Diagnostic and Interactive control systems 1 are important, yet distinct control systems

used by senior managers to direct middle manager attention and action (Simons 1990, 1995;

1
Simons (1995) also includes ‘boundary’ systems (codes of conduct, proscribed behaviors) and ‘belief’ systems
(values, credos, mission statements) as the two other control levers that organizations use for effective control. We
focus on ‘interactive’ and ‘diagnostic’ control systems because these systems (unlike boundary and belief systems)
provide feedback on strategy efficiency and effectiveness (Langfield- Smith 2006).

6
Garg et al. 2003; Kober et al. 2007). Diagnostic control systems are often described as

‘management by exception’ systems because they compare actual to targeted performance.

Performance targets are based upon metrics that are clearly understood within the organization

and usually represent key drivers of firm success. Comparative reports concentrate managerial

attention, especially when performance goals are missed, and trigger a search for solutions

(Greve 1998).

In contrast, interactive control systems are characterized by personal, regular and ongoing

dialogue and engagement by senior management with middle management. In anticipation of

this ongoing dialogue, middle managers attend to and dedicate effort toward understanding

strategic uncertainties. Through this interactive dialogue, which is inherently inquisitive and

information seeking, there is the ongoing establishment of goals, review of progress, and regular

follow-up. Interactive control systems generate broad attention and effort that is externally

focused and driven by creative search and inductive thinking (Simons 1990, 1991).

Prior research documents the influence of control systems on strategic behavior. For

instance, Simons (1994) documents the use of interactive and diagnostic control systems by top

managers to alter firm behavior in response to environmental change. In addition, prior research

has documented, in a survey of hospital CEOs, a positive relation between the use of interactive

control systems and strategic change (Abernethy and Brownell 1999); a relation between control

systems and executives’ ability to facilitate the acceptance of the need for strategic change (e.g.,

Ezzamel et al. 2004; Skærbæk and Tryggestad 2010); and a relation between control systems and

strategy formulation, implementation and modification in a management buyout context

(Bruining et al. 2004). Overall, these studies document the close relationship between control

systems in general and strategic change.

7
 However, diagnostic and interactive control systems represent different methods of

generating middle manager attention to organizational strategy that also affect the final

responsibility of the decision being made. We anticipate that these attention and responsibility

effects will interact with middle manager PEU and influence their willingness to recommend

strategic change.

Within a diagnostic system middle managers’ attention is directed toward specific

performance (often financial) metrics and indirectly the external factors that may impact these

metrics (Simons 1995; Ocasio 1997; Ocasio and Joseph 2005). In this way, the diagnostic

system’s focus on specific metrics allows the environment to be interpreted in relationship to the

firm. Thus, the metrics of the diagnostic system can act as an attention conduit toward

environmental issues and provide a structure to determine whether strategic change is necessary.

Additionally, within a diagnostic control system middle managers have the autonomy to

perform tasks in the manner they deem best (Christ et al. 2008). Senior management only

reviews the output from this work, without concern for how it is achieved. While this allows

middle managers to focus on the outcomes of their decisions it also makes them personally

responsible for the achievement of those outcomes. Due to this heightened responsibility middle

managers making such decisions are likely to more deeply consider the factors that might affect

their probability of success and therefore would scrutinize the market conditions to more clearly

understand and monitor the external factors that could impact those outcomes (Simons 1995;

Ocasio 1997; Ocasio and Joseph 2005).

Thus under a diagnostic system middle managers face increased personal responsibility

and an interpretive system that should help them understand the potential impact of

environmental changes on the firm’s important performance metrics.

8
 Unlike diagnostics systems, which are more management by exception, interactive

systems, are based upon the back and forth communication between middle management and

senior management. Designed to be a more interpersonal, free form and ongoing

communication, interactive systems by design do not focus managerial attention on specific

metrics but instead directs managerial attention toward issues raised through ongoing

conversations as they shift, flow and ebb. While interactive systems stimulate search on these

issues, that search has a tendency to be broad, un-focused, potentially random and driven by

inductive thinking (Simons 1995). Thus, while the diagnostic system provides a conduit to focus

middle manager attention and interpretation, the interactive system does not.

Further, within interactive systems, the personal responsibility of middle managers may

be reduced as decisions are not left up to individual managers, but are the result of joint effort

and ongoing communication. Middle managers may even rely to a greater extent on the

perceptions and decisions of senior managers instead of making careful assessments themselves.

Indeed, one of the characteristics of interactive control systems that differentiates them from

other types of control systems is that they provide a mechanism to focus on “information that

top-level managers have identified as potentially strategic” (Simons 1995, p 87). As a result,

responsibility for performance outcomes can be shared across all those involved in this

interactive dialogue. In this sense, interactive control systems could create a diffusion of

responsibility (Darley and Latane 1968) whereby managers are not motivated to change their

perceptions until prompted to do so by someone else.

Under an interactive control system, middle managers are faced with decreased personal

responsibility and an attention generating system that is more variable but also with less structure

for interpreting the impact of the information obtained. Given the attention, interpretation and

9
personal responsibility differences between interactive and diagnostic control systems, we expect

that middle managers operating under diagnostic control system will be more willing to

recommend strategic change. We expect that the increase in personal responsibility is more

likely to help middle managers identify critical information in the environment that can then be

better interpreted with ramifications for the firm under the diagnostic control system as opposed

to the interactive control system. Thus, we hypothesize:

H3: Managers operating under Diagnostic Control systems are more willing to
recommend strategic change than managers operating under Interactive Control Systems.

In addition to these direct effects, we also anticipate that strategic control systems will

interact with and influence the relationship between PEU and the perception of market

favorability. We expect that the attention, interpretation and responsibility effects of diagnostic

versus interactive controls systems described above is likely to affect middle managers with low

PEU differently from those with high PEU.

While the goal of interactive and diagnostic control systems are similar, the efficient and

long-term performance of the firm, they operate in different ways. As previously described,

interactive systems stimulate middle manager attention through the back and forth interpersonal

communication between middle management and senior management. It is through this

conversation that strategic issues are identified and explored. Within a diagnostic system middle

managers attention is directed toward specific performance (often financial) metrics and

indirectly the external factors that may impact these metrics (Simons 1995; Ocasio 1997; Ocasio

and Joseph 2005). Both of these control systems seek to stimulate middle manager attention to

important environmental issues in different ways. However, high PEU middle managers are

already sensitized to the uncertainty in the organizational environment, therefore the stimulative

10
effects of both control systems on additional environmental search would be muted. Therefore,

we expect both control systems to have little impact on high PEU middle managers.

For low PEU middle managers, both control systems would stimulate additional

environmental search. Within the interactive control system, this new information would be

more difficult to interpret for its organizational impact as the interactive system provides little

structure for this. Further, even if such information is interpreted in a way that suggests the need

for organizational change, the decreased personal responsibility inherent in the interactive system

is likely to reduce the middle manager’s impetus to action. Within the diagnostic system, this

new information would be easier to interpret as the metrics of the diagnostic system can act as an

attention conduit toward environmental issues and provide a structure to determine whether

strategic change is necessary. Further, the increased personal responsibility inherent in the

diagnostic control system means that such information when interpreted is more likely to lead to

organizational change recommendations by the middle manager. Thus, we hypothesize:

H4: The negative effect of PEU on the manager’s perception of market favorability is
moderated by the type of control system imposed. Specifically, managers with low PEU will
assess market stability and attractiveness lower under diagnostic control system as
compared to an interactive control system. And, managers with high PEU will assess market
stability and attractiveness no different under a diagnostic control system compared to an
interactive control system.

METHODOLOGY

Participants

Participants were 152 MBA students from a metropolitan state university. Participants’

demographic data is presented in Table 1. As reported in Table 1, participants had, on average,

8.93 years of professional work experience and 5.92 years of public company work experience.

On a seven-point scale where 1 indicates not at all familiar and 7 indicates very familiar,

11
participants indicated that their mean (SD) experience with identifying and providing responses

to strategic uncertainties was 5.23 (1.17) and their mean (SD) familiarity with using a planning

and budgeting system for identifying areas of strategic focus was 5.32 (1.12). This suggests that

participants had an appropriate level of knowledge and experience to perform the experimental

task.

[Insert Table 1]

Design and Independent Variables

The experiment is a 2x2 design in which PEU is a between-participant measured variable

that is then dichotomized into two levels (high or low) and control systems and incentives are

between participant variables manipulated at two levels: diagnostic or interactive control systems

and financial or social recognition incentives, respectively. The first independent variable, PEU,

is measured by asking participants to indicate their agreement on a seven-point scale (1 =

“strongly disagree” and 7 = “strongly agree”) with the following statement: “Atlanta radio

market (i.e., the hypothetical business unit described in our experimental case) faces significant

environmental uncertainty.” 2 These responses are then dichotomized into high and low PEU

based on the median value of the responses. 3

The second independent variable, control systems, is manipulated between participants at

two levels: diagnostic and interactive control systems. In the diagnostic condition, participants

are told that senior management relies on monthly reports to monitor the performance of the

radio station. When reviewing these reports, senior management focuses on actual to budget

2
All participants are given the same information regarding the external environment including some general
industry information and some radio market metrics that indicate the market is starting to change.
3
There is no significant difference between the four conditions for participants’ assessment of environmental
uncertainty (F = 0.56, p-value = 0.64). Thus, while PEU was measured after participants read the case materials, the
manipulations had no discernible impact on participants’ responses to the PEU question.

12
results and only becomes involved in the operations of ‘Atlanta Radio’ when the monthly reports

indicate that a performance goal is missed. In the interactive condition, participants are told that

senior management maintains ongoing personal involvement with the radio station manager,

helps establish new programs and milestones, conducts monthly reviews of progress and action

plans and follows up on reports.

Dependent Variables

The primary dependent variable is participants’ willingness to recommend change

(WTRC). Participants indicated their agreement on a seven-point scale (1 = “strongly disagree”

and 7 = “strongly agree”) with the following statement: As the programming manager, you

recommend to the Board of Directors that Atlanta Radio change its programming format to reach

a larger share of its target market.

Market favorability is a composite variable made up of 2 separate questions designed to

measure how stable and attractive the market is perceived to be by participants. Participants

indicated their agreement on the same seven-point scale with the following statements: “Atlanta

Radio’s market share is likely to remain stable” (market stability) and “The Rock-Adult

Contemporary segment that Atlanta Radio competes in is attractive” (market attractiveness).

Responses to these two questions are significantly correlated at the 0.01 level. We constructed a

composite measure, which we call “market favorability,” by averaging participants’ responses to

these two questions.

Procedures

Participants were emailed an on-line link to our research instrument. In the experimental

scenario, ‘Atlanta Radio’ is the business-unit, a local radio station that is a subsidiary of a larger

entity. Participants are asked to assume the role of the station manager at Atlanta Radio, and are

13
told that the radio station manager is responsible for station profitability and has full authority for

station strategy and operations. Participants are informed that the radio station is a subsidiary of a

publicly traded company and its strategic objective for the station is long-term profitability

which requires maintaining market share. They are told that market share depends on its choice

of programming format and that audience music and programming preferences can vary greatly.

Participants are then provided information that radio stations generally maintain a single program

format throughout the day and that although a radio station may switch to a new daily program

format, such changes occur infrequently as they are costly and create revenue uncertainty.

Given this scenario, participants are provided with an overview of the competitive

landscape, audience preferences, programming content and performance and market metrics.

Finally, participants are told that the parent company is engaged in its strategic planning process

and conducting an annual update to its strategic plan. Accordingly, senior management is

arranging a presentation for the Board of Directors and has asked the station manager to assess

station performance and make a recommendation regarding a program format change. After

considering the case information, participants indicated their recommendation regarding a

programming format change and then responded to several post-experimental and demographic

questions.

RESULTS

Manipulation Checks

Two questions were used to examine whether the manipulations of control systems and

incentives were successful. The control system manipulation check question asked participants

whether the information presented in this case stated that the executive vice president takes an

ongoing personal involvement in Atlanta Radio or only becomes personally involved in the

14
operations of a division when the monthly reports indicate that a performance goal is missed.

The incentive manipulation check question asked participants whether the executive vice

president personally appreciates and publicly recognizes individual performance on a regular

basis or whether he recognizes individuals for their performance with significant extra financial

payments. Thirty-one of 183 participants who completed the survey were dropped for incorrectly

answering either of the two questions or incorrectly answering both questions. The resulting final

sample totaled 152 participants. 4

Descriptive Statistics and Main Effects Testing

Our primary research interest for this study is the effect of PEU (an individual factor) and

control system type (an organizational factor) on middle managers’ willingness to recommend

change. Table 2, Panel A, reports the simple main effects of each of these independent variables

on willingness to recommend change. As shown, managers’ willingness to recommend change is

higher for managers with high PEU compared to those with low PEU (means = 4.68 vs. 4.21,

respectively) and this difference is statistically significant (p <0.05, one-tailed). Turning to

differences between control systems, managers under diagnostic control systems have a greater

willingness to recommend change than those working under interactive control systems (4.75 vs.

4.29, respectively) and this difference is also significant (p<0.05 one tailed). Panels B and C

report means and standard deviations for each variable in the theoretical model by condition.

<<Insert Table 2 about here>>

Tests of our Theoretical Model

4
Inclusion of those participants who failed the manipulation checks in our analyses does not substantively change
our results or alter the conclusions we draw.

15
 We use structural-based equation analysis to estimate our theoretical relationships

described in Figure 1. To assess the model, we conduct a test of goodness of fit. The

Comparative Fit Index (CFI), a measure of the proportion of improvement of the fit of our model

to the null model, is 0.952, which is above the generally accepted minimum value of 0.95 (Byrne

2001). We further confirm the model’s goodness of fit with a traditional Chi-square test

(x2=3.43, p = 0.18) (Byrne 2001). Thus, the model provides a good fit for describing the relations

in the data. The standardized path coefficients and statistical significance are presented for the

estimated model in Figure 2.

As predicted in H1, our results show that PEU is negatively associated with managers’

perceptions of market stability and attractiveness. That is, managers who perceive higher levels

of environmental uncertainty also believe that the market is less favorable (i.e., unstable and

unattractive), relative to those managers who perceive little environmental uncertainty (-0.35,

p<0.05). Consistent with H2, we find that middle managers’ perception of market favorability is

negatively associated with their willingness to recommend change (-0.19, p<0.01).

H3 predicts that middle managers under diagnostic control systems are more willing to

recommend strategic change than middle managers working under interactive control systems.

Consistent with this expectation, we find a significant, negative relationship between control

system type and willingness to recommend change (-0.14, p<0.05).

Finally, H4 predicts an interaction between PEU and control system such that managers

with low PEU will assess market favorability lower under diagnostic control systems compared

to interactive control systems, while managers with high PEU will not be affected by control

systems. Consistent with this expectation, we find a positive significant relationship between the

interaction term and market favorability (0.13, p<0.05). This finding suggests that diagnostic

16
control systems are more effective than interactive control systems in directing low PEU middle

managers’ attention toward the changing environmental conditions even when their own

organization’s performance has not started to decline. However, high PEU managers, who

already have a heightened awareness of market conditions are not affected by the control system

type.

Supplemental Analysis

Our theoretical development and results suggest that middle managers operating under

diagnostic and interactive control systems perceptions of the environment in which their

organization operates. We explore, in supplemental analysis, the potential mediating effect of

managers’ risk assessments under each control system. There is some evidence from the

marketing literature that outcome-based controls (such as diagnostic control systems) promote

risk-taking behavior (e.g., Sitkin and Pablo 1992). That is, because marketing managers are

incentivized to achieve certain outcomes regardless of the strategy used to reach them, they are

more likely to take risks that they might not take otherwise. 5 We contend that relative to

interactive control systems, managers subjected to diagnostic control systems will be more likely

to assess the benefits of a potential strategic change to outweigh its risks. This risk assessment

will, in turn, affect the likelihood that the manager recommends a change such that as the relative

benefit to risk trade-off increases, so will the likelihood that the manager recommends a strategic

change.

5
A propensity for greater risk-taking under a diagnostic control system could, of course, have dysfunctional
performance effects because outcome focused managers may take risks that do not pay off. However, it is not the
goal of this paper to determine whether recommended strategic changes are appropriate or successful, but rather to
identify characteristics that can influence managers’ willingness to recommend change. Future research should
explore whether there are further consequences of these recommendations.

17
 We measure experimental participants risk assessment based on their agreement on a seven-point

scale (1 = “strongly disagree” and 7 = “strongly agree”) with the following question: “The financial and

operational costs involved in making a radio station format change is worth the potential reward.” We

re-run our path analysis including risk assessment as a mediating variable between control system type

and willingness to recommend change. To assess the model, we conduct a test of goodness of fit. The

Comparative Fit Index (CFI), a measure of the proportion of improvement of the fit of our model to the

null model, is 1.00, which is above the generally accepted minimum value of 0.95 (Byrne 2001). We

further confirm the model’s goodness of fit with a traditional Chi-square test (x2=3.51, p = 0.48). Thus,

the model provides a good fit for describing the relations in the data. The standardized path coefficients

and statistical significance are presented for the estimated model in Figure 3.

<<Insert Figure 3 about here >>

In this new model, the significance of the original paths remain, with the exception of the

direct path from control system type to willingness to recommend change, which becomes

insignificant with the introduction of risk assessment as a mediator variable (-0.08, p = .135, one-

tailed).

DISCUSSION

In many modern organizations, middle managers are empowered to be agents of change

within the organization. These managers are likely to be more familiar with the day-to-day

operations and therefore their ability to identify necessary strategic change is critical. However,

there are many examples of organizations that do not execute strategic changes until long after

the external environment has changed and the organization’s financial performance has begun to

deteriorate. In this study, we investigate the extent to which an individual manager’s perception

18
of environmental uncertainty and the type of the control systems imposed by senior management

affect the manager’s willingness to recommend strategic change.

Overall, we document a positive association between PEU level and managers’

willingness to recommend strategic change. That is, the more uncertain a manager perceives

his/her environment to be, the more likely he/she is to recommend a strategic change. We find

that this relationship flows through the manager’s perception of market favorability (i.e., the

stability and attractiveness of the market in which the organization is operating).

Consistent with our expectations, PEU level interacts with the type of control system

used (diagnostic vs. interactive) to determine managers’ willingness to recommend strategic

change. Specifically, diagnostic systems (compared to interactive systems) help low PEU

managers focus upon more intently and better incorporate available information about the

external environment that may help him/her identify the need for strategic change. For high PEU

managers, diagnostic systems do not have such effects since these managers are already

sensitized to the risks in the external environment.

In supplemental analysis we find that the direct effect of control systems on willingness

to recommend change is fully mediated by managers risk assessment. In particular, under a

diagnostic control system, managers are more likely to believe the potential benefits of a

strategic change will outweigh the costs. This finding suggests that the type of control system

imposed can change managers’ perception of the risks faced by the organization.

The results of this study have both research and business implications. Our findings

suggest that the role of PEU should be considered when evaluating the effectiveness of control

systems for strategic decision making. Indeed, the results here suggest that (a) the contextual

19
component should include PEU and (b) control systems must be simultaneously considered

given their interactive effect.

Regarding the impact of our findings on business implications, as environments increase

in complexity and organizations become flatter, the ability of firms to adapt to changing

environmental conditions is dependent upon the actions of managers (Ocasio 1997; Ocasio and

Joseph 2005). Reporting systems and incentives are two critical levers senior managers have to

influence managers’ behavior. While control systems are usually established to encourage

specific levels of performance, those same systems influence the ability of managers to recognize

and internalize environmental issues and recommend action be taken accordingly. The current

study demonstrates that the effect of control system type is dependent upon the individual traint,

perceived environmental uncertainty. High PEU managers, who appear to already have a

heightened sensitivity to environmental changes, are generally less influenced by the attention

generating and focusing actions of control systems. However, diagnostic control systems seem

to help low PEU managers attend to the threats in the environment, thereby increasing their

willingness to recommend strategic change. Thus, given the effect of PEU and control systems, it

behooves senior managers to consider, and possibly influence, managers’ PEU level. Further, in

order to take advantage of high PEU managers’ WTRC as well as positively influencing that of

low PEU managers, diagnostic control systems appear most effective.

This study has several limitations that present opportunities for future research. First,

following Milliken (1987) and Tymon et al. (1998), we define PEU as an individual’s perception

of uncertainty that results from factors outside the organization. This definition of PEU is most

similar to state uncertainty – i.e., lack of understanding about how components of the

environment might be changing (Ashill and Jobber 2010). In addition to state uncertainty, Ashill

20
and Jobber (2010) measure PEU as a composite construct made up of subcomponents including

effect uncertainty (i.e., lack of understanding of cause and effect relationships) and response

uncertainty (i.e., inability to predict likely consequences). Given the effect of PEU documented

here, future research should examine how these PEU subcomponents interact with control

systems and incentives to impact managers’ WTRC.

Second, use of an experimental approach necessitates limiting the information set that is

available to participants to minimize participants’ time demands and simplifying design choices

to maximize the control over and precision of the research questions and analyses. In actual

firms, managers would normally have access to a far richer information set compared to our

experiment. Future research could examine the effect of control systems on strategic change in

an increasingly rich and realistic information setting.

Third, somewhat related to the previous point, in our study participants learned about the

type of control system imposed in their organization, but did not actually interact with that

control system. Thus, we focus on the type of control system, itself, conveys information to the

manager that can illicit different assessments of market favorability and risk, and ultimately

behavior. We cannot, however, comment on how managers’ use of the various control systems

may differ and how subsequent feedback provided from the system might influence their

willingness to recommend change. In particular, diagnostic systems, with their focus on

outcomes, are likely to induce managers to perform a focused search of the environmental

characteristics and their relationship to the firm. Therefore, the metrics of the diagnostic system

can act as a conduit to understand important environmental issues and to determine whether

strategic change is necessary. Alternatively, interactive systems by design do not focus attention

on specific metrics. While interactive systems stimulate search, that search has a tendency to be

21
broad, un-focused, potentially random and driven by inductive thinking (Simons 1995).

Although it is likely that these differences in how diagnostic and interactive control systems

influence information search would result in the same predictions as those tested in this study,

we leave this to future research to explore empirically.

Fourth, participants in our study are exposed to either a diagnostic control system or an

interactive control system. However, Simons’ levers of control theory (1995) maintains that

organizations are best served by using all four control levers – diagnostic, interactive, boundary

and belief systems – to stimulate innovation and creativity. The levers are supposed to work

together and reinforce one another. Future research should explore the interplay between

controls levers, including the dynamic tension brought about by the joint use of diagnostic and

interactive control systems, and their effect on managers’ willingness to recommend strategic

change.

Fifth, the participants in our study are graduate business students. While graduate

business students have been used to proxy for managers in prior studies (e.g., Kadous and Sedor

2004; Hannan et al. 2006; Kelly 2007, 2010; Jackson 2008) and those employed here have

adequate experience (i.e., an average of 8.9 years of professional experience) for the

experimental task, the possibility still exists that actual managers in their natural work setting

may behave differently. Future research should examine the interaction of control systems and

PEU on actual managers to further develop our understanding of the influence that important

system-level structures (such as control systems) and individual difference variables (such as

PEU) have on managers’ wiliness to recommend change.

22
 REFERENCES
Abernethy, M. A., and P. Brownell. 1999. The role of budgets in organizations facing strategic
change: an exploratory study. Accounting, Organizations and Society 24 (3): 189-204.
Agbejule, A. 2005. The relationship between management accounting systems and perceived
environmental uncertainty on managerial performance: A research note. Accounting and
Business Research 35: 295-305.
Ahuja, G., and C.M. Lampert. 2001. Entrepreneurship in the large corporation: A longitudinal
study of how established firms create breakthrough inventions. Strategic Management
Journal 22: 521–543.
Ashill, N. J., and D. Jobber. 2010. Measuring state, effect, and response uncertainty: Theoretical
construct development and empirical validation. Journal of Management 36 (5): 1278-1308.
Bruining, H., M. Bonnet, and M. Wright. 2004. Management control systems and strategy
change in buyouts. Management Accounting Research 15 (2): 155-177.
Byrne, B. M. 2001. Structural Equation Modeling with AMOS: Basic Concepts, Applications,
and Programing. Mahwah, NJ: Lawrence Erlbaum Associates.
Chenhall, R. H. 2003. Management control systems design within its organizational context:
Findings from contingency-based research and directions for the future. Accounting,
Organizations and Society 28 (2-3): 127-168.
Christ, M. H., K. L. Sedatole, K. L. Towry and M. Thomas. 2008. When formal controls
undermine trust and coopeartion. Strategic Finance (January):38-44.
Christ, M. H., K. L., Sedatole and K. L. Towry. 2012. Sticks and carrots: The effect of contract
frame on effort in incomplete contracts. The Accounting Review 87(6): 1913-1938.
Committee of Sponsoring Organizations of the Treadway Commission. 2004. Enterprise Risk
Management – Integrated Framework. www.coso.org
Cummins, C. 2011. Losing ground, Blackberry resets. The Wall Street Journal (Online), July 26,
B1.
Ezzamel, M., S. Lilley, and H. Willmott. 2004. Accounting representation and the road to
commercial salvation. Accounting, Organizations and Society 29 (8): 783-813.
Ferreira, A., and D. Otley. 2009. The design and use of performance management systems: An
extended framework for analysis. Management Accounting Research 20 (4): 263-282.
Fisher, C. 1996. The impact of perceived environmental uncertainty and individual differences
on management information requirements: A research note. Accounting, Organizations, and
Society 21: 361-369.
Garg, V. K., B.A. Walters, and R.L. Priem. 2003. Chief executive scanning emphases,
environmental dynamism, and manufacturing firm performance. Strategic Management
Journal 24: 725-744.
Greve, H. R. 1998. Performance, aspirations and risky organizational change. Administrative
Science Quarterly 43: 58-86.

23
Greve, H. R. 2003. A behavioral theory of R&D expenditures and innovations: Evidence from
shipbuilding. Academy of Management Journal 46 (6): 685-702.
Gul, F. A., and Y.M. Chia. 1994. The effects of management accounting systems, perceived
environmental uncertainty and decentralization on managerial performance: A test of three-
way interaction. Accounting, Organizations and Society 19: 413-426.
Hannan, R. L., F.W. Rankin, and K.L. Towry. 2006. The effect of information systems on
honesty in managerial reporting: A behavioral perspective. Contemporary Accounting
Research 23: 885-918.
Henri, J. F. 2006. Management control systems and strategy: A resource based perspective.
Accounting, Organizations and Society 31: 529-558.
Jackson, S. B. 2008. The effect of firms’ depreciation method choice on managers’ capital
investment decisions. The Accounting Review 83: 351-376.
Kadous, K., and L.M. Sedor. 2004. The efficacy of third-party consultation in preventing
managerial escalation of commitment: The role of mental representations. Contemporary
Accounting Research 21: 55-82.
Kelly, K. 2007. Feedback and incentives on nonfinancial value drivers: Effects on managerial
decision making. Contemporary Accounting Research 24: 523-556.
Kelly, K. 2010. Accuracy of relative weights on multiple leading performance measures: Effects
on managerial performance and knowledge. Contemporary Accounting Research 27 (2):
577-608.
Kressler, H. 2003 Motivate and reward: Performance appraisal and incentive systems for
business success. New York: Palgrave Macmillan.
Langfield-Smith, K. 1997. Management control systems and strategy: A critical review.
Accounting, Organizations and Society 22: 207-232.
Langfield-Smith, K. 2006. A review of quantitative research in management control systems and
strategy. In Handbooks of Management Accounting Research, eds. C.S. Chapman, A.G.
Hopwood, and M.D. Shields, 753-783. Oxford, UK: Elsevier.
Mangaliso, M. P. 1995. The strategic usefulness of management information as perceived by
middle managers. Journal of Management 21: 231–250.
Milliken, F. J. 1987. Three types of perceived uncertainty about the environment: State, effect,
and response uncertainty. Academy of Management Review 12: 133-143.
Nadkarni, S., and P.S. Barr. 2008. Environmental context, managerial cognition and strategic
action: An integrated view. Strategic Management Journal 29: 1395-1427.
Ocasio, W. 1997. Towards an attention-based view of the firm. Strategic Management Journal
18: 187-206.
Ocasio, W., and J. Joseph. 2005. An attention-based theory of strategy formulation: Linking
micro- and macro perspectives in strategy processes. In Advances in Strategic Management:
Strategy Process,eds. G. Szulanski, J. Porac, and Y. Doz, (Vol. 22) 39-61. Bingley, UK:
Emerald.

24
Sawyerr, O. 1993. Environmental uncertainty and environmental scanning activities of Nigerian
manufacturing executives: A comparative analysis. Strategic Management Journal 14: 287-
299.
Simons, R. 1990. The role of management control systems in creating competitive advantage:
New perspectives. Accounting, Organizations and Society 15: 127-143.
Simons, R. 1991. Strategic orientation and top management attention to control systems.
Strategic Management Journal 12: 49-62.
Simons, R. 1994. How new top managers use control systems as levers of strategic renewal.
Strategic Management Journal 15: 169–189.
Simons, R. 1995. Levers of Control. Boston, MA: Harvard Business School Press.
Sitkin, S. B. and A. L. Pablo. 1992. Reconceptualizing the determinants of risk behavior.
Academy of Management Review 17(1):9-38.
Skærbæk, P., and K. Tryggestad. 2010. The role of accounting devices in performing corporate
strategy. Accounting, Organizations and Society 35: 108-124.
Sorensen, J. B., and T.E. Stuart. 2000. Aging, obsolescence, and organizational innovation.
Administrative Science Quarterly 45: 81-112.
Tessier, S., and D. Otley. 2012. A conceptual development of Simons' levers of control
framework. Management Accounting Research 23 (3): 171-185.
Tripsas, M., and G. Gavetti. 2000. Capabilities, cognition, and inertia: Evidence from digital
imaging. Strategic Management Journal 21: 1147–1161.
Tymon, Jr., W. G., D. E. Stout, and K.N. Shaw. 1998. Critical analysis and recommendations
regarding the role of perceived environmental uncertainty in behavioral accounting research.
Behavioral Research in Accounting 10: 23-23.

25
 Figure 1: Theoretical Model

H1 (-) H2 (-) Willingness to Recommend

Perceived Environmental
Market Favorability Strategic Change
Uncertainty (PEU)

H4 (+) H3 (-)

Control Systems
(Interactive vs. Diagnostic)

Hypotheses:
H1: There is a negative relationship between a manager’s PEU and that manager’s perception of market favorability.
H2: Managers who perceive market favorability as high are less likely to recommend strategic change than managers who perceive market favorability
to be low.
H3: Managers operating under Diagnostic Control systems are more willing to recommend strategic change than managers operating under Interactive
Control Systems.
H4: The negative effect of PEU on the manager’s perception of market favorability is moderated by the type of control system imposed. Specifically,
managers with low PEU will assess market stability and attractiveness lower under diagnostic control system as compared to an interactive control
26
system. And, managers with high PEU will assess market stability and attractiveness no different under a diagnostic control system compared to an
interactive control system.
 Figure 2: Tests of Theoretical Model

-0.35 (p<0.05) -0.19 (p<0.01)

Perceived Environmental Willingness to Recommend
Market Favorability
Uncertainty (PEU) Strategic Change

0.13 -0.14
(p<0.05) (p<0.05)

Control Systems
(Interactive vs. Diagnostic)

27
 Figure 3: Test of Mediation Analysis

-0.35 (p<0.05) -0.215 (p<0.01)

Perceived Environmental Willingness to Recommend
Market Favorability
Uncertainty (PEU) Strategic Change

0.13 0.39
-0.08(p=.135)
(p<0.05) (p<0.01)

Control Systems -0.16 (p<0.01) Control Systems

(Interactive vs. Diagnostic) (Interactive vs. Diagnostic)

28
 TABLE 1
Descriptive Statistics of Demographic Data

Panel A: Demographic Data

N Percent
Participants with public company experience 98 65
Participants with no public company experience 54 35

Mean (SD)
Professional work experience (in years) 8.93 (6.77)
Public company work experience (in years) 5.92 (5.71)
Experience with identifying and providing responses to 5.23 (1.17)
strategic uncertaintiesa
Familiarity with using a planning and budgeting system 5.32 (1.12)
for identifying areas of strategic focusb

a
Participants reported their experience with identifying and providing responses to strategic uncertainties on a seven-
point scale where 1 = not very familiar and 7 = very familiar.

29
 TABLE 2
Descriptive Statistics and Main Effects Tests

Panel A: Test of Main Effects on Willingness to Recommend Changea

Sig
Low PEU High PEU F statistic (one-tailed)
High vs. Low PEUb 4.21 4.68 3.04 0.04

Sig
Diagnostic Interactive F statistic (one-tailed)
Diagnostic vs. Interactive Control 4.75 4.29 3.27 0.04

Panel B: High PEU

Recommend Market Risk
Means (standard deviation in italics) Change Favorabilityc Assessmentd
Diagnostic Control 4.76 3.63 4.52
1.55 1.19 1.46
Interactive Control 4.60 3.88 4.16
1.47 1.35 1.46
Total 4.68 3.76 4.34
1.50 1.27 1.47

Panel C: Low PEU

Recommend Market Risk
Means (standard deviation in italics) Change Attractiveness Assessment
Diagnostic Control 4.73 4.90 4.50
1.56 1.12 1.63
Interactive Control 3.69 4.48 3.69
1.69 1.18 1.90
Total 4.21 4.70 4.10
1.70 1.16 1.80

a
To measure Willingness to Recommend Change participants responded on a seven-point scale where 1 = strongly
disagree and 7 = strongly agree to the question, “As the programming manager, you would recommend to the Board
of Directors that Atlanta Radio change its programming format to reach a larger share of its target market.”
b
PEU is perceived environmental uncertainty and represents a within-participant independent variable that is
measured and dichotomized at two levels (high, low) based on participants’ agreement with the following statement:
Atlanta radio market faces significant environmental uncertainty. Participants reported their agreement on a seven-
point scale where 1 = strongly disagree and 7 = strongly agree. Responses of 4 or greater were categorized as High
PEU.
C
Market Favorability is a composite measure that combines participants’ responses (on a 7 point Likert scale) to the
following questions regarding market attractiveness and market stability, respectively: “The Rock - Adult

30
Contemporary segment that Atlanta Radio competes in is attractive,” and “Atlanta Radio’s market share is likely to
remain stable.”
d
Risk Assessment is based on participants response to the question: “The financial and operational costs involved in making a
radio station format change is worth the potential reward.”

31