Teemu Malmi by panniuniu


									Accounting controls as a part of the
organizational control package

Teemu Malmi
Helsinki School of Economics


• Why we should care
• What do we know today
• What should be studied next – a research agenda
• Two examples of recent research on control
• Some thoughts about the road map
• Concluding comments

Why we should care?

• At the end of the day, we should be able to tell, which
  accounting controls to use, how and in which
• In research terms, this is to demonstrate which accounting
  controls, and how used, produce desired performance in
  which circumstances
• Control refers to those practices intended to gain
  congruence between organization’s strategic and other
  goals and organizational actor’s goals and activities
• Hence, there are lot of potential substitutes and
  complements for accounting controls, making the study of
  accounting controls alone potentially misleading
• It seems that in the context of decision making, accounting
  may be studied alone more often than in the context of

Why we should care?

• Manager is ultimately interested in achieving desired
  outcomes (what ever they may be) and to make sure non-
  desired actions are not performed
• This is to say that s/he is not interested accounting
  controls as such, but those controls, accounting included,
  that are most effective
• To give advise on which accounting controls to use, how
  and in which circumstances, other controls available are
  likely to be more important than say, for example, some of
  the traditional contingency factors
• In research, we might be better off if we try to explain the
  success of certain accounting controls by including other
  controls used in studied organizations into our models


• Contingency research has attempted to explain the
  form, use and performance implications of various
• Also a lot of research on “novel” accounting methods
• Usually a single system studied at a time, e.g. BSC,
• Otley (1980) and others (e.g. Flamholtz, Das & Tsui,
  1985; Dent, 1990; Fisher, 1998; Chenhall, 2003)
  have called for the study of controls as a package
• Chenhall 2003 states that only studying specific
  systems at time has “the potential for serious model
  under specification”

In modeling terms from this

    Contingency               Accounting
    factors                   control(s)

    To this



Design   Control
  &      package                   Performance

What do we know

• Limited number of empirical studies so far
    –   Macintosh & Daft, 1987
    –   Simons, 1995
    –   Abernathy & Chua, 1996
    –   Alvesson & Kärreman, 2004
• We know that many different controls are normally
  applied in large organizations
• M&D, 1987: Managers may make choices between
  controls and interdependencies of organizational
  areas may be an important consideration in overall
  design of MCS

What do we know

• Simons suggest four levers of controls and argue that
  successful firms find balance between these
• Abernathy & Chua, 1996
    – Governance structures part of control package
    – Organizations may use other controls to substitute
      accounting controls
    – Control package design partly rational, partly
    – Institutional environment influence controls used
• Alvesson & Kärreman, 2004
    – Socio-ideological control systems are tied to technocratic
    – These controls do not operate as alternatives, but as
    – Technocratic control systems provide the vehicle to
      communicate the messages of the socio-ideological control


• Why we should care
• What do we know today
• What should be studied next – a research agenda
• Two examples of recent research on control
• Some thoughts about the road map
• Concluding comments

What should be studied next?

• Somewhat limited scope of what controls are
  included in these studies
• First issue is to discuss what should be included into
  a concept of control package, i.e. how do we define
  this construct of interest
• Number of authors have provided their definitions /
  classifications: Ouchi, Otley, Flamholtz, Simons,
  Merchant & Van der Stede, etc.

What should be studied next?

• We do not have broad empirical evidence on what
  kind of control packages organizations use, how do
  they use those, and why
• Neither do we know the performance implications of
  various packages and their potential contingencies
• Are some controls more likely to remain stable
  through times than some others?
• Do managers choose among number of controls at
  time or do they make decisions regarding one control
  system in time (suggesting some form of path
  dependency theories)

What should be studied next?

• Systems can substitute and/or complement each
    – Which ones can complement / substitute which ones
      successfully in which circumstances?
    – The concept of equifinality
• Relations seem to exists, but none of the studies
  have conceptualized these relations or analyzed how
  systems influence each other or organizational

Two examples of recent research on
control packages

• Brown, Malmi & Booth study on controls of a global
  consumer products company
• Baker, Brown & Malmi study on environmental
  strategy implementation in an Energy company

Brown et al. – background and motivation

• A study of BSC in a large successful multinational,
  BSC existed but was hardly making a difference
• Bewildering array of controls in use
• Why so many? Are they all truly necessary? Are
  some more important than the others? Which ones
  are really used to drive business?
• We ended up asking how this whole system works /
  operates? To answer this, it seemed that we need to
  study how these systems relate to each other

Loose Coupling Theory

• Orton & Weick, 1990, AMR
• Distinctiveness
    – Focus
    – Use
    – Components
• Responsiveness
    – Dependence
    – Directness
    – Strength
• Degree of distinctiveness and responsiveness
  determine the type of coupling

                                                  Values                         Culture

                         Long Range                                OGSM

Strategy   Objectives


                         Action                                        Budget                           Organisation
                         Planning                                                                       Outcomes
                                                                                                        Short Term

                        Dashboard                                                          Financial
                                                       Scorecard                           Measures


                                                                                     Policies &
                                      Organisational                Governance
                                     Values                   Culture

                        Long Range                OGSM

Strategy   Objectives


                        Action                       Budget              Organisation
                        Planning                                         Outcomes
                                                                         Short Term
Results / insights from the analysis so far:

• There seemed to be ”suboptimal” designs regarding e.g.
   transfer pricing, i.e. accounting based controls of P&L to
   assess BU managers performance
• Still the company has been very successful in the past
• Long range planning, budgeting, value statements and
   OGSM influenced behavior as well
-> Suboptimal (accounting) controls can exists in parts of
   MCS package without major effects on organizational
   performance when this can be compensated by other
   controls in the package
->The case provides support to the idea that in order to
   understand the effectiveness of any individual controls
   (e.g., accounting controls), an understanding of what other
   controls are coupled to this is necessary

Results / insights from the analysis so far:

• Tension between short and long run outcomes
• This firm seems to manage this tension by coupling
  OGSM and budgets to strategy and objectives, but
  decoupling them from each others
• Both OGSM and budget have an impact on activities,
  some of which help to achieve long range objectives,
  some short run objectives
-> Can some classic trade-offs / paradoxes in
  management actually be solved by certain control
  system package designs
-> It may be that traditional contingency factors have
  an impact on the couplings of different controls as
  opposed to / in addition to certain controls directly

Results / insights from the analysis so far:

• Senior management did not have a broad understanding
   of the extent of MCS package in place
-> we may question to what extent these control packages
   are (today) results of rational decision making (designed)

Baker et al: Implementing environmental
strategy with MCS package

• Could accounting and control systems help in solving
  environmental problems?
    – Critical research skeptical
    – Most efforts in research focused on reporting as
      opposed to controlling/managing internally
• A case study of a well known industry leader in
  Energy sector in Australia
• What controls are used to implement environmental
• We provide some evidence that this control package
  actually impact on organizational activities


  COMMITTEE                   E                                 PI
                              I                                  s
                         2    P                                 an
 ENVIRONMENT             n                                       d
   STEERING              d                                       M
  COMMITTEE         1s
                         Pl                                     an
                         a                                      ag
                    Pl   n                                       e
 ENVIRONMENT        a    ni                                      m
 SERVICES UNIT      n    n                                      en
     (ESU)          ni   g                 T                     t
                    n    D        P        o                    Pe
                    g    a        o                             rfo
SENIOR MANAGER      D    y        li
                                           p               E
                                           -               n              e
                    a             c        d    T               an    P
                                                           v              C
                    y              i       o    r               ce    D
                                                           ir             u
                                  e        w    a               As    S
                                                           o              lt
PROJECT MANAGER                   s    E   n    i
                                  a    I   B    n               ss
                                                           m               r
                                  n    A   u    i                m
                                                           e              e
                                  d        d    n               en
                                  P             g                t
SUPERINTENDENTS                   r
                                           g               t
                                           e               a
                                  o        ti               l
                                  c        n               A
                                  e        g               u
  FIELD STAFF                     d

                -                                   TIME
Results / insights so far

• A large number of controls in use to make sure
  environmental strategy / plan is implemented
• Accounting based controls / cybernetic controls do not
  play a major role, but are part of the package used
    – Difficult to measure and set clear targets, and even if that is
      possible, how do these targets compare to economic
      targets? Do managers make trade-offs to boost profits at the
      expense of activities directed towards saving the
• Organization made themselves accountable to
  stakeholders by involving them to the environmental
  strategy work
    – This is contrary to the claims of critical research which tends
      to argue accounting is used for signaling but is decoupled
      from activities; or that some systems are used for
-     management and some for signaling, but decoupled from
Results / insights so far

• These controls were used as complements, in this
  single case setting we can not really address the
  question whether all these are truly necessary to
  implement environmental strategy in this kind of a
• But this might suggest that using e.g. only BSC to
  translate environmental strategy into action might not
  be enough
-> Is it more likely that by relying on only one control
  tool we end up with situations where this single tool
  is decoupled from real activities

Some thoughts on the road map

• The result of this research on management controls as a
  package should be a theory of MCS
• It is not likely that such theory will be only economics,
  sociology or psychology based
• To explain why certain packages produce better
  performance may require insights from all these major
  theoretical approaches
• We may call it management control theory as it should
  explain (and predict) the designs (choices) of control
  packages and their associated outcomes given various
• Single empirical research pieces should make it clear how
  they contribute to this development of MCS theory

Some thoughts on the road map

• Is this a research program ending up of being “mile wide
  but only an inch deep”?
• By clever research design, not necessarily all controls in a
  package needs to be studied at the same time
• As package contains many elements, can survey
  instruments be designed and used to collect data
    – Large research groups aiming for large case samples one
      option for data collection
• How to measure culture based controls?
• Requires cross-disciplinary mind set and skills
• As it seems fairly obvious that control packages we will
  observe are not really a result of intentional design,
  perhaps less effort should be directed in explaining why
  they are as they are and more to what combinations work,
  how used and in which circumstances

Concluding comments

• Management accounting is an applied science, our
  research should help the effective management of
• To be able to say something meaningful regarding
  accounting controls requires an understanding of
  potential substitutes and complements
• Today we have little theory or empirical evidence on
• Despite some practical challenges, I believe this
  provides us a huge research opportunity as well as
  potentially strong identity as an academic discipline


To top