Applying the controllability principle andmeasuring divisional performance in UK companies
Research Executive Summaries Series
Vol. 1, No. 8
byProf. C Drury, Huddersfield UniversityH EL-shishini, Huddersfield University
ISSN 1744 - 7038 (online)
ISSN 1744 - 702X (print)
Applying the controllability principle and measuring divisional performance in UK companies
Aim of this executive summaryThis executive summary aims to provide a summary for practitioners of the authors’CIMA-funded research project published as a research report under the title ‘Divisional performance measurement:An Examination of the Potential ExplanatoryFactors’.
Key issues underlying the research projectThe key issue underlying the research project is whether or not divisional managersare held accountable for items which they cannot influence by their actions. Theconventional wisdom of management accounting, as reflected in textbooks,advocates the application of the ‘controllability principle,’ which is that theevaluation of a manager’s performance should be based only on those factors thatare under a manager’s control. The application of the controllability principle todivisional performance measurement results in the need to distinguish between theeconomic performance of divisions and the performance of divisional managers. It isadvocated that different performance measures should be used to evaluate theeconomic performance of the divisions and the performance of divisional managers.A separate divisional managerial performance measure applies the controllabilityprinciple by excluding those costs that cannot be controlled or influenced by adivisional manager whereas divisional economic measures generally include theallocation of uncontrollable costs based on the principle that, if the divisions wereindependent companies, they would have to bear such costs.
The limited empirical evidence, however, suggests that the allocation ofuncontrollable costs for evaluation of managerial performance is widespread andthat the controllability principle often does not appear to be applied in practice. It isapparent that the traditional two-fold classification of costs as being controllable ornon-controllable is too simplistic and that the application of the controllabilityprinciple lies along a continuum.At one extreme there is no application of thecontrollability principle with companies holding managers responsible for alluncontrollable factors.At the other extreme there is the full application of thecontrollability principle where companies tend to hold divisional managersresponsible only for controllable factors. In between these extremes managers maybe held accountable for some uncontrollable factors and not others.
The choice of appropriate measures of divisional managerial performance has beenwidely debated in the management accounting literature. One area of debate is theextent to which the controllability principle should be applied. Should differentperformance measures be used to evaluate the performance of divisional managersand the economic performance of the divisions or should a single measure be used forboth purposes? Another area of debate relates to the choice of appropriateperformance measures. The debate was concerned with which traditional financialmeasures (e.g. net profit before or after taxes, controllable profit, residual income,return on investment) should be used. Over the past decade new measures haveemerged such as economic value added® (EVA®) and the balanced scorecard.Non-financial measures have also been given more prominence and the relativeemphasis that should be given to financial and non-financial measures, and how theyshould be integrated, have been subject to debate.
Executive Summary
2
A copy of the full report isavailable for download fromthe CIMA website;www.cimaglobal.com
Applying the controllability principle and measuring divisional performance in UK companies3
In view of these developments, and the fact that most of theprevious empirical research was undertaken over 20 years agowhen the business environment and management practiceswere very different from those existing today, it wasconsidered appropriate to undertake empirical researchrelating to of the application of the controllability principleand divisional performance measurement in UK companies.
Aims of the research projectThe aims of the research project were therefore as follows:• to investigate how far financial and non-financial measures
are used in practice to evaluate the performance of divisionalmanagers;
• to identify the extent that the ‘controllability principle’ isapplied to factors that are identified as uncontrollable;
• to identify the influence of level of application of thecontrollability principle on the degree of satisfaction withthe divisional performance measurement system; and
• to investigate the relationship between the use of non-financial measures and the degree of satisfaction withthe performance measurement system.
Content of the research reportA large part of the research report describes the growth andimportance of divisionalised structures (not summarisedfurther here), the different financial measures that are used inpractice to evaluate divisional performance and the differentcategories of uncontrollable costs. It then describes, presentsand analyses the research findings.
Conduct of the research projectThe research project involved analysing the results of 124usable returned completed postal questionnaires frommanufacturing companies with an annual turnover over £100million. The analysis was based on annual sales turnover anddivisionalised structures (including listed and unlistedcompanies and subsidiaries or divisions of overseascompanies).The respondents were located in finance sections,at head office and in divisions.
Background to the researchTo understand why the research project focused on thespecified objectives it is appropriate to look at some of theunderlying information contained in the report.
Divisional performance measuresThe report addresses how divisional performance can bemeasured.
Distinguishing between economic and managerialdivisional performanceCorporate management require performance measures toassist in evaluating the economic performance of a division asa whole and its economic viability and future direction.Divisional profit in this case usually, but not always, includes allallocated central costs. In addition, measures are required toevaluate the performance of divisional managers. Forevaluating managerial performance conventional wisdomadvocates that performance evaluation should be limited tothe profit that is controllable or at least influenced by thatmanager’s actions. Examples of costs that would be excludedfrom controllable profit include foreign exchange ratefluctuations and allocated central administrative expenses.
The measure traditionally used to measure divisionalperformance has been return on investment (ROI), whichmeasures divisional profit against the level of investment inassets attributable to a division. However, if ROI is used toevaluate managerial divisional performance it can lead to sub-optimal decisions. For example, divisional managers maybe incorrectly motivated not to undertake a project with areturn in excess of the cost of capital simply because it has alower projected ROI than the current ROI for the division as awhole.To help overcome this problem conventional wisdomadvocated the use of controllable residual income (RI), whichis controllable profit less a charge for the cost of capitalemployed. Previous empirical studies, however, suggest that RIdoes not appear to be widely used in practice.
Limitations of financial performance measures The historical limitations of financial performance measurescan be summarised as follows:• they focus on short-term periods and encourage
short-termism;• they are lagging indicators, evaluating the outcomes of
management actions after a time period. Therefore it isdifficult to establish a relationship between managers’actions and reported financial results;
• they deal only with the current period, not the future; and• they use financial accounting information based on the
historical cost concept and thus tend to be poor estimates ofeconomic performance.
Given that the problems associated with the use of financialperformance measures, two possible methods of dealing withthem emerged in the early 1990’s.The first (economic valueadded) seeks to improve financial performance measures andthe second (the balanced scorecard) incorporates non-financial performance measures with financialperformance measures.
Economic value added (EVA)During the early 1990’s Stern Stewart and Co., a New York-based consulting firm, repackaged and refined residualincome in the form of economic value added® (EVA®).Theobjective of EVA is to develop a performance measure thataccounts for the ways in which corporate value can be addedor lost.Thus, by linking divisional performance to EVAmanagers are motivated to focus on increasing shareholdervalue.The EVA concept extends the traditional residual incomemeasure by incorporating adjustments to the divisionalfinancial performance measure for distortions introduced byGAAP.
EVA can be defined as:
EVA = Conventional divisional profit ± accountingadjustments - cost of capital charge on divisional assets
Adjustments are made to the chosen conventional divisionalprofit measure (e.g. controllable profit, net income) in order toreplace historic accounting data with a measure thatapproximates economic profit and asset values.Theseadjustments result in the capitalisation of many discretionaryexpenditures (e.g. research and development, marketing andadvertising) by spreading these costs over the periods in whichthe benefits are received.
Also by taking into account all the capital costs, economicvalued added attempts to show the amount of wealth abusiness created or destroyed in each period.
The balanced scorecardNon-financial measures, such as of competitiveness, productleadership, productivity, quality, delivery performance,innovation and flexibility have long been advocated as afurther way to prevent short-termism. Historicallyincorporating non-financial measures have presented twoproblems:• there are too many of them; and• they often tend to conflict with each other
The balanced scorecard, developed by Kaplan and Norton,seeks to integrate financial and non-financial performancemeasures and identify key ones that link to strategy. Measuresare developed based on four perspectives that seek to answerfour questions:
The balanced scorecard involves establishing major objectivesfor each of the four perspectives, translating each objectiveinto targeted performance measures and comparing actualperformance measures with the target measures.A criticalassumption of the balanced scorecard is that eachperformance measure is part of a cause-and-effectrelationship involving a linkage from strategy formulation tofinancial outcomes.
The balanced scorecard thus consists of two types ofperformance measures. The first consists of lagging measures.These are the financial (outcome) measures within thefinancial perspective that are the results of past actions. Theyshow the financial impact of the decisions as their impactmaterialises and this can be long after the decisions weremade. The second are leading measures that are the drivers offuture financial performance. These are the non-financialmeasures relating to the customer, internal business processand learning and growth perspectives.
Applying the controllability principle and measuring divisional performance in UK companies 4
Perspective
Customer
Internal businessprocesses
Learning and growth
Financial
Question
How do our customers see us?
What must we excel at?
Can we continue to improve and createvalue?
How do we look to our shareholders?
Applying the controllability principle and measuring divisional performance in UK companies5
Controllability and cost allocationsThe report addresses what is meant by controllability,uncontrollable factors and non-controllable costs, and whythese latter might be allocated when evaluating divisionalmanagerial divisional performance.
The controllability principle and non-controllable factorsThe controllability principle states that:‘managers should be held accountable only for the results thatthey can significantly influence.’
This suggests there is a need to distinguish betweencontrollable and uncontrollable factors. Uncontrollablefactors can be classified into the following categories:• economic and competitive factors, to which managers have
to react;• acts of nature, which are beyond management control;• costs that are uncontrollable by divisional managers such as
group head office general and administrative costs, taxes,interest and corporate costs; and
• divisional interdependencies whereby the actions ofdivisions impact on each other.
While the effects of acts of nature tend to be treated asuncontrollable, the treatment of the remaining factors variesin practice.
Distinguishing between controllable and uncontrollablecostsApplying the controllability principle is not an easy taskbecause control typically lies along a continuum with twoextremes: full control and no control; and in between there aredifferent degrees of control.Conventional wisdom provides thefollowing guidance for distinguishing between controllableand uncontrollable items:
1. If a manager can control the quantity and price paid for aservice then the manager is responsible for all theexpenditure incurred for the services and the expenditure isfully controllable.
2. If the manager can control the quantity of the service butnot the price paid for the service then the costs are partiallycontrollable and only that amount of difference betweenactual and budgeted expenditure that is due to usage shouldbe identified with the manager.
3. If the manager cannot control either the quantity or theprice paid for the services then the expenditure isuncontrollable and should not be identified with themanager.
A more simplistic general guideline is to ‘Hold managersaccountable for the performance areas you want them to payattention to.’
Various approaches can be used to apply the controllabilityprinciple. Divisional managerial performance can be evaluatedby comparing actual performance against budget culminatingwith a bottom line that represents the chosen divisional profitmeasure (e.g. controllable profit, net income, EVA).Controllability is imposed by making managers accountablefor the variance between actual and budgeted outcomes.Uncontrollability can be fully recognised by excluding non-controllable items from the performance report orpartially recognised by ensuring that the manager is not madeaccountable for the variance by assigning budgeted costsinstead of actual costs. In the former situation the divisionaltarget and actual profit measure is not reduced whereas thelatter results in divisional profit being reduced by the budgetedallocation.
If managers are allocated with the actual costs foruncontrollable items the controllability principle is clearly notbeing applied. However, if managers are allocated withbudgeted costs it could be argued that the controllabilityprinciple is being partially applied by not making managersaccountable for the variance. The allocation of budgeteduncontrollable costs can be viewed in these circumstances asrepresenting a means of increasing the target profit measureto ensure that divisional profits are sufficient to coverbudgeted allocated uncontrollable costs. Alternatively, insteadof relying on cost allocations a higher budgeted target for thechosen divisional performance measure that takes intoaccount these factors can be set.Thus, problems that may arisefrom allocating uncontrollable costs are avoided.
Reasons for the allocation of uncontrollable costsWhy do companies continue to allocate costs that areuncontrollable? Rationales for cost allocations have beenbased on behavioural and institutional factors. The followingbehavioural factors have been identified:• makes managers aware that such costs exist and must be
covered by divisional profits (although as indicated above thiscould be achieved just by increasing the level of target profit);
• stimulates managers to put pressure on resource centremanagers to control their costs;
• puts full business risk onto divisional managers, as if theywere managers of non-divisionalised companies; and
• induces divisional managers to take a greater interest in thecosts of shared resources.
Institutional reasons include:• legal requirements (e.g. taxation of subsidiaries);• the requirements of government contracts (i.e. based on full
cost);• tradition (allocation is done by accountants who regard cost
allocation as part of their tradition); and• to aid inter-division and inter-competitor comparison
to tie in with the link between top executive pay andmeasures of accounting profit.
Reasons for non-allocation of common resource costsCommon resource costs may not be allocated because:• the amounts are too small to warrant allocation;• they are not controllable by divisional managers;• the costs of making the allocations would exceed their
benefits;• divisional managers object to charges they cannot influence
and control; and• allocations are arbitrary and tend to distort divisional profits
unnecessary internal tension can be avoided by no allocation.
The research project findings
Measuring managerial and economic divisionalperformanceOf the responding companies, 38% do not apply thecontrollability principle as they do not distinguish betweenmanagerial and economic divisional performance.The sameperformance measure was used by 44% but different itemswere used within the same performance measures – thus theywere making at least some attempt to apply the controllabilityprinciple.The remaining companies (18%) sought to apply thecontrollability principle by using different performancemeasures for economic and managerial performance .
There is a strong preference for an absolute targetperformance measure rather than a ratio/percentage measurefor measuring managerial performance.The three mostimportant measures are:• target profit before charging interest on capital was
considered to be the most important measure by 55% of theorganisations;
• target profit after charging interest on capital (residualincome) was ranked as the most important measure by 14%of the respondents; and
• 9% ranked a version of EVA as the most important measure(EVA was used by 23% of the respondents, with another 11%planning to introduce it)
Interestingly, a target return on capital employed measure (i.e.ROCE / ROI) was ranked as the most important measure byonly 7% of the respondents. It would appear that there is astrong preference for the use of absolute performancemeasures rather than ratio measures. A possible explanationfor this observation is that central headquarters’ requiredivisional managers to focus on maximising an absolute valuerather than a percentage return.
The research confirms that both ‘head office’ and divisionalmanagers have considerable involvement in setting financialtargets.
Applying the controllability principle and measuring divisional performance in UK companies 6
Applying the controllability principle and measuring divisional performance in UK companies7
In attempting to find out more about uncontrollable costs theresearch project divides common resource costs into threecategories and indicates the percentage of companies whichdescribe themselves as falling into each category.The tablebelow describes these categories, and indicates howconventional wisdom advocates how the controllabilityprinciple should be applied.
Table 6.3 (see below) of the research report indicates howcommon resource costs with different degrees ofcontrollability and uncontrollable general/administrationcosts are allocated.This table presents clear evidence that themajority of companies do not apply the controllabilityprinciple to the assignment of common resource costs, asabout 55% allocate all or most uncontrollable commonresource costs to divisional managers when measuringdivisional performance.
Of the other non-controllable factors, the research shows thatabout 86% of the respondents took very little account of theeffect of uncontrollable environmental factors (such aschanging economic conditions, competitor actions andbusiness climates) and divisional interdependencies whenallocating costs. In addition, non-controllable factors were noteven taken into account informally, even by companies whichallocated uncontrollable costs in full.
Neither the location of the head office (UK or overseas), thefact that the company is listed on the stock exchange nor theinformal consideration of non-controllable factors were foundto explain the level of application of the controllabilityprinciple.
Using non-financial measuresMost respondents do not rely on financial measures alone toevaluate divisional performance: non-financial measures wereused by 78% of respondents.A balanced-scorecard approachwas very popular, being used by 42% of all respondents.Financial measures were, however, considered to be moreimportant than non-financial measures by approximately 70%of the respondents, although as the respondents were locatedin the finance function there may be bias here.
Divisional use of common resources and other costsCommon resources are defined as resources or servicesprovided centrally to two or more divisions. They include suchitems as data processing, research and development,marketing, training, purchasing, personnel, accounting,internal audit, legal services and planning.A great majority(95.2%) of respondents stated that divisions used commonresources, but 73% of respondents indicated that the costswere less than 10% of divisional turnover.
Common resources are distinguished from group business-sustaining general and administration costs.Thelatter are completely uncontrollable by divisional managers,and include such costs as depreciation on buildings andequipment at head office.
Application of the controllability principle to non-controllable factorsA large majority of companies (83%) allocate at least somecommon resource costs to divisional managers whencomputing performance measures, although the greater thedegree of decentralisation, the lower the amount of commonresources allocated.
The extent of costs allocated Controllable common Partially controllable Non-controllable Business-sustaining to divisions resource costs commons resource costs common resource costs corporate general
(full autonomy - (limited autonomy - (No autonomy - administrative costscategory 1) category 2) category 3)
No. % No. % No. % No. %
(a) None of the costs are allocated - - - - 23 31.1 18 15.5
(b) Only a minor portion of the costs 6 54.5 2 3.8 3 4.0 12 10.3are allocated
(c) A small but significant portion of 1 9.1 10 18.9 8 10.8 12 10.3the costs are allocated
(d) Most, but not all of the costs, are 3 27.3 19 35.8 11 14.9 25 21.6allocated
(e) All of the costs are allocated 1 9.1 22 41.5 29 39.2 49 42.3
Total 11 100 53 100 74 100 116 100
Table 6.3 The extent to which costs of common resources and business-sustaining corporate general administrative costs are allocated to divisions
Why do most companies allocate all or most non-controllable costs?A number of potential reasons are indicated for whycompanies allocate all or most non-controllable costs.Table6.5 from the research report shows the results derived from a7-point scale (ranging from a score of 7 indicating extremelyimportant to 1 indicating not at all important).
The most important reasons for allocating common resourcecosts were behavioural, and relate to an attempt to useallocations to make managers aware of the costs and to actas surrogates for the costs that would be incurred if thedivisions operated as independent companies. Measurementand institutional reasons were least important.
Applying the controllability principle and measuring divisional performance in UK companies 8
Table 6.5 Importance of factors influencing organisations to allocate the costs of shared resources
Rank
Reas
ons
Nat
ure
of th
e %
of r
espo
nses
%
of r
espo
nses
N
umbe
rSu
m (t
otal
scor
e)M
ean
Stan
dard
reas
ons
ratin
g 1
or 2
ratin
g 6
or 7
devi
atio
n
1**
To sh
ow d
ivis
iona
l man
ager
s the
tota
l cos
t of
B***
3.1
70.1
9756
35.
801.
15op
erat
ing
thei
r div
isio
ns (a
)*
2To
mak
e di
visi
onal
man
ager
s aw
are
that
such
B
5.1
62.9
9754
75.
641.
26co
sts e
xist
s and
mus
t be
cove
red
by
divi
sion
al p
rofit
s (g)
3D
ivis
iona
l man
ager
s wou
ld in
cur s
uch
cost
s if
B6.
247
.497
503
5.19
1.47
they
wer
e in
depe
nden
t uni
ts (m
)
4D
ivis
iona
l man
ager
s sho
uld
bear
the
full
B6.
243
.397
495
5.10
1.29
busi
ness
risk
as t
hey
wer
e m
anag
ers o
f no
n-di
visi
onal
ised
com
pani
es (c
)
5To
stim
ulat
e di
visi
onal
man
ager
s to
put p
ress
ure
B12
.345
.397
471
4.86
1.79
on re
sour
ces c
entr
e m
anag
ers t
o co
ntro
l th
eir c
osts
(j)
6To
indu
ce d
ivis
iona
l man
ager
s to
take
gre
ater
B
14.4
43.3
9746
84.
821.
71in
tere
st in
the
cost
s of s
hare
d re
sour
ces (
k)
7To
ena
ble
inte
r-di
visi
on o
r int
er-f
irm
I22
.629
.997
422
4.35
1.85
com
paris
ons t
o be
mad
e (e
)
8To
pro
vide
sign
als o
n th
e ef
ficie
ncy
of
B23
.726
.897
404
4.16
1.77
serv
ice
depa
rtm
ent t
hat p
rovi
des
shar
ed re
sour
ces (
f)
9To
stim
ulat
e di
visi
onal
man
ager
s to
B26
.824
.797
394
4.06
1.76
econ
omis
e in
thei
r usa
ge o
f sha
red
reso
urce
s (h)
10D
ivis
iona
l man
ager
s con
trol
the
usag
e of
B
29.9
18.6
9735
63.
671.
75th
e re
sour
ces (
l)
11D
istin
guis
hing
bet
wee
n co
ntro
llabl
e an
d D
24.7
8.2
9735
03.
611.
52un
cont
rolla
ble
requ
ires s
ubje
ctiv
e ju
dgem
ents
whi
ch c
an c
reat
e co
nflic
ts (b
)
12It
’s ex
trem
ely
diff
icul
t to
sepa
rate
con
trol
labl
e D
33.0
13.4
9734
63.
571.
73an
d un
cont
rolla
ble
elem
ents
(d)
13Be
caus
e co
st a
lloca
tions
are
par
t of a
I
50.5
11.3
9728
42.
931.
87co
mpa
ny’s
trad
ition
(i)
*Le
tter
s rep
rese
nt th
e ra
nkin
g of
the
ques
tionn
aire
**Th
e ra
nk w
as m
ade
in d
esce
ndin
g or
der
***
B=
beh
avio
ural
dim
ensi
ons,
I=
inst
itutio
nal d
imen
sion
s and
D=
diff
icul
ty o
f sep
arat
ing
cont
rolla
ble
from
non
-con
trol
labl
e el
emen
ts.
Why do some companies not allocate common resourcecosts?Although most respondents do make such allocations, theones who did not cited primarily that common corporate costsare not controllable by divisional managers, and that divisionalmanagers object to charges they can neither influence norcontrol.
Allocating the variance between budget and actual forcommon resource costsA further aspect of the application of the controllabilityprinciple was investigated by examining how the variancebetween budgeted and actual allocated uncontrollablecommon costs was dealt with. Within the uncontrollable costcategory 70% of divisional managers were not heldaccountable for the variance. This indicates that thecontrollability principle is applied here, by protecting managersfrom differences arising from inefficiencies occurring outsidetheir division.
Satisfaction with the performance measurement systemThe research project found that there was no generaldissatisfaction with the system. In particular, managers withgreater autonomy of common resource costs did not have agreater degree of satisfaction with the performancemeasurement system. Similarly, managers for whom thecontrollability principle was not applied were no more satisfiedor dissatisfied with the system. However, the greater the useof non-financial measures the less satisfied were therespondents with the system. Suggested reasons for this are:• preparing non-financial measures may involve some degree
of judgement or conflict between divisional managers andtop managers, reducing satisfaction, or
• emphasising both types of measure may lead to confusion,especially if they give conflicting messages.
Relation between applying the controllability principleand the use of non-financial measuresThere is no evidence to suggest that the lower the level ofapplication of the controllability principly, the greater the useof non-financial measures.
ConclusionThe research report provides useful insights into:• how far companies distinguish between managerial and
economic divisional performance;• the relative importance of financial and non-financial
measures;• the preferred financial measures for managerial performance;• the fact that most companies apply the controllability
principle in some situations at least, but not in others;• the fact that, rather than simply raising target profit to cover
the costs, companies prefer to allocate budgeteduncontrollable costs;
• the fact that the controllability principle is seen as importantbut is applied flexibly; and
• the impact of the use of the controllability principle and non-financial measures of the satisfaction with theperformance and measurement system.
9 Applying the controllability principle and measuring divisional performance in UK companies
Copyright © CIMA 2005First published in 2005 by:The Chartered Institute of Management Accountants26 Chapter StreetLondon SW1P 4NP
Printed in Great Britain
The publishers of this document consider that it is a worthwhile contribution to discussion, without necessarily sharing the views expressed which are those of the authors.
No responsibility for loss occasioned to any person acting or refraining from action as a result of any material in this publication can be accepted by the author or publishers.
All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, method or device, electronic(whether now or hereafter known or developed), mechanical, photocopying, recorded or otherwise, without the prior permission of the publishers.
Translation requests should be submitted to CIMA.
The Chartered Institute of Management Accountants26 Chapter StreetLondon SW1P 4NPT +44 (0)20 7663 5441F +44 (0)20 7663 5442EE [email protected]
CIMA (The Chartered Institute of Management Accountants) represents members and supportsthe wider financial management and business community. Its key activities relate to businessstrategy, information strategy and financial strategy. Its focus is to qualify students, to supportboth members and employers and to protect the public interest.
May 2004
REF: TE003V0705
Recommended
