See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/228301293

Divisional Performance Measurement: An Examination Of The Potential

Explanatory Factors

Article in SSRN Electronic Journal · March 2001

DOI: 10.2139/ssrn.268408

CITATIONS READS

22 7,292

1 author:

Hatem El-Shishini
University of Bahrain
12 PUBLICATIONS 72 CITATIONS

SEE PROFILE

All content following this page was uploaded by Hatem El-Shishini on 12 November 2017.

The user has requested enhancement of the downloaded file.

Divisional Performance Measurement: An Examination of the Potential
Explanatory Factors

Hatem El-shishini
Abstract

Surveys indicate that the majority of companies in the UK have adopted divisionalised organisational structure.
The manner in which divisions are controlled and evaluated is therefore of considerable importance. However,
much of the research relating to divisional performance measurement is approximately 20 years old.
The management accounting literature distinguish between the economic performance of a division and the
performance of its manager, advocating that the evaluation of a manager’s performance should consist only of
those factors under a manager’s control. However, empirical evidence showed that companies tend to hold
divisional managers responsible for uncontrollable factors (for the purpose of this research uncontrollable factors
include three categories: (1) uncontrollable common resources costs; (2) economic and competitive conditions
and (3) interdependencies among divisions).

This research aims to explain the factors that lead to hold divisional managers responsible for uncontrollable
items. In particular, the research addresses the following issues:
1. The extent to which companies use different measures to evaluate the economic performance of the
divisions and the performance of divisional managers;

2. The nature and content of the performance measures that are used to evaluate divisional managers’
performance;

3. The level of application of the controllability principle in terms of the above three categories of
uncontrollable factors;

4. The role of the potential explanatory factors in explaining the observed practice. They include behavioural,
institutional and organisational factors; and

5. The relative importance attached to non-financial performance measures that are used to evaluate divisional
managers’ performance.

A postal questionnaire survey has undertaken to the UK-based divisionalised companies during April and May
2000. A total usable questionnaire was 124. The preliminary research findings identify many different levels of
the application of the controllability principle. At one extreme some companies hold divisional managers
responsible for all uncontrollable factors. At the other extreme some companies seek to hold divisional managers
responsible only for controllable factors.

The paper reports on the preliminary research findings. Please do not quote without
the permission. Comments are welcome.

Correspondence address:
Hatem El- shishini
Huddersfield University Business School
Accountancy Department
Queensgate, Huddersfield, HD1 3DH, UK
Tel. +44(0) 1484 472840, Fax. +44(0) 1484 473062,
E-mail: h.m.shishini@hud.ac.uk
Introduction
Surveys in the UK (Scapens and Sale, 1982; Drury et al., 1993), USA (Reece and Cool, 1978) and
Australia and New Zealand (Skinner, 1990) indicate that the majority of companies in these countries
have adopted divisionalised organisational structures. The manner in which divisions are controlled
and evaluated is therefore of considerable importance.

During the 1970’s the appropriate measure of divisional performance was widely debated in the
academic management accounting literature (see Emmanuel and Otley, 1976 for a review of the
debate). The conventional wisdom that has emerged from this debate suggests that for investment
centres where divisional managers have significant authority for making capital investment decisions,
or those profit centres where divisional managers can influence significantly influence the investment
in working capital, residual income is the most appropriate financial measure of divisional
performance. For those profit centres where divisional managers cannot influence the investment in
working capital conventional wisdom advocates that return on investment or target absolute profit
(normally derived from assets employed in the division multiplied by a target return on investment)
should be used.
Until recently conventional wisdom, as portrayed in textbooks (e.g. Drury, 1996; Kaplan and
Atkinson, 1989; Horngren et al., 1999) advocated that residual income was the major financial
measure that should be used for evaluating divisional performance. More recently, residual income
has been repackaged and refined in the form of economic value added (EVA) and is now being
promoted by consultancy organisations. However, little information is available on the extent to
which EVA is used in organisations.
Previous research, prior to the emergence of EVA, suggests that residual income has not been
extensively used in practice. Reece and Cool (1978) in a USA Study reported that only 2% of the
responding companies relied solely on residual income and a further 28% used both residual income
and return on investment (ROI). The position in The UK is somewhat similar with Scapens and Sale
(1982) reporting that 37% of the responding companies used residual income whereas the
Australian/New Zealand study by Skinner (1990) reported only 7% usage. It is apparent from the
empirical studies that a diversity of divisional financial performance measures are used, the most
popular being return on investment (ROI), profit before interest and taxes and the ability to stay
within the budget.
Much of the research relating to divisional performance measurement is approximately 20 years old.
In addition, some important issues have not been addressed by previous research. The management
accounting literature (e.g. Merchant, 1998) distinguishes between the economic performance of a
division and the performance of its manager, advocating that the evaluation of a manager’s
performance should consist of only those factors under a manager’s control. Therefore, divisional
performance measures (whatever the measure in use) should include only the controllable items by
divisional managers. Merchant (1987; 1989) provides a practical analysis relating to the application
of the controllability principle. He identified three categories of uncontrollables: (1) uncontrollable
common resources costs; (2) economic and competitive conditions; (3) acts of nature such as
earthquakes and accidents. In practice, acts of nature are likely to be rare.

McNally (1980) and Skinner (1990) suggest that the application of controllability principle has
different levels and imply that it is likely to be a continuum. At one extreme there is no application
of controllability principle with companies holding divisional managers responsible for all
uncontrollable cost and including all the effects of environmental uncertainty and organisational
interdependencies. At the other extreme there is the full application of controllability principle where
companies tend to hold divisional managers responsible only on controllable factors. In between
these extremes managers may be held accountable for some uncontrollable and not for others.
Previous research has not given much attention to identifying where companies fall within this

2
continuum in the application of the controllability principle or the factors influencing the different
applications of the controllability principle.
Drawing off the literature (e.g. Choudhury, 1986; Merchant, 1989) sources of uncontrollables can be
classified into two main groups: internal and external sources. The internal sources are generated
inside the firm. Such factors include role conflict, task ambiguity, interdependencies and the
allocation of non-controllable costs. They arise from the vague specification of organisational
functions, authority and reward systems because it is sometimes difficult to place responsibility
across time and divisions. In contrast, external sources are mainly due to the environmental factors.

It has been debated in the literature whether or not divisional managers should be held accountable
for uncontrollable items. Advocators of the controllability principle argue that divisional performance
measurement should involve only those items controllable by divisional managers. Despite the
recommendation of management accounting literature, it would appear that many companies tend to
disregard controllability principle (Merchant, 1989). For example, a UK study by Ramadan (1985;
1989) and a New Zealand by Skinner (1990) reported that many companies allocated uncontrollable
common costs to divisions.
Based on the above discussion this research focuses on different categories of uncontrollables. They
include:
(1) uncontrollable common resource costs;
(2) economic and competitive conditions;
(3) interdependencies among divisions; and
(4) Business-sustaining corporate headquarter costs.
This paper will focus primarily on the treatment of the costs of common resources.
Costs of common resources
For the purpose of this research, the term ‘common resources’ applies to resources or services
provided by the head office for the benefit of two or more divisions within the organisation. Common
resources costs include central costs relating to activities such as data processing; research and
development; marketing services; training programmes; purchasing; personnel; accounting; internal
auditing; legal services; and group planning.1 Most of the previous research has tended to consider
the terms ‘indirect corporate costs’ and ‘central service costs’ as being synonymous with common
resource costs. The research has tended not to distinguish between common resource costs and
corporate business-sustaining costs. However, this research distinguishes between common resource
costs and general and administrative costs. The latter refers to corporate business-sustaining costs
that are incurred for the benefit of the company as a whole. Examples for such costs include
depreciation of the headquarters building and equipment, insurance of the head office and salaries
and compensation for head office senior corporate management.
Most of the survey studies relating to the allocation of the central service costs are over 20 years old.
The most notable USA studies are those by Reece and Cool (1978) and Fremgen and Liao (1981).
Reece and Cool reported that 40% of 594 surveyed companies calculated income using the same
method for both internal control and external reporting purposes; and approximately half of the
remaining 60% also allocated corporate administrative expenses to calculate income for managerial
purposes.
Fregman and Liao’s (1981) survey consisted of 123 large companies, which aimed to determine the
extent to which firms allocate their central costs to profit centres and to determine why and how these
allocations are made. The results indicated that over 80% of respondents allocated central services
costs for performance evaluation. They reported that the main reason for allocation was to remind
profit centre managers that central costs exist and divisions must make enough profit to cover them.
This reason was cited by 90% of the companies. For those companies that adopted a policy of non-

1
This definition and the illustrations were included within the questionnaire guidance notes.

3
allocation the main reason stated for adopting such a policy was that profit centre managers have no
control over such costs and they object to charges that they cannot influence and control.
In addition to the surveys Merchant (1989) conducted a field study of 12 corporation within the USA
which were diverse in size, type of business, capital and labour intensity, technology, growth,
customer bases, and the degree of diversification. He found that corporate general and administrative
expenses were assigned to profit centre managers in seven of the 12 corporations. With regards to
allocation bases, he found that costs were allocated using different bases such as total assets, number
of employees, or revenues.
Merchant (1989, P. 100) identified two reasons for cost allocations. The first was that allocating
general and administrative expenses can generate constructive conflict, between divisional managers
and corporate staff, which will help keep corporate expenses under control. The second was that
allocating costs facilitates evaluating profit centre managers’ on the basis of a closer approximation
of the full cost of their actions. In five of the 12 corporations, which did not allocate general and
administrative expenses, corporate managers believed that the profit centre managers had only
insignificant influence on these costs or that the allocations were arbitrary. Merchant concluded that
allocated corporate expenses should be related to different aspects. They include:
(1) cost drivers or the factors that cause the costs to be incurred;
(2) behavioural objectives (i.e. divisional managers exercise influence on the particular item of cost);
(3) decision making purposes at the divisional level; and
(4) whether or not divisional managers perceive their autonomy is greater if the costs are allocated to
them.
Ugras (1994) conducted a study of 159 companies in the USA that aimed to investigate the effects of
the organisational factors on the decision to allocate non-controllable costs. The study used a
classification of costs into controllable and non-controllable based on a combination of several
purposes, rather than just performance evaluation (i.e. decision analysis, cost-based pricing and
financial analysis). He found that the amount of non-controllable costs allocated was influenced by
various organisational factors: the cost and difficulty of observing managers’ actions; the vertical and
sequential hierarchy in the organisation, size of the firm; the diversity of the divisions; and the need
for co-ordination.
In the UK a study by Ramadan (1985:1989) examined the perceptions of top management in
divisionalised companies for central cost allocation for the purpose of performance evaluation. Based
on responses from 113 large UK companies, he found that the most important reason for cost
allocation was to evaluate divisional managers’ performance. Respondents (69 companies) who
allocated central costs for performance evaluation cited that the main reason was to make divisional
managers aware that central costs exist and must be covered by divisional profit. The remaining
respondents (44 companies) that did not allocate central costs reported that the main reasons were
that central costs were beyond divisional managers’ control and also that divisional managers object
to charges which they cannot influence and control. Concerning the relationship between cost
allocations and organisational factors, he found that cost allocation were more likely to exist in
companies with high degrees of interdependence between divisions. However, there were no
significant relationships between cost allocation and the other organisational variables (i.e. degree of
decentralisation, number of divisions and the cost of monitoring divisional managers’ performance).
Skinner (1990) conducted an empirical study in Australian and New Zealand organisations. He used
telephones conversations as a means for collecting the relevant information regarding performance
measures. His sample included 99 companies. He reported that companies allocated central services
for different reasons. The most common reason was to encourage profit centre managers to put
pressure on service department managers to do a better job of controlling their costs. In addition, he
found that companies, which allocated central services, used allocations to encourage questioning the
internal provision of the service or the level at which it is supplied internally. He also found that
companies used cost allocations for institutional purposes, such as making inter-division and inter-

4
firm comparisons. The rationale, for this, was that companies tended to have different types of
services provided centrally or locally by divisions, therefore the omission of central services' costs
would destroy comparability. Skinner also found that companies allocated central services for making
strategic decisions, at both corporate and divisional levels, such as changing (increasing, and
decreasing or even closure) the size of division. This reason implies that full cost allocation is a good
estimate of long-run incremental costs.
The extent to which divisional managers were able to influence in some way the central costs
allocated to them was also examined. He reported that 37 companies (in both countries) claimed that
divisional managers were able to influence all central costs; however, some companies (4 companies)
reported that divisional managers were able to influence only some costs (for example, services cost
but not administrative) and not others. On the other hand, 14 companies reported that divisional
managers could not influence central costs. Skinner (1990) went further in his analysis and examined
the types of influence; and, found two types of influence. The first type was that divisional managers
were able to put pressure on service department to do a better job of controlling central costs. The
second was that divisional managers could raise the question whether or not it would be better if an
existing service were to be supplied (wholly or partially) either from outside the company or within
each division.
Limitations of previous studies
The previous studies did not attempt to classify central costs/costs of common resources into their
controllable and non-controllable elements. A distinguishing feature of this research is that three
categories of the costs of common resources are identified. In addition, the research distinguishes
between the cost of common resources and group head office general and administrative costs. These
represent corporate business-sustaining costs that are unlikely to be influenced by divisional
managers.
Although previous studies have examined why companies allocate or do not allocate central costs
they have not incorporated many of the potential explanatory factors that have been identified in the
literature. Furthermore, many of the studies have not included Likert scale responses so that it has not
been possible to rank the potential explanatory factors for allocations or non-allocations. A further
limitation of previous studies is that that little attention has been given to the relationship between the
extent of the allocated costs and potential explanatory factors. Finally, most of the studies were
conducted over 20 years ago when the business environment and management practices/philosophies
were very different from those existing today.
Potential explanatory factors for holding divisional managers accountable for
uncontrollable factors
A number of potential explanatory factors have been identified in the literature for holding managers
accountable for uncontrollable costs. They include firm size, number of divisions, divisional
diversity, degree of decentralisation, divisional autonomy and the need for co-ordination.
Firm size
Many writers (e.g. Mercahnt, 1987; Ugras, 1994; Zimmerman, 1979; Biaman and Demski, 1980;
Holmstorm, 1982; Biaman and Noel, 1985; Antle and Smith, 1986; Maher et al., 1991) argue that the
difficulty of observing managers’ actions is a reason for holding managers accountable for
uncontrollable factors. Furthermore, Ugras (1994) argues that the difficulty of observing managers’
actions is likely to be high in larger firms. In addition, the management accounting literature suggests
that firms use cost allocations as a monitoring device as they become larger (Zimmerman, 1979;
Demski, 1981). Therefore, it can be argued that the greater the size of the organisation, the greater the
likelihood of the allocation of non-controllable costs.
Number of divisions
Following the argument that firm’s size is related to the difficulty of observing managers’ actions, it
can be argued that the number of divisions represents a proxy measure of firm size. Firms with

5
numerous divisions are more likely to have difficulty in keeping track of each division’s activities
and, in turn, cannot take into account the effect of uncontrollables. Williamson (1975) argues that
when a firm’s size increases, there will be great control losses. Ramadan (1985; 1989) hypothesised
that as the number of divisions within firms increase they tend to compensate for control losses over
the performance of divisional managers by allocating central costs. However, the empirical evidence
did not support his hypothesis. He found that there was no significant difference in the number of
divisions between the allocators and non-allocators of central costs for the purpose of performance
evaluation.
Divisional diversity
According to Ugras (1994), divisional diversity makes it more difficult to observe divisional
managers’ actions and, in turn, affects the extent to which uncontrollable factors are taken into
account when evaluating divisional managers’ performance. He argues that top management at firms
with diverse divisional activities may not be able to monitor each division’s actions. In addition, it
has been shown that companies tend to allocate non-controllable costs to get divisional managers take
them into account when they make their decisions. Similarly, it can be argued that the greater the
divisional diversity, the greater the likelihood of cost allocation and the less likelihood of taking into
account the effects of uncontrollable factors.
Degree of decentralisation
Drawing on Zimmerman’s (1979) argument that companies tend to allocate costs when it is difficult
to observe and measure divisional performance, Ugras (1994) argues that cost allocation is likely to
occur in organisations with numerous levels in their hierarchy. This implies that companies that hold
divisional managers accountable for uncontrollable factors are more likely to have higher degrees of
decentralisation.
Divisional autonomy
It can be argued that there are two views associated with the relationship between divisional
autonomy and uncontrollable factors. The first view is that which holds divisional managers
accountable for costs over which they have little or no control. Such a procedure encourages
divisional managers to cope with uncontrollable factors. Some researchers advocate this view on the
grounds that divisional managers should act like owners and consequently they should bear the risk in
terms of the uncontrollable factors (Bromwich and Walker, 1998). Merchant (1989) presents the
second view. He argues that some divisionalised companies cite that holding managers responsible
for uncontrollable factors contributes to their feeling of increased autonomy. This implies that the
greater the divisional autonomy, the greater the likelihood of divisional managers being held
accountable for uncontrollable factors.
The need for co-ordination
Another potential factor is the need for co-ordination. Chandler (1962) and Goold and Campell
(1987) argue that one of the centre’s (corporate management) roles is to co-ordinate divisions’
activities and wide-company strategies. The accounting literature also suggests that cost allocation
can be used as a means to achieve such co-ordination. According to Demski (1981), one of the
reasons for cost allocation is co-ordination by motivating various responsibility centres to act in the
firm’s best interest.
Rajan (1992) extended his analysis of cost allocation as a means to achieve co-ordination among
divisions. He developed an allocation scheme showing that common costs serve as a motivational
tool, which induces divisional managers to co-ordinate, their actions and act according to central
management (corporate management) desires. In addition, Ugras (1994) argues that cost allocation
for the purpose of performance evaluation is related to the organisation’s need for achieving co-
ordination among divisions. Such argument can be extended to cover uncontrollable factors. Based
on the above discussion it is hypothesised that the greater divisional autonomy, the greater the
likelihood of holding divisional managers accountable for uncontrollables.

6
Data collection
Empirical data was collected by means of postal questionnaire survey. Prior to undertaken the survey
the questionnaire was subject to extensive pre-testing stages that include interviews with both
practitioners and academic2.
In order to generalise the results of this research to the whole population, a random sample was
selected. The aim was to select a population consisting only of UK divisionalised companies.
Unfortunately, there is no database, which contains only divisionalised companies. In addition, there
were no reliable methods, which could be used to distinguish between divisionalised and non-
divisonalised companies prior to mailing the questionnaire. Thus the sample selection included both
divisionalised and non-divisionalised companies. To identify divisionalised companies the first
question in the questionnaire asked the respondents to specify whether their company was
divisionalised. Respondents employed in non-divisionalised companies were asked to complete only
the first four questions of Section A and return the questionnaire. The aim was to include only
divisionalised companies in the study.
It was decided that the selected sample should include subsidiaries (or divisions) of overseas
companies, listed and unlisted companies. Since it was considered possible that performance
measurement practices could be influenced by the practices of head office where this was located in
an overseas country, subsidiaries of overseas corporations were included in the sample. This
provided the opportunity to examine if there were any differences between the practices of
organisational units whose head office was located in the UK compared with those units with head
offices located outside the UK. Both listed and unlisted companies were included in the sample
because they may use different performance measures, or attach different levels of importance to the
measures used. Including both types of organisations provided the opportunity to examine if the
observed practices differed.
Only manufacturing organisations were included in the sample because of the difficulty of designing
a single questionnaire that was applicable to both manufacturing and service organisations.
Consideration was given to producing two questionnaires, one for manufacturing and the other for
service organisations. However, it is extremely difficult to make generalisations about service
organisations because of their distinctive features. Hospitals are very different from banks, and banks
are quite different from universities. Therefore it was considered that it would not be possible to
design a single questionnaire that would be applicable to all types of service organisations.
Having determined the criteria to be used to select the sample it was necessary to identify an
appropriate database. It was decided to use the ‘One Source’ database which includes detailed
information relating to over 360,000 public and private companies. The information is contained on
two disks, volume one and two. Volume one includes companies with a turnover of £500,000 or more
and volume two includes companies with a turnover of less than £500,000. Since it was felt that size
was a good indicator for divisionalised companies, it was decided to use volume one. The second
reason for using this database was that it provided information relating to sales turnover, number of
employees, directors’ names and accountants’ names. This information was of vital importance for
identifying potential respondents.
Random sampling was used to select the sampling units in this study as there was no reason to
believe that non-divisionalised companies were not distributed randomly across the population. Given
the low response rate that was achieved in the pilot study, a random sample of 750 companies with an
annual sales turnover in excess of £100 million was selected from the population. Companies in
excess of £100 million were selected because there is a higher probability that companies with a sales
turnover of below £100 million will be non-divisionalised. In addition, the research focused on

2
I wish to thank Professors Otley (University of Lancaster) and Hopper (University of Manchester)
and a number of PhD students from Huddersfield University Business School for their helpful
comments on the early versions of the questionnaire.

7
organisations that have established formal performance reporting control systems. Smaller
organisations might rely more on direct observations, action and social controls rather than formal
performance measurement systems.
It was important that the respondents to the questionnaire should have a good knowledge of their
organisation's performance measurement systems and the potential explanatory variables. It was
decided to address the questionnaire directly to the group finance director, or to the individual at the
position of highest level of financial responsibility in each group head office, and to head of the
accounting function if the company was a subsidiary. The name of the group finance director or the
head of the accounting function was identified for nearly all of the identified respondents as it was
considered that this was likely to increase the response rate. The letter accompanying the
questionnaire also included a request for the recipients to pass the documentation to their appropriate
colleague if they thought that they had been incorrectly identified because they did not have sufficient
knowledge relating to the performance measurement system in their organisation. The job titles of all
of the persons completing the questionnaire were examined carefully prior to including the responses
in the analysis. The analysis of responses by job titles indicated that 55% were completed by finance
directors and the remainder by financial controllers or management accountants.
A total of 251 questionnaires were returned from the sample of 750 companies. This included 18
returned not completed with a letter stating that the company no longer employed the respondent. A
further 94 were returned with a covering letter explaining why they had not been completed the
questionnaire. Most of the responses indicated that it was not company policy to participate in
surveys or lack of time due to work pressures. The remaining 139 were completed. Of the 139
completed questionnaires 11 companies were not organised into divisions according to the definition
of a division given in the guidance notes in the first page of the questionnaire. Therefore the number
of returned completed questionnaires from divisionlised companies was 128 (including 4 unusable
questionnaires). There are various ways of measuring the response rate. According to de Vaus (1990,
P.99), a common way of computing the response rate is to use the following formula:
Response rate = Number of completed and returned / N in sample – (Ineligible + Unreachable)
Therefore the response rate = 128/ 750 – (18+11) = 18%
Given that it was not possible to distinguish in the initial sample between divisionalised and non-
divisionalised companies it is likely that some of the respondents were non-divisionalised and
therefore ineligible for inclusion in the sample of potential respondents. Therefore the above
response rate is likely to be understated.
Questionnaire content
In the first page of the questionnaire, guidance notes were provided to facilitate answering some of
the questions and to state definitions of those terms which might differ amongst the organisations.
The questions sought to obtain the following information:
1. The extent to which companies use different measures to evaluate the economic performance of
the divisions and the performance of divisional managers;
2. The level of application of the controllability principle;
3. The explanatory factors of the different levels of application of the controllability principle; and
4. The relative importance attached to non-financial performance measures.
Wherever possible a seven-point scale was used to measure the variables. The questionnaire was
divided into seven sections. The first section, section A, contained questions relating to respondents’
companies or organisational unit. Section B focused on the performance measures that are used to
evaluate the economic performance of the divisions and the performance of divisional managers.
Questions relating to the treatment of common corporate costs were incorporated in section C. This
section aimed primarily to examine whether or not the companies allocated different categories
common resource costs and general administrative business-sustaining costs for the purpose of
measuring divisional managerial performance. In addition, the extent of divisional autonomy for

8
different categories of common resources provided by corporate headquarters and the role of cost
allocations were examined. Section D collected data on the potential explanatory factors for holding
divisional managers responsible for uncontrollable factors. The fifth section, Section E, aimed to
determine whether or not economic value added was used to evaluate divisional managerial
performance as the theory advocated. Economic value added is a fairly recent innovation and it is of
interest to ascertain the extent to which companies are adopting this innovation. Section F focused
on the role of non-financial performance measures and the use of the balanced scorecard in
evaluating divisional managerial performance. In addition, this section included question relating to
the extent to which respondents were satisfied with their divisional performance measurement. The
final section obtained relevant demographic information relevant to the respondents' background in
order to test for non-response bias.
Multi-item questions were extensively used to gather information on the potential explanatory
variables so as to capture the different characteristics (or dimensions) of the individual concepts
representing the variables. When an individual indicates his or her own attitude (or opinion) relating
to an object on some scales, a substantial element of intuitive judgement is involved, no matter how
precise the rating instructions and no matter how will trained the individual. Such judgement in the
use of rating scales makes the ratings vulnerable to bias (Judd et al, 1991). To overcome such
potential problems the respondents were asked to respond to multiple-item questions and scores
assigned to the response of each item were then combined and divided by the number of the items in
order to arrive to an overall mean score for the variable. Foster and Swenson (1997) also recommend
the use of the mean of the scores in measuring a construct. They claim that a composite score has the
advantage over an individual single question when either: (1) the variable being measured contains
multi-dimensional aspects requiring several different questions to capture the multi-dimensional
aspects, or (2) there is measurement error in an individual question that is minimised by aggregating
individual question into a composite. In order to test whether the variables measure the same
construct Cronbach’s alpha was computed. For all of the variables measured using multi-item
questions Cronbach’s alpha exceeded 0.6.
Findings relating to the use of financial and non-financial measures
Pursuing the line of argument that companies may use different performance measures to evaluate
divisional managerial performance and the economic performance, the respondents were asked to
indicate which of the following methods best described the use of performance measures in their
companies:
(1) the same performance measures are used for evaluating the performance of divisional managers
and the economic performance of the divisions but different items are included within the
performance measure (44%);
(2) different performance measures are used for evaluating the performance of divisional managers
and the economic performance of the divisions (18%);
(3) Identical performance measures are used for evaluating the performance of divisional mangers
and the economic performance of the divisions (38%).
The figures in the parentheses signify the percentage responses to the question (N=124). Thus, 62%
of the companies distinguish between the economic performance of the divisions and the managerial
performance of the divisional mangers. Thus, variations do exist in the application of the
controllability principle. It should also be noted that the above distinction as a method of examining
the controllability principle has not been investigated by previous studies.
Given that more than one financial measure may be used to evaluate the performance of divisional
managers the respondents were given a list of measures that have been identified in the literature for
measuring overall divisional performance. They were asked to rank in order of importance the three
most important measures. The results for the most important ranking are summarised in Table 1. It
can be seen that target profit before charging interest on capital was considered to be the most
important measure by 55% of the organisations. The second highest percentage (14%) was attributed

9
to target profit after charging interest on capital (residual income). The widely cited target return on
capital employed measure was ranked as the most important measure by only 7% of the respondents.
The question also provided the opportunity for respondents to insert measures other than those listed.
Of the 124 respondents to this question, 8 respondents inserted an unlisted measure as the most
important financial measure. Measures specified included gross contribution margin, return on sales,
sales volume and internal growth, revenue growth and target sales.
Economic value added (EVA) is a fairly recent innovation and the questionnaire sought to ascertain
the extent of use and its relative importance. EVA was ranked as the most important financial
measure by 11 (9%) of the respondents. With regard to EVA usage 28 respondents (23%) indicated
that it was used as a method of evaluating the performance of divisional managers and one
respondent indicated that it was used at a higher level. The responses also indicated that 12% of the
respondents plan to introduce EVA within the next two years and 65% indicated that they had no
plans to use EVA. Of the 28 respondents currently using EVA, 16 indicated that they made
adjustments to accounting numbers for distortions introduced by generally accepted accounting
principles and 12 did not make any adjustments.
In recent years increasing attention has been given in the literature to incorporating into the
performance measurement system those non-financial performance measures that provided feedback
on the key variables that are required to compete successfully in a global economic environment
(Kaplan and Norton, 1996). In particular, a balanced-scorecard approach has been advocated. The
responses indicated that most organisations do not rely only on financial measures to evaluate
divisional performance. Non-financial measures were used to evaluate divisional performance by 97
(78%) of the respondents. These respondents were asked to specify the approaches they adopted to
incorporate non-financial measures. Of the 97 respondents specifying that they incorporated non-
financial measures3:
• 55% adopted the balanced-scorecard approach;
• 7% used the Tableau de Bord;
• 18% used EFQM 4;
• 14% used other approaches;
• 25% stated that they used none of the listed items and did not specify the approach they adopted
for incorporating non-financial measures.
Thus 43% (N=53) of all of the 124 respondents used the balanced scorecard to evaluate divisional
performance. Eight respondents specified other approaches. They included customer performance
and satisfaction measures (3), key performance indicators (3), market share (1), safety targets (1),
quality and productivity (1), and employee satisfaction (1).
Finally, the respondents were asked to indicate on a seven-point scale the importance attached to
financial versus non-financial measures for evaluating divisional performance. A scale of 1 (financial
measures are considerably more important than non-financial measures) to 7 (non-financial measures
are considerably more important than financial measures) was used. The mid-point of 4 was anchored
'They are about the same importance.' Of the 97 respondents using both financial and non-financial
measures to evaluate divisional performance, 71% entered a score of less than 4 (thus indicating that
financial measures were more important), 18% entered a score of 4 (equal importance) and 11% a
score of above 4 (non-financial measures were more important). Given that all of the respondents
were located within the finance function it could be argued that they have a biased view relating to
the relative importance of the financial measures compared with non-financial measures.
Nevertheless, the findings tend to reject the views that Kaplan and Norton have observed and
objected to. That is, that financial measures should be de-emphasised on the grounds that by making
fundamental improvements to the key non-financial measures the financial measures will take care of

3
The responses add up to more than 100% because some respondents specified that they used more than one approach
4
The term ‘EFQM’ refers to the European Foundation for Quality Management’s Excellence Model.

10
themselves. In other words, financial success should be the logical consequence of favourable key
non-financial measures.
The costs of common resources and the application of the controllability
principle
The respondents were asked to indicate whether some of the costs of common resources were
allocated to divisional managers prior to computing the performance measures. Of the 118
organisation responding to this questions, 98 (83%) stated that they allocated some of the costs of
common resources to divisional managers prior to computing performance measures. Only 20 (17%)
stated that they did not allocate such costs to divisional managers for the purpose of performance
evaluation.
In order to assess the extent of controllability of the cost of common resources allocated to divisions
the respondents were asked to specify whether or not common corporate resources fell within the
following categories:
1. Divisional managers can determine the quantity acquired and the price paid because they have the
authority to purchase the services either inside or outside the organisation. In other words, they
have full autonomy over determining from where to acquire the service and the price paid for the
services (11%);
2. Divisional managers cannot determine the prices paid for the services because they do not have
the authority to purchase the service from outside the organisation. However, they can determine
the quantity of common resources that are consumed and thus influence the amount of coats that
are allocated to them (55%); and
3. Divisional managers cannot determine either the price or quantity they are charged for the
resources. They are allocated with a fixed sum irrespective of usage (56%).
The numbers in the parentheses indicate the percentage of respondents (N=98) indicating that
corporate common resources fell within the specified category. For analysis purposes the three
categories described above were classified as follows:
Category 1 - Full autonomy and controllable;
Category 2 - Partial autonomy and partly controllable; and
Category 3 - No autonomy and non-controllable.
Applying the controllability principle controllable costs should either be allocated using cause-and-
effect allocations or direct charges where the service consumed can be directly measured. The
controllable element of the partly controllable costs should be allocated and the non-controllable
element not allocated. Non-controllable costs should not be allocated to divisions for measuring
divisional managerial performance.
In order to determine the extent to which organisations allocate costs with different potential levels of
controllability to divisions, the respondents were asked to indicate the extent to which the cost of
common resources for each of the three different autonomy/controllability categories were allocated
to divisions for the purpose of measuring divisional managerial performance. Table 3 summaries the
responses. Given that controllable costs can be subject to either direct charges (no allocations) or
cause-and-effect allocations, and that the total respondents with common resources falling in this
category was 11 the responses for the controllability category (i.e. category 1) are not presented.
Earlier in this paper it was argued that business-sustaining corporate general administrative costs
should be distinguished from the costs of common resources because they are incurred for the group
as a whole rather than being viewed as joint resources that fluctuate according to the demand for
them. Therefore, the respondents were asked in a separate question to indicate the extent to which
these costs were allocated to divisions. The responses to this question are also presented in Table 3.
It is apparent from Table 3 that a significant majority of organisations (70% for non-controllable
category 3 costs and 64 % for business-sustaining corporate general administrative costs) allocate

11
most or all non-controllable costs for measuring divisional managerial performance. Table 3
indicates that 26% of the respondents either did not allocate, or only allocated a minor portion of the
costs, in respect of business-sustaining corporate general administrative costs. The corresponding
figures for limited controllability (category 2) and non-controllable (category 3) common resource
costs were 4% and 11%.
It is possible that the three-fold classification of corporate resources costs may not fully capture the
extent of autonomy that divisional mangers have in using corporate resources. Therefore a check
question was included in the questionnaire in order to provide additional indications for the degree of
control over the use of common resources. The respondents were asked to indicate the extent to
which divisional managers were prohibited from using an outside service when such a service was
provided internally. The responses are presented in Table 4. The table shows that 57% of the
responding companies had limited autonomy to use the outside service, 30% had no autonomy and
13% had substantial autonomy.
In addition, the respondents were asked to indicate the extent to which divisional managers can
negotiate the amount allocated (or the price paid) for the costs of common resources. On a scale of 1
(not at all) to 7 (to a considerable extent) 48% of the responding organisations entered a score of 1 or
2 and only 8% entered a score of 6 or 7. This mean score and standard deviation were 3.15 and 1.77
respectively. The responses to this question provides further support to indicate that the majority of
divisions had limited or no autonomy in determining the costs of common resources that are allocated
to them.
The research clearly indicates that organisations generally allocate all or most non-controllable costs
to divisions for the evaluation of divisional management. However, this practice can be viewed as a
mechanism for increasing the target profit/profitability levels for the divisions to ensure that
divisional target profit levels are sufficient to cover a share of corporate costs. Adopting this
interpretation the allocation of corporate costs that are non-controllable by divisional managers can
be viewed as a tax levy that aims to recover a share of corporate costs (Bromwich and Bhimani,
1994). The counter-argument to this point of view is that is that if corporate management wishes to
inform managers that divisions must be profitable enough to cover not only their own operations but
corporate expenses as well, it is preferable to set a high budgeted controllable profit target that takes
into account these factors. Divisional managers can then concentrate on increasing controllable profit
by focusing on those costs and revenues that are under their control and not be concerned with the
costs that they cannot control.
Importance of factors influencing organisations to allocate the costs of shared
resources
The findings indicate that most organisations allocate the cost of shared resources and it has
suggested that this could be viewed as a mechanism for increasing the target profit/profitability levels
for the divisions to ensure that divisional target profit levels are sufficient to cover a share of
corporate costs. The extensive use of cost allocations suggests that setting a higher budgeted
controllable profit target that is sufficient to cover a share of corporate costs, but which does not
involve allocations of central costs, is not adopted by most organisations. It is therefore of interest to
examine the potential factors that are likely to influence organisations to allocate the costs of shared
resources.
The respondents were asked on a-7 point scale to indicate the importance attached to list of 13 factors
relating to the allocation of the costs of common resources. To ascertain the relative importance
attached to each factor the means, standard deviations and the total scores are presented in Table 5. A
threefold classification was also used to classify the factors by behavioural, institutional and
measurement factors (the latter relating to the difficulty of separating controllable and uncontrollable
elements). It can be seen from Table 6 that the behavioural reasons had the highest five positions
according to their importance rating; followed by institutional reason and the measurement reasons
were ranked eleventh and twelfth.

12
The top four rankings were:
1. To show divisional managers the total costs of operating their divisions (a)
2. To make divisional managers aware that such costs exist and must be covered by divisional
profits (g)
3. Divisional managers would incur such costs if they were independent units (m)
4. Divisional managers should bear the full business risk as they were managers of non-
divisionalised companies (c)

These rankings suggest that the most important reasons for allocating the cost of common resources
related to allocations being used as a mechanism for making divisional managers aware of the total
costs of operating their divisions and the costs that must be covered by divisional profits. Two of the
reasons in Table 6 relate to measurement problems. They are:
• Distinguishing between controllable and uncontrollable requires subjective judgements which
can create conflicts (b)
• It is extremely difficult to separate controllable and uncontrollable elements (d)
As indicated above these items were respectively ranked eleventh and twelfth. The responses
therefore suggest that difficulty in isolating the non-controllable items is a relatively unimportant
factor in the decision to allocate the cost of shared resources to divisions. The other important
observation from Table 6 is that ‘allocations being undertaken because they had become embedded as
part of company tradition’ was ranked as the least important item.
It should be noted that other studies have reported similar findings. The Fremgen and Liao (1981,
P.1) USA study reported that about 90% of the responding companies allocated indirect corporate
costs to remind profit centre managers that indirect costs exist and that profit centre earnings must be
adequate to cover some share of those costs. In the UK, Ramandan (1985, P.173) found that the most
important reason (in 38 companies out of 69) was to make divisional managers aware that central
overhead costs exist and must be covered by divisional profits. Therefore, the most important reasons
in this study are similar to the ones reported by the previous studies even though this study
incorporated many reasons, which were not pursued by the previous studies. Also, the ranking order
was used in order to determine the relative importance attached to each reason.
Importance of factors influencing organisations not to allocate the costs of shared resources
The factors influencing organisations not to allocate the costs of shared resources were also explored.
The respondents were asked to indicate where their organisation did not allocate some of the costs of
common corporate resources, how important was a list of factors in arriving at a decision not to
allocate the costs. Table 6 shows the rank order of the factors influencing non-allocation. The
highest ranked factor is that common corporate costs are not controllable by divisional managers and
the second highest is that divisional managers object to charges they cannot influence and control. It
is apparent from the responses that factors relating to the application of the controllability principle
are the main reason for not allocating common costs
Summary and conclusion
The survey findings indicate that the majority of companies did not use identical measures for
evaluating the performance of divisional managers and the economic performance of the divisions.
Target profit before charging interest on capital was considered to be the most important measure
used to evaluate the performance of divisional managers by 55% of the organisations, target profit
after charging interest on capital (residual income) was considered the most important measure by
14% of the organisations. The widely cited target return on capital employed measure was ranked as
the most important measure by only 7% of the respondents. EVA was used by 23% (28) of the
respondents as a method of evaluating the performance of divisional managers. A further 11% of the
respondents planned to introduce EVA within the next two years. The balanced scorecard was used
to evaluate divisional performance by 43% of the respondents.

13
The survey findings indicated that a significant majority of organisations (70% for non-controllable
common resource costs and 64 % for business-sustaining corporate general administrative costs)
allocate most or all non-controllable costs for measuring divisional managerial performance.
However, the findings indicated that the controllability principle was applied. It was concluded that
the observed widespread allocation of central costs could be viewed as a mechanism for increasing
the target profit/profitability levels to ensure that divisional target profit levels are sufficient to cover
a share of corporate costs. The findings imply that the alternative mechanism for achieving this
objective involving setting a higher budgeted controllable profit target that is sufficient to cover a
share of corporate costs, but which does not involve allocations of central costs, was not adopted by
most organisations.
The responses relating to the factors that influence organisations to allocate the costs of shared
resources tended to support the above interpretation of the role of cost allocations. The dominant
factors related to allocations being used as a mechanism for making divisional managers aware of the
total costs of operating their divisions and that the corporate costs must be covered by divisional
profits. The use of allocations to stimulate divisional managers to put pressure on resource centre
managers to control their costs and to take a greater interest in the cost of shared resources were also
considered to be extremely important factors by the respondents. The findings indicate that the
behavioural reasons are much more important than institutional and measurement reasons.
This paper has concentrated on the descriptive findings of the research. Hypotheses have also been
developed that examines how potential explanatory factors influence the extent to which
organisations allocate uncontrollable costs. Data has been collected on three categories of
uncontrollable costs (partially controllable common resource costs, uncontrollable common resource
costs and business-sustaining corporate general administration costs). In addition, data has been
collected on (1) the extent to which companies exclude the effects of uncontrollable environmental
factors (e.g. changing economic conditions, competitors’ actions); and (2) the extent to which
divisional interdependencies are taken into account when evaluating divisional managerial
performance. Thus 5 alternative measures of the dependent variable have been collected. Potential
explanatory variables (the independent variables) for which data has been collected include measures
relating to firm size, number of divisions, divisional diversity, degree of decentralisation, co-
ordination need, divisional autonomy, divisional interdependence, the role of informality and
perceived environmental uncertainty. Correlation coefficients between the dependent variables and
the independent variable will be computed and the possibility of aggregating the data to enable
appropriate non-parametric hypotheses tests to be undertaken is being examined. The
appropriateness of applying regression analysis to data that has been measured on an ordinal scale is
also currently under consideration. The presentation of the findings relating to these issues will be
the subject of future papers.
There is a need for future research to examine how allocations of the costs of common resources are
determined. In particular the following issues need to be explored. Do allocations involve fixed
charges or rely on allocation bases? What mechanisms are used? For example, is transfer pricing
used to allocate some of the costs of common resources? Are the bases that are used to determine the
amount of the allocations made explicit to divisional managers? Can managers indirectly influence
the allocations by actions relating to the allocation bases attributable to the divisions?

14
Table 1: Ranking of three most important financial measures

Financial measure Most important Second most Third most

ranking important ranking important
ranking
No. Percent No. Percent No. Percent
(a) Achievement of a target rate of
return on capital employed 9 7.3 21 18.1
41 41
(b) A target profit after charging
interest on capital employed 18 14.5 11 9.5
5 5
(residual income)
(c) A target profit before charging
interest on capital employed 68 54.7 23 19.8
5 5
(d) A target economic value added
figure 11 8.9 8 6.9
10 10
(e) A target cash flow figure 10 8.1 45 38.8 27 27
(f) Other 8 6.5 8 6.9 12 12
Total 124 100.0 116 100.0 100 100

Table 2: The costs of common resources as a percentage of divisional turnover

% common Frequency Percent Cumulative percent
resources
0-5% 52 44.4 44.4
6-10% 34 29.1 73.5
11-15% 22 18.8 92.3
16-20% 5 4.3 96.6
21-25% 1 0.9 97.4
26-30% 2 1.7 99.1
Over 30% 1 0 .9 100.0
Total 117 100.0

15
Table 3: The extent to which costs of common resources and business-sustaining
corporate general administrative costs are allocated to divisions

The extent of costs Partially Non-controllable Business-

allocated to divisions controllable costs costs sustaining
(category 2) (category 3) corporate general
administrative
costs
No. Percent No Percent No. Percent
.
(a) None of the costs are _ _ 3 5.6 18 15.5
allocated
(b) Only a minor portion of 2 3.8 3 5.6 12 10.3
the costs are allocated
(c ) A small but significant 10 18.9 8 14.8 12 10.3
portion of the costs are
allocated
(d) Most, but not all of the 19 35.8 11 20.4 25 21.6
costs, are allocated.
(e) All of the costs are 22 41.5 29 53.6 49 42.2
allocated
Total 53 100 54 100 116 100

Table 4: Extent to which divisional managers are prevented from using an outside
service when the service is provided internally
Frequency Percent
(a) divisional managers have substantial 1 1.0
freedom to purchase the service externally,
and in practice they exercise this option.
(b) divisional managers have substantial 1 1.0
freedom to purchase the service externally,
but in practice they do not exercise this
option.
(c) divisional managers have substantial 11 11.2
freedom to use an outside service, and they
exercise this option on some occasions.
(d) divisional managers have very limited 56 57.1
freedom to use an outside service, and in
most cases they must use the central
services.
(e) divisional managers must use the central 29 29.6
services in all the cases
Total 98 100.0

16
Table 5: Importance of factors influencing organisations to allocate the costs of shared resources
Rank Reasons Nature of % rating 1 or % rating 6 or N Sum (total Mean Standard
the 2 7 score) deviation
reasons
1** To show divisional managers the total costs of B*** 3.1 70.1 97 563 5.80 1.15
operating their divisions (a)*
2 To make divisional managers aware that such costs B 5.1 62.9 97 547 5.64 1.26
exist and must be covered by divisional profits (g)
3 Divisional managers would incur such costs if they B 6.2 47.4 97 503 5.19 1.47
were independent units (m)
4 Divisional managers should bear the full business B 6.2 43.3 97 495 5.10 1.29
risk as they were managers of non-divisionalised
companies (c)
5 To stimulate divisional managers to put pressure on B 12.3 45.3 97 471 4.86 1.79
resources centre managers to control their costs (j)
6 To induce divisional managers to take greater B 14.4 43.3 97 468 4.82 1.71
interest in the costs of shared resources (k)
7 To enable inter-division or inter-firm comparisons I 22.6 29.9 97 422 4.35 1.85
to be made (e)
8 To provide signals on the efficiency of service B 23.7 26.8 97 404 4.16 1.77
department that provides shared resources (f)
9 To stimulate divisional managers to economise in B 26.8 24.7 97 394 4.06 1.76
their usage of shared resources (h)
10 Divisional managers control the usage of the B 29.9 18.6 97 356 3.67 1.75
resources (l)
11 Distinguishing between controllable and D 24.7 8.2 97 350 3.61 1.52
uncontrollable requires subjective judgements
which can create conflicts (b)
12 It’s extremely difficult to separate controllable and D 33.0 13.4 97 346 3.57 1.73
uncontrollable elements (d)

13 Because cost allocations are part of a company’s I 50.5 11.3 97 284 2.93 1.87
tradition (i)
* letters represent the ranking in the questionnaire.
** The rank was made in descending order*** B= behavioural dimensions, I= institutional dimensions and D= difficulty of separating controllable from non-controllable
elements

17
Table 6: Importance of factors influencing the decision not to allocate the cost of common
resources

Rank Reasons % rating % rating N Sum Mean SD

1 or 2 6or 7
1 Common corporate costs are not 13.2% 54.7% 53 274 5.17 1.96
controllable by divisional
managers (b)*
2 Divisional managers object to 24.5% 30.2% 53 227 4.28 1.96
charges they cannot influence
and control (d)
3 The costs of making the 34% 26.4% 53 208 3.92 2.16
allocations would exceed their
benefits (c)
4 Allocations are arbitrary and 37.7% 34% 53 206 3.98 2.22
tend to distort divisional profit
(e)
5 The amount of common 49.1% 17% 53 163 3.08 2.04
corporate costs is too small to
warrant allocation (a)
6 Unnecessary internal tension can 47.2% 13.2% 53 161 3.04 1.93
be avoided by no allocation (f)

* letters represent the ranking in the questionnaire.

18
 References
Antle, R. and Demski, J.S., 1988. The controllability principle in responsibility accounting, The
Accounting Review, LXIII, October, 700-718.
Antle, R. and Smith, A., 1986. An empirical investigation of the relative performance evaluation of
corporate executives, Journal of Accounting Research, Vol.8, No.2, 201-221.
Baiman, S. and Demski, J.S., 1980. Economically optimal performance evaluation and control
systems, Journal of Accounting Research, Supplement, 184-220.
Bromwich, M. and Bhimani, A., 1994. Management accounting: Pathways to progress, Chartered
Institute of Management Accountants, London.
Bromwich, M. and Walker, M., 1998. Residual income past and future, Management Accounting
Research, Vol.9, 391-419.
Chandler, Alfred, 1962. Strategy and Structure, Cambridge: MIT Press.
Choudhurry, N., 1986. Responsibility accounting and controllability, Accounting and Business
Research, Vol. 16, 189-198.
De Vaus, D.A., 1990. Surveys in Social Research, London: UWIN HYMAN.
Demski, J.L., 1981. Cost allocation games. Edited in Moriraity, S., Joint cost allocations, University
of Oklahoma Press, 142-173.
Drury, Colin, 1996. Management & cost accounting, Fourth Edition, London: Thomson.
Drury, C., Braund,S., Obsorne, P. and Tayles, M., 1993. A survey of management accounting
practices in UK manufacturing companies, Chartered Association of Certified Accountants.
Emmanuel, C. and Otley, D., 1976. The usefulness of residual income, Journal of Business Finance
and Accounting, Summer, 43-51.
Foster, G. and Swenson D. W., 1997. Measuring the success of activity-based cost management and
its determinates, Journal of Management accounting Research, Vol.9, 104-141.
Fremgen, J. M. and Liao, S.S., 1981. The allocation of corporate indirect cots, New York: National
Association of Accountants.
Goold, M. and Campbell, A., 1987. Strategies and styles-The role of the centre in managing
diversified corporations, Balckwell Publisher.
Horngren, C.T., Bhimani,A., Foster, G. and Datar, S.M., 1999. Management and cost accounting,
Prentice Hall.
Holmstrom, B., 1982. Moral hazard in teams, Bell Journal of Economics, Autumn, 324-340.
Judd, C.M., Smith, E.R. and Kidder, L.H., 1991. Research methods in social relations, Fort Worth:
Harcourt Brace Jovanovich.
Kaplan, R. S and Atkinson, A.A., 1989. Advanced management accounting, Prentice Hall.
Kaplan, R.S. and Norton, D.P., 1996. Translating strategy into action: The Balanced Scorecard,
Harvard Business School Press, Boston, Massachusetts.
Maher, M. W. and Stickney, C.P. , Weil, R.L. and Davidson, S., 1991. Managerial accounting, 4th
edition, New York: Harcourt Brace Jovanovich.
McNally, G.M., 1980. Responsibility accounting and organisational control: some perspectives and
prospects, Journal of Business Finance and accounting, Vol. 7, 165-181.
Merchant, K.A., 1987. How and Why firms disregard the controllability principle, In: Field study in
management accounting , Bruns, W.J. and Kaplan, R.S., Harvard Business School Press, 316-338.

19
 Merchant, K.A., 1989. Rewarding results: profit center managers, Harvard Business School Press.
Merchant, K.A., 1998. Modern management control systems-Text and cases, Prentice Hall.
Rajan, M.V., 1992. Cost allocation in multiaget setting, Accounting Review, Vol.87, No. 3, 527-545.
Ramadan, Sayel, 1985. The allocation of central overhead costs for the purpose of performance
evaluation, Unpublished PhD thesis, University Of Glasgow.
Ramadan, Sayel, 1989. The rationale for cost allocation: a study of UK divisionalised companies,
Accounting and Business Research, Vol.20, No.77, 31-37.
Reece, J.S. and Cool, W.R., 1978. Measuring investment center performance, Harvard Business
Review, May/June, 28-46; 174-176.
Scapens, W. Robert and Sale, Timothy J., 1982. Financial control of divisional capital investment,
The Institute of Cost and Management Accountants.
Skinner, R. C., 1990. The role of profitability in divisional decision making and performance
evaluation, Accounting and Business Research, Vol. 20, No.78, 135-141.
Suh, Y., 1987. Collusion and noncontrollable cost allocation, Journal of Accounting Research,
Vol.25, Supplement, 22-46
Suh, Y., 1988. Noncontrollable costs and optimal performance measurement, Journal of Accounting
Research, Vol. 26, No.1, 154-168.
Ugras, J.Y., 1994. Factors affecting allocation of noncontrollable costs for performance evaluation
use: a survey, In: Advances in management accounting Vol.3, Epstein, M.J. (ed.), Greenwich: JAI
Press, 255-278.
Williamson, Oliver E, 1975. Markets and Hierarchies: analysis and antitrust implications, New
York: Free Press.
Zimmerman, J.L., 1979. The costs and benefits of cost allocation, Accounting Review, July, 504-521.

20

View publication stats