Feature article

Legitimacy theory or managerial

reality construction? Corporate
social disclosure in Marks and
Spencer Plc corporate reports,
1969–1997
David J. Campbell
Abstract
This paper sets out to make some comment on the debate surrounding
explanatory theories for the phenomenon of voluntary social disclosure.
It notes that two explanations appear from the literature to be most
prominent: legitimacy theory and political-economy of accounting expla-
nations, both of which are consistent with a stakeholder understanding
of the organization-society relationship. The published annual corporate
reports of the British retailer Marks and Spencer Plc are analyzed over the
period 1969–1997 inclusive with a view to providing insight into the
causes of variability in the volume of social disclosure. The scope of sam-
ple selection in previous empirical studies is discussed and commented
upon. Methods of data capture used in this empirical analysis are dis-
cussed. The paper finds that whilst the expected upward trend in CSR is
notable, the more interesting feature in the longitudinal analysis is the
variability in disclosure between chairmen’s terms in office. It is argued
that marginal variability of disclosure can be explained by the varying
perceptions of reality of the successive chairmen. The limitations of exist-
ing theories as explicators of the observed phenomenon are briefly dis-
cussed.

Introduction
The phenomenon of corporate social reporting (CSR) has precipitated
a substantial body of research. The literature in the field seems to fall
broadly into two general categories. The first category comprises those
papers that discuss CSR and its significance in describing the relation-
ship between organizations and their constituencies (see for example
Morgan 1988; Pallot 1991), and those which discuss the issues sur-
rounding the social responsibility and organizational performance of
which CSR is said to be a partial discharge (see for example Brooks

Address for correspondence: Professor David J. Campbell, Department of Strategic

Management, University of Northumbria at Newcastle, Ellison Building, Ellison
Place, Newcastle Upon Tyne, NE1 8ST, UK. Email: david.campbell@
unn.ac.uk

 Blackwell Publishers Ltd. 2000, 108 Cowley Road, Oxford OX4 1JF, UK and 350 Main Street, Malden,
MA 02148, USA.
Accounting Forum Vol 24 No 1 March 2000 81

1986). The second category (as Preston 1981, observed), comprises

those papers that report upon empirical CSR studies in surveys of vary-
ing size and with a wide range of objectives in mind. In respect to this
second tranche of literature (those reporting on empirical studies), the
samples chosen vary in both cross-sectional and longitudinal measure.
Studies have been published on CSR in several countries, but there
seems to have been a concentration on studies of CSR in the major First
World economies, especially in the US, the UK, Canada, New Zealand,
Australia and in the Scandinavian countries.
This paper begins by briefly summarizing the most prominent expla-
nations for CSR from the relevant literature. Within the inquiry into
the organization-society interface, brief descriptions of the two most
prominent explanations for CSR are entered into and commented upon.
The possibility that CSR may be influenced by the personalities of senior
officers is introduced and briefly discussed. Next, sample selection is
discussed, particularly in respect to longitudinal ‘depth’ and cross-
sectional ‘width’. The advantages and disadvantages of the various sam-
pling strategies are briefly discussed. The paper then discusses the
methods that have been adopted in respect to capturing data on CSR
from corporate reports and a justification is offered for the data capture
method adopted in this paper. Next, the paper reports on the ‘raw’
findings of the empirical analysis. A range of categories of CSR disclos-
ure are reported upon and other relevant comments are made that
inform the enquiry upon which the paper is based. Finally, conclusions
are drawn from the analysis and comment is offered on the explanations
for CSR in Marks and Spencer Plc from the findings of the empirical
survey.

Explanations for CSR

Those who have written on CSR agree that it is a phenomenon which
it is difficult to find explanations for. It has been noted that the more
cogent explanations for CSR may have a connection with the stake-
holder debate that has been taking place in other areas of business
research (see for example Donaldson and Preston 1995; Azzone et al
1997). However, (as Gray, Kouhy, and Lavers 1995a, point out)
accounting research has been slower than some business disciplines to
explain disclosure phenomenon in the light of stakeholder theory (albeit
with its critics—see for example Den Uyl 1984 or for a more general
critique of social accounting see Wildavsky 1994).
A relatively small number of explanations have emerged as the most
commonly-discussed and the point has been made that they are not
necessarily mutually exclusive (on the contrary, Gray et al 1995a sug-
gest that are mutually supportive). It is not within the remit of this
 Blackwell Publishers Ltd. 2000
82 Campbell

paper to provide a comprehensive review of these explanations, but an

overview is offered to draw the distinction between these and the main
hypothesis of this paper.

Most of the literature has reflected upon explaining CSR in terms of

two broad themes. The political economy of accounting explanations
(see for example Cooper and Sherer 1984; Tinker 1980; Burchell et al
1980) arise from the discussion about power and information asymmet-
ries that subsist between agents and other constituencies. Explanations
of just how the effects upon CSR are exerted is not immediately obvi-
ous, but the redress of the asymmetry is seen as a central part of the
discharge of agents’ ‘responsibility’ to the non-economic constituencies.
The nebulous nature of the political economy explanation renders it
inaccessible to empirical testing until a harder definition is synthesized.

In terms of conventional orthodoxy, however, the purpose of account-

ing has traditionally been to facilitate responsible stewardship of inves-
tors’ funds and rational decision-making by representing the agents’
perception of reality, whether economic, sociological or technological,
to the principals (see Ijiri 1965; Horngren 1977; Cooper and Kaplan
1991 and Chandler 1991). Such a conventional understanding of the
raison d’être of accounting raises the question of which (or, more likely,
whose) understanding of reality is represented in conventional account-
ing—particularly in the extent to which management think they have
an element of accountability in respect to those elements for which they
account. Given the information and power asymmetries that subsist
between company agents and other constituencies (including
principals), there may be a case to argue (and this is the focus of this
paper) that the succession of company senior agents may usher in
slightly different perceptions of the reality that is to be communicated
via the non-mandatory sections of corporate reports. This is not an
enquiry into the macro understanding of accounting reality per se (see
for example Hines 1988; Morgan 1988), but rather, an enquiry into
whether, at the margins (and this point is stressed), agents’ perceptions
of local (i.e. specific to their company) reality may influence social dis-
closure as a discharge of their perception of their accountability. This
hypothesis is tested in this paper and is discussed at the end.

Complementary (according to Gray et al 1995a) to the political econ-

omy understanding of CSR is legitimacy theory. According to legit-
imacy theory, organizations exist in society under an expressed or
implied social contract. The early thinkers in the field (such as Hobbes,
Locke and Rousseau) viewed the social contract primarily as a political
theory insofar as it explained the supposed relationship between a
government and its constituencies. In the modern era, the principle of
social contract has been interpreted and developed for business insti-
 Blackwell Publishers Ltd. 2000
Accounting Forum Vol 24 No 1 March 2000 83

tutions by Rawls (1971, albeit implicitly), Donaldson (1982), Dowling

and Pfeffer (1975), Suchman (1995) and others.
Dowling and Pfeffer (1975) argue that organizations exist within a
super-ordinate social system (see also Parsons 1960) within which,
organizations enjoy legitimacy insofar as their activities are congruent
with the broad goals or acceptations of the super-ordinate system.
Accordingly, changes in the value system resident within the super-
ordinate social system become the pre-eminent cause of cultural change
within organizations and of change in the manner in which the organi-
zation relates to society. It is here that we encounter the link with CSR –
legitimacy theory would suggest that social disclosure can be used to
narrow the legitimacy ‘gap’ between how the organization wishes to be
perceived and how it actually is. Maurer (1971) discusses legitimacy
in terms of justification (of certain behaviours) whilst Suchman (1995)
describes it in terms of manipulation and garnering societal support.
In essence, legitimacy theory suggests a convergence of the type and
volume of disclosure with the concerns of the most influential stake-
holders. In this regard, legitimacy theory is, in principle, empirically
testable and some writers have attempted to report on this (see for
example Guthrie and Parker 1989).
Apart from the two foregoing explanations for the presence and varia-
bility of CSR, the literature contains few other serious explanations. It
is likely that signalling theory and agency theory may be applicable to
the phenomenon, although there have been no serious attempts to do
so thus far (see Watts and Zimmerman 1986; Jensen and Meckling
1976; Morris 1987; Bar-Yosef and Livnat 1984; Downes and Heinkel
1982; Fama 1980; Haugen and Senbet 1979).
This paper advances a novel explanation for changes in the volume of
CSR. Tattenbaum and Schmidt (1973) identify the personal value sys-
tems of key managers as an important determinant of their leadership
style and presumably, the resulting corporate culture (see also Thomas
1989). This notion seems intuitively reasonable, although the extent to
which senior managers’ personal values ‘infect’ culture and the organi-
zations attitude to social reporting will vary from company to company.
There would seem to be a case, therefore, for an empirical study to test
for the extent to which the personal values of senior managers affect
the company’s attitude to CSR.
The potential problems of setting up such a study are considerable. The
first and most obvious problem is to identify which senior officer whose
succession to test disclosure against. The second is the uncertainty of
the precise process of corporate report compilation within the company.
It may be that in some companies, senior management have little of no
input into the amount of social disclosure. This paper, however, makes
 Blackwell Publishers Ltd. 2000
84 Campbell

the pragmatic presupposition that strategic-level managers do have an

input, but this matter is probably a researchable topic in its own right.
This paper sets out to provide some insights into the explanations for
CSR by specifically testing for the influence of senior officers on the
company’s volume of CSR. It is suggested that if the succession of senior
management officers can be shown with a reasonable degree of confi-
dence to have an influence on CSR, then there may be a basis for chal-
lenging the hitherto most discussed explanations of CSR.

Sample selection
The summary of twenty-five years of CSR research by Mathews (1997)
shows the large number of empirical studies that have been undertaken
by researchers in this area. A detailed survey of the sample taken for
each of these surveys remains to be published, but for the purpose of
providing a background to this paper, a large number of prior empirical
studies were examined.
One of the most striking features of the samples selected for analysis is
that most empirical CSR studies have privileged cross-sectionality over
longitudinality. ‘Deep’ longitudinal studies are conspicuous by their
scarcity whilst other papers privilege cross sectionality over longitudi-
nality, some of whom select one year only.

Method
This paper set out to establish a data set from which some insights into
the explanations into CSR could be made. Accordingly, and specifically
in order to test for trends, a priority was placed upon longitudinality.
One company was chosen and as many back issues as were available
were obtained for the analysis. For the company in question (Marks
and Spencer Plc), most corporate reports were available in hard copy
whilst a minority were read on microfiche from files obtained from UK
Companies House. The survey consists of all corporate reports between
1969 and 1997 (inclusive), excepting the years 1971 and 1972 which
were unavailable to the author.

Source of data for analysis

The matter of which company communications to include in an analysis
is one of the most difficult. In principle, a study of CSR would include
all forms of data reaching the public (see Gray et al. 1995b), but in
practice, this is rarely possible.
The overwhelming majority of researchers have taken the pragmatic
view that the annual corporate report can be accepted as an appropriate
 Blackwell Publishers Ltd. 2000
Accounting Forum Vol 24 No 1 March 2000 85

source of a company’s attitudes towards social reporting. This is for

two reasons:
1. the company has total editorial control over the document
(excluding the audited section);
2. it is usually the most widely distributed of all public documents
produced by the company.
To optimise resolution, data was recorded at three locations in the each
document; in the chairman’s statement, in the directors’ report and in
‘other’ locations.

Company selection
The selection of Marks and Spencer Plc as the subject of the analysis
was partly serendipitous, but it was thought that this was a company
that might be worthy of analysis due to a number of company-
specific features:
앫 it has a reputation in the UK as a retailer of premium, safe products;
앫 it has explicitly stated that it considers itself to have a social
responsibility (see for example Marks and Spencer Plc corporate
report 1997, p. 5);
앫 in common with many other well-known high street names, it argu-
ably has good reason to be rather more sensitive to opinion than
businesses in some other commercial sectors;
앫 it has enjoyed a reputation for many years as a good employer with
a paternalistic management regime;
앫 it has consistently promoted to senior positions from within, thus
reducing any dilution of senior management culture and conviction
by the appointment of outsiders.
In contrast to some other sectors (e.g. petrochemicals), most observers
would not recognise high street retailers as being a business sector with
a great deal to justify in terms of alleged ‘bad behaviour’. However,
given the company’s reputation and the sensitivity of retailers in general
to social opinion in a competitive market, it was decided that the com-
pany would present an opportunity for a longitudinal analysis of CSR
spanning a number of potential trends and switch points in societal,
political and economic opinion, such as the increase in ethical aware-
ness as an issue in retailing, as well as several changes in chairmanship.

Hypotheses and switch points

The purpose of this survey is explicitly to provide some insight into the
causes of changes in CSR in the company in question. However, given
 Blackwell Publishers Ltd. 2000
86 Campbell

that only one company is studied, any conclusions drawn must obvi-
ously be heavily qualified.
A cross-sectional sample of one in any population of any size is unlikely
to provide a statistically significant or generalizable analysis. Rather
than attempt to draw indefensible conclusions in respect of confirming
or rebutting a hypothesis with such a sample (this being a criticism of
Guthrie and Parker, 1989), it is stated at the outset that the analysis
in question is designed to provide indications and to provoke further
discussion of any findings.
Accordingly, the method is designed to chart CSR over the period in
question with a view to identifying both trends in presentation and dis-
closure against time, and to report any switch points at which trends
changed direction. Given the sample size, it would probably be inappro-
priate to systematically test for association with internal or external
events which may have had an antecedental effect upon the change in
disclosure although some comments are made in respect to coincidence
with some socio-political events.
However, it was felt that some insights into management attitude to
CSR might come to light if variability between senior officers’ suc-
cessions could be demonstrated to be statistically significant. The matter
of which senior officer to test against (e.g. chairman, chief executive,
etc.) is problematic. In the case of Marks and Spencer Plc, the managing
director/chief executive roles are split between several executive direc-
tors so it was thought that testing against one or more joint-CEO would
muddy the analysis somewhat. The most obvious senior officer to test
against thus becomes the company chairman, which, in the case of ‘M
and S’ is an executive role. The chairmanships are superimposed on the
longitudinal graphs in this paper to show changes in CSR patterns
within and between the respective terms of office.

Content analysis—what to study

The three social categories were selected because they are all, in some
way, dependent upon the activities and policies of the company.
Accordingly, customers as a social constituency are excluded (because
the direction of dependence is the other way round).
The database into which the data was entered contained seven rel-
evant fields.
1. year end of corporate report;
2. total pages in corporate report;
3. total non-financial pages in corporate report;
4. identity of chairman;
 Blackwell Publishers Ltd. 2000
Accounting Forum Vol 24 No 1 March 2000 87

5. employee welfare, including equal opportunities, but excluding

financial information on employee pensions;
6. community involvement including charitable work sponsored;
7. environmental disclosure including energy conservation.

Methods of data capture

It has been observed (Gray et al 1995b) that data capture and measure-
ment in the field of CSR falls into two broad categories—the number
of disclosures (e.g. frequency of mentions of category x) and amount
of disclosure. The majority or researchers in the CSR field have chosen
to capture data based upon the amount because it gives a richer data
set and this paper adopts the same approach. Both number of words
and page proportions were recorded for each piece of social infor-
mation – a more labour-intensive approach than that taken by other
researchers but one intended to give a higher level of resolution. The
problem of measuring what is said in social disclosure (the ‘quality’
rather than just the quantity) is one which most researchers have found
it convenient to dismiss. This is because it is almost always ‘good news’
that is being communicated. The surrogate for measuring the com-
pany’s response to social issues is that the company discloses at all and
by how much. That it doesn’t account for qualitative assessment is a
acceptable limitation that most empirical work has allowed.

Methods of data recording and analysis

In order to record and subsequently analyse the data collected, a data-
base was set up in Microsoft Excel in which statistical associations
could be drawn and graphical representations of trends and switch-
points generated. In order to ensure internal consistency, data capture
was undertaken throughout the investigation by the author.
The data was entered into fields referring to the three specific locations
within the corporate reports (chairman’s statement, directors’ report
and other). When the data entry was completed, the mathematical abili-
ties of the package were employed to generate sums of each category
of social information for disclosure in the whole corporate report.
Ratios were calculated to show how numerical data were related to
each other as a function of time, for example by calculating total page
proportions of CSR as a percentage of total non financial pages in the
corporate report document.

Findings
Physical and aesthetic presentation
One of the most immediately noticeable features of the ‘M and S’ cor-
porate reports over the thirty-year period is the change in the physical
 Blackwell Publishers Ltd. 2000
88 Campbell

presentation of the documents themselves. Two trends are potentially

noteworthy: and increase in the length of the document with the passing
of years and an increase in the complexity of the presentation. In the
early years of the study, the reports were typically based upon a pro
forma design and were 19 pages long. With the accession of Lord
Raynor to the chairmanship in 1985, photographs appeared for the first
time. The length increased to a peak of 80 pages in the 1990s.

Total voluntary social disclosure against time

It is not surprising to report that the aggregate social reporting increased
over the period in question (see for example Spicer 1978, Dierkes and
Antal 1985) but the increase was not consistent—fluctuations can be
seen within the upward trend. The introduction of photographs in 1985
is responsible for the upturn in page proportions at that time (see Figure
1). The word count and page proportion count measures are relatively
comparable, with the greatest periods of increase coming in the 1980s
(Figures 1 and 2).

Employment-issues disclosure
In every year of the period under analysis, ‘M and S’ made some volun-
tary disclosure in respect of its employees and their welfare. The Chair-
man’s Statement has consistently contained a word of thanks to the
company’s employees, but as this was in addition to customary thanks
to customers and suppliers, and was perfunctory in content, this content
was excluded from the word and page count. The Director’s Report

Figure 1 – Social disclosure (by category and total) by page proportion

count.
 Blackwell Publishers Ltd. 2000
Accounting Forum Vol 24 No 1 March 2000 89

Figure 2 – Social disclosure (in all categories) by word count.

introduced a statement on employee issues in 1982, the wording of

which remained substantially unchanged in subsequent years (despite
any changes that may have occurred in societal opinion). A statement
on equal opportunities is included in the Director’s Report entry but
its content has also remained almost unchanged since its initial intro-
duction except for the inclusion of sexual orientation in 1997 as a basis
upon which the company will not discriminate (see also Adams, Coutts
and Harte, 1995).
The high fluctuations that occurred in employee welfare disclosure in
the later years of the period arise mainly because of the variations in
the space committed to the matter in separate internal employee reports.
A more sporadic approach to employee disclosure was exhibited in the
Chairman’s Statement (Figure 4) and in other sections of the report
(Figure 3). From the late 1970s onwards, a separate (though internal
to the corporate report) Employee Report was included. In later years,
the Employee Report was supported by one or more annotated photo-
graphs which increased the page proportion count but obviously not
the word count.
The contrast in employee welfare disclosure between Chairmen Raynor
and Greenbury demonstrates the two men’s emphasis on social disclos-
ure (Figure 4). Although the employee welfare word count in Sieff’s
and Raynor’s statements were variable, it is noticeable that Greenbury
declined the opportunity to make any employee disclosures at all after
his first year in office. It is probably reasonable to assume that the chair-
man’s attitude towards employee welfare disclosure would be reflected
in his own statement, and this is supportive of the hypothesis that the
 Blackwell Publishers Ltd. 2000
90 Campbell

Figure 3 – Social disclosure by word count—employee welfare.

Figure 4 – Words on employee welfare in chairman’s statement (note that

1972 and 1973 are missing).

chairman’s attitude towards social responsibility is a major determinant

of the volume of CSR.

Community disclosures
The voluntary social disclosures in respect of communities are perhaps
one of the more interesting features of this analysis because of the com-
pany’s explicit claims to recognise its social responsibility. The 1996
report contains the following statement:
‘Our commitment to social responsibility is shown by the full part we
play in the community . . .’ (Annual report and Accounts, 1996, p. 4)
 Blackwell Publishers Ltd. 2000
Accounting Forum Vol 24 No 1 March 2000 91

Prior to 1975, the company made no reference to the communities in

which it operated which is an interesting observation given that in 1995,
the corporate report stated that,
‘Marks and Spencer Plc has always acknowledged a corporate
responsibility to society . . .’ (Annual report and Accounts, 1995,
p. 30). Author’s emphasis.

Over the period of this study, the amount of disclosure on community

relations has fluctuated significantly (see Figure 5) and until 1986, dis-
closure was almost entirely located within the chairman’s statement.
After this date, community disclosure in the chairman’s statement
decreased as a significant increase was accounted for by the introduc-
tion of a separate Community Report. This innovation coincided with
the accession to the chairmanship of Lord Raynor.
Within this category of social disclosure, a number of social ‘campaigns’
may be noteworthy including an entry in the 1994 report stating the
company’s opposition to Sunday trading (in response to Parliamentary
proceedings in respect of this matter) and a statement in the 1981 chair-
man’s statement (during a recession and at the time when unemploy-
ment hit a high of three million) that,
‘We have a responsibility to help young people who have difficulty in
finding jobs.’ (Annual report and Accounts, 1981)
This remarkably socially aware 1981 attitude to unemployment in gen-
eral was not repeated in any subsequent corporate report and may have
consequently been a partial response to the high levels of unemployment
at the time.

Figure 5 – Social disclosure by word count—community and community

relations.
 Blackwell Publishers Ltd. 2000
92 Campbell

Environmental disclosure
For the purposes of this investigation, environmental disclosure was
conveniently taken to mean any mention of environmental responsi-
bility or any mention of energy conservation, energy efficiency or waste
reduction. Given the longitudinal sample under analysis, it is perhaps
not altogether surprising that this broad definition drew in some disclos-
ures that were not necessarily referring to environmental disclosure in
the contemporary understanding of the term.
When taken to mean reference to the natural environment, it can be
reported that the first explicit disclosure on the environmental behav-
iour of the company was in 1989. The volume of disclosure then sub-
sequently increased (as a trend-line) from 1989 up until the end of the
period under analysis (1997). During the 1970s, however, the company
made voluntary disclosures on its energy and waste performance in a
period before the natural environment became a topical issue (see Wor-
cester 1995, 1996). In context, we may postulate that such disclosures
were in response to a number of external political, economic and socio-
logical events including the mid 1970s oil price shock, the miners’
strike, the three-day week, etc. The primary concern on the part of the
company in making such disclosures (making the above postulate) was
not so much to respond to environmental opinion, but rather to respond
to a societal climate of energy shortages and its concomitant austerity.
Figure 6 shows the company’s environmental disclosure by word-count.

Figure 6 – Social disclosure by word count—environment.

 Blackwell Publishers Ltd. 2000
Accounting Forum Vol 24 No 1 March 2000 93

Discussion
Consistency with legitimacy explanations
In order to test to see the extent to which the findings are consistent
with legitimacy theory, we need to examine the extent to which the
company’s disclosures are concerned with justification (Maurer 1971),
garnering societal support (Suchman 1995) or generally rising or falling
as public opinion changes on the key subjects discussed in the disclos-
ures.
This study has not made a detailed textual examination of the social
disclosures that it has reported upon in terms of their volumes. Accord-
ingly, the point is conceded that the company may have made some
disclosures in respect to self justification or in order to garner societal
support. Whether or not this has occurred is not central to the main
theme of this paper. It is also possible that the variations in CSR over
the period of the study accord to some extent with changes in societal
opinion (such as the disclosures reported above in response to oil price
rises and high unemployment).
Testing for any of these features using an empirical instrument is neces-
sarily fraught with potential difficulties due to the problem of providing
a reliable mechanism of charting changes in societal opinion (or of cir-
cumscribing and defining the value system resident within the super-
ordinate social system).
This is a conspicuous omission in the literature that discusses legitimacy
theory—it is all very well to discuss it in vacuo (i.e. in abstraction), but
it is an altogether more challenging research task to arrive at an empiri-
cal method for accurately charting opinion change over the course of
several decades.
So can we say that the findings are consistent with legitimacy theory?
The answer is that they might be—but this is not the point. The signifi-
cant variability of disclosure suggests, however, that legitimacy theory
is not a total explanation of the phenomenon observed in this study.
We can in fact trace some trends in CSR from the ‘M and S’ corporate
reports that may be partially explained by legitimacy theory. There is,
for example, some documented evidence of a rise in some aspects of
environmental concern over recent years (Worcester 1996). In then dis-
covering that ‘M and S’ included explicit statements on environment
from 1989 onwards may suggest an apparent consistency with legit-
imacy theory, which it may well be (i.e. an apparent consistency). The
acknowledgement of some responsibility for the unemployed in 1981
(see above) may also be offered in evidence for this theory.
There is, however, some evidence that some aspects of ‘M and S’ dis-
 Blackwell Publishers Ltd. 2000
94 Campbell

closure may actually run counter to societal opinion. Political

donations, small though they were, were given directly to the Conserva-
tive Party near to the time when the Party’s popularity began to decrease
after the 1992 general election and ‘Black Wednesday’ in the same year.
The company made its opposition to Sunday trading clear in the early
1990s when Parliament (presumably believing it to be a popular
measure) were legislating for it. A company seeking approval
(legitimacy) from society would be unlikely to espouse unpopular causes
on a regular basis.

Managerial reality construction as an explicator

This paper does not reject the hypothesis that social disclosure in the
‘M and S’ corporate reports may be, at least in part, consistent with
a legitimacy understanding of CSR. It does, however, suggest that in
the case of this company (and this observation is not necessarily
generalizable), another explanation is a rather more plausible one.
Over the thirty-year period of the survey, ‘M and S’ had four chairmen,
all of whom were promoted from within and each of whom had pre-
viously served for several years as executive officers. It is shown by
this paper that the key switch points in the aggregate volume of social
disclosure coincide closely with the points of succession of consecu-
tive chairmen.
Two ‘page proportion count’ ratios were calculated for each year.
Firstly, the percentage of CSR information by page count over total
pages in the corporate report gives some indication of the emphasis in
the document on its attitude to social reporting. Secondly, the percent-
age was calculated of CSR information by page count over all pages in
the corporate reports after the number of mandatory financial pages
were subtracted (note that ‘non-financial pages’ is a field in the CSEAR
database—see Gray et al 1995b). The latter ratios may provide a higher
level of meaning insofar as it is an indication of the company’s attitude
to CSR in respect to other, mainly discretionary disclosure.
Both of these ratios were plotted against time. The points of chairmen’s
succession are superimposed over the graph (Figure 7). Even a cursory
glance at this Figure reveals a striking coincidence.
It would appear (prima facie) from Figure 7 that two key successions
of chairman may be in part responsible for the apparent switch points
in the company’s observable attitude to CSR. The succession from J
Edward Sieff to Sir Marcus Sieff (later Lord Sieff of Brimpton) was
apparently without incident in respect of CSR (in all categories). In
1985, however—the first published corporate report under The Lord
Raynor—we witness not only a ‘glossy’ format document complete with
 Blackwell Publishers Ltd. 2000
Accounting Forum Vol 24 No 1 March 2000 95

Figure 7 – Percentage of corporate report given to CSR by page pro-

portion total.

photographs for the first time, but also an immediate reversal in Sieff’s
declining CSR from a high in 1981. As The Lord Raynor’s chairman-
ship progressed, CSR increased year-on-year reaching its zenith at his
retirement in 1991. At this point—the succession of Sir Richard Green-
bury—CSR showed an immediate decline. Any speculation as to what
effect the personality of the chairman may have had on CSR would, of
course, be both premature and necessarily uninformed. It is clearly evi-
dent, however, that the switch points are both marked and closely
coincident with the two most recent chairmen’s successions.
Another indication of the chairman’s personal influence on the volume
of CSR may be seen from his personal direct input into the amount

Table 1 – CSR disclosure in the chairmanships of the three most recent ‘M and S’
chairs. Observations are CSR in all categories by page count over total non-financial
pages in corporate report. Each observation represents the calculation for a year

Sieff Raynor Greenbury

4.64 5.09 16.50

2.94 7.27 9.37
9.06 11.48 10.73
10.31 13.41 11.43
7.33 13.93 13.37
8.43 14.34 11.32
9.13 9.33
9.8
15
12.81
9.45

 Blackwell Publishers Ltd. 2000

96 Campbell

from the chairman’s statement. Figure 8 shows the word count of CSR
in all categories in the chairman’s statements over the period in ques-
tion. The same trend observed above is also in evidence here. The stark-
est switch point is that between Lord Raynor’s retirement in 1990 and
Richard Greenbury’s accession in 1991. In three of the seven years of
Greenbury’s tenure, he chose to make no social disclosure in his chair-
man’s statement at all. There is little doubt that one reason for this may
be the dedication of other parts of the report to social matters, but
the different stress on social matters between Raynor and Greenbury is
probably illustrated within this measure.

Figure 8 – Pages CSR in all categories by word count in the chairman’s

statement (note that 1972 and 1973 reports are missing).

Conclusions
In attempting to synthesise an explanation for CSR changes in the case
of this longitudinal study, there are two observations to be made.
Firstly, the overall upward linear trend over the period from the mid
1970s to the late 1990s is ostensibly consistent with a legitimacy under-
standing of CSR assuming that the items addressed in the disclosures
are of increasing concern to the public over the period (see Worcester
1995 and 1996). However, the observation that CSR has increased
since the 1970s is not new (see for example Gray et al 1995a and
1995b).
Secondly, the fact that there are significant variabilities in CSR volume
as a percentage of non-financial total between chairmen, may indicate
some variations in the perception of reality by ‘M and S’ under the
regimes of the successive chairmen. This paper is unable to generalise
 Blackwell Publishers Ltd. 2000
Accounting Forum Vol 24 No 1 March 2000 97

this observation beyond the cross-sectional sample of one that has been
its subject, but it is suggested that marginal variations (at least) in CSR
volume within a broader upward trend may indicate different percep-
tions of reality on the part of the reporting chairmen.
Admittedly, this observation rests in large part upon the presuppo-
sitions that (a) the chair substantially imprints his perception of reality
on the senior officers sufficient to affect the volume of CSR, and (b)
that the chairman is the only internal independent variable that has an
effect on the volume of CSR. In the case of ‘M and S’ , the latter of
these two presuppositions may have some credibility owing to the fact
that other senior Board positions are shared (i.e. the company operates
with three-to-four joint managing directors).
In conclusion, the findings of this study suggest that contractarian
theories may not be totally adequate as explicators for changes in cor-
porate social reporting. Political economy explanations do not easily
lend themselves to empirical study. These have not been tested for,
although most discussions suggest that within this explanation, simi-
larities between disclosure and societal values would be expected if the
explanation is to be given any credence (see Woodward et al, 1998).
There does, however, appear to be a managerial influence on the com-
pany’s attitude to voluntary social disclosure and this is not readily
explained by the meta theories espousing stakeholder theory (see also
Ullmann, 1985).
Data capture by page count has the limitation of being, of necessity,
inexact—not in the counting itself (which is a systematic procedure) but
in the extent to which pages can be taken to be a good indicator of
social disclosure. Accordingly, this paper makes no claim other than
the results may be indicative of that to which it refers.
It is argued, however, that the findings are sufficient justification for
widening the enquiry to see if it can be observed in a wider sample of
companies. A wider cross-sectional sample over a similar longitudinal
period would provide a richer data set upon which to develop this poss-
ible explanation for CSR.

Acknowledgements
The author is indebted to a number of people for input and advice
before and during the preparation of this paper. The usual disclaimer
applies. Helpful feedback was received from delegates at the BAA-
ICAEW Doctoral Colloquium (University of Lancaster), Jan 1998, at
the CSEAR Summer School (University of Dundee), Sept. 1998, from
Prof. David Owen (University of Sheffield), from Claire Marston
(University of Durham) and from several colleagues at the University
 Blackwell Publishers Ltd. 2000
98 Campbell

of Northumbria at Newcastle – Philip Shrives, Geoff Moore, Barrie

Craven, Rob Gausden and Simon Lillystone. Funding in support of the
research was provided by the University of Northumbria at New-
castle, UK.

References
Adams, C. A., Coutts, A. and Harte, G. (1995). Corporate equal
opportunities (non-) disclosure. British Accounting Review. 27, 87–108.
Azzone, G., Brophy, M., Noci, G., Welford, R. and Young, W. (1997). A
stakeholder’s view of environmental reporting. Long Range Planning, 30
(5): 699–709.
Bar-Yosef, S. and Livnat, J. (1984). Auditor selection: an incentive signalling
approach. Accounting and Business Research. (Autumn).
Brooks, L. J. (1986). Canadian Corporate Social Performance, Hamilton,
Ontario, The Society of Management Accountants of Canada.
Burchell, S., Clubb, C., Hopwood, A., Hughes, J. and Nahapiet, J. (1980).
The role of accounting in organizations and societies. Accounting,
Organizations and Society 5 (27).
Chandler, I., Fetter, R. and Newbold, R, (1991). Cost accounting and
budgeting, in Freeman, R. (ed.). DRGs. Ann Arbor: Health
Administration Press.
Cooper, R. and Kaplan, R. (1991). The design of cost management systems.
NJ: Prentice Hall.
Cooper, D. J. and Sherer, M. J. (1984). The value of corporate accounting
reports: arguments for a political economy of accounting. Accounting,
Organizations and Society. 9 (3/4): 207–232.
Den Uyl, D. J. (1984). The new crusaders: The corporate social
responsibility debate. Studies in Social Philosophy and Policy. 5.
Dierkes, M. and Antal, A. B. (1985). The usefulness and use of social
reporting information. Accounting, Organizations and Society. 10 (1):
29–34.
Donaldson, T. (1982). Corporations and Morality. NJ: Prentice Hall.
Donaldson, T. and Preston, L. E. (1995). The stakeholder theory of the
corporation: concepts, evidence and implications. Academy of
Management Review. 20 (1): 65–91.
Dowling, J. and Pfeffer, J. (1975). Organizational legitimacy: Social values
and organizational behavior. Pacific Sociological Review. 18 122–136.
Downes, D. H. and Heinkel, R. (1982). Signalling and the valuation of
unseasoned new issues. Journal of Finance. March.
Fama, E. F. (1980). Agency problems and the theory of the firm. Journal of
Political Economy. April.
Gray, R., Kouhy, R. and Lavers, S. (1995a). Corporate social and
environmental reporting. A review of the literature and a longitudinal
study of UK disclosures, Accounting, Auditing and Accountability
Journal. 8 (2): 47–77.
 Blackwell Publishers Ltd. 2000
Accounting Forum Vol 24 No 1 March 2000 99

Gray, R., Kouhy, R. and Lavers, S. (1995b). Methodological themes:

Constructing a research database of social and environmental reporting
by UK companies . Accounting, Auditing and Accountability Journal. 8
(2): 78–101.
Guthrie, J. E. and Mathews, M. R. (1985). Corporate social accounting in
Australasia. Research in Corporate Social Performance and Policy. 7
251–77.
Guthrie, J. E. and Parker, L. D. (1989). Corporate social reporting: a
rebuttal of legitimacy theory. Accounting and Business Research. 9 (76):
343–52.
Guthrie, J. E. and Parker, L. D. (1990). Corporate social disclosure practice:
a comparative international analysis. Advances in Public Interest
Accounting. 3 159–76.
Haugen, R. A. and Senbet, L. W. (1979). New perspectives of informational
asymmetry and agency relationships. Journal of Accounting and
Quantitative Analysis. November.
Hines, R. (1988). Financial accounting: in communicating reality, we
construct reality. Accounting, Organizations and Society. 13 (3): 251–
61.
Horngren, C. T. (1977). Cost accounting: a managerial emphasis. NJ:
Prentice Hall.
Ijiri, Y. (1965). Management goals and accounting for control. NY: North
Holland.
Jensen, M. C. and Meckling, W. H. (1976). Theory of the firm: managerial
behaviour, agency costs and ownership structure. Journal of Financial
Economics. October.
Mathews, M. R. (1997). Twenty five years of social and environmental
accounting research—is there a silver jubilee to celebrate? Accounting,
Auditing and Accountability Journal. 10 (4) 481–531.
Maurer, J. G. (1971). Readings in organizational theory: Open systems.
Approaches. NY: Random House.
Morgan, G. (1988). Accounting as reality construction: towards a new
epistemology of accounting practice. Accounting, Organizations and
Society. 13 (5) 477–485.
Morris, R. D. (1987). Signalling, agency theory and accounting policy
choice. Accounting and Business Research. 18 (69) 47–56.
Pallot, J. (1991). The legitimate concern with fairness: a comment.
Accounting, Organizations and Society. 16 (2) 201–208.
Parsons, T. (1960). Structure and process in modern societies. NY: Free
Press.
Preston, L. E. (1981). Research on corporate social reporting: directions for
development. Accounting, Organizations and Society. 6 (3) 255–262.
Rawls, J. (1971). A Theory of Justice. Cambridge, MA: Harvard University
Press.
Spicer, B. H. (1978). Investors, corporate social performance and
information disclosure: an empirical study. The Accounting Review, LIII
(1).
 Blackwell Publishers Ltd. 2000
100 Campbell

Suchman, M. C. (1995). Managing legitimacy: strategic and institutional

approaches. Academy of Management Review. 20 (3) 571.
Tattenbaum, R. and Schmidt, W. H. (1973). How to choose a leadership
pattern. Harvard Business Review. May–June. 162–75, 178–180.
Thomas, A. P. (1989). The effects of organizational culture on choices of
accounting methods. Accounting and Business Research. 19 (76) 363–
378.
Ullmann, A. A. (1985). Data in search of a theory: A critical examination of
the relationships among social performance, social disclosure and
economic performance of US firms. Academy of Management Review.
10 (4) 540–57.
Watts, R. L. and Zimmerman, J. L. (1986). Positive Accounting Theory. NJ:
Prentice Hall.
Wildavsky, A. (1994). Accounting for the environment. Accounting,
Organizations and Society. 19. (4/5) 461–481.
Woodward, D., Edwards, P. and Birkin, F. (1998). Some evidence on
executives views of corporate social responsibility and its reporting.
Proceedings from the International Conference on Accounting and
Governance. 23–24 Oct. 1998. Naples, Italy.
Worcester, R. M. (1995). Business and the environment—the predictable
shock of Brent Spar. HRH The Prince of Wales’ Business and The
Environment Programme. University of Cambridge Programme for
Industry. 18 September 1995.
Worcester, R. M. (1996). Greening the millennium—public opinion and the
environment. (A MORI research paper). Greening the Millennium
Conference. Centre for the Study of Environmental Change, Lancaster
University, UK. 27 September 1996.

Appendix—Marks and Spencer Plc,

chairmen since 1969
1969–73 J. E. Sieff
1973–85 Sir Marcus Sieff/Lord Sieff of Brimpton
1985–91 The Lord Raynor
1991–99 Sir Richard Greenbury

 Blackwell Publishers Ltd. 2000