The construction of auditability MBA rankings and assurance in practice

Description
For much of the past decade, the audit profession has been enjoined to enter new and novel fields of assurance services.
This call implies the importation of constructs from traditional attest financial audits into new domains to provide elevated
levels of assurance for information-users and decision-makers

The construction of auditability: MBA rankings
and assurance in practice
q
Clinton Free
*
, Steven E. Salterio, Teri Shearer
School of Business, Goodes Hall, Queen’s University, Kingston ON, Canada K7L 3N6
Abstract
For much of the past decade, the audit profession has been enjoined to enter newand novel ?elds of assurance services.
This call implies the importationof constructs fromtraditional attest ?nancial audits into newdomains toprovide elevated
levels of assurance for information-users and decision-makers. Little is presently known about the process by which audit
scope, practices and communications about the work done are developed in these new ?elds. This paper attempts to shed
light on these issues through an in-depth ?eld study of KPMG’s ‘‘audit” of the Financial Times MBA rankings. The audit
project is argued to import legitimacy to data provided by International Business Schools as well as imbue a derived legit-
imacy to the Financial Times rankings. At an operational level, audit planning, procedures and communicatedwritten con-
clusions emerge as a much more negotiated and adaptive practice than rhetoric might suggest.
Ó 2008 Elsevier Ltd. All rights reserved.
Introduction
In the past 25 years in much of the English
speaking world, the label of ‘‘audit” has come to
be used with growing frequency to describe a wide
range of practices. Especially in the pre-Enron era,
assurance services have been widely heralded as a
remedy to stagnating audit fees and as the future
of the public accounting ?rms (Elliott, 1995),
enabling the penetration of the concepts and ter-
minology of traditional attest auditing into arenas
outside of ?nancial accounting.
1
Although the
competitive environment of public accounting
0361-3682/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.aos.2008.02.003
q
We appreciate the comments and suggestions of David
Cooper, Yves Gendron, Wai-Fong Chua and participants at
the 2007 Alternative Perspectives in Accounting Research
Conference, Universite´ Laval, Que´bec City. We thank Andrea
Davis for her excellent research assistance. We also thank
Michelle Podhy and Patrick Gaudet of KPMG Canada for
assistance with this research project and for comments on this
paper. Please note that the interpretations herein are those of
the paper’s authors and not those of KPMG, its partners or
sta?. The authors are solely responsible for any errors.
*
Corresponding author. Tel.: +1 613 533 3255; fax: +1 613
533 6589.
E-mail address: [email protected] (C. Free).
1
In 1996, the AICPA’s Elliott Committee identi?ed six areas
(risk assessment, business performance measurement, informa-
tion systems reliability, electronic commerce, healthcare per-
formance measurement and elder care) considered to have high
potential for revenue growth for assurance providers (AICPA,
1996; Elliott, 1995).
www.elsevier.com/locate/aos
Available online at www.sciencedirect.com
Accounting, Organizations and Society 34 (2009) 119–140
?rms has been transformed (Gri?n & Lont, 2005)
in the aftermath to a number of high-pro?le ?nan-
cial scandals such as those of Enron, Worldcom
and Tyco, the public accounting profession retains
a crucial role in regulation and supervision systems
and continues to extend its in?uence into diverse
domains of organizational and societal life.
Like any practice, auditing is characterised by
both programmatic (normative) and technological
(operational) elements (Power, 1997). The former
elements relate to ideas and concepts which shape
the mission of practice and draw on policy objec-
tives existing in the political sphere. At the level
of programmes, broad goals are formulated with
practices implicitly assumed to be capable of serv-
ing these goals. The programmatic level is the level
at which audits, for example, are demanded by
regulatory agencies, uncertain social constituencies
and policy discourse. In contrast, operational ele-
ments are the concrete tasks and routines which
constitute the world of the practitioner. The oper-
ational bedrock of auditing lies in samples, analyt-
ical methods and procedures (e.g., Knechel,
Salterio, & Ballou, 2007). At the level of opera-
tions, practitioners constantly debate the e?ciency
of di?erent methods and seek to engender cost-e?-
cient solutions to the daunting challenge of provid-
ing assurance over matters that can be considered
inherently subjective (e.g., Bell, Peecher, & Solo-
mon, 2005; Robson, Humphrey, Khalifa, & Jones,
2007).
Following Power (1997), we argue that there is
a loose coupling between the level of program-
matic appeal and the level of audit practice.
2
Power (1997) refers to this as ‘‘the essential obscu-
rity” of auditing and regards this as a constitutive
feature of practices like auditing whose criteria of
e?ectiveness are opaque – and which therefore
invest heavily in due process. Indeed, the power
of the idea of ‘‘audit” – and its ready exportability
from the ?nancial audit context – rests on a vague-
ness about its scope and meaning and its capacity
to lend legitimacy to the new context.
There has been little empirical research examin-
ing new auditable contexts in practice, especially in
the private sector (Cooper & Robson, 2006). Fur-
ther, there have been even fewer empirical studies
of the substance of evolving discretionary assur-
ance services.
3
In a rare ?eld-based contribution,
Radcli?e (1998, 1999) has documented the nature
of ‘‘e?ciency” or ‘‘value for money” audits in
the public sector of one Canadian province, illus-
trating the way that auditing can become infused
with the wider programmatic aims of public sector
reform. In a related study, Gendron, Cooper, and
Townley (2007) employ actor-network theory to
describe and interpret the processes by which state
auditors establish themselves as legitimate arbiters
of performance in the public sector. Their work
provides a rare window on the fragile and negoti-
ated process of network-building that ultimately
enabled state auditors to claim a publicly legiti-
mate role in the implementation of new public
management. However, in these studies it was
uncontested that the state auditor at least nomi-
nally had the right to audit hence the research
focused on the subject matter to be audited, why
those subjects (and not others) were chosen and
how the auditor constructed the claims to expertise
necessary to support their legitimate involvement.
In a di?erent context, but also using actor-net-
work theory, Gendron and Barrett (2004) examine
attempts by North American accounting institutes,
in an unstable and uncertain coalition with some of
the then Big 5 accounting ?rms, to cultivate a mar-
ket in eCommerce assurance through a longitudinal
?eld study. They ?nd that the attempted top down
planned expansion of the audit domain, mobilizing
an alliance of professional institutes andsome of the
2
This notion of a loose coupling connects with the ‘‘expec-
tations gap” – the gap between external auditors’ understanding
of their role and duties and the expectations of various user
groups and the general public (Epstein & Geiger, 1994;
McEnroe & Martin, 2001). We argue that the expectations
gap is itself as much a resource for the auditing ?eld as it is a
problem. A loose coupling between the potential and opera-
tional capacity of audits supports a certain ambiguity about the
audit process and its meaning, which in turn supports practi-
tioner discretion and buttresses practitioner power.
3
Assurance services are de?ned as ‘independent professional
services that improve the quality of the information or its
context for decision makers.’ (AICPA, 1996). The ‘audit’, from
the viewpoint of professional accounting institutes, refers to the
audit of ?nancial statements whereas the terms ‘attest’ and
‘assurance’ refer to broader engagements beyond ?nancial
statements (Knechel et al. 2007, pp. 12–13).
120 C. Free et al. / Accounting, Organizations and Society 34 (2009) 119–140
large audit ?rms, re?ects a trial and error process
that was both fragile and dynamic.
While these studies provide insights into the way
audits are constructed in non-traditional contexts,
the antecedents and character of the assurance pro-
ject they identify are unlikely to be at work in our
more market-oriented context. Speci?cally, there
is no established ‘right to audit’ as in the public sec-
tor where new accountability regimes are fostered
by a new public management approach to govern-
ment. Our context also does not re?ect a top down
industry wide planning exercise that attempts to
mobilize a broad constituency of leading audit
?rms to establish a new market space. Hence we
are examining the establishment of an ‘‘audit” in
a novel private sector context where the auditor is
engaged by the Financial Times to audit the data
provided by the business schools that enter into
the Financial Times (FT) MBA rankings. Thus,
the ?rst research question is: what are the anteced-
ents and character of an assurance project in a
‘‘new” unregulated context (as opposed to the reg-
ulated market for ?nancial attest audits and other
non-traditional areas examined in the literature)?
A related concern is the way that assurance
operations are devised and enacted in practice. In
recent years, a concern with the disconnect
between management theory and managerial
action has given rise to the so-called ‘‘practice”
approach in the management literature (Ahrens
& Chapman, 2005, 2007; Lounsbury, 2007).
Recent contributions in this ?eld have considered
strategy as an adaptive and recursive practice (Jar-
zabkowski, 2004), management control as an array
of activity delineated through processes of distrib-
uted cognitions (Ahrens & Chapman, 2005, 2007)
and technology as a duality of structure and
agency (Orlikowski, 2000). All of these studies
are concerned with the way in which actors inter-
act with the social and physical features of context
in everyday behaviours that constitute ‘‘practice”.
In new assurance contexts, the actual practice of
determining scope and generating assurance state-
ments remains largely unexplored by researchers in
accounting (Cooper & Robson, 2006).
This operations issue invokes the ideas of best
practice, e?ciency and legitimacy. As Power
(1997, p. 11) puts it, ‘‘the issue is how technical
routines like the sampling of purchase invoices,
the circularization of debtors, the assessment of
inherent risks and so on actually come to be
regarded as constituting ‘reasonable’ practice”.
Audits seek to draw general conclusions from a
limited examination of the domain under investi-
gation; however, despite the potential to use statis-
tically credible foundations for sampling,
economic pressures have led to a drive to obtain
assurance from fewer inputs, and hence into evi-
dence types that are not as easily assessed using
sampling means (Bell, Marrs, Solomon, & Tho-
mas, 1997; Bell et al., 2005; Curtis & Turley,
2007). To this end, two exploratory qualitative
studies by Humphrey and Moizer (1990) and
Fischer (1996) suggest that the design of ?nancial
attest audit work is resolved in situ and that prac-
tice is socially constructed. And in a ?eld study of
the introduction of the business risk audit, Curtis
and Turley (2007) document the di?culties in
translating a new concept developed at the admin-
istrative level into actual audit techniques at the
practitioner level. In a similar vein, the second
research question of this paper is: how do auditors
and clients negotiate and determine audit scope,
testing procedures and communication of work
done in non-traditional engagements?
Due to a number of di?culties in conducting
?eld research in the area, ?eld studies of the gap
between the explosion of programmatic demands
of auditing and the more localized stories of under-
lying operational capability drawn from the ?eld
are rare (Power, 2000). This article addresses calls
for researchers to enter the ?eld to examine practice
in auditing (Cohen, Krishnamoorthy, & Wright,
2002; Gendron, 2000, 2001, 2002) as well as various
other accounting realms (see, for example, Chua,
1995; Hopwood, 1983). It draws on ?eld research
relating to the negotiation and implementation of
a non-traditional engagement between KPMG
and the FT. The data for this research were gath-
ered in July 2004 by one of the authors with the
aid of a research assistant and with the consent of
the client ?rm (the FT). Complete and virtually
unfettered access was granted to the working
papers dating back to the origins of the engage-
ment in 2001. The two primary engagement team
members, Michele Podhy (the partner) and Patrick
C. Free et al. / Accounting, Organizations and Society 34 (2009) 119–140 121
Gaudet (the manager), made themselves available
for both formal and informal interviews about
the events leading up to the engagement and the
process of carrying out the engagements over the
period from 2001 to 2004.
Our interpretations of the ?eld data are broadly
informed by Power’s (1997) observations about
the relationships between auditing, accountability,
and the programmatic aspirations of neo-liberal
governments (Foucault, 1991; Miller & Rose,
1990; Rose, 1992). This study makes three princi-
pal contributions. First, it provides ?eld evidence
of the power of audit as a competitive resource
capable of importing legitimacy to entities across
a variety of domains. Second, the study elaborates
on the notion of the legitimacy produced in the
audit process. Through a detailed study of the
unique relationships created by the engagement,
the notions of a?rmatory and derived legitimacy
are explicated. Thirdly, the study provides evi-
dence of the negotiated nature of audit-scope
determination, testing and reporting done to third
parties in new assurance domains, as the entities
involved attempt to balance external credibility
with the practicalities of cost, time and available
information.
The remainder of the paper is structured as fol-
lows. The next section presents an overview of cur-
rent MBA rankings and the growing, primarily
critical, literature surrounding published business
school rankings. Section 3 discusses the ?eld work
methodology and is followed by an overview of the
context and content of the ‘‘audit” engagement
between KPMG and the FT. Two aspects of the
engagement are then discussed in detail. The ?rst
is the character of the assurance project, including
the way legitimacy accrues to both the business
school community and the FT. The second is the
negotiated and adaptive nature of audit planning
and testing. Section 7 discusses the major ?ndings
of the paper and the ?nal section summarizes the
major themes of the paper and identi?es a number
of avenues for future research.
MBA and business school rankings
[Numerical scoring systems of business
schools are] the equivalent of simply giving
every woman a rating of 1–10 and saying
we do not have to date. Just marry the one
with the best score.
– John Katzman, president of the Princeton
Review, quoted in Thompson (2000).
Industry monitors, rankings and league tables
have mushroomed in the education sector across
a variety of locations in the past twenty years.
In the realm of business school education alone,
there are now dozens of rankings available that
assess program quality, research output, corpo-
rate social responsibility and environmental ori-
entation
4
; each attempting to signal or in?uence
the perceptions of important external audiences
(Mills, Weatherbee, & Colwell, 2006). More
broadly, the rapid growth of rankings and
league tables as performance measures o?ers
considerable opportunities for audit ?rms to
respond to increasing demands for assurance.
Audit services potentially o?er a means of dif-
ferentiating the growing number of rankings
providers.
5
Business school and MBA rankings have
been both castigated and stoutly defended, cast
variously as introducing ‘‘a kind of democracy
in the management development industry”
(Bickersta?e & Ridgers, 2007, p. 65) and
‘‘biased, bogus and ?awed” (Welch, 2002). In
September 2004, Harvard Business School and
Wharton Business School announced that they
would no longer fully cooperate with news
media creating such rankings, refusing to release
current and former graduate students’ contact
information to BusinessWeek for the magazine’s
biennial survey of MBA programs (Yee, 2004).
According to a Harvard spokesperson, ‘‘our
interest is not in restricting information, but in
improving the usefulness and transparency of
that information . . . The media have not paid
particular attention to the rigour of their
4
See, for example,http://www.businessresearch.ca/ andhttp://www.beyondgreypinstripes.org/results/welcome.cfm.
5
The relatively slow uptake is likely related to short-term
resource issues as accounting ?rms respond to the burgeoning
demand for their services in the wake of the Sarbanes-Oxley
legislation.
122 C. Free et al. / Accounting, Organizations and Society 34 (2009) 119–140
method or the real needs of students” (Yee,
2004).
6
Business school rankings rose to prominence in
1988 when the editors of BusinessWeek released
the magazine’s ?rst ever business school rankings.
Until that point, rankings had been limited to less
systematic lists of ‘‘top” business schools compiled
by the deans of schools (often based on research
reputation), recent MBA graduates (based on
schools they had never attended) or CEOs (often
based on schools they had never visited). The stated
goal of BusinessWeek’s ranking was to focus busi-
ness schools on customer satisfaction as measured
by MBA graduates’ perceptions of their education
quality and the perceptions of recruiters who
recruited on those campuses. Since BusinessWeek
released its ranking in 1988, the number and scope
of rankings throughout the world have burgeoned.
While these rankings were initially restricted to the
United States, rankings are now available for every
conceivable corner of the globe. Arguably, the most
in?uential and widely discussed ranking models are
the annual FT ranking and the biennial ranking
produced by BusinessWeek.
7
The FT ranking is
based on two surveys, one sent to business schools
eligible for inclusion and a second to recent gradu-
ates of the schools involved.
8
An independent count
of faculty research at forty primarily academic
research journals is also included in the composite
scores. Table 1 provides a summary of the major
business school rankings currently published.
To date, research relating to MBA and business
school rankings has been focused around two key
themes. The ?rst relates to the veracity and statisti-
cal properties of the rankings. In a study of the
rankings provided by the US News & World Report
and BusinessWeek, Dichev (1999) notes that
changes in ranking, in both magazines, have strong
tendencies to mean revert, meaning that a school
that has moved up in the ranking one year is most
likely to move down in the next ranking issue and
vice versa. He also notes a lack of co-movement
between the two ranking lists that would be
expected if they measured the same latent construct.
Table 1
Leading MBA/Business School Rankings by date of commencement
Publication Survey respondents Geographical coverage Inaugural ranking
BusinessWeek Graduates and corporate recruiters Ranks business schools internationally
every two years
1988
US News & World Report Academics (presidents, provosts, and
deans of admission) for the purpose
of peer assessment
Ranks graduate schools every year,
including Business schools
1990
The Financial Times Business Schools and Graduates Ranks international MBA and EMBA
programs every year
1999
Forbes Magazine Graduates only Biennial ranking based on return on
investment, de?ned as compensation
?ve years after graduation minus tuition
and the forgone salary during school.
2000
The Wall Street Journal Corporate recruiters only Ranks international MBA programs
every year
2001
6
In a statement released online in response to Harvard and
Wharton’s decision, Business Week said its biennial rankings
help students make informed decisions by providing ‘‘objec-
tive”, ‘‘un?ltered information” about each school. The state-
ment concluded ‘‘just as investors today are clamoring for more
transparency on the part of the companies, so should students
expect a similar degree of openness and cooperation from the
very schools that nurture new business leaders”.
7
Downes (2000) employs a regression model (with an R-
squared of 0.48) of the Business Week rankings in 1996 and
1999 to argue that ‘‘changes from one ranking to the next have
a tremendous e?ect on applications – approximately 3% for
each place. Cornell’s increase of 10 places would be predicted to
lead to an increase of 30% and it was actually 35.2%. Darden’s
decrease of 6 places predicts an 18% decline and it was actually
17.9%. So much for believing that our potential students cannot
be swayed by a weekly news rag ...”.
8
Criteria for inclusion in the survey are: a full-time MBA
program that has been running for at least three years and a
return of at least 20 per cent on the alumni questionnaire.
C. Free et al. / Accounting, Organizations and Society 34 (2009) 119–140 123
In a similar way, Dichev (2001) uses a simple regres-
sion model and covariance matrices to provide sup-
port for the claim that the US News annual ranking
of national universities and liberal arts colleges has
a strong tendency to mean revert in the space of
three years, and that approximately 70–80 per cent
of the variation in annual ranking changes is transi-
tory and reversible. Both Downes (2000) and Welch
(2002) note the signi?cant and positive correlation
between business school rankings and business
school size (measured in terms of enrolment).
A second stream of inquiry has presented pre-
scriptive arguments about the implications of shifts
occasionedinbusiness school operationbyrankings
(Mintzberg, 2004; Pfe?er & Fong, 2004). A number
of commentators in this area have pointed to dys-
functional outcomes of business school evaluations.
Zimmerman (2001) argues that a ?xation on ratings
has caused schools to divert resources from invest-
ment inknowledge creation, suchas doctoral educa-
tion and research, to short-term strategies aimed at
improving rankings, such as placement o?ces and
public relation campaigns. The result, Zimmerman
concludes, is a looming critical faculty shortage
and a narrowing of curricula that threaten the pre-
eminence of American management education.
Similarly, DeAngelo, DeAngelo, and Zimmerman
(2005) staunchly criticise the short-term ‘‘quick ?x,
look good” packaging fostered by a preoccupation
with ranking criteria.
9
In a wide-ranging review of
business rankings in Europe, Wedlin (2006) argues
that the FT rankings have been instrumental in
de?ning what constitutes legitimate management
education in Europe as well as re-drawing bound-
aries between Europe and the United States:
The Financial Times rankings was driven by
a desire among European schools to redraw
the boundaries of the ?eld of management
education, largely in response to a perceived
threat that American business schools were
dominating and the American rankings were
setting boundaries that excluded European
schools. The rankings provided a means for
European schools to place themselves in the
group and to redraw the boundaries of the
?eld to include them, thus to create and to
re-create positions and prototypes in the ?eld
(Wedlin, 2006, p. 158).
Despite the widespread view that the FT takes care
of European interests better than other rankings
and rankers, Wedlin (2006) argues that rankings
have led to the construction of models or tem-
plates of management education in Europe that
reinforce an American model of the business
school.
Using qualitative methods, Zell (2001) and Els-
bach and Kramer (1996) investigate the impact of
rankings on business school operations, arguing
that rankings have resulted in a ‘‘customer-
focused” and ‘‘revenue-driven” orientation that
poses a threat to many schools’ perceptions of
highly valued, core identity attributes. This emer-
gent body of literature has yielded important
insights into the operation and impact of rankings
in business schools throughout the globe.
Methodology
In light of the emerging state of the research
domain and the phenomena under study, ?eld
research is well suited to the research questions
being investigated. In this study, data were col-
lected from three main sources: (i) semi-structured
interviews, (ii) auditor–client written and verbal
communications (as documented in the working
papers), audit working papers and formal reports
to the FT and (iii) secondary sources such as news
reports, press releases and academic articles. We
had access to relevant documents from the initial
engagement (i.e., the 2002 engagement) as well as
the 2003 and 2004 engagements. In total, 15 days
of full-time ?eld work were spent by the author
and a research assistant collecting and cataloging
the raw materials for the study. Three formal
semi-structured interviews were conducted with
both Michelle Podhy, the assurance partner on
the engagement, and Patrick Gaudet, a senior man-
ager, at KPMG Canada (with the consent of Della
Bradshaw, the editor of the FT business education
section who is responsible for the rankings project).
9
It is notable that of the top 100 global MBA programs in the
FT 2006 ranking, 96 refer to their position in rankings on their
websites, through press releases, or in brochures (Bradshaw,
2007b, p. 54).
124 C. Free et al. / Accounting, Organizations and Society 34 (2009) 119–140
The formal interviews were transcribed by an
independent transcription expert and later coded
by two of the authors. During the period of ?eld
work, the researchers were allowed complete
access to all working papers from the ?rst year
of the engagement (approximately 1000 pages) as
well as any additional working papers they
requested from subsequent years with the sole con-
dition that the identity of schools and that of their
personnel remain con?dential. In total, we esti-
mate we read over two thousand pages of working
papers, engagement letters, and other correspon-
dence and we were able to copy selected portions
of each provided schools’ names and personnel
were blacked out on copies. Frequent informal dis-
cussions about the meaning of individual working
papers and audit programs, as well as the way
terms were applied and used took place through-
out the ?eld research stage, especially with the
engagement manager Patrick Gaudet and his sta?.
This represents an unprecedented level of access in
a study that is not anonymized.
In parallel to this primary data collection, press
releases, newspaper clippings and academic jour-
nal articles were compiled into a large database
and used to corroborate interpretations. The trian-
gulation made possible by multiple data-collection
methods provides for stronger substantiation of
constructs and theoretical development. The prior
literature and emerging theoretical constructs
served to direct the investigation of the phenome-
non by means of theoretical sampling. Theoretical
sampling is an iterative process whereby speci?c
sampling decisions evolve during the research pro-
cess in light of the emerging conceptual structure
until theoretical saturation is achieved. Emerging
codes in the archival and interview material
included (but were not limited to) the following:
negotiation, adaptation, procedures, assurance
wording and reputation.
The data analysis undertaken adopts three
principal methods proposed by Eisenhardt
(1989) and Anderson (1995). The ?rst method
of analysis involved arranging the interview tran-
scripts and notes into a chronological order. The
transcripts and notes were superimposed on one
another and common and unique perceptions of
events were identi?ed. This process highlighted
areas where further investigation was required.
FT articles and records were used to corroborate
event chronology and prevailing business press
opinion.
The second method of data analysis involved
dissecting and re-organizing the original transcripts
and notes around the key themes identi?ed: the
programmatic (normative) nature of assurance ser-
vices and the technological (operational) elements
of assurance practice. As these themes were only
loosely coupled with time, this process was critical
in identifying the major issues in each theme.
Finally, the insights identi?ed were compared with
the existing literature to identify the extent of
congruence to previous theoretical research.
Whilst primary ?eld data collection took place
over an intensive period of ?fteen days, there has
been an ongoing dialogue between one of the
authors and the KPMG team over this engage-
ment. The team has read numerous versions of
the related teaching case (Davies & Salterio,
2007) and provided their views on the events and
the interpretations of events. Michelle and Patrick
have also been present in the classroom twice when
the case has been discussed, made presentations
based on their involvement with the engagement
and the case study to student groups, and have
read the early drafts of this paper. In sum, a num-
ber of steps were taken to ensure the trustworthi-
ness of this qualitative, naturalistic ?eld study.
Table 2 presents an overview of e?orts directed
at enhancing the trustworthiness of our data anal-
ysis using the schema developed by Basu, Dir-
smith, and Gupta (1999).
Context: the engagement
The FT is one of the leading international busi-
ness newspapers, with worldwide circulation in
excess of 10 million readers daily. It is an operating
division of the Pearson Group (which also com-
prises Pearson Education and Penguin Publish-
ing). Given the FT’s status as ‘‘the world’s most
international business newspaper”, in the late
1990s its editors saw an opportunity to jump into
the growing market of business school rankings
by concentrating on an international comparison
C. Free et al. / Accounting, Organizations and Society 34 (2009) 119–140 125
of MBA programs. On January 25, 1999 the FT
published its ?rst ranking of what it considered
to be the top 50 MBA programs in Europe and
North America. Harvard was ?rst placed overall,
edging out Columbia by 5.3 ranking points (both
rank and total score were provided in the initial
years). Scores were measured on a relative basis,
with the school that scored the highest in a cate-
gory awarded 100 points, and the schools that fol-
lowed being given a proportion of the 100 points
based on how they scored in comparison to the
highest-ranked school in the category.
On the day that the rankings were published,
Della Bradshaw (editor of the FT business educa-
tion section and responsible for the ranking pro-
ject) acknowledged that ‘‘no system of ranking
was perfect”. Indeed, she reported that one of
the loudest criticisms of such rankings had become
that some schools may be ‘‘cheating and lying” in
their data submissions. Certainly, allegations of
data manipulation involving the BusinessWeek
rankings had become well publicized. In 1998,
BusinessWeek determined that students at ?ve
schools, Dartmouth, Duke, Purdue, The Univer-
sity of Texas at Austin and Washington Univer-
sity, had urged their fellow classmates to enhance
their responses in the hope of moving up the rank-
ings (Reingold & Habal, 1998).
10
Shortly after-
wards, cynicism attached to the candour of
submissions prompted the Dean of the Irvine
Graduate School of Management, University of
California, to sign a statement attesting the accu-
racy of the Irvine books, patterned after recently
imposed statements required of American chief
executive o?cers and chief ?nancial o?cers in
their ?lings of corporate documents with the US
Securities and Exchange Commission. She also
challenged her counterparts across the US to pub-
licly attest the honesty of reporting of the numbers
used in business school rankings – to a resounding
silence.
In 2000, the FT again published its rankings
using the same methodology but expanded its cov-
erage to include the top 75 schools worldwide. In
late 2000, as Della was beginning to prepare the
2001 rankings, she became particularly concerned
with the legitimacy of the rankings. Since the FT
had noted the increasing in?uence of business
school rankings, its consulting statisticians decided
to change the approach used to produce the rank-
ings to a normalized scoring system, which they
felt would yield a more accurate overall result.
To add even more rigour to the ranking process,
Della considered whether she should engage a
public accounting ?rm to audit the data that the
schools provided. After making some initial inqui-
ries with ?rms in the Greater London (UK) area,
she was disappointed – not one of the ?rms felt
that they could perform such an engagement. In
late October 2000, Della learnt that an assurance
partner at KPMG Canada, Michelle Podhy, had
been involved with several alternative assurance
engagements including ‘‘auditing” the data that
Table 2
Criteria and procedures for enhancing trustworthiness of data
analysis
Criterion for
trustworthiness
a
Procedures undertaken in this study
Credibility In-depth interviews with representatives
of the auditor (KPMG) with consent of
the client and access to client thoughts
via early drafts of memorandum and
proposed news articles about the ‘‘audit”
Triangulation of observations:
Use of public archival material
O?cial private correspondence
between KPMG and FT and
Working papers from KPMG
Segregation of data gathering and
data interpretation functions
Field debrie?ng with respondents
Field diary and memos
Transferability Thick description of ?eld observations
Reporting of ?eld observations to other
auditors and students
Dependability and
con?rmability
Triangulation of observations
Audit trail from data to ?ndings and
copious use of primary sources
a
These criteria have been adapted from Basu et al. (1999).
10
At Washington University in St. Louis, students distributed
‘‘mock surveys” with BusinessWeek’s logo, indicating how
rankings were calculated and advising students that the
BusinessWeek survey was not the appropriate forum for
criticism. At Texas, several student government leaders circu-
lated a memo reminding students how important it was to keep
the school competitive in the rankings. At Dartmouth, students
were enjoined to rank the school 9 (on a 1–10 scale) if they
hated the school and 10 if they loved it.
126 C. Free et al. / Accounting, Organizations and Society 34 (2009) 119–140
the Queen’s School of Business (in Kingston,
Ontario, Canada) used in marketing its various
programs.
Della e-mailed Michelle on 2 November 2000.
Michelle told Della that she might be able to per-
form such an engagement, but that it was too late
for it to take place for the 2001 rankings. Given
the engagement would have to be proposed,
approved, planned and executed, Michelle stated
that the two-and-a-half month time frame (the
FT published its rankings in mid-January) was
insu?cient to carry out such an engagement. Della
promised to call again next year earlier to arrange
the engagement.
In October 2001, Michelle was again
approached by Della. Della was in the early stages
of analyzing the submitted data from the business
schools for the 2002 rankings. The conversation
started where it had been left o? the year before:
could KPMG conduct an ‘‘audit” or some other
form of assurance engagement over the business
schools’ responses to the FT survey? Before devel-
oping an engagement proposal, Michelle received
more information from Della about the rankings
and the measures to be used for the 2002 rankings
(see Table 3). Della wanted KPMG to ‘‘audit” all
the business school-provided information used to
calculate the relative scores. Table 3 lists the items
used to calculate the rankings in 2002. These crite-
ria have been virtually unchanged since 2002.
Michelle called upon a senior KPMG manager,
Patrick Gaudet, to help her consider the engage-
ment proposal. Under the terms of the proposed
engagement, the entity under ‘‘audit” would be
the business school submitting the data to the
FT, even though the client who engaged KPMG
was the FT. As participation in the FT survey is
voluntary, neither the FT nor KPMG has a con-
tract with the various business schools.
11
Accord-
ingly, Michelle and Patrick were faced with many
initial challenges. These included determining the
set of professional standards that would govern
the engagement, the appropriate procedures to
be employed and the logistical challenge of visiting
25–35 schools per year in a relatively tight time
frame (i.e. the months of November and
December).
Michelle and Patrick concluded that Section
9100 of the CICA Assurance Handbook entitled
‘‘Reports on the results of applying speci?ed audit-
ing procedures to ?nancial information other than
?nancial statements” would be the most appropri-
ate framework.
12
This section allows the auditor to
use only very speci?c audit procedures over speci?c
data, with a speci?ed, standard sample size – thus
meeting the ‘‘audit” objective in a way that satis?ed
the requirements of both KPMG and the FT.
The logistical challenge of the engagement was
heightened by the fact that Della wanted the same
engagement team to visit every school in order to
provide consistency of application of the ‘‘audit”
procedures and provide for continuity over time.
In 2002, ten schools were selected for an initial
‘‘audit”. Because of geographical proximity, and
their history with KPMG (KPMG has audited
the University’s ?nancial statements for many
years), Queen’s University was chosen to be the
?rst school for the ‘‘audit.” The remaining nine
schools were selected on the basis of proximity to
Kingston, Ontario. In defending this approach,
Della commented that selecting schools randomly
from the list of participating business schools
would have ‘‘laid those schools open to suspicions
that they had exaggerated their data”. That is,
there may have been a perception that selection
was in fact not random but rather on the basis
11
In subsequent years, a term and condition for a business
school to participate in the survey was to allow KPMG access
to the school in order to ‘‘audit” the data they submitted to the
FT.
12
CICA HB Section 9100 is similar to the International
Federation of Accountants (IFAC) International Audit and
Assurance Standards Board Related Services Standard 4400
and the US American Institute of Certi?ed Publication
Accountants Attestation Standard AT 201. CICA HB Section
5025, which deals with attest or audit level assurance engage-
ments in general, was also considered by Michelle and Patrick.
However, they concluded as it would be very di?cult to provide
the requisite audit level assurance over the data that underlie
the FT ranking tables given the time constraints and nature of
the data. Thus, they concluded that a speci?ed procedure
engagement was the most they could o?er. Further, KPMG had
some experience using a similar type of exception reporting at
both the Queen’s School of Business and the Province of
Ontario’s Department of Health (in reporting on various
hospitals’ Y2K compliance).
C. Free et al. / Accounting, Organizations and Society 34 (2009) 119–140 127
of suspicion. She further commented that ‘‘the
geographical approach had two further advanta-
ges: it made the audits easier to complete and
meant that schools throughout the top 100 table
were included” (Bradshaw, 2002).
During the 2002, 2003, and 2004 engagements,
many immaterial errors were found and reported
to Della in the very extensive and detailed reports
that were provided by KPMG on each and every
school visited (an average report ran 12–15 pages
per school, reporting each procedure performed
and its results, along with any KPMG team obser-
vations). She required every school where an
exception was found, however minor, to correct
their submissions. Interestingly, most of the errors
that the engagement team found were more con-
Table 3
2002 FT Ranking Measures
128 C. Free et al. / Accounting, Organizations and Society 34 (2009) 119–140
servative interpretations of the questions resulting
in the schools under-reporting on relevant criteria.
One school, however, when it learned it had been
chosen for ‘‘audit” withdrew from the survey with
no explanation.
13
Characterizing the new assurance project
An inherent con?ict of interest in the traditional
audit of ?nancial statements is that the entity
under audit pays the public accounting ?rm for
the audit on the entity’s ?nancial statements for
the bene?t of third parties (Goldman & Barlev,
1974; Moore, Tetlock, Tanlu, & Bazerman,
2006). This issue has exercised researchers (e.g.
Chaney & Philipick, 2002; Gunz & McCutcheon,
1991) and regulators throughout the globe (in the
United States, the Sarbanes-Oxley Act of 2002 rep-
resents the latest legislative approach to managing
this inherent con?ict). To this end, there is a small
but growing academic literature suggesting alter-
native mechanisms for appointing (Mayhew &
Pike, 2004) or compensating (Ronen, 2002) the
auditor to avoid this inherent con?ict of interest.
Proposals for such alternative compensation
arrangements include having third parties, such
as insurance companies, hire auditors to carry
out the audit of ?nancial statements as a term
and condition of underwriting ?nancial statement
insurance for a public company (Bhattacharjee,
Moreno, & Yardley, 2005).
Our setting mirrors in some ways these alterna-
tive appointment and compensation proposals.
The FT pays for the ‘‘audit” of the data it
requested from the entity under ‘‘audit”, the busi-
ness school, but the only contractual obligation
the auditor has is to the FT.
14
In other words,
the inherent con?ict of interest found in the ?nan-
cial audit is not present. Fig. 1 characterizes the
features of the FT Rankings engagement and con-
trasts it to a conventional ?nancial statement
audit.
The nature of these engagement relationships,
in contrast to the traditional attest audit, provides
a number of insights into the production of legiti-
macy through the operation of ‘‘auditing”.
The production of legitimacy
The incursion of assurance services into non-
traditional markets is consistent with a number
of observed shifts in programmatic emphasis at
the political and the social levels. At the political
level, numerous scholars have documented a grow-
ing commitment, within advanced liberal democ-
racies, to the programmatic commitments of neo-
liberalism (Rose, 1992; Rose & Miller, 1992).
Neo-liberalism, as the label given to the dominant
governmental rationality of contemporary
advanced liberal democracies, is predicated on
the belief that the proper role of the state is to cre-
ate a space of ‘‘regulated freedom” (Rose & Miller,
1992, p. 158), within which free and self-determin-
ing citizens assume active responsibility for their
choices and their life outcomes. Neo-liberalism
also entails a related social commitment to
enhanced accountability by and for organizations
and individuals.
As Power (1997) has observed, the appeal of
auditing within the broader context of neo-liber-
alism derives from its capacity to instantiate
13
The withdrawal of a school from the rankings as a result of
being selected for audit ?ows naturally from the signalling
hypothesis (Datar, Feltham, & Hughes 1991; Melumad &
Thoman 1990; Spence, 1973). Under signalling, in order to
signal credibly the signaller (i.e. the school) must incur costs
and if they do not incur them then they are disciplined by either
being revealed as a ‘‘cheater” by the audit or forced to withdraw
from the ‘‘market” (i.e. the rankings). All schools that
participate in the rankings must commit to incur the costs of
both reporting the data and maintaining data in a form
susceptible to audit.
14
The tripartite nature of the relations between the Univer-
sities, KPMG and the FT presented numerous challenges
relating to privacy, policies and accountability which emerged
and re-emerged throughout the course of the engagement. In
the working papers dated January 16, 2003, Patrick Gaudet
noted ‘‘in some jurisdictions, due to laws or internal policies
concerning privacy of information, it was di?cult to obtain
supporting documentation. In such instances this was docu-
mented in the ?ndings as well where appropriate alternative
audit procedures were performed . . . Some of the business
schools were also unsure of what types of student activities
should be included as overseas projects or study tours. KPMG
reviewed the reasonableness of the activities included and if
questions arose the FT was contacted for further clari?cation”.
C. Free et al. / Accounting, Organizations and Society 34 (2009) 119–140 129
these programmatic political aims by construct-
ing new and novel relations of accountability,
and by inscribing these within systems that per-
mit the discharge of accountability through
third-party audits. Indeed, Power (1997, p. 134)
considers the capacity of audit to ‘‘operationalize
and realize accountability” to be the ‘‘core pro-
grammatic value” of auditing. Audit accom-
plishes this by embedding relevant aspects of
the individual’s or the organization’s actions into
formalized systems that can then be validated by
internal or external audit (Power, 1997, p. 53).
Once individual or organizational action has
been captured in a system that is amenable to
audit, then audit becomes a means – perhaps
the primary means – of lending legitimacy to
the individual or the organization. As Power
argues, this legitimacy accrues to the audited
entity without regard to the actual e?cacy of
the audit to instantiate the speci?c programmatic
aims for which it was undertaken.
These legitimacy arguments help explain why
the most common business school response, at
least publicly, was enthusiasm for the prospect of
an assurance mechanism that would ensure that
competing schools report truthfully.
15
Deszo
Horvath, Dean of the Schulich School (York Uni-
versity, Canada), claimed: ‘‘We at the Schulich
School of Business welcomed the introduction of
the audit as a regular part of the annual FT survey.
I believe the use of auditing will be good for all of
the schools involved in the annual ranking” (Brad-
shaw, 2002). In her discussion of the ‘‘audit” pro-
ject published with the release of the 2002
rankings, Della further commented:
Other schools said the audit helped them
decide how better to collect the data in
future. Some non-participating schools lob-
bied to be included in this year’s audit while
others suggested they might pay for their
own audit to be carried out with an approved
auditor on an annual basis (Bradshaw, 2002).
These self-design processes aimed at making busi-
ness schools more readily amenable to audit proce-
Financial statements
and related
disclosures
Written opinion in
accordance with
GAAS
Conformity
of Financial
Statements
With GAAP
Accountable Party:
Firm Management
Auditor
User
Audit fee
Audit
Accountable Party:
Firm Management
Auditor
User
procedures
Panel A: Conventional Attest Audit Panel B: Financial Times Rankings Assurance
Engagement
Raw data
Written
conclusion (based on
negotiated criteria)
Veracity of
Data Submitted
Accountable Party:
Business Schools
Assurer: KPMG
Financial Times
Assurance fee
Audit procedures
Users
Summary data in the
form of rankings table
Specification of
data sought
Fig. 1. The relationship between parties in a conventional ?nancial statement audit compared with the Financial Times Rankings
assurance engagement.
15
Privately there appeared to be some initial reluctance among
the inaugural auditees to participate in the audit until there was
assurance that peer schools had agreed to cooperate. Hence, as
noted in footnote 9, the FT made inclusion in the rankings from
2003 onwards conditional on schools agreeing to be audited.
130 C. Free et al. / Accounting, Organizations and Society 34 (2009) 119–140
dures are a rational response in the face of an audit
and represent a key element in the process of aud-
itability (see Power, 1996).
Apparently then, business schools view the
‘‘audit” as bestowing or enforcing legitimacy on
the actions of their competitors, thus ensuring ‘‘a
level playing ?eld” across all participating schools.
With respect to the ‘‘audit” of their own reporting,
however, business schools seem only to perceive
what might be termed an ‘‘a?rmatory legitimacy”;
that is, a?rmation that their reputation remains
untarnished in the face of a reputational risk that
would be realized if the KPMG ‘‘audit” were to
reveal exceptions in their reported data.
16
Com-
menting on this, Michelle explains that ‘‘[the
schools] are very careful. And I think that, again,
on their part, there is a reputational risk thing
the same as we worry about at KPMG. They are
just not willing to go there. And the schools who
do not want us in [have] dropped out. . .” (Inter-
view, 14 July 2004). Presumably for similar rea-
sons, schools that were noti?ed of the prospect
of an upcoming ‘‘audit” engaged in ‘‘self-audit”
or double checking of their data. When the
KPMG team arrived, discrepancies were typically
tabulated and presented to the team for veri?ca-
tion. And if the reception given the team can be
taken as an indication, it appears that participat-
ing business schools – even the most elite among
them – take the FT ‘‘audit” very seriously. As
Michelle describes it, ‘‘[Patrick] has been every-
where. And they treat him like, they get out their
kid gloves and whatever Patrick wants, he gets,
because they are trying to be nice” (Interview, 14
July 2004).
Thus, our evidence suggests that participating
schools attend seriously to the requirements of
the ‘‘audit”, and that they do so because they fear
a loss of legitimacy if the ‘‘audit” should reveal
exceptions. This evidence is consistent with the
proposition that audits confer reputational bene-
?ts (e.g., legitimacy, safety, e?ciency, etc.) upon
audited organizations. What makes this observa-
tion particularly surprising in this context, how-
ever, is that, under the terms of the engagement,
only Della Bradshaw at the FT is apprised of
any exceptions in the audited data. As Della, in
turn, requires only that the school correct the data
prior to the publication of the rankings, the repu-
tational risk apparently perceived by the partici-
pating schools seems unmerited; the results of the
audit procedures are never made public. Appar-
ently, the concept of legitimacy is so thoroughly
enmeshed in the programmatic appeal of audit
that it cannot readily be dismissed from consider-
ation, even in those instances where it may not
be at issue.
Our ?eld study also suggests that the legitima-
tion bene?ts bestowed by the audit can extend
beyond those organizations whose accounts are
audited. As previously discussed, the FT engage-
ment di?ers from a conventional attest audit in
that the accountable party (i.e., the auditee) does
not purchase the audit services on their own
behalf. As conventionally conceived, audits confer
reputational bene?ts (‘‘legitimacy”) on the organi-
zations whose accounts are audited, thus suggest-
ing that legitimacy is one of the ‘‘products” that
the organization purchases with its audit dollars
(see Fig. 1, Panel A). In the FT case, however,
the purchasing organization is distinct from the
‘‘audited” organization, thus raising questions
about what the FT is buying with the KPMG
‘‘audit.”
On the basis of our ?eld work, it appears that an
aura of legitimacy accrues from the ‘‘audit”, not
just to the auditees, but to the FT itself. Della has
publicly pointed to the value of having the rankings
audited (Bradshaw, 2002). By harnessing auditing
as a competitive resource, the FT encourages read-
ers to view their rankings as possessed of greater
integrity than rankings derived from unaudited
data. But such legitimacy seems, at best, a derived
legitimacy, one that takes its meaning from the
primary legitimacy bestowed by the ‘‘audit” on
the participating schools (see Fig. 1, Panel B).
16
That is, while the participating schools evidently embrace
the belief that the ‘‘audit” bestows legitimacy on the data
supplied, there seems to be an asymmetry in how this legitimacy
is perceived to accrue to themselves vis-a`-vis their competitor
schools. With respect to their competitors, ‘‘audit” is seen to
actively grant legitimacy (where previously legitimacy was open
to question) by ensuring the accuracy of the reported data.
With respect to themselves, however, the ‘‘audit” is conceived
as an a?rmatory mechanism; a ‘‘clean audit” leaves intact a
pre-existing legitimacy, while the discovery and communication
of reporting errors pose a threat to this legitimacy.
C. Free et al. / Accounting, Organizations and Society 34 (2009) 119–140 131
This reveals that the power of audit and its
basic acceptance by society across a range of di?er-
ent contexts may be even greater than might have
been previously supposed. The mere mention of
audit apparently confers legitimacy, not only on
the audited entities, but also secondarily on the
entity that employs the data from the audited enti-
ties (that is, a legitimacy that is derived from the
audit of other distinct entities). This occurs despite
the fact that the FT compiles its rankings using a
combination of audited business school data,
unaudited data from the same source and unau-
dited data that the FT collects itself, and despite
the fact that the unaudited data far outweighs
the audited data in terms of relative contribution
to the ranking (63% of the ranking methodology
is not subject to audit procedures, see Table 3).
It also occurs despite the fact that the most impor-
tant elements of the rankings, the selection of data
to be included and the algorithm that combines
these data into a single rank, are neither the
responsibility of, nor audited by, KPMG.
In sum, audit appears to have a robust capabil-
ity to imbue legitimacy, not just to entities whose
accounts are audited, but also – at least in some
circumstances – to those who make use of these
audited data. Our study thus suggests that, as the
audited entities, business schools are su?ciently
concerned about the power of ‘‘audit” to a?ect
their reputations that they attend seriously to it,
even where there is limited evident basis for a rep-
utational e?ect. And as a user of the audited data,
the FT enjoys a derived legitimacy that imbues its
rankings with an aura of integrity that extends well
beyond the limited reach of the ‘‘audit”. This
robust capability is even more remarkable when
one considers that it was a matter of some dispute
whether this particular ‘‘audit” could even be per-
formed within the institutional framework of
auditing. We now turn to the construction of the
‘‘audit” engagement itself.
Consensus building in audit planning, testing and
reporting
A striking feature of the emerging engagement
was consensus building and adaptation across
each major element in the construction of the
‘‘audit”: planning the audit scope, designing the
audit procedures and crafting of the public report
of the ‘‘audit”. This section treats these three for-
mative areas – audit planning, testing and report-
ing – in turn.
Audit scope deals with the extent of the subject
matter that the auditor will examine. A review of
KPMG’s correspondence with the FT reveals there
was considerable negotiation in planning the
‘‘audit” and attempting to resolve the tension
between external credibility and pragmatism. The
following exchanges typify this process:
For our purposes, there are clearly some
areas where there is little controversy. For
example, it is quite hard to get the number
of women faculty wrong! But when we talk
about the percentage of international stu-
dents and faculty I feel some schools may
be miscalculating. There may be a case,
therefore, for auditing just these two catego-
ries for a number of schools . . .
– E-mail from Della Bradshaw to Michelle
Podhy (2 November 2000).
I don’t believe we will be able to audit the
information related to the students employed
three months after graduation. This will be
di?cult to audit as each school may have
varying ways of measuring this and they
may not be susceptible to audit procedures
. . .
– E-mail from Patrick Gaudet to Della Brad-
shaw (11 November 2001).
The limited scope of the ‘‘audit” (salary related
data, for example, was not examined even though
it comprises the key component of the FT rankings
constituting 40% of ranking scores
17
) is particu-
larly noteworthy given that the ranking process it-
self has been criticized for failing to take into
17
Indeed these data are collected directly by the FT so, in
principle, it should be more susceptible to audit procedures
than data collected from business schools. Furthermore, it is
not raw salary data that are employed but salary data that are
transformed by a process of ‘‘adjustment for salary variation
between industry sectors”; a process that is not disclosed by the
FT making this heavily weighted measure rather opaque.
132 C. Free et al. / Accounting, Organizations and Society 34 (2009) 119–140
account key, albeit more subjective, measures of
student experience and curriculum. For example,
as Thompson (2000) puts it, ‘‘do the big shot pro-
fessors actually teach? . . . Are there unknown pro-
fessors who are better teachers? Or is it the
graduate students who teach? What is the intellec-
tual atmosphere like on campus?”
In addition to negotiating audit scope, audit
testing procedures were also the outcome of nego-
tiation both internally and externally with the FT.
Working papers and internal KPMG team corre-
spondence pointed to the emergent nature of the
design of testing procedures.
I may need access to this type of information
[HR ?les and private student ?les], however,
there may be alternate sources documenta-
tion available for testing in the business
school. I completed one of the business
school audits yesterday and I was able to ?nd
source documentation within the business
school. For example:
– I may be able to test whether a faculty
member has a PhD. or not by reviewing the
published faculty biographies in the business
schools academic calendar or another
publication.
– I may be able to test whether a faculty or
advisory board member is an international
member from e-mails con?rming this directly
from the individual faculty and advisory
board members.
– I may be able to test whether a student is
enrolled in the MBA program, their gender
and their nationality from information con-
tained on student data forms in the school
of business or a listing from the University
registrar’s department.
– KPMG working paper (29 November
2001).
Our audit tasks are about obtaining support-
ing documentation and we sort of agreed
with Della what that supporting documenta-
tion can be.
– Interview with Patrick Gaudet (14 July
2004).
Additionally, sampling in terms of both sample
size and sampling technique (selection of schools
on the basis of geography – a form of ‘‘block”
sampling long discouraged in auditing textbooks
(e.g. Chapter 16 of Knechel et al., 2007) for most
sampling applications) appeared to be set more
symbolically to appease the client rather than the
outcome of an assessment of risk taking into ac-
count the assertions being tested.
We agreed to a speci?c sample size that was
consistent for each of the schools. Because
one of the things that was important sort of
from a logistical perspective was that each of
these audits no matter how large the MBA
program was, that we would be able to com-
plete an audit in a day. So if you have a school
like the Wharton School that has 400 stu-
dents, we decided that the business editor just
wanted a sample size big enough that she was
comfortable that the information was accu-
rate. So we weren’t taking a look at sample
size saying OK we’re going to look at 20% of
the students, it was just a ?xed number. We’re
going to look at this number of students on a
randombasis at all schools . . . Soa lot of times
that’s ten or ?fteen [students].
– Interview with Patrick Gaudet (21 July
2004).
When asked about the one-day ?eld work
limit, Michelle Podhy replied ‘‘that’s just how we
priced and sort of what worked” in the context
of the time available (November and December),
the sta?ng demands of the FT (the requirement
that the same KPMG team performs the ‘‘audit”
each year), and budget issues within the FT. This
provides evidence that resource availability
constrains audit practices even in this novel
domain.
As audit ?les were created, checklists created
and ticked, and performance measured and moni-
tored, ‘‘susceptibility to audit procedures” became
the primary driver of the work’s scope. From
KPMG’s viewpoint, the preferred focus of proce-
dures was largely ‘‘factual” in nature, with the
requirement that the schools have credible data
from which to draw their questionnaire responses.
The following comment is typical of this auditor
preference for a more limited scope and easily ver-
i?ed procedures:
C. Free et al. / Accounting, Organizations and Society 34 (2009) 119–140 133
Table 4
Examples of Speci?ed Audit Procedures Relating to the Financial Times, 2004 MBA Survey, November 11, 2003
Area Purpose Procedure
Faculty Veri?cation of the percentage
of faculty who have a
doctoral degree
KPMG will agree the total number of faculty with doctoral degrees from the
listing to the amount submitted to the Financial Times, 2004 MBA Survey
KPMG will recalculate the percentage of faculty who have a doctoral degree
based on the listing obtained from the business school
KPMG will trace ?ve faculty members who have doctoral degrees
to supporting documentation in the academic institution to verify
that they are female faculty
Board
members
Veri?cation of the percentage
of international Board
members
KPMG will agree the total number of international Board
members from the listing to the amount submitted to the
Financial Times, 2004 MBA Survey
KPMG will recalculate the percentage of international Board
members based on the listing obtained from the business school
KPMG will trace ?ve International Board members from the
listing obtained from the business school to supporting
documentation in the academic institution to verify that they are
international Board members
MBA
students
Veri?cation of the percentage
of female students
KPMG will agree the total number of female students from the listing to
the amount submitted to the Financial Times, 2004 MBA Survey
KPMG will recalculate the percentage of female students based
on the listing obtained from the business school
KPMG will trace ?ve female students of the MBA programme to
supporting documentation in the academic institution to verify
that they are female students
Graduate
employment
Veri?cation of the percentage
of graduates who had accepted
a job o?er or were in
employment three months
after graduation
KPMG will obtain a listing of graduates with an accepted job o?er
or were employed three months after graduation and agree the
total to the total number of employed graduates submitted to the
Financial Times, 2004 MBA Survey
KPMG will recalculate the percentage of graduates with an
accepted job o?er or employed three months after graduation
based on the listing obtained from the business school
KPMG will request con?rmations from ten graduates for each
business school from the listing requesting con?rmation of their
employment status from selected business schools on a trial basis.
KPMG will select two business schools in the United Kingdom
and two business schools in the United States
Language
requirements
Veri?cation of the number of
spoken languages required for
MBA programme students
upon graduation
KPMG will obtain the language requirements to graduate for MBA
programme students from the academic institution and agree them
to the number submitted to the Financial Times, 2004 MBA Survey
International
studies
Veri?cation of the amount of
international exchanges
completed by students of the
last graduating class of the
MBA programme
KPMG will obtain a listing of MBA programme students from
the last graduating class who have completed exchanges with
international business schools and agree the number of students
to the number submitted to the Financial Times, 2004 MBA Survey
KPMG will recalculate the percentage of students who completed
international exchanges
KPMG will agree two international exchanges to supporting
documentation in the academic institution
Doctoral
Graduates
Veri?cation of the number
of doctoral graduates of the
business school
KPMG will trace two doctoral graduates of the business school
to supporting documentation in the academic institution
134 C. Free et al. / Accounting, Organizations and Society 34 (2009) 119–140
. . . if it’s student salaries then that’s not clear
and we . . . wanted to be able to audit it and
not just take the word of management or
that particular client as you know. We
wanted wherever we could to get third party
veri?cation . . . there is certainly a number of
things there that are more soft and we knew
right o? the bat we couldn’t do anything with
that throughout the process.
– Interview with Michelle Podhy (14 July
2004).
To date KPMG’s objections to the feasibility of
auditing certain measures have not led to any
changes in the measures used in the FT ranking
methodology.
18
However, the ‘‘audit” has necessi-
tated the tightening of some of the de?nitions used
by the FT in constructing the rankings:
A good example of that is international sta-
tus. At a lot of the American schools they
are considering students with a green card
as American students whereas from sort of
the perspective of everyone else, they would
be an international student. When I was in
England last fall, one of the common things
was if they had a French student because of
the EU they would consider a French student
as a domestic student whereas the FT wanted
to treat that student as a foreign student
because they are not from that country. So
that was one of the things where they started
creating additional sort of de?nitions and
tightening up what the criteria is with the
school as we are coming across these anom-
alies as we’re visiting the schools.
– Interview with Patrick Gaudet (14 July
2004).
Table 4 outlines typical examples of the proce-
dures that were the outcome of this negotiation
of the nature and extent of audit scope. As can
be seen, the language of ?nancial audits – ‘‘docu-
mentation”, ‘‘veri?cation”, ‘‘con?rmations”,
‘‘trace”, ‘‘agree” – permeated work notes and the
entire engagement.
Indeed, the use of technically precise audit ter-
minology in the reporting about the engagement
was the source of some debate between Della
and Patrick and Michelle. A draft article describ-
ing the nature of the auditing process for the
2002 was edited from ‘‘. . .it was decided to use
KPMG in Kingston to audit the pilot 10 schools”
to ‘‘. . .it was decided to use KPMG in Kingston to
perform speci?ed audit procedures at the pilot 10
schools (emphasis in original)”. Seven changes
were made in total, with the majority along similar
lines, replacing the terms ‘‘revise their submis-
sions” with ‘‘review their submissions” and
‘‘audit” to ‘‘speci?ed audit procedures”. These
changes re?ect the tension between Della seeking
a broader statement and Michelle seeking a more
narrowly de?ned statement consistent with the
professional standards (CICA Handbook Section
9100) she had proposed as governing the
engagement.
. . . we got her to change audit to de?ning
what it was more than just using audit.
Because it’s not an audit. I mean we have
to be very careful;. This is not an audit in
the sense of the assurance world and what
an audit stands for. Because we have not
audited the data at Harvard and we can’t
give an opinion. We have not given an opin-
ion . . . And that’s lost on many, as you
know.
– Interview with Michelle Podhy (14 July
2004).
It seems that the opaqueness of audit practice to
users – the ‘‘audit blackbox” – means that quali?-
cations and caveats may have limited resonance
with users, thus contributing to the inherent expec-
tation gap.
19
In the actual disclosure on the 2004 FT survey,
Michelle (the KPMG audit partner) sought to com-
municate the following three points: (1) an audit
was not conducted over the input data and no opin-
ion was expressed; (2) that KPMG performed
procedures on behalf of the FT and reported
the results of those procedures to the FT, and (3)
that KPMG does not take responsibility for the
18
It is an open question as to whether over time audits will
have a normalizing and standardizing e?ect on ranking systems
themselves.
19
We thank an anonymous reviewer for this insight.
C. Free et al. / Accounting, Organizations and Society 34 (2009) 119–140 135
measures used in the rankings themselves nor does
KPMG participate in the actual aggregation of
the data into ranks. Communicating the precise
details of this message to users represents a daunt-
ing challenge and again entailed a process of nego-
tiation between the FT and KPMG, especially in
the initial years of the engagement. This sort of
report content negotiation harkens to the negotia-
tions that are encountered by auditors of ?nancial
statements when they deal with problematic ?nan-
cial accounting and disclosure issues that could lead
to the auditor modifying the wording of the audit
report if the issues are not resolved appropriately
(Gibbins, McCracken, & Salterio, 2007).
The published rankings in 2004 included an addi-
tional column entitled ‘‘Audit Year
*
”. This column
listed the most recent year that KPMG sta? visited
the school to perform the speci?ed audit proce-
dures. A footnote, designated by the ‘‘
*
” at the bot-
tom of the ranking table stated the following:
*
KPMG reported on the results of obtaining
evidence and applying speci?ed audit proce-
dures relating to selected data for the Finan-
cial Times 2004 MBA survey ranking for
selected business schools. Inquiries into the
process can be made be contacting Michelle
Podhy and Patrick Gaudet of KPMG by e-
mail at [email protected]. The speci?ed
audit procedures were carried out during
November and December 2003. The audit
date denotes the survey for which speci?ed
audit procedures were conducted.
The technical references to ‘‘evidence” and ‘‘speci-
?ed audit procedures” corresponds with Van Maa-
nen and Pentland’s (1994, p. 54) conclusion that
‘‘audit reports are a symbol of legitimacy” which
do not so much communicate information as ‘‘give
o?” legitimacy by virtue of a rhetoric of ‘‘neutral-
ity, objectivity, dispassion, expertise.” The di?-
culty in communicating the nature of the
engagement was underlined by Michelle Podhy:
I’m getting e-mails daily still . . . [on] how do
you come up with [the rankings criteria]?
They think that we are responsible for the
ranking because our name is in the paper
to say, and what we have done is star the
schools that are selected for audit and then
also there is our footnote as to which parts
of the data we have applied the speci?ed
audit procedures to. We are very careful to
remind people we have not done an audit,
of course, because we cannot give an opinion
on any of this. The client just said, you
know, tell me how many females there are,
they said there’s 30, go and see if there’s
30. It just speci?ed audit procedure and we
just report any exceptions if there are any.
– Interview with Michelle Podhy (14 July
2004).
Discussion and implications
The ?rst research question of the paper was to
explore the antecedents and character of an assur-
ance project in a ‘‘new” unregulated context. The
?eld study presented here suggests that the sym-
bolic capital of the notion of the audit is a central
element of the demand for audit and assurance ser-
vices in new marketplaces. Power (1997) argues
that the fashion of auditing signi?es a distinctive,
if unevenly distributed, phase in the development
of advanced economic systems as they grapple
with the production of new or enhanced risks
and uncertainties, the erosion of social trust, ?scal
crisis and the need for control. He groups these
forces under the broad rubric ‘‘audit society”. In
this environment, it is likely that ‘‘audits” will
increasingly be seen as a competitive or discursive
resource able to distinguish ‘‘reliable” providers of
information from others.
It is clear from the ?eld study that a major dri-
ver in the FT decision to engage KPMG was the
suspicion attached to MBA rankings (akin to the
fraud suspicion inherent in ?nancial statement
audits) and a desire to distinguish the FT’s rank-
ings and import the legitimacy associated with tra-
ditional attest audits. In this sense, the
introduction of an ‘‘audit” can be seen as a form
of brand management by the FT in the competi-
tive game between providers of rankings. This is
not to say, however, that Della viewed the ‘‘audit”
as a purely symbolic exercise undertaken only to
create an illusion of legitimacy; rather, Della’s
decision to engage KPMG needs to be understood
136 C. Free et al. / Accounting, Organizations and Society 34 (2009) 119–140
as emerging out of a broader social context in
which suspicion of dishonest reporting was wide-
spread, and in which ‘‘audit” is accorded an
almost mystical power to discern the truth. Given
this context, that Della turned to ‘‘audit” to pro-
vide the source of the FT’s di?erentiating legiti-
macy speaks both to her own faith in the
e?cacy of ‘‘audit” and to the social legitimacy
accorded audit practice. Indeed, it is the social
belief in the potency of audit that e?ectively
converts it into a competitive or discursive
resource to be deployed by information providers.
In introducing the 2007 rankings, Della claimed
‘‘it has become accepted that audited data are
the way forward – they enable applicants and
recruiters to make valid comparisons” (Bradshaw,
2007a). As Power (2000, p. 117) summarizes,
‘‘being audited per se is a badge of legitimacy”.
Legitimacy has long been recognized as a cen-
tral concept in auditing. Organizations perceived
to be legitimate ?nd it easier to attract economic
resources (in our context, student applications
and fees that go with their enrolment) and gain
the support required for successful operation.
Although the concept of legitimacy has been
invoked in a range of auditing studies and some
audit researchers have even referred to legitimacy
as a ‘‘theory” (e.g. Mobus, 2005; Ogden & Clarke,
2005), there has been little delineation of the con-
cept. Unique features of the FT Rankings engage-
ment provide insights into the nature of legitimacy
produced by the operation of auditing and the
power of auditing. In particular, the nature of
the relationships created allows us to distinguish
between two di?erent types of legitimacy: a?rma-
tory legitimacy and derived legitimacy. Schools
participating in the ‘‘audit” appeared to see the
audit process as bestowing legitimacy on the dubi-
table data provided by competitor schools while at
the same time, with respect to their own submis-
sion, simply a?rming their pre-existing legitimacy.
That is, from each school’s perspective, the process
simply endowed them with an a?rmatory legiti-
macy (which only a publicly communicated ‘‘fail-
ed” audit could jeopardize). In contrast, the
party paying for the ‘‘audit”, the FT, enjoyed a
legitimacy derived from the application of a set
of limited audit procedures to a set of third parties,
which re?ects the power of the term ‘‘audit” in
modern society. Future research could usefully
seek to further delineate di?erent types of
legitimacy.
The second research question was to investigate
the way that auditors and clients negotiate and
determine the scope, procedures and communica-
tions about the ‘‘audit” when applied to non-tradi-
tional entities and relationships. An analysis of
working papers and correspondence between the
FT and KPMG suggests that the conduct of the
‘‘audit”, both in terms of audit planning and audit
testing, was the outcome of an extensive negotia-
tion and set of compromises between external
credibility and pragmatic imperatives – rather than
a strict conformance with the blueprints, idealiza-
tions, de?nitions and concepts often publicly
invoked by audit practitioners. In terms of engage-
ment scope, the design of audit procedures, and
communication of work done, issues were resolved
via negotiation between the FT and KPMG as well
as in situ by Patrick Gaudet and Michelle Podhy
within KPMG. Auditability must be constructed
as a series of legitimate yet pragmatic techniques.
The assurance project then emerges as a much
more negotiated and interactive practice than gen-
erally presented. While auditing in general has to
be made to look as if it works incontrovertibly,
there was evidence of both controversy and prior-
itization of that which can be measured in audit-
able terms over that which is perhaps more
ambiguous (and potentially useful to aspiring stu-
dents). What emerges then from the ?eld data is a
view of auditing as a collection of negotiated and
highly adapted pragmatic routines which may
add credibility to the rankings, but in a way that
cannot be easily communicated to users. Audit
procedures, like the statistical underpinnings of
the rankings themselves, are opaque enough that
very few users outside of the FT can ?gure out
exactly how they work, yet clear enough to convey
legitimacy.
20
20
Power (2007) notes a similar tendency within the profession
to reduce the history of audit practice to a simplistic account of
progressive improvement in audit technique, thus providing a
‘‘schematic narrative” that enables practitioners’ self-under-
standing of the past and the future of audit practice.
C. Free et al. / Accounting, Organizations and Society 34 (2009) 119–140 137
Conclusion and future research
Given the increasing prevalence of new realms
for auditing in the UK, Australia, New Zealand,
Europe and North America, there is a need for
more detailed empirical and comparative studies
on the extension of audit-type practices into new
arenas, including their e?ects and consequences
(Power, 2003, p. 387). New objects and practices
are continually being subjected to audit. In many
cases, this is a non-trivial development: as in the
case of business school rankings, the very process
of performing the ‘‘audit” itself often renders a
construction of performance as much as a mea-
surement of it.
This article aims to respond to calls for studies
addressing the complex ‘‘back stage” of audit prac-
tice in its social and organizational context (Power,
2003, p. 380). It makes three principal contribu-
tions. First, it provides ?eld evidence of a new audit
context that speaks to the power of audit as a com-
petitive resource capable of importing legitimacy to
entities across a variety of domains. Second, the
paper elaborates on the notion of the legitimacy
produced in the audit process. Through an analysis
of the unique relationships created by the engage-
ment, the notions of a?rmatory and derived legit-
imacy are introduced. Finally, the paper provides
evidence of the social construction of assurance
and the recursive development of audit scope, test-
ing procedures and reporting.
A number of implications and avenues for fur-
ther research can be drawn from the empirical data
and analysis presented here. Speci?cally, the ?eld
study provides support for the notion of the ‘‘audit”
as a competitive or discursive resource in new mar-
ketplaces and provides a counterpoint to rational-
ized accounts of the audit judgment process. In
this sense, ?eld research in emerging assurance mar-
kets has considerable potential to enrich a research
literature dominated by experimental psychology
and analytical economics. Moreover, although
beyond the scope of the present paper, the parallels
between business school rankings and the nature
and impact of value-for-money audits in the public
sector (see Radcli?e, 1998, 1999) are stark. Where
organizations do not have clear measures of pro-
ductivity which relate their inputs to their outputs,
the audit of e?ciency and e?ectiveness is in fact a
process of de?ning and operationalizing measures
of performance for the audited entity. That is, the
e?ciency and e?ectiveness of organizations are
not so much veri?ed as constructed around the
audit process itself. Future researchers could use-
fully examine the way in which making e?ectiveness
auditable is closely bound up with de?ning perfor-
mance and installing a management system to mea-
sure that performance.
References
Ahrens, T., & Chapman, C. S. (2005). Management control
systems and the crafting of strategy: A practice-based view.
In C. S. Chapman (Ed.), Controlling strategy: management,
accounting, and performance measurement. Oxford: Oxford
University Press.
Ahrens, T., & Chapman, C. S. (2007). Management accounting
as practice. Accounting, Organizations and Society, 32(1–2),
1–27.
AICPA (American Institute of Certi?ed Public Accountants).
(1996). Report of the AICPA Special Committee on Assur-
ance Services (Elliott Committee). New York: AICPA.
Anderson, S. W. (1995). A framework for assessing cost
management system changes: the case of activity based
costing implementation at General Motors, 1986–1993.
Journal of Management Accounting Research, 7(1), 1–51.
Basu, O., Dirsmith, M., & Gupta, P. (1999). The coupling of the
symbolic and the technical in an institutionalized context:
The negotiated order of the GAO’s audit reporting process.
American Sociological Review, 64(4), 506–526.
Bell, T. B., Marrs, F. O., Solomon, I., & Thomas, H. (1997).
Auditing organizations through a strategic systems lens: The
KPMG business measurement process. Montvale, NJ:
KPMG Peat Marwick.
Bell, T. B., Peecher, M. E., & Solomon, I. (2005). The 21st
century public company audit: Conceptual elements of
KPMG’s global audit methodology. Montvale, NJ: KPMG
International.
Bhattacharjee, S., Moreno, K., & Yardley, J. (2005). Auditors
as underwriters: An alternative framework. International
Journal of Auditing, 9(1), 1–19.
Bickersta?e, G., & Ridgers, B. (2007). Ranking the business
schools. Journal of Management Development, 26(1), 61–66.
Bradshaw, D. (2002). How to put together a 2100-piece data
jigsaw. Financial Times, January 17. Available at: .
Accessed 3.2.2007.
Bradshaw, D. (2007a). Auditing: Aim for accuracy. Financial
Times, January 17. Available at: . Accessed 7.02.
2007.
138 C. Free et al. / Accounting, Organizations and Society 34 (2009) 119–140
Bradshaw, D. (2007b). Business school rankings: The love–hate
relationship. Journal of Management Development, 26(1),
54–60.
Chaney, P. K., & Philipick, K. L. (2002). Shredded reputation:
The cost of audit failure. Journal of Accounting Research,
40(4), 1221–1245.
Chua, W. F. (1995). Experts, networks and inscriptions in the
fabrication of accounting images: A story of the represen-
tation of three public hospitals. Accounting, Organizations
and Society, 20(2/3), 111–145.
Cohen, J., Krishnamoorthy, G., & Wright, A. (2002). Corpo-
rate governance and the audit process. Contemporary
Accounting Research, 19(4), 573–593.
Cooper, D., & Robson, K. (2006). Accounting, professions and
regulation: Locating the sites of professionalization.
Accounting Organizations and Society, 31(4/5), 415–444.
Curtis, E., & Turley, S. (2007). The business risk audit – A
longitudinal case study of an audit engagement. Accounting
Organizations and Society, 32(4/5), 439–461.
Datar, S. M. G. A., Feltham, G., & Hughes, J. (1991). The role
of audits and audit quality in valuing new issues. Journal of
Accounting and Economics, 14(1), 3–49.
Davies, A., & Salterio, S. (2007). Financial Times business
school rankings: A non-traditional assurance case in three
parts. Accounting Perspectives, 6(1), 95–113.
DeAngelo, H., DeAngelo, L., & Zimmerman, J. (2005). What’s
really wrong with U.S. business schools? July. Available at:
.
Dichev, I. (1999). How good are business school rankings? The
Journal of Business, 72(2), 201–213.
Dichev, I. (2001). News or noise? Estimating the noise in the US
News university rankings. Research in Higher Education,
42(3), 237–266.
Downes, D. (2000). Does BusinessWeek ranking matter? The
MBA Newsletter, 8(9), 5–10.
Eisenhardt, K. M. (1989). Building theories from case study
research. Academy of Management Review, 14(4), 532–550.
Elliott, R. K. (1995). The future of assurance services: Impli-
cations for academia. Accounting Horizons, 9(4), 118–127.
Elsbach, K. D., & Kramer, R. M. (1996). Members’ responses
to organizational identity threats: Encountering and coun-
tering the Business Week rankings. Administrative Science
Quarterly, 41(3), 442–476.
Epstein, M. J., & Geiger, M. A. (1994). Investor views of audit
assurance: Recent evidence of the expectation gap. Journal
of Accountancy, 177(1), 60–65.
Fischer, M. (1996). Realizing the bene?ts of new technologies as
a source of audit evidence: An interpretive ?eld study.
Accounting, Organizations and Society, 21(2/3), 219–242.
Foucault, M. (1991). Governmentality. In G. Burchell, C.
Gordon, & P. Miller (Eds.), The foucault e?ect: Studies in
governmentality (pp. 87–104). London: Harvester
Wheatsheaf.
Gendron, Y. (2000). Openness to context-based research: The
gulf between the claims and actions of Big Six ?rms in the
USA. Accounting, Auditing and Accountability Journal,
13(2), 175–196.
Gendron, Y. (2001). The di?cult client-acceptance decision in
Canadian audit ?rms: A ?eld investigation. Contemporary
Accounting Research, 18(2), 283–310.
Gendron, Y. (2002). On the role of the organization in auditors’
client-acceptance decisions. Accounting, Organizations and
Society, 27(7), 659–684.
Gendron, Y., & Barrett, M. (2004). Professionalization in
action: Accountants’ attempt at building a network of
support for the WebTrust seal of assurance. Contemporary
Accounting Research, 21(3), 563–602.
Gendron, Y., Cooper, D., & Townley, B. (2007). The construc-
tion of auditing expertise in measuring government perfor-
mance. Accounting, Organizations and Society, 32(1–2),
101–129.
Gibbins, M., McCracken, S., & Salterio, S. (2007). The
auditor’s negotiation strategy selection: Nature of the
auditor–client management relationship and ?exibility of
initial accounting position, Working Paper. Queen’s Uni-
versity, 2007.
Goldman, A., & Barlev, B. (1974). The auditor-?rm con?ict of
interests: Its implications for independence. The Accounting
Review, 49(4), 707–718.
Gri?n, P. & Lont, D. (2005). The e?ects of auditor dismissals
and resignations on audit fees: Evidence based on SEC
disclosures under Sarbanes-Oxley. Working paper, Univer-
sity of California at Davis. Available at SSRN: . Accessed 1.05.2007.
Gunz, S., & McCutcheon, J. (1991). Some unresolved ethical
issues in auditing. Journal of Business Ethics, 10(10), 777–785.
Hopwood, A. (1983). On trying to study accounting in the
contexts in which it operates. Accounting, Organizations and
Society, 8(2/3), 287–305.
Humphrey, C., & Moizer, P. (1990). From techniques to
ideologies: An alternative perspective on the audit function.
Critical Perspectives on Accounting, 1(3), 217–238.
Jarzabkowski, P. (2004). Strategy as practice: Recursiveness,
adaptation, and practices-in-use. Organization Studies,
25(4), 529–560.
Knechel, R., Salterio, S., &Ballou, B. (2007). Auditing: Assurance
and risk (3rd ed.). Cincinnati, OH: Thomson Southwest.
Lounsbury, M. (2007). Institutional rationality and practice
variation: New directions in the institutional analysis of
practice. Accounting, Organizations and Society.
doi:10.1016/j.aos.2007.04.001.
Mayhew, B. W., & Pike, J. E. (2004). Does investor selection of
auditors enhance auditor independence? The Accounting
Review, 80(1), 289–314.
McEnroe, J., & Martens, S. (2001). Auditors’ and investors’
perceptions of the ‘‘expectation gap”. Accounting Horizons,
15(4), 345–358.
Melumad, N. D., & Thoman, L. (1990). On auditors and the
courts in adverse selection setting. Journal of Accounting
Research, 28(1), 77–120.
Miller, P., & Rose, N. (1990). Governing economic life.
Economy and Society, 19(1), 1–31.
Mills, J. H., Weatherbee, T., & Colwell, S. (2006). Ethnosta-
tistics and sensemaking: Making sense of university and
C. Free et al. / Accounting, Organizations and Society 34 (2009) 119–140 139
business school accreditation and rankings. Organizational
Research Methods, 9(4), 491–515.
Mintzberg, H. (2004). Managers not MBAs: A hard look at the
soft practice of managing and management development. San
Francisco, CA: Berrett-Koehler Publishers.
Mobus, J. L. (2005). Mandatory environmental disclosures in a
legitimacy theory context. Accounting, Auditing and
Accountability Journal, 18(4), 492–517.
Moore, D. A., Tetlock, P. E., Tanlu, L., & Bazerman, M. H.
(2006). Con?icts of interest and the case of auditor
independence: Moral seduction and strategic issue cycling.
Academy of Management Review, 31(1), 10–29.
Ogden, S., & Clarke, J. (2005). Customer disclosures, impression
management and the construction of legitimacy: Corporate
reports in the UK privatised water industry. Accounting,
Auditing and Accountability Journal, 18(3), 313–345.
Orlikowski, W. J. (2000). Using technology and constituting
structures: A practice lens for studying technology in
organizations. Organization Science, 11(4), 404–428.
Pfe?er, J., & Fong, C. T. (2004). The business school ‘business’:
Some lessons from the US experience. The Journal of
Management Studies, 41(8), 1501–1520.
Power, M. (1996). Making things auditable. Accounting,
Organizations and Society, 21(2/3), 289–315.
Power, M. (1997). The audit society: Rituals of veri?cation.
Oxford: Oxford University Press.
Power, M. (2000). The audit society – Second thoughts.
International Journal of Auditing, 4(1), 111–119.
Power, M. (2003). Auditing and the production of legitimacy.
Accounting, Organizations and Society, 28(4), 379–394.
Power, M. (2007). Business risk auditing – Debating the history
of its present. Accounting, Organizations and Society, 32(4/
5), 379–382.
Radcli?e, V. (1998). E?ciency audit: An assembly of rational-
ities and programmes. Accounting, Organizations and Soci-
ety, 23(4), 377–410.
Radcli?e, V. (1999). Knowing e?ciency: The enactment of
e?ciency in e?ciency auditing. Accounting, Organizations
and Society, 24(4), 333–362.
Reingold, J., & Habal, H. (1998). How we kept the data
unsullied. Business Week, 19(October), 94.
Robson, K., Humphrey, C., Khalifa, R., & Jones, J. (2007).
Transforming audit technologies: Business risk audit meth-
odologies and the audit ?eld. Accounting, Organizations and
Society, 32(4/5), 409–438.
Ronen, J. (2002). Policy reforms in the aftermath of accounting
scandals. Journal of Accounting and Public Policy, 21(4/5),
281–286.
Rose, N. (1992). Governing the enterprising self. In P. Heelas &
P. Morris (Eds.), The values of the enterprise culture: The
moral debate. London: Routledge.
Rose, N., & Miller, P. (1992). Political power beyond the state:
Problematics of government. British Journal of Sociology,
43(2), 173–205.
Sarbanes-Oxley Act of 2002. (2002). An act to protect investors
by improving the accuracy and reliability of corporate
disclosures made pursuant to the securities laws, and for
other purposes. Library of Congress. .
Spence, M. (1973). Job market signalling. Quarterly Journal of
Economics, 87(3), 355–374.
Thompson, N. (2000). Playing with numbers. Washington
Monthly(September).
Van Maanen, J., & Pentland, B. (1994). Cops and auditors: The
rhetoric of records. In S. Sitkin & R. Bies (Eds.), The
legalistic organization. Thousand Oaks, California: Sage.
Wedlin, L. (2006). Ranking business schools: Forming ?elds,
identities and boundaries in international management edu-
cation. Cheltenham: Edward Elgar.
Welch, I. (2002). The 2000 business week rankings of business
schools: Why they are both harmful and wrong. Available
at: .
Accessed 7.11.2006.
Yee, C. (2004). Ranking methods under ?re. University Wire,
4(September), 1.
Zell, D. (2001). The market-driven business school: Has the
pendulum swung too far? Journal of Management Inquiry,
10(4), 324–338.
Zimmerman, J. L. (2001). Can American business schools
survive? September 5. Simon School of Business Working
Paper No. FR 01-16. Available at: . Accessed 12.10.2006.
140 C. Free et al. / Accounting, Organizations and Society 34 (2009) 119–140

doc_982405280.pdf