Description
This paper examines the role of accounting in society by looking at the consumption of accounting signs during the
financial restructuring of a corporation. The paper builds upon insights from prior research on accounting as simulacrum
and hyperreality. It examines how accounting numbers serve as reconfigurable signs that construct appropriate
‘‘crises”, motivate government intervention, and marshal stakeholders towards solutions
Fearful asymmetry: The consumption of accounting signs
in the Algoma Steel pension bailout
Cameron Graham
*
Schulich School of Business, York University, 4700 Keele Street, Toronto, Ontario, Canada M3J 1P3
Abstract
This paper examines the role of accounting in society by looking at the consumption of accounting signs during the
?nancial restructuring of a corporation. The paper builds upon insights from prior research on accounting as simula-
crum and hyperreality. It examines how accounting numbers serve as recon?gurable signs that construct appropriate
‘‘crises”, motivate government intervention, and marshal stakeholders towards solutions. The incident at the heart
of this study is the 2001 bailout of the Algoma Steel pension plan by the Ontario government. The incident demon-
strates how accounting technologies are required both for the production of accounting signs and for their consump-
tion. The paper asks how the production and consumption of accounting signs is di?erent from that of other
communication signs, what role consumers of accounting signs play in determining their meaning, and what di?erence
this makes in how corporate pension plans are protected by government. It concludes that the structures and mecha-
nisms surrounding the consumption of accounting signs enable di?erent stakeholders to in?uence the production of
meaning at the moment when accounting signs are consumed, changing the way that risk and wealth are redistributed,
and shaping government intervention.
Ó 2008 Elsevier Ltd. All rights reserved.
Introduction
In this paper, I draw on the work of Baudrillard
to examine the consumption of accounting signs by
various stakeholders during a corporate ?nancial
crisis. Baudrillard (1972, 1973) developed provoca-
tive semiological theories about the production and
the consumption of signs in society. I ask how the
production and consumption of accounting signs
is di?erent from the production and consumption
of other signs, and what di?erence this makes in
the role of accounting in society. I use a Canadian
pension accounting example to explore how
accounting signs are used by management, govern-
ment, creditors, and other corporate stakeholders.
In 2001, the severely underfunded pension plan of
Algoma Steel was taken over by the Ontario gov-
ernment. The analysis of documents pertaining
to this incident demonstrates how accounting
0361-3682/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.aos.2008.01.001
*
Tel.: +1 416 736 2100x77926; fax: +1 416 736 5687.
E-mail address: [email protected]
Available online at www.sciencedirect.com
Accounting, Organizations and Society 33 (2008) 756–782
www.elsevier.com/locate/aos
technologies structure the production and con-
sumption of accounting signs, exacerbating di?er-
ences in how stakeholders are able to consume
these signs, triggering and limiting government
intervention, and shifting the boundary between
the private and public sectors.
My analysis draws upon the work of Macintosh,
Shearer, Thornton, and Welker (2000), who dis-
cussed ?nancial accounting theory using Baudril-
lard’s notions about signs (see Baudrillard, 1983;
Baudrillard, 1995, 1988). This approach asserts
that accounting signs have become divorced from
any underlying objective reality. With respect to
pension accounting, I argue that Canadian GAAP
for de?ned-bene?t pensions has codi?ed a system
of arbitrary signs. The study of Algoma Steel dem-
onstrates how these accounting signs can be con?g-
ured to create representations of pension plans that
are useful to corporations in managing government
and other stakeholders. I also argue that inequali-
ties amongst stakeholders in their ability to con-
sume accounting signs contribute to di?erences in
the power they are able to exert over the meaning
of those signs. The study will be of interest to
accounting researchers because it expands, with
concrete examples, our understanding of how
accounting functions as a language, and in particu-
lar how accounting di?ers from other languages at
the point where its signs are consumed.
This study contributes to our understanding of
the role of accounting information in de?ning the
boundary between the private and public sectors.
It provides speci?c evidence of how the production
and consumption of accounting information, as a
highly structured form of communication, alters
the very institutional setting in which that commu-
nication occurs. The transference of pension liabil-
ities from private corporations to government, and
the intervention by government in private ?nancial
crises, are examples of how the boundary between
the private and the public sector can shift.
The study is particularly relevant because Alg-
oma Steel was an early adopter of revised Cana-
dian pension disclosure standards in 2000.
Similar changes to pension disclosure standards
have been implemented by the International
Accounting Standards Board under IAS 19 (IASB,
2004) and the Financial Accounting Standards
Board under FAS 158 (FASB, 2006). These
changes have provoked concern (e.g. Brooksbank,
2006; Liddle, 2002; Pension Changes Could Cost
$180b, 2006; Squeeze on Pensions Could Tighten,
2006). This study helps shed timely light on the
impact of such changes by examining a ‘‘pension
crisis” triggered in part by revisions to Canadian
pension disclosure standards.
Retracing Baudrillard’s steps
The underlying thesis of this paper is that
accounting, as a system of signs, can be e?ectively
critiqued by approaching it from the perspective
of linguistic theories about the function of signs
in today’s society. Prior research has spent surpris-
ingly little time examining accounting from such a
perspective, given the prominence of linguistic the-
ory in philosophy in the 20th century. Pioneers in
linguistic or literary approaches to accounting
research include Belkaoui (1978, 1980), who
asserted that accounting is a language and explored
the cognitive implications of using that language;
Lavoie (1987), who explored the hermeneutics of
economic decision making with accounting infor-
mation; Arrington and Francis (1989), who intro-
duced postmodern linguistic theory to the
accounting literature; and Boland (1989), who
showed how a hermeneutic approach could break
down the dichotomy between subjectivism and
objectivism. Although many critical accounting
articles are indebted to linguistic theory in a general
way, explicitly linguistic approaches to accounting
research after 1989 are unfortunately few.
1
The one
1
I am setting aside the excellent stream of research on
accounting and visual images (Daly & Schuler, 1998; Graves,
Flesher, & Jordan, 1996; Preston, Wright, & Young, 1996;
Preston & Young, 2000), which draws on linguistic theorists. In
the pages of AOS, among the few explicitly linguistic papers, as
opposed to papers that merely have textual data, are Cooper
and Puxty’s (1994) textual analysis of a professional accounting
journal article, and Francis’s (1994) exploration of the herme-
neutics of auditing. Linguistic treatments of accounting are
both more frequent and more recent outside AOS (e.g.,
Armstrong, 2000; Evans, 2004; Everett, 2004; Macintosh &
Baker, 2002; Macintosh & Shearer, 2000; McGoun, 1997;
Mouck, 1994; Walters-York, 1996).
C. Graham/ Accounting, Organizations and Society 33 (2008) 756–782 757
I engage here, by Macintosh et al. (2000), draws on
Baudrillard’s notions of simulacrum, implosion,
and hyperreality to discuss ?nancial accounting
theory. Macintosh et al. (p. 14) argue that Baudril-
lard’s radical semiological theories are relevant to
accounting because they deal with changes in lan-
guage, information technology, communication
and media that fundamentally a?ect accounting.
2
They use Baudrillard’s insights to develop a cri-
tique of accounting based on the production and
consumption of information. This approach is par-
ticularly appropriate for the present study because
the resolution of the 2001 Algoma Steel crisis
hinged on how the company’s accounting ‘‘signs”
were produced and consumed.
In semiotic terms (see, for example, Saussure,
1961), signs consist of a word or picture – the sig-
ni?er – that evokes a concept – the signi?ed. The
signi?er and signi?ed together act as a sign point-
ing to some underlying thing – the referent. The
word ‘‘corporation”, for instance, evokes the
notion of an incorporated company and can be
used to refer to a speci?c organization. In tradi-
tional accounting theory, signs have an objective
referent: ‘‘net income” measures a real surplus of
a company’s economic activity. Macintosh et al.
argue, however, that accounting signs have lost
their objective referent. They take up Baudrillard’s
(1983) provocative suggestion that signs now take
their meaning only from their relationship to other
signs in the communicative system. As noted by
Macintosh et al. (p. 40), Baudrillard argued
(1983, pp. 11–12) that signs have passed through
four historical phases. In the ?rst phase, signs were
a re?ection of a profound reality. In the second,
signs masked or denatured this profound reality.
In the third, signs masked the absence of a pro-
found reality. And, in the fourth, which according
to Baudrillard is the phase of today, signs precede
reality.
Macintosh et al. equate the ?rst phase to pre-
historic accounting, and trace the accounting
notions of capital and income through Baudril-
lard’s other phases. Accounting, they maintain, is
now without objective referent, and they give sev-
eral supporting examples, including accounting for
executive stock options and earnings management.
They argue that, just as in Baudrillard’s examples
of the hyperreal society, accounting signs now pre-
cede the reality they purport to represent, creating
that reality through their sign value: accounting
signs have gained independence from the real. As
Baudrillard put it, ‘‘from now on signs will
exchange among themselves exclusively, without
interacting with the real . . . [they are] at last free
for a structural or combinatory play that succeeds
the previous role of determinate equivalence”
(Baudrillard & Poster, 1988, p. 125); quoted in
(Macintosh et al., 2000, p. 40).
Macintosh et al. argue (p. 40) that Baudrillard’s
characterization totalizes the present age and
ignores the way that accounting signs still main-
tain a connection to an underlying, albeit social,
reality. They ask, ‘‘why is it that accounting infor-
mation is used extensively in economic and social
relations when, according to our Baudrillardian
analysis, it does not refer to any real objective
realm, and has had changing referents over time?”
(p. 40). They point to the clean surplus model (Fel-
tham & Ohlson, 1995) as a partial answer, suggest-
ing that as long as capital and income articulate in
a clean surplus relationship, it does not matter
how each is de?ned. The accounting notions of
capital and income take their sign values from
the system of signs in which they are embedded,
rather than from any underlying objective ‘‘thing”
represented by capital or income. In this sense, the
signs are arbitrary.
If accounting is a system of arbitrary signs –
that is, a language – then it is amenable to linguis-
tic theoretical analysis. However, caution is
required. Accounting di?ers from English and
other vernaculars in several socially important
respects. It is formally de?ned by a professionally
dominated process, for instance. Its use is partially
professionalized. It has peculiar technical features
2
Mattessich (2003) has criticized Macintosh et al. and
Baudrillard from the perspective of objective realism. Mattess-
ich, who seems to be irritated by Baudrillard’s refusal to play
the scienti?c language game, clings to the notion that science –
and hence accounting – can represent objectively the reality of
which it is a part. This stance has been discredited by many
philosophers, notably Rorty (1991). Mattessich’s philosophy
has been critiqued by Archer (1998) and Mouck (2004), but
Macintosh et al. have not yet responded to his treatment of
Baudrillard.
758 C. Graham/ Accounting, Organizations and Society 33 (2008) 756–782
pertaining to its numeric and calculative nature
that lend it a false aura of objectivity and precision
(cf. Tinker, Merino, & Neimark, 1982; Tinker,
1991) and enforce a certain arithmetic grammar
over what can be said. Above all – and this is
something Macintosh et al. neglect in their inter-
pretation of Baudrillard for accounting – account-
ing functions as a system of signs primarily in the
?nancial markets, not the product markets. The
?nancial markets are unlike the mass consumer
market featured in Baudrillard’s analysis of signs.
In order to explain the implications of this more
fully, I want to retrace the development of Baudril-
lard’s notions of simulacra and hyperreality.
Baudrillard’s analysis of signs began with his
e?orts to update Marxist theory to account for
the shift in emphasis in Western society from pro-
duction to consumption (Baudrillard, 1968, 1970).
He formalized his theories in a series of essays
published as For a Critique of the Political Econ-
omy of the Sign (1972 [trans. 1981]). In these
essays, Baudrillard draws a contrast between sym-
bolic exchange and consumption (1981, pp. 63–
65). An example of symbolic exchange is the giving
of a gift. A gift is more than an object, it is sym-
bolic of a unique relationship between people. A
commodity, on the other hand, is a sign of rela-
tions of production. It takes its meaning not from
a unique relationship between people but from its
‘‘di?erential relation to other signs” (p. 66). What
Baudrillard means by this is that the basis of con-
sumption is meaningful social exchange, not the
brute satisfaction of individual needs. He com-
pares consumption to language. Language, he
says, cannot be understood on an utterance-by-
utterance basis because individual utterances are
only intelligible in the context of language. So con-
sumption cannot be understood as the satisfaction
of individual needs, but must be understood as a
social system for producing meaning through dif-
ferences in the sign value of commodities (p. 75).
Baudrillard’s most formal moments come when
he compares the functional logic of use value, the
economic logic of exchange value, the di?erential
logic of sign value, and the logic of symbolic
exchange (pp. 123–129). In teasing out the connec-
tions between these logics, Baudrillard comes to
the conclusion that, just as exchange value domi-
nates use value in our economic system, so the sig-
ni?er dominates the signi?ed in our signi?cation
system. The principle of economic circulation is
the production of equivalence, exchange value.
The principle of circulation in our system of
signi?cation is the production of di?erence, the
signi?er. Thus, in order to understand the con-
sumption of objects completely, says Baudrillard,
one must understand them as signs, not only as
commodities.
Baudrillard’s critique reaches its zenith in the
essay for which the book is named. Here, Baudril-
lard argues that the problems of Marxist thinking
derive from an arti?cial separation of the economy
and culture, paralleled in the arti?cial separation
of the commodity and the sign, requiring ‘‘magical
thinking” regarding ideology to reunite them (pp.
142–145). He makes two matching claims: that
because the logic of the commodity is at the heart
of the sign, signs can function as both exchange
value and use value; and that because the structure
of the sign is at the heart of the commodity, com-
modities can function inherently as signs (p. 146).
The commodity functions as a system of commu-
nication governing social interaction, reducing
the symbolic to the form of the sign and rational-
izing all exchange (p. 147). The key point in this
essay is Baudrillard’s recognition that the moment
of interconversion between commodity and sign,
when meaning is generated from the commodity
and the sign is commodi?ed, is the moment of con-
sumption, not production (p. 147).
3
This is what
de?nes today’s society, and what escapes tradi-
tional and neo-Marxist analysis with its focus on
production and its arti?cial separation of economy
and culture. What Baudrillard is proposing is a
more general critique of political economy.
Baudrillard makes a subtle point here that will
prove important in the discussion of Algoma Steel
3
This may seem fairly straightforward to those familiar with
the hermeneutical precepts of Heidegger (1962) and Gadamer
(1976), who argued that meaning is created when a text is read,
and that the meaning will go beyond the original intentions of
the author (cf. Lavoie, 1987; Boland, 1989). Baudrillard’s
insight is to link this line of thought to the production and
consumption of the commodity, and to recognize the commod-
ity as a sign.
C. Graham/ Accounting, Organizations and Society 33 (2008) 756–782 759
that follows. He argues that the arbitrariness of the
sign does not have to do with the referent alone. It
is not the sign that is arbitrarily connected to the
referent. It is the signi?er that is arbitrarily con-
nected to the signi?ed/referent pair. The signi?ed
and referent together form the content of the sign,
under the ‘‘aegis” of the signi?er (p. 151). The
implication of this is that the referent, the ‘‘objec-
tive” thing that we often presume to be re?ected in
the sign, is actually not independent of the sign:
‘‘In a profound sense, the referent is the re?ection
of the sign . . .” (p. 151). Baudrillard’s later notions
of simulacrum, implosion, and hyperreality derive
from this insight.
This critique of the sign positions Baudrillard
to break with Marxism. Baudrillard suggests, for
instance, that Marx’s analysis of the commodity
was incomplete because he never developed a cri-
tique of use value (pp. 128–129). The actual break
did not come, however, until Baudrillard’s next
book, The Mirror of Production (1973 [trans.
1975]). In this he criticized Marx for reducing
people to producers, life to economic production,
by accepting the central and unfortunate proposi-
tion of political economy: that people are labour
power. To Baudrillard, people are no better o?
being pegged as homo faber than they are as homo
economicus (1975, p. 49). He illustrates this by
pointing to Marx’s imposition of a ‘‘mode of pro-
duction” on primitive societies. He argues that
Marx has only adopted and reproduced the fun-
damental concepts of political economy, and that
it was necessary to look next at the mode of sig-
ni?cation and a critique of the sign (p. 51). He
argues that Marx took the economy and produc-
tion as axiomatic, and is therefore unable to
account for how the principle of production is
itself produced (p. 66). He goes so far as to say
that Marx’s anachronistic analysis of primitive
and feudal societies results in a ‘‘theoretical, polit-
ical and strategic miscomprehension of capitalist
formations themselves” (p. 107).
Baudrillard (1975, pp. 119–129) addresses these
limitations by examining how monopoly capital-
ism di?ers from competitive capitalism. Monopoly
capitalism, he argues, represents a revolutionary
change in capitalism, not merely an extension.
Exchange is no longer just abstracted, it is opera-
tionalized, coded, through the production of
meaning and di?erence. What matters is not the
control of the means of material production, but
control of the code (p. 122). While the consump-
tion of signs produces di?erence, the process is
not that of social di?erentiation through conscious
consumption to consolidate one’s class status.
Rather, the process is unconscious (pp. 122–123)
because it is programmed. Consumption is no
longer contingent, supply and demand are no
longer dialectically related; competition, supply
and demand remain only as myths to support the
system (p. 125). Consumption is now driven by
planned socialization, by advertising, for example,
which provides not only the answers but also the
questions. In this system, signi?ers are divorced
from their contents, both the signi?ed and the ref-
erent, to play freely in self-referential exchange
with other signi?ers. The system thus neutralizes
all contradictions by abolishing referents (p.
129), that is, by collapsing or imploding the di?er-
ence between the signi?ed and the referent. The
radical nature of Baudrillard’s critique comes to
life when he shows how his analytic approach
explains what Marxist analysis cannot, namely
the relegation of entire groups of people – stu-
dents, races, women – to positions outside the
realm of production. These groups cannot be
aligned with workers in revolt against those who
control the means of production. They must revolt
against the code (pp. 131–141).
This, then, is the radical and intensely theorized
backdrop against which Baudrillard’s (1983)
notions of simulacra, implosion, and hyperreality
must be viewed. Macintosh et al. (2000, pp. 14–
16) provide a solid summary of these three
notions, but they do not connect them to their the-
oretical derivation. Brie?y, the simulacrum is the
sign that has been divorced from its referent, and
takes its meaning only from its relation to other
signs. Implosion is the collapse of the boundary
between two concepts or realms, and of the di?er-
ence between sign and referent. Hyperreality is the
state of society in which simulation and simulacra
dominate. When Macintosh et al. use these con-
cepts to explore accounting, they achieve little
more than a recapitulation of traditional account-
ing history in trendy terms. By picking up Baudril-
760 C. Graham/ Accounting, Organizations and Society 33 (2008) 756–782
lard at the point where terms like simulacra and
hyperreality emerge, and ignoring the prior devel-
opment of his thought, they reinvigorate an ana-
lytical model of accounting theory but threaten
nothing. Baudrillard, they claim (pp. 44–45), is
obsessed with the sign world and marginalizes
relations of power. Such an accusation is incom-
prehensible given even the abbreviated chronology
of his theoretical development provided above.
What then to do about Baudrillard and
accounting? I will argue in what follows that
Baudrillard must be taken just seriously enough.
He is a provocateur, and must not under any cir-
cumstances be taken too literally.
4
There are
glaring weaknesses in what he says, such as a
failure to ground his notion of ‘‘symbolic
exchange” empirically, which makes it seem
somewhat nostalgic; a failure to de?ne more pre-
cisely the ‘‘code” or identify who controls it and
how; as well as a tendency towards fatalism and
nihilism in later works (e.g. Baudrillard, 1993).
However, he provokes us to rethink our assump-
tions, and provides a compelling vocabulary for
discussing society that is rooted in his own dis-
section of Marx. Implications regarding relations
of power can quite readily be brought out when
using his vocabulary to understand present-day
accounting. This is what I will do in the follow-
ing case study. I will use the incident of the
?nancial collapse of Algoma Steel in 2001 and
the subsequent bailout of its pension plan by
the Ontario government to show how accounting
signs function in Baudrillard’s terms. I will, how-
ever, also show how Baudrillard’s analysis is lim-
ited for accounting by certain features in the
mode of production and the mode of consump-
tion of accounting signs.
Before I proceed with my discussion of the Alg-
oma case, I want to clarify what is distinct about
the form of accounting signs. Semiotics, as noted
above, describes the sign as a signi?er/signi?ed
pair pointing to a referent. The accounting sign
has a peculiar form, in that accounting informa-
tion consists of a system of equations. The basic
form of the accounting sign is ‘‘Label = Value”,
for example ‘‘Accounts Receivable = $100,000”.
These signs can be built up into higher level signs,
such as ‘‘Current Assets = $1,000,000”, and still
higher ones, ‘‘Total Assets = $10,000,000”. In
?nancial accounting, these basic forms are
arranged according to the logic of the double
entry. Certain events in the course of business
are selectively recorded in variations on the form
‘‘Debits = Credits”. This is an equation of equa-
tions, each side consisting of one or more basic
signs in the form ‘‘Label = Value”. The combina-
tion and recombination of the basic form and the
transactional form gives rise to other equations at
the level of the ?nancial statement proper, such as
‘‘Assets = Liabilities + Owners’ Equity” and
‘‘Revenue – Expenses = Net Income”. It seems
clear that the form I have called the basic one is
no more basic than the transactional form or the
higher level forms of the balance sheet and income
statement, but I will call it ‘‘basic” for expository
purposes, because it is the form most consistently
used throughout the Algoma Steel legal
proceedings.
One must avoid concluding that the ‘‘Label”
in the basic accounting equation is the signi?er
and the ‘‘Value” is the signi?ed, or worse, the ref-
erent. The entire equation ‘‘Label = Value” is a
signi?er. The form of the accounting sign supports
the attempt to impose a single meaning on the
sign, to eliminate its symbolic potential and reduce
it to the unequivocal (Baudrillard, 1981, p. 149).
For the sign-equation suggests ?nality, as it seems
to contain in the monetary amount its own refer-
ent. Yet the monetary amount is but a sign of
value. It takes its meaning from its relationship
to other signs of value, for instance, the values that
were assigned to that label in previous ?nancial
statements, or the values that are assigned to a
similar label in other companies’ ?nancial state-
ments, or the values that were expected by ana-
lysts. It also takes its meaning from its relation
to other signs of value in the same set of ?nancial
statements, such as the meaning of one liability in
relation to other liabilities. These meanings, as will
be seen, are heavily contested, and are not strictly
related to the number assigned by the producer of
the accounting sign.
4
The tedious literalism of Schoonmaker’s (1994) critique of
Baudrillard is an example of what can happen. The metaphor-
ical qualities of terms like ‘‘code” and ‘‘binary” completely
escape her.
C. Graham/ Accounting, Organizations and Society 33 (2008) 756–782 761
Data sources for a study of accounting signs
In basic outline, the Algoma Steel crisis and
bailout ran as follows. In 2000 and 2001, the com-
pany encountered ?nancial di?culties. It faced
severe cash ?ow problems, and its total liabilities
exceeded all its assets, even without including the
company’s obligations towards its pension plans
and other employee future bene?ts plans.
5
These
were non-contributory, de?ned-bene?t plans cov-
ering substantially all employees. They were
underfunded by almost $800M (Algoma Steel,
2001a). The size of the funding de?cit and its swift
growth alarmed investors and creditors. The com-
pany ?led for court protection, prompting the
Ontario government to step in. Eventually, an
agreement was reached between the province, the
company and its various stakeholders, in which
the company was restructured and a substantial
portion of the pension liability was assumed by
the government (Algoma Steel, 2001b).
In order to expose the function of accounting
signs in this series of events, a variety of related data
sets must be examined. The public discourse sur-
rounding the Algoma Steel pension plan incident
is one set. It supports an analysis of howaccounting
numbers were consumed and interpreted in this
contentious pension incident. This discourse is rep-
resented by articles fromCanadian newspapers dur-
ing 2001, when the crisis made headlines. The
Canadian Newsstand database was searched for all
references to Algoma Steel during 2001, and the rel-
evant articles
6
were read to determine howaccount-
ing signs were taken up and represented in the
media. This representation includes not only spe-
ci?c citations of individual accounting numbers,
but the connection made in the media between Alg-
oma’s accounting numbers and consequent political
and ?nancial developments in the case.
The history of pension disclosures made by Alg-
oma Steel is a second data set. This information
allows us to reconstruct the way the pension fund
was presented, such that it was transformed using
accounting signs from a politically negligible pri-
vate fund of wealth into a necessary target of gov-
ernment intervention. Annual reports for Algoma
Steel going back to the company’s 1991 restructur-
ing will be included; this cuto? permits
comparability by ignoring pre-restructuring signs
produced by the company.
Court documents from the bankruptcy proceed-
ings form a third data set. From six ?ling boxes
containing all the paper records from the proceed-
ings, 418 pages of information relevant to arrange-
ments with creditors were obtained. These pages
were selected by the author and a research assistant
by reviewing every document in the ?ling boxes and
pulling out everything substantive related to credi-
tor arrangements. The many routine documents
that form the legal paper trail, such as notices of
papers being served on witnesses, were ignored.
Documents available from other sources, such as
drafts of the ?nal agreement, were also ignored.
Many documents in the ?ling boxes were routinely
updated before the courts on a monthly basis or
better, such as statements from the bankruptcy
trustees. Of these, only one or two examples each
were retained. Key documents obtained from these
boxes included testimony from the Chief Financial
O?cer of the company, testimony from the court-
appointed mediator, and copies of letters sent to
key stakeholders towards the end of the settlement
negotiations.
Supplementing these data sets are the publicly
available Ontario government documents pertain-
ing to the bailout of Algoma Steel. These are avail-
able from electronic archives on the Ontario
government website. The documents include the
Pension Bene?ts Act (1990), related regulations,
various government news releases, and successive
revisions to the agreements developed between
stakeholders under the auspices of the govern-
ment’s pension intervention facilities.
5
Under Canadian regulations, future bene?ts to employees
are divided into two categories for both funding and accounting
purposes. Bene?ts corresponding to the notion of deferred
wages are properly called ‘‘pension bene?ts”, while other
bene?ts pertaining to extended health bene?ts, dental coverage,
and life insurance are given the unwieldy label ‘‘other employee
future bene?ts”. In this study, ‘‘pension plan” is used to refer
generally to both kinds of plans. Where necessary, ‘‘pension”
and ‘‘non-pension” future bene?ts are distinguished.
6
There were 416 articles on Algoma Steel in the database
from 2001. As many as half of them were duplicates or closely
related stories due to the use of wire services by multiple
newspapers.
762 C. Graham/ Accounting, Organizations and Society 33 (2008) 756–782
All the data are examined in the context of Sec-
tion 3461 of the CICA Handbook, which governs
the disclosure of future bene?ts for employees.
This permits us to understand the syntactic
features of the accounting signs, and to see how
the pension accounting signs derive their meaning
from their linguistic context. While interviews and
other local data would enhance the richness of the
study, the opportunity to pursue such sources will
be left for future research.
In order to understand the pension accounting
information within its full social context, other
data sources are used to establish the speci?c his-
torical antecedents of the pension incident. McDo-
wall’s (1984) narrative history of Algoma Steel has
been consulted to develop a description of the eco-
nomic history of the company and its impor-
tance to Ontario. Two other histories (Beatty &
Schachter, 2002; Nishman, 1995) were valuable
in describing the 1991 restructuring of the com-
pany. This restructuring led to the employee own-
ership situation that pertained at the time of the
‘‘pension crisis” in 2001.
The Algoma Steel bailout has been deliberately
selected because of its sheer magnitude and the
unprecedented size of the intervention by the pro-
vincial government.
7
This extraordinary event gen-
erated a rich availability of data. Because of this,
the case helps reveal social con?icts and account-
ing problems that may be latent, but are perhaps
undetectable, in many other pension funding cases.
Certain aspects of the Algoma Steel situation are
unusual, however. The steel industry presents
speci?c highly-institutionalized features: the indus-
try is heavily unionized, de?ned-bene?t pension
plans are prevalent, and wages are relatively high.
While workers surrendered future wages and
certain pension bene?ts during the incidents
examined here, the institutional support for these
workers nonetheless far exceeds that found in
many other areas of the Canadian economy.
Because of this, Algoma Steel represents in some
ways a ‘‘best case” scenario for Canadian work-
ers. Furthermore, this ‘‘crisis” primarily illus-
trates the social impact of pension accounting
when things go badly. This is important to note
because de?ned-bene?t plans in Canada create
asymmetries of risk and reward. Funding sur-
pluses typically accrue to workers but employers
are responsible for any funding de?cits.
8
Subse-
quent events have seen Algoma Steel enjoy a resur-
gence, exacerbating the asymmetries.
The Algoma Steel pension crisis and bailout
Algoma Steel is no stranger to crisis, having
gone through a series of collapses, rescues and
recoveries since its establishment at the turn of
the 20th century (McDowall, 1984, pp. 4–6). Given
the railway subsidies, land grants, special con-
tracts, tax breaks, and production bounties
Algoma received from the government during its
?rst 50 years (McDowall, pp. 33–38, 162, and
165), it is not farfetched to suggest that Algoma
owed its early existence to government support.
During the Second World War, government poli-
cies encouraging steel production stimulated an
expansion at Algoma that continued through
to the 1970s (McDowall, pp. 235–257). How-
ever, similar government policies in other coun-
tries in the 1980s led to global overproduction of
steel (Nishman, 1995, pp. 8–9), resulting in
substantial losses and a large debt load at Algoma.
Eventually, in 1991, Algoma Steel declared
insolvency and sought court protection from its
creditors.
The provincial government of the day was pro-
labour,
9
and was under political pressure to help
secure jobs. The federal government refused to
step in, leaving the provincial government to co-
7
Imperilled corporate pensions have been central features of
other major bankruptcy crises in Canada, notably Air Canada
in 2003–2004. However, what distinguishes the Algoma Steel
case is the Ontario government’s intervention speci?cally to
relieve the company of its pension liability.
8
Depending on the terms of the plan, funding surpluses may
be withdrawn by the company with the consent of workers.
This can happen, for example, as a result of labour negotiations
if the workers trade part of the surplus for other contractual
bene?ts.
9
The provincial government in Ontario in 1991 was formed
by the New Democratic Party, led by Bob Rae. The New
Democratic Party in Canadian politics has traditionally advo-
cated social democratic policies and enjoyed labour support.
C. Graham/ Accounting, Organizations and Society 33 (2008) 756–782 763
ordinate a task force involving major stakehold-
ers (Nishman, 1995, pp. 14–16). A mediated
settlement was reached based largely on the work-
ers’ proposal for a form of employee ownership.
10
Voting shares amounting to 60% of control of the
company were issued to an employee trust, with
the remaining shares publicly traded. In exchange,
workers agreed to a $200 million wage reduction,
bene?t reductions, and substantial job cuts (Beatty
& Schachter, 2002, pp. 30–31; Nishman, 1995, pp.
19–22).
This 1991 restructuring of Algoma Steel accom-
plished two important things. First, it established
the company as an employee-owned enterprise.
This development carried signi?cant implications
for the governance of the company. Management
in the new company was to be ‘‘independent”
(Nishman, 1995, p. 25), presumably meaning that
it was to operate independently of the majority
owners (the employees) in making strategic deci-
sions. Despite acquiring majority ownership and
a minority representation on the board, unionized
workers regarded the $200 million wage reduction
as a contract concession, not as a takeover move
(DaPrat, 2005a; Nishman, 1995, p. 25).
The second thing the 1991 restructuring accom-
plished was the entrenchment of the provincial
government’s role in ensuring the survival of the
company in times of crisis. Direct involvement
by both provincial and federal governments in
the establishment and growth of Algoma Steel
had already provided clear precedent for govern-
ment intervention in the company. The 1991 pro-
vincial government intervention rea?rmed this
line of action, and underscored the importance of
Algoma Steel to the Ontario economy. Hence, by
the time the pension crisis of 2001 arose, the com-
pany had been ?rmly established as a site of gov-
ernment action.
In the mid-1990s, management embarked on a
major capital expansion, the construction of a
large new production complex, and initiated a
major re?nancing of the company. Management’s
?nancial projections for the project e?ectively cir-
cumscribed the choices available to the union.
According to the union president, the employees
believed they would inevitably lose control of
the company if the production facility went ahead
(DaPrat, 2005a). The employees therefore entered
into new contractual negotiations with the com-
pany, and converted most of their 60% interest
in the company into improved pensions and other
bene?ts, retaining only 20% ownership (DaPrat,
2005a). Management then issued $500 million of
?rst mortgage notes to acquire funds for the con-
struction project (Algoma Steel, 1996, p. 4; 1997,
Note 7; 2000, Note 6). The holders of these notes
became one of the in?uential stakeholder groups
in the 2001 bankruptcy negotiations (Keenan,
2001).
This new capital structure generated impor-
tant consequences for the company. For exam-
ple, while ?xed assets rose from $323 million
in 1994 to $932 million in 1997, denoting the
construction of the new complex (Algoma Steel,
1996, p. 7; 1997, p. 24), the conclusion of con-
struction meant that expenses the company
had been capitalizing now hit the income state-
ment. The new ?xed assets began to be amor-
tized, including the interest on the ?rst
mortgage notes capitalized during construction
(Algoma Steel, 2001a, p. 26). In addition, a
weakening Canadian dollar created foreign
exchange losses for the company, as the ?rst
mortgage notes were denominated in US dollars.
The company reported large losses beginning in
1998. By 2000, the company had run out of
cash and was staying a?oat through short-term
borrowing. The company blamed its situation
on declining sales resulting from oversupply, a
reduction in demand, and the dumping of for-
eign steel onto the Canadian market (Algoma
Steel, 2000, p. 2).
10
The settlement converted common shares held by Dofasco
and the bonds held by most creditors into preferred shares.
(Dofasco, a major competitor in the Ontario steel industry, was
at this time majority owner of Algoma.) In the process,
Dofasco’s claim on Algoma dropped from $224 million to $69
million. O?setting this, Dofasco picked up a useful $150 million
tax loss and was exempted from pension liabilities. The
employees’ shares carried special rights to approve any funda-
mental changes to the company, and employees could elect a
minority (5 of 13) of the company’s directors. The provincial
government provided $110 million in loan guarantees, reduced
railway freight rates for the company, and waived environmen-
tal liability for the creditors and preferred shareholders.
764 C. Graham/ Accounting, Organizations and Society 33 (2008) 756–782
The 2001 pension crisis at Algoma Steel
On April 23, 2001, the company sought court
protection. In an a?davit ?led by the company
on this date (Ontario Superior Court, 2001a), the
Chief Financial O?cer blamed the company’s pre-
dicament on the dumping of foreign steel, as
before, but now also blamed Algoma’s ?nancial
expenses, and the lengthy implementation of its
new production complex (p. 4). The CFO o?ered
many accounting representations as evidence of
the company’s plight. He provided as exhibits the
unaudited ?nancial statements as at March 31,
2001, as well as a copy of the company’s December
31, 2000 annual report. In the body of his a?davit,
he drew attention to the company’s net losses for
these two dates, as well as its total assets and prin-
cipal debts for March 31, 2001. Included in the
principal debts were the company’s credit facility,
the ?rst mortgage notes, its trade payables, wages
and deductions payable, accrued vacation pay,
and various pension debts and post-retirement
bene?ts obligations. These debts are shown in
Table 1, taken verbatim from the a?davit. What
is interesting is that the a?davit lists all the
accounting numbers except the pension-related
ones as they appear on the ?nancial statements,
that is, under a going-concern assumption. The
pension liabilities, however, are shown on a
wind-up basis.
The impact of this di?erence in treatment is dra-
matic. Pension debts total $769 million, instead of
the $428.4 million shown in the March 31, 2001
?nancial statements (Ontario Superior Court,
2001a, Exhibit A). They amount to approximately
half the listed ‘‘principal debts”. Although most of
these numbers were initially reported in the press
(e.g., Van Alphen, 2001a), in later media reports
it was the pension shortfall that was highlighted
(e.g., Algoma Steel’s $625-M Pension Shortfall,
2001; Daw, 2001; Van Alphen, 2001b). In part,
this was due to the immense size of the pension
number relative to Algoma’s other accounting
numbers, such as equity ($236 million as at the
2000 year end) and even annual revenue ($1106
million).
There are several other possible reasons why so
much attention was given to the pension aspect of
Algoma’s situation. Elderly pensioners and older
workers were perceived as vulnerable. Algoma’s
average worker was around 50 years of age, with
upwards of 25 years of service; approximately 600
of themwere due or able to retire in the next 3 years
(DaPrat, 2005b). To a certain extent, moves to ‘‘res-
cue” workers and pensioners would be received pos-
itively (Steelworkers Hail Appointment, 2001). In
contrast, a direct intervention on behalf of the com-
pany as a whole would have been interpreted as tak-
ing sides against competitors who were equally
important to the Ontario economy.
However, this does not fully explain the pension
?xation of the media. I would argue that the pen-
sion numbers drew attention because they were
easily repositioned and recon?gured in ways that
allowed them to take on new meanings. That is,
even though the pension plans had not changed
materially, the pension disclosures could arbi-
trarily be given new meanings just by rearranging
the signi?ers. This was not as easily done with
the other accounting numbers o?ered by the
CFO in his a?davit.
11
Table 1
Froma?davit of Algoma Steel’s Chief Financial O?cer (Ontario
Superior Court, 2001a, p. 5): Algoma’s principal debts (on an
unconsolidated basis) as at March 31, 2001 are as follows
Cdn. Millions
(a) Credit facility $146
(b) Notes
(i) Principal amount (US $349 million) $551
(ii) Accrued interest $14
(c) Trade payables $41
(d) Wages and employee deductions payable $13
(e) Accrued vacation pay $36
(f) Other
(i) Accrued pension liability on a
wind-up basis (estimated)
$503
(ii) Pension plan indexing obligations
on a wind-up basis (estimated)
$121
(iii) Accrued post-retirement bene?t
obligations (estimated)
$145
11
Pension disclosures are arguably the most easily recon?g-
ured ?nancial accounting signs, but they are not categorically
distinct. The increasing emphasis on fair value in accounting
standards, particularly for ?nancial instruments, suggests that it
will only become easier for managers to rearrange other
signi?ers at will.
C. Graham/ Accounting, Organizations and Society 33 (2008) 756–782 765
The speci?c con?gurations of pension plan sig-
ni?ers that were used will be described below when
we review Canadian pension accounting disclosure
standards. For now, I want to focus on the signi?-
cance of this facility with pension signi?ers. For
this is where Baudrillard’s insight about the arbi-
trariness of signs comes into play. It is not that
the sign (e.g., the stated amount of the company’s
funded future obligation to employees) is arbi-
trarily connected to the referent (e.g., the future
cash ?ows when those obligations are paid).
Rather, it is that the signi?er (e.g., pension liabil-
ity = $X) is arbitrarily connected to the signi?ed
(the notion that the company has a future obliga-
tion to its employees). As the signi?ers are
rearranged, new signi?eds emerge from the rela-
tionships amongst the signi?ers. The notion of
the company’s obligation to its employees, how
large it is, how well or poorly funded it is – these
are all signi?eds that change as the signi?ers are
rearranged.
12
Yet because stakeholders react to
this notion in complex ways, exercising varying
degrees of power at the moment of consuming
the sign and thus shaping both the meaning of
the sign and the outcome of the restructuring pro-
cess, the referent is never independent of the signi-
?ed. The sign precedes the referent, and shapes it.
Hence, as the company changes the signi?er it uses,
the referent is deliberately altered. The company
could not completely control the reactions of the
stakeholders, but it could and did attempt to
impose boundaries on how its obligations could
be understood. Anything vague, anything warm
and fuzzy about its long-term obligations to its
aging workforce, was given precision by the sign
that denoted it. As Baudrillard repeatedly argued
(e.g., Baudrillard, 1981, p. 149), the sign is not
equivocal and ambiguous like a symbol is, it is
univocal; the production of the sign is an attempt
to exclude unwanted interpretations. The struggle
for Algoma Steel’s future hinged on the meaning
of these signs, and control of the ability to produce
them was a crucial advantage for management, an
asymmetry that was institutionalized in the market
for ?nancial information despite the partial owner-
ship position of the workers and the huge claims on
the company by the notesholders.
However, as Baudrillard noted about the com-
modity, it is not the moment of production that
is meaningful but the moment of consumption.
So for the accounting sign. Management con-
trolled the production of the signs, but not their
consumption. Hence, we must look closely at the
mode of consumption of this information: the
moment of consumption of accounting informa-
tion is highly structured, and is simultaneously a
moment of reproduction.
The consumption of accounting signs in the
Algoma Steel ‘‘crisis” involved two important
recon?gurations. First, the signs were discon-
nected from their traditional location in audited
?nancial statements prepared for the ?nancial
markets, and re-embedded in legal documents.
Although the accounting numbers came from the
same accounting systems that produced the
audited ?nancial statements, the company dis-
pensed with the imprimatur of its auditors and dis-
closed unaudited accounting information under
the authority of the legal system, that is, by having
its CFO state the numbers under oath. Note that
this pertains to all the accounting numbers in the
a?davit, not just the pension ones. The e?ective-
ness of this redeployment by the company there-
fore contradicts Macintosh et al. (2000, p. 41),
who maintained that the sign value of accounting
statements about capital and income derives from
their certi?cation by public accounting ?rms. The
Algoma Steel a?davit was not audited. The
accounting numbers it contained functioned as
signs under two related systems of signi?cation.
As accounting signs, they helped present Algoma’s
situation dispassionately as a technical, calculative
problem amenable to technical, calculative solu-
tions. As signs embedded in the legal sign of ‘‘a?-
davit”, they took on an urgency that attracted
both media and government attention. For
instance, they served to reveal to the premier of
12
Note that multiple possible signi?ers interact with multiple
possible signi?eds. There are many ways of signifying the
obligation/liability, and each one has multiple connotations.
This does not change the logic of the sign form (Baudrillard,
1981, pp. 149–150). The production of signs is about the
attempt to impose univocal meanings, not about simplicity.
766 C. Graham/ Accounting, Organizations and Society 33 (2008) 756–782
the province both the necessity and the possibility
of government intervention (cf. Perkel, 2001).
13
The second and simultaneous recon?guration
of Algoma’s accounting signs was particular to
the pension numbers. To function so well as an
alarm, the pension shortfall had to be both a
departure from prior signs produced by the com-
pany, and yet consistent with those prior signs. If
it was not a departure, if it was not ‘‘new”, it could
not have attracted attention. But if it was not
somehow consistent, it would not be believable.
In other words, the system of signs must permit
a new con?guration of signi?ers that is utterable
and understandable, yet previously unspoken.
From an accounting perspective, this means
that the pension numbers released in the a?davit
had to match prior disclosures under Canadian
GAAP, yet Canadian GAAP had to have permit-
ted the new message to have remained unspoken
until this time. To understand how this could be,
it is necessary to review what disclosures Canadian
GAAP requires and permits regarding company
pension plans.
Pension accounting disclosures under Canadian
GAAP
Accounting for pensions in Canadian GAAP is
governed by Section 3461 of the CICA Handbook,
which deals with ‘‘Employee Future Bene?ts”.
Subsection 3461.150 states that the objective of
the given disclosure requirements is to provide
?nancial statement users with information about
the e?ect of employee future bene?ts on the ?nan-
cial statements, and about the obligations and
assets related to de?ned-bene?t plans such as Alg-
oma Steel’s. The subsection also states that the
information is to be useful in understanding the
costs, risks and uncertainties related to the com-
pany’s pension plan obligations. Users should be
able to use the information to make investment
(‘‘resource allocation”) decisions and assess man-
agement’s stewardship. A distinction is made in
the disclosure guidelines between ‘‘pension bene-
?ts” and ‘‘other employee future bene?ts”. Com-
panies in Canada are legally required to manage
these two classes of bene?ts in separate plans,
and Section 3461.151 requires that they be dis-
closed separately.
14
The large shortfall in Algoma’s pension and
post-retirement bene?ts plans developed quickly
during 2000 and 2001. This had been set up by
two decisions by management during the 1990s.
The ?rst was the apparent decision to invest the
company’s pension fund assets in equity markets,
‘‘apparent” because it is di?cult to know from
the accounting disclosures how much of the assets
were invested in this way, and in which speci?c
stocks. The company’s ?nancial statements indi-
cate that the actual return on plan assets plum-
meted in ?scal 2001 from a 5-year average of
$123 million per annum to a negative return of
$18 million.
15
These results were typical of many
North American companies that had heavily
invested their pension assets in equities during
the expansive markets of the late 1990s and were
exposed when the dot-com bubble burst in 2000
(Fore, 2004; Joss, 2004).
16
The relative amount
of Algoma’s investment loss was not huge, only
2% of assets, and was o?set by continued contribu-
tions to the plan. However, the reversal of fortune
in the markets was dramatic, and would have had
13
Furthermore, news articles from this period reinforced the
legitimacy of government intervention by recalling the role of
the Ontario government in resolving the company’s previous
‘‘crisis” in 1991 (e.g. Ottawa Provides Loan Guarantee, 2001;
Perkel, 2001; Van Alphen, 2001; Watson, 2001).
14
Algoma’s a?davit made a further distinction, listing
pension indexing liabilities separately from pension obligations
and other future bene?ts obligations. Indexing provisions of
corporate pension plans were not protected by the Ontario
government. See the section on the Pension Bene?t Guarantee
Fund below.
15
While the equity markets plummeted in 2000, a negative
‘‘actual return on plan assets” ?gure did not appear until
Algoma’s 2001 ?nancial statements. The recognition of the
negative return may have been delayed due to timing of the
asset valuation, although this should not have been the case,
since Section 3461 (Exhibit I of Example 1, Situation I) implies
that the 2000 calculation should have been based on the fair
value of plan assets on Algoma’s year-end date of December 31,
2000. Algoma, unfortunately, did not disclose how it arrived at
its ?gures.
16
Algoma’s key competitors, Stelco and Dofasco, stated
similar changes in return on pension assets in the same period.
Dofasco’s return dropped from $86.5 million in 2000 to $17.9
million in 2001 (Dofasco, 2001). Stelco’s return dropped from
$317 million to $11 million for the same years (Stelco, 2001).
C. Graham/ Accounting, Organizations and Society 33 (2008) 756–782 767
abrupt consequences for pension funding decisions
if the company had not ?led for court protection
(cf. Fore, 2004, p. 7).
The tentativeness of the statements in the pre-
ceding paragraph is quite deliberate. The impossi-
bility of being more conclusive points to a speci?c
problem in Canadian GAAP. The extent to which
plan assets may have been invested in dot-com
stocks, for example, remains unknown because
Canadian GAAP in 2001 did not require such a
disclosure,
17
and Algoma’s management chose
not to exceed GAAP requirements. This left the
accounting signs that were produced bereft of a
meaningful context. Like a word uttered in isola-
tion rather than as part of a sentence, these signs
wanted the context of other signs—such as
asset allocation—that would enrich their meaning.
The second management decision leading to
Algoma’s sudden pension shortfall was the choice
of a high discount rate for future bene?ts during
the 1990s. This policy had to be abandoned when
the company adopted Section 3461. These new
standards came into e?ect on January 1, 2000,
the beginning of Algoma Steel’s ?scal year, and
replaced Section 3460, which had covered pension
costs and obligations.
18
According to Accounting
Standards Board of Canada (2002) and Estey
(1999), the purpose of Section 3461 was to rectify
a perceived de?ciency in pension cost accounting,
and to harmonize this section of the handbook
with US GAAP. The new section changed the dis-
count rate used to evaluate future cash ?ows from
a pension plan. Previously, reporting entities could
use a rate determined by management’s best esti-
mate of the long-term rate of return on the pension
fund’s assets. The new standard required discount-
ing either at the market interest rates on high
quality debt instruments whose cash ?ows approx-
imated the plan’s cash ?ow, or at the rate implicit
in whatever present amount could be paid to settle
the plan’s obligations (Accounting Standards
Board of Canada, 2002; Estey, 1999). The new
standards also speci?ed that an entity should mea-
sure plan assets at fair value; compute the expected
return on plan assets using a long-term rate of
return on the market value of assets; match the
cost of bene?ts to the periods in which the
employee earned them, up to the date at which
the employee became entitled to receive them; rec-
ognize actuarial gains and losses using the ‘‘corri-
dor” approach (discussed below and in
Appendix); and observe new guidance on plan set-
tlements, curtailments, terminations, accrued ben-
e?t assets, and multi-employer plans. Signi?cantly,
the new standards required all these pension
accounting guidelines to be applied to accounting
for non-pension future bene?ts, which had previ-
ously been treated on a cash basis (Estey, 1999).
As a result of adopting these standards, in 2000
and 2001 the Algoma Steel plan endured dramatic
unrecognized actuarial losses, in stark contrast to
its prior pattern of unrecognized actuarial gains.
The glossary of Section 3461 de?nes actuarial
gains and losses in a de?ned-bene?t plan as
changes in the value of the accrued bene?t obliga-
tion and plan assets resulting either from experi-
ence being di?erent than assumed or from
changes in actuarial assumptions. In the case of
Algoma Steel, the key assumption that changed
was the discount rate selected by management to
calculate the present value of the company’s future
bene?t obligations. In adapting to the new GAAP
guidelines, management shifted these rates from
the steady 8% it had used since the 1991 restructur-
ing, to 7% in 2000 and 6.5% in 2001.
19
This was
precisely the e?ect of Section 3461 that had been
predicted in the Canadian accounting profession’s
17
Even now, Section 3461.155(b)(ii) requires only that a
company disclose its asset mix in the broadest terms, by
showing the percentage of assets held in general categories such
as equity securities, debt securities and real estate. More
detailed information is contained in actuarial valuation reports
?led with federal and provincial pension authorities. These are
available to plan members upon request, but are not available
to the general public.
18
Comments in this paper regarding Section 3461 are not
intended to suggest that Section 3461 is either ‘‘better” or
‘‘worse” than Section 3460. By o?ering speci?c critiques of
Section 3461, I do not mean to suggest that the social impacts
of pension accounting in the Algoma Steel case would not have
occurred, or would have been more felicitous, under Section
3460. They would simply have been di?erent. As Macintosh
et al. (2000, p. 16) note, Baudrillard’s ‘‘orders of simulacra” do
not imply sequential improvement.
19
These were the rates used for the pension obligation. For the
much smaller non-pension future bene?ts obligation, the rates
were 7% for 2000 and 6.75% for 2001.
768 C. Graham/ Accounting, Organizations and Society 33 (2008) 756–782
leading practitioner magazine (Tishco?, 1999).
These seemingly small adjustments in discount rate
caused unrecognized actuarial losses on non-pen-
sion future bene?ts to swell from a 5-year average
of $14 million to $75 million in 2001. Even more
dramatic was the e?ect on the pension side, which
had enjoyed unrecognized actuarial gains averag-
ing $115 million per year from 1994 to 1999. These
gains disappeared in 2000, and became an unrec-
ognized actuarial loss of $267 million in 2001.
While random ?uctuations in mortality or morbid-
ity experience may have contributed to this nega-
tive result, a shift of this magnitude was more
likely caused by the change in discount rates; pro-
jected bene?ts calculations have been shown to be
highly sensitive to changes in discount rate (Duf-
resne, 1993).
20
Again, it is impossible to be certain
here: Algoma did not disclose any details about its
actuarial losses, nor did Section 3461 require it to
do so.
21
Actuarial calculations underlying the pension
accounting disclosures informed the funding deci-
sions that produced eventual funding shortfalls.
These endogenous e?ects can be seen in Fig. 1.
In this chart, each column indicates the total fund-
ing shortfall for pensions and other post-retire-
ment bene?ts in a given year. This shortfall is
called ‘‘Funded status” in the notes to the ?nancial
statements. However, in Algoma’s notes (follow-
ing CICA Handbook guidelines) the amount for
pensions and the amount for other post-retirement
bene?ts are never added together as they are here.
Each column in the chart is composed of several
smaller amounts, including the on-balance sheet
‘‘accrued liability” and the o?-balance sheet
‘‘unrecognized actuarial gains and losses”. The
chart demonstrates that prior to 2000, actuarial
gains o?set the accrued liability substantially;
thereafter actuarial losses exacerbated the liability.
During this time period Algoma Steel’s man-
agement adopted a curious format for disclosing
the pension plan’s funding status. The CICA
Handbook’s illustrative examples for such disclo-
sures are fairly straightforward. Nonetheless,
Algoma chose a more convoluted presentation.
Table 2 shows the funding status disclosures con-
tained in Note 10 to Algoma’s 2001 ?nancial
statements.
Rather than breaking the plan’s funding status
into its components, the Algoma disclosures tell
the story of how the accrued liability portion of the
pension plan obligation is calculated. They begin
with large negative numbers and o?set them with
a series of mainly positive numbers, to arrive at a
smaller negative number. The implication is that
the reader should focus on this smaller number.
The disclosure is arrangedtonot toexplainthe over-
all obligation, but to explain it away. The reader is
guided to pay attention only to the portion of the
pension obligation that has been recognized as a lia-
bility by the company, the portion that appears on
the balance sheet. The overall pension obligation
appears less signi?cant—has a diminished sign
value—and diminished social consequences. This
reinforces the prior e?ect of relegating the pension
obligation disclosure to the notes.
In contrast, the examples given in Section 3461
show the funding status as the di?erence between
plan obligations and plan assets. If this format
had been followedby Algoma, the notes wouldhave
included a table such as Table 3. Note that in Table
3, it is clear that the funding status is the amount
being explained. That is, the plan de?cit is ‘‘the bot-
tom line”. Yet even in this format recommended by
the Handbook, the total obligation for pension and
non-pension bene?ts is never stated.
20
Algoma was not alone in its experience with the new
standards. One of Algoma’s key competitors, Stelco, dropped
its discount rate from 8% in 1998 and 1999 to 7% in 2000, and
to 6.75% in 2001; its actuarial gain of $15 million on all de?ned-
bene?t plans in 2000 changed to a loss of $181 million in 2001
(Stelco, 1999, 2001). Algoma’s other key competitor, Dofasco,
actually raised its discount rate from 6.75% in 2000 to 7% in
2001; its actuarial losses on all de?ned-bene?t plans dropped
from $72.8 million in 2000 to $34.9 million in 2001 (Dofasco,
2001). The widespread uncertainty caused by the changes in
discount rate mandated by Section 3461 prompted the AcSB
(2002) to issue a statement on ‘‘Measurement Uncertainty and
Employee Bene?t Plans”.
21
The impact of the adoption of Section 3461 by Algoma was
clouded by the simultaneous adoption of other new accounting
standards. The e?ect of Section 3461 on Algoma’s bottom line
was a decrease in retained earnings of $40 million. However the
concurrent adoption of new income tax accounting standards
simultaneously increased retained earnings by $81 million
(Algoma Steel, 2000, p. 22).
C. Graham/ Accounting, Organizations and Society 33 (2008) 756–782 769
Algoma Steel’s disclosure thus appears designed
to downplay the company’s obligation for
employee future bene?ts, focussing attention
instead on the accrued bene?t liability. Yet the
accrued bene?t liability is the most heavily ‘‘con-
structed” number in Section 3461’s implicit model.
It is the one number in the model that is least
directly representative of any underlying objective
reality that the model might purport to represent.
(Readers unfamiliar with Section 3461 are referred
to Appendix.) The model stipulates that the
accrued bene?t liability is the di?erence between
the accrued bene?t obligation and the aggregate
of the plan assets and any unamortized costs.
Thus, any unamortized costs are e?ectively
deducted from the funding status to determine
the accrued liability shown on the balance sheet.
The implicit model therefore constrains the entity
from fully disclosing the funding status of its pen-
sion plan in audited statements.
E?ects of the model in the Algoma Steel pension
crisis
The express intent of the model in slowly
amortizing pension costs is to dampen the random
-$200
$0
$200
$400
$600
$800
$1.000
1997 1998 1999 2000 2001
Total accrued retirement obligation
Contributions and payments after measurement
date
Unrecognized prior service costs
Unrecognized net retirement obligation
Unrecognized actuarial losses (gains)
Fig. 1. E?ect of actuarial gains and losses on Algoma Steel pension plan obligation. Source: Algoma Steel Inc. annual reports, 1997–
2001.
Table 2
Disclosure of funded status
Reconciliation of pension funding status 2001 2000
Funded status (552) (287)
Unrecognized actuarial losses (gains) 267 3
Employer contributions
after measurement date
3 3
Unrecognized net pension obligation (2) (3)
Unrecognized prior service costs 6 7
Accrued pension liability (278) (277)
Reconciliation of post-employment
funding status
2001 2000
Funded status (245) (196)
Unrecognized prior service costs 0 0
Unrecognized actuarial losses (gains) 75 32
Bene?ts paid after measurement date 1 1
Unrecognized net bene?t obligation 12 12
Accrued post-employment
bene?t obligation [sic]
(157) (151)
Source: Algoma Steel Inc., 2001 ?nancial statements, Note 10.
770 C. Graham/ Accounting, Organizations and Society 33 (2008) 756–782
short-termnoise that is inherent in the mortality and
morbidity experience of the plan, and in the market
price of the ?nancial assets of the plan. Tothis extent,
the model is arguably useful. Short-term noise, the
argument runs, can have an unwarranted impact
on the funding status of the plan, and since the
expected value of the noise is zero over time, the
model should smooth out these short-term e?ects.
However, the unwarranted impact of certain
‘‘noise-related” parameters arises from the model
itself. Market-based parameters like ?nancial securi-
ties prices and the discount rates mandated by Sec-
tion 3461 a?ect the two main aggregate values in
the model, which are the fair value of plan assets
and the accrued bene?t obligation. Funding status
is simply the di?erence between these two values. If
these large values are ‘‘normally” roughlyequal, then
a relatively small percentage change in one of them
can cause the ‘‘normally” small di?erence to grow
by many multiples of itself. Yet changes to these
main aggregate values are not always small, being
so sensitive to some of the model’s parameters. In a
sense, therefore, by implementing extensive smooth-
ing, the model is protecting its users from itself.
The argument in favour of smoothing, as frail
as it is when parameters take random walks, is par-
ticularly ?awed when it comes to non-random ?uc-
tuations in plan parameters. In the case at hand,
the change in discount rates mandated by Section
3461 was singular and non-random. The impact of
the adoption of Section 3461 at Algoma Steel,
given management’s choice in prior years of high
discount rates that were no longer permissible,
was to drive up the pension obligation. This hap-
pened at the very time plan assets were falling in
value. Yet because the net change in these numbers
was classi?ed under the model as an actuarial loss,
it was metered out to the balance sheet only
through the corridor rule, in tiny increments and
a year delayed. Thus, a mechanism designed to
smooth random ?uctuations was being used to
cope simultaneously with a permanent restatement
of plan assets and the structural adjustment of the
market to plummeting dot-com valuations. Subse-
quent random ?uctuations in experience and the
market cannot be counted upon to undo this.
The signi?cant changes in Algoma’s pension
obligation that arose in 2001 resulted in two di?er-
ent sets of accounting signs being produced. When
the company ?led for bankruptcy, one set of signs
was produced that highlighted the pension funding
shortfall. At the year end, after the ‘‘crisis” had
been resolved, a second set of signs was produced
that covered the same period of time but presented
a much di?erent picture. In the second set of signs,
contained in the company’s annual report, the over-
all pension plan obligation was relegated to the
notes of the ?nancial statements, as mandated by
the CICA guidelines. Thus, Algoma’s management
disregarded the demand for accounting informa-
tion that was evident in the way the media took
up the pension obligation ?gure when the crisis ?rst
broke.
22
This illustrates how the production of
accounting signs is dependent on the conditions of
their consumption. At the ?scal year end, Algoma
management no longer addressed the court, from
whom it had sought protection, but the share-
holder, from whom it sought con?dence. The Man-
agement Discussion and Analysis in the annual
report was thus well within the requirements of
Canadian GAAP, but entirely unresponsive to
other stakeholder needs, when it failed to mention
the enormous actuarial losses. In fact, the MD&A
stated only that during the year the accrued pension
liability had risen from$277 million to $278 million.
The implicit model of Section 3461 accom-
plished two contradictory things for Algoma Steel.
First, it constrained evidence of volatility for years,
only to release it in a torrent when discount rate
regulations changed. Second, it was delivered into
the hands of management two ready sets of ?gures,
one for minimal recognized portions of future lia-
22
Congruence between media attention and the interests of
other stakeholders, such as investors, should not be taken for
granted.
Table 3
Format for disclosure of funded status suggested by Section
3461 of CICA handbook
Pension
bene?ts
Other
bene?ts
2001 2000 2001 2000
Fair value of plan assets 1018 1101 11 10
Accrued bene?t obligation 1570 1388 256 206
Funded status – plan de?cit (552) (287) (245) (196)
C. Graham/ Accounting, Organizations and Society 33 (2008) 756–782 771
bilities and another for their greater unrecognized
whole. All the accounting signs in the a?davit were
taken straight from Algoma’s accounting system,
but assets and non-pension liabilities had only
going-concern estimates associated with them.
The pension numbers, in contrast, were easily
turned on their heads to show the (un)funded sta-
tus of the plans. These numbers could be said to
represent the wind-up value of the pension plans
without further work being necessary.
Structured consumption of signs and the resolution
of the crisis
Algoma Steel’s disclosure decisions and the
underlying calculative techniques of Section 3461
must be evaluated in terms of the network of social
relations in which Algoma Steel was embedded.
Certainly, the market for Algoma Steel’s securities
was impacted by the calculations, at least to the
extent that the calculations were inseparable from
the 2001 restructuring process. According to Tor-
onto Stock Exchange archival records (available
on MSN.com), the price of Algoma shares fell from
$0.40 to as low as $0.22 on April 23, 2001, when the
company ?led for court protection. After a brief
rebound, prices declined from $0.35 in late April
to $0.25 in mid-October, and plummeted to the
$0.10 range after the restructuring plan was
announced on October 24, 2001. However, the mar-
ket was but one of the social relations pertaining to
Algoma Steel. Algoma’s pension accounting also
a?ected relations between Algoma and its workers,
and between both these parties and the government.
With the workers, Algoma was engaged in a
rather circular exercise involving capital forma-
tion. As discussed above, when the 1991 restruc-
turing took place, the workers gained control of
60% of the company. While this stake was reduced
to 20% in 1994–1995, the relevant governance
structures remained unchanged. Management
was considered independent, and the workers were
both owners and employees. The relationship
between management and workers-as-owners was
mediated by the board of directors and the trust
that held the employee shares. The relationship
between management and workers-as-employees
was mediated by the union, but only partially,
because not all employees were union members.
Because of the circularity of the company–
worker relationship, it is inappropriate to draw
on facile explanations of the disclosure choices
made by management regarding the pension plan.
Since workers had representatives on the board,
one would have expected them to have full access
to information. However, management did not dis-
close all that it could, nor indeed all that it was
required to under the company’s articles of incor-
poration. Under cross-examination during the
court proceedings, the company’s ‘‘Chief Restruc-
turing O?cer” said that despite the articles specify-
ing that the union’s representative was to receive all
Board material, the company had not divulged any
‘‘con?dential” information to the union regarding
the restructuring. The workers were not given any
information about the restructuring that was not
shared equally with other stakeholders (Ontario
Superior Court, 2001b, pp. 6–8). In understanding
the situation they were in, the workers were depen-
dent on management and upon the expertise of the
accountants hired by management. The result was
a passive acceptance of the company’s representa-
tions: when asked why the union was not more
vocal during the events of 2001, a present union
leader said, ‘‘Well, management told us the com-
pany was going bankrupt” (DaPrat, 2005a).
Algoma’s adoption of the new pension account-
ing guidelines in 2000 triggered a change in its rela-
tionship with the provincial government. With the
removal of the latitude to nominate a discount rate,
management was unable to continue constructing
its preferred picture of a healthy pension plan.
When plan assets failed to deliver the large returns
of the 1990s, and the pension obligation ballooned
due to the lower discount rate, the pension account-
ing model e?ectively constructed a new reality for
the company. Algoma was now a company with a
severely underfunded pension plan. This was the
handhold the provincial government needed in
order to intervene in the company once more.
The pension bene?ts guarantee fund
What made it feasible for the government of
Ontario to intervene was its possession of the
772 C. Graham/ Accounting, Organizations and Society 33 (2008) 756–782
requisite technologies for intervention. In 1990, the
provincial Pension Bene?ts Act had established in
Ontario the Pension Bene?ts Guarantee Fund
(PBGF) to protect workers from the risks of
private company pension plans. This fund covered
pension bene?ts, including worker pensions, spou-
sal pensions, and plant-closure bene?ts, but did not
cover non-pension post-retirement bene?ts such as
health care. The fund was ?nanced by annual con-
tributions from employers with pension plans. By
2001, however, the fund had accumulated only
$200–240 million in assets, less than half the
amount of Algoma Steel’s pension de?cit (Daw,
2001; Keenan, 2001; Province of Ontario, 2001).
The fact that the PBGF was nevertheless used
in the Algoma Steel bailout suggests that when it
comes to e?ecting the goals of government, it is
not necessarily the amount of funding that mat-
ters, but the existence and availability of appropri-
ate technologies. Embedded in the PBGF were a
set of accounting technologies that enabled the
government to act. Using these technologies, the
Financial Services Commission of Ontario, which
administers the fund, was able to collect, track,
and enforce contributions from employers. The
funds collected were aggregated and invested.
PBGF assets were invested in cash equivalents
and bonds (Province of Ontario, 2001), not in
stocks and other riskier securities. The fund was
designed to be supplemented by loans from the
provincial government when necessary. The PBGF
received pension plan assets in exchange for
absolving companies of a limited portion of their
pension obligations. The liability assumed by the
PBGF was for the amount of pensions earned by
Ontario-based employees, excluding most escalat-
ing or indexing provisions.
23
Under the terms of the Pension Bene?ts Act,
plans with over $500 million in assets were able
to ?le for exemption from certain funding solvency
regulations (Pension Bene?ts Act, 1990, s. 5 & s.
37). This privilege was withdrawn (cf. Reg. 203/
02, 2002, s. 1) shortly after the Algoma plan bail-
out, but Algoma had quali?ed for this exemption
and hence had not been observing the funding sol-
vency regulations. Nonetheless, the Algoma pen-
sion plan was covered by the PBGF.
The existence of the PBGF was crucial to the
government’s ability to consume the accounting
signs that had raised the alarm about the Algoma
Steel pension plan. While the meaning of any sign
depends on the conditions of its consumption, for
accounting signs those conditions are highly tech-
nical. This distinguishes accounting from other
languages. The PBGF, as noted, was not in a posi-
tion to accept the entire pension obligation of Alg-
oma. However, under the Pension Bene?ts Act and
the terms negotiated in the 2001 settlement, the
PBGF assumed not only most of the obligation,
but most of the plan assets. Hence, the net new lia-
bility taken on by province of Ontario, backed by
the assets of the PBGF and any loans to the PBGF
that would be necessary, was limited to the fund-
ing de?cit, and was further reduced by the deletion
of indexing and other bene?ts under the negotiated
settlement. Given that the existing plan assets
would, in the event of Algoma Steel’s complete
failure, have gone to settle at least a portion of
the pension obligation, the intervention of the gov-
ernment through the PBGF must be evaluated on
the basis of the incremental bene?t provided to the
pensioners, plus the bene?ts realized by other
stakeholders.
Terms of the 2001 plan of arrangement
The incremental bene?t provided to workers
and pensioners was mixed. While the pension plan
was saved from default, the plan assets would,
without government intervention, have provided
65% of the accrued pension obligation at the end
of 2001 (56% of the total future bene?ts obliga-
tion).
24
Under the terms of the bailout, the existing
pensions of retirees were protected, but as noted
23
The PBGF liabilities excluded the following: ‘‘(a) any
escalated adjustment, (b) excluded plant-closure bene?ts, (c)
excluded permanent layo? bene?ts, (d) special allowances other
than funded special allowances, (e) consent bene?ts other than
funded consent bene?ts, (f) prospective bene?t increases, (g)
potential early retirement window bene?t values, and (h)
pension bene?ts and ancillary bene?ts payable under a quali-
fying annuity contract” (Pension Bene?ts Act, 1990, s. 2).
24
At the beginning of 2001, plan assets would have provided
even more: 79% of the accrued pension obligation (70% of the
total future bene?ts obligation). These calculations are derived
from Algoma’s 2001 annual report.
C. Graham/ Accounting, Organizations and Society 33 (2008) 756–782 773
the Pension Bene?ts Act did not protect any index-
ing of these pensions. Bene?ts not guaranteed by
the PBGF were moved over to a new supplemen-
tary plan to be funded by Algoma Steel, with
whatever risks that entailed.
The technical capacities of the PBGF a?ected
the provincial government’s ability to respond to
the Algoma Steel crisis, and were an important
factor in how Algoma’s accounting signs played
out in court. However, the consumption of
accounting signs in the legal proceedings was not
shaped merely by technical matters. The terms of
the Plan of Arrangement were negotiated by the
company and its stakeholders through a long
and arduous process. The court documents reveal
that the settlement was a?ected strongly by the rel-
ative power that the stakeholders were able to
exert on the negotiations. While the provincial
government had the technical capacity to consume
certain accounting signs, other stakeholders lack-
ing such elaborate technical capacities also in?u-
enced the negotiations. Each stakeholder was
framed and denominated by its accounting status
as a creditor.
25
Creditors were grouped into clas-
ses, such as unsecured creditors who had the
option of taking either a $2500 ?at settlement or
a prorated share of a ?xed amount of money set
aside for this class. The power of stakeholders
was not a function of the class’s total claim or
the individual stakeholder’s claim, but of the dis-
tribution of the claims within the class. The City
of Sault Ste. Marie was in a class on its own. Pen-
sioners were divided into two classes, according to
their indexed and non-indexed bene?ts claims; the
large total bene?ts claim being well distributed
amongst many pensioners, these people exercised
very little individual power. The notesholders were
di?erent from the pensioners in that the distribu-
tion of notes was uneven, a minority of noteshold-
ers holding a majority of the dollar amount of the
notes.
In the ?nal terms of arrangement, listed in
Table 4, the notesholders did well. They came
away with 75% ownership of the company plus a
considerable fraction of the original debt still being
owed to them in new notes. The notesholders were
able to gain this result because they were able to
defeat earlier drafts of the plan. The plan had to
be approved by a double majority of each class
of creditor (i.e., by a majority of the creditors rep-
resenting a majority of the dollar value). The ?rst
draft of the plan of arrangement called for a 9%
interest rate on the bulk of the new notes. This
was raised to 9.5% in the second draft of the agree-
ment, which was approved by everyone except
25
Because the proceedings took place under Ontario’s Com-
panies’ Creditors Arrangement Act, stakeholders lacking the
status of creditor were excluded from the negotiations. It is
interesting to see legislation formulated speci?cally around
accounting constructs.
Table 4
Final terms of arrangement
Terms of the plan of arrangement
All existing voting shares were deleted, including the employees’ 20% ownership stake
Employees received 20% of the new common shares, which were no longer held in trust and had no special rights
Interim ?nancing provided by banks during Algoma’s court protection was entirely secured
Notes holders received, pro rata, new notes totalling US$187.5 million, plus 75% of new common shares. (Existing notes had
amounted to US$349 million, plus US$47 million in accrued interest)
Pension claimants received a replacement pension plan
Unsecured claimants received, pro rata, 5% of the new common shares, or could elect to receive up to $2,500 to settle their claim
The City of Sault Ste. Marie received $5 million in future payments, secured by Algoma’s real property
The federal and provincial governments received all income taxes owing
A new collective bargaining agreement was enforced, featuring $153 million wage and bene?t reductions, reduced vacations, reduced
pension bene?ts, and job cuts
Pension payments to existing Ontario pensioners were taken over by the PBGF, along with related plan assets. Pension payments to
non-Ontario pensioners remained the obligation of Algoma Steel, to be paid from general revenues or a new pension plan
Pensions of current workers were secured by a claim on the remaining plan assets, and also by a $100 million claim on the ?xed assets
of the company, subordinate to the claims of the banks and notes holders
Source: (Algoma Steel, 2001b, p. 13; 2002, p. 30).
774 C. Graham/ Accounting, Organizations and Society 33 (2008) 756–782
24% of the notesholders. Because they held 84% of
the dollar value of the notes, they were able, at the
last minute, to prevent the second draft from being
approved. This triggered emergency meetings
between management and the stakeholders to res-
cue the plan (Ontario Superior Court, 2001d, p. 5).
The ?nal version of the plan of arrangement shows
an 11% interest rate on the new notes. This is
another indication of how the consumption of
accounting signs is structured: because the pension
obligation was widely dispersed amongst the pen-
sioners, the larger notesholders were more easily
able to a?ect how the accounting information
would be interpreted and translated into stakes
in the restructured company. The accounting signs
that denominated the creditors’ claims on the com-
pany presumed to impose a speci?c valuation on
each claim. Yet the meaning of these signs is deter-
mined at the time of their consumption during the
legal procedures and negotiations. The meaning is
shaped by legal features such as the de?nitions and
procedures of the Companies’ Creditors Arrange-
ment Act, by technical features such as the PBGF
and Section 3461 of the CICA Handbook, by the
distribution of power amongst the creditors and
by their exercise of power.
Denouement
Following the 2001 Plan of Arrangement, the
company enjoyed a resurgence in both sales and
stock price. Investors received a good return,
employees continued to work, and retirees contin-
ued to receive their pensions. But the complexities
of the resolution prevent an unequivocal judge-
ment of its success, even with the bene?t of hind-
sight. Employee wages were lower than before,
and retirees’ pensions were no longer indexed. In
addition, the resolution to Algoma’s 2001 pension
crisis may well have damaged the relationship
between employees and retirees. According to the
present union leader (DaPrat, 2005a), in 2001 the
retirees felt they had no choice but to go along
with the agreement being negotiated amongst the
stakeholders. They felt they would gain no advan-
tage by speaking out against the company’s posi-
tion. According to the union leader, while the
union’s primary goal during the 2001 negotiations
was to protect retirees’ existing pension bene?ts,
the retirees ended up angry about the settlement
because they lost indexing. Their anger was exacer-
bated in 2004 when Algoma returned to prosperity
and the current workers received a $10,000
advance on pro?t sharing as part of a new con-
tract.
26
The retirees received no such bonus, yet
they felt they had sacri?ced in 2001 in order to
help turn the company around (DaPrat, 2005a).
Fearful asymmetry
The transplantation of accounting signs from
Algoma Steel’s ?nancial statements to the legal
context of the a?davit demonstrates how the
meaning of accounting signs is a?ected by their
mode of reproduction. The subsequent struggle
between the stakeholders demonstrates how the
meaning of accounting signs is a?ected by their
mode of consumption. Reproduced in their new
legal setting, all the accounting signs took on an
urgency that transcended audited statements.
These signs were an attempt by management to
assert both the need for court protection and the
distribution of liability. The disclosures made by
Algoma Steel in the a?davit were simply insu?-
cient for anyone to see how some crucial numbers
were derived, a situation engendered by informa-
tion asymmetries, the opacity of Section 3461,
and the arrayed expertise of Algoma’s actuaries
and accountants. Yet despite these advantages,
management was unable to impose an undisputed
meaning on their accounting signs.
The most important accounting signs in the
legal process were those that denominated the
creditors. The legal status of participants was not
independent of these signs. Each claim on Algoma
Steel was initially predicated on Algoma Steel’s
accounting calculation of the amount owing. To
dispute this would entail producing an invoice or
other accounting records that could signify a
greater claim. For tradespeople, this might be rel-
atively simple, supposing for instance that Algoma
26
The union leader who provided the information did not
state whether this was an average amount or a ?at amount per
worker.
C. Graham/ Accounting, Organizations and Society 33 (2008) 756–782 775
had failed to process an invoice correctly. For pen-
sioners, however, disputing the stated pension lia-
bility would be extremely di?cult. It would involve
a calculation so complex as to be beyond the aver-
age individual’s capability to perform, and even
beyond the pensioners’ collective ability. They
were, for all practical purposes, completely at the
mercy of the company’s calculations.
However, there is no indication in the court doc-
uments or the media coverage that Algoma’s
accounting calculations were themselves conten-
tious. The contest came after the accounting signs
were accepted under oath, and reproduced into
the court proceedings. The lengthy negotiations
from April until December of 2001 between man-
agement and the creditors, which also involved
the union making wage concessions, was a process
of giving the accounting signs a precise legal mean-
ing, of translating them into new legal and ?nancial
arrangements. In this direct sense, the accounting
signs preceded the reality they seemed to describe.
The signs described a set of claims, yet the claims
had no e?ect until the signs were consumed and
rendered meaningful. The signs suggested an equiv-
alence – that the pensioners’ claimon the company,
for instance, was roughly equivalent to the notes-
holders’ claim on the company ($503 million com-
pared to $551 million) – yet each class of creditors
required the accounting signs to translate into a
unique new set of ?nancial relationships. For
example, in due course the sign ‘‘Notes pay-
able = $551 million” came to mean 75% of the
new shares and an entire new set of notes payable.
This is not obvious or inherent in the sign at the
moment of reproduction. It is the result of contes-
tation during an extended process of consumption.
The legal status of the creditors, established
through accounting records, gave each class of
participants the same opportunity to in?uence
the outcome. However, the ability to operate the
system of calculations and the system of legal pro-
cedures was not evenly distributed amongst these
participants. The pensioners had no collective
voice; a common legal advisor had to be provided
by the courts (Ontario Superior Court, 2001c, p.
151). As noted, because the pensions were fairly
evenly distributed amongst the pensioners, there
was no small group of them that could hold up
the restructuring proceedings. Unlike the notes-
holders, the pensioners were passive consumers
of Algoma Steel’s accounting signs, and hence suf-
fered more severely the consequences of the asym-
metries of the mode of reproduction.
These information asymmetries would have
been worse had they not been partially addressed
27
by the new discount rates mandated in Section
3461. The guidelines for discount rates in Section
3461 take some of the previous latitude out of
the hands of management, making the accounting
somewhat more transparent. However, there is a
cost to this. The new guidelines stipulate that the
discount rates should now follow the market.
Thus, an opaque but relatively stable discount rate
is replaced by a rate that is arguably more trans-
parent, but also more volatile. The risks of the
market replace moral hazard.
The opacity of Section 3461 might suggest that
the accounting profession plays a hegemonic role
in pension accounting in Canada, that somehow
they control ‘‘the code” that for Baudrillard sym-
bolized the programmatic way that signs take on
meaning in our society. However, other features
of the Algoma Steel narrative contradict this. Alg-
oma management was able to hire accounting
expertise to exploit its information asymmetries,
putting the accounting profession in a subservient
role. Also, the accounting signs in the company’s
2001 a?davit that triggered governmental actions
lacked the accounting profession’s imprimatur.
27
Asymmetries are still evident in the impossibility of recon-
structing certain intermediate results in the present study. For
example, the discount rates disclosed in the 2001 Algoma
statements seem wrong. If the disclosed rate of 6.5% is used, the
pension-related portion of unamortized net actuarial loss at
year end works out to $250 million. Since the funding de?cit of
the pension plan was $552 million, a di?erence of $296 million
would be computed for the accrued pension liability. However,
the accrued pension liability was only $278 million. This is the
amount that would be computed with an 8.1% rate. While the
working papers for the disclosure would likely be more complex
than the general model based on Section 3461 that is described
here, the discrepancy is nonetheless irresolvable given Algoma’s
existing disclosures. Thus the complexity of Section 3461 turns
disclosures of pension liabilities into ‘‘black boxes” impenetra-
ble to ?nancial statement users, and renders the expertise of
professional accountants indispensable to managers of compa-
nies that must submit to the requirements of the CICA
Handbook.
776 C. Graham/ Accounting, Organizations and Society 33 (2008) 756–782
The Algoma Steel events, therefore, would call
into question the way that Macintosh et al.
(2000) privilege the accounting profession in their
discussion of accounting as simulacrum and
hyperreality. They argue that the sign value of an
accounting sign derives from its authentication
by an accountant: certi?ed signs impart ‘‘a sense
of exogeneity and reliability to society at large”
(Macintosh et al., 2000, p. 41). It is unclear, how-
ever, what system boundaries Macintosh et al. are
assuming when they refer to exogeneity. The
accounting profession is no more exogenous to
the social systems that produce accounting signs
than booksellers or newspaper editors are to the
social systems that produce vernacular language
signs. The insistence on the exogeneity of profes-
sional certi?cation unnecessarily undermines the
Baudrillardian perspective Macintosh et al. pro-
vide on accounting. For consistency with Baudril-
lard, I would argue that an accounting sign takes
its sign value from its position within the system
of signs in which it appears, not from pseudo-exo-
geneity. This system of signs, however, must be
understood more comprehensively than Macin-
tosh et al. suggested. The system context of
accounting signs is not limited to clean surplus
models or pension accounting models. The pen-
sion numbers in the a?davit, for example, took
their sign value not only from their relation to
other accounting signs, but also from the legal
signs in which they were embedded and indeed
from the credibility of Algoma Steel management
itself. This sign value was ampli?ed when the
media picked up the numbers and repeated them
widely.
If certi?ed accounting signs provide anything
to social and commercial discourse, it is not exo-
geneity but relative stability. This is because cer-
ti?ed accounting signs are produced, at least
nominally, according to standards that take con-
siderable time and e?ort to change. While the
standards setting process can be in?uenced by
those with the power to do so, the process is
institutionalized to the degree that it is largely
conservative and resistant to change. This results
in accounting standards that change much less
quickly than accounting signs can be changed,
making certi?ed signs relatively stable against
the backdrop of economic commotion and
babble.
This does not mean that accounting signs are
not arbitrary. But it is important to clarify the
notion of arbitrariness used by Baudrillard. Exam-
ples provided by Macintosh et al. (2000, pp. 41–
42), such as the use of ‘‘red” on a tra?c light to
mean stop and ‘‘green” to mean go, illustrate not
arbitrariness but conventionality. The original
choice of red instead of, say, purple, was possibly
quite arbitrary, but this does not exhaust Baudril-
lard’s notion of arbitrariness in language. The traf-
?c-light example only evokes the image of
someone following the convention of stopping at
a red light. However, consider for a moment a
world in which drivers of large, powerful trucks
are armed with devices that can control tra?c
lights for their own bene?t, switching them to
green as they approach intersections. In such a
world, these truck drivers would be producers of
tra?c signals, and the masses of other drivers
and pedestrians would be mere consumers. This
is the world of accounting today. The Algoma
Steel crisis illustrates the considerable latitude that
corporations have in controlling the production of
accounting signs. This latitude is built into the
models of Section 3461 of the CICA Handbook,
and it is available to corporations because they
dominate the resource of accounting expertise.
Bringing this matter of power into the equation is
necessary in order to comprehend what Baudril-
lard was getting at. At another point in their article
Macintosh et al. provide a wonderful illustration
of the di?erential e?ects of power, General Elec-
tric’s use of earnings models to make strategic
decisions (pp. 32–33). Their example shows how
the production of accounting signs creates the real-
ity that accounting then pretends to re?ect. Not
everyone has the power to produce such signs.
This is clear in the case of Algoma Steel. Man-
agement was able to produce accounting signs that
mattered, that made a di?erence in the social
world. However, this social di?erence was not pro-
duced until the moment when these accounting
signs were consumed. Consumption, as we saw in
our re-examination of Baudrillard’s earlier writ-
ings, is a social system for producing meaning.
Yet workers and pensioners were largely passive
C. Graham/ Accounting, Organizations and Society 33 (2008) 756–782 777
consumers of these signs, while other creditors and
government bureaucrats had greater capacity to
shape how the accounting signs would be inter-
preted. The function of accounting signs in Alg-
oma Steel’s social relations thus went beyond the
benign agreements on meaning that Macintosh
et al. discussed, when they said that institutionally
produced accounting information is:
. . .an arbitrary but essential feature of a
hyperreal world. Without it, everything
would seem to depend on everything else
and interaction would be problematic if not
impossible. (pp. 32–33)
What Macintosh et al. miss, however, is that
social interactions are problematic. This is true
for government, for investors, for banks, and, no
doubt, even for management. But it is especially
true for the relatively powerless, such as elderly
pensioners. The relations of power in society
become institutionalized in things like Section
3461, and render di?cult, for the average citizen,
the operation of the system of signs by which soci-
ety is governed.
This helps explain what Macintosh et al. (p. 44)
noticed about accounting information, that it does
not fully behave as Baudrillard predicted signs
should in a hyperreal society, and become
absorbed, de-politicized and neutralized by the
masses. The Algoma Steel incidents suggest that
the reason it does not behave in this way is that
the consumers of accounting information are not
‘‘masses”. The consumption of accounting infor-
mation is structured and institutionalized, just as
its production is. There are no undi?erentiated,
passive masses absorbing accounting signs.
Rather, there are relatively institutionalized and
programmatic interfaces waiting to receive the cor-
rect combination of accounting signs in order that
the technologies of governance and wealth distri-
bution be triggered into action. This overstates
the mechanistic aspects of accounting communica-
tion, of course. Nonetheless, the point remains
that accounting signs are consumed primarily by
technically structured corporate and bureaucratic
agents, rather than by masses of consumers.
Hence, when the underfunding of Algoma
Steel’s pension plan was revealed, the tools of pen-
sion accounting permitted Algoma’s management
to construct a new picture of its pension plans,
and to marshal creditors and institutionalized
resources towards a negotiated solution. The Alg-
oma Steel events illustrate that at least in Cana-
dian society, with its relatively decentralized form
of government, accounting signs serve to de?ne
and to shift the boundary between government
and the private sector. The intervention of govern-
ment whereby pension assets and liabilities were
taken over into the PBGF was a temporary col-
lapse, or in Baudrillard’s terms, an implosion, of
the boundary between government and private
sector. The reproduction and recon?guration of
accounting signs in this cases was structured by
institutional mechanisms, regulations, and the role
of expertise. These structures are at the root of the
uneven distribution in Canadian society of the
ability to operate the system of accounting signs.
Conclusion
As Macintosh et al. (2000, p. 45) state, ‘‘We
are concerned with describing the nature of
accounting signs, with unearthing how they came
to be produced and with why they subsequently
come to be taken for granted as a reality of their
own”. The present study has provided speci?c
evidence of how accounting, as a system of
signs, structures and enables the public protec-
tion of private pensions. We have seen how the
institutions and techniques of pension accounting
create a system of wealth distribution and risk
distribution amongst investors, creditors, man-
agement, employees, the government, and other
stakeholders. We have seen, in particular, how
pension accounting under Canadian GAAP has
served to mobilize these institutions in times
when the system is represented as ‘‘failing”.
The system of accruals built into Section 3461
permitted the company’s pension funding prob-
lem to remain unspoken, isolated as a pension
de?cit note rather than being aggregated with
the company’s overall performance numbers.
This permitted the company’s immediate stake-
holders, including investors and creditors, to
avoid the brunt of the new reality induced by
778 C. Graham/ Accounting, Organizations and Society 33 (2008) 756–782
Section 3461. By applying the corridor rule and
con?ning the ‘‘damage” to the company’s pen-
sion fund, the situation was successfully proble-
matized as a pension crisis, amenable to
government intervention through accounting
technologies embodied in the Pension Bene?t
Guarantee Fund. The ‘‘failure” of the system is
thus seen to be the very point at which all its
?exibility and strength was brought into play.
The crisis point—constructed, packaged and pre-
sented by accounting—served as the pivot point
where risks and wealth allocations were success-
fully renegotiated.
The possibilities provided by Section 3461 for
this pivoting and recon?guration were created not
by direct legislation and state control, but by an
institutionalized di?usion of social governance.
The role of accounting signs in this pervasive form
of governance is contingent on the particular
regimes of power and knowledge in which the signs
are situated. I have argued that these signs should
be viewed as simulacra, arbitrary signs that take
their sign value from the system of signs in which
they circulate, and that these arbitrary signs serve
to shift the boundary between government and
the private sector. By focussing on the accounting
sign, the present study highlights how governmen-
tal programs like the PBGF function as institution-
alized consumers of accounting signs. When, as in
the Algoma Steel situation, these accounting signs
are produced by corporations and others outside
of government, and are taken up discursively to
trigger programmatic government interventions,
the power to govern is e?ectively dispersed. This
suggests that the power to govern in today’s demo-
cratic society is, at least in part, negotiated through
and mediated by accounting signs. The Algoma
Steel pension bailout shows how unevenly the
capacity to in?uence this negotiation is distributed
in Canadian society, because of information asym-
metries, the role of accounting expertise, and the
arbitrariness of accounting signs.
Acknowledgements
The comments of Dean Neu, Je? Everett, Teri
Shearer, Hussein Warsame, Daphne Taras, Mike
Welker, Dan Thornton, Steve Salterio, and Eliza-
beth Farrell are gratefully acknowledged, as are
the suggestions of two anonymous reviewers. This
paper bene?ted from the insights of Yves Gendron
and participants at the Alternative Perspectives on
Accounting Conference, held at Laval University
in March 2007. Funding was provided by SSHRC
through a Doctoral Fellowship, and by the Schu-
lich School of Business. The able research assis-
tance of Mona Mann is gratefully acknowledged.
Appendix. The implicit model of Section 3461
Section 3461 of the CICA Handbook uses an
accounting model that brings together actuarial
calculations of future bene?t payments with ?nan-
cial valuations of assets set aside to fund those
payments. The crucial accounting question being
answered by the model (cf. 3461.001) is this: given
the actuarial and ?nancial calculations, what is the
proper way to recognize, measure and disclose the
costs of employee future bene?ts? The principle
applied in the guidelines is that the cost of future
employee bene?ts should be recognized in the per-
iod in which the employee provided the service
that earned her the bene?ts.
28
The explicit basis
for this principle is that ‘‘Bene?t plans are consid-
ered part of an employee’s compensation arrange-
ment” (3461.002). Just as wages are considered an
expense to be matched to the period in which cor-
responding revenues were earned, so future bene-
?ts should be matched to the same period. This
can occur through the recognition of the liability
to pay those bene?ts, or the recognition of the cost
itself, either as an expense or capitalized in an asset
such as inventory (3461.002, Footnote 1).
The attribution mechanisms at the heart of this
model involve matching distinct units of the future
bene?ts to speci?c periods of service by employees,
and calculating by actuarial methods the present
value of each unit for the period in which it
accrued. The actuarial methods are black-boxed
by the Handbook. That is, they are treated as an
28
The guidelines point out that some future bene?ts are
recognized when an event occurs, such as the application of the
employee for disability bene?ts (3461.003).
C. Graham/ Accounting, Organizations and Society 33 (2008) 756–782 779
independent and unquestionable given. Two sepa-
rate accrual attributions are involved. In the ?rst,
the units of future bene?t are calculated as earned
on a period by period basis. By this process, the
accrued bene?t obligation of the company is calcu-
lated. This obligation is recalculated each year,
based on updated actuarial assumptions, employee
data, and other input parameters.
29
During the
course of a year, the obligation for future bene?ts
increases by the cost of bene?ts earned by employ-
ees during the year (the current service cost), and
decreases by the amount of any bene?t payments
actually made. In addition, interest accrues to the
employees on the outstanding portion of the obli-
gation, the portion that had been earned before the
period began and which has not yet been paid to
the employees. When the experience of all these
increases and decreases does not match the actuar-
ial projections, the di?erence is calculated as an
actuarial gain or loss on the accrued bene?t obliga-
tion. An additional actuarial gain or loss is calcu-
lated on the plan assets, being the di?erence
between expected and actual return on assets.
The total of these two actuarial gains or losses is
not recognized immediately in the entity’s ?nancial
statements. Rather, the model assumes that these
actuarial gains and losses are random, and will o?-
set themselves over time. In order to give the entity
time for these ?uctuations to net out, and so avoid
short-term shocks to its ?nancial statements, a sec-
ond accrual attribution is used.
In the second accrual attribution of the model,
a portion of the costs of the bene?t obligation is
allocated back to the current period. By this pro-
cess, the accrued bene?t liability of the company
is calculated. The so-called ‘‘corridor” rule is the
primary method for making this attribution.
Under the corridor rule, the net actuarial gain or
loss, which may ?uctuate widely from year to year
due to changes in mortality experience, market
values for assets, and so on, is dampened by accu-
mulating it from year to year o? the balance sheet,
and amortizing a portion of this accumulation to
the ?nancial statements each year. The calculation
of the amortization amount is convoluted. First,
only part of the net actuarial gain or loss is subject
to amortization. If the net actuarial gain or loss
does not exceed 10% of the greater of the accrued
bene?t obligation and the fair value of plan assets,
no amortization is necessary. If it does exceed this
10% ?gure, then only the amount by which it
exceeds the 10% ?gure is subject to amortization.
This excess amount is to be spread out over the
remaining service life of the average employee.
That is, if most employees have 8 years to go
before collecting their bene?ts, then only 1/8th of
the excess amount is recognized on the entity’s
?nancial statements this year. Furthermore, since
the calculation is performed using the net actuarial
gain or loss as at the beginning of the year, the
impact of a ?uctuation on the ?nancial statements
is not only dampened but delayed.
Besides the corridor rule for actuarial gains and
losses, several other calculations are used to attri-
bute the bene?t obligation back to the current per-
iod as a pension cost. The current service cost, the
interest cost on the obligation, and the simple
amortization of any changes to the plan—i.e.,
changes to accounting policies (transitional obliga-
tions) or retroactive changes to promised bene?ts
(past service costs)—are all components of the
pension cost.
The pension cost recognized every year is accu-
mulated. The contributions the entity makes to the
pension plan are deducted from this accumulation.
The result is the accrued bene?t liability that
appears on the entity’s balance sheet.
References
Accounting Standards Board of Canada. (2002). Employee
future bene?ts: Accounting Standards Board of Canada.
Downloaded December 19, 2004.
Algoma Steel. (1996). Annual report. Sault Ste. Marie, Ontario:
Algoma Steel Inc.
Algoma Steel. (1997). Annual report. Sault Ste. Marie, Ontario:
Algoma Steel Inc.
Algoma Steel. (2000). Annual report. Sault Ste. Marie, Ontario:
Algoma Steel Inc.
29
Pension plans are governed by federal legislation for entities
registered nationally, and by provincial legislation for entities
registered provincially. This pension legislation typically
requires a complete actuarial recalculation only every few
years. Extrapolations from the prior recalculation are permitted
in intervening years.
780 C. Graham/ Accounting, Organizations and Society 33 (2008) 756–782
Algoma Steel. (2001a). Annual report. Sault Ste. Marie,
Ontario: Algoma Steel Inc.
Algoma Steel. (2001b). Notice of proceedings and meetings of
certain creditors of Algoma Steel Inc. and information
circular with respect to a plan of arrangement and reorga-
nization. Sault Ste. Marie, Ontario: Algoma Steel Inc.
Algoma Steel. (2002). Annual report. Sault Ste. Marie, Ontario:
Algoma Steel Inc.
Algoma Steel’s $625-M Pension Shortfall Key to Restructuring.
(2001). Sudbury Star, August 24, p. A.10.
Archer, S. (1998). Mattessich’s critique of accounting: A
review article. Accounting and Business Research, 28(4),
297–316.
Armstrong, P. (2000). Designing accountability: The manage-
rial semiotics project. Critical Perspectives on Accounting,
11(1), 1–22.
Arrington, C. E., & Francis, J. R. (1989). Letting the chat out of
the bag: Deconstruction, privilege and accounting research.
Accounting, Organizations and Society, 14(1–2),
1–28.
Baudrillard, J. (1968). Le Syste` me des Objets. Paris: Gallimard.
Baudrillard, J. (1970). La Socie´ te´ de Consommation; Ses
Mythes, Ses Structures. Paris: Gallimard.
Baudrillard, J. (1972). Pour Une Critique de L’e´ conomie
Politique du Signe. Paris: Gallimard.
Baudrillard, J. (1973). Le Miroir de La Production; Ou,
L’illusion Critique du Mataˆ erialisme Historique. Tournai:
Casterman.
Baudrillard, J. (1975). The mirror of production (M. Poster,
Trans.). St. Louis, Missouri: Telos Press.
Baudrillard, J. (1981). For a critique of the political economy of
the sign (C. Levin, Trans.). St. Louis, Missouri: Telos Press.
Baudrillard, J. (1983). Simulations. New York City:
Semiotext(e).
Baudrillard, J. (1993). Symbolic exchange and death. London:
Sage.
Baudrillard, J. (1995). Simulacra and simulation. Ann Arbor:
University of Michigan Press.
Baudrillard, J., & Poster, M. (1988). Jean Baudrillard: Selected
writings. Stanford, CA: Stanford University Press.
Beatty, C. A., & Schachter, H. (2002). Employee ownership: The
new source of competitive advantage. Etobicoke, Ontario:
John Wiley & Sons.
Belkaoui, A. (1978). Linguistic relativity in accounting.
Accounting, Organizations and Society, 3(2), 97–104.
Belkaoui, A. (1980). The impact of socio-economic accounting
statements on the investment decision: An empirical
study. Accounting, Organizations and Society, 5(3), 263–
283.
Boland, R. J. Jr., (1989). Beyond the objectivist and the
subjectivist: Learning to read accounting as text. Account-
ing, Organizations and Society, 14(5–6), 591–604.
Brooksbank, D. (2006). Accounting rules ‘Threaten Swiss
Pension System’. Investment and Pensions Europe (24
January 2006).
Cooper, C., & Puxty, A. (1994). Reading accounting writing.
Accounting, Organizations and Society, 19(2), 127–146.
Daly, B. A., & Schuler, D. K. (1998). Rede?ning a certi?ed
public accounting ?rm. Accounting, Organizations and
Society, 23(5–6), 549–567.
DaPrat, M. (2005a). Personal communication, January 31.
DaPrat, M. (2005b). Personal communication, February 10.
Daw, J. (2001). Failure could wipe out pension guarantee fund;
Provincial kitty has $230 million for Aid. Toronto Star,
April 25, E.01.
Dofasco. (2001). Annual report. Hamilton, Ontario: Dofasco
Inc.
Dufresne, D. (1993). Some aspects of statement of Financial
Accounting Standards No. 87. Actuarial Research Clearing
House (Vol. 2).
Estey, N. A. (1999). Accountable for your promises. CA
Magazine.
Evans, L. (2004). Language, translation and the problem of
international accounting communication. Accounting,
Auditing & Accountability Journal, 17(2), 210–248.
Everett, J. (2004). Exploring (False) dualisms for environmental
accounting praxis. Critical Perspectives on Accounting,
15(8), 1061–1084.
FASB. (2006). FAS 158: Employers’ accounting for de?ned
bene?t pension and other postretirement plans. Norwalk,
Connecticut: Federal Accounting Standards Board. Downloaded February 28,
2007.
Feltham, G. A., & Ohlson, J. A. (1995). Valuation and clean
surplus accounting for operating and ?nancial activities.
Contemporary Accounting Research, 11(2), 689–731.
Fore, D. (2004). Changes in accounting practices will drive
pension paradigm shift. Pension Research Council Working
Paper. Philadelphia: Pension Research Council, University
of Pennsylvania.
Francis, J. R. (1994). Auditing, hermeneutics, and subjectivity.
Accounting, Organizations and Society, 19(3), 235–269.
Gadamer, H. (1976). Philosophical hermeneutics. Berkeley:
University of California Press.
Graves, O. F., Flesher, D. L., & Jordan, R. E. (1996). Pictures
and the bottom line: The television epistemology of U.S.
annual reports. Accounting, Organizations and Society,
21(1), 57–88.
Heidegger, M. (1962). Being and time (J. Macquarrie & E.
Robinson, Trans.). New York: Harper & Row.
IASB. (2004). Summaries of International Financial
Reporting Accounting Standards: IAS 19, employee bene-
?ts. London: International Accounting Standards Board.
Down-
loaded February 25, 2006.
Joss, P. N. (2004). The earnings implications of pension expense:
A stochastic analysis of ten Canadian companies. Unpub-
lished Masters thesis, University of Calgary, Calgary.
Keenan, G. (2001). Noteholder votes deal a blow to Algoma
restructuring plan. Toronto Star, December 7.
Lavoie, D. (1987). The accounting of interpretations and the
interpretation of accounts: The communicative function of
‘‘the language of business”. Accounting, Organizations and
Society, 12(6), 579–604.
C. Graham/ Accounting, Organizations and Society 33 (2008) 756–782 781
Liddle, J. (2002). New accounting and supervisory standards
could prove unpopular with pension funds themselves.
European Pensions & Investment News (114).
Macintosh, N. B., & Baker, C. R. (2002). A literary theory
perspective on accounting: Towards heteroglossic account-
ing reports. Accounting, Auditing & Accountability Journal,
15(2), 184–222.
Macintosh, N. B., & Shearer, T. (2000). The Accounting
Profession Today: A Poststructuralist Critique. Critical
Perspectives on Accounting, 11(5), 607–626.
Macintosh, N. B., Shearer, T., Thornton, D. B., & Welker, M.
(2000). Accounting as Simulacrum and Hyperreality: Per-
spectives on Income and Capital. Accounting, Organizations
and Society, 25(1), 13–50.
Mattessich, R. (2003). Accounting Representation and the
Onion Model of Reality: A Comparison with Baudrillard’s
Orders of Simulacra and His Hyperreality. Accounting,
Organizations and Society, 28(5), 443–470.
McDowall, D. (1984). Steel at the Sault: Francis H. Clergue, Sir
James Dunn and the Algoma Steel Corporation: 1901–1956.
Toronto: University of Toronto Press.
McGoun, E. G. (1997). Hyperreal Finance. Critical Perspectives
on Accounting, 8(1-2), 97–122.
Mouck, T. (1994). Corporate Accountability and Rorty’s
Utopian Liberalism. Accounting, Auditing & Accountability
Journal, 7(1), 6–30.
Mouck, T. (2004). Institutional Reality, Financial Reporting
and the Rules of the Game. Accounting, Organizations and
Society, 29(5-6), 525–541.
Nishman, R. F. (1995). Worker ownership and the restructuring
of Algoma Steel in the 1990s. Kingston, Ontario: Industrial
Relations Centre, Queen’s University.
Ontario Superior Court. (2001a). A?davit of Glen Percival
Manchester, Sworn April 23, 2001. Court File No. 01-CL-
4115. Toronto, Ontario.
Ontario Superior Court. (2001b). Cross-examination of Harold
Stephen, Sworn October 22, 2001. Court File No. 01-CL-
4115. Toronto, Ontario.
Ontario Superior Court. (2001c). Motion returnable Novem-
ber 9. 2001 Regarding meetings of creditors and claims
procedure. Court File No. 01-CL-4115. Toronto,
Ontario.
Ontario Superior Court. (2001d). Ninth report of Ernst &
Young Inc. in its capacity as court-appointed monitor of
Algoma Steel Inc., December 7, 2001. Court File No. 01-
CL-4115. Toronto, Ontario.
Ottawa Provides $50-M Loan Guarantee for Algoma. (2001).
Sudbury Star, December 23, 2001, A.1.FRO.
Pension Bene?ts Act. (1990). Revised statutes of Ontario.
Pension Changes Could Cost $180b: Accounting Organiza-
tion’s Strict New Rules Could Cost Billions for Companies
O?ering De?ned Pension Plans. (2006). Downloaded February 3, 2006.
Perkel, C. (2001). Taxpayers may pay for Algoma; Steel ?rm’s
fall would have huge impact: Harris. Toronto Star,
September 7, 2001, NE.07.
Preston, A. M., Wright, C., & Young, J. J. (1996). Imag[in]ing
annual reports. Accounting, Organizations and Society,
21(1), 113–137.
Preston, A. M., & Young, J. J. (2000). Constructing the global
corporation and corporate constructions of the global: A
picture essay. Accounting, Organizations and Society, 25(4–5),
427–449.
Province of Ontario. (2001). Ontario Government appoints
representative to protect pensioners – News release, August
31, 2001. Toronto: Province of Ontario. Down-
loaded December 12, 2003.
Reg. 203/02. (2002). Ontario regulations.
Rorty, R. (1991). Science as solidarity, objectivity, relativism,
and truth: Philosophical papers (Vol. 1). Cambridge, UK:
Cambridge University Press, pp. 35–45.
Saussure, F. d. (1961). Course in general linguistics. London:
Owen.
Schoonmaker, S. (1994). Capitalism and the code: A critique of
Baudrillard’s third order simulacrum. In D. Kellner (Ed.),
Baudrillard: A critical reader (pp. 168–188). Oxford, UK:
Blackwell.
Squeeze on Pensions Could Tighten: Overhaul in Accounting
Rules Will Cost Workers. (2006). Downloaded February 3, 2006.
Steelworkers Hail Pension Fund Appointment. (2001). Edmon-
ton Journal, September 1, 2001, G.8.
Stelco. (1999). Annual report. Hamilton, Ontario: Stelco Inc.
Stelco. (2001). Annual report. Hamilton, Ontario: Stelco Inc.
Tinker, A. M., Merino, B. D., & Neimark, M. D. (1982). The
normative origins of positive theories: Ideology and
accounting thought. Accounting, Organizations and Society,
7(2), 167–200.
Tinker, T. (1991). The accountant as Partisan. Accounting,
Organizations and Society, 16(3), 297–310.
Tishco?, N. (1999). An actuarial action plan, CA Magazine.
Van Alphen, T. (2001a). Insolvent Algoma Steel gains bankruptcy
court’s protection; Major employer in Sault Forecasts loss this
year of $200 Million. Toronto Star, April 24, 2001, E.01.
Van Alphen, T. (2001b). Pension aid crucial to restructuring of
Algoma Steel; government’s response to woes ’disappoint-
ing’. Toronto Star, August 24, 2001, E.01.
Walters-York, L. M. (1996). Metaphor in accounting discourse.
Accounting, Auditing & Accountability Journal, 9(5), 45–70.
Watson, T. (2001). Ottawa may give millions for bailout of
Algoma: Ontario will not help. National Post, August 25,
2001, C.1.FRO.
782 C. Graham/ Accounting, Organizations and Society 33 (2008) 756–782
doc_204874543.pdf
This paper examines the role of accounting in society by looking at the consumption of accounting signs during the
financial restructuring of a corporation. The paper builds upon insights from prior research on accounting as simulacrum
and hyperreality. It examines how accounting numbers serve as reconfigurable signs that construct appropriate
‘‘crises”, motivate government intervention, and marshal stakeholders towards solutions
Fearful asymmetry: The consumption of accounting signs
in the Algoma Steel pension bailout
Cameron Graham
*
Schulich School of Business, York University, 4700 Keele Street, Toronto, Ontario, Canada M3J 1P3
Abstract
This paper examines the role of accounting in society by looking at the consumption of accounting signs during the
?nancial restructuring of a corporation. The paper builds upon insights from prior research on accounting as simula-
crum and hyperreality. It examines how accounting numbers serve as recon?gurable signs that construct appropriate
‘‘crises”, motivate government intervention, and marshal stakeholders towards solutions. The incident at the heart
of this study is the 2001 bailout of the Algoma Steel pension plan by the Ontario government. The incident demon-
strates how accounting technologies are required both for the production of accounting signs and for their consump-
tion. The paper asks how the production and consumption of accounting signs is di?erent from that of other
communication signs, what role consumers of accounting signs play in determining their meaning, and what di?erence
this makes in how corporate pension plans are protected by government. It concludes that the structures and mecha-
nisms surrounding the consumption of accounting signs enable di?erent stakeholders to in?uence the production of
meaning at the moment when accounting signs are consumed, changing the way that risk and wealth are redistributed,
and shaping government intervention.
Ó 2008 Elsevier Ltd. All rights reserved.
Introduction
In this paper, I draw on the work of Baudrillard
to examine the consumption of accounting signs by
various stakeholders during a corporate ?nancial
crisis. Baudrillard (1972, 1973) developed provoca-
tive semiological theories about the production and
the consumption of signs in society. I ask how the
production and consumption of accounting signs
is di?erent from the production and consumption
of other signs, and what di?erence this makes in
the role of accounting in society. I use a Canadian
pension accounting example to explore how
accounting signs are used by management, govern-
ment, creditors, and other corporate stakeholders.
In 2001, the severely underfunded pension plan of
Algoma Steel was taken over by the Ontario gov-
ernment. The analysis of documents pertaining
to this incident demonstrates how accounting
0361-3682/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.aos.2008.01.001
*
Tel.: +1 416 736 2100x77926; fax: +1 416 736 5687.
E-mail address: [email protected]
Available online at www.sciencedirect.com
Accounting, Organizations and Society 33 (2008) 756–782
www.elsevier.com/locate/aos
technologies structure the production and con-
sumption of accounting signs, exacerbating di?er-
ences in how stakeholders are able to consume
these signs, triggering and limiting government
intervention, and shifting the boundary between
the private and public sectors.
My analysis draws upon the work of Macintosh,
Shearer, Thornton, and Welker (2000), who dis-
cussed ?nancial accounting theory using Baudril-
lard’s notions about signs (see Baudrillard, 1983;
Baudrillard, 1995, 1988). This approach asserts
that accounting signs have become divorced from
any underlying objective reality. With respect to
pension accounting, I argue that Canadian GAAP
for de?ned-bene?t pensions has codi?ed a system
of arbitrary signs. The study of Algoma Steel dem-
onstrates how these accounting signs can be con?g-
ured to create representations of pension plans that
are useful to corporations in managing government
and other stakeholders. I also argue that inequali-
ties amongst stakeholders in their ability to con-
sume accounting signs contribute to di?erences in
the power they are able to exert over the meaning
of those signs. The study will be of interest to
accounting researchers because it expands, with
concrete examples, our understanding of how
accounting functions as a language, and in particu-
lar how accounting di?ers from other languages at
the point where its signs are consumed.
This study contributes to our understanding of
the role of accounting information in de?ning the
boundary between the private and public sectors.
It provides speci?c evidence of how the production
and consumption of accounting information, as a
highly structured form of communication, alters
the very institutional setting in which that commu-
nication occurs. The transference of pension liabil-
ities from private corporations to government, and
the intervention by government in private ?nancial
crises, are examples of how the boundary between
the private and the public sector can shift.
The study is particularly relevant because Alg-
oma Steel was an early adopter of revised Cana-
dian pension disclosure standards in 2000.
Similar changes to pension disclosure standards
have been implemented by the International
Accounting Standards Board under IAS 19 (IASB,
2004) and the Financial Accounting Standards
Board under FAS 158 (FASB, 2006). These
changes have provoked concern (e.g. Brooksbank,
2006; Liddle, 2002; Pension Changes Could Cost
$180b, 2006; Squeeze on Pensions Could Tighten,
2006). This study helps shed timely light on the
impact of such changes by examining a ‘‘pension
crisis” triggered in part by revisions to Canadian
pension disclosure standards.
Retracing Baudrillard’s steps
The underlying thesis of this paper is that
accounting, as a system of signs, can be e?ectively
critiqued by approaching it from the perspective
of linguistic theories about the function of signs
in today’s society. Prior research has spent surpris-
ingly little time examining accounting from such a
perspective, given the prominence of linguistic the-
ory in philosophy in the 20th century. Pioneers in
linguistic or literary approaches to accounting
research include Belkaoui (1978, 1980), who
asserted that accounting is a language and explored
the cognitive implications of using that language;
Lavoie (1987), who explored the hermeneutics of
economic decision making with accounting infor-
mation; Arrington and Francis (1989), who intro-
duced postmodern linguistic theory to the
accounting literature; and Boland (1989), who
showed how a hermeneutic approach could break
down the dichotomy between subjectivism and
objectivism. Although many critical accounting
articles are indebted to linguistic theory in a general
way, explicitly linguistic approaches to accounting
research after 1989 are unfortunately few.
1
The one
1
I am setting aside the excellent stream of research on
accounting and visual images (Daly & Schuler, 1998; Graves,
Flesher, & Jordan, 1996; Preston, Wright, & Young, 1996;
Preston & Young, 2000), which draws on linguistic theorists. In
the pages of AOS, among the few explicitly linguistic papers, as
opposed to papers that merely have textual data, are Cooper
and Puxty’s (1994) textual analysis of a professional accounting
journal article, and Francis’s (1994) exploration of the herme-
neutics of auditing. Linguistic treatments of accounting are
both more frequent and more recent outside AOS (e.g.,
Armstrong, 2000; Evans, 2004; Everett, 2004; Macintosh &
Baker, 2002; Macintosh & Shearer, 2000; McGoun, 1997;
Mouck, 1994; Walters-York, 1996).
C. Graham/ Accounting, Organizations and Society 33 (2008) 756–782 757
I engage here, by Macintosh et al. (2000), draws on
Baudrillard’s notions of simulacrum, implosion,
and hyperreality to discuss ?nancial accounting
theory. Macintosh et al. (p. 14) argue that Baudril-
lard’s radical semiological theories are relevant to
accounting because they deal with changes in lan-
guage, information technology, communication
and media that fundamentally a?ect accounting.
2
They use Baudrillard’s insights to develop a cri-
tique of accounting based on the production and
consumption of information. This approach is par-
ticularly appropriate for the present study because
the resolution of the 2001 Algoma Steel crisis
hinged on how the company’s accounting ‘‘signs”
were produced and consumed.
In semiotic terms (see, for example, Saussure,
1961), signs consist of a word or picture – the sig-
ni?er – that evokes a concept – the signi?ed. The
signi?er and signi?ed together act as a sign point-
ing to some underlying thing – the referent. The
word ‘‘corporation”, for instance, evokes the
notion of an incorporated company and can be
used to refer to a speci?c organization. In tradi-
tional accounting theory, signs have an objective
referent: ‘‘net income” measures a real surplus of
a company’s economic activity. Macintosh et al.
argue, however, that accounting signs have lost
their objective referent. They take up Baudrillard’s
(1983) provocative suggestion that signs now take
their meaning only from their relationship to other
signs in the communicative system. As noted by
Macintosh et al. (p. 40), Baudrillard argued
(1983, pp. 11–12) that signs have passed through
four historical phases. In the ?rst phase, signs were
a re?ection of a profound reality. In the second,
signs masked or denatured this profound reality.
In the third, signs masked the absence of a pro-
found reality. And, in the fourth, which according
to Baudrillard is the phase of today, signs precede
reality.
Macintosh et al. equate the ?rst phase to pre-
historic accounting, and trace the accounting
notions of capital and income through Baudril-
lard’s other phases. Accounting, they maintain, is
now without objective referent, and they give sev-
eral supporting examples, including accounting for
executive stock options and earnings management.
They argue that, just as in Baudrillard’s examples
of the hyperreal society, accounting signs now pre-
cede the reality they purport to represent, creating
that reality through their sign value: accounting
signs have gained independence from the real. As
Baudrillard put it, ‘‘from now on signs will
exchange among themselves exclusively, without
interacting with the real . . . [they are] at last free
for a structural or combinatory play that succeeds
the previous role of determinate equivalence”
(Baudrillard & Poster, 1988, p. 125); quoted in
(Macintosh et al., 2000, p. 40).
Macintosh et al. argue (p. 40) that Baudrillard’s
characterization totalizes the present age and
ignores the way that accounting signs still main-
tain a connection to an underlying, albeit social,
reality. They ask, ‘‘why is it that accounting infor-
mation is used extensively in economic and social
relations when, according to our Baudrillardian
analysis, it does not refer to any real objective
realm, and has had changing referents over time?”
(p. 40). They point to the clean surplus model (Fel-
tham & Ohlson, 1995) as a partial answer, suggest-
ing that as long as capital and income articulate in
a clean surplus relationship, it does not matter
how each is de?ned. The accounting notions of
capital and income take their sign values from
the system of signs in which they are embedded,
rather than from any underlying objective ‘‘thing”
represented by capital or income. In this sense, the
signs are arbitrary.
If accounting is a system of arbitrary signs –
that is, a language – then it is amenable to linguis-
tic theoretical analysis. However, caution is
required. Accounting di?ers from English and
other vernaculars in several socially important
respects. It is formally de?ned by a professionally
dominated process, for instance. Its use is partially
professionalized. It has peculiar technical features
2
Mattessich (2003) has criticized Macintosh et al. and
Baudrillard from the perspective of objective realism. Mattess-
ich, who seems to be irritated by Baudrillard’s refusal to play
the scienti?c language game, clings to the notion that science –
and hence accounting – can represent objectively the reality of
which it is a part. This stance has been discredited by many
philosophers, notably Rorty (1991). Mattessich’s philosophy
has been critiqued by Archer (1998) and Mouck (2004), but
Macintosh et al. have not yet responded to his treatment of
Baudrillard.
758 C. Graham/ Accounting, Organizations and Society 33 (2008) 756–782
pertaining to its numeric and calculative nature
that lend it a false aura of objectivity and precision
(cf. Tinker, Merino, & Neimark, 1982; Tinker,
1991) and enforce a certain arithmetic grammar
over what can be said. Above all – and this is
something Macintosh et al. neglect in their inter-
pretation of Baudrillard for accounting – account-
ing functions as a system of signs primarily in the
?nancial markets, not the product markets. The
?nancial markets are unlike the mass consumer
market featured in Baudrillard’s analysis of signs.
In order to explain the implications of this more
fully, I want to retrace the development of Baudril-
lard’s notions of simulacra and hyperreality.
Baudrillard’s analysis of signs began with his
e?orts to update Marxist theory to account for
the shift in emphasis in Western society from pro-
duction to consumption (Baudrillard, 1968, 1970).
He formalized his theories in a series of essays
published as For a Critique of the Political Econ-
omy of the Sign (1972 [trans. 1981]). In these
essays, Baudrillard draws a contrast between sym-
bolic exchange and consumption (1981, pp. 63–
65). An example of symbolic exchange is the giving
of a gift. A gift is more than an object, it is sym-
bolic of a unique relationship between people. A
commodity, on the other hand, is a sign of rela-
tions of production. It takes its meaning not from
a unique relationship between people but from its
‘‘di?erential relation to other signs” (p. 66). What
Baudrillard means by this is that the basis of con-
sumption is meaningful social exchange, not the
brute satisfaction of individual needs. He com-
pares consumption to language. Language, he
says, cannot be understood on an utterance-by-
utterance basis because individual utterances are
only intelligible in the context of language. So con-
sumption cannot be understood as the satisfaction
of individual needs, but must be understood as a
social system for producing meaning through dif-
ferences in the sign value of commodities (p. 75).
Baudrillard’s most formal moments come when
he compares the functional logic of use value, the
economic logic of exchange value, the di?erential
logic of sign value, and the logic of symbolic
exchange (pp. 123–129). In teasing out the connec-
tions between these logics, Baudrillard comes to
the conclusion that, just as exchange value domi-
nates use value in our economic system, so the sig-
ni?er dominates the signi?ed in our signi?cation
system. The principle of economic circulation is
the production of equivalence, exchange value.
The principle of circulation in our system of
signi?cation is the production of di?erence, the
signi?er. Thus, in order to understand the con-
sumption of objects completely, says Baudrillard,
one must understand them as signs, not only as
commodities.
Baudrillard’s critique reaches its zenith in the
essay for which the book is named. Here, Baudril-
lard argues that the problems of Marxist thinking
derive from an arti?cial separation of the economy
and culture, paralleled in the arti?cial separation
of the commodity and the sign, requiring ‘‘magical
thinking” regarding ideology to reunite them (pp.
142–145). He makes two matching claims: that
because the logic of the commodity is at the heart
of the sign, signs can function as both exchange
value and use value; and that because the structure
of the sign is at the heart of the commodity, com-
modities can function inherently as signs (p. 146).
The commodity functions as a system of commu-
nication governing social interaction, reducing
the symbolic to the form of the sign and rational-
izing all exchange (p. 147). The key point in this
essay is Baudrillard’s recognition that the moment
of interconversion between commodity and sign,
when meaning is generated from the commodity
and the sign is commodi?ed, is the moment of con-
sumption, not production (p. 147).
3
This is what
de?nes today’s society, and what escapes tradi-
tional and neo-Marxist analysis with its focus on
production and its arti?cial separation of economy
and culture. What Baudrillard is proposing is a
more general critique of political economy.
Baudrillard makes a subtle point here that will
prove important in the discussion of Algoma Steel
3
This may seem fairly straightforward to those familiar with
the hermeneutical precepts of Heidegger (1962) and Gadamer
(1976), who argued that meaning is created when a text is read,
and that the meaning will go beyond the original intentions of
the author (cf. Lavoie, 1987; Boland, 1989). Baudrillard’s
insight is to link this line of thought to the production and
consumption of the commodity, and to recognize the commod-
ity as a sign.
C. Graham/ Accounting, Organizations and Society 33 (2008) 756–782 759
that follows. He argues that the arbitrariness of the
sign does not have to do with the referent alone. It
is not the sign that is arbitrarily connected to the
referent. It is the signi?er that is arbitrarily con-
nected to the signi?ed/referent pair. The signi?ed
and referent together form the content of the sign,
under the ‘‘aegis” of the signi?er (p. 151). The
implication of this is that the referent, the ‘‘objec-
tive” thing that we often presume to be re?ected in
the sign, is actually not independent of the sign:
‘‘In a profound sense, the referent is the re?ection
of the sign . . .” (p. 151). Baudrillard’s later notions
of simulacrum, implosion, and hyperreality derive
from this insight.
This critique of the sign positions Baudrillard
to break with Marxism. Baudrillard suggests, for
instance, that Marx’s analysis of the commodity
was incomplete because he never developed a cri-
tique of use value (pp. 128–129). The actual break
did not come, however, until Baudrillard’s next
book, The Mirror of Production (1973 [trans.
1975]). In this he criticized Marx for reducing
people to producers, life to economic production,
by accepting the central and unfortunate proposi-
tion of political economy: that people are labour
power. To Baudrillard, people are no better o?
being pegged as homo faber than they are as homo
economicus (1975, p. 49). He illustrates this by
pointing to Marx’s imposition of a ‘‘mode of pro-
duction” on primitive societies. He argues that
Marx has only adopted and reproduced the fun-
damental concepts of political economy, and that
it was necessary to look next at the mode of sig-
ni?cation and a critique of the sign (p. 51). He
argues that Marx took the economy and produc-
tion as axiomatic, and is therefore unable to
account for how the principle of production is
itself produced (p. 66). He goes so far as to say
that Marx’s anachronistic analysis of primitive
and feudal societies results in a ‘‘theoretical, polit-
ical and strategic miscomprehension of capitalist
formations themselves” (p. 107).
Baudrillard (1975, pp. 119–129) addresses these
limitations by examining how monopoly capital-
ism di?ers from competitive capitalism. Monopoly
capitalism, he argues, represents a revolutionary
change in capitalism, not merely an extension.
Exchange is no longer just abstracted, it is opera-
tionalized, coded, through the production of
meaning and di?erence. What matters is not the
control of the means of material production, but
control of the code (p. 122). While the consump-
tion of signs produces di?erence, the process is
not that of social di?erentiation through conscious
consumption to consolidate one’s class status.
Rather, the process is unconscious (pp. 122–123)
because it is programmed. Consumption is no
longer contingent, supply and demand are no
longer dialectically related; competition, supply
and demand remain only as myths to support the
system (p. 125). Consumption is now driven by
planned socialization, by advertising, for example,
which provides not only the answers but also the
questions. In this system, signi?ers are divorced
from their contents, both the signi?ed and the ref-
erent, to play freely in self-referential exchange
with other signi?ers. The system thus neutralizes
all contradictions by abolishing referents (p.
129), that is, by collapsing or imploding the di?er-
ence between the signi?ed and the referent. The
radical nature of Baudrillard’s critique comes to
life when he shows how his analytic approach
explains what Marxist analysis cannot, namely
the relegation of entire groups of people – stu-
dents, races, women – to positions outside the
realm of production. These groups cannot be
aligned with workers in revolt against those who
control the means of production. They must revolt
against the code (pp. 131–141).
This, then, is the radical and intensely theorized
backdrop against which Baudrillard’s (1983)
notions of simulacra, implosion, and hyperreality
must be viewed. Macintosh et al. (2000, pp. 14–
16) provide a solid summary of these three
notions, but they do not connect them to their the-
oretical derivation. Brie?y, the simulacrum is the
sign that has been divorced from its referent, and
takes its meaning only from its relation to other
signs. Implosion is the collapse of the boundary
between two concepts or realms, and of the di?er-
ence between sign and referent. Hyperreality is the
state of society in which simulation and simulacra
dominate. When Macintosh et al. use these con-
cepts to explore accounting, they achieve little
more than a recapitulation of traditional account-
ing history in trendy terms. By picking up Baudril-
760 C. Graham/ Accounting, Organizations and Society 33 (2008) 756–782
lard at the point where terms like simulacra and
hyperreality emerge, and ignoring the prior devel-
opment of his thought, they reinvigorate an ana-
lytical model of accounting theory but threaten
nothing. Baudrillard, they claim (pp. 44–45), is
obsessed with the sign world and marginalizes
relations of power. Such an accusation is incom-
prehensible given even the abbreviated chronology
of his theoretical development provided above.
What then to do about Baudrillard and
accounting? I will argue in what follows that
Baudrillard must be taken just seriously enough.
He is a provocateur, and must not under any cir-
cumstances be taken too literally.
4
There are
glaring weaknesses in what he says, such as a
failure to ground his notion of ‘‘symbolic
exchange” empirically, which makes it seem
somewhat nostalgic; a failure to de?ne more pre-
cisely the ‘‘code” or identify who controls it and
how; as well as a tendency towards fatalism and
nihilism in later works (e.g. Baudrillard, 1993).
However, he provokes us to rethink our assump-
tions, and provides a compelling vocabulary for
discussing society that is rooted in his own dis-
section of Marx. Implications regarding relations
of power can quite readily be brought out when
using his vocabulary to understand present-day
accounting. This is what I will do in the follow-
ing case study. I will use the incident of the
?nancial collapse of Algoma Steel in 2001 and
the subsequent bailout of its pension plan by
the Ontario government to show how accounting
signs function in Baudrillard’s terms. I will, how-
ever, also show how Baudrillard’s analysis is lim-
ited for accounting by certain features in the
mode of production and the mode of consump-
tion of accounting signs.
Before I proceed with my discussion of the Alg-
oma case, I want to clarify what is distinct about
the form of accounting signs. Semiotics, as noted
above, describes the sign as a signi?er/signi?ed
pair pointing to a referent. The accounting sign
has a peculiar form, in that accounting informa-
tion consists of a system of equations. The basic
form of the accounting sign is ‘‘Label = Value”,
for example ‘‘Accounts Receivable = $100,000”.
These signs can be built up into higher level signs,
such as ‘‘Current Assets = $1,000,000”, and still
higher ones, ‘‘Total Assets = $10,000,000”. In
?nancial accounting, these basic forms are
arranged according to the logic of the double
entry. Certain events in the course of business
are selectively recorded in variations on the form
‘‘Debits = Credits”. This is an equation of equa-
tions, each side consisting of one or more basic
signs in the form ‘‘Label = Value”. The combina-
tion and recombination of the basic form and the
transactional form gives rise to other equations at
the level of the ?nancial statement proper, such as
‘‘Assets = Liabilities + Owners’ Equity” and
‘‘Revenue – Expenses = Net Income”. It seems
clear that the form I have called the basic one is
no more basic than the transactional form or the
higher level forms of the balance sheet and income
statement, but I will call it ‘‘basic” for expository
purposes, because it is the form most consistently
used throughout the Algoma Steel legal
proceedings.
One must avoid concluding that the ‘‘Label”
in the basic accounting equation is the signi?er
and the ‘‘Value” is the signi?ed, or worse, the ref-
erent. The entire equation ‘‘Label = Value” is a
signi?er. The form of the accounting sign supports
the attempt to impose a single meaning on the
sign, to eliminate its symbolic potential and reduce
it to the unequivocal (Baudrillard, 1981, p. 149).
For the sign-equation suggests ?nality, as it seems
to contain in the monetary amount its own refer-
ent. Yet the monetary amount is but a sign of
value. It takes its meaning from its relationship
to other signs of value, for instance, the values that
were assigned to that label in previous ?nancial
statements, or the values that are assigned to a
similar label in other companies’ ?nancial state-
ments, or the values that were expected by ana-
lysts. It also takes its meaning from its relation
to other signs of value in the same set of ?nancial
statements, such as the meaning of one liability in
relation to other liabilities. These meanings, as will
be seen, are heavily contested, and are not strictly
related to the number assigned by the producer of
the accounting sign.
4
The tedious literalism of Schoonmaker’s (1994) critique of
Baudrillard is an example of what can happen. The metaphor-
ical qualities of terms like ‘‘code” and ‘‘binary” completely
escape her.
C. Graham/ Accounting, Organizations and Society 33 (2008) 756–782 761
Data sources for a study of accounting signs
In basic outline, the Algoma Steel crisis and
bailout ran as follows. In 2000 and 2001, the com-
pany encountered ?nancial di?culties. It faced
severe cash ?ow problems, and its total liabilities
exceeded all its assets, even without including the
company’s obligations towards its pension plans
and other employee future bene?ts plans.
5
These
were non-contributory, de?ned-bene?t plans cov-
ering substantially all employees. They were
underfunded by almost $800M (Algoma Steel,
2001a). The size of the funding de?cit and its swift
growth alarmed investors and creditors. The com-
pany ?led for court protection, prompting the
Ontario government to step in. Eventually, an
agreement was reached between the province, the
company and its various stakeholders, in which
the company was restructured and a substantial
portion of the pension liability was assumed by
the government (Algoma Steel, 2001b).
In order to expose the function of accounting
signs in this series of events, a variety of related data
sets must be examined. The public discourse sur-
rounding the Algoma Steel pension plan incident
is one set. It supports an analysis of howaccounting
numbers were consumed and interpreted in this
contentious pension incident. This discourse is rep-
resented by articles fromCanadian newspapers dur-
ing 2001, when the crisis made headlines. The
Canadian Newsstand database was searched for all
references to Algoma Steel during 2001, and the rel-
evant articles
6
were read to determine howaccount-
ing signs were taken up and represented in the
media. This representation includes not only spe-
ci?c citations of individual accounting numbers,
but the connection made in the media between Alg-
oma’s accounting numbers and consequent political
and ?nancial developments in the case.
The history of pension disclosures made by Alg-
oma Steel is a second data set. This information
allows us to reconstruct the way the pension fund
was presented, such that it was transformed using
accounting signs from a politically negligible pri-
vate fund of wealth into a necessary target of gov-
ernment intervention. Annual reports for Algoma
Steel going back to the company’s 1991 restructur-
ing will be included; this cuto? permits
comparability by ignoring pre-restructuring signs
produced by the company.
Court documents from the bankruptcy proceed-
ings form a third data set. From six ?ling boxes
containing all the paper records from the proceed-
ings, 418 pages of information relevant to arrange-
ments with creditors were obtained. These pages
were selected by the author and a research assistant
by reviewing every document in the ?ling boxes and
pulling out everything substantive related to credi-
tor arrangements. The many routine documents
that form the legal paper trail, such as notices of
papers being served on witnesses, were ignored.
Documents available from other sources, such as
drafts of the ?nal agreement, were also ignored.
Many documents in the ?ling boxes were routinely
updated before the courts on a monthly basis or
better, such as statements from the bankruptcy
trustees. Of these, only one or two examples each
were retained. Key documents obtained from these
boxes included testimony from the Chief Financial
O?cer of the company, testimony from the court-
appointed mediator, and copies of letters sent to
key stakeholders towards the end of the settlement
negotiations.
Supplementing these data sets are the publicly
available Ontario government documents pertain-
ing to the bailout of Algoma Steel. These are avail-
able from electronic archives on the Ontario
government website. The documents include the
Pension Bene?ts Act (1990), related regulations,
various government news releases, and successive
revisions to the agreements developed between
stakeholders under the auspices of the govern-
ment’s pension intervention facilities.
5
Under Canadian regulations, future bene?ts to employees
are divided into two categories for both funding and accounting
purposes. Bene?ts corresponding to the notion of deferred
wages are properly called ‘‘pension bene?ts”, while other
bene?ts pertaining to extended health bene?ts, dental coverage,
and life insurance are given the unwieldy label ‘‘other employee
future bene?ts”. In this study, ‘‘pension plan” is used to refer
generally to both kinds of plans. Where necessary, ‘‘pension”
and ‘‘non-pension” future bene?ts are distinguished.
6
There were 416 articles on Algoma Steel in the database
from 2001. As many as half of them were duplicates or closely
related stories due to the use of wire services by multiple
newspapers.
762 C. Graham/ Accounting, Organizations and Society 33 (2008) 756–782
All the data are examined in the context of Sec-
tion 3461 of the CICA Handbook, which governs
the disclosure of future bene?ts for employees.
This permits us to understand the syntactic
features of the accounting signs, and to see how
the pension accounting signs derive their meaning
from their linguistic context. While interviews and
other local data would enhance the richness of the
study, the opportunity to pursue such sources will
be left for future research.
In order to understand the pension accounting
information within its full social context, other
data sources are used to establish the speci?c his-
torical antecedents of the pension incident. McDo-
wall’s (1984) narrative history of Algoma Steel has
been consulted to develop a description of the eco-
nomic history of the company and its impor-
tance to Ontario. Two other histories (Beatty &
Schachter, 2002; Nishman, 1995) were valuable
in describing the 1991 restructuring of the com-
pany. This restructuring led to the employee own-
ership situation that pertained at the time of the
‘‘pension crisis” in 2001.
The Algoma Steel bailout has been deliberately
selected because of its sheer magnitude and the
unprecedented size of the intervention by the pro-
vincial government.
7
This extraordinary event gen-
erated a rich availability of data. Because of this,
the case helps reveal social con?icts and account-
ing problems that may be latent, but are perhaps
undetectable, in many other pension funding cases.
Certain aspects of the Algoma Steel situation are
unusual, however. The steel industry presents
speci?c highly-institutionalized features: the indus-
try is heavily unionized, de?ned-bene?t pension
plans are prevalent, and wages are relatively high.
While workers surrendered future wages and
certain pension bene?ts during the incidents
examined here, the institutional support for these
workers nonetheless far exceeds that found in
many other areas of the Canadian economy.
Because of this, Algoma Steel represents in some
ways a ‘‘best case” scenario for Canadian work-
ers. Furthermore, this ‘‘crisis” primarily illus-
trates the social impact of pension accounting
when things go badly. This is important to note
because de?ned-bene?t plans in Canada create
asymmetries of risk and reward. Funding sur-
pluses typically accrue to workers but employers
are responsible for any funding de?cits.
8
Subse-
quent events have seen Algoma Steel enjoy a resur-
gence, exacerbating the asymmetries.
The Algoma Steel pension crisis and bailout
Algoma Steel is no stranger to crisis, having
gone through a series of collapses, rescues and
recoveries since its establishment at the turn of
the 20th century (McDowall, 1984, pp. 4–6). Given
the railway subsidies, land grants, special con-
tracts, tax breaks, and production bounties
Algoma received from the government during its
?rst 50 years (McDowall, pp. 33–38, 162, and
165), it is not farfetched to suggest that Algoma
owed its early existence to government support.
During the Second World War, government poli-
cies encouraging steel production stimulated an
expansion at Algoma that continued through
to the 1970s (McDowall, pp. 235–257). How-
ever, similar government policies in other coun-
tries in the 1980s led to global overproduction of
steel (Nishman, 1995, pp. 8–9), resulting in
substantial losses and a large debt load at Algoma.
Eventually, in 1991, Algoma Steel declared
insolvency and sought court protection from its
creditors.
The provincial government of the day was pro-
labour,
9
and was under political pressure to help
secure jobs. The federal government refused to
step in, leaving the provincial government to co-
7
Imperilled corporate pensions have been central features of
other major bankruptcy crises in Canada, notably Air Canada
in 2003–2004. However, what distinguishes the Algoma Steel
case is the Ontario government’s intervention speci?cally to
relieve the company of its pension liability.
8
Depending on the terms of the plan, funding surpluses may
be withdrawn by the company with the consent of workers.
This can happen, for example, as a result of labour negotiations
if the workers trade part of the surplus for other contractual
bene?ts.
9
The provincial government in Ontario in 1991 was formed
by the New Democratic Party, led by Bob Rae. The New
Democratic Party in Canadian politics has traditionally advo-
cated social democratic policies and enjoyed labour support.
C. Graham/ Accounting, Organizations and Society 33 (2008) 756–782 763
ordinate a task force involving major stakehold-
ers (Nishman, 1995, pp. 14–16). A mediated
settlement was reached based largely on the work-
ers’ proposal for a form of employee ownership.
10
Voting shares amounting to 60% of control of the
company were issued to an employee trust, with
the remaining shares publicly traded. In exchange,
workers agreed to a $200 million wage reduction,
bene?t reductions, and substantial job cuts (Beatty
& Schachter, 2002, pp. 30–31; Nishman, 1995, pp.
19–22).
This 1991 restructuring of Algoma Steel accom-
plished two important things. First, it established
the company as an employee-owned enterprise.
This development carried signi?cant implications
for the governance of the company. Management
in the new company was to be ‘‘independent”
(Nishman, 1995, p. 25), presumably meaning that
it was to operate independently of the majority
owners (the employees) in making strategic deci-
sions. Despite acquiring majority ownership and
a minority representation on the board, unionized
workers regarded the $200 million wage reduction
as a contract concession, not as a takeover move
(DaPrat, 2005a; Nishman, 1995, p. 25).
The second thing the 1991 restructuring accom-
plished was the entrenchment of the provincial
government’s role in ensuring the survival of the
company in times of crisis. Direct involvement
by both provincial and federal governments in
the establishment and growth of Algoma Steel
had already provided clear precedent for govern-
ment intervention in the company. The 1991 pro-
vincial government intervention rea?rmed this
line of action, and underscored the importance of
Algoma Steel to the Ontario economy. Hence, by
the time the pension crisis of 2001 arose, the com-
pany had been ?rmly established as a site of gov-
ernment action.
In the mid-1990s, management embarked on a
major capital expansion, the construction of a
large new production complex, and initiated a
major re?nancing of the company. Management’s
?nancial projections for the project e?ectively cir-
cumscribed the choices available to the union.
According to the union president, the employees
believed they would inevitably lose control of
the company if the production facility went ahead
(DaPrat, 2005a). The employees therefore entered
into new contractual negotiations with the com-
pany, and converted most of their 60% interest
in the company into improved pensions and other
bene?ts, retaining only 20% ownership (DaPrat,
2005a). Management then issued $500 million of
?rst mortgage notes to acquire funds for the con-
struction project (Algoma Steel, 1996, p. 4; 1997,
Note 7; 2000, Note 6). The holders of these notes
became one of the in?uential stakeholder groups
in the 2001 bankruptcy negotiations (Keenan,
2001).
This new capital structure generated impor-
tant consequences for the company. For exam-
ple, while ?xed assets rose from $323 million
in 1994 to $932 million in 1997, denoting the
construction of the new complex (Algoma Steel,
1996, p. 7; 1997, p. 24), the conclusion of con-
struction meant that expenses the company
had been capitalizing now hit the income state-
ment. The new ?xed assets began to be amor-
tized, including the interest on the ?rst
mortgage notes capitalized during construction
(Algoma Steel, 2001a, p. 26). In addition, a
weakening Canadian dollar created foreign
exchange losses for the company, as the ?rst
mortgage notes were denominated in US dollars.
The company reported large losses beginning in
1998. By 2000, the company had run out of
cash and was staying a?oat through short-term
borrowing. The company blamed its situation
on declining sales resulting from oversupply, a
reduction in demand, and the dumping of for-
eign steel onto the Canadian market (Algoma
Steel, 2000, p. 2).
10
The settlement converted common shares held by Dofasco
and the bonds held by most creditors into preferred shares.
(Dofasco, a major competitor in the Ontario steel industry, was
at this time majority owner of Algoma.) In the process,
Dofasco’s claim on Algoma dropped from $224 million to $69
million. O?setting this, Dofasco picked up a useful $150 million
tax loss and was exempted from pension liabilities. The
employees’ shares carried special rights to approve any funda-
mental changes to the company, and employees could elect a
minority (5 of 13) of the company’s directors. The provincial
government provided $110 million in loan guarantees, reduced
railway freight rates for the company, and waived environmen-
tal liability for the creditors and preferred shareholders.
764 C. Graham/ Accounting, Organizations and Society 33 (2008) 756–782
The 2001 pension crisis at Algoma Steel
On April 23, 2001, the company sought court
protection. In an a?davit ?led by the company
on this date (Ontario Superior Court, 2001a), the
Chief Financial O?cer blamed the company’s pre-
dicament on the dumping of foreign steel, as
before, but now also blamed Algoma’s ?nancial
expenses, and the lengthy implementation of its
new production complex (p. 4). The CFO o?ered
many accounting representations as evidence of
the company’s plight. He provided as exhibits the
unaudited ?nancial statements as at March 31,
2001, as well as a copy of the company’s December
31, 2000 annual report. In the body of his a?davit,
he drew attention to the company’s net losses for
these two dates, as well as its total assets and prin-
cipal debts for March 31, 2001. Included in the
principal debts were the company’s credit facility,
the ?rst mortgage notes, its trade payables, wages
and deductions payable, accrued vacation pay,
and various pension debts and post-retirement
bene?ts obligations. These debts are shown in
Table 1, taken verbatim from the a?davit. What
is interesting is that the a?davit lists all the
accounting numbers except the pension-related
ones as they appear on the ?nancial statements,
that is, under a going-concern assumption. The
pension liabilities, however, are shown on a
wind-up basis.
The impact of this di?erence in treatment is dra-
matic. Pension debts total $769 million, instead of
the $428.4 million shown in the March 31, 2001
?nancial statements (Ontario Superior Court,
2001a, Exhibit A). They amount to approximately
half the listed ‘‘principal debts”. Although most of
these numbers were initially reported in the press
(e.g., Van Alphen, 2001a), in later media reports
it was the pension shortfall that was highlighted
(e.g., Algoma Steel’s $625-M Pension Shortfall,
2001; Daw, 2001; Van Alphen, 2001b). In part,
this was due to the immense size of the pension
number relative to Algoma’s other accounting
numbers, such as equity ($236 million as at the
2000 year end) and even annual revenue ($1106
million).
There are several other possible reasons why so
much attention was given to the pension aspect of
Algoma’s situation. Elderly pensioners and older
workers were perceived as vulnerable. Algoma’s
average worker was around 50 years of age, with
upwards of 25 years of service; approximately 600
of themwere due or able to retire in the next 3 years
(DaPrat, 2005b). To a certain extent, moves to ‘‘res-
cue” workers and pensioners would be received pos-
itively (Steelworkers Hail Appointment, 2001). In
contrast, a direct intervention on behalf of the com-
pany as a whole would have been interpreted as tak-
ing sides against competitors who were equally
important to the Ontario economy.
However, this does not fully explain the pension
?xation of the media. I would argue that the pen-
sion numbers drew attention because they were
easily repositioned and recon?gured in ways that
allowed them to take on new meanings. That is,
even though the pension plans had not changed
materially, the pension disclosures could arbi-
trarily be given new meanings just by rearranging
the signi?ers. This was not as easily done with
the other accounting numbers o?ered by the
CFO in his a?davit.
11
Table 1
Froma?davit of Algoma Steel’s Chief Financial O?cer (Ontario
Superior Court, 2001a, p. 5): Algoma’s principal debts (on an
unconsolidated basis) as at March 31, 2001 are as follows
Cdn. Millions
(a) Credit facility $146
(b) Notes
(i) Principal amount (US $349 million) $551
(ii) Accrued interest $14
(c) Trade payables $41
(d) Wages and employee deductions payable $13
(e) Accrued vacation pay $36
(f) Other
(i) Accrued pension liability on a
wind-up basis (estimated)
$503
(ii) Pension plan indexing obligations
on a wind-up basis (estimated)
$121
(iii) Accrued post-retirement bene?t
obligations (estimated)
$145
11
Pension disclosures are arguably the most easily recon?g-
ured ?nancial accounting signs, but they are not categorically
distinct. The increasing emphasis on fair value in accounting
standards, particularly for ?nancial instruments, suggests that it
will only become easier for managers to rearrange other
signi?ers at will.
C. Graham/ Accounting, Organizations and Society 33 (2008) 756–782 765
The speci?c con?gurations of pension plan sig-
ni?ers that were used will be described below when
we review Canadian pension accounting disclosure
standards. For now, I want to focus on the signi?-
cance of this facility with pension signi?ers. For
this is where Baudrillard’s insight about the arbi-
trariness of signs comes into play. It is not that
the sign (e.g., the stated amount of the company’s
funded future obligation to employees) is arbi-
trarily connected to the referent (e.g., the future
cash ?ows when those obligations are paid).
Rather, it is that the signi?er (e.g., pension liabil-
ity = $X) is arbitrarily connected to the signi?ed
(the notion that the company has a future obliga-
tion to its employees). As the signi?ers are
rearranged, new signi?eds emerge from the rela-
tionships amongst the signi?ers. The notion of
the company’s obligation to its employees, how
large it is, how well or poorly funded it is – these
are all signi?eds that change as the signi?ers are
rearranged.
12
Yet because stakeholders react to
this notion in complex ways, exercising varying
degrees of power at the moment of consuming
the sign and thus shaping both the meaning of
the sign and the outcome of the restructuring pro-
cess, the referent is never independent of the signi-
?ed. The sign precedes the referent, and shapes it.
Hence, as the company changes the signi?er it uses,
the referent is deliberately altered. The company
could not completely control the reactions of the
stakeholders, but it could and did attempt to
impose boundaries on how its obligations could
be understood. Anything vague, anything warm
and fuzzy about its long-term obligations to its
aging workforce, was given precision by the sign
that denoted it. As Baudrillard repeatedly argued
(e.g., Baudrillard, 1981, p. 149), the sign is not
equivocal and ambiguous like a symbol is, it is
univocal; the production of the sign is an attempt
to exclude unwanted interpretations. The struggle
for Algoma Steel’s future hinged on the meaning
of these signs, and control of the ability to produce
them was a crucial advantage for management, an
asymmetry that was institutionalized in the market
for ?nancial information despite the partial owner-
ship position of the workers and the huge claims on
the company by the notesholders.
However, as Baudrillard noted about the com-
modity, it is not the moment of production that
is meaningful but the moment of consumption.
So for the accounting sign. Management con-
trolled the production of the signs, but not their
consumption. Hence, we must look closely at the
mode of consumption of this information: the
moment of consumption of accounting informa-
tion is highly structured, and is simultaneously a
moment of reproduction.
The consumption of accounting signs in the
Algoma Steel ‘‘crisis” involved two important
recon?gurations. First, the signs were discon-
nected from their traditional location in audited
?nancial statements prepared for the ?nancial
markets, and re-embedded in legal documents.
Although the accounting numbers came from the
same accounting systems that produced the
audited ?nancial statements, the company dis-
pensed with the imprimatur of its auditors and dis-
closed unaudited accounting information under
the authority of the legal system, that is, by having
its CFO state the numbers under oath. Note that
this pertains to all the accounting numbers in the
a?davit, not just the pension ones. The e?ective-
ness of this redeployment by the company there-
fore contradicts Macintosh et al. (2000, p. 41),
who maintained that the sign value of accounting
statements about capital and income derives from
their certi?cation by public accounting ?rms. The
Algoma Steel a?davit was not audited. The
accounting numbers it contained functioned as
signs under two related systems of signi?cation.
As accounting signs, they helped present Algoma’s
situation dispassionately as a technical, calculative
problem amenable to technical, calculative solu-
tions. As signs embedded in the legal sign of ‘‘a?-
davit”, they took on an urgency that attracted
both media and government attention. For
instance, they served to reveal to the premier of
12
Note that multiple possible signi?ers interact with multiple
possible signi?eds. There are many ways of signifying the
obligation/liability, and each one has multiple connotations.
This does not change the logic of the sign form (Baudrillard,
1981, pp. 149–150). The production of signs is about the
attempt to impose univocal meanings, not about simplicity.
766 C. Graham/ Accounting, Organizations and Society 33 (2008) 756–782
the province both the necessity and the possibility
of government intervention (cf. Perkel, 2001).
13
The second and simultaneous recon?guration
of Algoma’s accounting signs was particular to
the pension numbers. To function so well as an
alarm, the pension shortfall had to be both a
departure from prior signs produced by the com-
pany, and yet consistent with those prior signs. If
it was not a departure, if it was not ‘‘new”, it could
not have attracted attention. But if it was not
somehow consistent, it would not be believable.
In other words, the system of signs must permit
a new con?guration of signi?ers that is utterable
and understandable, yet previously unspoken.
From an accounting perspective, this means
that the pension numbers released in the a?davit
had to match prior disclosures under Canadian
GAAP, yet Canadian GAAP had to have permit-
ted the new message to have remained unspoken
until this time. To understand how this could be,
it is necessary to review what disclosures Canadian
GAAP requires and permits regarding company
pension plans.
Pension accounting disclosures under Canadian
GAAP
Accounting for pensions in Canadian GAAP is
governed by Section 3461 of the CICA Handbook,
which deals with ‘‘Employee Future Bene?ts”.
Subsection 3461.150 states that the objective of
the given disclosure requirements is to provide
?nancial statement users with information about
the e?ect of employee future bene?ts on the ?nan-
cial statements, and about the obligations and
assets related to de?ned-bene?t plans such as Alg-
oma Steel’s. The subsection also states that the
information is to be useful in understanding the
costs, risks and uncertainties related to the com-
pany’s pension plan obligations. Users should be
able to use the information to make investment
(‘‘resource allocation”) decisions and assess man-
agement’s stewardship. A distinction is made in
the disclosure guidelines between ‘‘pension bene-
?ts” and ‘‘other employee future bene?ts”. Com-
panies in Canada are legally required to manage
these two classes of bene?ts in separate plans,
and Section 3461.151 requires that they be dis-
closed separately.
14
The large shortfall in Algoma’s pension and
post-retirement bene?ts plans developed quickly
during 2000 and 2001. This had been set up by
two decisions by management during the 1990s.
The ?rst was the apparent decision to invest the
company’s pension fund assets in equity markets,
‘‘apparent” because it is di?cult to know from
the accounting disclosures how much of the assets
were invested in this way, and in which speci?c
stocks. The company’s ?nancial statements indi-
cate that the actual return on plan assets plum-
meted in ?scal 2001 from a 5-year average of
$123 million per annum to a negative return of
$18 million.
15
These results were typical of many
North American companies that had heavily
invested their pension assets in equities during
the expansive markets of the late 1990s and were
exposed when the dot-com bubble burst in 2000
(Fore, 2004; Joss, 2004).
16
The relative amount
of Algoma’s investment loss was not huge, only
2% of assets, and was o?set by continued contribu-
tions to the plan. However, the reversal of fortune
in the markets was dramatic, and would have had
13
Furthermore, news articles from this period reinforced the
legitimacy of government intervention by recalling the role of
the Ontario government in resolving the company’s previous
‘‘crisis” in 1991 (e.g. Ottawa Provides Loan Guarantee, 2001;
Perkel, 2001; Van Alphen, 2001; Watson, 2001).
14
Algoma’s a?davit made a further distinction, listing
pension indexing liabilities separately from pension obligations
and other future bene?ts obligations. Indexing provisions of
corporate pension plans were not protected by the Ontario
government. See the section on the Pension Bene?t Guarantee
Fund below.
15
While the equity markets plummeted in 2000, a negative
‘‘actual return on plan assets” ?gure did not appear until
Algoma’s 2001 ?nancial statements. The recognition of the
negative return may have been delayed due to timing of the
asset valuation, although this should not have been the case,
since Section 3461 (Exhibit I of Example 1, Situation I) implies
that the 2000 calculation should have been based on the fair
value of plan assets on Algoma’s year-end date of December 31,
2000. Algoma, unfortunately, did not disclose how it arrived at
its ?gures.
16
Algoma’s key competitors, Stelco and Dofasco, stated
similar changes in return on pension assets in the same period.
Dofasco’s return dropped from $86.5 million in 2000 to $17.9
million in 2001 (Dofasco, 2001). Stelco’s return dropped from
$317 million to $11 million for the same years (Stelco, 2001).
C. Graham/ Accounting, Organizations and Society 33 (2008) 756–782 767
abrupt consequences for pension funding decisions
if the company had not ?led for court protection
(cf. Fore, 2004, p. 7).
The tentativeness of the statements in the pre-
ceding paragraph is quite deliberate. The impossi-
bility of being more conclusive points to a speci?c
problem in Canadian GAAP. The extent to which
plan assets may have been invested in dot-com
stocks, for example, remains unknown because
Canadian GAAP in 2001 did not require such a
disclosure,
17
and Algoma’s management chose
not to exceed GAAP requirements. This left the
accounting signs that were produced bereft of a
meaningful context. Like a word uttered in isola-
tion rather than as part of a sentence, these signs
wanted the context of other signs—such as
asset allocation—that would enrich their meaning.
The second management decision leading to
Algoma’s sudden pension shortfall was the choice
of a high discount rate for future bene?ts during
the 1990s. This policy had to be abandoned when
the company adopted Section 3461. These new
standards came into e?ect on January 1, 2000,
the beginning of Algoma Steel’s ?scal year, and
replaced Section 3460, which had covered pension
costs and obligations.
18
According to Accounting
Standards Board of Canada (2002) and Estey
(1999), the purpose of Section 3461 was to rectify
a perceived de?ciency in pension cost accounting,
and to harmonize this section of the handbook
with US GAAP. The new section changed the dis-
count rate used to evaluate future cash ?ows from
a pension plan. Previously, reporting entities could
use a rate determined by management’s best esti-
mate of the long-term rate of return on the pension
fund’s assets. The new standard required discount-
ing either at the market interest rates on high
quality debt instruments whose cash ?ows approx-
imated the plan’s cash ?ow, or at the rate implicit
in whatever present amount could be paid to settle
the plan’s obligations (Accounting Standards
Board of Canada, 2002; Estey, 1999). The new
standards also speci?ed that an entity should mea-
sure plan assets at fair value; compute the expected
return on plan assets using a long-term rate of
return on the market value of assets; match the
cost of bene?ts to the periods in which the
employee earned them, up to the date at which
the employee became entitled to receive them; rec-
ognize actuarial gains and losses using the ‘‘corri-
dor” approach (discussed below and in
Appendix); and observe new guidance on plan set-
tlements, curtailments, terminations, accrued ben-
e?t assets, and multi-employer plans. Signi?cantly,
the new standards required all these pension
accounting guidelines to be applied to accounting
for non-pension future bene?ts, which had previ-
ously been treated on a cash basis (Estey, 1999).
As a result of adopting these standards, in 2000
and 2001 the Algoma Steel plan endured dramatic
unrecognized actuarial losses, in stark contrast to
its prior pattern of unrecognized actuarial gains.
The glossary of Section 3461 de?nes actuarial
gains and losses in a de?ned-bene?t plan as
changes in the value of the accrued bene?t obliga-
tion and plan assets resulting either from experi-
ence being di?erent than assumed or from
changes in actuarial assumptions. In the case of
Algoma Steel, the key assumption that changed
was the discount rate selected by management to
calculate the present value of the company’s future
bene?t obligations. In adapting to the new GAAP
guidelines, management shifted these rates from
the steady 8% it had used since the 1991 restructur-
ing, to 7% in 2000 and 6.5% in 2001.
19
This was
precisely the e?ect of Section 3461 that had been
predicted in the Canadian accounting profession’s
17
Even now, Section 3461.155(b)(ii) requires only that a
company disclose its asset mix in the broadest terms, by
showing the percentage of assets held in general categories such
as equity securities, debt securities and real estate. More
detailed information is contained in actuarial valuation reports
?led with federal and provincial pension authorities. These are
available to plan members upon request, but are not available
to the general public.
18
Comments in this paper regarding Section 3461 are not
intended to suggest that Section 3461 is either ‘‘better” or
‘‘worse” than Section 3460. By o?ering speci?c critiques of
Section 3461, I do not mean to suggest that the social impacts
of pension accounting in the Algoma Steel case would not have
occurred, or would have been more felicitous, under Section
3460. They would simply have been di?erent. As Macintosh
et al. (2000, p. 16) note, Baudrillard’s ‘‘orders of simulacra” do
not imply sequential improvement.
19
These were the rates used for the pension obligation. For the
much smaller non-pension future bene?ts obligation, the rates
were 7% for 2000 and 6.75% for 2001.
768 C. Graham/ Accounting, Organizations and Society 33 (2008) 756–782
leading practitioner magazine (Tishco?, 1999).
These seemingly small adjustments in discount rate
caused unrecognized actuarial losses on non-pen-
sion future bene?ts to swell from a 5-year average
of $14 million to $75 million in 2001. Even more
dramatic was the e?ect on the pension side, which
had enjoyed unrecognized actuarial gains averag-
ing $115 million per year from 1994 to 1999. These
gains disappeared in 2000, and became an unrec-
ognized actuarial loss of $267 million in 2001.
While random ?uctuations in mortality or morbid-
ity experience may have contributed to this nega-
tive result, a shift of this magnitude was more
likely caused by the change in discount rates; pro-
jected bene?ts calculations have been shown to be
highly sensitive to changes in discount rate (Duf-
resne, 1993).
20
Again, it is impossible to be certain
here: Algoma did not disclose any details about its
actuarial losses, nor did Section 3461 require it to
do so.
21
Actuarial calculations underlying the pension
accounting disclosures informed the funding deci-
sions that produced eventual funding shortfalls.
These endogenous e?ects can be seen in Fig. 1.
In this chart, each column indicates the total fund-
ing shortfall for pensions and other post-retire-
ment bene?ts in a given year. This shortfall is
called ‘‘Funded status” in the notes to the ?nancial
statements. However, in Algoma’s notes (follow-
ing CICA Handbook guidelines) the amount for
pensions and the amount for other post-retirement
bene?ts are never added together as they are here.
Each column in the chart is composed of several
smaller amounts, including the on-balance sheet
‘‘accrued liability” and the o?-balance sheet
‘‘unrecognized actuarial gains and losses”. The
chart demonstrates that prior to 2000, actuarial
gains o?set the accrued liability substantially;
thereafter actuarial losses exacerbated the liability.
During this time period Algoma Steel’s man-
agement adopted a curious format for disclosing
the pension plan’s funding status. The CICA
Handbook’s illustrative examples for such disclo-
sures are fairly straightforward. Nonetheless,
Algoma chose a more convoluted presentation.
Table 2 shows the funding status disclosures con-
tained in Note 10 to Algoma’s 2001 ?nancial
statements.
Rather than breaking the plan’s funding status
into its components, the Algoma disclosures tell
the story of how the accrued liability portion of the
pension plan obligation is calculated. They begin
with large negative numbers and o?set them with
a series of mainly positive numbers, to arrive at a
smaller negative number. The implication is that
the reader should focus on this smaller number.
The disclosure is arrangedtonot toexplainthe over-
all obligation, but to explain it away. The reader is
guided to pay attention only to the portion of the
pension obligation that has been recognized as a lia-
bility by the company, the portion that appears on
the balance sheet. The overall pension obligation
appears less signi?cant—has a diminished sign
value—and diminished social consequences. This
reinforces the prior e?ect of relegating the pension
obligation disclosure to the notes.
In contrast, the examples given in Section 3461
show the funding status as the di?erence between
plan obligations and plan assets. If this format
had been followedby Algoma, the notes wouldhave
included a table such as Table 3. Note that in Table
3, it is clear that the funding status is the amount
being explained. That is, the plan de?cit is ‘‘the bot-
tom line”. Yet even in this format recommended by
the Handbook, the total obligation for pension and
non-pension bene?ts is never stated.
20
Algoma was not alone in its experience with the new
standards. One of Algoma’s key competitors, Stelco, dropped
its discount rate from 8% in 1998 and 1999 to 7% in 2000, and
to 6.75% in 2001; its actuarial gain of $15 million on all de?ned-
bene?t plans in 2000 changed to a loss of $181 million in 2001
(Stelco, 1999, 2001). Algoma’s other key competitor, Dofasco,
actually raised its discount rate from 6.75% in 2000 to 7% in
2001; its actuarial losses on all de?ned-bene?t plans dropped
from $72.8 million in 2000 to $34.9 million in 2001 (Dofasco,
2001). The widespread uncertainty caused by the changes in
discount rate mandated by Section 3461 prompted the AcSB
(2002) to issue a statement on ‘‘Measurement Uncertainty and
Employee Bene?t Plans”.
21
The impact of the adoption of Section 3461 by Algoma was
clouded by the simultaneous adoption of other new accounting
standards. The e?ect of Section 3461 on Algoma’s bottom line
was a decrease in retained earnings of $40 million. However the
concurrent adoption of new income tax accounting standards
simultaneously increased retained earnings by $81 million
(Algoma Steel, 2000, p. 22).
C. Graham/ Accounting, Organizations and Society 33 (2008) 756–782 769
Algoma Steel’s disclosure thus appears designed
to downplay the company’s obligation for
employee future bene?ts, focussing attention
instead on the accrued bene?t liability. Yet the
accrued bene?t liability is the most heavily ‘‘con-
structed” number in Section 3461’s implicit model.
It is the one number in the model that is least
directly representative of any underlying objective
reality that the model might purport to represent.
(Readers unfamiliar with Section 3461 are referred
to Appendix.) The model stipulates that the
accrued bene?t liability is the di?erence between
the accrued bene?t obligation and the aggregate
of the plan assets and any unamortized costs.
Thus, any unamortized costs are e?ectively
deducted from the funding status to determine
the accrued liability shown on the balance sheet.
The implicit model therefore constrains the entity
from fully disclosing the funding status of its pen-
sion plan in audited statements.
E?ects of the model in the Algoma Steel pension
crisis
The express intent of the model in slowly
amortizing pension costs is to dampen the random
-$200
$0
$200
$400
$600
$800
$1.000
1997 1998 1999 2000 2001
Total accrued retirement obligation
Contributions and payments after measurement
date
Unrecognized prior service costs
Unrecognized net retirement obligation
Unrecognized actuarial losses (gains)
Fig. 1. E?ect of actuarial gains and losses on Algoma Steel pension plan obligation. Source: Algoma Steel Inc. annual reports, 1997–
2001.
Table 2
Disclosure of funded status
Reconciliation of pension funding status 2001 2000
Funded status (552) (287)
Unrecognized actuarial losses (gains) 267 3
Employer contributions
after measurement date
3 3
Unrecognized net pension obligation (2) (3)
Unrecognized prior service costs 6 7
Accrued pension liability (278) (277)
Reconciliation of post-employment
funding status
2001 2000
Funded status (245) (196)
Unrecognized prior service costs 0 0
Unrecognized actuarial losses (gains) 75 32
Bene?ts paid after measurement date 1 1
Unrecognized net bene?t obligation 12 12
Accrued post-employment
bene?t obligation [sic]
(157) (151)
Source: Algoma Steel Inc., 2001 ?nancial statements, Note 10.
770 C. Graham/ Accounting, Organizations and Society 33 (2008) 756–782
short-termnoise that is inherent in the mortality and
morbidity experience of the plan, and in the market
price of the ?nancial assets of the plan. Tothis extent,
the model is arguably useful. Short-term noise, the
argument runs, can have an unwarranted impact
on the funding status of the plan, and since the
expected value of the noise is zero over time, the
model should smooth out these short-term e?ects.
However, the unwarranted impact of certain
‘‘noise-related” parameters arises from the model
itself. Market-based parameters like ?nancial securi-
ties prices and the discount rates mandated by Sec-
tion 3461 a?ect the two main aggregate values in
the model, which are the fair value of plan assets
and the accrued bene?t obligation. Funding status
is simply the di?erence between these two values. If
these large values are ‘‘normally” roughlyequal, then
a relatively small percentage change in one of them
can cause the ‘‘normally” small di?erence to grow
by many multiples of itself. Yet changes to these
main aggregate values are not always small, being
so sensitive to some of the model’s parameters. In a
sense, therefore, by implementing extensive smooth-
ing, the model is protecting its users from itself.
The argument in favour of smoothing, as frail
as it is when parameters take random walks, is par-
ticularly ?awed when it comes to non-random ?uc-
tuations in plan parameters. In the case at hand,
the change in discount rates mandated by Section
3461 was singular and non-random. The impact of
the adoption of Section 3461 at Algoma Steel,
given management’s choice in prior years of high
discount rates that were no longer permissible,
was to drive up the pension obligation. This hap-
pened at the very time plan assets were falling in
value. Yet because the net change in these numbers
was classi?ed under the model as an actuarial loss,
it was metered out to the balance sheet only
through the corridor rule, in tiny increments and
a year delayed. Thus, a mechanism designed to
smooth random ?uctuations was being used to
cope simultaneously with a permanent restatement
of plan assets and the structural adjustment of the
market to plummeting dot-com valuations. Subse-
quent random ?uctuations in experience and the
market cannot be counted upon to undo this.
The signi?cant changes in Algoma’s pension
obligation that arose in 2001 resulted in two di?er-
ent sets of accounting signs being produced. When
the company ?led for bankruptcy, one set of signs
was produced that highlighted the pension funding
shortfall. At the year end, after the ‘‘crisis” had
been resolved, a second set of signs was produced
that covered the same period of time but presented
a much di?erent picture. In the second set of signs,
contained in the company’s annual report, the over-
all pension plan obligation was relegated to the
notes of the ?nancial statements, as mandated by
the CICA guidelines. Thus, Algoma’s management
disregarded the demand for accounting informa-
tion that was evident in the way the media took
up the pension obligation ?gure when the crisis ?rst
broke.
22
This illustrates how the production of
accounting signs is dependent on the conditions of
their consumption. At the ?scal year end, Algoma
management no longer addressed the court, from
whom it had sought protection, but the share-
holder, from whom it sought con?dence. The Man-
agement Discussion and Analysis in the annual
report was thus well within the requirements of
Canadian GAAP, but entirely unresponsive to
other stakeholder needs, when it failed to mention
the enormous actuarial losses. In fact, the MD&A
stated only that during the year the accrued pension
liability had risen from$277 million to $278 million.
The implicit model of Section 3461 accom-
plished two contradictory things for Algoma Steel.
First, it constrained evidence of volatility for years,
only to release it in a torrent when discount rate
regulations changed. Second, it was delivered into
the hands of management two ready sets of ?gures,
one for minimal recognized portions of future lia-
22
Congruence between media attention and the interests of
other stakeholders, such as investors, should not be taken for
granted.
Table 3
Format for disclosure of funded status suggested by Section
3461 of CICA handbook
Pension
bene?ts
Other
bene?ts
2001 2000 2001 2000
Fair value of plan assets 1018 1101 11 10
Accrued bene?t obligation 1570 1388 256 206
Funded status – plan de?cit (552) (287) (245) (196)
C. Graham/ Accounting, Organizations and Society 33 (2008) 756–782 771
bilities and another for their greater unrecognized
whole. All the accounting signs in the a?davit were
taken straight from Algoma’s accounting system,
but assets and non-pension liabilities had only
going-concern estimates associated with them.
The pension numbers, in contrast, were easily
turned on their heads to show the (un)funded sta-
tus of the plans. These numbers could be said to
represent the wind-up value of the pension plans
without further work being necessary.
Structured consumption of signs and the resolution
of the crisis
Algoma Steel’s disclosure decisions and the
underlying calculative techniques of Section 3461
must be evaluated in terms of the network of social
relations in which Algoma Steel was embedded.
Certainly, the market for Algoma Steel’s securities
was impacted by the calculations, at least to the
extent that the calculations were inseparable from
the 2001 restructuring process. According to Tor-
onto Stock Exchange archival records (available
on MSN.com), the price of Algoma shares fell from
$0.40 to as low as $0.22 on April 23, 2001, when the
company ?led for court protection. After a brief
rebound, prices declined from $0.35 in late April
to $0.25 in mid-October, and plummeted to the
$0.10 range after the restructuring plan was
announced on October 24, 2001. However, the mar-
ket was but one of the social relations pertaining to
Algoma Steel. Algoma’s pension accounting also
a?ected relations between Algoma and its workers,
and between both these parties and the government.
With the workers, Algoma was engaged in a
rather circular exercise involving capital forma-
tion. As discussed above, when the 1991 restruc-
turing took place, the workers gained control of
60% of the company. While this stake was reduced
to 20% in 1994–1995, the relevant governance
structures remained unchanged. Management
was considered independent, and the workers were
both owners and employees. The relationship
between management and workers-as-owners was
mediated by the board of directors and the trust
that held the employee shares. The relationship
between management and workers-as-employees
was mediated by the union, but only partially,
because not all employees were union members.
Because of the circularity of the company–
worker relationship, it is inappropriate to draw
on facile explanations of the disclosure choices
made by management regarding the pension plan.
Since workers had representatives on the board,
one would have expected them to have full access
to information. However, management did not dis-
close all that it could, nor indeed all that it was
required to under the company’s articles of incor-
poration. Under cross-examination during the
court proceedings, the company’s ‘‘Chief Restruc-
turing O?cer” said that despite the articles specify-
ing that the union’s representative was to receive all
Board material, the company had not divulged any
‘‘con?dential” information to the union regarding
the restructuring. The workers were not given any
information about the restructuring that was not
shared equally with other stakeholders (Ontario
Superior Court, 2001b, pp. 6–8). In understanding
the situation they were in, the workers were depen-
dent on management and upon the expertise of the
accountants hired by management. The result was
a passive acceptance of the company’s representa-
tions: when asked why the union was not more
vocal during the events of 2001, a present union
leader said, ‘‘Well, management told us the com-
pany was going bankrupt” (DaPrat, 2005a).
Algoma’s adoption of the new pension account-
ing guidelines in 2000 triggered a change in its rela-
tionship with the provincial government. With the
removal of the latitude to nominate a discount rate,
management was unable to continue constructing
its preferred picture of a healthy pension plan.
When plan assets failed to deliver the large returns
of the 1990s, and the pension obligation ballooned
due to the lower discount rate, the pension account-
ing model e?ectively constructed a new reality for
the company. Algoma was now a company with a
severely underfunded pension plan. This was the
handhold the provincial government needed in
order to intervene in the company once more.
The pension bene?ts guarantee fund
What made it feasible for the government of
Ontario to intervene was its possession of the
772 C. Graham/ Accounting, Organizations and Society 33 (2008) 756–782
requisite technologies for intervention. In 1990, the
provincial Pension Bene?ts Act had established in
Ontario the Pension Bene?ts Guarantee Fund
(PBGF) to protect workers from the risks of
private company pension plans. This fund covered
pension bene?ts, including worker pensions, spou-
sal pensions, and plant-closure bene?ts, but did not
cover non-pension post-retirement bene?ts such as
health care. The fund was ?nanced by annual con-
tributions from employers with pension plans. By
2001, however, the fund had accumulated only
$200–240 million in assets, less than half the
amount of Algoma Steel’s pension de?cit (Daw,
2001; Keenan, 2001; Province of Ontario, 2001).
The fact that the PBGF was nevertheless used
in the Algoma Steel bailout suggests that when it
comes to e?ecting the goals of government, it is
not necessarily the amount of funding that mat-
ters, but the existence and availability of appropri-
ate technologies. Embedded in the PBGF were a
set of accounting technologies that enabled the
government to act. Using these technologies, the
Financial Services Commission of Ontario, which
administers the fund, was able to collect, track,
and enforce contributions from employers. The
funds collected were aggregated and invested.
PBGF assets were invested in cash equivalents
and bonds (Province of Ontario, 2001), not in
stocks and other riskier securities. The fund was
designed to be supplemented by loans from the
provincial government when necessary. The PBGF
received pension plan assets in exchange for
absolving companies of a limited portion of their
pension obligations. The liability assumed by the
PBGF was for the amount of pensions earned by
Ontario-based employees, excluding most escalat-
ing or indexing provisions.
23
Under the terms of the Pension Bene?ts Act,
plans with over $500 million in assets were able
to ?le for exemption from certain funding solvency
regulations (Pension Bene?ts Act, 1990, s. 5 & s.
37). This privilege was withdrawn (cf. Reg. 203/
02, 2002, s. 1) shortly after the Algoma plan bail-
out, but Algoma had quali?ed for this exemption
and hence had not been observing the funding sol-
vency regulations. Nonetheless, the Algoma pen-
sion plan was covered by the PBGF.
The existence of the PBGF was crucial to the
government’s ability to consume the accounting
signs that had raised the alarm about the Algoma
Steel pension plan. While the meaning of any sign
depends on the conditions of its consumption, for
accounting signs those conditions are highly tech-
nical. This distinguishes accounting from other
languages. The PBGF, as noted, was not in a posi-
tion to accept the entire pension obligation of Alg-
oma. However, under the Pension Bene?ts Act and
the terms negotiated in the 2001 settlement, the
PBGF assumed not only most of the obligation,
but most of the plan assets. Hence, the net new lia-
bility taken on by province of Ontario, backed by
the assets of the PBGF and any loans to the PBGF
that would be necessary, was limited to the fund-
ing de?cit, and was further reduced by the deletion
of indexing and other bene?ts under the negotiated
settlement. Given that the existing plan assets
would, in the event of Algoma Steel’s complete
failure, have gone to settle at least a portion of
the pension obligation, the intervention of the gov-
ernment through the PBGF must be evaluated on
the basis of the incremental bene?t provided to the
pensioners, plus the bene?ts realized by other
stakeholders.
Terms of the 2001 plan of arrangement
The incremental bene?t provided to workers
and pensioners was mixed. While the pension plan
was saved from default, the plan assets would,
without government intervention, have provided
65% of the accrued pension obligation at the end
of 2001 (56% of the total future bene?ts obliga-
tion).
24
Under the terms of the bailout, the existing
pensions of retirees were protected, but as noted
23
The PBGF liabilities excluded the following: ‘‘(a) any
escalated adjustment, (b) excluded plant-closure bene?ts, (c)
excluded permanent layo? bene?ts, (d) special allowances other
than funded special allowances, (e) consent bene?ts other than
funded consent bene?ts, (f) prospective bene?t increases, (g)
potential early retirement window bene?t values, and (h)
pension bene?ts and ancillary bene?ts payable under a quali-
fying annuity contract” (Pension Bene?ts Act, 1990, s. 2).
24
At the beginning of 2001, plan assets would have provided
even more: 79% of the accrued pension obligation (70% of the
total future bene?ts obligation). These calculations are derived
from Algoma’s 2001 annual report.
C. Graham/ Accounting, Organizations and Society 33 (2008) 756–782 773
the Pension Bene?ts Act did not protect any index-
ing of these pensions. Bene?ts not guaranteed by
the PBGF were moved over to a new supplemen-
tary plan to be funded by Algoma Steel, with
whatever risks that entailed.
The technical capacities of the PBGF a?ected
the provincial government’s ability to respond to
the Algoma Steel crisis, and were an important
factor in how Algoma’s accounting signs played
out in court. However, the consumption of
accounting signs in the legal proceedings was not
shaped merely by technical matters. The terms of
the Plan of Arrangement were negotiated by the
company and its stakeholders through a long
and arduous process. The court documents reveal
that the settlement was a?ected strongly by the rel-
ative power that the stakeholders were able to
exert on the negotiations. While the provincial
government had the technical capacity to consume
certain accounting signs, other stakeholders lack-
ing such elaborate technical capacities also in?u-
enced the negotiations. Each stakeholder was
framed and denominated by its accounting status
as a creditor.
25
Creditors were grouped into clas-
ses, such as unsecured creditors who had the
option of taking either a $2500 ?at settlement or
a prorated share of a ?xed amount of money set
aside for this class. The power of stakeholders
was not a function of the class’s total claim or
the individual stakeholder’s claim, but of the dis-
tribution of the claims within the class. The City
of Sault Ste. Marie was in a class on its own. Pen-
sioners were divided into two classes, according to
their indexed and non-indexed bene?ts claims; the
large total bene?ts claim being well distributed
amongst many pensioners, these people exercised
very little individual power. The notesholders were
di?erent from the pensioners in that the distribu-
tion of notes was uneven, a minority of noteshold-
ers holding a majority of the dollar amount of the
notes.
In the ?nal terms of arrangement, listed in
Table 4, the notesholders did well. They came
away with 75% ownership of the company plus a
considerable fraction of the original debt still being
owed to them in new notes. The notesholders were
able to gain this result because they were able to
defeat earlier drafts of the plan. The plan had to
be approved by a double majority of each class
of creditor (i.e., by a majority of the creditors rep-
resenting a majority of the dollar value). The ?rst
draft of the plan of arrangement called for a 9%
interest rate on the bulk of the new notes. This
was raised to 9.5% in the second draft of the agree-
ment, which was approved by everyone except
25
Because the proceedings took place under Ontario’s Com-
panies’ Creditors Arrangement Act, stakeholders lacking the
status of creditor were excluded from the negotiations. It is
interesting to see legislation formulated speci?cally around
accounting constructs.
Table 4
Final terms of arrangement
Terms of the plan of arrangement
All existing voting shares were deleted, including the employees’ 20% ownership stake
Employees received 20% of the new common shares, which were no longer held in trust and had no special rights
Interim ?nancing provided by banks during Algoma’s court protection was entirely secured
Notes holders received, pro rata, new notes totalling US$187.5 million, plus 75% of new common shares. (Existing notes had
amounted to US$349 million, plus US$47 million in accrued interest)
Pension claimants received a replacement pension plan
Unsecured claimants received, pro rata, 5% of the new common shares, or could elect to receive up to $2,500 to settle their claim
The City of Sault Ste. Marie received $5 million in future payments, secured by Algoma’s real property
The federal and provincial governments received all income taxes owing
A new collective bargaining agreement was enforced, featuring $153 million wage and bene?t reductions, reduced vacations, reduced
pension bene?ts, and job cuts
Pension payments to existing Ontario pensioners were taken over by the PBGF, along with related plan assets. Pension payments to
non-Ontario pensioners remained the obligation of Algoma Steel, to be paid from general revenues or a new pension plan
Pensions of current workers were secured by a claim on the remaining plan assets, and also by a $100 million claim on the ?xed assets
of the company, subordinate to the claims of the banks and notes holders
Source: (Algoma Steel, 2001b, p. 13; 2002, p. 30).
774 C. Graham/ Accounting, Organizations and Society 33 (2008) 756–782
24% of the notesholders. Because they held 84% of
the dollar value of the notes, they were able, at the
last minute, to prevent the second draft from being
approved. This triggered emergency meetings
between management and the stakeholders to res-
cue the plan (Ontario Superior Court, 2001d, p. 5).
The ?nal version of the plan of arrangement shows
an 11% interest rate on the new notes. This is
another indication of how the consumption of
accounting signs is structured: because the pension
obligation was widely dispersed amongst the pen-
sioners, the larger notesholders were more easily
able to a?ect how the accounting information
would be interpreted and translated into stakes
in the restructured company. The accounting signs
that denominated the creditors’ claims on the com-
pany presumed to impose a speci?c valuation on
each claim. Yet the meaning of these signs is deter-
mined at the time of their consumption during the
legal procedures and negotiations. The meaning is
shaped by legal features such as the de?nitions and
procedures of the Companies’ Creditors Arrange-
ment Act, by technical features such as the PBGF
and Section 3461 of the CICA Handbook, by the
distribution of power amongst the creditors and
by their exercise of power.
Denouement
Following the 2001 Plan of Arrangement, the
company enjoyed a resurgence in both sales and
stock price. Investors received a good return,
employees continued to work, and retirees contin-
ued to receive their pensions. But the complexities
of the resolution prevent an unequivocal judge-
ment of its success, even with the bene?t of hind-
sight. Employee wages were lower than before,
and retirees’ pensions were no longer indexed. In
addition, the resolution to Algoma’s 2001 pension
crisis may well have damaged the relationship
between employees and retirees. According to the
present union leader (DaPrat, 2005a), in 2001 the
retirees felt they had no choice but to go along
with the agreement being negotiated amongst the
stakeholders. They felt they would gain no advan-
tage by speaking out against the company’s posi-
tion. According to the union leader, while the
union’s primary goal during the 2001 negotiations
was to protect retirees’ existing pension bene?ts,
the retirees ended up angry about the settlement
because they lost indexing. Their anger was exacer-
bated in 2004 when Algoma returned to prosperity
and the current workers received a $10,000
advance on pro?t sharing as part of a new con-
tract.
26
The retirees received no such bonus, yet
they felt they had sacri?ced in 2001 in order to
help turn the company around (DaPrat, 2005a).
Fearful asymmetry
The transplantation of accounting signs from
Algoma Steel’s ?nancial statements to the legal
context of the a?davit demonstrates how the
meaning of accounting signs is a?ected by their
mode of reproduction. The subsequent struggle
between the stakeholders demonstrates how the
meaning of accounting signs is a?ected by their
mode of consumption. Reproduced in their new
legal setting, all the accounting signs took on an
urgency that transcended audited statements.
These signs were an attempt by management to
assert both the need for court protection and the
distribution of liability. The disclosures made by
Algoma Steel in the a?davit were simply insu?-
cient for anyone to see how some crucial numbers
were derived, a situation engendered by informa-
tion asymmetries, the opacity of Section 3461,
and the arrayed expertise of Algoma’s actuaries
and accountants. Yet despite these advantages,
management was unable to impose an undisputed
meaning on their accounting signs.
The most important accounting signs in the
legal process were those that denominated the
creditors. The legal status of participants was not
independent of these signs. Each claim on Algoma
Steel was initially predicated on Algoma Steel’s
accounting calculation of the amount owing. To
dispute this would entail producing an invoice or
other accounting records that could signify a
greater claim. For tradespeople, this might be rel-
atively simple, supposing for instance that Algoma
26
The union leader who provided the information did not
state whether this was an average amount or a ?at amount per
worker.
C. Graham/ Accounting, Organizations and Society 33 (2008) 756–782 775
had failed to process an invoice correctly. For pen-
sioners, however, disputing the stated pension lia-
bility would be extremely di?cult. It would involve
a calculation so complex as to be beyond the aver-
age individual’s capability to perform, and even
beyond the pensioners’ collective ability. They
were, for all practical purposes, completely at the
mercy of the company’s calculations.
However, there is no indication in the court doc-
uments or the media coverage that Algoma’s
accounting calculations were themselves conten-
tious. The contest came after the accounting signs
were accepted under oath, and reproduced into
the court proceedings. The lengthy negotiations
from April until December of 2001 between man-
agement and the creditors, which also involved
the union making wage concessions, was a process
of giving the accounting signs a precise legal mean-
ing, of translating them into new legal and ?nancial
arrangements. In this direct sense, the accounting
signs preceded the reality they seemed to describe.
The signs described a set of claims, yet the claims
had no e?ect until the signs were consumed and
rendered meaningful. The signs suggested an equiv-
alence – that the pensioners’ claimon the company,
for instance, was roughly equivalent to the notes-
holders’ claim on the company ($503 million com-
pared to $551 million) – yet each class of creditors
required the accounting signs to translate into a
unique new set of ?nancial relationships. For
example, in due course the sign ‘‘Notes pay-
able = $551 million” came to mean 75% of the
new shares and an entire new set of notes payable.
This is not obvious or inherent in the sign at the
moment of reproduction. It is the result of contes-
tation during an extended process of consumption.
The legal status of the creditors, established
through accounting records, gave each class of
participants the same opportunity to in?uence
the outcome. However, the ability to operate the
system of calculations and the system of legal pro-
cedures was not evenly distributed amongst these
participants. The pensioners had no collective
voice; a common legal advisor had to be provided
by the courts (Ontario Superior Court, 2001c, p.
151). As noted, because the pensions were fairly
evenly distributed amongst the pensioners, there
was no small group of them that could hold up
the restructuring proceedings. Unlike the notes-
holders, the pensioners were passive consumers
of Algoma Steel’s accounting signs, and hence suf-
fered more severely the consequences of the asym-
metries of the mode of reproduction.
These information asymmetries would have
been worse had they not been partially addressed
27
by the new discount rates mandated in Section
3461. The guidelines for discount rates in Section
3461 take some of the previous latitude out of
the hands of management, making the accounting
somewhat more transparent. However, there is a
cost to this. The new guidelines stipulate that the
discount rates should now follow the market.
Thus, an opaque but relatively stable discount rate
is replaced by a rate that is arguably more trans-
parent, but also more volatile. The risks of the
market replace moral hazard.
The opacity of Section 3461 might suggest that
the accounting profession plays a hegemonic role
in pension accounting in Canada, that somehow
they control ‘‘the code” that for Baudrillard sym-
bolized the programmatic way that signs take on
meaning in our society. However, other features
of the Algoma Steel narrative contradict this. Alg-
oma management was able to hire accounting
expertise to exploit its information asymmetries,
putting the accounting profession in a subservient
role. Also, the accounting signs in the company’s
2001 a?davit that triggered governmental actions
lacked the accounting profession’s imprimatur.
27
Asymmetries are still evident in the impossibility of recon-
structing certain intermediate results in the present study. For
example, the discount rates disclosed in the 2001 Algoma
statements seem wrong. If the disclosed rate of 6.5% is used, the
pension-related portion of unamortized net actuarial loss at
year end works out to $250 million. Since the funding de?cit of
the pension plan was $552 million, a di?erence of $296 million
would be computed for the accrued pension liability. However,
the accrued pension liability was only $278 million. This is the
amount that would be computed with an 8.1% rate. While the
working papers for the disclosure would likely be more complex
than the general model based on Section 3461 that is described
here, the discrepancy is nonetheless irresolvable given Algoma’s
existing disclosures. Thus the complexity of Section 3461 turns
disclosures of pension liabilities into ‘‘black boxes” impenetra-
ble to ?nancial statement users, and renders the expertise of
professional accountants indispensable to managers of compa-
nies that must submit to the requirements of the CICA
Handbook.
776 C. Graham/ Accounting, Organizations and Society 33 (2008) 756–782
The Algoma Steel events, therefore, would call
into question the way that Macintosh et al.
(2000) privilege the accounting profession in their
discussion of accounting as simulacrum and
hyperreality. They argue that the sign value of an
accounting sign derives from its authentication
by an accountant: certi?ed signs impart ‘‘a sense
of exogeneity and reliability to society at large”
(Macintosh et al., 2000, p. 41). It is unclear, how-
ever, what system boundaries Macintosh et al. are
assuming when they refer to exogeneity. The
accounting profession is no more exogenous to
the social systems that produce accounting signs
than booksellers or newspaper editors are to the
social systems that produce vernacular language
signs. The insistence on the exogeneity of profes-
sional certi?cation unnecessarily undermines the
Baudrillardian perspective Macintosh et al. pro-
vide on accounting. For consistency with Baudril-
lard, I would argue that an accounting sign takes
its sign value from its position within the system
of signs in which it appears, not from pseudo-exo-
geneity. This system of signs, however, must be
understood more comprehensively than Macin-
tosh et al. suggested. The system context of
accounting signs is not limited to clean surplus
models or pension accounting models. The pen-
sion numbers in the a?davit, for example, took
their sign value not only from their relation to
other accounting signs, but also from the legal
signs in which they were embedded and indeed
from the credibility of Algoma Steel management
itself. This sign value was ampli?ed when the
media picked up the numbers and repeated them
widely.
If certi?ed accounting signs provide anything
to social and commercial discourse, it is not exo-
geneity but relative stability. This is because cer-
ti?ed accounting signs are produced, at least
nominally, according to standards that take con-
siderable time and e?ort to change. While the
standards setting process can be in?uenced by
those with the power to do so, the process is
institutionalized to the degree that it is largely
conservative and resistant to change. This results
in accounting standards that change much less
quickly than accounting signs can be changed,
making certi?ed signs relatively stable against
the backdrop of economic commotion and
babble.
This does not mean that accounting signs are
not arbitrary. But it is important to clarify the
notion of arbitrariness used by Baudrillard. Exam-
ples provided by Macintosh et al. (2000, pp. 41–
42), such as the use of ‘‘red” on a tra?c light to
mean stop and ‘‘green” to mean go, illustrate not
arbitrariness but conventionality. The original
choice of red instead of, say, purple, was possibly
quite arbitrary, but this does not exhaust Baudril-
lard’s notion of arbitrariness in language. The traf-
?c-light example only evokes the image of
someone following the convention of stopping at
a red light. However, consider for a moment a
world in which drivers of large, powerful trucks
are armed with devices that can control tra?c
lights for their own bene?t, switching them to
green as they approach intersections. In such a
world, these truck drivers would be producers of
tra?c signals, and the masses of other drivers
and pedestrians would be mere consumers. This
is the world of accounting today. The Algoma
Steel crisis illustrates the considerable latitude that
corporations have in controlling the production of
accounting signs. This latitude is built into the
models of Section 3461 of the CICA Handbook,
and it is available to corporations because they
dominate the resource of accounting expertise.
Bringing this matter of power into the equation is
necessary in order to comprehend what Baudril-
lard was getting at. At another point in their article
Macintosh et al. provide a wonderful illustration
of the di?erential e?ects of power, General Elec-
tric’s use of earnings models to make strategic
decisions (pp. 32–33). Their example shows how
the production of accounting signs creates the real-
ity that accounting then pretends to re?ect. Not
everyone has the power to produce such signs.
This is clear in the case of Algoma Steel. Man-
agement was able to produce accounting signs that
mattered, that made a di?erence in the social
world. However, this social di?erence was not pro-
duced until the moment when these accounting
signs were consumed. Consumption, as we saw in
our re-examination of Baudrillard’s earlier writ-
ings, is a social system for producing meaning.
Yet workers and pensioners were largely passive
C. Graham/ Accounting, Organizations and Society 33 (2008) 756–782 777
consumers of these signs, while other creditors and
government bureaucrats had greater capacity to
shape how the accounting signs would be inter-
preted. The function of accounting signs in Alg-
oma Steel’s social relations thus went beyond the
benign agreements on meaning that Macintosh
et al. discussed, when they said that institutionally
produced accounting information is:
. . .an arbitrary but essential feature of a
hyperreal world. Without it, everything
would seem to depend on everything else
and interaction would be problematic if not
impossible. (pp. 32–33)
What Macintosh et al. miss, however, is that
social interactions are problematic. This is true
for government, for investors, for banks, and, no
doubt, even for management. But it is especially
true for the relatively powerless, such as elderly
pensioners. The relations of power in society
become institutionalized in things like Section
3461, and render di?cult, for the average citizen,
the operation of the system of signs by which soci-
ety is governed.
This helps explain what Macintosh et al. (p. 44)
noticed about accounting information, that it does
not fully behave as Baudrillard predicted signs
should in a hyperreal society, and become
absorbed, de-politicized and neutralized by the
masses. The Algoma Steel incidents suggest that
the reason it does not behave in this way is that
the consumers of accounting information are not
‘‘masses”. The consumption of accounting infor-
mation is structured and institutionalized, just as
its production is. There are no undi?erentiated,
passive masses absorbing accounting signs.
Rather, there are relatively institutionalized and
programmatic interfaces waiting to receive the cor-
rect combination of accounting signs in order that
the technologies of governance and wealth distri-
bution be triggered into action. This overstates
the mechanistic aspects of accounting communica-
tion, of course. Nonetheless, the point remains
that accounting signs are consumed primarily by
technically structured corporate and bureaucratic
agents, rather than by masses of consumers.
Hence, when the underfunding of Algoma
Steel’s pension plan was revealed, the tools of pen-
sion accounting permitted Algoma’s management
to construct a new picture of its pension plans,
and to marshal creditors and institutionalized
resources towards a negotiated solution. The Alg-
oma Steel events illustrate that at least in Cana-
dian society, with its relatively decentralized form
of government, accounting signs serve to de?ne
and to shift the boundary between government
and the private sector. The intervention of govern-
ment whereby pension assets and liabilities were
taken over into the PBGF was a temporary col-
lapse, or in Baudrillard’s terms, an implosion, of
the boundary between government and private
sector. The reproduction and recon?guration of
accounting signs in this cases was structured by
institutional mechanisms, regulations, and the role
of expertise. These structures are at the root of the
uneven distribution in Canadian society of the
ability to operate the system of accounting signs.
Conclusion
As Macintosh et al. (2000, p. 45) state, ‘‘We
are concerned with describing the nature of
accounting signs, with unearthing how they came
to be produced and with why they subsequently
come to be taken for granted as a reality of their
own”. The present study has provided speci?c
evidence of how accounting, as a system of
signs, structures and enables the public protec-
tion of private pensions. We have seen how the
institutions and techniques of pension accounting
create a system of wealth distribution and risk
distribution amongst investors, creditors, man-
agement, employees, the government, and other
stakeholders. We have seen, in particular, how
pension accounting under Canadian GAAP has
served to mobilize these institutions in times
when the system is represented as ‘‘failing”.
The system of accruals built into Section 3461
permitted the company’s pension funding prob-
lem to remain unspoken, isolated as a pension
de?cit note rather than being aggregated with
the company’s overall performance numbers.
This permitted the company’s immediate stake-
holders, including investors and creditors, to
avoid the brunt of the new reality induced by
778 C. Graham/ Accounting, Organizations and Society 33 (2008) 756–782
Section 3461. By applying the corridor rule and
con?ning the ‘‘damage” to the company’s pen-
sion fund, the situation was successfully proble-
matized as a pension crisis, amenable to
government intervention through accounting
technologies embodied in the Pension Bene?t
Guarantee Fund. The ‘‘failure” of the system is
thus seen to be the very point at which all its
?exibility and strength was brought into play.
The crisis point—constructed, packaged and pre-
sented by accounting—served as the pivot point
where risks and wealth allocations were success-
fully renegotiated.
The possibilities provided by Section 3461 for
this pivoting and recon?guration were created not
by direct legislation and state control, but by an
institutionalized di?usion of social governance.
The role of accounting signs in this pervasive form
of governance is contingent on the particular
regimes of power and knowledge in which the signs
are situated. I have argued that these signs should
be viewed as simulacra, arbitrary signs that take
their sign value from the system of signs in which
they circulate, and that these arbitrary signs serve
to shift the boundary between government and
the private sector. By focussing on the accounting
sign, the present study highlights how governmen-
tal programs like the PBGF function as institution-
alized consumers of accounting signs. When, as in
the Algoma Steel situation, these accounting signs
are produced by corporations and others outside
of government, and are taken up discursively to
trigger programmatic government interventions,
the power to govern is e?ectively dispersed. This
suggests that the power to govern in today’s demo-
cratic society is, at least in part, negotiated through
and mediated by accounting signs. The Algoma
Steel pension bailout shows how unevenly the
capacity to in?uence this negotiation is distributed
in Canadian society, because of information asym-
metries, the role of accounting expertise, and the
arbitrariness of accounting signs.
Acknowledgements
The comments of Dean Neu, Je? Everett, Teri
Shearer, Hussein Warsame, Daphne Taras, Mike
Welker, Dan Thornton, Steve Salterio, and Eliza-
beth Farrell are gratefully acknowledged, as are
the suggestions of two anonymous reviewers. This
paper bene?ted from the insights of Yves Gendron
and participants at the Alternative Perspectives on
Accounting Conference, held at Laval University
in March 2007. Funding was provided by SSHRC
through a Doctoral Fellowship, and by the Schu-
lich School of Business. The able research assis-
tance of Mona Mann is gratefully acknowledged.
Appendix. The implicit model of Section 3461
Section 3461 of the CICA Handbook uses an
accounting model that brings together actuarial
calculations of future bene?t payments with ?nan-
cial valuations of assets set aside to fund those
payments. The crucial accounting question being
answered by the model (cf. 3461.001) is this: given
the actuarial and ?nancial calculations, what is the
proper way to recognize, measure and disclose the
costs of employee future bene?ts? The principle
applied in the guidelines is that the cost of future
employee bene?ts should be recognized in the per-
iod in which the employee provided the service
that earned her the bene?ts.
28
The explicit basis
for this principle is that ‘‘Bene?t plans are consid-
ered part of an employee’s compensation arrange-
ment” (3461.002). Just as wages are considered an
expense to be matched to the period in which cor-
responding revenues were earned, so future bene-
?ts should be matched to the same period. This
can occur through the recognition of the liability
to pay those bene?ts, or the recognition of the cost
itself, either as an expense or capitalized in an asset
such as inventory (3461.002, Footnote 1).
The attribution mechanisms at the heart of this
model involve matching distinct units of the future
bene?ts to speci?c periods of service by employees,
and calculating by actuarial methods the present
value of each unit for the period in which it
accrued. The actuarial methods are black-boxed
by the Handbook. That is, they are treated as an
28
The guidelines point out that some future bene?ts are
recognized when an event occurs, such as the application of the
employee for disability bene?ts (3461.003).
C. Graham/ Accounting, Organizations and Society 33 (2008) 756–782 779
independent and unquestionable given. Two sepa-
rate accrual attributions are involved. In the ?rst,
the units of future bene?t are calculated as earned
on a period by period basis. By this process, the
accrued bene?t obligation of the company is calcu-
lated. This obligation is recalculated each year,
based on updated actuarial assumptions, employee
data, and other input parameters.
29
During the
course of a year, the obligation for future bene?ts
increases by the cost of bene?ts earned by employ-
ees during the year (the current service cost), and
decreases by the amount of any bene?t payments
actually made. In addition, interest accrues to the
employees on the outstanding portion of the obli-
gation, the portion that had been earned before the
period began and which has not yet been paid to
the employees. When the experience of all these
increases and decreases does not match the actuar-
ial projections, the di?erence is calculated as an
actuarial gain or loss on the accrued bene?t obliga-
tion. An additional actuarial gain or loss is calcu-
lated on the plan assets, being the di?erence
between expected and actual return on assets.
The total of these two actuarial gains or losses is
not recognized immediately in the entity’s ?nancial
statements. Rather, the model assumes that these
actuarial gains and losses are random, and will o?-
set themselves over time. In order to give the entity
time for these ?uctuations to net out, and so avoid
short-term shocks to its ?nancial statements, a sec-
ond accrual attribution is used.
In the second accrual attribution of the model,
a portion of the costs of the bene?t obligation is
allocated back to the current period. By this pro-
cess, the accrued bene?t liability of the company
is calculated. The so-called ‘‘corridor” rule is the
primary method for making this attribution.
Under the corridor rule, the net actuarial gain or
loss, which may ?uctuate widely from year to year
due to changes in mortality experience, market
values for assets, and so on, is dampened by accu-
mulating it from year to year o? the balance sheet,
and amortizing a portion of this accumulation to
the ?nancial statements each year. The calculation
of the amortization amount is convoluted. First,
only part of the net actuarial gain or loss is subject
to amortization. If the net actuarial gain or loss
does not exceed 10% of the greater of the accrued
bene?t obligation and the fair value of plan assets,
no amortization is necessary. If it does exceed this
10% ?gure, then only the amount by which it
exceeds the 10% ?gure is subject to amortization.
This excess amount is to be spread out over the
remaining service life of the average employee.
That is, if most employees have 8 years to go
before collecting their bene?ts, then only 1/8th of
the excess amount is recognized on the entity’s
?nancial statements this year. Furthermore, since
the calculation is performed using the net actuarial
gain or loss as at the beginning of the year, the
impact of a ?uctuation on the ?nancial statements
is not only dampened but delayed.
Besides the corridor rule for actuarial gains and
losses, several other calculations are used to attri-
bute the bene?t obligation back to the current per-
iod as a pension cost. The current service cost, the
interest cost on the obligation, and the simple
amortization of any changes to the plan—i.e.,
changes to accounting policies (transitional obliga-
tions) or retroactive changes to promised bene?ts
(past service costs)—are all components of the
pension cost.
The pension cost recognized every year is accu-
mulated. The contributions the entity makes to the
pension plan are deducted from this accumulation.
The result is the accrued bene?t liability that
appears on the entity’s balance sheet.
References
Accounting Standards Board of Canada. (2002). Employee
future bene?ts: Accounting Standards Board of Canada.
Downloaded December 19, 2004.
Algoma Steel. (1996). Annual report. Sault Ste. Marie, Ontario:
Algoma Steel Inc.
Algoma Steel. (1997). Annual report. Sault Ste. Marie, Ontario:
Algoma Steel Inc.
Algoma Steel. (2000). Annual report. Sault Ste. Marie, Ontario:
Algoma Steel Inc.
29
Pension plans are governed by federal legislation for entities
registered nationally, and by provincial legislation for entities
registered provincially. This pension legislation typically
requires a complete actuarial recalculation only every few
years. Extrapolations from the prior recalculation are permitted
in intervening years.
780 C. Graham/ Accounting, Organizations and Society 33 (2008) 756–782
Algoma Steel. (2001a). Annual report. Sault Ste. Marie,
Ontario: Algoma Steel Inc.
Algoma Steel. (2001b). Notice of proceedings and meetings of
certain creditors of Algoma Steel Inc. and information
circular with respect to a plan of arrangement and reorga-
nization. Sault Ste. Marie, Ontario: Algoma Steel Inc.
Algoma Steel. (2002). Annual report. Sault Ste. Marie, Ontario:
Algoma Steel Inc.
Algoma Steel’s $625-M Pension Shortfall Key to Restructuring.
(2001). Sudbury Star, August 24, p. A.10.
Archer, S. (1998). Mattessich’s critique of accounting: A
review article. Accounting and Business Research, 28(4),
297–316.
Armstrong, P. (2000). Designing accountability: The manage-
rial semiotics project. Critical Perspectives on Accounting,
11(1), 1–22.
Arrington, C. E., & Francis, J. R. (1989). Letting the chat out of
the bag: Deconstruction, privilege and accounting research.
Accounting, Organizations and Society, 14(1–2),
1–28.
Baudrillard, J. (1968). Le Syste` me des Objets. Paris: Gallimard.
Baudrillard, J. (1970). La Socie´ te´ de Consommation; Ses
Mythes, Ses Structures. Paris: Gallimard.
Baudrillard, J. (1972). Pour Une Critique de L’e´ conomie
Politique du Signe. Paris: Gallimard.
Baudrillard, J. (1973). Le Miroir de La Production; Ou,
L’illusion Critique du Mataˆ erialisme Historique. Tournai:
Casterman.
Baudrillard, J. (1975). The mirror of production (M. Poster,
Trans.). St. Louis, Missouri: Telos Press.
Baudrillard, J. (1981). For a critique of the political economy of
the sign (C. Levin, Trans.). St. Louis, Missouri: Telos Press.
Baudrillard, J. (1983). Simulations. New York City:
Semiotext(e).
Baudrillard, J. (1993). Symbolic exchange and death. London:
Sage.
Baudrillard, J. (1995). Simulacra and simulation. Ann Arbor:
University of Michigan Press.
Baudrillard, J., & Poster, M. (1988). Jean Baudrillard: Selected
writings. Stanford, CA: Stanford University Press.
Beatty, C. A., & Schachter, H. (2002). Employee ownership: The
new source of competitive advantage. Etobicoke, Ontario:
John Wiley & Sons.
Belkaoui, A. (1978). Linguistic relativity in accounting.
Accounting, Organizations and Society, 3(2), 97–104.
Belkaoui, A. (1980). The impact of socio-economic accounting
statements on the investment decision: An empirical
study. Accounting, Organizations and Society, 5(3), 263–
283.
Boland, R. J. Jr., (1989). Beyond the objectivist and the
subjectivist: Learning to read accounting as text. Account-
ing, Organizations and Society, 14(5–6), 591–604.
Brooksbank, D. (2006). Accounting rules ‘Threaten Swiss
Pension System’. Investment and Pensions Europe (24
January 2006).
Cooper, C., & Puxty, A. (1994). Reading accounting writing.
Accounting, Organizations and Society, 19(2), 127–146.
Daly, B. A., & Schuler, D. K. (1998). Rede?ning a certi?ed
public accounting ?rm. Accounting, Organizations and
Society, 23(5–6), 549–567.
DaPrat, M. (2005a). Personal communication, January 31.
DaPrat, M. (2005b). Personal communication, February 10.
Daw, J. (2001). Failure could wipe out pension guarantee fund;
Provincial kitty has $230 million for Aid. Toronto Star,
April 25, E.01.
Dofasco. (2001). Annual report. Hamilton, Ontario: Dofasco
Inc.
Dufresne, D. (1993). Some aspects of statement of Financial
Accounting Standards No. 87. Actuarial Research Clearing
House (Vol. 2).
Estey, N. A. (1999). Accountable for your promises. CA
Magazine.
Evans, L. (2004). Language, translation and the problem of
international accounting communication. Accounting,
Auditing & Accountability Journal, 17(2), 210–248.
Everett, J. (2004). Exploring (False) dualisms for environmental
accounting praxis. Critical Perspectives on Accounting,
15(8), 1061–1084.
FASB. (2006). FAS 158: Employers’ accounting for de?ned
bene?t pension and other postretirement plans. Norwalk,
Connecticut: Federal Accounting Standards Board. Downloaded February 28,
2007.
Feltham, G. A., & Ohlson, J. A. (1995). Valuation and clean
surplus accounting for operating and ?nancial activities.
Contemporary Accounting Research, 11(2), 689–731.
Fore, D. (2004). Changes in accounting practices will drive
pension paradigm shift. Pension Research Council Working
Paper. Philadelphia: Pension Research Council, University
of Pennsylvania.
Francis, J. R. (1994). Auditing, hermeneutics, and subjectivity.
Accounting, Organizations and Society, 19(3), 235–269.
Gadamer, H. (1976). Philosophical hermeneutics. Berkeley:
University of California Press.
Graves, O. F., Flesher, D. L., & Jordan, R. E. (1996). Pictures
and the bottom line: The television epistemology of U.S.
annual reports. Accounting, Organizations and Society,
21(1), 57–88.
Heidegger, M. (1962). Being and time (J. Macquarrie & E.
Robinson, Trans.). New York: Harper & Row.
IASB. (2004). Summaries of International Financial
Reporting Accounting Standards: IAS 19, employee bene-
?ts. London: International Accounting Standards Board.
Down-
loaded February 25, 2006.
Joss, P. N. (2004). The earnings implications of pension expense:
A stochastic analysis of ten Canadian companies. Unpub-
lished Masters thesis, University of Calgary, Calgary.
Keenan, G. (2001). Noteholder votes deal a blow to Algoma
restructuring plan. Toronto Star, December 7.
Lavoie, D. (1987). The accounting of interpretations and the
interpretation of accounts: The communicative function of
‘‘the language of business”. Accounting, Organizations and
Society, 12(6), 579–604.
C. Graham/ Accounting, Organizations and Society 33 (2008) 756–782 781
Liddle, J. (2002). New accounting and supervisory standards
could prove unpopular with pension funds themselves.
European Pensions & Investment News (114).
Macintosh, N. B., & Baker, C. R. (2002). A literary theory
perspective on accounting: Towards heteroglossic account-
ing reports. Accounting, Auditing & Accountability Journal,
15(2), 184–222.
Macintosh, N. B., & Shearer, T. (2000). The Accounting
Profession Today: A Poststructuralist Critique. Critical
Perspectives on Accounting, 11(5), 607–626.
Macintosh, N. B., Shearer, T., Thornton, D. B., & Welker, M.
(2000). Accounting as Simulacrum and Hyperreality: Per-
spectives on Income and Capital. Accounting, Organizations
and Society, 25(1), 13–50.
Mattessich, R. (2003). Accounting Representation and the
Onion Model of Reality: A Comparison with Baudrillard’s
Orders of Simulacra and His Hyperreality. Accounting,
Organizations and Society, 28(5), 443–470.
McDowall, D. (1984). Steel at the Sault: Francis H. Clergue, Sir
James Dunn and the Algoma Steel Corporation: 1901–1956.
Toronto: University of Toronto Press.
McGoun, E. G. (1997). Hyperreal Finance. Critical Perspectives
on Accounting, 8(1-2), 97–122.
Mouck, T. (1994). Corporate Accountability and Rorty’s
Utopian Liberalism. Accounting, Auditing & Accountability
Journal, 7(1), 6–30.
Mouck, T. (2004). Institutional Reality, Financial Reporting
and the Rules of the Game. Accounting, Organizations and
Society, 29(5-6), 525–541.
Nishman, R. F. (1995). Worker ownership and the restructuring
of Algoma Steel in the 1990s. Kingston, Ontario: Industrial
Relations Centre, Queen’s University.
Ontario Superior Court. (2001a). A?davit of Glen Percival
Manchester, Sworn April 23, 2001. Court File No. 01-CL-
4115. Toronto, Ontario.
Ontario Superior Court. (2001b). Cross-examination of Harold
Stephen, Sworn October 22, 2001. Court File No. 01-CL-
4115. Toronto, Ontario.
Ontario Superior Court. (2001c). Motion returnable Novem-
ber 9. 2001 Regarding meetings of creditors and claims
procedure. Court File No. 01-CL-4115. Toronto,
Ontario.
Ontario Superior Court. (2001d). Ninth report of Ernst &
Young Inc. in its capacity as court-appointed monitor of
Algoma Steel Inc., December 7, 2001. Court File No. 01-
CL-4115. Toronto, Ontario.
Ottawa Provides $50-M Loan Guarantee for Algoma. (2001).
Sudbury Star, December 23, 2001, A.1.FRO.
Pension Bene?ts Act. (1990). Revised statutes of Ontario.
Pension Changes Could Cost $180b: Accounting Organiza-
tion’s Strict New Rules Could Cost Billions for Companies
O?ering De?ned Pension Plans. (2006). Downloaded February 3, 2006.
Perkel, C. (2001). Taxpayers may pay for Algoma; Steel ?rm’s
fall would have huge impact: Harris. Toronto Star,
September 7, 2001, NE.07.
Preston, A. M., Wright, C., & Young, J. J. (1996). Imag[in]ing
annual reports. Accounting, Organizations and Society,
21(1), 113–137.
Preston, A. M., & Young, J. J. (2000). Constructing the global
corporation and corporate constructions of the global: A
picture essay. Accounting, Organizations and Society, 25(4–5),
427–449.
Province of Ontario. (2001). Ontario Government appoints
representative to protect pensioners – News release, August
31, 2001. Toronto: Province of Ontario. Down-
loaded December 12, 2003.
Reg. 203/02. (2002). Ontario regulations.
Rorty, R. (1991). Science as solidarity, objectivity, relativism,
and truth: Philosophical papers (Vol. 1). Cambridge, UK:
Cambridge University Press, pp. 35–45.
Saussure, F. d. (1961). Course in general linguistics. London:
Owen.
Schoonmaker, S. (1994). Capitalism and the code: A critique of
Baudrillard’s third order simulacrum. In D. Kellner (Ed.),
Baudrillard: A critical reader (pp. 168–188). Oxford, UK:
Blackwell.
Squeeze on Pensions Could Tighten: Overhaul in Accounting
Rules Will Cost Workers. (2006). Downloaded February 3, 2006.
Steelworkers Hail Pension Fund Appointment. (2001). Edmon-
ton Journal, September 1, 2001, G.8.
Stelco. (1999). Annual report. Hamilton, Ontario: Stelco Inc.
Stelco. (2001). Annual report. Hamilton, Ontario: Stelco Inc.
Tinker, A. M., Merino, B. D., & Neimark, M. D. (1982). The
normative origins of positive theories: Ideology and
accounting thought. Accounting, Organizations and Society,
7(2), 167–200.
Tinker, T. (1991). The accountant as Partisan. Accounting,
Organizations and Society, 16(3), 297–310.
Tishco?, N. (1999). An actuarial action plan, CA Magazine.
Van Alphen, T. (2001a). Insolvent Algoma Steel gains bankruptcy
court’s protection; Major employer in Sault Forecasts loss this
year of $200 Million. Toronto Star, April 24, 2001, E.01.
Van Alphen, T. (2001b). Pension aid crucial to restructuring of
Algoma Steel; government’s response to woes ’disappoint-
ing’. Toronto Star, August 24, 2001, E.01.
Walters-York, L. M. (1996). Metaphor in accounting discourse.
Accounting, Auditing & Accountability Journal, 9(5), 45–70.
Watson, T. (2001). Ottawa may give millions for bailout of
Algoma: Ontario will not help. National Post, August 25,
2001, C.1.FRO.
782 C. Graham/ Accounting, Organizations and Society 33 (2008) 756–782
doc_204874543.pdf