Accounting in markets, hierarchies and networks: The role of accounting in the transnation

Description
We examine the changing role of accounting in the development of the international postal
system between 1840 and the emergence of the Universal Postal Union (UPU) in 1875. We
use the distinction between mundane and opportunistic transaction costs to explain why
accounting disappeared as a coordinating mechanism as postal transactions migrated from
spot market exchanges, through bilateral contracts (treaties) between nations, into a network
of domestic post offices coordinated by the UPU. Our analysis refines the application
of transaction cost economics to the understanding of the role of management accounting
in different governance mechanisms.

Accounting in markets, hierarchies and networks: The role of accounting
in the transnational governance of postal transactions
Alan J. Richardson
*
, Eksa Kilfoyle
Schulich School of Business, York University, Toronto, Ontario, Canada
a b s t r a c t
We examine the changing role of accounting in the development of the international postal
system between 1840 and the emergence of the Universal Postal Union (UPU) in 1875. We
use the distinction between mundane and opportunistic transaction costs to explain why
accounting disappeared as a coordinating mechanism as postal transactions migrated from
spot market exchanges, through bilateral contracts (treaties) between nations, into a net-
work of domestic post of?ces coordinated by the UPU. Our analysis re?nes the application
of transaction cost economics to the understanding of the role of management accounting
in different governance mechanisms.
Ó 2009 Elsevier Ltd. All rights reserved.
Introduction
Williamson (1996) suggests that there exists a set of
discrete governance mechanisms through which transac-
tions can be completed. Each governance mechanism has
its own capabilities and limitations such that different
types of transactions are handled by speci?c governance
mechanisms at lowest cost. The polar types of governance
mechanisms are markets and hierarchies but hybrid forms
of governance are common. These hybrid governance
mechanisms include relationships, alliances and networks.
Hybrid governance mechanism may have different infor-
mation and control requirements than markets and hierar-
chies. Hopwood (1996), among others, has expressed the
need to understand the implications of these types of gov-
ernance mechanisms for management accounting informa-
tion and control systems:
To date accounting research has largely ignored such
changes and their implications for ?nancial decision
making and control. Having earlier given little or no
consideration to the informational implications of
matrix structures and the ?nancial aspects of project
oriented forms of organization, the accounting research
community is largely continuing to be satis?ed with its
?xation on the traditional hierarchical organization
(Hopwood, 1996, pp. 589–590).
In response to this call for a better understanding of the
role of accounting in inter-organizational relationships,
accounting researchers, drawing on a variety of theoretical
perspectives, have examined how accounting information
is used to facilitate ‘‘cooperative coordination” (Hakansson
& Lind, 2004) within hybrid structures and the impact of
these cooperative undertakings on management account-
ing practices. In doing so, some researchers looked at the
constitutive role of management accounting information
by probing into ‘‘how accounting is a ‘force’ – an actor –
in establishing and developing inter-organizational rela-
tionships” (Mouritsen & Thrane, 2006, p. 242) or explored
the enactment of accounting controls in alliances that are
part of larger networks and the effect of these alliances
on accounting practices (Chua & Habib, 2007). Others,
drawing on economic theories, including transaction cost
economics (TCE), have explored the governance role of
the management accounting information and controls in
alliances and networks (Anderson & Dekker, 2005; Dekker,
2004; Hakansson & Lind, 2004).
In studying the role of management accounting systems
within speci?c governance structures, however, these stud-
ies do not problematize the change in use of accounting as
0361-3682/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.aos.2009.04.002
* Corresponding author.
E-mail address: [email protected] (A.J. Richardson).
Accounting, Organizations and Society 34 (2009) 939–956
Contents lists available at ScienceDirect
Accounting, Organizations and Society
j our nal homepage: www. el sevi er. com/ l ocat e/ aos
transactions migrate from one governance mechanism (i.e.
markets or hierarchies) to another (i.e. networks). To over-
come this gap in our understanding of the existence and
role of accounting techniques in inter-organizational rela-
tionships, we explore the effect of a series of changes in
governance structure on the use of accounting information.
We identify a setting, the international postal market, in
which we are able to trace ‘‘longue durée” changes (Braudel
& Matthews, 1982) in the use of accounting as transactions
migrate from spot market exchanges to bilateral treaties/
contracts between national postal providers and, ?nally,
to a network of national postal providers coordinated by
the Universal Postal Union (UPU). Unlike the empirical
observations in extant accounting literature on networks,
in this setting, accounting technologies are eliminated
when transactions move from bilateral agreements to a
network in an attempt to facilitate interaction among par-
ticipants. Concurring with Dekker (2004) on the limitations
of current interpretations of TCE to explain the use of
accounting in networks, we expanded on the use of TCE
in the accounting literature by introducing the concepts
of ‘‘mundane” and ‘‘opportunistic” transaction costs
(Baldwin, 2008). This distinction allows us to theorize the
disappearance of accounting as transactions shift from the
market into network forms of governance.
This paper is structured as follows. We begin with a dis-
cussion of current work on the role of accounting in mar-
kets, hierarchies and networks. A key assumption of this
literature is that management accounting information
emerges as a replacement for market prices when transac-
tions migrate from the market to other forms of gover-
nance. We identify gaps in this literature: to date the
literature has not examined changes in accounting when
transactions shift from one governance structure to an-
other and do not account for instances where accounting
is eliminated as a result of this transition. We then identify
a key elaboration of transaction cost economics that has
not been used in the accounting literature: the distinction
between mundane and opportunistic transaction costs.
Next the use of the case method to explore accounting in
the international postal system is discussed. In particular,
we argue for the use of case analysis as a theory testing
method when theories make non-probabilistic predictions
and as a theory elaboration method through its emphasis
on mechanisms and processes. We then describe our data
set. Our main results are presented in three sections
describing the state of the postal system in the time period
1840–1875, the transactional issues that arose, and, ?nally,
the change in rules and accounting procedures that were
implemented in the ?rst UPU Convention in 1875. Our dis-
cussion and conclusion draw the lessons from this case for
the application of TCE to accounting phenomena.
The role of accounting in markets, hierarchies and
networks
Accounting in markets and hierarchies
Neo-institutional economics, and transaction cost eco-
nomics in particular, provide an important framework for
understanding many accounting phenomenon. This frame-
work relies on two main insights. First, Coase (1937) devel-
oped the insight that markets and ?rms were alternative
ways of completing transactions and that there are costs
to the use of either mechanism. The idea that the market
mechanism was not costless, in particular, provided Coase
with a way to reconceptualize the nature of the ?rm. The
?rm would arise if it could complete transactions at lower
cost than in the market. For example, where complex inter-
dependent tasks are required to produce some good, it may
be more effective to hire skilled craftsmen as wage labour
controlled by a central authority (the ?rm) rather than
have each craftsman attempt to contract with all others
for speci?c work. However, there must be increasing costs
to the use of the ?rm and authoritative control to complete
transactions that limits the size of the ?rm (otherwise, a
single ?rm would internalize all of some forms of transac-
tions). Coase (1937) thus provides a role for management
inside the ‘‘black box” production function envisioned by
classical economics and provides a rationale for the exis-
tence and size limits of ?rms.
1
Second, Williamson (1975) generalized these observa-
tions by suggesting that markets and hierarchies (i.e.
?rms/government agencies
2
) were alternative governance
mechanisms that had a contingent cost advantage in com-
pleting transactions depending on the nature of those trans-
actions (characterized by the ‘‘critical dimensions” of asset
speci?city, frequency, and uncertainty, Williamson, 1979, p.
239). This leads to the ‘‘discriminating alignment” hypothesis
that transactions will be allocated to particular governance
mechanisms in order to minimize the cost of completing
those transactions.
Transactions, which differ in their attributes, are
aligned with governance structures, which differ in
their cost and competence, so as to effect a discriminat-
ing—mainly a transaction cost-economizing—result
(Williamson, 1996, p. 12).
Williamson’s work proved to be particularly useful for
challenginganti-trust laws byshowing that vertical integra-
tion in some industries was economically rational rather
than simply an attempt to create monopoly power. In
general, empirical support has been greatest for the rela-
tionship between asset speci?city and the choice of gover-
nance form but inconsistent for other dimensions of
transactions (Shelanski & Klein, 1995). Williamson’s
(1975) insight has also been used in the accounting litera-
ture to suggest that management control systems, as a key
aspect of governance systems, would also vary according
to the characteristics of transactions (Anderson & Dekker,
2005; Spekle, 2001). It has also informed studies of the
1
Santos and Eisenhardt (2005) provide a review of alternative concep-
tions of the formation of organizational boundaries.
2
Williamson explicitly recognizes government agencies/departments as
a valid governance structure subject to transaction cost analysis: ‘‘Trans-
action cost economics views the public agency as a candidate mode of
governance that is well-suited for some purposes, poorly suited for others”
(Williamson, 1999, p. 311). The term ‘‘hierarchy” is used to refer to both
private and public sector bureaucracies in which resources are allocated by
management.
940 A.J. Richardson, E. Kilfoyle / Accounting, Organizations and Society 34 (2009) 939–956
governance structures adopted to minimize the cost of
transactions across organizational boundaries (Anderson &
Dekker, 2005; Dekker, 2004).
In spite of the relevance of these core insights to
accounting, Coase (1990) lamented the (relative) lack of
use of institutional economics in accounting. From his
point of view the ?rm existed as an alternative governance
mechanism to the market and within this governance
mechanism accounting information played the same role
as prices in allocating resources. He saw accounting aca-
demics as being well placed to contribute to and drawfrom
transaction cost economics.
In. . . the ?rm, costs do not, in the main, arise directly
out of the operations of the market but are computed
and provided by the accounting system. While outside
the ?rm prices and therefore costs are explicit (because
of the demands of others for resources) and are deter-
mined by the operations of the market, within the ?rm
there are explicit costs for exactly the same reason but
they are provided by the accounting system. This inter-
nal system takes the place of the pricing system of the
market (Coase, 1990, p. 11).
This view of the role of accounting as a substitute for
prices as transactions are moved out of markets has been
noted by others.
Without market prices for these goods, the ?rm must
rely on relatively costly and inef?cient methods of gen-
erating. . . accounting prices to perform internal calcula-
tions (Klein, 1996, p. 4).
Firms engage in non-market contracting, on their own
internal terms (here labeled ‘quasi-prices’). These are
not market prices and need not resemble market prices
in either form or magnitude; and market prices are not
suf?cient to guide ?rms’ decisions. Hence, a whole range
of institutional phenomena, including the sets of ‘trans-
fer prices’, standard costs and the numbers on ?rms’
income statements and balance sheets, are hypothe-
sized as being neither market prices nor estimates of
market prices. In this model, accounting is viewed as a
specialist function for providing information that assists
?rms in establishing their quasi-prices, or even for pro-
viding the quasi-prices themselves (Ball, 1989, p. 3).
Just as the ?rm is an alternative to the market system,
accounting in this setting is an alternative to market
prices (Emanuel, Wong, & Wong, 2003, p. 154).
Thus accounting historians such as Johnson (1975),
Johnson (1983), Kaplan (1984), Spraakman and Davidson
(1998) and Spraakman and Wilkie (2000) draw on Coase
(1937) and Williamson (1975) as theoretical prolegome-
non to the discussion of the emergence of management
accounting.
3
They see management accounting techniques
as arising in concert with the development of large scale
organizations to provide a substitute set of information
to market prices on which managers could make resource
allocation decisions and to evaluate performance.
The simple story of management accounting informa-
tion arising to replace market price information when
transactions are removed from the market may be based
on a simplistic view of the nature of markets. Markets
are based on contracts and accounting information is
used to support contracting even in a market setting.
4
Rosen (1988) contrasts the completion of production in
a setting where all relationships are based on con-
tracts (historically this was known as the ‘‘putting out”
system of production) with a hierarchy where wage
labour is used without specifying the tasks to be com-
pleted:
The number of prices necessary to manage it can be
very large indeed. However, a simpler mechanism
may be available; one person retains all residual rights,
assembles the appropriate team of workers on a
contractual basis, assigns them to their most productive
positions in the ?rm, and monitors their work. The
terms of these contracts must specify standards for
the quality and quantity of work, as well as employ-
ment conditions regarding working hours and regular-
ity of employment, these non-price dimensions of
contracts being necessary to internalize technological
dependencies among workers. Financial terms of con-
tracts are constrained by competition for workers in
the labour market. Concentrating control in this way
and establishing a wage system may be a less complicated
way of achieving ef?ciency than designing and monitoring
an elaborate accounting system and calculating the indi-
vidualized prices required by a decentralized internal
transfer-pricing mechanism. (Rosen, 1988, emphasis
added)
While Rosen (1988) is not an accountant and his paper
focuses on labour markets, this paragraph captures the
dilemma; compared with the record-keeping needed for
a complex multi-stage production process run by con-
tracts through the market, a hierarchical system of control
might need less accounting rather than more. If this is
so, then the basic insight drawn from Coase (1937) in
the accounting literature that management accounting
information develops to replace market prices may be
in error or, at least, such statements may be over
generalized.
It is not uncommon to see accounting used in contracts
between the ?rm and external stakeholders to monitor and
report on contingencies or to track contract performance
over time (Watts & Zimmerman, 1986). So from this per-
spective, accounting is part of the contractual controls
used between arm’s length contracting parties as well as
being used within the ?rm to control operations. As Coase
(1937, p. 391) notes ‘‘It is true that contracts are not elim-
inated when there is a ?rm but they are greatly reduced”.
Consequently, the relative use of accounting in markets
(contracts) and hierarchies or networks to complete equiv-
alent transactions is the key empirical issue that arises
from this theoretical perspective.
3
See Hopwood (1987), Jones and Dugdale (2001), and MacDonald and
Richardson (2002) for alternative interpretations of the emergence of
management accounting.
4
Basu, Kirk, and Waymire (2008) suggests, based on ethnographic data
across time and societies, that record-keeping is a necessary precursor for
the development of markets and other governance structures.
A.J. Richardson, E. Kilfoyle / Accounting, Organizations and Society 34 (2009) 939–956 941
Accounting in networks
A further complication to the use of transaction cost
economics to study accounting is the rise of hybrid gover-
nance structures that combine the characteristics of mar-
ket and hierarchy. Miller, Kurunmaki, and O’Leary (2007)
note that hybrid forms are not a single, stable class of gov-
ernance structures and hence the forms of accounting that
go along with them are also in a state of ?ux. In spite of this
caveat, however, one hybrid form of governance has
emerged as a focus of empirical work: networked organi-
zations (Powell, 1990; Podolny and Page, 1998).
Networks may be de?ned as
. . .any collection of actors (N > 2) that pursue repeated,
enduring exchange relations with one another and, at
the same time, lack a legitimate organizational author-
ity to arbitrate and resolve disputes that may arise dur-
ing the exchange (Polodny and Page, 1998, p. 59).
The existence of ‘‘enduring exchange relationships” dif-
ferentiates the network from markets in which transac-
tions are completed on a ‘‘spot” basis while the lack of a
‘‘legitimate organizational authority” differentiates the
network from hierarchical governance structures (this is
not to imply that a network of actors cannot voluntarily
create rules for resolving disputes but these are an out-
come of the network and not a priori to the network or
the result of authoritative intervention).
The original formulation of transaction cost economics
was based on the comparison of a market of individual fac-
tors of production (i.e. labourers, owners of resources or
capital) versus a ?rm. Given the complexity of modern pro-
duction technologies and the economies of scale of large
organizations, the role of isolated factors of production in
the economy has decreased and ?rms/organizations are
now the dominant governance structure (Presthus, 1962).
The market is now populated by ?rms dealing with other
?rms rather than individuals.
Within this milieu, the issue may no longer be the rela-
tive transaction cost ef?ciency of ?rms versus markets
(Coase, 1990, p. 11) rather it becomes a question of when
a ?rm must network with other ?rms to overcome the lim-
itations of its own governance structure. This requires
viewing the value chain as a system and inquiring into
the costs of systemic coordination as well as the costs of
each unit within the system of production.
What it does mean, if I am right, is that. . . ‘to explain the
institutional structure of production in the system as a
whole it is necessary to uncover the reasons why the
cost of organizing particular activities differs among
?rms’.. . . If economists are to study the determinants
of the costs of organizing various activities within ?rms,
they will have to call in the assistance of accountants
since the costs of organizing clearly depend on the ef?-
ciency of the accounting system. (Coase, 1990, p. 11)
To summarize, the limits of the ability of the ?rm to
internalize transactions may not be resolved by employing
market mechanisms particularly as organizations come to
dominate the economy; rather the resolution may be to
arrange ?rms within network structures that allow a
‘‘domesticated” market to arise (Arndt, 1979). A ‘‘domesti-
cated” market is one in which repeat exchanges occur on
the basis of negotiated rules between parties. The rules
developed in domesticated markets typically combine
market mechanisms (i.e. contracts enforced by the courts)
and hierarchical mechanisms (i.e. rules that are created
and enforced through a voluntary enforcement mechanism
created by the network participants).
The conceptualization of networks as hybrid gover-
nance structures that have characteristics of both markets
and hierarchies has been adopted by numerous authors in
the accounting literature (e.g. Anderson & Dekker, 2005;
Cooper & Slagmulder, 2004; Dekker, 2004; Hakansson &
Lind, 2004). These studies are limited, in terms of the is-
sues explored here, by their focus on the role of accounting
within existing networks rather than exploring the changes
in the use of accounting before and after the formation of
the network, or more precisely, by focusing on the role of
accounting in completing transactions that have migrated
from one form of governance to another. The lack of con-
trol for the uses of accounting in market contracting or
within hierarchies for a given set of transactions makes it
dif?cult to make valid inferences about the causal role of
network formation on the use of accounting. The theoreti-
cal relationship between accounting and networks cannot
be addressed by these studies.
The role of mundane and opportunistic transaction costs
The view of accounting from a transaction cost econom-
ics perspective has been dominated by Williamson’s
emphasis on opportunism as the key problem to be over-
come by governance systems (e.g. Tiessen & Waterhouse,
1983). This emphasis differs from Coase’s view of transac-
tion costs and has led accounting researchers to ignore a
dimension of transaction costs that we believe provides
important insights into the role of accounting. This is the
distinction between opportunistic and mundane transac-
tion costs.
Williamson has frequently used the metaphor of trans-
action costs as a ‘‘friction” in economic exchanges (Klaes,
2000).
Although failures can be and often are assessed with
respect to a frictionless ideal, my concern. . . is with
comparative institutional choices. Only to the extent
that frictions associated with one mode of organization
are prospectively attenuated by shifting the transac-
tion. . . to an alternative mode can a failure be said to
exist. Remediable frictions thus constitute the condi-
tions of interest. (Williamson, 1975, p. 20)
In mechanical systems we look for frictions. . . The eco-
nomic counterpart of friction is transaction cost (Wil-
liamson, 1985, p. 1).
But the ‘‘frictions” that concern Williamson are those
associated with opportunism, bounded rationality and
small numbers bargaining. He is concerned with the ability
of one party in an exchange to cheat another and transac-
tion costs represent the losses due to this problem and the
costs of mechanisms to reduce this possibility. This
approach to transaction costs underlies most uses of TCE
942 A.J. Richardson, E. Kilfoyle / Accounting, Organizations and Society 34 (2009) 939–956
in the accounting literature: accounting is associated with
the monitoring and bonding processes that reduce agency
costs.
Several authors have noted, however, that Coase and
Williamson de?ne transaction costs in different ways
(Baldwin, 2008; Baldwin & Clark, 2003; Langlois, 2006;
Pessali & Fernandez, 1999). Coase is concerned with the
costs of using a governance mechanism (e.g. the cost of
?nding someone to transact with, negotiating a price, writ-
ing a contract etc.) while Williamson is concerned with the
costs of opportunistic behavior (incentive effects and
agency costs). Baldwin (2008) refers to these two sets of
costs as ‘‘mundane transaction costs” and ‘‘opportunistic
transaction costs” respectively. Even in the absence of
opportunism, in order to complete transactions it is neces-
sary to de?ne the properties of the good or service to be
exchanged, measure the properties of the goods or services
actually exchanged, and arrange for compensation be-
tween parties. The costs associated with de?ning, measur-
ing and compensating for transactions are mundane
transaction costs. In the accounting literature to date we
have focused exclusively on the effect of accounting on
opportunistic transaction costs but the effect of accounting
on, or as, mundane transaction costs has been ignored.
In contrast to the conventional engagement with TCE to
understand the role of accounting within a single gover-
nance mechanism, our empirical work focuses on the
changes in the use of accounting as transactions migrate
from one governance mechanism to another, speci?cally
from markets to networks, over an extended period of
time. Our focus on equivalent transactions moving
between governance mechanisms allows us to develop a
re?ned understanding of the role of accounting in these
settings compared with studies of accounting within a sin-
gle governance mechanism. We conceptualize the change
in governance mechanism as an effort to minimize transac-
tion costs examining the in?uence of both mundane and
opportunistic transaction costs on the redesign of the
governance mechanism. Our expanded conceptualization
of transaction costs allows us to re?ne the interpretation
of Coase’s and Williamson’s theories in accounting.
Method
We undertake a longitudinal case study of the interna-
tional postal system between 1840 and 1875. There is a
long tradition in institutional economics of using case anal-
ysis to resolve theoretical issues. This arose largely because
of the use of stylized institutional facts by marginalist eco-
nomics authors attempting to improve the face validity of
theoretical statements. Institutional economists have
examined the details of these anecdotes and frequently
found that real institutions have evolved in ways to deal
with problems that standard economics argues are ‘‘mar-
ket failures”. For example, it was argued that lighthouses,
designed to protect shipping in dangerous waters, are an
example of a public good that must be provided by the
state because the free rider problem associated with the
market would result in the underproduction of this good
and hence a loss of social welfare. Coase (1974) undertook
an historical study of lighthouses in the UK and found that
they were being provided for hundreds of years in a free
market. Ship owners using the ports at the terminus of
the shipping lanes protected by the lighthouses formed a
voluntary organization to operate lighthouses and collect
fees to pay their costs. No government regulation was
needed. A second example was the case of apiaries (bee
keeping) to illustrate the problem of externalities. Many
crops require bees to pollinate their ?owers and bee keep-
ers need ?owers for their bees to make honey. The argu-
ment was that since in each case there is a positive
externality that has no market price, there would be an un-
der-supply of bees compared with the needs of farmers to
pollinate their crops and an under-supply of honey com-
pared with market demand. Cheung (1973) undertook a
study of the actual relationship between farmers and api-
aries and demonstrated that well developed social norms
ensured that both pollination services and honey were pro-
duced in ef?cient quantities.
5
The use of case studies in the service of theory testing
can take one of two forms. First, if theory makes non-prob-
abilistic statements about the phenomenon, then a case
study can provide unequivocal falsi?cation of the theory
(i.e. if a theory predicts that black swans do not exist, ?nd-
ing one black swan disproves the theory). The studies by
Coase (1974, 2000) and Cheung (1973) are of this nature.
Since each example had been used to demonstrate the abso-
lute need for a non-market (state) solution, illustrating that
the participants had solved the issue without resort to state
intervention disproved the original assertion. Second, case
studies may provide multiple degrees of freedomfor testing
theory (Campbell, 1975). The use of a case is appropriate
where the context provides a clear situation to which the
theory applies and multiple dimensions of the case can be
used to test the predictions of the theory. The situation we
examine allows for the interdependence of observations
to be used to test a range of predictions of TCE inaccounting.
The distinction between theory–testing and theory–
development in case studies does not imply that a case
may be used for only one purpose. In particular, if a case
disproves a theory’s prediction, the case may also be useful
to develop boundary conditions for the original theory or
to suggest extensions (Eisenhardt, 1989). Williamson
(2005, p. 19), speaking speci?cally of historical studies in
business notes:
. . .there is no better way to discover the need for re?ne-
ments, corrections, and extensions than to do research
on puzzling phenomena for which the microanalytics
have yet to be carefully worked out.
We approach the present case with this rich view of the
potential of case studies. First, we believe that this case
provides clear data on the assumption that management
accounting information is developed to replace market
5
There is a negative correlation between the cost to the farmer of
‘‘renting” the hives and the total amount of honey produced. Thus if the
hives are placed in a ?eld that allows ef?cient honey production but
pollination is not needed, then no fee is paid to the bee-keeper; if the hives
are placed in a ?eld where pollination is essential but little honey is
produced, then a signi?cant fee is paid to the bee-keeper.
A.J. Richardson, E. Kilfoyle / Accounting, Organizations and Society 34 (2009) 939–956 943
price information when transactions are internalized with-
in hierarchical or network structures. Second, the case pro-
vides multiple perspectives on this issue since it deals with
a number of postal operators who came together to form
one network. The experience of each national postal oper-
ator who entered into the Universal Postal Union provides
additional insight into the phenomenon. Finally, the case
provides us with an opportunity to develop the application
of transaction cost economics to accounting phenomena in
a context where equivalent transactions migrate between
governance mechanisms over time.
6
Data
The UPU was created in 1875 as a venue for national
postal operators to negotiate the terms under which mail
would be exchanged between nations and the revenues
from postage would be divided among those involved in
providing the service from the sender in one country along
the value chain to the ?nal recipient in another country.
7
The UPU also served as a data clearing house to ensure that
information about mail volumes and other factors affecting
postal service were shared among participants in the
network. We examine the period from 1840 to 1875 to
identify the issues facing the ef?cient development of
international mail systems; these issues provided the con-
text in which the change in governance from spot market
exchanges to bilateral treaties and then to a network struc-
ture occurred. The earlier date, 1840, used as a cut-off for
the period prior to the formation of the UPU re?ects the
creation of the modern business model for postal services
in the UK with the passage of the Penny Post Act in that
year. The reforms brought about by the Penny Post Act in-
cluded prepayment of postage, the creation of postage
stamps and a drastic reduction in postal rates to encourage
greater volume of mail (Richardson, forthcoming). This was
a unilateral act on the part of the UK at a time when post-
age was normally collected on delivery of mail from the re-
cipient. Thirty-?ve years later the basic principles of UK
postal reform would become embedded in the Constitution
of the General Postal Union (now Universal Postal Union,
UPU). We end our sample period with the creation of the
?rst UPU Convention in 1875. The basic provisions of this
treaty remained stable until the 1960s consistent with
the conjecture that they represented an ef?cient solution
to the problems of the international mail system.
8
We examine the period between 1840 and 1875 to bet-
ter understand the accounting and business problems
associated with the international ?ow of mail and to cap-
ture the transition between discrete governance struc-
tures. Our data consists of bilateral postal treaties signed
between these dates, the ?rst UPU multi-lateral conven-
tion, and a small secondary literature. The treaties were
collected from the Royal Mail Archives (London, UK), the
archives of the UPU (Berne, Switzerland) and on-line
sources including the Yale University Law School Avalon
Project, and various web sites dealing with philatelic issues
(in particular the US Philatelic Classics Society/American
Philatelic Society). There is no catalogue of postal treaties
available for this time-period so we cannot claim that this
list is comprehensive or representative. On the other hand,
the decision to place these treaties in the public domain is
unrelated to our study and we do not anticipate any partic-
ular bias from the sample used. The treaties that we exam-
ine include documents in both French and English
9
and
involve countries that would become founding members of
the UPU and others that remained outside the UPU for some-
time after its formation.
The international postal system prior to the creation of
the UPU was fragmented and complex. It operated partially
on voluntary agreements that were subject to opportunism
that could not be resolved and partially on the basis of
bilateral treaties that attempted to provide accountability
and improve performance of the postal system. We will
discuss the evolution of accounting procedures within
the international mail system as re?ected in private agree-
ments and treaties prior to the formation of the UPU and
contrast these with the rules implemented by the UPU.
We will argue that the UPU was, at least in part, an institu-
tional solution to an accounting problem.
The focus on an international market involving state
agencies provides the additional bene?t that national gov-
ernment regulations had no direct effect on the structure
of the market, i.e. it is extra-territorial and therefore not
subject to legal intervention (Keohane, 1982, 1984). The
key effect of government regulation was to create national
postal monopolies and to regulate domestic postal rates.
National governments, however, could not unilaterally
6
Our view of history in this context is one of contingency not progress.
The migration of transactions between governance mechanisms re?ects
practical choices by actors in the face of cost and incentive functions
shaped by speci?c institutional structures among a host of other factors
affecting their choices. These institutional structures are endogenous, i.e.
they are shaped by actors as they enact their reality, and, hence, we do not
regard the UPU as a unique, inevitable or permanent outcome. Finally, the
logic used by actors in making their choices is affected by the broader
institutional logic or ‘‘regime” (Jones & Dugdale, 2001) in which the entire
process is embedded. These broader concerns are methodologically brack-
eted in order to focus on the insights of TCE for understanding management
accounting practice.
7
The original signatories of the UPU were ‘‘Germany, Austria-Hungary,
Belgium, Denmark, Egypt, Spain, The United States of America, France,
Great Britain, Greece, Italy, Luxemburg, Norway, The Netherlands, Portugal,
Roumania, Russia, Servia, Sweden, Switzerland, and Turkey” (the order of
countries and spelling is shown as they appear in the treaty).
8
As with all social phenomena, multiple interpretations of events are
possible. The use of transaction cost economics in this setting is consistent
with Keohane’s (1982, 1984) work on the origins of international regimes.
His work has been criticized, as might ours, for ignoring issues of power,
culture and the path dependence of social institutions. Williamson’s work
has been criticized as being overly atomistic and voluntaristic (Granovetter,
1985, 2005). Given the extensive literature that has critiqued Williamson’s
original formulation and Keohane’s application of transaction economics in
international relations, these views will not be covered here but are
recognized (see also Hopper & Armstrong, 1991; MacDonald & Richardson,
2002). We begin with the view that the situation we examine is consistent
with TCE, this theory has been used to make predictions of accounting
phenomena, and we believe that the setting we examine contributes to a
critical understanding of the potential application of TCE to accounting
phenomena.
9
French was the language of diplomacy from the 1700s until it was
replaced by English after the Second World War. The operating language of
the UPU, for example, was French until 1969.
944 A.J. Richardson, E. Kilfoyle / Accounting, Organizations and Society 34 (2009) 939–956
determine the structure of the international postal regime;
this required negotiations with other sovereign govern-
ments. The existence of postal monopolies in most coun-
tries and the nationalization of postal services, i.e. the
post of?ce was a government agency/department, did
facilitate the negotiation of an international solution to
the problems that were arising in a fragmented interna-
tional postal market.
In most settings, the development of any governance
mechanism is dependent upon the existence of the state
both as a direct regulator of business activity (e.g. provid-
ing rules regarding acceptable organizational forms, man-
dated disclosures etc.) and as the source of authority for
a court system that arbitrates voluntary agreements (con-
tracts). There is a complementarity between state-created
rules and voluntary rules of the type envisioned within
TCE in the sense that, at a minimum, the state provides a
means of enforcing voluntary contracts and places con-
straints on the conditions under which contracts may be
created (e.g. by prohibiting the use of coercion and requir-
ing disclosure of material facts affecting a contract) (North,
1990). One setting in which the impact of the state may be
reduced, allowing for more powerful tests of the insights of
transaction cost economics, is in situations where transac-
tions occur internationally particularly in transactions be-
tween state agencies. This setting gives priority to
voluntary agreements (treaties) that are self-enforcing
since the enforcement of contracts by any one agency be-
comes problematic because of jurisdictional disputes and
con?icts of laws between countries.
The postal treaties included in our database are listed in
Table 1. To supplement our direct examination of postal
treaties we have also relied on Hargest (1971) and Courtis
(2004). Hargest (1971) provides an analysis of postal trea-
ties and practices between the US and various European
countries during the period from 1845 to 1875. His work
Table 1
Postal treaties (1840–1875) consulted.
Agreement between the General Post Of?ce of London and the Post Of?ce of Hamburg, January 27, 1841
Postal Convention between US and Great Britain, December 15, 1848
Postal Convention between US and Bremen, February 6, 1849
US Treaty with the Hawaiian Islands, December 20, 1849 (this is a treaty concerning trade and communications;
postal issues are dealt with in Article 15)
Postal Convention between US and Canada, March 25, 1851
Postal Convention between US and Prussia, April 26, 1852
Postal Convention between US and Bremen, August 4, 1853
Postal Convention between the Hawaiian Kingdom and the French Protectorate Government of Tahiti, November 24, 1853
Postal Convention between US and Prussia, August 29, 1855
Postal Convention between Great Britain and France, September 24, 1856
Postal Convention between US and France, April 1, 1857
Postal Convention between US and Hamburg, June 30, 1857
Agreement between the General Post Of?ce of London and the Post Of?ce of Hamburg, June 29, 1859
Agreement between the General Post Of?ce of London and the Post Of?ce of Hamburg, March 26, 1860
Postal Convention between US and Belgium, December 21, 1860
Postal Convention between US and Prussia, April 1861
Postal Convention Between US and the Republic of Mexico, December 11, 1861
Postal Convention between US and the Republic of Guatemala, July 16, 1862
Agreement between the General Post Of?ce of London and the Post Of?ce of Hamburg, December 9, 1862
Postal Convention between US and Italy, July 8, 1863
Postal Convention between US and Venezuela, June 26, 1865
Postal Convention between US and Great Britain, November 11, 1865
Postal Convention between US and Great Britain, June 18, 1867
Postal Convention between US and Belgium, August 21, 1867
Postal Convention between US and the Netherlands, September 26, 1867
Postal Convention between US and Switzerland, October 11, 1867
Postal Convention between US and North German Union, October 21, 1867
Postal Convention between US and Italy, November 8, 1867
Postal Convention between US and Great Britain, November 24, 1868
Postal Convention between US and Switzerland, March 26, 1869
Postal Convention between US and Italy, May 25, 1869
Postal Convention between US and Italy, February 8, 1870
Postal Convention between US and Belgium, March 1, 1870
Postal Convention between US and Brazil, March 14, 1870
Postal Convention between US and the Hawaiian Kingdom, May 4, 1870
Postal Convention between US and British Columbia, June 9, 1870
Postal Convention between US and the Netherlands, June 15, 1870
Postal Convention between US and Salvador, October 5, 1870
Postal Convention between US and New Zealand, October 5, 1870
Postal Convention between US and German Empire, March 31, 1871
Postal Convention between US and Ecuador, May 9, 1871
Postal Convention between US and Great Britain, July 27, 1871
Postal Convention between US and Newfoundland, December 1, 1872
Postal Convention between US and the United Kingdoms of Sweden and Norway, May 26, 1873
Postal Convention between US and France, April 28, 1874
Postal Convention between the Colonial Government of New South Wales and the Hawaiian Kingdom, July 1, 1874
Universal Postal Union Convention, 1875
A.J. Richardson, E. Kilfoyle / Accounting, Organizations and Society 34 (2009) 939–956 945
focuses particularly on the markings used on letters to
track changes in postal rates, routings and the distribution
of revenues. Courtis (2004), although not referring to Har-
gest’s (1971) work, provides a parallel analysis of postal
markings used on letters between Australia and the UK be-
tween 1843 and 1876. We used these sources to develop
an expanded set of business issues affecting the post dur-
ing our focal time period.
An overview of postal services between 1840 and 1875
In 1840 the UK redesigned its postal system to require
prepayment of postage rather than, as was then customary,
charging the recipient for the cost of delivering the letter.
This allowed much simpler delivery methods and a dra-
matic reduction in postal rates (Richardson, forthcoming).
After 1840 postage rates were set to encourage but not re-
quire prepayment of postage. Thus a prepaid letter would
be delivered for one rate and an unpaid letter would be
delivered at a higher rate. International mail posed a par-
ticular problem. As Table 2 indicates, after postal reform
but before the universal acceptance of prepayment for
mail, there were four potential combinations of postal pol-
icies between originating and destination posts. In cases
(1), (3) and (4), there must be a negotiation between the
posts to divide the total postage charged or received. Only
in case (2) can the posts operate independently, i.e. with
mail physically transferred between states on a spot basis
without an accounting for the costs incurred. To further
complicate this system, even in countries where prepay-
ment was available, it was not usually mandatory. Custom-
ers could choose to prepay or not depending on their
assessment of the relative costs in the receiving and send-
ing jurisdictions and based on their ability to pay the post-
age at the time of mailing. Thus, at the customer’s
discretion, countries could move from case (1) to case (3)
and from case (2) to case (4) for some proportion of their
total mail ?ow.
Key issues in the early postal system
The compensation exchanged for services provided by
one national post of?ce to deliver the mail originating with
another national post of?ce required complex negotiations
dealing with a series of contractual problems that arose in
bilateral relationships. We identify seven of these prob-
lems and provide a brief discussion of the characteristics
and implications for the postal system. These problems
formed the context in which the bilateral treaties de-
scribed in the next section arose.
Variation in currencies. The delivery of mail at this time
could take months between distant locations. This meant
that, particularly for prepaid mail, there was both an in?a-
tion risk and a currency exchange risk that had to be borne
by the postal system. The uncertainty of costs could be re-
duced by creating bilateral postal treaties which speci?ed
the basis of exchange but this simply shared the exchange
rate risk between the post of?ces. Postal treaties were
established and maintained for an indeterminate period
but usually measured in years before being renegotiated.
In addition, the settlement of accounts between countries
was done, at most, on a quarterly basis but typically only
annually. At the time of negotiation, revenue sharing rules
were based on an assumed rate of exchange between the
two currencies. If one currency appreciated against the
other, they would ?nd that the share of revenue from the
other country represented a decreasing payment for ser-
vices while their payment to the other country would pro-
vide a higher return than expected. Currency ?uctuations
during this period could severely undermine the revenue
sharing between countries.
Variation in weight scales. The use of both imperial and
metric weight scales, and different thresholds on these
scales at which increments in postage would be charged,
proved to be a problem for some exchanges of mail. For
example, mail exchanged between the US and France
encountered this problem when the US rate of postage
was based on a ½ ounce letter while the French based their
postage on a 15 g letter (the exact equivalence is ½ oun-
ce = 14.175 g). This slight discrepancy in the standard ap-
plied to postage could create circumstances where a
letter might be charged single postage in France but would
require double postage in the US (e.g. for a letter weighing
between 14.175 and 15 g). The US postmaster recognized
that this difference in weight scales was resulting in an un-
der payment by France to the US (Gough, 2005).
The choice of weight scales was also affected by the
technology of paper production. The French, for example,
were routinely using lighter weight paper stock than Ger-
many so that the French preference for a 10 g letter rate
between these two countries would result in much higher
costs in Germany than France for an equal volume of mail.
The problem was also exacerbated by the decision to use
Table 2
Combinations of postal policies between 1840 and 1875.
Postal system at destination
Prepaid Cash on delivery
Postal system at point of origin
Prepaid (1) Originator retains postage paid in compensation for outbound
logistics and pays destination post of?ce for delivery services at
normal mail rates; destination post of?ce delivers mail without
additional charge
(2) Originator retains postage paid in compensation for outbound
logistics; mail is transferred without charge to the destination post
of?ce; destination charges normal domestic rate for delivery
Cash on
delivery
(3) Originator receives no compensation for outbound logistics from
customer; outbound mail is sold to the destination post of?ce who
charges higher unpaid rate for delivery
(4) Originator receives no compensation for outbound logistics from
customer; outbound mail is sold to the destination post of?ce;
destination charges normal domestic rate for delivery
946 A.J. Richardson, E. Kilfoyle / Accounting, Organizations and Society 34 (2009) 939–956
weight as a proxy for the postage collected on prepaid
mail. If, on average, French letters weighed less than Ger-
man letters, even where both were within the limits for
single postage, then for the same gross weight (and there-
fore the same estimate of postage revenue), French mail
would represent more items and a higher delivery cost
than German mail. On balance, this would result in Ger-
many subsidizing the delivery of French mail in Germany.
The use of postal rates as diplomatic tools. Hargest (1971, p.
4) suggests: ‘‘Postal rates. . . were conceived to have func-
tions beyond those of paying for a service or of producing
a revenue. Postal rates could be used as an instrument of
diplomacy; they could be used to promote trade; or. . . they
could be used as a protective tariff”. This observation was
primarily in reference to the UK who, in spite of reduced
postage for domestic mail, continued to charge much high-
er rates for some international mail while subsidizing mail
?ows between the UK and its colonies. This encouraged
domestic/imperial trade and correspondence at the ex-
pense of cross-border trade and correspondence. Although
beyond the scope of this paper, postal rates also varied be-
tween letters and other forms of mail. For example, news-
papers and magazines received privileged rates based on
the negotiation of publishers with the post of?ce and sup-
ported politically as helping to develop the intellectual and
moral strength of the nation. However, this privileged rate
also meant that there were incentives for people to use
newspapers and books as a cover for more personal com-
munications. This meant that the post of?ce had to create
standards to differentiate generic newspapers from per-
sonal communications on newspapers and to monitor for
violations of these rules. This raises the transaction costs
associated with the postal system overall.
Issues in the transit of mail across third jurisdictions. The
transport of mail from one country to another might, in
some cases, be most ef?ciently achieved by transit across
the territory of a third country. For example, mail to the
western coast of South America from Europe might be best
sent to New York, transported by train to San Francisco and
then by boat to its ?nal destination. In these cases the
country providing transport would charge for their ser-
vices. Note that foreign countries did not have the option
of negotiating with private freight companies to make
these transfers after national postal operators were created
and given a monopoly. This provided certain national posts
with unique opportunities. In Europe in the early 1800s, for
example, Belgium found that it could bene?t by acting as a
transit point for mail between France and Prussia and be-
tween rest of Europe and the UK. This central location
within the postal system allowed some posts to charge
amounts that were very pro?table. France and Prussia,
for example, found that they needed to establish a postal
treaty even though they had been at war and not reestab-
lished other normal relations because of the high cost of
exchanging mail via Belgium. These situations arose be-
cause of the opportunity cost of alternative routings. The
country that had a territorial advantage in transportation
routes could raise its prices up to the cost of the next best
available route.
Variation in international carrier costs. Courtis (2004) pro-
vides a detailed description of accountancy marks that
were recorded on letters travelling between the UK and
Australian colonies during this period. These markings, in
red or black, signaled the distribution of revenues among
parts of the value chain. One complication facing clerks
creating these records was that the distribution depended
upon the carrier transporting mails, in particular, whether
the carrier was under contract to the post of?ce or a pri-
vate vessel. The UK had established a law requiring private
vessels to transport mail on demand. This allowed mail to
be sent on the next available departure rather than being
limited to the schedules of contract ships. The private ves-
sels were compensated for carrying mail at a ?xed rate in
the UK although in the US the amount paid would vary
depending on the type of ship (e.g. sail versus steam, Har-
gest, 1971). The delivery of each letter thus had a unique
cost/price depending on the precise routing used.
Variation in domestic carrier costs. The variation in the cost
of international transport between two countries (i.e. port-
to-port) was at times less than the variation in cost within
a country, particularly for large countries such as the US.
Initially some countries attempted to charge postage based
on the actual distance travelled. This approach however
was not tractable, i.e. the mundane transaction costs of
measuring and charging postage per mile was too high
compared with the value of the service, and soon gave
way to simpli?cations.
10
The Franco–Prussian Treaty of
1817, for example, divided France and Prussia into ‘‘ray-
ons” or districts based on the distance from the border ex-
change point (Whiteside, 2002). In the US, domestic mail
was charged at two rates: one rate for mail travelling less
than 300 miles and a second rate for distances greater than
300 miles. Hill (1837), contrary to these systems, advo-
cated a single rate to any point within the UK based on
his ?nding that the variation in cost was trivial (this was
based on two assumptions, that the mail coaches were
travelling under capacity so additional mail had zero mar-
ginal cost, and that the variation in cost was smaller than
the minimum amount that could be billed (i.e. an integer
programming problem)).
Lack of internal controls/auditability. Courtis (2004) notes
that the use of letters as accounting records within the
postal system meant that there was no audit trail once
the letter was delivered. The letters were marked to indi-
cate the distribution of postal charges between the origi-
nating post of?ce, shipper and destination post of?ce.
Summary accounts were created from these markings
but the markings were turned over to the recipient of the
letter and lost from the system. If a letter was not deliver-
able, the envelope would be returned and the markings
used to reverse the original entries. There was, however,
no means to verify the original entries.
The set of issues described above suggests that if the ex-
change of mail between national postal operators was to
10
This is the problem of the ‘‘?neness” of the system. There is a tradeoff
between the costs of increasing the ?neness of a costing system and the
bene?ts from better decision-making/resource allocation.
A.J. Richardson, E. Kilfoyle / Accounting, Organizations and Society 34 (2009) 939–956 947
be based on cost, then the contracts written between oper-
ators would have to take into account multiple factors
including: currency ?uctuations, variation in the mode
and routing of shipments, the volume of mail and its com-
position, the ?nal destination of mail and the distances be-
tween the port of entry to a country and the ?nal
destination of the mail, and ancillary non-market issues
such as political interference in mail rates.
From the perspective of TCE, the transactions at issue
are frequent (the volume of international mail was increas-
ing exponentially during this period, see Courtis, 2004, pp.
386–387) and undertaken under conditions of uncertainty
(particularly with respect to exchange rates, transportation
costs and the composition of the mail). There was very lit-
tle asset speci?city in this market with the exception of
some mail transit facilities provided by third-parties for
mail crossing their jurisdiction. The sorting facilities were
created to support domestic mail services and so are not
subject to ‘‘hold-up” to international mail. The transporta-
tion services were provided by regular commercial ship
and railroad lines as a marginal cost activity although some
special purpose facilities were built as the volume of mail
increased. The main issue was simply the high costs of
monitoring, reporting and reconciling the provisions of
contracts between countries; these are mundane transac-
tion costs.
Attempts to deal with market problems through bilateral
treaties
Nations exchanging signi?cant volumes of mail at-
tempted to deal with these complexities in the spot market
by negotiating bilateral treaties that reduced the frequency
of negotiation and resolved some of the uncertainties of
market exchanges. For example, the UK and Hamburg
established a postal treaty in 1859 for letter mail under
½ ounce (in the UK) or under 1 zoll loth (16 g, in Hamburg).
The treaty between the UK and Hamburg speci?ed that the
UK post of?ce pay the Hamburg post of?ce 3d for every
prepaid letter originating in the UK and 4d for every un-
paid letter originating in Hamburg (and a reciprocal agree-
ment for other mail). The 1860 treaty created a more
complex revenue sharing rule:
(1) For prepaid letters (6d) originating in the UK, the UK
receives 4½d and Hamburg received 1
1
2
d.
(2) For prepaid letters (5 silver groschen (sg)) originat-
ing in Hamburg, Hamburg receives 1
3
12
sg and the
UK received 3
9
12
sg.
(3) Unpaid letters were charged 8d. in the UK and 7 sg
in Hamburg with the UK receiving
3
4
of this amount
and Hamburg receiving
1
4
.
(4) Ship owners who conveyed the mail between the UK
and Hamburg were paid 1d. All of these costs were
paid by the UK post (or reimbursed to Hamburg if
paid by them).
The rewritten treaty recognizes the UK’s dominance of
commercial shipping and shifts more revenue to the UK.
This latter aspect may re?ect differences in cost or simply
the greater negotiating power of the UK. Note, for example,
that after allowing for shipping costs the UK received more
of the total revenue than Hamburg under this treaty.
The 1862 treaty between the UK and Hamburg intro-
duces accounting procedures to capture and reconcile
these revenue sharing agreements. The treaty requires
monthly accounting reports, agreed to by both parties,
and quarterly payment of the fees agreed to by the parties.
It also provided a table of postal rates and revenue sharing
rules for mail ?owing through the UK and Hamburg from
other countries (see Fig. 1). Note in this table the complica-
tions that arise when mail is ?owing between nations with
different postal policies (pay-on-delivery versus prepaid
postage). Even with this treaty, a signi?cant amount of
accounting was required to track letter ?ows and costs to
allow the reconciliation and payment. A similar treaty be-
tween the US and Hamburg in 1857 required reconciliation
of postage revenues on a ‘‘letter-by-letter” basis.
The 1862 UK-Hamburg treaty also allowed the UK to
forward mail to Hamburg through Belgium, with which it
had a separate treaty, should it be more cost effective to
do so. The introduction of accounting into these contracts
arises because of the spatial and temporal separation of
technological transactions (i.e. physical mail) from the
?nancial transactions that completed the exchange. Rather
than have money change hands each time mail changed
hands, accounting allowed the parties to aggregate the
many small, frequent transactions into a more cost ef?-
cient contracting structure.
Even though accounting simpli?ed the costs of transac-
tions compared with a spot market, the complexity of the
accounting/bookkeeping within the international postal
system was clearly of concern. Accounting represented a
deadweight loss on postal transactions.
11
The US Postmas-
ter General reported to Congress in 1861 that the costs of
incoming and outgoing mail tended to be roughly equal
(Table 3) and that signi?cant accounting work was being
done with virtually no impact on the money owing
between countries (Gough, 2005).
This ?nding was re?ected in US postal treaties begin-
ning with the Mexican postal treaty of 1861. Mexico at that
time collected its postage on delivery of a letter hence let-
ters from Mexico also entered the US without postage. US
mail to Mexico, even though the US used stamps for
domestic mail, were sent without postage. Article 3 of
the Treaty between Mexico and the US speci?ed that:
Upon all letters, newspapers, printed pamphlets, or
other printed matter received in the United States of
America from Mexico by sea, there will be charged by
the United States such rates of inland postage as are
now, or may hereafter be, established by the laws of
11
The precise accounting procedures used change as a consequence of the
change in governance mechanism. In spot markets, each letter was treated
as a job and actual costs were associated with the letter as a ‘‘marking” on
the face of the letter. This level of detail gradually gave way to estimates of
costs based on 6 steps: (1) estimating the weight of mail transferred
between countries (sampling); (2) estimating the number of pieces of mail
per unit of weight (sampling); (3) estimating the postage paid per unit of
mail (sampling of classes of mail); (4) calculating the imbalance in mail
?ows (inbound versus outbound), (5) reaching agreement between parties
on each estimate; and (6) arranging compensation according to an agreed
exchange rate.
948 A.J. Richardson, E. Kilfoyle / Accounting, Organizations and Society 34 (2009) 939–956
the United States, which shall be collected at the place
of destination, and shall belong exclusively to the Uni-
ted States of America, and vice versa, upon all letters,
newspapers, printed pamphlets, or other printed matter
received in Mexico fromthe United States of America by
sea, there will be charged by Mexico such rates of
inland postage as are now, or may hereafter be, estab-
lished by the laws of Mexico, which shall be collected
at the place of destination, and shall belong exclusively
to Mexico.
Similarly, the Treaty between the US and Newfound-
land in 1872, Article 3, speci?ed:
No accounts shall be kept between the Post Depart-
ments of the two countries upon the international cor-
respondence, written or printed, exchanged between
them, but each Department shall retain to its own use
all the postages which it collects thereon.
Both Newfoundland and the US used prepaid postage so
this clause reverses the onus for payment fromthe recipient
to the originator of the letter but in both the US-Newfound-
land and US-Mexico treaties, the requirement for account-
ing for revenues and revenue sharing was eliminated.
The US/Mexican postal treaty also speci?ed that mails
could cross each territory on route to a third destination
without cost (i.e. this service was provided as a marginal
addition to the second country’s internal mail distribution
system and/or based on the assumption that the ?ow of
mail across each country would be approximately equal
in cost allowing each country to barter their services).
Article 7 speci?es that:
The United Mexican States engage to grant to the
United States of America the transit, in closed mails,
free from any postage, duties, imposts, detention, or
examination whatever, through the United Mexican
Fig. 1. Postal rates for mail between the UK and Hamburg originating in or destined to other countries.
Table 3
Report to congress by the US Postmaster General (1861).
a
Direction Volume Value of mail (US dollars)
From USA to Europe 3,086,121 $675,997.29
From Europe to USA 3,059,700 $686,039.41
Difference (USA–Europe) 26,421 À$10,042.12
Differential 0.009 (9/10th %) À0.015 (1½ %)
a
This re?ects mail ?ow to the UK and France and involves assumptions about exchange rates in addition to measured postal volumes. The small
discrepancy between ?ows may be due to measurement errors rather than re?ecting true differences in the values of mail. (Source:http://www.rpsl.org.uk/
cpu/index.html Accessed 04.08).
A.J. Richardson, E. Kilfoyle / Accounting, Organizations and Society 34 (2009) 939–956 949
States, or any of their possessions or territories, of let-
ters, newspapers, printed pamphlets, or other printed
matter, forwarded from the United States of America,
or any their possessions or territories, to an other pos-
session or territory of the United States of America, or
to an foreign country, or from any foreign country, or
possession or territory of the United States of America,
their possessions or territories.
The United States of America on their part, engage to
grant to the United Mexican States the transit in closed
mails, free from any postage, duties, imposts, detention,
or examination whatever, through the United States of
America, or any of their possessions or territories, or let-
ters, newspapers, printed pamphlets, or other printed
matter, forwarded from the United Mexican States, or
any of their possession or territories, to any other Mex-
ican possession or territory, or to any foreign country, or
from any foreign country, or Mexican possession or ter-
ritory, to the United Mexican States, their possessions
or territories.
The elimination of transit charges for Mexican and US
mails in each others territories and the retention without
reconciliation of international postage paid in each country
largely eliminated the accounting associated with the
cross-border postal system. These clauses may be con-
trasted with the Treaty between the US and UK negotiated
in 1868 that speci?es in Article 6 that accounts are to be
kept and that the revenues collected by each party be
pooled and divided equally. These treaties re?ect apparent
differences between the US and UK regarding their beliefs
about the realized equality of cost and revenue ?ows. The
treaty between the UK and US requires accounting records
but the equal distribution of revenues presumes an equal-
ity of costs between the countries.
The UK however was also aware of the complexities of
accounting within the international mails. This was re-
?ected in a simpli?cation of revenue sharing under the An-
glo-French treaty of 1843 (Courtis, 2004). This treaty
introduced the use of rubber stamps rather than hand
markings on letters to indicate the distribution of reve-
nues. Only three markings were used representing a letter
fully paid to its destination (PD), paid to the frontier of the
receiving country (PP) or unpaid (P). This system elimi-
nated complex accounting for different carriers and rout-
ings from one country to the other.
The changes in the bilateral treaties reviewed above
suggest that the ?rst approach to reducing transaction
costs was to change from a unit (job) costing approach to
a batch (process) costing approach. By reducing the num-
ber of distinct accounting entries (e.g. by using a small
number of distance classes rather than actual mileage trav-
elled by each letter, or by requiring that postage be either
fully prepaid or fully postage due) the cost of record-keep-
ing could be substantially reduced. As contemporary activ-
ity based costing studies have made clear, however, this
results in cross-subsidization of different services and,
potentially, suboptimal investment and operating deci-
sions. The second approach was to eliminate accounting
where it could be demonstrated that mail ?ows were
approximately balanced such that no signi?cant exchange
of resources resulted from the detailed accounting for
transactions.
High mundane transaction costs of operating the inter-
national postal system through the market were suf?cient
to motivate a search for alternative governance structures
for the completion of these transactions. The fact that the
postal operators were government agencies meant that a
pure hierarchical solution was not possible – this would re-
quire each government to cede sovereignty to a single
transnational agency that would operate both the interna-
tional and domestic postal systems. The solution that
emerged was to create a network of domestic postal oper-
ators whose transactions would be subject to an agreed set
of rules that could reduce the transaction costs of the mar-
ket. The exact nature of these rules however is indetermi-
nate theoretically, i.e. transaction cost economics focuses
on changes between discrete modes of governance but
the detailed rules embedded in those modes of governance
will be speci?c to the nature of the goods or services ex-
changed. We thus examine the actual rules created and
consider the role of accounting within the network and
consequences of these rules on the functioning of the inter-
national postal system.
The uses of accounting in the ?rst UPU treaty
The UPU arose after an initial set of meetings in 1863 in
Paris called at the request of the US Postmaster General to
discuss the possibility of simplifying the international
postal system along the lines of their growing number of
bilateral treaties. This meeting was attended by fourteen
countries. The meeting did not result in speci?c agree-
ments and further progress was limited by the US civil
war (1861–1865) and the Franco–Prussian war (1870–
1871). A second meeting was held in Berne Switzerland
at the behest of the German Postmaster General who pre-
pared an initial document based on the postal agreements
between states of the German federation. This meeting
was attended by 22 countries and resulted in the creation
of the Universal Postal Union (Williamson, 1930).
The UPU Treaty in 1875 contained a remarkable clause
that states (emphasis added):
Each Administration shall keep the whole of the sums
which it collects. . . Consequently, there will be no
necessity on this head for any accounts between
the several Administrations of the Union. Neither
the senders nor the addressees of letters and other
postal packets shall be called upon to pay, either in
the country of origin or in that of destination, any tax
or postal duty other than those contemplated by the
Articles above mentioned.
The UPU had eliminated any accounting between mem-
bers of the postal network with regard to the services pro-
vided in one country to deliver the mail originating in
another country. It achieved this by (a) setting a uniform
rate of postage to be charged by all countries for interna-
tional mail denominated in a stable currency and with a
?xed weight scale for letter mail, (b) making the assump-
tion that the volume of mail and characteristics of the mail
950 A.J. Richardson, E. Kilfoyle / Accounting, Organizations and Society 34 (2009) 939–956
exchanged between countries was essentially equal
12
, and
(c) treating all members of the international postal network
as a single jurisdiction. In part, this approach to international
mail was due to Hill (1837) who had suggested at the time of
reform of the UK postal system that international mail is
delivered without cost in the UK and that international mail
from the UK is charged twice the domestic rate to compen-
sate for the unpaid inbound delivery of mail.
The use of a common rate of postage meant that ?ows
of mail could be accounted for by bulk weight rather than
by tracking the postage attached to individual letters. The
use of a single currency (the French centime) as the basis
of this postage meant that all postage rates, regardless of
the domestic currency, could be determined against a
known standard. The adoption of a ?xed international
postal rate across different countries eliminates the possi-
bility of arbitrage between countries, i.e. people shipping
mail to a country with low postage rates for onward deliv-
ery to a third country. This phenomenon has become
known as ‘‘remail”. Similarly by adopting the gram as the
of?cial unit of weight, problems associated with the con-
version between different postal scales disappeared.
These elements of standardization would have signi?-
cantly reduced the accounting burden facing the posts
but it would not have been suf?cient to eliminate the need
for accounting information. The assumption of equal ?ows
of mail between all postal partners was also required. This
is a heroic assumption because it assumes not only that
there is an equal volume of mail; it assumes also that the
composition of the mail is also the same. For example, if
the number of items per kilogram of mail varies between
countries even though the number of kilograms is the
same, then one of the countries will be required to spend
more to deliver the mail to its ?nal address than the other
country.
The ?nal assumption/policy required is that the UPU
treat its members as a single jurisdiction. By this time most
countries had adopted the norm that domestic mail was
delivered for a uniform rate regardless of the distance trav-
elled. This policy was not an unreasonable approximation
in densely populated and geographically compact coun-
tries such England where the policy originated. In coun-
tries such as Australia, Canada, India and the United
States, however, this assumption might not be reasonable.
The ?xed domestic mail rate used by many countries was
set to approximate a weighted-average of all possible pairs
of origins/destinations within the postal system. The UPU
policy extended that logic to the international system.
The international mail rate was chosen to approximate a
weighted-average of global mail ?ows and the ‘‘cost” of
those deliveries. This approach means that the postage
for letters does not depend on the routing used to send a
letter between two points. It should be noted, however,
that the ‘‘cost” was negotiated during the ?rst UPU conven-
tion without the bene?t of cost data. The basis for this
decision requires further analysis of historical materials.
The only accounting that remained within the UPU
treaty was to compensate countries that acted as carriers
for mail passing across their territory to a third country.
The country providing the service would keep a record of
the weight of mail transported and could claim a ?xed
fee virtually independent of distance (two rates could be
charged: one for under 750 miles, and a second rate for
longer distances):
The despatching Of?ce shall pay to the Administration
of the territory providing the transit, the sum of 2 francs
per kilogramme for letters...
This payment may be increased to 4 francs for letters. . .
when a transit is provided of more than 750 km in
length over the territory of one Administration.
It is understood, however, that in any case in which the
transit is already actually gratuitous or subject to lower
rates, those conditions shall be maintained.
Whenever a transit shall take place by sea over a dis-
tance exceeding 300 nautical miles within the district
of the Union, the Administration by or at the expense
of which this sea-service is performed shall have the
right to a payment of the expenses attending this
transport.
The members of the Union engage to reduce those
expenses as much as possible. The payment which the
Of?ce providing the sea-conveyance may claim on this
account from the despatching Of?ce shall not exceed
6 francs 50 centimes per kilogramme for letters...
In no case shall these expenses be higher than these
now paid. Consequently, no payment shall be made
upon the postal sea routes on which nothing is paid at
the present time.
Even for these costs the UPU took steps to reduce the
accounting required. Instead of using actual weights, the
treaty speci?ed that the payment would be based on a
sample conducted during a two week period. The regula-
tions that implemented the treaty stipulated that this sam-
ple data would be collected every three years. The
approach used assumes that the pattern of mail ?ow be-
tween countries was very stable both within a year and
over a period of years.
Discussion
At the beginning of our sample period the international
postal system, composed of agents transacting without any
governing agreements, attempted to operate on a cost
recovery basis. National postal operators delivering mail
originating with other national postal operators were
reimbursed depending on weight, distance, routing, cur-
rency of payment, mode of transportation, and timing of
payment. In addition, opportunism associated with coun-
tries and individual postal clients taking advantage of vari-
ations in weight scales, currency ?uctuations and unique
transportation resources were creating identi?able ex post
problems that, in an extra territorial setting, were dif?cult
to resolve. The high mundane transaction costs of operat-
ing on a spot market basis led countries exchanging signif-
icant volumes of mail to negotiate agreements that would
12
This is a cost/bene?t issue and depends on the costs of setting up
systems to monitor the composition and volume of mail ?ows versus the
materiality of the amount that would be paid between countries to correct
for the imbalance in mail ?ows.
A.J. Richardson, E. Kilfoyle / Accounting, Organizations and Society 34 (2009) 939–956 951
reduce the uncertainty of the exchanges and the need for
repetitive bargaining. Even with bilateral treaties in place,
however, the required level of detail in accounting for cost-
ing and revenue sharing became a signi?cant deadweight
cost on the operation of the postal system.
The bilateral postal treaties created between 1840 and
1875 provide evidence of adjustments to reduce mundane
transaction costs based on the parties’ experience with
costs. Initially, these treaties reduced costs by eliminating
the need for repetitive bargaining by establishing contract
terms that would apply to a period of time and reduced
uncertainty by specifying the exchange rates that would
be used to settle accounts. Even these contract features
however resulted in signi?cant accounting costs and fur-
ther innovations to reduce such costs were attempted.
The most extreme adjustments are evident in the US postal
treaties that eliminated accounting for transit costs and
revenue sharing based on the implicit assumptions that
(a) the volumes of mail exchanged between and travelling
through each country was approximately the same, and (b)
that the costs of service in each jurisdiction was approxi-
mately the same or that differences were immaterial com-
pared with the mundane transaction costs of identifying,
measuring and compensating for those differences.
The transition from bilateral treaties to the multilateral
UPU Convention further reduced transaction costs. The
convention eliminated much of the potential for opportun-
ism/arbitrage by standardizing the currency and weight
scales on which postal rates were based and by integrating
the transit system so that mail moved throughout the
world as if it were a single jurisdiction. The largest cost
saving however was mainly due to the reduction in mun-
dane transaction costs. Standardization and the creation
of a single postal territory removed the need for much of
the accounting for mail ?ows between countries that pre-
viously existed. International mail ?ows were monitored
by weight rather than by cost incurred or revenues col-
lected. Each national postal operator retained all postage
collected within their territory and provided delivery ser-
vices for foreign mail without charge to the originating
postal system.
The UPU Convention was a relational contract that pro-
vided a framework within which members could develop
norms for the operation of international postal system
(cf. Ouchi, 1980). It lowered the probability of opportunis-
tic behavior by creating ‘‘a shadow of the future” (Baldwin,
2008, p. 171) in dealings between member states but really
was premised on a ‘‘logic of con?dence and good faith”
(Meyer & Rowan, 1977) that opportunism between mem-
ber states was a minimal risk and that all states shared a
common interest in improving the volume and ef?ciency
of global mail ?ows. The creation of the UPU network di-
vided mail transactions into ?nancial and technical com-
ponents. The technical component, i.e. the physical
delivery of mail, was internalized within the network.
The ?nancial component, i.e. the collection of revenue
and incurring of costs, was internalized within each na-
tional postal operator. The UPU created, in essence, a barter
system between postal operators where services within a
country were traded for services in a second country with-
out any reconciliation of value while charging domestic
users of the international mail system for the outbound
logistics of their own international mail and the inbound
logistics of an equivalent volume of mail coming from
abroad.
The barter economy is sometimes discussed in TCE as
an example of a high transaction cost exchange that is
eliminated through the institution of ‘‘money” i.e. a med-
ium of exchange that is agreed to represent value within
a community. However, if we adopt Baldwin’s (2008) per-
spective, the barter arrangement can be interpreted as a
relational contract that governs the transfers within the
network to minimize opportunistic costs and reduce cer-
tain mundane transactions costs such as those incurred
by detailed accounting. The use of barter and countertrade
to overcome ?nancial market failures is gaining more
interest among economists who have observed its rise
among formerly planned economies where the lack of
trade credit and a sound banking system has forced com-
panies to seek alternatives to money-based exchange
(Mirus & Yeung, 1993). The UPU network brought together
a group of countries with a common need to provide reci-
procal services that were perceived to be equal in nature
and so could be held to be the same without resort to ex-
plicit mechanisms of valuation that would result in trans-
action costs exceeding the actual bene?ts gained from
these services. Following the introduction of new agents
and new communication technologies, this assumption
that the services were of equal value would ultimately re-
emerge as a challenge to the network.
It is important to note that the elimination of account-
ing between member states did not eliminate the need
for accounting within member states. In fact the opposite
is more likely since the internalization of the ?nancial
component of the transaction within each state meant that
they could gain by improving the ef?ciency of their ser-
vices independently of such decisions by other members
of the UPU. Recall that the UPU standardized international
postal rates. This meant that the revenue gained for inter-
national mail was ?xed, i.e. it was unrelated to the costs of
service within a state. If the member state could improve
the ef?ciency of outbound and inbound logistics for inter-
national mail, the member state would retain all ?nancial
gains through operating ef?ciencies.
In summary, the postal treaties show consistent evi-
dence of voluntary agreements to reduce mundane trans-
action costs and speci?cally to reduce the cost of
accounting by (a) standardizing key measurement scales
for currency and weight, (b) using averages to represent
a range of cost variation, (c) moving to a barter system
where the value of services was not monetized and (d)
omitting cost/revenue reconciliations where experience
showed that these costs/revenues were approximately
evenly distributed across exchange partners. While some
of these changes re?ected concerns about opportunism,
the key concern explicitly given by actors and implicit in
many of the changes implemented was simply to eliminate
the mundane transaction costs, primarily accounting
record-keeping, associated with operating the interna-
tional mail system.
The case we describe here is of great theoretical interest
in elaborating the interpretation of transaction cost
952 A.J. Richardson, E. Kilfoyle / Accounting, Organizations and Society 34 (2009) 939–956
economics in accounting. There are four key insights: (1)
management accounting information does not arise simply
to replace market price information and need not be pres-
ent in all non-market governance structures; (2) the role of
accounting in governance is a question of relative empha-
sis, which requires an explicit comparative framework in
order to appreciate the change in the use of accounting
as a transaction migrates from one governance structure
to another; (3) the choice of governance structure is re-
lated to transaction costs but may, in some cases, be driven
by mundane transaction costs rather than concern for
opportunism; and, (4) the creation of network governance
structures may be an explicit attempt to reduce the need
for accounting between organizations re?ecting a net-
work’s status as a relational contract.
Most studies in accounting using a TCE framework sug-
gest that management accounting information is devel-
oped when market transactions are internalized in ?rms
and networks (Johnson, 1983; Kaplan, 1984). Our study
provides a setting in which transactions are internalized
in order to reduce the accounting that was necessary be-
tween parties previously engaged in market transactions.
This apparent contradiction requires further theoretical
development. In particular, we return to the origins of con-
cern with transactions and transaction costs in the work of
Commons (1934) and Coase (1937) to elaborate the nature
of transaction costs (mundane and opportunistic). This
broader conceptualization by Langlois (2006) and Baldwin
(2008) of transaction cost economics provides insights into
the empirical ?ndings above.
The implicit assumption in accounting studies using TCE
has been that transactions completed in the markets do not
require accounting information. All transactions are com-
pleted on the basis of market prices: no mundane transac-
tion costs are incurred and opportunistic transaction costs
are insigni?cant. This view ignores the fact that markets,
with the exception of the simplest of spot market ex-
changes, are based on contracts between parties. These
contracts incur mundane transaction costs to de?ne and
measure the goods or services exchanged and compensate
parties according to the terms of the contract. Accounting
information is frequently called on to perform these func-
tions. Furthermore, markets are rarely complete hence
complex production processes operating through the mar-
ket will require signi?cant costs to negotiate contracts for
non-standard goods. Since accounting information is part
of market contracting, the theoretical claim that it arises
when transactions are internalized within ?rms is not sup-
ported empirically. Our work, consistent with Coase’s
(1937) observation on the relative use of contracts between
markets and ?rms, supports the view that the relative
emphasis on accounting measures will vary between
markets and other governance structures but this is an
empirical matter depending on, among other things, trans-
action cost differences. This view allows for the possibility
that the amount of accounting may decrease when transac-
tions migrate from the market into other forms of
governance.
Our work is also consistent with Williamson’s (1975, p.
20) emphasis on ‘‘remedial frictions” in the choice of gov-
erning institution. Studies that examine the role of
accounting in a particular governance system without
considering the actual or counterfactual, i.e. a theoretical
baseline, alternative to this system will be unable to ex-
plain the way in which accounting is used (or not used).
The design of the governance mechanism is theoretically
related to the alternative foregone as a means of remediat-
ing transaction costs. Williamson (1991) has emphasized
that TCE empirical work must be comparative in order to
identify the cost savings that a change in governance
mechanism is intended to generate. The lack of accounting
in the UPU international postal network can only be ex-
plained when the costs of accounting in previous modes
of governance are considered.
The predominant view of the role of accounting in gov-
ernance has been that accounting is part of the monitoring
and bonding processes that reduce opportunism. This
builds on Williamson’s view of opportunism as the key
source of costs and bene?ts of different governance mech-
anisms. Williamson’s use of the concept of transaction
costs stands at odds with Coase’s concern with the mun-
dane transaction costs of using a governance mechanism.
The key difference is whether the costs are incurred in
de?ning, creating and exchanging property rights versus
incurring costs to protect property rights in the face of
opportunism (Langlois, 2006, p. 1390; Baldwin, 2008).
The setting that we explore is one in which mundane
transaction costs were far more signi?cant than opportu-
nistic transaction costs. The changes in the governance
structure used for international mail – from market ex-
changes, to bilateral treaties and to the creation of a net-
work of national postal providers – appear to have been
undertaken primarily to reduce mundane transaction
costs. The relative cost of opportunism among state actors
in this setting was small compared with the costs of track-
ing mail and associating transit and delivery costs with
individual pieces of mail.
Our ?ndings regarding the changes in the use of
accounting within the international postal system are con-
sistent with several other observations of the use of
accounting within networks. Several authors have been
surprised by the lack of accounting systems at the network
level of governance. Hakansson and Lind (2004), for exam-
ple, found that no new accounting systems were designed
at the network level:
[. . .] neither collective accounting information, e.g.
inter-organizational budgets and cost behavior analy-
ses, or transparency through open book accounting
and target costing approach is used in the relationship.
Thus, the decision makers are not supplied with more
formalized accounting information about the relation-
ship or the counterpart. Instead it is through a system-
atic combining of accounting in the formation of
organizational units with partly overlapping account-
ability that established accounting methods support
the relationship formation (p. 67).
Mouritsen and Thrane (2006), drawing on ANT, concep-
tualize accounting as being part of the network rather than
a technology or a governance structure of the network.
They state that accounting works through ‘‘self-regulating”
or ‘‘orchestrating” mechanisms that stabilize and develop
A.J. Richardson, E. Kilfoyle / Accounting, Organizations and Society 34 (2009) 939–956 953
the network, ‘‘a fragile accomplishment because its form,
in principle, has no strategic apex and hierarchy” (p.
242). Orchestrating mechanisms, such as strategy and rela-
tionship building, ‘‘help develop the network as an entity
with a common objective” (p. 268). Self-regulating mecha-
nisms, on the other hand, are transfer prices, taxes and fees
that ‘‘‘facilitate ‘frictionless’ network interaction” (p. 268).
Self-regulating mechanisms are oriented towards the
exploitation of existing complementary relations and
typically allow unobtrusive ?nancial ?ows between
partners. Through pre-set transfer prices and fees,
self-regulating mechanisms help partners to rally
around projects and customers without having to
debate the distribution of proceeds from these engage-
ments. This is important because much of the ideology
of network participation is sharing and equality, and
this makes discussions about pricing and ?nancial rela-
tions dif?cult as they are seen to concern the brutality
of the market. Questions about the distribution of ?nan-
cial ?ows and concerns about ownership can disrupt
caring and supporting relations. Self-regulation mecha-
nisms allow the interactions and discussions between
partners to be primarily about matters concerning the
offering to an external customer and thus more about
knowledge and competencies than about ?nancial rela-
tions (Mouritsen & Thrane, 2006, p. 273).
Accounting technologies play a facilitator role by mini-
mizing the ‘‘friction” (Mouritsen & Thrane, 2006, p. 267), or
in TCE terms by minimizing transaction costs. They ?nd
that accounting calculations were problematic but not a fo-
cus of effort:
Calculation was a problem. No central entity was in
place to survey the calculation of costs and there was
ample room to debate the accuracy of actual costs. This
was acknowledged, as the quote indicates, but it also
indicates that this problem, even if real, was to be dis-
counted and made to go away by insisting that calcula-
tion was there not primarily to develop accurate costs,
but to make sure that the systems of transfer-prices
could go on. The quote indicates that the frailty of cal-
culation should not be a central discussion, but it was
more interesting to look into what the calculation could
accomplish in terms of new business to the network.
The frailties and ambiguities of accounting were dis-
missed—the ambition was to make them irrelevant—by
focusing on the possible desirable effects of an accounting
system that could increase interaction and thus increase
the total amount of bene?cial ?nancial effects of network
cooperation. The system of fees and transfers provided
a self-regulating network system because it could min-
imize discussions on the distribution of money in day-
to-day activities and make the ?rms concentrate on
their possible interaction. The transfer pricing system
was fragile because each partner had to accept that
the calculation of cost was a personal affair, but as long
as the cost was within reason its size could be accepted (p.
255–256, emphasis added).
The dif?culty in calculating the costs was a concern but
there was a strong attempt to not make a detailed cost
calculation a big issue because it was more important
that the system of costs and transfer prices helped
interaction to increase than to make transfer prices
authentic. They could be appropriate and acceptable
without a strong accounting regime. The effectiveness
of transfer-prices could be problematised, but this issue
was skirted and left behind in most day to day situa-
tions (p. 257).
It is implicit in these quotes that different types of
transaction costs come into play when imperfect calcula-
tions function as self-regulating mechanisms within net-
works. For example, there is undoubtedly a trade off
between the costs of not having these calculations versus
the cost of having calculations with varying degrees of
imperfections. Mouritsen and Thrane’s study falls short,
however, of providing insights into how the relative mag-
nitude of these costs affects the use of accounting as a self
regulating mechanism: a question that this study of the
international postal network has addressed.
In this study, a network was created because (a) there
were structural impediments to the integration of the
international postal system within a single hierarchical
governance system and (b) the use of market mechanisms
to complete international postal transactions was too
costly. The network emerged as a relational contract be-
tween organizations and consisted of a set of rules that
eliminated the need for detailed calculations in the trans-
fer of mail between countries and administrative proce-
dures for the resolution of any issues that might emerge
from the implementation of these rules. The immediate ef-
fects of the introduction of the UPU were to reduce the
average cost of international postal service, to improve
the speed of service and to encourage greater volumes of
mail to be exchanged. These features had ancillary bene?ts
to the nations involved in terms of economic development
and resource allocation, the integration of political net-
works (particularly colonies) and the encouragement of
Diaspora. Although the rules would ultimately set the
stage for new forms of opportunism as the technologies
and actors involved in international communication chan-
ged, during the period we examine, the effect on interna-
tional communications was dramatic.
Conclusion
We have examined the role of accounting in transna-
tional governance focusing on the restructuring of the
international postal system between 1840 and 1875. Dur-
ing this period international mail transactions were ini-
tially completed as market transactions but this quickly
gave way to bilateral contracts to reduce the frequency
and cost of contracting. An examination of a time series
of bilateral treaties between countries demonstrates a
growing recognition of the deadweight losses associated
with accounting procedures to track mail ?ows and to di-
vide revenue among stages of the value chain according
to an approximation of the actual cost of delivery of each
letter. The attempts to simplify accounting in bilateral
treaties culminated in the creation of the UPU Convention
which created a network of national postal organizations
954 A.J. Richardson, E. Kilfoyle / Accounting, Organizations and Society 34 (2009) 939–956
and eliminated accounting between national posts. The
UPU Convention, and associated structures for administra-
tion of postal agreements, constitutes a relational contract
in which assumptions about mutuality of purpose, equality
of mail ?ows and the integrity of sovereign states, and the
standardization of key attributes of the mail, substituted
for detailed accounting for costs and revenues.
Our case analysis supports the conclusion that the
changes in the governance structure for international mails
were largely an attempt to reduce mundane transaction
costs where accounting for the cost and revenue of mail
?ows was the major component of such costs. Contrary
to much of the extant research on accounting from a trans-
action cost perspective, accounting information was not
used to replace market prices to coordinate activities and
accounting was not used primarily as a means to reduce
opportunism between parties. Our work, although differ-
ing in its theoretical framing, con?rms several studies of
accounting in networks that ?nds that accounting may
be reduced or absent in network governance structures.
The case illustrates several key points which re?ne our
interpretation of the role of accounting within TCE. First,
accounting was associated with the use of contracts in the
early international postal market. It was not the case that
accounting emerged simply to replace market prices as
transactions migrated fromthe market to other governance
structures. Our work con?rms the importance of adopting
the comparative institutional approach advocated by Wil-
liamson (1991) and establishing the baseline use of
accounting in one discrete governance system before
assessing the use in a second governance system. The ‘‘fail-
ure” of any governance structure, i.e. the migration of trans-
actions to another governance structure, will only occur if
there are ‘‘remedial” transaction costs, i.e. transaction costs
that can be avoided or reduced in an alternative governance
structure. The empirical issue is the relative use and mode
of use of accounting within different governance structures
but this can only be assessed when compared with an ac-
tual or counterfactual, i.e. theoretical, alternative.
Second, networks are often conceptualized as hybrids
between markets and hierarchies. Our study suggests that
networks are a distinct form of governance structure in
which organizations are embedded, speci?cally in our case
national postal operators continue to function as indepen-
dent hierarchies embedded in a network that governs their
interaction with each other. The concept of organizations
being embedded in a network allows for the possibility
that the use of accounting at the network level is indepen-
dent of the use within members of the network. In our
case, the reduction in the use of accounting within the
international postal network did not reduce the need for
accounting within the national postal operators that com-
prised the network. The barter arrangement implicit in the
UPU’s lack of accounting for mail ?ows between national
postal operators created incentives for each postal opera-
tor to improve the ef?ciency of their internal delivery sys-
tem. The adoption of a ?xed postage rate for international
postage meant that national postal operators could achieve
a surplus or pro?t by reducing the cost of providing the
service within their own borders. However, the barter
system did not provide incentives for trading partners to
achieve equivalent cost savings and, in the long run, the
difference in the cost ef?ciency of different national postal
systems created incentives for ‘‘remail” and the emergence
of private mail carriers that undermined the assumptions
on which the UPU network was premised.
Third, contrarytomost interpretations of TCEinaccount-
ing that management accounting arises to replace market
prices when transactions migrate from the market to other
governance mechanisms, we ?nd that governance mecha-
nisms such as bilateral contracts and networks can be asso-
ciated with a reduction and even elimination of accounting.
We suggest that prior interpretations the role of accounting
from a TCE perspective have focused on the opportunistic
transaction costs envisioned by Williamson as the key
source of costs and bene?ts of governance mechanisms.
An alternative perspective on transaction costs based on
Coase (1937) and developed by Baldwin (2008) differenti-
ates between mundane and opportunistic transaction costs.
Mundane transaction costs are the costs of de?ning, mea-
suring and compensating for transactions and are incurred
even in the absence of opportunism. We suggest that the
change in the use of accounting in the evolving governance
structures affecting the international mail system was dri-
ven primarily by mundane transaction costs. We have iden-
ti?ed studies that have observed an absence or a decrease in
use of accounting (c.f. Hakansson &Lind, 2004; Mouritsen &
Thrane, 2006) but, contrary to our work, they have not ad-
dressed the mechanisms underlying these observations.
We posit that accounting technologies will be employed
within organizational structures only when the mundane
transaction costs incurred by their use are less than the
opportunistic transaction costs avoided by their use and if
the total cost of opportunistic and mundane transaction
costs is less than the bene?ts gained from the transaction.
Acknowledgements
This work has been supported by a grant from the Social
Sciences and Humanities Research Council of Canada. The
authors thank the staff of the Royal Mail Archives, London,
and the UPU Archives, Berne, for providing access to mate-
rials. An earlier version of the paper was presented at the
World Congress of Accounting Historians and at HEC-Paris.
The authors acknowledge the helpful comments of partic-
ipants and particularly Marcia Annisette, Rachel Basker-
ville, Gary Carnegie, Marta Macias, Laurent Tanguy and
the reviewers.
References
Anderson, S. W., & Dekker, H. C. (2005). Management control for market
transactions: The relation between transaction characteristics,
incomplete contract design and subsequent performance.
Management Science, 51(12), 1734–1752.
Arndt, J. (1979). Toward a concept of domesticated markets. Journal of
Marketing, 43(4), 69–75.
Baldwin, C. Y. (2008). Where do transactions come from? Modularity,
transactions, and the boundaries of ?rms. Industrial and Corporate
Change, 17(1), 155–195.
Baldwin, C. Y., & Clark, K. B. (2003). Where do transactions come from? A
perspective from engineering design. Working paper. Harvard
Business School. Accessed 23.03.08.
A.J. Richardson, E. Kilfoyle / Accounting, Organizations and Society 34 (2009) 939–956 955
Ball, R. (1989). The ?rm as a specialist contracting intermediary: Appli-
cation to accounting and auditing. Working paper. University of
Chicago.
Accessed 20.03.08.
Basu, S., Kirk, M., & Waymire, G. (2008). Accounting institutions emerge
very early in an economy’s development. Working paper. Emory
University. Accessed 03.09.
Braudel, F., & Matthews, S. (1982). On history. Chicago: The University of
Chicago Press.
Campbell, D. T. (1975). Degrees of freedom and the case study.
Comparative Political Studies Issue, 8(2), 178–198.
Cheung, S. N. S. (1973). The fable of the bees: An economic investigation.
Journal of Law and Economics, 16(1), 11–33.
Chua, W. F., & Habib, M. (2007). The effect of network ties on accounting
controls in a supply alliance: Field study evidence. Contemporary
Accounting Research, 24(1), 47–86.
Coase, R. H. (1937). The nature of the ?rm. Economica, New Series, 4(16),
386–405.
Coase, R. H. (1974). The lighthouse in economics. Journal of Law and
Economics, 17(2), 357–376.
Coase, R. H. (1990). Accounting and the theory of the ?rm. Journal of
Accounting and Economics, 12(1–3), 3–13.
Coase, R. H. (2000). The acquisition of ?sher body. Journal of Law and
Economics, 43(1), 15–31.
Commons, J. R. (1934). Institutional economics: Its place in political
economy. New York: Macmillan Company.
Cooper, R., & Slagmulder, R. (2004). Interorganizational cost management
and relational context. Accounting Organizations and Society, 29(1),
1–26.
Courtis, J. K. (2004). Red and black interpostal accountancy marks.
Accounting Forum, 28(4), 385–402.
Dekker, H. C. (2004). Control of interorganizational relationships:
Evidence on appropriation concerns. Accounting Organizations and
Society, 29(1), 27–49.
Eisenhardt, K. M. (1989). Building theories from case study research. The
Academy of Management Review, 14(4), 532–550.
Emanuel, D., Wong, J., & Wong, N. (2003). Ef?cient contracting. Accounting
and Finance, 43(2), 149–166.
Gough, J. P., (2005). The evolution of the postal service in the era of the
UPU. Accessed 04.08.
Granovetter, M. (1985). Economic action and social structure. The
American Journal of Sociology, 91(3), 481–510.
Granovetter, M. (2005). The impact of social structure on economic
outcomes. Journal of Economic Perspectives, 19(1), 33–50.
Hakansson, H., & Lind, J. (2004). Accounting and network coordination.
Accounting Organizations and Society, 29(1), 51–72.
Hargest, G. E. (1971). History of letter post communication between the
United States and Europe, 1845–1875. Smithsonian studies in history and
technology. Washington: Smithsonian Institution Press.
Hill, R. (1837). Post of?ce reform: Its importance and practicability. London:
Charles Knight and Co.
Hopper, T., & Armstrong, P. (1991). Cost accounting, controlling labour,
and the rise of the conglomerates. Accounting Organizations and
Society, 16(4), 405–438.
Hopwood, A. G. (1987). The archaeology of accounting systems.
Accounting, Organizations and Society, 12(3), 207–234.
Hopwood, A. G. (1996). Looking across rather than up and down: On the
need to explore the lateral processing of information. Accounting,
Organizations and Society, 21(6), 589–590.
Johnson, H. T. (1975). Management accounting in an early integrated
industrial: E. I. duPont de Nemours Powder Company, 1903–1912.
The Business History Review, 49(2), 184–204.
Johnson, H. T. (1983). The search for gains in markets and ?rms: A review
of the historical emergence of management accounting systems.
Accounting, Organizations and Society, 8, 139–146.
Jones, T. C., & Dugdale, D. (2001). The concept of an accounting regime.
Critical Perspectives on Accounting, 12(1), 35–63.
Kaplan, R. S. (1984). The evolution of management accounting. The
Accounting Review, 59(3), 390–418.
Keohane, R. O. (1982). The demand for regimes. International Organization,
36(2), 325–355.
Keohane, R. O. (1984). After hegemony: Cooperation and discord in the world
political economy. Princeton University Press.
Klaes, M. (2000). The history of the concept of transaction costs: Neglected
aspects. Journal of the History of Economic Thought, 22(2), 191–216.
Klein, P. G. (1996). Economic calculation and the limits to organization.
The Review of Austrian Economics, 9(2), 3–28.
Langlois, R. N. (2006). The secret life of mundane transaction costs.
Organization Studies, 27(9), 1389–1410.
MacDonald, L. D., & Richardson, A. J. (2002). Alternative perspectives on
the development of American management accounting: Relevance
lost induces a renaissance. Journal of Accounting Literature, V21,
120–156.
Meyer, J. W., & Rowan, B. (1977). Institutional organizations: Formal
structure as myth and ceremony. American Journal of Sociology, V83,
340–363.
Miller, P., Kurunmaki, L., & O’Leary, T. (2007). Accounting, hybrids and the
management of risk. Accounting, Organizations and Society, 33(7–8),
942–967.
Mirus, R., & Yeung, B. (1993). Why countertrade? An economic
perspective. International Trade Journal, 7(4), 409–433.
Mouritsen, J., & Thrane, S. (2006). Accounting, network complementarities
and the development of inter-organisational relations. Accounting,
Organizations and Society, 31, 241–275.
North, D. (1990). Institutions, institutional change, and economic
performance. New York: Cambridge University Press.
Ouchi, W. G. (1980). Markets, Bureaucracies, and Clans. Administrative
Science Quarterly, 25(1), 129–141.
Pessali, H. F., & Fernandez, R. G. (1999). Institutional economics at
the micro level? What transaction costs theory could learn from
the original institutionalism. Journal of Economic Issues, 33(2),
265–275.
Polodny, J. M., & Page, K. L. (1998). Network forms of organization. Annual
Review of Sociology, 24, 57–76.
Powell, W. W., (1990). Neither market nor hierarchy: Network forms of
organization. In B.M. Staw, & L.L. Cummings, (Eds.), Research in
organizational behavior (Vol. 12, pp. 295–336).
Presthus, R. (1962). The organizational society: An analysis and a theory.
New York: Knopf.
Richardson, A. J. (forthcoming). Organizational founding, strategic
renewal and the role of accounting: Management accounting
concepts in the formation of the Penny Post. Journal of Management
Accounting Research, 20.
Rosen, S. (1988). Transactions costs and internal labor markets. Journal of
Law, Economics & Organization, 4(1), 49–64.
Santos, F. M., & Eisenhardt, K. M. (2005). Organizational boundaries and
theories of organization. Organization Science, 16(5), 491–508.
Shelanski, H. A., & Klein, P. G. (1995). Empirical research in transaction
cost economics: A review and assessment. Journal of Law, Economics &
Organization, 11(2), 335–361.
Spekle, R. F. (2001). Explaining management control structure variety: A
transaction cost economics perspective. Accounting, Organizations and
Society, 26, 419–441.
Spraakman, G., & Davidson, R. (1998). Transaction cost economics as a
predictor for management accounting practices at the Hudson’s Bay
Company, 1860 to 1914. Accounting History, 3(4), 69–101.
Spraakman, G., & Wilkie, A. (2000). The development of management
accounting at the Hudsonís Bay Company, 1670–1820. Accounting
History, 5(1), 59–84.
Tiessen, P., & Waterhouse, J. H. (1983). Towards a descriptive theory of
management accounting. Accounting, Organizations and Society, 8(2–
3), 251–267.
Watts, R. L., & Zimmerman, J. L. (1986). Positive accounting theory.
Prentice-Hall Inc.
Whiteside, J. H. (2002). The Prussian–Franco postal treaty of 1817. Danzig
Report 117(1), 3–13.
Williamson, F. H. (1930). The international postal service and the
Universal Postal Union. Journal of the Royal Institute of International
Affairs, 9(1), 68–78.
Williamson, O. E. (1975). Markets and hierarchies: Analysis and antitrust
implications. New York: The Free Press.
Williamson, O. E. (1979). Transaction cost economics: The governance of
contractual relations. Journal of Law and Economics, 22(2), 233–261.
Williamson, O. E. (1985). The economic institutions of capitalism. New York:
Free Press.
Williamson, O. E. (1991). Comparative economic organization: The
analysis of discrete structural alternatives. Administrative Science
Quarterly, V36, 269–296.
Williamson, O. E. (1996). The mechanisms of governance. Oxford, UK and
New York: Oxford University Press.
Williamson, O. E. (1999). Public and private bureaucracies: A transaction
cost economics perspective. The Journal of Law, Economics, &
Organization, 15(1).
Williamson, O. E. (2005). Business history and the economics of
organization: The governance perspective. In International society for
new institutional economics conference. Barcelona.
956 A.J. Richardson, E. Kilfoyle / Accounting, Organizations and Society 34 (2009) 939–956

doc_604883642.pdf