Source:  Review of Social Economy, Dec 2000 v58 i4 p455.

    Title:  What is a Market? On the Methodology of a Contested Concept.
   Author:  Eckehard F. Rosenbaum

 Subjects:  Markets (Economics) - Analysis
            Austrian school of economics - Beliefs, opinions and attitudes

  Business Collection:  129Y0034
Electronic Collection:  A69390309
                   RN:  A69390309


Full Text COPYRIGHT 2000 Routledge

Abstract Some economists find markets everywhere and assume that they emerge
spontaneously once a set of necessary conditions such as well-defined property
rights is fulfilled. Others emphasise the role of organisations and contend
that markets are actually less dominant. But despite claims to the contrary,
the market concept is hardly analyzed in depth. Nor are there serious attempts
to examine empirically where markets exist. Against this background, the paper
addresses the question of what is a market? It begins by illustrating how the
literature has hitherto defined the concept of a market. On the basis of
methodological considerations, which center on the subject matter of economic
analysis, the paper provides a revised conceptualisation of markets in terms
of those conditions under which stylised facts about relative prices can be
observed. The final part of the paper discusses the link between the
conceptualisation of a market and the informational role of relative prices
highlighted by the Austrian Sc hool.

Keywords: market, institution, methodology, Austrian approach

INTRODUCTION

Many economists tend to find markets almost everywhere on Earth and in
history. Indeed, some believe that even the formation of intimate social
relations such as marriages is governed by markets (cf. Becker 1976). Moreover
markets are frequently assumed to emerge spontaneously once a set of necessary
conditions, including, amongst other things, well-defined property rights and
liberalized prices, is fulfilled (e.g. Aslund 1995; EBRD 1996). The essence of
this view is beautifully reflected by Oliver Williamson's dictum that 'in the
beginning, there were markets' (Williamson 1983: 20).

Evidently, the above picture differs sharply from that drawn by Herbert Simon.
He has posed the question of how a Martian would describe the institutional
structure of economic life on Earth, were it to observe our planet from its
spaceship (Simon 1991). Distinguishing between organizations (green areas),
market relationships (red lines) and contractual relationships (blue lines),
this Martian would most likely report back to its home base, or so Simon
muses, that on the economic map of the Earth, green areas of various sizes
dominate (except for rural areas in India, Africa or China perhaps), in turn
to a varying degree connected by either red and blue lines on the Western
hemisphere and large parts of Asia, and--prior to 1989--by mostly blue lines
on the North Eastern hemisphere. On Simon's account, then, markets do not
appear to be as dominant as some economists would have it, nor are markets and
organizations the only governance mechanisms that can be found within national
economies (cf. Hollingworth and Lindberg 1985).

In a sense, however, these contrasting views are merely the most visible
symptoms of a much deeper problem. No less an authority than Ronald Coase has
observed that the role of markets in modern economic theory is even more
opaque than that of enterprises (Coase 1988). Thus, even though its frequent
use in economic discourse seems to suggest that there must be some agreement
on the meaning and content of the concept of a market, closer inspection
immediately reveals a wide range of meanings and contents. This diversity is
in turn rooted in the fact that the market concept itself is hardly if ever
analyzed in a systematic fashion with a view to identifying the constituting
or essential elements of a market (cf. North 1977). Indeed, the term 'market'
has nowadays-strong metaphorical connotations in that many uses do not make
reference to the etymological origins of the term. "Market" denotes both
demand and supply, buyers and sellers, competition and exchange, but none of
these uses refers directly to markets in the way common parlance sometimes
still does, namely to markets as the socioeconomic phenomenon which takes
place in the marketplace of a city or town. Given the paramount role ascribed
to markets in economic analysis, what are we to make of this diversity? To be
sure, McCloskey and others have repeatedly pointed to the rhetorical nature of
economic discourse, rightly arguing that economics is a conversation about
economic arguments (McCloskey 1986). And in such a conversation, metaphors are
as legitimate (and necessary) as in any other conversation, their specific
meaning being shaped by the respective context and the background knowledge of
the speaker. But, evidently, a metaphor presupposes a non-metaphorical
meaning. While there may be situations when we speak of the market but have
something else in mind for which we lack a proper expression, in some cases at
least, the notion of a market is not used rhetorically and thus has to be
given some substance.

An important example in this respect is the already mentioned literature on
the transformation of post-socialist countries (e.g. World-Bank 1996). This
literature epitomizes transformation as the replacement of plans by markets.
Therefore, it is the market itself that is of interest now. Moreover, in order
to make the transition from plan to market empirically verifiable and thus in
order to assess the progress of transformation, one has to give criteria for
the existence (or absence) of a market. Yet such criteria are virtually
absent, the implicit assumption being that markets emerge spontaneously
whenever appropriate preconditions are present, or, what is even more
questionable (cf. Kornai 1994), that markets came into being as soon as
central planning had been abolished. As a consequence, economists appear to
find themselves in a situation where the victory of the market marks "the end
of history" (Fukuyama 1992), but where it seems difficult to be really
confident that (or to what extent) markets do act ually exist. Moreover,
lacking as we do a theory explaining the emergence of markets which goes
significantly beyond product cycle reasoning (for an account along such lines
see Brezinski and Fritsch 1997), we do not know where to look for newly
emerging markets either, or why some markets may fail to materialise.

The need for a more detailed specification of the market concept is thus
particularly pressing in a normative context. Economists or politicians who
endorse markets must specify where and when a market does in fact exist and
where and when it is absent. Unless they are able to do so, their policy
recommendations could neither be evaluated in relation to the purported
objectives of market creation nor tested with respect to the empirical
implementation of a market. In fact, much the same goes for economic theory.
Hypotheses about the functions and the properties of markets (e.g.
establishment of prices, allocation of resources, efficiency) which claim to
have empirical validity presuppose that the researcher outlines the
characteristics of the social object "market" for which the hypothesised
relationship or property is to hold. Arrow and Hahn's (1971) General
Competitive Analysis, for example, provides no explicit definition of what a
market is and only a highly abstract account of how markets function. Inde ed,
it is not until page 348 that they acknowledge that they have taken the
"existence of markets... for granted." Furthermore, lacking a substantive
account of markets, economic research would be in clear danger of ignoring the
widespread existence of rival coordination mechanisms between economic
entities, thereby failing to address the question of why markets are replaced
or circumvented by other governance mechanisms such as industrial networks or
cooperatives. [1]

Against this background, the purpose of the present paper is to discuss a
number of difficulties that have to be tackled when attempts are made to
define the market. On this basis, I shall develop a definition that makes it
possible to identify markets empirically against the background of rival
social forms such as firms, central planning or occasional exchange
transactions. Not surprisingly perhaps, the endeavor reveals, or so I shall
claim, that markets cannot be said to occur as often as some economists would
have it if the notion is to retain any discriminatory force.

The paper is organized as follows. In section I, I shall briefly illustrate
how economists and other social scientists have attempted to define the
concept of a market and identify several key issues. On the basis of a number
of methodological considerations discussed in section II, section III then
takes up the task of providing a revised conceptualization of markets and
explores its operationalisation against the background of rival social forms.
Section IV discusses the suggested definition of a market against the
background of the institutionalist perspective on markets and the Hayekian
analysis of markets as information processing devices. Section V concludes and
summarizes.

I. THE RECEIVED VIEW: SETTING THE STAGE

The diversity and ambiguity of the notion of a market in common parlance
corresponds with few and rather diverse attempts to define the notion in more
detail. In order to illustrate this diversity, I shall proceed by providing a
brief account of the received view, thereby highlighting a number of key
issues, which the following methodological discussion has to address.

For a start, it seems helpful to distinguish between roughly three categories
of definitions. The first category, observational definitions, comprises
definitions, which refer to some empirical phenomenon, often together with one
or several stylized facts about prices and/or commodities. Some (Neo-)
classical economists can be included here, for they saw the market as
synonymous with either a marketplace or a geographical area in which exchanges
of the same commodity take place (Swedberg 1994). Thus, according to Cournot,
a market is "a region in which buyers and sellers are in such frequent
intercourse with each other that the prices of the same goods tend to equality
easily and quickly" (Cournot 1897, quoted in Hodgson 1988: 173). But even more
recently, it is possible to find definitions, which focus on markets as
locations. Lipsey for instance has defined a market as 'an area over which
buyers and sellers negotiate the exchange of a well-defined commodity' (Lipsey
1983: 69). His definition thus echoes the traditional notion of a marketplace,
even though the area in question is now (apparently) much larger than a
medieval square or the site of a trade fair. Moreover, his definition also
highlights the notion of a well-defined rather than unspecified commodity. And
this too can be regarded as an empirical fact to be singled out.

The reference to a specific locale is lost in definitions which see the market
as synonymous with (the willingness to) exchange, as in the case of Gravelle
and Rees. For these authors, "a market exists whenever two or more individuals
are prepared to enter into an exchange transaction, regardless of time or
place" (Gravelle and Rees 1992: 3). And Jevons, more than a century earlier,
takes the market "to mean any body of persons who are in intimate business
relations and carry on extensive transactions in any commodity" (Jevons 1871,
quoted in Hodgson 1988: 173). For all apparent differences, an observable
phenomenon, namely exchange, remains at the heart of these definitions. The
same applies to Marshall's thinking about markets, even though it has changed
over time (Swedberg 1994). In his earlier definition, that which can be found
in his Principles of Economics, Marshall conceives of the market in terms of
demand and supply and takes converging local prices (again an empirical
criterion) to be the hallmark of a unified market (Marshall 1938). In his
later Industry and Trade, by contrast, Marshall focuses more on the social
organization of a market when he writes that "in all its various
significations, a 'market' refers to a group or groups of people, some of whom
desire to obtain certain things, and some of whom are in a position to supply
what the others want" (Marshall 1919: 182; cf. Swedberg 1994). Marshall's
latter definition thus comes closer to the category of structural definitions,
which I shall discuss in a moment.

In contrast to observational definitions, functional definitions focus on what
the market does rather on what, from an empirical point of view, the market
is. Evidently, therefore, policy proposals which endorse the market as a
superior allocation mechanism do so on the basis of a functional rather than
an observational definition of the market. Within the orthodox tradition, at
least two views can be distinguished. According to the first, the market is
essentially regarded as an allocation mechanism (or more metaphorically as an
'invisible hand' as in Adam Smith's work) while according to the second,
closely related view, the concept of a market is equated with the
determination of relative prices by demand and supply (see Barnett 1991). Both
definitions reflect the analytical role ascribed to the market in orthodox
economic theory. Their underlying understanding culminates in general
equilibrium theory as pioneered by Walras, where the market is synonymous with
(the intersection of) unobservable and merely hypothesized demand and supply
(curves) and devoid of any institutional, spatial or social features, and
where each individual market is part of a set of interrelated markets that
make up the economy as a whole and that is seen as jointly determining a
vector of relative prices and an associate allocation of commodities (Debreu
1959). From this perspective, the pricing function and the allocative function
of markets are two sides of the same coin. A third view, which appears
somewhat more fundamental, ascribes to markets the function of providing a
natural order or an equilibrium of social activities, thereby having some
resonance with Hayek's notion of spontaneous order.

Heterodox approaches generally differ from neoclassical theory with respect to
the analytical role of demand and supply factors for the explanation of prices
and quantities. The former see prices and quantities, outside auction markets
at least, as being determined by different although possibly interdependent
factors, for instance some variety of mark-up or "administrative" pricing and
effective demand (cf. Means 1938; Okun 1981; Lavoie 1992). As a consequence,
the market qua demand and supply is amended or replaced by institutions,
production technology (Sraffa 1960) and/or income effects, its analytical role
being confined to an arena where these factors shine through as it were in the
form of specific prices and quantities. The importance of the market then lies
in the fact that without real exchanges, we would have price tags rather than
prices and/or quantities asked for or offered rather than quantities traded.

Likewise functional are Austrian definitions of the market, but with a
different function at their core. In particular Hayek emphasizes the ability
of markets to aggregate and disseminate dispersed and particular knowledge and
information in a way, which is superior to, say central planning (Hayek 1945).
I shall take up this issue in section IV.

The last category comprises structural definitions. These definitions draw
attention the underlying and hence not immediately observable structure of a
market, emphasising the alleged mechanisms and structures that give rise to
market phenomena. Hodgson's definition, for instance, belongs to this
category, for he sees the market as "a set of social institutions in which a
large number of commodity exchanges of a specific type regularly take place,
and to some extent are facilitated and structured by those institutions.
Exchange...involves contractual agreements and the exchange of property
rights, and the market consists in part of mechanisms to structure, organize,
and legitimate these activities" (Hodgson 1988: 174, my emphasis; similarly
North 1990, Richter 1996 and Furubotn 1996). In like manner,
conceptualizations of markets in terms of networks of stable exchange
relationships (e.g. Fourie 1991 or Snehota 1993) can be said to belong to the
third category provided these networks perform some analytical role in
understanding market phenomena. Last but not least, recall Marshall's later
definition.

At this point, I should emphasize that hardly any definition fits nicely in
only one category. While the notion of "extensive transactions" in the case of
Jevons would seem to single out his definition for the first category, for
instance, his emphasis on "intimate business relationships" draws attention to
some underlying social structure. The same applies to Hodgson's definition,
which, despite its institutionalist flavor, contains a number of empirically
observable elements, which make the definition come closer to some of the
observational definitions. Likewise, the network approach to markets
highlights the importance of social structures but ties these structures in
turn to observable events. Finally, functional definitions have structural
connotations in the sense that the typical market is characterized by
(usually) a multitude of buyers and/or sellers rather than a monopolist facing
a monopsonist.

Bearing these caveats in mind, the above categorization nevertheless suggests
that definitions are congenial to specific purposes, which can be rationalized
against the background of the research program within which they are
introduced. As for observational definitions, one may suspect for instance
that, in principle, the objective is to single out for further investigation
some empirical phenomenon. In the case of functionalist definitions, by
contrast, a great deal of the analysis has already been carried out. For it is
precisely a result of this analysis that we can now ascribe a specific
function to the market. At the same time, the particular function attributed
to the market provides a guiding line for further research, e.g. as regards
market failure. Structural definitions, finally, belong to an intermediate
position in that, accepting as they do the function of markets in principle,
their concern is to open the black box of the disembedded and
deinstitutionalized market of neoclassical theory, and t o provide a better
understanding of the working of the market by highlighting some central
structural features of markets.

On this view, then, the reported lack of a common definition of the market can
in part be explained by the purpose for which each definition is introduced.
And since each definition belongs to a specific part of economic discourse, it
has first of all to be judged in its own terms. It would therefore be
nonsensical to criticize structural definitions for being empirically
ambiguous, as it would be problematic to look for rich empirical content in
functional definitions.

Still, do at least observational definitions suffice to identify markets
empirically? On the whole, the answer seems to be negative. [2] For instance,
what is the region where, according to Cournot, buyers and sellers meet? And
is "intercourse" supposed to denote more than exchange, e.g. a certain degree
of mutual trust? A similar critique applies to definitions, which focus on
exchange. For exchanges which fit these definitions occur in various contexts
and forms, many of them located beyond the confines of the economic sphere
narrowly conceived. Using these definitions, we would have to conclude that
markets are indeed everywhere and that, consequently, the market concept lacks
discriminatory force. Thus, despite the cautionary remark that precedes the
quotation from Gravelle and Rees, the concept of markets held by (some)
economists is not only much more general than its common understanding (this
in itself would be hard to criticize), at times the concept is so general that
it becomes indistinguishable f rom exchange itself, and hence redundant. [3]

As to the remaining two categories of definitions, there are also significant
problems. Thus, functions can only be ascribed to previously specified objects
because a functional definition presupposes an account of the object to which
the function is ascribed (Searle 1995). Hence functional definitions may be
rightly criticized if such an account is absent or incomplete. Similarly,
structural definitions presuppose an account of the object or phenomenon whose
underlying structure is to be identified. Otherwise the market concept as such
would remain empty. One could hold, of course, that the market is synonymous
either with its function or with a particular structure or set of
institutions. But both defenses do not hold enough water. The first implies
that markets vanish once they cease to function properly while the second
requires that a set of institutions be identified that can be found in each
and every market. As far as I have reviewed the institutionalist literature,
no such identification has been ca rried out so far. Indeed, it seems
difficult to conceive of any specific set of institutions (in Hodgson's sense)
that is common to each and every market and that would suffice to define a
market. [4] The only institution, which appears to be almost universal, is
money. Yet barter trade or the use of commodity rather than fiat money suggest
that markets can exist in the absence of money. The same goes for property
rights safeguarded by the state. While it is clear that the notion of exchange
presupposes a distinction between "mine and thine", property rights do not
have to be enforced by the state. In fact, for centuries international trade,
and prior to the emergence of modem states even intranational trade could not
rely on legal support, but had to be based on mutual trust and reciprocity,
often linked to kinship or religious ties (cf. Greif 1989; Greif 1994).

For these reasons, the approach to be followed in the present paper seeks to
develop a largely formal and a priori conceptualization of markets. The
methodological groundwork for this task will be laid in the following section.

II. ON THE METHODOLOGY OF DEFINING A MARKET

Of course, whatever definition of a market we choose, it is neither true nor
false, only appropriate for the purpose at hand. And as indicated in the
preceding section, a definition of the market may have various purposes. So
defining a market is essentially a normative enterprise that has to be judged
by its usefulness for the chosen purpose rather than its truth or falsity. But
what purpose should a definition of the market fulfil, and by which principles
is its 'usefulness' to be judged? Given the aforementioned economic policy
rationale, one criterion is obvious from the start. Any definition of a market
should make it possible to identify markets among other social forms. [5] This
mode of separation is comparative and contrastive rather than idiosyncratic,
meaning that markets are to be distinguished from other specific forms of
interaction, notably from hierarchies, central planning and perhaps casual
exchange transactions. In other words, the definition of a market must
simultaneously indicate why a fi rm or a hierarchy of planners is not a
market. [6] Moreover, although it is often used interchangeably with the
notion of a market, exchange is a key concept in virtually all definitions of
the market. Hence if these terms are not supposed to denote the same
thing--which, as indicated above, would render the market concept
pointless--then there must be an empirically identifiable rationale for the
use of the term market in addition to its denoting exchange. And whatever this
rationale may be, it should also be reflected in the definition of a market.

In order to find empirical criteria for the existence of a market, it would
seem sufficient, then, to choose one of the definitions of the first category,
suitably specified so as to take the aforementioned criticism on board. Yet
doing so raises the further question of how this observational definition and
its concomitant empirical criteria are linked to the market mechanism and/or
the underlying structure of a market. What reasons do we have to assume, given
what economists claim to know about markets, that certain empirical
observations are an indicator of the workings of some underlying market
mechanism? Indeed, precisely because it is the (economic policy) rationale for
having or introducing markets that markets do perform certain highly valued
functions, the former question cannot be neglected when markets are to be
defined. This leads to a second criterion our market definition has to fulfil:
The definition should be conceptually linked to the analysis of market
phenomena, functional or structural, th ereby indicating why certain empirical
observations are an indicator of the workings of some underlying (market)
mechanism.

When this question is to be answered, basically two issues have to be
addressed. First, are there any reasons to suggest that some particular social
phenomenon with generalized features is going on at all? In other words, are
there reasons to suggest that something is taking place that can be theorized
(in the sense of making general law-like statements about it)? Second, what
justifies the link to the alleged properties and functions of markets?

To address the first issue, note that the subject matter of social scientific
explanation is (usually) not any kind of strict event regularity of the form
"whenever x, then y (under conditions c)", but rather what could be termed a
demiregularity (Lawson 1997) or a stylized fact, i.e. a recurrent conjunction
of types of events. In economics, stylized facts usually include the
allocation of resources to various purposes or the distribution of income and
wealth, to name just a few. The reason for focusing on stylized facts rather
than strict event regularities is that, in contrast to laboratory settings,
the social world is open. Social situations are generally characterized by a
multitude of overlapping, sometimes interacting and possibly counterbalancing
forces. Consequently, a mechanism whose effects may be observed at one point
in time or in some place is overshadowed or offset under different conditions.
What we can reasonably look for in the social realm is therefore a situation
where one (or a very limi ted number of) mechanism(s) shine(s) through as it
were by producing a number of correlated events. But this means in ontological
terms that it is precisely the observation of some regularity or pattern in
the flux of events that invites the conjecture that this regularity or this
pattern did not occur at random but was the result of the workings of some
mechanism or causal relationship which we may try to identify in the course of
research.

Second, what justifies the link to the alleged properties and functions of
markets? As far as market phenomena are concerned, reviewing the subject
matter of much of economic analysis in market contexts reveals that stylized
facts commonly refer to relative prices and quantities of exchanged
commodities including the relationship between quantities and prices, and the
change of quantities and prices over time. It is an obvious (stylised) fact,
for instance, that the prices of many commodities exhibit a rather low degree
of volatility, but some prices are more volatile than others (cf. Carlton
1989). In a similar vein, relative prices are quite stable over time (even
though prices of services tend to increase relative to those of manufactures,
cf. Baumol et al. 1989), and the demand for a commodity often decreases as its
price increases.

Note though that these stylized facts are related to market forms rather than
markets per se. Thus, if markets constitute empirical phenomena in their own
right, then markets cannot be synonymous with whatever particular phenomena
take place in them, e.g. with observed relative prices. Instead, there must be
certain stylized facts, which can be said to characterize markets in addition
to, and in a more fundamental way than, those facts we may observe in specific
markets. But what are these stylized facts? To answer this question, I shall
ask under which conditions the stylized facts that are apparently analyzed as
market phenomena (whatever their content) can be observed in the first place.
Moreover, I shall claim that these conditions are themselves stylized facts
precisely because they refer to the form or type of the phenomena whose
content is to be explained. In short, it is specific types or forms of
exchanges that a amenable to economic analysis as stylized facts. And it is
these types that I take to b e market phenomena. Consequently, by identifying
the nature of these stylized facts, I shall be able to develop general
criteria for identifying markets.

Clearly, such an empirical but methodologically grounded definition of the
market seems to presuppose an essentialist position in the sense that all
markets qua objects of social analysis must have something in common which
firms or hierarchies lack. It goes beyond the scope of the present paper to
discuss the plausibility of essentialism in depth. Suffice to say however that
Wittgenstein's famous criticism of the "craving for generality" (Wittgenstein
1960) does not undermine essentialism. Wittgenstein rightly points out that
there is no reason to assume that all entities that fall under a general term
must have something in common. Yet the only conclusion that can be drawn from
Wittgenstein's argument is that one cannot assume in advance that some concept
is characterized by a set of essential properties, not that there are no
essential properties of objects at all (cf. O'Neill 1998). I should also
hasten to add that the foregoing considerations do no intend to explain the
existence of a market by its func tion(s) for economic analysis. Rather, the
approach adopted here is meant to be transcendental in the sense that it seeks
to explore the preconditions under which a social object called market can be
theorized in the manner currently en vogue in economics.

III. TOWARDS A DEFINITION OF THE MARKET

In the present section I shall discuss which forms of exchange should count as
market exchange, given the above criteria. Subsections 1 and 2 focus on a
number of formal criteria of market exchange. Subsection 3, by contrast, deals
with competition as a necessary expression of the impersonal nature of market
exchange, and hence focuses on the main motivation underlying market exchange
as opposed to other forms of exchange. The discussion will proceed by
contrasting the market with five alternative social forms, namely firms and
organizations, central planning, bargaining, casual exchange, and the exchange
of gifts. For a summary of the discussion see Table 1.

1. Voluntary and Specified Exchange

The Argument

The definitions reviewed in section 1 have one thing in common. With the noted
exception of the Austrian approach (but see Shand 1984, who includes
exchange), they all consider exchange to be the hallmark of a market. In fact,
markets are regarded in some cases as nothing but exchange. Note however that
it is not exchange per se, but voluntary and specified exchange that, for the
methodological reasons outlined above, can qualify as market exchange and that
can hence be analyzed as a stylised fact. To see why it is not exchange as
such, let me explain what I mean by "voluntary and specified exchange". For
exchange to be specified, it is not necessary that the two parts of the
exchange take place simultaneously. Indeed, many exchanges stretch over an
extensive period of time. Rather, specificity means that the mutual agreement
on the exchange includes a substantive specification of both parts of the
transaction. The criterion thus excludes forms of exchange (e.g. the exchange
of gifts) where the specificity of the exchange is not given in substantive
terms because the transaction is based on (at most) the mere expectation of
some compensation or reward that is still to be decided upon. Exchanges within
firms or organizations also often lack specificity, e.g. when the employment
contract does not give a complete description of the duties and
responsibilities of the employee. What is important in the present context is
that, without specificity, we could not know what exactly is exchanged for
what, nor, consequently, could we know and analyze the (implicit) price.
Specificity is therefore an epistemological precondition for the economic
analysis of exchange.

At first sight, the condition of specificity would seem to exclude experience
goods from the province of markets, since exchanges which involve goods whose
characteristics become only apparent in the course of their consumption or use
(e.g. second-hand cars) cannot be fully specified ex ante. The condition of
regularity and typification, which are discussed in more detail below, imply
though that market exchanges are not singular events. Consequently, economic
agents can form expectations, either directly or through vicarious learning,
about the characteristics of such goods, which can then form the basis for
specifying exchange. In the case of used cars, for instance, quality has
proven to decrease with time. Hence, the age of a used car may be taken as an
indicator of its quality. The subject matter of the exchange, then, is not a
used car per se but a used car of specific age. The case of experience goods
nevertheless illustrates that the notion of a price becomes problematic if
agents have only fuzzy ide as about what they are going to exchange.

The last remark indicates that there are forms of exchange where some
components of the exchange are well specified, e.g. working hours and wage in
a labor contract, while others are left open or implicit. Since the form and
timing of compensation is specified at least in part, these exchanges are not
merely gifts. But they also differ from the discrete exchange framework which
dominates neoclassical economics and which is sometimes even assumed in
heterodox approaches. This framework presupposes not only that the exchange is
disembedded from social relations (in the sense that the identity of
transacting partners is irrelevant) but also that it is instantaneous and
always fully specified. Against this, the theory of relational contracts
(Goldberg 1998; MacNeil 1974; MacNeil 1985) emphazises that in an uncertain
and essentially open environment, complete contracts would be impossible while
(long-term) contracts cannot be dissociated from the social relations in which
they are embedded. As in the case of exper ience goods, the notion of a price
becomes problematic, and so does of course any attempt at measurement and
aggregation of quantities, because agents cannot tell in advance what exactly
they are going to exchange and whether two exchange acts concern truly the
same commodity. For the economist, this means that the focus of the analysis
has to shift from the content of the exchange (price and quantity) and those
factors which determine the content (supply, demand, income, technology etc.)
to its form, i.e. to an analysis of those aspects of the exchange which have
been formalized and those which have been knowingly left open or contingent
upon future events (see for example Goldberg 1998).

Having accepted the methodological basis of our market definition, we are thus
led to conclude that extensive relational contracts, even if they fulfil the
remaining criteria to be specified below, cannot easily be included in the
realm of market exchange as a specific object of analysis. This is not to say,
of course, that relational contracts are excluded from market exchange, or
that they are economically insignificant. The point is that their nature
sometimes calls for a different kind of analysis, depending on the degree of
uncertainty involved.

Note that specificity as understood here should not be confounded with the
sociological concept of reciprocity. For sociologists reciprocity is given if
the exchanged commodities have the same value and they sometimes invoke
monopolistic power as an instance where reciprocity is violated (cf. Heinemann
1976). But this view seems mistaken because, provided the exchange is
voluntary (see the next paragraph), the participants to the exchange have only
agreed on the relative value of the commodities to be exchanged, i.e. on the
terms of trade, no matter if, or how much, monopolistic power is involved.
That is, they have agreed on how much of commodity x each participant is
willing to sacrifice in order to obtain y, and thus on the price. Nothing is
thereby implied for the absolute valuation of each commodity, either by the
individual or by society as a whole. For assessing absolute valuation a common
standard of value would be required which is valid throughout society. [7] Nor
should specificity be confounded wi th (the absence of differences between)
subjective preferences over goods. Since it is only intelligible to speak of
varying preferences if the subject matter of these preferences is the same,
specificity is an intersubjective element, which is presupposed whenever
statements about preferences are made.

The second criterion for exchange to qualify as market exchange is its being
voluntary. This condition, while difficult to pin down precisely, is
indispensable in order to rule out forced (in voluntary) forms of exchange
such as central planning, where the terms of the exchanges are ordered from
above. A basic element of voluntarily can be seen in the possibility of
"exit", either by choosing another exchange, if alternatives are available
("partial exit"), or by abandoning the exchange situation altogether, if no
alternatives are available ("total exit"). The latter means that the agent to
whom an exchange offer may help to operationalize voluntarity. For the moment,
suffice to say that the question to what extent exchanges are voluntary
depends as much on the structure of the exchange situation as on the subject
matter of the exchange. Monopolistic power, that is, can, but does not
necessarily need to, preclude voluntarity.

Note, finally, that bargaining (but not simply haggling about the price in the
presence of competitors or more formal negotiations which aim at specifying
the precise terms of an exchange) is often an involuntary form of exchange if
it occurs in situations where no alternative trading partners are available
(or at least conceivable) for both sides so that exit is not an easy option.
The participants in the exchange cannot credibly threaten to withdraw in order
to deal with somebody else. Similar considerations hold for central planning
and other hierarchical allocation mechanisms, of course, since transactions
are ordered from above and no exit option is usually available. In both cases
agents have to resort to "voice." [8] Hence, voluntarity is an important
criterion by which markets can be distinguished from allocation mechanisms,
which resort to power and authority, or where circumstances force participants
to reach an agreement.

Operationalization

The development of criteria for the existence of markets would be of little
use, if it were impossible to link these criteria to observable events. In
principle, at least, this seems to be possible for both properties of
exchanges.

The criterion of voluntarity can be operationalized in terms of the costs of
not entering into a particular exchange transaction. Thus any particular
exchange is involuntary if the possibility of exit is foreclosed because the
costs, including the psychological costs, of not carrying out a particular
exchange, that is, the costs of the status quo or the terms of alternative
exchanges, are prohibitive. Thus petty exchanges with a monopolist/monopsonist
could still qualify as voluntary as do, of course, exchanges where substitutes
or other suppliers/sellers are readily available. For instance,
telecommunications used to be a monopoly in many countries until recently, yet
I would still regard the decision to have (or use) a telephone as voluntary,
be it because it is mostly not a matter of life or death whether I have or use
a phone, be it because alternatives are available (writing a letter) whose
existence reduces the opportunity costs of telecommunication. The requirement
to have one's house connected to the sewer, by contrast, constitutes a form of
involuntary exchange in that opting out is usually precluded by law. The same
seems to be true for many sections of the labor market where job seekers have
no choice but to accept whatever they are offered so that a substantial amount
of power may be conferred to employers. The extent to which an exchange is
voluntary thus depends on the available freedom of choice. Incidentally,
Friedman makes a similar point when he endorses competitive markets on the
grounds that both sellers and buyers can choose between their trading partners
(Friedman 1962).

Blackmail, on the other hand, does not qualify as voluntary exchange. We
cannot choose not to exchange our life for whatever material goods we possess
with the person who blackmails us. Of course, what matters is how the agent
herself perceives the situation, i.e. whether she thinks that exit is possible
rather than whether an omniscient observer knows that exit is possible. The
latter aspect may complicate the issue, but only to the extent that official
statistics do not suffice and have to be replaced or amended by information
(gathered through interviews for instance) about how agents construct their
environment. Perhaps this is also what some sociologists have in mind when
they refer to valuational reciprocity in (some forms of) exchange. Reciprocity
would then mean that the opportunity costs for both parties are relatively
low.

In a similar vein, official statistics may tell us little about the degree to
which an exchange is specified. Yet it is straightforward to examine exchange
contracts with a view to investigating the degree to which they cover all
aspects that are of interest for buyer and seller (again as perceived by those
involved in the exchange).

2. Typification and Regularity

The Argument

By distinguishing between voluntary and involuntary, specified and
non-specified exchange, we can exclude forms of exchange (gifts, blackmail but
also firms and organizations) that cannot be analyzed as "economic exchange"
proper or that are the result of bargaining processes or hierarchical
governance. Still, not all instances of exchange selected by these criteria
allow meaningful analysis. As already noted, typified and regular exchanges
enable economic agents to form expectations about the characteristics of
experience goods. In the present section, I shall argue that typification and
regularity have to be seen as essential elements of markets for other reasons
as well.

To begin with, recall that if exchanges can be characterized by the implied
"prices" and if these prices are to be analyzed, then, in keeping with the
above characterization of the subject matter of economic explanation, there
must be some property of these prices--a (set of) stylized fact(s)--that
requires explanation. For stylized facts (or demi-regularities) to be
observable, however, the characteristics of the commodities in question and
their prices cannot be wholly idiosyncratic across exchanges. In other words,
both prices and commodities must be similar in a significant number of cases,
where the similarity of commodities is a necessary (but not sufficient)
precondition for meaningful similar prices to be observable. Without
similarity, we would have a multitude of heterogeneous and unconnected
exchanges, which have nothing in common, that could serve as the starting
point for further generalized analysis. At the same time, regularity suggests
that the factors which underlie different exchanges are si milar.

Of course, the condition of commodities being sufficiently similar is usually
fulfilled for many natural resources or other bulk materials. For manufactured
commodities, by contrast, things are more complicated as even apparently
homogenous commodities such as clothes reveal upon closer inspection a
multitude of features. The question then arises as to which features of a
commodity can be neglected (abstracted from) in order to include the commodity
in question into a category such as "middle price white shirts" whose price
may then exhibit certain stylized facts, and which features are essential to
the commodity. This cannot be done from an armchair's perspective but requires
the analyst to examine how buyers and sellers themselves perceive the
respective commodity. [9] Lumping together commodities on the basis of an
often arbitrary selection of physical properties is a widespread habit among
statisticians, but research in the sociology and psychology of consumption
reveals that in reality physical character istics hardly play the role that is
accorded to them in the construction of statistics and, consequently, in
economic analysis (cf. Rosenbaum 1999).

In a similar vein, the researcher has to decide whether all observed prices
have to be considered in the search for stylized facts or whether only
sufficiently similar prices are included. After all, price differences between
otherwise homogenous commodities may themselves be something that has to be
explained. This is particularly evident once a potential market stretches over
a large geographical area. Here two observed prices may belong to two separate
markets.

Exchange, then, is typical if the primary subject matter of exchange (good or
service) can be regarded as remaining largely unchanged across a number of
exchanges or, in the case of barter, if both parts of the exchange are similar
across a number of exchanges. And only by virtue of exchanges being typical is
it possible to observe stylized facts about prices. Note that these
considerations also hint at criteria for distinguishing between markets. Thus
if a group of exchanges can be described in terms of certain stylized facts,
while another group of exchanges is characterized by another set of stylized
facts, then these two groups obviously belong to, or constitute, different
markets (even though the commodity in question is basically homogenous).
Stigler and Sherwin, for instance, use the degree of correlation between
prices for grain in various locations (as such a stylized fact) in the US in
order to determine the extent of the respective market (Stigler and Sherwin
1985).

The condition of regularity basically means that similarities between
commodities remain constant over a certain period of time or, at least, that
they do not decrease rapidly. This condition is related to the condition of
typification just discussed because a number of exchanges which exhibits
certain features (e.g. roughly the same characteristics) is hardly ever
carried out simultaneously. Even so, regularity is a criterion in its own
right. Not only must phenomena last for some time in order to be observable,
the relevance of any social scientific explanation also depends on the
persistence of the phenomenon that is to be explained. To be sure, it is at
times the change in some variable, rather than the persistence of its
numerical value, that calls for analysis. But even then the change has to
stand out in the flux of events, which, after all, comprises all sorts of
changes, by exhibiting a certain degree of stability. In other words, change
can only be observed if the change itself remains at least part ly unchanged
or if the change takes place against a background that exhibits a certain
degree of stability.

These considerations apply not only to the characteristics of the commodity
itself, but also (but in a somewhat more limited sense) to its price. The
deviation of the observed price from the mean value, for instance, may be of
limited relevance if this is part of the fluctuation of prices around the
mean. But things are clearly different once a new mean value is being
established in the course of time. This is not to suggest, of course, that the
volatility of prices cannot constitute an explanandum in its own right, nor
that prices are required to remain constant or sticky for an extended period
of time. Rather regularity means that, first, prices are not so volatile as to
render it impossible to specify what exactly is the price that is to be
explained, i.e. if we can observe more then merely white noise. Second, it
means that in order to link observed prices causally with a set of potential
explaining factors, both prices and explaining factors must persist (often
parallely) for some period of time, the len gth of which depends on the type
of commodity in question.

Operationalization

The operationalization of regularity and typification requires the economist
to make a judgment informed by partly extra-economic considerations (physical,
biological, psychological etc.) about the significance of differences between
observed exchanges. Since such a judgment is likewise necessary for compiling
all sorts of statistics, it does not seem to pose problems, which exceed
those, encountered whenever non-identical items are to be collected under a
single heading. There are no reasons to suggest after all that the
distinctions used in official statistics always correlate with a meaningful
distinction between markets, both geographically and substantially. And as
with the definition of a market in general, the operationalization of
individual criteria is unlikely to be achieved once and for all. Since theory
and empirical application are mutually dependent, we should expect that some
operationalizations have to be revised in the light of a closer examination of
the market they helped to identify in the first place. But this is not an
argument against operationalization per Se.

3. Competition

The Argument

Competition between either sellers or buyers or both is the fourth
characteristic of a market. In contrast to neoclassical theory, however,
competition as understood here refers only in a very limited sense to the
structure of a market in terms of number of buyers and sellers on each side.
Rather, competition is seen, following Georg Simmel, as a form of indirect
conflict which is not directed at the opponent but consists of a parallel
effort, attempting to surpass an opponent by offering opportunities for
exchange which are preferred by other buyers or sellers (cf. Simmel 1903).
Since such efforts presuppose that economic agents can create and, at least
temporarily, maintain informational asymmetries regarding preferences and
technologies, 'perfect competition' in the neoclassical sense with its
idealizing assumption of perfect information precludes Simmelian competition.
The neoclassical notion of competition and the Simmelian understanding thus
only agree with respect to the claim that true monopolies and monopsonies
without the possibility of exit preclude competition because (in contrast to
auctions and tenders with only one seller or buyer) neither potential buyers
nor potential sellers compete with each other. [10]

The criterion of competition is important because it excludes not only
monopolies and monopsonies without exit (here competition and voluntarity
overlap conceptually) [11] but also forms of exchange which, even though they
are voluntary, specified, regular and typified, have purposes other than
personal gain, e.g. maintaining or establishing social relations via gifts.
[12] Competition is thus in a way the alter ego of the impersonal nature of
market exchange in modern economies emphasized by Max Weber. This becomes
evident if we consider the role played by the identity of the exchange
partners including his or her status, social function etc. vis a vis his or
her actions, i.e. the degree of embeddedness of the exchange (cf. Granovetter
1985). In other words, if we focus on what the competitor is as opposed to
what he or she does.

In the most extreme case, both buyers and sellers are unknown to each other
and so is their social status. In such a situation buyers and seller do not
share any personal bonds. They only acquire a degree of mutual familiarity
(and may also develop more intimate personal relationships) in the course of
obtaining and formulating offers for potential exchanges (this is an aspect of
the "vergesellschaftende" effect of competition emphasized by Simmel (1903).
Even so, their respective identity does not influence the decision with whom
to trade nor does it influence the terms of the trade. What matters are only
the features of an offer relative to the features of other offers and the
correspondence of offers with one's own desires and preferences. If, on the
other hand, maintaining or establishing social relations motivate exchange,
then the identity of the trading partners as well as the identities of fellow
buyers or sellers are of utmost importance and have to be known prior to the
actual transaction. For it i s only with respect to these identities that the
full content of the exchange can be determined.

Clearly Simmelian competition and neoclassical oligopoly, in particular game
theoretic models, have much in common in that both depart from the framework
of perfect competition by conceptualizing competition as some form of
strategic interaction. But there are two important differences. First, on
Simmel's account, interaction with competitors and customers is not an
indicator of imperfection as in neoclassical economics, where interaction
comes to an end when prices are given, but the hallmark of competition.
Without interaction, there is no competition whatsoever. Second, Simmel
acknowledges the social nature of economic agents and the way in which
different forms of interaction affect the embeddedness of economic agents.

As with specificity, competition as understood here contributes to the
possibility of analyzing exchange in largely economic terms, i.e. by
abstracting from determinants of the outcomes of exchange that have to do with
idiosyncratic relations between individual buyers and sellers. For such an
analysis would evidently be impossible, or at least lack explanatory weight,
if observed prices were merely surface phenomena without real significance to
the parties to the exchange. To be more specific, the notion of a price would
loose much of its significance as a decision variable for agents, if the
latter cease to see the acquisition of goods or services as the main purpose
of exchange transactions. The insignificance of the identity of buyers and
sellers in competition highlights an important difference between market
exchange and bargaining. Market exchange is not only different from bargaining
because the latter usually precludes exit but also because bargaining is often
shaped by pre-existing social relations, including trust and power, which are
welded to the identity of those who bargain.

Note moreover that competition as understood here likewise helps to
distinguish between exchange on markets and exchange in the context of central
planning. For while central planning involves various kinds of typified and
regular exchanges, buyers and sellers do not compete with each other (except
for attempts to bribe superiors in order to ensure low output targets and high
input allocations). Finally, competition also distinguishes markets from
hierarchies and organizations in which interactions are coordinated through
commands backed by authority and power and in which competition is often
deliberately restricted so as to safeguard the interests of the organization
as a whole as opposed to those of individuals.

Operationalization

The criterion of competition may seem to be more difficult to operationalize
than the criteria discussed above. The reason is that interpersonal relations,
or their absence, are notoriously difficult to observe. As a proxy, however,
one may take obviously suboptimal but persistent exchanges (given the
information actually available to agents with bounded rationality) as
indicators for the existence of factors (such as personal bonds) which prevent
agents from exploiting all opportunities for exchange, and hence as indicators
for limits to competition. Another possible indicator for competition could be
the amount of resources spent by both buyers and sellers for collecting and
disseminating information about potential exchanges.

4. Degrees of Marketness

The above criteria should not be taken to represent discrete borderlines,
which tell us exactly when a market exists and when it doesn't. Rather, they
represent characteristics by which real world situations have to be judged and
evaluated. Thus what matters is not the presence or absence of competition but
its degree, as it is the degree of typification and regularity that can be
observed rather than the question of whether we are dealing with a truly
homogenous commodity such as tin. The same goes for specificity and
voluntarity. Few exchanges are fully specified, and even if they are there may
be an element of uncertainty involved because one cannot exactly know ex ante
what one is to get in return. Similarly, voluntarity may be reduced but not
necessarily absent in a monopoly. Following the terminology of Block exchange
situations can be said to exhibit different degrees of "marketness" which are
in turn characterized by differences in the degree to which the above criteria
are realised (Block 1990). Cons equently, an economy is not composed of
markets on the one hand and non-market exchange situations on the other, but
may include a variety of exchange situations from wholly unstructured singular
exchanges at one end of the spectrum, through to "idealtypical" markets at the
other.

In a similar vein, not every market economy exhibits the same degree of
"marketness" across time and space. Neither is the number of markets in an
economy constant over time, nor is its number equal to that in other
economies. The concept of degrees of marketness is thus similar to the
neoclassical notion of market perfectness in that perfectness too is not a
dichotomous criterion. In contrast to the neoclassical understanding, however,
the present account does not focus on the extent to which agents can influence
relative prices but on the existence and suitability of prices and quantities
as the subject matter of economic analysis. Obviously, a lower degree of
marketness reduces the extent two which prices and quantities can become the
objects of analysis.

IV. MARKETS, INSTITUTIONALIZATION, AND INFORMATION

In section III, I introduced and defended three broad criteria by which
markets can be distinguished from other exchange situations. Let me repeat the
main points. Voluntary and specified exchange as well as competition would
appear to be quite straightforward criteria; the first in order to
discriminate between forms of exchange that are open-ended, as it were (e.g.
gifts, marriages), and exchanges that are more specified and thus involve a
price; the second in order to exclude forms of exchange that are motivated by
reasons which have to do with the identities of the trading partners rather
than the opportunity for exchange he or she offers (the latter aspect also
excludes familial relations from the province of the market). Moreover, both
bargaining processes and central planning are characterized by the absence of
competition. Typification and regularity, finally, are imposed in order to
bound the range of exchanges, both cross-sectionally and intertemporally so as
to clarify what exactly (which observed exchange) is to constitute the focus
of economic enquiry.

Routines and Institutions

Note that typification and regularity can also be seen as the most basic
elements of the institutional character of markets, if one conceives of
institutions, in the most fundamental sense, as routines (cf. Lawson 1997).
That is, markets constitute institutions because they involve certain routines
(typical exchanges) that are repeated over time (regularity). In this respect,
the present approach adopts a broader understanding of institutions, which is
not confined to institutions as sets of rules with or without enforcement
mechanism (North 1990). Rather than investigating the institutional framework
of exchange, the present account thus emphasises the institutionalized form of
exchange.

Clearly, many markets can also be regarded as institutions in a more specific
sense in that market clearing and the determination of prices follow
(procedural) rules or in that there is a number of auxiliary institutions
and/organizations which facilitate certain types of exchange. This, it would
seem, is the view Hodgson adopts (cf. Hodgson 1988). It is at this point,
therefore, that the social and cultural embeddedness of economic interaction
has to be considered in more detail. For Hodgson's auxiliary institutions or
organizations, which serve to overcome problems relating to moral hazard,
trust and asymmetric information, are likely to be shaped by the social and
cultural setting. There are no reasons to suggest after all that similar
problems are always solved by using the same set of institutions. In fact, the
available evidence on modem capitalist societies indicates that there is not
only a wide range of institutional forms in use (e.g. systems of corporate
governance and finance, cf. Grosfeld 1994; Zysman 1983), but when it comes to
addressing specific problems within the economic sphere, there is also a
tendency to adopt institutional or organizational forms which originate
outside the economic system and which have to do with the culturally
impregnated "Wirschaftsstil" (Schefold 1994) of a society. The German habit of
seeking consensual solutions to social conflicts is perhaps a case in point.

Clearly, the impact of cultural factors varies across societies. At one
extreme, certain forms of exchange are tabooed or so heavily overshadowed by
cultural influences that economic factors can hardly be detected, at the other
extreme culture provides little more than the background of transactions in
the form of general norms of behavior. From this perspective, Polanyi's "Great
Transformation" (Polanyi 1995) can be interpreted as a process whereby
cultural factors are pushed to the rear while letting "market forces" play a
more prominent role.

Prices and Information

Importantly, institutionalized exchange as routines is a precondition for
prices to fulfil the informational role which figures so prominently in the
works of Hayek and other Austrians (cf. Hayek 1945; Kirzner 1997). For unless
exchanged commodities exhibit an adequate degree of similarity, the range of
prices at which this commodity is traded is likely to be too large to convey
any information to other economic agents. By the same token, prices must
display a certain degree of stability over time (across exchanges) in order
for price changes to embody significant information, notwithstanding any
information possibly contained in the phenomenon of volatility or dispersion
itself (see below). The more volatile prices, the less reliable is whatever
information on demand or technology is encapsulated in any particular price or
in any particular price change. Volatile or dispersed prices are thus less
able to coordinate decentralized decisions. Thus in contrast to the Austrian
approach, the present account argues that it is precisely because of the fact
that markets involve specific, regular and typified exchanges that market
prices are able to provide the information, which individuals then acquire,
store and use as economically relevant knowledge. This is partly acknowledged,
if only implicitly, by Hayek when he refers to tin, i.e. to a homogeneous
commodity, and similar commodities with one price, in order to illustrate his
point about the informational role of the price system, and when he talks
about an apparently permanent increase in the price of tin rather than
short-run fluctuations (Hayek 1945). However, Hayek does not discuss how
prices of heterogeneous commodities can convey information in a similar
manner. Hence, by failing to fully acknowledge the connection between the
degree to which exchange is institutionalized or routinized and the
informational content of prices, this version of the Austrian approach is
prone to overstate the informational role of the price system. A corollary of
this view is that not every observable price conveys the same amount of
information to producers and customers.

There is of course another sense in which the informational role or the price
system can be understood in Hayek's work (cf. Kirzner 1992). Accordingly,
prices do not only coordinate economic activity because they are already so
adjusted that decisions on their basis become self-enforcing (equilibrium
prices), but also because disequilibrium prices reveal to market participants
that altered decisions may be advantageous in the future. Thus disequilibrium
prices provide incentives to agents to modify their behavior so that,
potentially, prices approach equilibrium values (Kirzner 1992). Note though
that there are no immanent reasons why disequilibrium prices should be uniform
if trading takes place in a decentralized fashion. This brings us back to the
methodological issues discussed above. Since there may be a multitude of wrong
prices, there is no stylized fact or demiregularity which could become the
subject matter of analysis in the same way as (uniform) equilibrium prices
can.

IV. CONCLUDING REMARKS: IN SEARCH OF MARKETS

If one adopts the above elements of markets, then the examples outlined in the
introduction indicate that the term "market," as it is used in these examples,
refers to particular components of markets (sellers, buyers) or to certain
characteristics of markets (competition, exchange). But to assert that a
market in the fuller sense of the term does indeed exist in each case would
mean jumping to conclusions. For instance, to claim, as Gary Becker does, that
a market for marriages exists (cf. Becker 1976), obscures the fact that
marriages are largely unspecified exchanges and hence violate the condition of
specificity. In a similar vein, it does not make much sense to talk about a
market for dams or fighter airplanes (but perhaps radar components or
generators), since many of these exchanges are neither standardized nor (as in
the case of military equipment) always subject to competition. Of course, this
critique applies a fortiori to authors who endorse (without further
elaboration) the existence of a religiou s market on the grounds that
individual denominations function as religious firms and thus collectively
constitute a religious market (cf. Iannaccone 1998).

While these examples seem largely uncontested, there is another case where
speaking of a market proper or of a collection of markets seems doubtful. The
example I have in mind is the labor market. We may not only question the
existence of a labor market because, as some post-Keynesians argue (cf. Lavoie
1992), wages and working conditions are predominantly determined by social
customs and norms rather than demand and supply conditions. There are also
reasons to suggest that labor markets fulfil only in part the criteria for
markets which I have developed above. While competition, regularity and
typification would seem common features of many segments of the labor "market"
both voluntarity and specificity are more problematic. As to voluntarity, it
is clear that the opportunity costs of not accepting an exchange offer (job)
can be extremely high, at least in the absence of unemployment benefits or
similar means of subsistence. Furthermore, labor contracts are prime examples
of relational contracts (cf. Goldbe rg 1998) in that, due to their long-term
nature, it is impossible to specify ex ante all rights and duties in the
contract. At the same time, employer and employee may make specific yet
unpredictable investments in the course of their relationship, which change
the content of the work done. Hence, while both the wage and basic features of
the job are usually known in advance to employer and employee, many other
aspects of their relationship will only become known much later, contingent as
they often are upon external circumstances and events. Consequently, if both
parties to the exchange have only incomplete knowledge about the nature of one
commodity then the commonplace understanding of a price as the relative
amounts of two commodities (one possibly being money) as well as the notion of
a market for one type of commodity is called into question.

This is not to say, of course, such exchanges are economically insignificant,
nor that they could not be analyzed. The point is that methods of analysis
which presuppose specificity because they seek to explain prices and/or
quantities seem inadequate.

Eckehard F. Rosenbaum was born in 1967 and received his PhD from Cambridge
University in 1998. His research has focused on methodological issues in
welfare economics, transport economics and on the development of markets in
Central and Eastern Europe. From 1996 to 1999, he was research fellow at the
Frankfurt Institute for Transformation Studies in Frankfurt (Oder). He
currently works at the Federal Ministry of Economics and Technology, Berlin.

ACKNOWLEDGEMENTS

Previous drafts of this paper were presented at the joint graduate workshop of
the University of Economics, Poznan, and the European University Viadrina,
Frankfurt (Oder), at the Economics Colloquium in Frankfurt, and at the Seminar
of the Evolutionary Economics Unit, Max Planck Institute for Research into
Economic Systems, Jena. I would like to thank the participants for most
helpful comments and suggestions. I would also like to thank Michael Fritsch,
Geoff Hodgson, Dieter Schmidtchen and two anonymous referees for generously
commenting on earlier drafts. The views expressed in this paper are those of
the author.

(1.) In particular networks have recently received a lot of attention. See the
contributions in Grabher and Stark (1997)).

(3.) This is also evident in an otherwise thoughtful paper which provides
various reasons for distinguishing between markets and other forms of exchange
but then hesitates to acknowledge that the notion of market is redundant once
the concept of a market is synonymous with exchange (cf. Heinemann 1976).

(2.) cf. Hodgson (1988).

(4.) Of course, if one sees institutions as routines then there are certain
routines or roles (the role of buyer or seller for instance) which appear to
be ubiquitous and which are also necessary for markets to emerge. Yet these
roles are also played in the case of casual exchange acts. Hence they do not
suffice to identify markets.

(5.) This objective is similar to that of Menard (1995), who seeks to
differentiate markets as institutions from institutions as markets.

(6.) Such a contrastive mode of identification is similar in spirit to a
contrastive mode of explanation where the relevant question is not "Why did x
occur?" but "Why did x rather than y occur? (cf. Lipton 1993).

(7.) In technical terms, a common standard of value would require a cardinal
measure of value plus unit comparability.

(8.) In reality, the distinction between bargaining and central planning is
blurred since the setting up of plans usually involves extensive bargaining
between planners and enterprises.

(9.) Note again that perception is not the same as evaluation. The latter
presupposes the former and perceptions are less likely to differ across
individuals than are evaluations. Indeed, for communication to be possible,
perceptions must overlap to a certain degree.

(10.) A public utility such as water is presently a case in point, for buyers
do not compete with each other, nor do they have the possibility of exit.

(11.) In the present context, a monopoly/monopsony is defined in terms of a
specific characteristic that is supplied or demanded by only one supplier or
buyer, respectively. Thus BR may have been the only railway company in Britain
prior to privatization, but it was not the only supplier of transport
services, private ears, buses and, more recently, planes being close
substitutes. Hence BR was not a monopoly.

(12.) For a discussion of motives other than personal gain narrowly conceived
see Heinemann 1976.

REFERENCES

Arrow, K. J. and Hahn, F. H. (1971) General Competitive Analysis, San
Francisco: Holden-Day.

Aslund, A. (1995) How Russia Became a Market Economy, Washington, DC:
Brookings Institution.

Barnett, V. (1991) "Conceptions of the market among Russian economists: a
survey," Soviet Studies 44(6): 1087-1098.

Baumol, W. J., Batey Blackman, S. A. and Wolff, E. N. (1989) Productivity and
American Leadership: The Long View, Cambridge, Mass: MIT Press.

Becker, G. S. (1976) The Economic Approach to Human Behavior, Chicago:
University of Chicago Press.

Block, F. L. (1990) Postindustrial Possibilities: A Critique of Economic
Discourse, Berkeley: University of California Press.

Brezinski, H. and Fritsch, M. (1997) "Spot-markets, hierarchies, networks, and
the problem of economic transition" in H. Brezinski and M. Fritsch (eds) The
Emergence and Evolution of Markets, Cheltenham, UK: Edward Elgar: 7-19.

Carlton, D. (1989) "The theory and the facts of how markets clear," in R.
Schmalensee and R. Willig (eds) Handbook of Industrial Organisation,
Amsterdam: North-Holland: 909-946.

Coase, R. (1988) The Firm, the Market and the Law, Chicago: University of
Chicago Press.

Cournot, A. (1897) Research into the Mathematical Principles of the Theory of
Wealth, New York: Macmillan.

Debreu, G. (1959) Theory of Value--An Axiomatic Analysis of Economic
Equilibrium, New York: Yale.

EBRD (1996) Transition Report 1996, London: EBRD.

Fourie, F. C. V. N. (1991) "The nature of the market: a structural analysis,"
in G. M. Hodgson and E. Screpanti (eds) Rethinking Economics--Markets,
Technology and Economic Evolution, Aldershot: Edward Elgar: 40-57.

Friedman, M. (1962) Capitalism and Freedom, Chicago: University of Chicago
Press.

Fukuyama, F. (1992) The End of History and the Last Man, London: Hamilton.

Goldberg, V. P. (1998) "Relational Contract" in P. Newman (ed.) The Palgrave
Dictionary of Economics and the Law, London: Macmillan Reference Ltd: 289-293.

Grabher, G. and Stark, D. (eds) (1997) Restructuring Networks in
Post-Socialism: Legacies, Linkages, and Localities, Oxford: Oxford University
Press.

Granovetter, M. (1985) "Economic action and social structure: a theory of
embeddedness," American Journal of Sociology 91(3): 481-510.

Gravelle, H. and Rees, R. (1992) Microeconomics. 2nd ed., London: Longman.

Greif, A. (1989) "Reputation and coalitions in medieval trade: evidence on the
Maghribi traders," Journal of Economic History 49: 857-882.

Greif, A. (1994) "Cultural beliefs and the organisation of society: a
historical and theoretical reflection on collectivist and individualist
societies," Journal of Political Economy 102(4): 912-950.

Grosfeld, I. (1994) "Financial Systems in Transition: Is there a Case for a
Bank Based System?," CEPR Discussion Paper, 1062, Centre for Economic Policy
Research, London.

Hayek, F. V. (1945) "The use of knowledge in society," American Economic
Review 35(4): 519-530.

Heinemann, K. (1976) "Elemente einer Soziologie des Marktes," Kolner
Zeitschrift fur Soziologie und Sozialpsychologie 28: 48-67.

Hodgson, G. (1988) Economics and Institutions--A Manifesto for a Modern
Institutional Economics, Cambridge: Polity Press.

Hollingworth, J. R. and Lindberg, L. N. (1985) "The governance of the American
economy: the role of markets, clans, hierarchies and associative behavior,"
Discussion Paper, IIM/LMP 85-8, Social Science Research Centre Berlin, Berlin.

Iannaccone, L. R. (1998) "Introduction to the economics of religion," Journal
of Economic Literature XXXVI(3): 1465-1496.

Jevons, W. S. (1871) The Theory of Political Economy, London: Macmillan.

Kirzner, I. M. (1992) The Meaning of the Market Process--Essays in the
Development of Modern Austrian Economics, London: Routledge.

Kirzner, I. M. (1997) "Entrepreneurial discovery and the competitive market
process: an Austrian approach," Journal of Economic Literature XXXV: 60-85.

Kornai, J. (1994) "Transformational recession: the main causes," Journal of
Comparative Economics 19: 39-63.

Lavoie, M. (1992) Foundations of Post-Keynesian Economic Analysis, Aldershot:
Edward Elgar.

Lawson, T. (1997) Economics and Reality, London: Routledge.

Lipsey, R. G. (1983) Positive Economics, 6th edn., London: Weidenfeld and
Nicolson.

Lipton, P. (1993) Inference to the Best Explanation, London: Routledge.

MacNeil, I. R. (1974) "The many futures of contracts," Southern California Law
Review 47: 691-816.

MacNeil, I. R. (1985) "Reflections on relational contract," Zeitschrift fur
die gesamte Staatswissenschaft 141: 541-546.

Marshall, A. (1919) Industry and Trade, London: Macmillan.

Marshall, A. (1938) Principles of Economics, London: Macmillan.

McCloskey, D. N. (1986) The Rhetoric of Economics, Madison: University of
Wisconsin Press.

Means, G. C. (1938) "Notes on inflexible prices," American Economic Review
26(1): 23-35.

Menard, C. (1995) "Markets as institutions versus organizations as markets?
Disentangling some fundamental concepts," Journal of Economic Behavior and
Organization 28(2): 161-182.

North, D. C. (1977) "Markets and other allocation systems in history," Journal
of European Economic History 6: 703-716.

North, D. C. (1990) Institutions, Institutional Change and Economic
Performance, Cambridge: Cambridge University Press.

Okun, A. M. (1981) Prices and Quantities: A Macroeconomic Analysis,
Washington, DC: The Brookings Institution.

O'Neil, J. (1998) The Market: Ethics, Knowledge and Politics, London:
Routledge.

Polanyi, K. (1995) The Great Transformation: politische und okonomische
Ursprunge von Gesellschaften und Wirtschaftssystemen, Frankfurt (Main):
Suhrkamp.

Richter, R. (1996) "Die Neue Institutionenokonomik des Marktes," Lectiones
Jenenses 5, Max-Planck-Institut zur Erforschung von Wirtschaftssystemen, Jena.

Richter, R. and E. Furubotn (1996) Neue Institutionenokonomik: Ein Einfuhrung
und kritische Wurdigung, Tubingen: J.C.B. Mohr (Paul Siebeck).

Rosenbaum, E. F. (1999) "Against naive materialism--culture, consumption and
inequality," Cambridge Journal of Economics 23(3): 317-336.

Schefold, B. (1994) Wirtschaftsstile--Band 1: Studien zum Verhaltnis von
Okonomie und Kultur, Frankfurt (Main): Fischer.

Searle, J. R. (1995) The Construction of Social Reality, London: Allen Lane.

Shand, A. H. (1984) The Capitalist Alternative: An Introduction to
Neo-Austrian Economics, Brighton: Harvester Wheatsheaf.

Simmel, G. (1903) "Soziologie der Konkurrenz," Neue Deutsche Rundschau 14(4):
1009-1023.

Simon, H. S. (1991) "Organizations and markets," Journal of Economic
Perspectives 5(2): 25-44.

Snehota, I. (1993) "Market as network and the nature of the market process" in
D. Deo (ed.) Advances in International Marketing, Greenwich CT: JAI Press:
31-41.

Sraffa, P. (1960) Production of Commodities by Means of Commodities,
Cambridge: Cambridge University Press.

Stigler, G. J. and Sherwin, R. A. (1985) "The extent of the market," Journal
of Law & Economics XXVIII: 555-585.

Swedberg, R. (1994) "Markets as social structures" in N. J. Smelser and R.
Swedberg (eds) The Handbook of Economic Sociology, Princeton, NJ: Princeton
University Press: 255-282.

Williamson, O. E. (1983) Markets and Hierarchies: Analysis and Antitrust
Implications, New York: Macmillan.

Wittgenstein, L. (1960) Blue Book, Oxford: Blackwell.

World-Bank (1996) World Development Report 1996: From Plan to Market, New
York: Oxford University Press.

Zysman, J. (1983) Governments, Markets, and Growth, Ithaca: Cornell University
Press.
               The market in contrast to other social forms
                   Market    Firm/          Central      Bargaining  Casual
                          Organization     Planning                 Exchange
Voluntarity         Yes       Yes             No             No       Yes
Specificity         Yes        No             Yes           Yes       Yes
Regularity          Yes       Yes             Yes            No        No
  and Typification          (mostly)       (mostly)
Competition         Yes        No             No             No       Yes
                                       (or very limited)
                   Gifts
Voluntarity         Yes
Specificity         No
Regularity          No
  and Typification
Competition         No

                                -- End --

Delete | Reply | Reply to All | Forward | Redirect | Blacklist | Message Source | Save as | Print

Back to INBOX

Move