__________________ ____________________  

Market Failure and Uncertainty


In this essay I would like to argue that market failure is
the reason for hierarchy. I will
argue that market failure leads to uncertainty and that
uncertainty leads to incomplete
contracts. To explore the notion of incomplete contracts I
have used the ideas of Milgrom
and Roberts. They describe how incomplete contracts can
lead to a "relational contract" or
an "implicit contract. I will argue that relational
contracts can lead to the formation of
an organisation. Once an organisation is established it
differs from other equally
organisations by its social capital. This social capital, I
will argue, leads again to
uncertainty and again to incomplete contracts. To prove
this I will use repeated game theory
and the folk theorem. I believe that the uncertainty within
an organisation leads to
implicit contracts that capture the reputation of
cooperation of the members of the
organisation. In conclusion I will argue that this
reputation of cooperation is corporate
In the first part of 'Managerial Dilemma' Miller argues
that hierarchy, at least in theory,
exists by the grace of market failure. Market failure is
described by Miller (1992) as:
"(...) market equilibria that fail to allocate resources
efficiently" (p.17). Miller states
that market failure is caused by market power, information
asymmetries and (team)
externalities. However if I take a closer look at these
reasons for market failure I will
argue that they can be the reason to form an organisation,
but not nesecarily have to be.
Market power by a small number of parties increases the
amount of uncertainty about how
these parties are planning to use this power in ongoing or
future transactions when faced by
unforeseen contingencies. Information asymmetry -when one
party is deprived of vital
information available to the other side- also raises the
amount of uncertainty about how the
other party is likely to react to unforeseen contingencies.
The third reason for market
failure (externalities) can also be seen as driver for
uncertainty since transactions in the
marketplace can have an impact on third parties.
So far, I have argued that market failure leads to
uncertainty. The general assumption
presented in the lectures is that uncertainty leads to
incomplete contracts and
interdependency between the parties involved in a
transaction. This takes me to the next
step of my argument. This general assumption makes parties
to design their contracts
recognising that they cannot possibly be perfectly adapted
to all possible future
circumstances. Milgrom and Roberts (1992) describe how this
can lead to a "relational
contract" or an "implicit contract". 
A relational contract is defined as: "A contract that
specifies only the general terms of a
relationship and specifies mechanisms for decision making
and dispute resolution"(p.602). An
that are not legally enforceable
but that the parties consider to be binding on one
another's conduct" (p.599).
An employment contract is a good example of a relational
contract, because it delegates
authority to the employer to direct the employees actions
rather than describing precisely
the work that has to be done in every contingency. In this
stage of my argument I would like
to take the notion of relational contracts further. 
If an exchange in the market is more costly than the same
exchange in an organisational
structure, due to the degree of uncertainty caused by
market failure, one of the parties can
decide to add the relation to its organisation structure.
Especially if we assume that based
on rationality a contract will occur within whatever
infrastructure minimises its costs.
This notion combined with the fact that most people dislike
having their incomes dependent
on random factors and are therefore "risk averse". I argue
that uncertainty caused by
market failure can lead to relational contracts such as an
employment contract and therefore
can lead to the formation of an organisation. I put
emphasis on the word can because if the
parties concerned in the transaction are all "risk neutral"
and there are no wealth effects
it is more likely that the same uncertainty will lead to
the building of trust between the
parties by means of implicit contracts in order to prevent
that one of the parties behaves
"opportunistic" (e.g. monopolises the contracts residual).
Later on I will return to the
notion of implicit contracts when I discuss the role they
play in social capital.
Until this point I have argued that market failure leads to
uncertainty and that that 

uncertainty can lead to relational contracts and therefore
can lead to the formation 

of an organisation. Now I would like to argue that once an
organisation is formed 

uncertainty within the organisation can lead to implicit
contracts. Furthermore I would like
to argue that these implicit contracts can lead to social
capital and therefore can lead to
a cooperative corporate culture.
How can two organisation that have the same capital, human
capital, technology, operating
environment and the same incentive system differ in
performance? In the lectures we argued
that in the absence of non-standard utilities, one answer
can be that they differ in "social
In class we defined social capital as follows: "B has
the extent that she has either pre-existing or can generate
at 'low costs' relationships
from 'others' whereby B can draw (with high probability)
upon an insecure commitment of
resources from others (i.e. in the absence of collateral at
zero interest rate)." 
A will invest in the social capital of B only if she
believes there exists high probability
that B will discharge the debt at an appropriate time and
in a self-enforcing manner. In
other words A must not feel uncertain about B returning her
'help'. This brings me back to
the first notion that unforeseen contingencies can
influence transactions. Earlier I have
showed that uncertainty can lead to incomplete contracts
and that these incomplete contracts
can lead to hierarchical transactions. However, I will
argue here that it will lead to
implicit contracts because I believe that the existence and
the level of social capital
within the firm depends on the reputation of cooperation of
the individuals within an
organisation. Cooperation will be defined as occurring when
individuals in a social dilemma
select alternatives that are not rewarded by the formal
incentive system but that result in
Pareto efficient outcomes (Miller 1992). The other party
will only select these alternatives
if it believes that it will be reciprocated. Kreps (1984)
argues that the source of this
faith is the "reputation" of the other party.
In order to get a clear picture about reputation, Kreps
uses repeated game theory and the
folk theorem. Repeated game theory can best be approached
from a one-sided version of the
Prisoners' dilemma game; the trust game (see figure 1). If
this game is played only once and
noncooperative both of the parties have a dominant strategy
to not trust. The Nash
equilibrium will therefore be a $0 pay-out to both A and B.
From an economical point of view
this is a Pareto inefficient outcome.
A Pareto efficient outcome can be achieved by a
"tit-for-tat" strategy in repeated
prisoner's dilemma games. A tit-for-tat multiperiod
strategy is defined as playing the
cooperative (dominated) alternative in the first play of
the game and thereafter mimicking
the other player's previous choice. In other words the
trust-honour arrangement allows for
the possibility of self-enforcing if the game is repeated
with sufficiently high
Unfortunately, even when the game is repeated with
sufficiently high probability a Pareto
efficient outcome is not inevitable. The folk theorem
states that in any repeated Prisoners'
dilemma game, there are an infinite number of outcomes that
are sustainable as long-run
equilibria by rational, self-interested actors. Imagine for
instance what will happen if B
states that he will violate A's trust once every three
times the game is played. And
furthermore, if A ever retaliates by not trusting B, player
B will be less likely to trust A
again. A's dominant strategy will still be trusting B all
of the times, as long as B does
not violate A's trust more than once every three times.
This is just one example of the wide
range of equilibria of a repeated Prisoner's dilemma. 
Even in a repeated game the player's have an incentive to
specify and enforce the terms of
the transaction and therefore allow the transaction cost to
"eat" some of the residual
gained by the transaction. This is where reputation comes
in. Reputation means that the
trusted party will honour that trust because not honouring
it would preclude or
substantially limit opportunities to engage in future
valuable transactions. To make this
point clear I will use the social capital example that I
used earlier. Say that A invests in
the social capital of B and B does not reciprocate. If this
noncooperation of B is
observable A will never invest in the social capital of B
anymore, and furthermore she will
refuse to reciprocate B's future investments in her social
capital. This observable failure
to reciprocate reduces B's social capital to zero. If we
take this "bilateral reciprocity"
up to a corporate level we can speak of "generalised
reciprocity". This means that B depends
on several A's to build up his social capital. This
generalised reciprocity is much more
efficient because it reduces the chance that any A is not
available at the moment that B
needs her help. On the other hand, the shift towards
generalised reciprocity exacerbates
"free-riding" problems as B will attempt not to reciprocate
all A's and as A's will attempt
not to reciprocate B.
I have argued that the existence and the level of social
capital within the firm depends on
the reputations of cooperation of the individuals within
the organisation or, in other
words, the implicit contracts. And I have argued that the
social capital is more efficient
when it exists out of generalised reciprocity. Furthermore
I have showed with the use of
repeated game theory that there is a possibility that
individual behaviour in a coordination
game is rationally constrained by social conventions and
norms. The viability of this
convention depends on the level of generalised reciprocity
(Miller 1992). 
I believe, to conclude my argument, that corporate culture
is a mechanism of generalised
reciprocity. This takes me back to Miller when he states:
"The term 'corporate culture'
denotes these mutually reinforcing expectations in a firm.
A 'cooperative' corporate culture
is one in which each player expects all others to cooperate
and to enforce the norm of


Quotes: Search by Author