<<

. 14
( 25 .)



>>

successful decision-making is always true. Moreover, an induction-based
decision-making as well as the essential role of knowledge. More
knowledge involves a very static (since it is timeless) concept of
important, an adequate explanation of the evolution of institutions must be
knowledge, one which begs the question as to why there should ever be a
based on a theory which explicitly gives institutions a broader role than is
change in long-run variables. This methodological problem can be
allowed by seeing them as merely static constraints on the choices of any
overcome by explicitly recognizing the role of the decision-maker™s
individual decision-maker. I will outline a theory of institutions which will
knowledge and by recognizing that changes are usually the result of
form a basis for an adequate explanation of institutional dynamics.
systematic failures due to reliance on false knowledge, rather than of
Although my theory will not necessitate giving up the fundamental
systematic successes based on necessarily true knowledge.
assumption of rational decision-making, it will show that all neoclassical
Second, if the ultimate basis for any explanation of the changes of the
theories of institutional change are very special cases.
institutional constraints is outside the neoclassical explanation, then the
To begin, I would like to note that the critical issues of the adequacy of
pro-neoclassical view cannot be sustained. As noted before, to avoid
the knowledge available to a decision-maker and the methodological role
circularity every explanation of any set of variables requires the recognition
of institutions are not independent. The reason is simple. One of the roles
of one or more exogenous variables. It should be obvious, then, that
that institutions play is to create knowledge and information for the
without a change in at least one exogenous variable (e.g. in an ideological
individual decision-maker. In particular, institutions provide social
perspective in North™s theory), the long-run neoclassical economy is static,
knowledge which may be needed for interaction with other individual
since there is no reason for a change in the endogenous variables (such as
decision-makers.9 Thus, the following theory of institutions emphasizes the
institutional constraints) once the optimum values of the institutional
primary role of social institutions, namely, to institutionalize social
˜constraints™ have been successfully established. If, for example, the
knowledge. However, for an adequate dynamic theory, I will avoid the
optimizing changes in the endogenous constraint variables are to be
presumption of successful decision-making; thus, in particular, I will not
explained as the result of changes in the exogenous ideology variable, then
assume that the social knowledge is correct, even though it may be durable.
by definition of ˜exogenous™ (not explained within), that change in
But I go too fast. Let me proceed very deliberately by putting my theory in
ideology must be explained outside the neoclassical explanation of
the form of explicit propositions.
institutions “ an exogenous ideology cannot be an object of optimizing
choice. But even worse, if one wishes to make ideology an endogenous Proposition 1. All sociological acts are based on expectations of
variable in a neoclassical model, then another new exogenous variable expectations. Specifically, all interactive decision-making involves the
actor™s knowledge of the other individuals™ knowledge. 10
must be invented. Of course, having to invent a stream of new exogenous
variables as the neoclassical programme progresses merely means that one
The significance of this proposition lies primarily in the conceivable alter-
is marching down the long road of the infinite regress.
natives, such as the actor™s direct questioning of the other individuals. 11
These methodological considerations reveal, I think, the inherent
poverty of every neoclassical programme for explaining the evolution of Proposition 2. All social problems result from conflicts over
the organizational structure (institutions) of an economy as the dynamic expectations (or knowledge), which in turn result from the lack of
consequences of constrained optimization. Specifically, these acceptable limits on the range of expectations (at either source).
considerations call into question the adequacy of the decision-maker™s
The significance of this proposition is dependent on the first and would
knowledge by questioning the presumed success of the intended
mean little without it. Since most of our everyday experience involves
optimization. They also question the neoclassical view of the nature of
previously solved social problems, it would be fairly difficult to give a pure
institutions which, for methodological reasons, views them as static
description of any social problem apart from its assumed solution. Thus, I
constraints facing the short-run optimizer.
© LAWRENCE A. BOLAND
120 Principles of economics Knowledge and institutions in economic theory 121
turn directly to solved social problems. If all institutions were viewed as essentially concrete, it would lead to the
view called ˜institutional holism™ (sometimes called ˜collectivism™). 14
It should be clear that, based on the second proposition, all solutions to
social problems involve the limits on expectations. There are basically two The theory formed here views institutions as social conventions which
different ways of limiting expectations: (1) narrowing the range of possible can be influenced by individual members of the society but which also
options (with prohibitions, taboos, and so forth), and (2) increasing the extend (in terms of time or space) beyond the individuals and thereby can
likelihood of particular possible options (with norms, standards, guides, influence the individuals either as constraints or as instruments of change.
conventions, and so forth). This brings me to my third, fourth and fifth How the institutions can be influenced depends on the institutions designed
propositions. to deal with that problem (such as election rules). This theory can best be
understood in terms of a sequence of events or steps.
Proposition 3. All social institutions exist to solve social problems.
Step 1. A society faces a problem for which there is at least one
Proposition 4. All social institutions can be divided into two categories:
conceivable solution.
consensus institutions, which exist as socially accepted solutions to
specific problems (or to a set of problems), and concrete institutions, Step 2. A consensus is formed around one particular solution, thereby
which exist to solve social problems resulting from relying on consensus establishing a consensus institution.
institutions (e.g. common agreements) to solve problems.
The establishment of the consensus may depend on a political process.
Proposition 5. All concrete institutions are attempts to manifest the In the modern urban world, a consensus is virtually impossible to achieve.
extent of a society™s learning, that is, they are a society™s social One can easily see that the institutions of political parties and platforms are
knowledge. parts of a solution to the problem of forming a consensus. Specifically, a
platform ties together a set of problems for each of which a consensus for a
And, as a corollary of the fifth proposition, I note:
particular solution cannot be obtained. To construct a consensus, every
Proposition 5a. The sole job of a concrete institution is to represent a party member agrees to support all planks in the platform, even though he
given particular consensus institution (or system of institutions). or she may not be interested in every plank.
There are many examples of concrete institutions; the American Step 3. It is recognized that the solution of Step 2 has inherent
Constitution is the most obvious, and legal contracts are the most common. methodological difficulties because a consensus institution is limited in
Consensus institutions are much less obvious, but one can identify all terms of space and time.
˜unwritten laws™ and ˜gentlemen™s agreements™ as common examples. 12
In particular, the solution of Step 2 will be limited to the members that
Propositions 1, 2 and 3 form a static theory of institutions. That is, one
form the consensus in terms of both their life-span and their number. For
can explain the existence of an institution by explaining the problem for
example, in this semester™s seminar, everyone may know what to expect of
which the institution was intended to be (or accepted as) a solution. 13 Such
one another in terms of operating rules, but next semester (or in any other
problems include those discussed by North and others. One can also
seminar at the same time) there will be a new set of students who may not
explain the continuance of the institutions by explaining the current
know what to expect. Thus, every semester a new consensus will have to be
problem for which the members of the society think the institution is a
reached. The fact that there is no carry-over from one period (or place) to
solution. In both cases the individual members may be mistaken, either in
another is in effect another social problem for which some form of
terms of the competence of the solution (as it may not do the job) or in
durability is the only solution.
terms of the realities of the problem (it may be a false problem or an
impossible one to solve). Step 4. The society establishes a concrete institution to represent the
The addition of Propositions 4, 5 and 5a allows for a dynamic theory of consensus of Step 2; however, the durability or concreteness of the
institutions. More technically, these propositions form what has been called institution is merely another consensus institution.
˜institutional individualism™ [see Agassi 1974; Boland 1982a, Chapter 2]. If
Durability is the essential ingredient for a truly dynamic model, even if the
all institutions are considered to be essentially of the consensus type, it
durability is not exogenous.
would lead to the view which Agassi called ˜psychologistic individualism™.
© LAWRENCE A. BOLAND
122 Principles of economics Knowledge and institutions in economic theory 123
Step 5. In the future, the succeeding consensus is formed partly as a more important, this view of institutions is inherently static. Once the
result of the existing concrete institutions and partly as a result of the institutions have been established, there can be no real institutional change,
existing social problems, and so forth. hence changes in other endogenous variables cannot be explained within
the given institutional structure. This view™s static nature, combined with
In other words, when Step 4 has been reached, the succeeding generations
its emphasis on power politics, leads its proponents to make political
are taught how to solve their social problems by teaching them about the
mistakes. For example, this view™s proponents often oppose the
existing (concrete) institutions. Of course, the process involves to a great
establishment of an undesirable (concrete) institution because they fear the
extent teaching them what their problems ˜are™. Note that concreteness may
rigidity of its concreteness even though it can usually be shown that a
present other social problems, which in turn are solved by a higher level of
concrete institution (such as a written rule) is easier to change than a
concrete institutions (e.g. an ombudsman). Some societies may wish to
consensus institution (an unwritten rule). 15
prevent any further changes. Others may design their institutions so that
The second way to avoid the problem of ˜circular causation™ is to say
they can be easily altered in order to be able to adapt to changing
that consensus institutions (which underlie any concrete institutions) are
circumstances. Whether a concrete institution actually possesses the
the only real institutions. Moreover, there may be more than one way to
intended durability is an important question of dynamics, but the form of
represent a consensus institution; thus, changes in concrete institutions do
concreteness is still only a consensus institution. In other words, concrete
not imply changes in consensus institutions or social conditions. This
institutions continue to exist only because we allow them to exist. As
alternative has the advantage of avoiding collectivist dogma, but the
individuals, we can choose to ignore them or persuade others to ignore
disadvantage of viewing all social change entirely as a matter of persuasion
them. There may be certain social or personal costs involved in such a
(such as ˜Madison Avenue™ advertising techniques). Of course, with this
stance, but it clearly is an option open to every member of a society.
view, changes in social conditions are very slow whenever communication
Clearly, with this theory the question of social change becomes very
is very controlled (e.g. ˜one should not talk about such things™). But there is
delicate because of the seemingly indeterminate nature of the structural
a more serious methodological problem. It is virtually impossible to know
relationship between problems and solutions at both static and dynamic
when a consensus institution has changed, and thus an operational
levels. The structural relationship at issue is an instance of ˜circular
explanation of social change becomes impossible. Any theory (such as
causation™. Simultaneously, in the process of teaching (or socializing) new
Marshall™s) which explains long-run changes in prices as the consequences
members of a society, the prior existence of an institutionalized solution is
of changes in social conditions (consensus institutions) is inherently
used as evidence of the importance of certain social problems, but the
untestable!
existence of the solution is in turn justified on the basis of the prior
Neoclassical theories of institutional change can be seen to be variants
existence of the social problem. Such a symbiotic relationship may lead to
of the theory represented by Propositions 1 through 5a. But being basically
a very static society if the ˜elders™ are skilled at socializing. It also raises
concerned with the individual decision-maker, every neoclassical theory
certain difficulties with regard to the concept of a change in ˜social
would have to view real changes as those in consensus institutions;
conditions™, including the existing institutions. My presentation of a
however, such changes may (have to) be brought about by changes in
hypothetical sequence which would lead to a concrete institution presumed
concrete institutions. It should be clear that most modern societies provide
the existence of a consensus institution. But, given the symbiotic
specific institutions which make orderly changes or the creation of other
relationship, can the consensus institution be changed without a change in
institutions possible. The legislative bodies of most Western democracies
the concrete institution?
are an example. In fact, the changeability of any institution is a problem for
This methodological problem for the explanation of social change is
which the rigidity of other institutions provides the solution. It should be
usually avoided, but not solved, in one of two ways. The first way to avoid
noted that those institutions whose role is to provide information (such as
˜circular causation™ is to view all concrete institutions (such as the laws that
norms, guidelines and legal limits) are effective only to the extent that they
constrain individual choices) as the only real institutions. Although this
are stable. Thus, the changeability of such institutions compromises their
view has the advantage of being clear-cut and more appealing to common
knowledge role [see further, Newman 1976].
sense, it also has the methodological disadvantage of leading its proponents
The critical issue with any neoclassical variant, as noted earlier, is
to view all matters of social change as matters of only power politics. But
whether a chosen concrete institution is, in fact, a successful representation
© LAWRENCE A. BOLAND
124 Principles of economics Knowledge and institutions in economic theory 125
of a given consensus institution (e.g. whether it adequately represents the response variations but only cover up the failure systematically to explain
them accurately.16
given ideology). Kenneth Arrow™s (im)possibility theorem [1951/63] might
easily be seen as an argument against the possibility of (complete) success In this chapter I have extended this dynamic issue of false knowledge to
in every social situation. Specifically, one cannot guarantee a successful the question of institutions. I have argued that institutions provide essential
social decision mechanism (a concrete institution) which will always knowledge to individual decision-makers. If that institutional knowledge is
represent the society™s welfare function (a consensus institution). false, there is another reason for change. The only difference between
Similarly, there is the critical issue of the adequacy of the solution over institutional knowledge and knowledge in general is that the former (like
which the consensus is formed. Does the given ideology, for example, capital) takes longer to change. In other words, institutional knowledge
solve the social problems that exist? People may think the market system may be durable, and its durability may create problems. Even though an
can solve all social problems, but that does not prove that it can. It is only a institution may successfully represent social knowledge that is true for one
conjecture, the truth of which is neither proven nor provable. For example, period of time, its durability may extend to a period for which it is false.
Arrow [1974] has argued that one essential ingredient for social interaction Thus, since institutional knowledge is durable, it is likely to be false.
(which includes doing business in the market as well as within the firm) is Moreover, the existence of false institutional knowledge is a reason for
simple trust but the existence of a market for trust would be a virtual change and, because change takes time, false knowledge is a continuing
contradiction. reason why the success assumption of neoclassical explanations is often
unrealistic.
In this part I have discussed three widely recognized but allegedly
TIME, KNOWLEDGE AND SUCCESSFUL INSTITUTIONS
neglected elements in neoclassical economics. To the extent that these
The neoclassical programme for explaining the evolution of an economy™s elements are essential, proper consideration of them can surely improve
institutions is quite compatible with my simple theory of the epistemologi- neoclassical explanations. In the next three chapters I will discuss
cal role of institutions. However, once one recognizes that neoclassical additional ways by which new elements might be included. Each of them
programmes (Marshallian or otherwise) presume successful decision- represents a major departure from neoclassical methodology but it will
making and hence, for continuing success over time, that every individual remain an open question whether they represent impossible avenues for the
must possess correct knowledge (which includes accurate representations possible repair of neoclassical theory. I will argue that Keynes clearly
of relevant consensus institutions), it becomes clear that a neoclassical wished to recognize missing elements in Marshall™s economics which
theory is a special case of my version of institutionalism presented here. would make long-run equilibrium explanations rather precarious. And as
That is, in my theory, when the consensus institutions do succeed in always one can find lurking about proponents of the alleged necessity to
accurately representing those solutions, then (and only then) are my theory give neoclassical economics a transfusion of psychology to make it
and a neoclassical theory of institutional change completely compatible. realistic. Into these murky waters I will venture the need to address the
Neoclassical theories are incompatible with my theory whenever any methodology of the individual decision-maker on the grounds which were
individual™s knowledge is not correct (i.e. not true). But, incompatibility is introduced in Chapter 6. Each chapter involves a claim that there is one or
not the important issue here. As has been argued elsewhere [e.g. Hayek more missing elements in every neoclassical explanation.
1937/48; Hicks 1976], the existence of false knowledge is an essential
ingredient in any dynamic theory of economic decision-making. If all
NOTES
knowledge were true (including knowledge about the future), then there
1 This theory of institutions was developed in an undergraduate sociology class
would be no reason for (disequilibrium) change without changes in one or
that I taught in 1968. It was subsequently reported in Boland [1979b] and is
more exogenous givens. If one is going to explain change, the source of the
partially reprinted here by special permission of the copyright holder, the
change cannot be exogenous. Thus, it has been argued, dynamic theories Association for Evolutionary Economics.
must recognize false knowledge (and explain why it might be false). 2 Exogeneity is, of course, defined as the purported intrinsic property of certain
Furthermore, a theory of dynamic behaviour must specify the systematic variables of a model within which they cannot be explained (i.e. they are not
influenced by changes in endogenous or other exogenous variables of the
way each individual responds to the discovery that his or her knowledge is
model).
false. Stochastic theories, their popularity notwithstanding, do not explain
© LAWRENCE A. BOLAND
126 Principles of economics Knowledge and institutions in economic theory 127
3 For a more detailed discussion of the methodological role of exogeneity and the Schumpeter use the term ˜methodological individualism™ to actually mean the
requirements of determinant explanations, see Boland [1989, Chapter 6]. stronger psychologistic individualism. It would be best to reserve the term
4 Consequently, in terms of the logic of solvability, it does not matter whether a ˜institutional individualism™ to indicate the form of individualism that allows
formal constraint is socially given or is a parameter of nature (e.g. available exogenous variables beyond the limits of natural givens and psychological
resources). states of the individual.
5 To avoid circularity, it must be remembered that there still have to be some 15 Similarly, when in power, this view™s proponents waste much time or many
givens which do not endogenously change within or with the generation. resources on superficial changes, that is, on those which change (concrete)
6 For the given values of the exogenous variables, if the current choices of values appearances without altering the underlying consensus.
for the endogenous variables are such that there exist incentives for changes in 16 For a more elaborate discussion of the methodological problems with stochasti-

<<

. 14
( 25 .)



>>