Hayekian Expectations: Theory and Empirical Applications (1993)

CONSTITUTIONAL POLITICAL ECONOMY

it( is likely to lead" (1952: 121). Thus any conscious being lives "as duch in a world of expectation as in a world of 'fact"' (1952: 121).

3. Coherent Expectations

/ In Hayek's theory the semi-permanent map offers a stable framework . And though the model involves the more ustment of cognitive activity to the environ- as constantly adapting to new experiences s, provides an account of cognitive activity onstrained kind. el of the world cannot exist somehow

tly of prior knowledge. If Hayek's theory is correct, then ividual's accumulated stock of knowledge-and by that we mean ts, theories, and interpretations derived from the abstract and les governing cognitive activity-serves to establish boundary ns for the construction of new knowledge. In this way, the f the individual at the cognitive level to dramatically depart

theories is subject to definite, though

ayek's theory exhibits a subtle tension between those mechanisms restrain expectation revision and those tending to move the toward a closer fit with reality. As noted earlier, Hayek

acc unts for knowledge and expectation revision through "reclassifi- cat on." By virtue of the mind's ability to reclassify (i.e., employ a di rent set of rules), new conjectures-literally new theories of reality can be constructed that replace previously held ones. Learning for .; Hayek represents not simply the acquisition of more data points but more importantly the adoption of new (generally tacit) generative rule . Reality is seen differently because a new set of rules have con- stru ted it differently.

hile theories of reality may be consciously selected, the rules of det rmination governing cognitive activity operate, according to Hay ! k, at a supra-conscious (tacit) level. Thus, the opportunity to

effect some kind of fundamental revision in cognitive function- coincide with changing these deep-seated rules of mental

But the further removed such rules are from the conscious activity, the more difficult it is to make such a change.

model may induce certain kinds of changes in the

~ A Y E K I A N EXPECTATIONS: THEORY AND EMPlRlCAL APPLICATIONS

and reclassi- processes, according to

less conscious-rules for clas-

difficult to alter (Hayek 1952: chap. VI). hen we turn to the vast array of sensory data to which the individual

is lexposed, whether by choice or coincidence, another avenue for nstrained change" presents itself. Though Hayek's theory requires the mind construct theories of reality, it is the perception of sensory that figuratively puts the process in motion. The flow of incoming as noted earlier, is subjected to classification by the mind; yet

of reality when the model or map fails to generate the necessary correspondence

the individual expects and what reality appears to be. that expectations are disappointed, a gap exists between

knowledge and actions relative to the environment. individual to generate new conjectures and theories Hayek (1978: 43), what we know and expect are to the mind. Given the structure of the mind and

governing its operation, our knowledge and sensory inputs whose source is external to us. experiences are dependent on external stimuli

Any individual's knowledge and patterned stimuli of its external possible for an individual to set

reality. What the individual environment within which

endogenous to the are endoge-

7 14 other words. Hayek's theory, if true, puts an end to Cartesian dualism.


the influence of family, social norms, conventions, and cultural tradi- tians. Hayek has argued that much of this sort of data is not quantifiable or even collectable. If what we know, given our cognitive apparatus, is contingent on such complex arrays of particulars, much of which we know only tacitly through our daily habits and practices, then knowledge for Hayek has a dimension which is always social and not exclu- sively personal. Because of this broadly sculpted endogenous conception of knowledge, an Hayekian account of knowledge must show it to be rooted in the social order.

4 . Competitive Expectations

Our "knowledge of the external world" Hayek claims, "consists in the action patterns" that "stimuli tend to evoke, or, with special refer- ence to the human mind, that which we call knowledge is primarily a system of rules of action assisted and modified by rules indicating equiva- lences or differences of various combinations of stimuli" (1978: 41). Hayek warns us against thinking of the "formation of abstractions" as "actions of the mind." We should think of the formation of concepts "as something which happens to the mind, or that alters that structure of relationships which we call the mind, and which consists of the system of abstract rules which govern it operation" (1978: 43).

Hlayek's remarks about the formation of abstractions reflect his opin- ion that we should liken the mind to "a system of abstract rules of action" (1 978: 43). And "every appearance of a new rule (or abstraction) constitutes a change in that system, something which," Hayek holds, "its own operations cannot produce but which is brought about by extraneous factors" (1978: 43). Experience induces change in the taxo- nomic rules that constitute the mind. But to what end is this change directed?

For Hayek, the structure and mechanisms of the cognitive apparatus function chiefly as a method to allow the individual to get a better fit with the environment. The generation of knowledge and the actions induced by expectations involve acybernetic relationship between mind and environment. This relationship enables the individual to adapt to changes in the world through an ongoing evolution of knowledge: adaptive modification of the individual's knowledge produces adaptive modi- fications of his actions. Indeed, what an individual "knows" is not so much how things are as what to When action is inconsistent with

8 See "Notes on the Evolution of Systems of Rules of Conduct" in Hayek (1%7).

environment, the individual changes response patterns. He changes at he does. This implies a change in his expectations and his theories

n the Hayekian view of cognition, there is no guarantee that the 's knowledge and actions will correspond adequately to exter- . As noted earlier, Hayek holds that the mind's classificatory nstructs its representation of the external world "in a partial

imperfect manner" (1952: 145). The "model" which is thereby ed provides "only a very distorted reproduction of the relation-

ng in that world" and thus the "classification of these events en prove to be false, that is, give rise to expectations which orne out by events" (1952: 145). This gives rise to "reclassi- which the mind rearranges sensory experiences into new s that allow better predictions to be made about real it^.^

that are "fit" tend to survive while those that are eded out. The criterion of fitness is confirmation of

indicated by the success of one's actions. The selection ng at this individual level has a necessary counterpart

rating at the social level as well. It is the social order which provides individual with the relevant context for appraising his theories.

those persons whose theories are "fitter" are more likely to social success than persons with relatively poor theories.

more consistent with social reality give a competitive the individuals holding them.'"

the mechanisms ensuring this competitive weeding out expectations is an important dimension of economics, such inquiries ordinarily are individually and interactively work and,

life, in the "extended order" (Hayek of the composite actions of separate

to expectation. These separate

9 or Hayek (1952: chap. Vl) this process operates at both the conscious and supra- cons ious (or tacit) levels. Also see Agonito (1975) and Weimer (1982).

10 he selection process occurring at the individual and social levels has a counterpart, accor ing to Hayek, at the inter-social level too. In his controversial theory of group select on, Hayek (1988) argues that those societies with the better rules of conduct will tend t achieve greater success. Consideration of this issue falls outside the scope of this p I per.

1 3 13

CONSTITLITIONAL POLITICAL ECONOMY

,lctions may or may not f i t together well. To understand when we may .:xpect market actions to be coordinated and when not, we need to ,~nderstand how individuals formulate c:xpectations and how those c.:.xpectaitions generate observed patternis of social phenomena. For ,I-iayek, the mechanisms which determin~e expectation formation and III . , \V these expectations generate social patterns are chiefly competitive.

The competitive process among expectations occurs, i n Hayek's \vr-itings, at three levels. There is competition between ideas within the individual, between the ideas of separate individuals with] n society, ; ,r~d bet ween the ideas of societies. At all three levels the logic of >election is the same. The better ideas lead to actions that better pro- r.tole :it~rvival and success. Succ:ess causes the spread of the ideas that spa,m them.

5. Huyekian Expectations

In the preceding discussion we have shc~wn that Hayek provides an a,:count of expectation formation that builcls on his philosophy of mind. The cognitive problem Hayek sets out to resolve is identical to the sc:lcial problem he ha? addressed over the years: how do cornlplex phe- nl.)rr,erla like the mind and markets resolve inherent limitations on k ,-~olwlcdge?

I'n the cognitive domain the individual's construction of a coherent interpretation of reality, both as it is and as it might be, emerges from rL l'es governing the operation of the mind. This interpretation, as Hayek points out, is incomplete and thus the beliaviors it supports will not gt rierate full conformity and con~patibility with the environment. The 6 6 ~.r:stlessness" which motivates action thus has an episten?ological b2 .;i> in Hayek and helps to bring into sharper relief the nec:essity of leiii-ning and acting by trial and error. The Hayekian view of expectation formation emphasizes the individual as a complex phenomenlon func- ticining in a complex social environment. And in both domains the order thiil. emerges is a byproduct of the rules that are followed. Just as so'.:ieties stumble upon or somehow come to rely on consfitutional sy,;lems, so too the mind may be understoocl as a kind of constitutional :syr,;tern.

The main elements of this system and of the theory of expectations i t >8t~stains involve four central points. First, expectations are formed in he context of ignorance about reality. The epistemological uncer- tai~ity in Hayek's theory is not simply (and in some sense, trivially) a

H A Y ICKIAN FXPH-7 A7 I O N S 7 HEORY A N D EMPIRICAL APPI-ICATIONS

result of uni-directional calendar time, a~s emphasized by Shackle and Il,;lchm~ann, but also a result of abstract and inherent constraints on w ~ a t can be known cognitively. We do no't know the world a:; i t "really" is. Rather, the mind constructs a representation of the external world \,vhich is necessarily incomplete. Second, an individual's expectations dcrive from a mental classificatory apparatus. The mind constructs th~:ories of reaiity by using rules of determination to organize and interpret sensory data. Such rules operate at the conscious and tacit levels of cognitive activity. Expectations are formed as a re.sult of rule poverncd creativity. The individual's "nnap" of the worlcl limits the range of "models" he may employ.

'Third, cognitive activity functions as a mechanism of adaptation. .1~ "goodness of fit" criterion permits the individual to appraise the ust:fuln~ess of theories of reality and the expectations they generate. 1:)ir;appointed expectations induce revision. Expectations atre self-cor- ~rt:ctive in that inappropriate expectations are subject to a weeding out I-)rc~cc:ss at the individual and social levels. And fourth, expectations ;ri;.e formed endogenously as a necessary consequence of the role which sensory inputs play in Hayek's cognitive theory. Hayek's theory does 11c.1t casl individuals as cognitive islands who can step outside them- :,cllies epistemologically or divorce therrlselves cognitively from the \r:iaer social reality they not only inhabit but to which they must also adapt. Thus, expectations will ordinarily correspond to the habits, practices, norms, and traditions of the societ:y or, in other words, to the 5ocial rules governing action. For Hayek, these rules are products of sc~cial and even, in the case of the sensory order, biological levolution. I n this evolutionary view there is no reaslon to see in the u,ncertainty cl' the future a special cause for discoordination of actions. Expecta- tll:)ns, on the contrary, have a tendency tow,ard coherence and coordina- t~on . The evolutionary selection processes at work among thoughts and a,::tions tiend to select "fit'' expectations artd to weed out "unfit" ones.

1r1 our discussion thus far, we have attempted only to outline in a h~ghly abstract manner the general principles of an Hayekian theory of expec~.ations. These principles are unconnected to any specific envi- ronrnent in which individuals might find themselves and to which they nl.:ecl to adjust. That the theory of expectations advanced here finds t t ~ c n ~ to be "coherent" and "competitive" applies to the individual level of action but does not take account explicitly of the social out- colnes such behaviors may in fact generate. Specification ofthe environ- mcnt or context of constraint provides the opportunity to examine


thc: implications of an Hayekian theory of expectations when certain .:onclitions prevail. If. as has been argued above, expectations are rules .'if action geared toward an individual's siiccessful adaptation to the ~:nl:ii-onment, then Ihe rules that are selected and employed will bear ;rjrne correspondence to the environment. Hence the selection of rules i:;iltnot be divorced from the context of constraint to which individuals Ir.~ust submit and adapt. In the next two sections of this paper we ~,.xilrnine the implications for the selection of rules by individuals under iliff'erent assumptions about the social context of constraint. Individu- i l l s ' t:xpectations and behaviors are thus situationally endogenized in O I I ~ tr-eatment by virtue of individuals adapting behavior to the existing 'filter-ing conditions" or social rules of thegame. Consequently, market

t 11.lt:c~)mes will be gen~erated that are not only unintended but, depending ( 1 1 these filtering conditions, also may be disorderly and arguably 1 lrldesirh~ble.

11. Filtering Conditions

I . Expc.ctations and the System Constraint

According to Hayek (1978: 41) our knowledge "consists in . . . action p~ttterms'" and it is indeed "primarily a system of rules of action." The f I,.iyek:ian theory of economic expectations we have outlined above is a Ihl:ory about the colgnitive basis of such rules and how such rules (or ,gn,~ipings of them) emerge from the process of social interaction. " 'The rules tending to generate successful actions have a competitive ;~Jvantage in the marketplace and will be "reproduced" in the sense tk;.it the successful actor will be imitated by others and that he will also fc l low the rule again in the future. Individuals embody their expectation:; in rules governing their conduct and will produce interactively, ac,.ortling to Hayek, patterns of social order as a consequence of all in~~li\~itiuals following various habits, practices, norms. and traditions.

I 11 I:; esxntial to note that Hayek argues that a full specification of the rules governing co::ni!i~s functioning and that of any complex phenomena is subject to precise but un ,:)e:~fiable explanatory constraints. As we noted above, for example, the rules deter- n~ilinj: the process of classification at the cognitive level are themselves largely tacit o r ~ I J [ I I - : ~ - C ~ I ~ S C I ~ U S . See Hayck (1952: chap. 8; 1%7: chap. I ) . Similar explanatory con- sfr i ints on rules also apply at the social level. See. for example, Hayek (1967: chap. 3 ; 1') !. <:h21p. 2 ) .

HAYEKIAN EXPECTATIONS: THEORY AND EMPIRICAL APPLICATIONS

In short, Hayek claims that order is a byproduct of rule-governed bdhavior and that the kinds of observed social patterns are reflections of the rules that are followed.

As we have learned from Gilbert Ryle (1949), to "know" something may mean to "know that" or to "know how." Knowledge, in other words, may exist as thoughts or as skills and routines. So too with eqpectations, which are, after all, only a form of knowledge. Following Shackle, we may choose to see expectations primarily as thoughts, as ideas, as visions of the imagination. We prefer, however, to follow Hgyek here and emphasize expectations as rules of conduct, as skills anid routines, as propensities to act. This distinction may be highlighted by noting, for example, that Robert Shiller's interview studies do not subgest that much "rationality" goes into stock trading. And yet tests oflthe efficient market hypothesis either confirm the theory or show only marginal violations of it. This is because the "rationality" of "elxpectations" in the stock market is a product of the rules traders rely oni rules which induce certain propensities, and not traders' superior inslight. Thus, what we call economic expectations may be understood, indeed, primarily as propensities to act."

It matters if economic expectations are thoughts or propensities. If thei thoughts of the mind, as argued by Shackle, are fickle and more or less protean, then economic expectations (as thoughts and not as prdpensities) may be fickle and unreliable as well. Alternatively, if exlpectations are understood as propensities embodied in rules of action, selected for their survival value and adapted by competitive market forces to fit the environment in which they are to operate, we would exdect them to be more stable and reliable. Within the social domain, the observed social structures and the patterns of results generated by the social order are attributable to the adaptation of individual rules of corlduct to the wider social realm and also to the coordination achieved ambng its members because all followed explicit and implicit social rulds." The order we observe in the social realm is in Hayekian terms priqcipally a byproduct of rule-following. Thus, it seems reasonable to suggest that the kinds of order or patterns observed in society (or more

12 Also see Gray (1984: 14f.) and Kukathas (1990: 52-54). 13 Rules operating at the cognitive and social levels are not claimed by Hayek to

be "fixed." Rather, Hayek's cognitive and social theory is fully within the area of "evdlutionary epistemology," as developed by Popper, Bartley, Campbell, and others.

CONSTI'I'UTIONAL POLITI('AL< ECONOMY

c,pc:cifically in the catallaxy) will reflect the kinds of rules governing l h c t rnarket activities of individuals." 11-1 short, the context of con- '.. trsint--here, the rules governing market activities-carries implica- t i ox for the functioning of the market.

'This question has received attention by Langlois (1986) ar~d Langlois ;11:1d Koppl (1991). They develop a way to examine market behaviors l l v employing a "system constraint." The system constraint is the clegree lo which "single-exit" models are appropriate and a measure ol'the latitude that (ideal-typical) actors in the market may deviate from optimality. The tight system constraint of competitive markets imposes 1 osts upon actors who deviate from the discipline of the market. Devia- tior~s an: punished with losses and, for the recalcitrant, exit from the riial-ket. The looser the system constrain:t, the more likely i t is that (itieal-typical) actor:; will behave in idio.syncratic and unpredictable

i.l)'S.

When actors are s,ubject to a tight system constraint, their actions n11.1:,t manifest the sorts of propensities that prevent losses and with- dr.;iwal from the market. Alternatively. when actors are subject to a loose system constraint. the evolutionary selection process is weak. ~ C - I particular set of rules or propensities is selected by the market pi-c.)c:ess; individuals may in such cases employ rules that foster "any- [!,in;: goes" sorts of actions and which may in the aggregate generate jc)ci,al outcomes that appear to follow no obvious pattern.

I n extireme cases of a tight system constraint, the neoclassical a!s,umption of "as-if maximizing" is appropriate. In such cases, some vt',rslon of the "perfect knowledge" assumption of neoclassical theory m..ty also be usefully e:mployed. But even if the assumptions of rational m,.lxlmizing and perf'ect knowledge are sometimes appropriate, we sh(:luld not infer that market traders are perfectly informed or rationally

1 .4 sc)ci..il spontaneous order, according to Hayek, will exhibit patterns or rregularities dei rvativc from the order's constituent elenients obeying certain general rules of conduct. Set., for t:xample. Hayek (1973: chap. 2). Although such rules are necessarily finite, the di\.:l-sty which the rules allow is not: as a result, the system, though it is constrained bv .-l:r:ain rules, has the potential to produce novel results. The evolution of !.he system i\ described ,as rule-bounded creativity. As we have suggested here, the mind i:; appropri- atelv understood as a hponlaneous order in precisely the same sense as is the market ord1.r. That Hayek's conception of social spontaneol.ls order shares similarities with lanl:~~a~;e and especially with science as rule-bound creative processes is an area that fallt hcyond the scope of this paper.

zing. Assumptions such as perfect foresight and rational maxim- entify a property of markets and their institutional structures. why it may be sensible, as Koppl (1991) suggests, to look for ns under which rationality is more likely and conditions under

not that people are less rational under so under others. It is that in certain

nces the economic context or system is more "rational" in f the applicability of maximizing models and in other circum-

a loose system constraint, the radical subjectivist ctability is appropriate. In such cases, the actor's ded periods bear no systematic relation to the t and exhibit no tendency to consistency or

in authority, who hear voices in the air," may ed be unwitting slaves to the ideas of past thinkers (Keynes 1936:

But no one knows what they will do with those ideas. constraint tight or loose? For evolution to k changes in the environment with the sur- iduals comprising the social order. These

the process by which knowledge is pro- rding to Campbell (1974), adaptation to lection processes and mechanisms that n and processes that preserve or propa-

lutionary pattern of social orders will ing tendencies of innovation and tradi- e character of that interaction and the

I CONSTITUTIONAL POLITICAL ECONOMY

s of outcomes thereby generated hinge on the framework of rules action. Here, we wish to highlight two elements integral to

of the system constraint and to the efficient operation of adoption of rules of action: stable constitutional rules and atomistic competition.

~ 2. Stable Constitutional Rules

"constitutional constraints" in Buchanan's (1987; 1991), change too often and by too much, then the general

environment within which groups of rules compete will lack necessary to select rules that induce actions consistent

scarcities.16 Our Hayekian perspective, then, gives theme taken up by the proponents of constitutional

stability in the rules of market competition. Hayek stability of constitutional rules. He gave it

cia1 real estate markets. This expansion, however, came to an abrupt halt when

1980's, it is also important to recognize that frequent overhauling of the tax

as 20 CC

TI

t h a c tic trl Hi W(

SO'

co rul co. art

is mc is, ha1

1

sec off;

I ' issu disc fron follc nize men

I t of pi betv the I

.YEKIAN EXPECTATIONS: THEORY AND EMPIRICAL APPLICATIONS

condition of "true laws that they be known and certain" (1960: Without such stability, he argued, reliable expectations about the

equences of our actions could not be formed (Hayek 1973: 106-1 lo)." e is no rule of law if the laws change often and dramatically. ie "miscarriage of the democratic ideal" against which Hayek rails, :laim that democratic governments serve "the several interests of ~glomerate of numerous groups," led him to propose as a constitu- I reform the creation of a nomorherae, a legislative body cons- ed to "stating the general rules ofjust conduct" (Hayek 1973: 98). ~roposal was an attempt to find a meta-constitutional constraint that d induce stability in the other constitutional constraints governing I cooperation.

3. Atomistic ~ o m ~ e t i t i o n ' ~

he market includes actors who are more or less immune from the ,etitive pressures of profit and loss, then the natural selection of will be inoperative with regard to these actors and the system raint facing them will be loose. In such cases the privileged actors .ee to act idiosyncratically. e obvious case for such a violation of the atomicity assumption mopoly. But the freedom of the monopolist is relative. If the ~polist fears potential competition, he may behave himself, that t consistently and without guile or opportunism. If, on the other his monopoly position is secure, he is more likely to be capricious.

:oa is an example of the first type of monopolist. In spite of having ed over 90% of the virgin ingot market, the company consistently :d high quality at low prices. In the famous anti-trust case involving

he dynamic inconsistency of optimal plans when rules are not followed is an f particular relevance to the actions of policymakers. If, for example, the Fed is onary, monetary policy that attempts to insulate the economy in the short run togenous disturbances will be destabilizing in the long run. Requiring the Fed to a known rule will avoid this time inconsistency problem. Friedman (1%2) recog- his problem, though Kydland and Prescott (1977) provide a more complete treat- llso see Klein (1990). ur conception of "atomistic competition" bears no particular relation to the model competition or to those approaches that seek to establish a clear correspondence n numbers of putative competitors and "economic performance." We endorse ion of competition as a rivalrous discovery process along the lines developed by ( 1948) and Kirzner (1973).

1 CONSTITUTIONAL POLITICAL ECONOMY

, Judge Learned Hand argued that Alcoa's consistency in antici- m in a timely and economical

al pieces of evidence that Alcoa ted to maintain a monopoly and thereby violated United States

loser to the second type of monopoly is the case offranchise bidding. f Oliver Williamson (1976) and others have shown that the ng scheme of Demsetz does not create the competitive ined because of the problem of incomplete contracts.

inner gets a contract. But the contract can and will renegotiated. These two facts may combine to give the franchise

unity from competitive pressures. Besides the obvi- rtunism with guile" created by such market power, on of the evolutionary selection of rules of action. ns in such a market are the dubious thoughts and le individuals, not propensities to act culled by

is only in the latter case that we have a real violation of the condition oes the fewness of numbers spell hich some privileged actor(s) has

er has been investigated by Koppl (1991), Koppl and Langlois 4), and Koppl and Yeager (1992) under the heading of the theory

idered in the next section.

I 111. Big Players

that in markets with a tight system constraint, there e in which individual expectations do not matter. Whatever hts associated with the propensities of market transactors, the ofit and loss acts to ensure that market prices and quantities

appropriately to changes in relative scarcities. However, it is e when the system constraint is loose.

e system constraint is loose because of a failure of atomism, one has, in the short run at least, power. And the decision discretion in the use of that power may disrupt the process selection that occurs in markets. Koppl (1990, 1992) and Yeager (1992) have developed a theory of the "Big Player" rates some of the ways expectations matter when the system

Big Player is "anyone who influences the results of the market ess by exercising his discretionary judgment while remaining him- wholly or largely immune from the discipline of profit and loss"

and Yeager 1992: 5). Finance Ministers and central bankers are pica1 Big Players. To be a Big Player it is not enough to be big.

e or less immune from the discipline of profit and loss must be actively engaged in discretionary decision making that es the results of the market process. To be a Big Player, you

st have power and use it. When markets are subject to the discretion- interventions of a Big Player, the filter of profit and loss no longer

with the same efficiency. As Koppl and Yeager point out, the nary actions of Big Players, such as discretionary government ions, "focus[es] entrepreneurial attention on the politics of the " But, "these are typically less knowable than more narrowly

es." Thus the presence of a Big Players "reduces ability of markets and the reliability of expectations" and

skill and makes luck count for more" (Koppl and Yeager

a Big Player is active, habits and routines that induce actions to underlying scarcities are less likely to generate profits s less likely to survive. Habits and routines that induce nsistent with underlying scarcities are less likely to bring re thus less likely to be weeded out by natural selection. rvive. When the filter of profit and loss has been weakened . nce of a Big Player, the results of the market process predictable. These results come to depend less and less tional givens-that is, the norms or rules ordinarily fol- nderlying scarcities. The results, instead, come to depend ictable thoughts and actions of individuals.

ection operates in a thick market without the intru- ayer, we do not need to know much about the expecta- rket transactors. Theories such as the efficient markets apply. When the process of natural selection is dis- layer, then we may be able to improve our understand- ements by studying the particular rules, "reactions

eories of market transactors. This, of course, is ert Shiller (1989) has begun to do.I9


theory sheds light on current controversies in the f asset prices. Standard economic theory predicts that the price et tends to equal its present value. This insight has led some s of "efficient markets" to predict that the prices of assets thick markets will always equal the best available calculation t value and that, in consequence, the asset's returns will rtingale.20 Other economists (e.g., DeLong, er.al. 1990) have t even such markets as the New York Stock Exchange are "irrational" actors called "noise traders" (Black 1986; Summers 1990)) who induce movements away from present

. These movements cannot be modeled as a martingale process. ith noise traders, prices may be subject to fads and band-

(Topol 1991; Scharfstein and Stein 1990). re on asset prices includes many empirical studies com-

t market and noise trader models. These tests have ous results (Cochrane 1991; LeRoy 1989; West 1988;

he Big Players theory provides a somewhat new the theory and empirics of asset prices. The theory egree of noise trader herding in an asset market is a f the degree of Big Player influence (Koppl 1990; and Yeager 1992). When Big Players are not much rices should behave in ways close to the predictions

of e cient market models. When Big Players flex their muscles, there may be (irrational) "bubbles" in asset prices.

U ing Heiner's theory of reliable behavioral routines (Heiner 1983, 1986 1 1990), Koppl and Yeager argue that under Big Player influence

~ Conclusion

asset Ttey

to

thr/

duce do.

prices will be subject to herding and thus tend to create bubbles. argue that the herding induced by Big Players should create the

irregular waves Mandelbrot (1972) calls "aperiodic cycles." They used rescaled range analysis, a technique of time-series analysis appropriate

the study of long-memory processes, to test for these aperiodic ~yc les .~ ' Their study of the international value of the Russian ruble in

19th century suggest that Big Players increase the tendency of the acctumulated series of asset-price changes to deviate from trend. Such deviations may be interpreted as speculative bubbles. Big Players intro-

a kind of wedge between what people do and what they should In this sense, Big Players introduce a wedge between epistemic

knowledge and social reality.

21 or descriptions of this technique see Greene and Fielitz (1977). Lo (1991). Mandel- brot (1972). f Mandelbrot and Wallis (1968, I%%, 196%) and Wallis and Matalas (1970).

In

tations actim in draws The thus "Big suck we that ensue

Tt cal in th

filtering

desc

this paper we have outlined a theory of expectations based on Hayek's theories of mind and market. According to this theory, expec-

are the product of evolutionary selection that favor rules of adapted to survival. Such expectations are primarily embodied

"knowing how" more than "knowing that." This evolutionary stance our attention to the filtering conditions of stability and atomicity.

failure of either condition creates a loose "system constraint" and a loose link between environment and expectation. The theory of Players" gives us an example of the empirical consequences of a failure. The Big Player theory shows that the Hayekian theory

lave proposed is testable. It has, moreover, policy implications suggest the desirability of designing constitutional constraints that

the atomicity of the market process and the stability of the rules governing that process.

e Big Player theory suggests a possible direction for future empiri- work in an Hayekian tradition. If the perspective we have adopted

s paper is warranted, then "neoclassical" economics and "radical Keynesianism" may be seen in a sense as limiting cases. When certain

conditions-certain institutions and constitutional constraints-are in place, the results of the market process may be

-ibed by "neoclassical" models of rational maximizing; action will

1 CONSTlTUTIONAL POLITICAL ECONOMY

redictable to the observing economist. To the extend that these ns fail, the market process will be influenced by animal spirits,

and the like; action will seem unpredictable to the observing

are "Keynesian." The evolutionary logic of an Hayekian econom- rovides the general theoretical foundations for the special empirical

such as the Big Player theory, that identify the precise causes equences of deviation from the neoclassical model. Finding rical limits to the application of neoclassical models is an research project. Similarly, finding the observable constitu-

ditions that tend to induce "kaleidic" or "Keynesian" behav-

the results of the market process. While this may seem a ult, economists have often neglected an important dimension age. It is widely understood that different constitutional

aints create different incentives and thus different patterns of hts a different line of

ction mechanisms and thus different mechanisms for the nd distribution of knowledge in society. There are episte- ences to the choice of constraints. We believe this is an

underdeveloped part of the research program of constitu-

REF RENCES P Agon

3: Baddy, Black, Boett Buchz

Buchqnan,

Buchanan, A r b ~ :

Burrel and

to, R. (1975) "Hayek Revisited: Mind as a Process of Classification." Behaviorism 62-171.

W. W. (1984) The Retreat to Commitment. LaSalle: Open Court. 2nd. ed. F. (1986) "Noise." The Journal of Finance 41: 529-543.

te, P. J . (1993) Why Peristroika Failed. New York and London: Routledge. nan, J. M. (1975) The Limits 0fLiberty:BetweenAnarchy and Leviathan. Chicago:

University of Chicago Press. J. M. (1987) "Constitutional Economics." In: Eatwell, J.; Milgate, M.; and

Nevvman, P. (eds) The New Palgrave. London: Macmillan. J. M. (1991) The Economics and the Ethics of Constitutional Order. Ann

The University of Michigan Press. , D. B. (1%7) "Obeying Rules and Following Instructions." In: Crosson, F.J., Sayre, K.M. (eds) Philosophy and Cybernetics. New York: Simon and Schuster.

rnal of Monetary Economics 27: 463-485. ng, J. B. . et a/ . (1990) "Noise Trader Risk in Financial Markets." Journal of irical Economy 98: 703 -748.


Optimal Plans." Journal of Polirical Economy 85: 473-491.

d) Economics as a Process: Essays in rhe New Instirulional Economics. New York:

onomic Perspectives 4: 117-134.

, D. S., and Stein, J . C. (1990) "Herd Behavior and Investment." American c Revieu~ 80: 465479.

Hayekian Expectations: Theory and Empirical Applications (1993)

Documents