Caldwell: Hayek and Socialism Hayek and Socialism

Journal of Economic LiteratureVol. XXXV (December 1997), pp. 1856−1890

Caldwell: Hayek and Socialism

Hayek and Socialism

BRUCE CALDWELLUniversity of North Carolina at Greensboro

I have received many helpful comments on this paper from participants at the History ofEconomics Society meetings at the University of British Columbia, at seminars given at theUniversity of Georgia, at the Austrian Colloquium at New York University, at the York Uni-versity-University of Toronto Workshop on the History of Economics, and at the Duke Univer-sity History of Political Economy Workshop. Extensive comments were also received fromPeter Boettke, William Butos, Neil De Marchi, Greg Ransom, and George Selgin. All remain-ing errors are the responsibility of the author.

I. Introduction

MOST ECONOMISTS know thatFriedrich A. Hayek was a life-long

opponent of socialism. But who were hisopponents, what were his arguments,and how did he come to develop them?

In the 1930s Hayek attacked the eco-nomic feasibility of socialism, drawingon arguments from an earlier German-language debate and taking to task anumber of separate proposals for a so-cialist society. This drew a responsefrom Oskar Lange, who advocated mar-ket socialism. The ensuing battle withLange and others led Hayek to de-velop a distinctive “knowledge-based”critique of socialism.

Just before the onset of World War IIHayek began work on a series of articleswhose ultimate result was the publica-tion in 1944 of The Road to Serfdom,his most famous (and in some quarters,notorious) book. The Road to Serfdomcontains Hayek’s political critique of so-cialism, but also the seeds of more posi-tive work. The fruit was his descriptionand defense of an alternative liberal

utopia in such works as The Constitu-tion of Liberty (1960) and Law, Legisla-tion and Liberty (1973−79).

During the war years Hayek also be-came fascinated with questions of meth-odology. This led him, apparently in-congruously, to publish a book on thefoundations of psychology. The SensoryOrder (1952a) is, particularly amongeconomists, Hayek’s least appreciatedbook, yet in a letter to an economistprior to its publication he described itas “the most important thing I have yetdone . . .” (letter to John Nef, Nov. 6,1948). The book provides a theoreticalbasis for the “limitations of knowledge”theme that has been recurrent inHayek’s work. This claim in turn im-plies limits on the ambitions of socialistplanners and other “rationalist con-structivists,” and as such constitutes an-other set of arguments against social-ism. It also implies limits on theambitions of economists, and from anAustrian viewpoint, may help to explainwhy the models of economists have sooften misled them about the prospectsfor socialism.

1856

The paper attempts to answer thequestions provided at the outset, to sur-vey Hayek’s various arguments againstsocialism and to provide some historicalbackground on why he developed themwhen he did.

Recent interpretive literature pro-vides the impetus for the emphasis onhistorical context. Since the events of1989, the Soviet central planning modelhas largely been abandoned by aca-demic advocates of socialism, and a re-newed interest in market socialism hastaken its place. Recent discussions ofmarket socialism by economic theorists,both proponents and opponents, typi-cally draw on insights derived from the“economics of information.” This newperspective views general equilibriumapproaches as inadequate, if not mis-leading, for understanding economicphenomena. Because many of the ear-lier debates on market socialism uti-lized a general equilibrium framework,they are from the perspective of theeconomics of information otiose.

A new interpretation of the debatesover socialism, one that highlights thecontributions of the economics of infor-mation, has begun to emerge (PranabBardhan and John Roemer, eds. 1993,pp. 3−9; Joseph Stiglitz 1994). These in-terpretations are designed to providecontext for recent work, but they typi-cally run into difficulties when they tryto characterize Hayek’s contribution.Though it is evident that Hayek (likethe information theorists) was critical ofusing static general equilibrium modelsfor assessing the merits and limitationsof socialism, it seems clear (and this de-spite his frequent invocation of the con-cept of “knowledge”) that Hayek didnot directly participate in the develop-ment of the economics of information.

How, then, to characterize Hayek’scontribution? From a modern perspec-tive, his work seems “fuzzy” (Louis

Makowski and Joseph Ostroy, in Bard-han and Roemer, eds. 1993, p. 82).Many may well be tempted to concludewith Jànos Kornai (in Bardhan and Roe-mer 1993, p. 63) that “the warnings of aMises or a Hayek about market social-ism” should be viewed as “brilliantguesses” rather than as anything resem-bling “scientific propositions,” or withRobert Heilbroner (1990, p. 1098) that“the successes of the farsighted seemaccounted for more by their prescient‘visions’ than by their superior analy-ses.”

In the final section an alternative,more positive interpretation of Hayek’scontribution is offered. As he becameever more deeply entangled in the de-bates over socialism, Hayek decidedthat a more integrative approach to thestudy of complex social phenomena wasnecessary, that standard economicanalysis taken alone might itself be in-adequate, if not misleading, for under-standing the problems of socialism. Tobe sure, his decision to branch out wasbased in part on his recognition of thelimitations of the static “equilibriumtheory” of his day. But his emphases onthe market as a discovery process, hisconcern with the limits of human cogni-tion, and above all his fascination withquestions of knowledge, what John Gray(1986, p. 134) has aptly called his “epis-temological turn,” led Hayek to insightsthat may well be different in kind fromthose advanced by proponents of theeconomics of information.

If this alternative account is correct,it means at a minimum that those whowould try to read history backwards,who would assess Hayek’s contributionsolely in terms of whether and how heanticipated the later literature on theeconomics of information, will perforcemisunderstand his arguments againstsocialism. So one point of the final sec-tion is to plea for a less ahistorical his-

Caldwell: Hayek and Socialism 1857

tory. But it may also be the case thatHayek actually provided an indepen-dent set of claims against market social-ism, one that has been missed by mod-ern information theorists. If so, thenHayek’s apparently “fuzzy” argumentsmay well prove to be of more than“merely historical” interest today.

II. Hayek and the Market Socialists1

Friedrich August von Hayek was bornin Vienna on May 8, 1899. Followingwar service he entered the University ofVienna, where he completed degrees in1921 and 1923. After 14 months ofstudy in the U.S., Hayek returned toAustria in 1925. He spent the rest ofthe decade studying monetary historyand developing a theory of the trade cy-cle. Hayek’s life and the academic mi-lieu of Vienna in the 1920’s is docu-mented in Hayek (1984, Introduction;1992, Prologue; 1994, pp. 47−72) andEarlene Craver (1986).

In the spring of 1931 Hayek was in-vited by Lionel Robbins to deliver a se-ries of lectures on the trade cycle at theLondon School of Economics (LSE).The next year he was appointed to theTooke Chair of Economic Science andStatistics at the LSE, a position hewould hold until he moved to theUnited States in 1950, where he ulti-mately accepted a position on the Com-mittee on Social Thought at the Univer-sity of Chicago. During the early 1930sHayek had exchanges with John May-nard Keynes and Piero Sraffa on mone-tary and trade cycle theory and withFrank Knight on capital theory (Hayek1994, pp. 75−98; 1995). His debateswith the socialists began in 1935 withthe publication of Collectivist EconomicPlanning: Critical Studies on the Possi-bilities of Socialism. The book con-

tained translations of four articles withintroductory and concluding essays pro-vided by Hayek as editor.

A. Prelude—Mises and the German Language Debates

The most important of the translatedcontributions was an article by Ludwigvon Mises that had originally appearedin 1920. The collapse of the Germanand Austro-Hungarian Empires at theend of the First World War opened thedoor for a variety of socialist proposalsfor reorganizing society. Prior to thewar Mises was known as a monetarytheorist and for his unyielding devotionto liberalism. He was as good a personas any to challenge the socialists; in sodoing he initiated the German-languagesocialist calculation debate. DavidSteele (1992, ch. 4) identifies some ofMises’ predecessors, and a number ofwriters (Trygve Hoff, 1949; JudithMerkle 1980, ch. 6; Don Lavoie 1985;Günther Chaloupek 1990) examine thedebates as well as the specifics of vari-ous proposals.

The proposal that most provokedMises was made by the sociologist andphilosopher Otto Neurath. Neurath isremembered today as a member of theVienna Circle of logical positivists andas the inventor of ISOTYPE, the Inter-national System of Typographical Pic-ture Education. Before World War Ihe began to make a reputation as a pro-ponent of a new academic subfield,“war economy.” He also participated,along with Mises, in Eugen von Böhm-Bawerk’s famous economics seminar.Among the other seminar participantswere Joseph Schumpeter, Otto Bauer,who would lead the Austrian SocialistDemocratic party in the 1920s, andRudolf Hilferding, one of the leadingMarxian theoreticians of the 20th cen-tury.

According to Neurath, during peace-1 Parts of this section are adapted from the in-

troduction I wrote as editor for Hayek (1997).

1858 Journal of Economic Literature, Vol. XXXV (December 1997)

time, production in market economiesis driven by the search for profits, butthis leads to recurrent periods of over-production and unemployment. In war-time, by contrast, production is nolonger driven by profit-seeking, and thewar effort ensures that productive ca-pacity is always fully utilized. Anothercharacteristic of the war economy is thesuppression of the price system, whichis replaced by extensive planning of ma-terials management from the center.This is all to the good, because for Neu-rath the monetary system, the searchfor profits, and the disorderliness ofcapitalist production all go hand inhand.

Neurath argued that the central plan-ning that emerges within war econo-mies should continue in peacetime. Heproposed that a “natural accountingcenter” be set up to run the economy asif it were one giant enterprise. Mostcontroversially, he insisted that moneywould be unnecessary in the newplanned order: because productionwould be driven by objectively deter-mined needs rather than by the searchfor profits, all calculation regarding theappropriate levels of inputs and outputcould be handled in “natural” physicalterms. In Neurath’s opinion, attemptsto employ monetary calculations withina planned society would render impossi-ble scientific economic management,which had to be conducted in terms of“real” physical quantities.2

Mises (1978, pp. 39−41, 133) plainly

had a strong negative reaction to Neu-rath personally, and as a monetary theo-rist he found his economic claims fan-tastic. (As Chaloupek 1990, pp. 668−70,has shown, most socialists also soon re-jected Neurath’s proposals for a money-less economy.) But he wanted to make amore general case against socialism,one that could be used against otherproposals as well, including those thatretained a role for some form of money.Mises took as a starting premise thatunder most forms of socialism “produc-tion-goods” (factors of production) areowned by the state, and that as suchthere is no market for them. But thisbasic feature of socialism has substan-tial consequences:

because no production-good will ever becomethe object of exchange, it will be impossibleto determine its monetary value. Moneycould never fill in a socialist state the role itfills in a competitive society in determiningthe value of production-goods. Calculation interms of money will here be impossible.(Mises, in Hayek, ed. 1935, p. 92)

Even if money is retained, in the so-cialist state no prices for factors of pro-duction exist. As such, socialist manag-ers have no way to tell when choosingamong a huge array of technologicallyfeasible input combinations which areeconomically feasible. Without someknowledge of relative scarcities, theyare left “groping in the dark.” As Misesput it: “Where there is no free market,there is no pricing mechanism; withouta pricing mechanism, there is no eco-nomic calculation” (in Hayek, ed. 1935,p. 111).

B. Hayek’s Initial Arguments

In his introductory chapter Hayek re-counted the debates that had takenplace in the German language literaturea decade before. In his conclusion heturned to the current scene.

English socialism in the 1930s was a

2 For more on his views, see Neurath (1973, ch.5), Nancy Cartwright et al. (1996, ch. 1), and Ag-nes Miklòs-Illès (1996). The resurgence of interestin Neurath among philosophers is chiefly due tohis pluralistic and anti-foundationalist approach tounified science. Austrian economists focused onhis physicalism, his insistence that, to be meaning-ful, scientific terms must refer to observable phe-nomena. The Austrian economists’ opposition topositivism and scientism in the social sciences de-rives in part from their arguments with such oppo-nents.


mixed bag. The Fabians, whose leadersincluded Sidney and Beatrice Webb andGeorge Bernard Shaw, had been argu-ing for an evolutionary style of socialismsince the late 1880s. The British LabourParty, formed in 1906, officially en-dorsed socialism in its platform. Labourhad prevailed in the general election of1929, but had faltered with Britain’sabandoning of gold in 1931 and spentthe remainder of the decade regroup-ing. Though interest in Guild Socialism,a form of syndicalism, had declined fol-lowing World War I, academic advo-cates like R. H. Tawney and G. D. C.Cole were still active. Barbara Woottonwas the Director of Tutorial Studies atthe University of London, and underher influence a “tutorial version of his-tory,” one which emphasized the dete-rioration of the position of the workingclass under capitalism, was taught inadult education courses throughoutBritain. Maurice Dobb was the leadingspokesman among economists for Marx-ism, and a number of prominent naturalscientists favored communism. Finally,there existed a nascent interest in mar-ket socialism, dubbed “pseudo-competi-tion” by Hayek.3 With no idea which ofhis many opponents might respond,Hayek offered a diversity of criticismsof socialism.

He began with a review of “the Rus-sian experiment,” basing his criticismson the writings of a Russian émigréBoris Brutzkus. This would serve as acounterweight to a more appreciativeassessment of the experiment offeredby the Webbs, whose massive study (itfilled two volumes) was titled SovietCommunism: A New Civilization? when

it was published in 1935. With incred-ibly bad timing, they chose to drop thequestion mark in the 1937 edition.

Next Hayek took up the argument ofHenry D. Dickinson (1933), whoclaimed that Mises was wrong, that ra-tional calculation under socialism was atleast theoretically possible. Because anyeconomy could be formally representedby a Walrasian system of equations,Dickinson claimed that on a theoreticallevel there is no difference betweencapitalism and socialism: In a capitalistsystem the equations are “solved” bythe market, whereas in a socialist sys-tem they could be solved by the plan-ning authorities.

In his rebuttal, Hayek enumeratedmany difficulties associated with “themathematical solution,” or any regimethat relied on formulating and solvinga giant system of equations for therelevant prices and quantities. Hementioned the staggering amount ofinformation that would need to begathered; the immense difficulty of for-mulating the correct system of equa-tions; the hundreds of thousands ofequations that would then need to besolved, not just once but repeatedly;and the inability of such a system toadapt to change.

Should socialist authorities decide to“solve” the system using a trial and er-ror method, an approach also men-tioned by Dickinson, other problemswould arise. The most important ofthese is the inability of any price-chang-ing mechanism to replicate the auto-matic adjustments that occur in a com-petitive free market system in responseto underlying changes in supply and de-mand:

Almost every change of any single pricewould make changes of hundreds of otherprices necessary and most of these otherchanges would by no means be proportionalbut would be affected by the different de-

3 See Elizabeth Durbin (1985) on British social-ism in the interwar years, Gary Werskey (1978) onsocialism among the natural scientists, and the es-says in Philip Bean and David Whynes, eds. (1986)on Wootton. Hayek, ed. (1954) may be read as aresponse to the tutorial version of history.


grees of elasticity of demand, by the possibili-ties of substitution and other changes in themethod of production. (Hayek, ed. 1935, p.214)

Next Hayek took up the arguments ofthe Marxist Dobb (1933), who notedthat if consumption decisions were alsosubjected to central control, most of theproblems associated with central plan-ning would be alleviated. Hayekpointed out that the abrogation of con-sumer sovereignty implied in such anapproach would presumably be repel-lant to most Britons, and that even un-der such a regime, prices to help guideproduction would still be necessary.

In the last half of his chapter Hayekexplicitly engaged market socialism. Be-cause no concrete proposals were yet onthe table, he had to imagine the formsof market organization that his oppo-nents might propose. One possible ar-rangement is for managers of monopo-lized industries to be directed toproduce so that prices covered marginalcosts, thereby duplicating the results ofcompetitive equilibrium. Hayek’s mostoriginal argument here is that in thereal world (as opposed to the staticworld of perfect competition models) itis typically difficult to know exactlywhat “true” marginal costs are (Hayek,ed. 1935, pp. 226−31). In a market so-cialist regime in which firms within anindustry do compete, a different prob-lem arises: in decisions concerning capi-tal allocation, central planners wouldhave to take over the role played bythousands of entrepreneurs in a marketsystem (Hayek, ed. 1935, pp. 233−37).Hayek clearly considered this to be adisadvantage, but did not specify thenature of the problem. Finally, henoted that the absence of private own-ership in the means of production cre-ates incentive problems for managers,who will put off making difficult deci-sions and who will tend toward risk-

aversion in making investment decisions(Hayek, ed. 1935, pp. 235, 237).

Though Hayek’s two early pieces con-tain fleeting glimpses of his mature po-sition, the modern reader must work tofind them. As Israel Kirzner has con-vincingly argued, the socialist calcula-tion debate served as a

catalyst in the development and articulationof the modern Austrian view of the market asa competitive-entrepreneurial process of dis-covery. . . . it was through the give-and-takeof this debate that the Austrians gradually re-fined their understanding of their own posi-tion. (Kirzner 1988, p. 1; cf. Lavoie 1985)

C. Lange’s Rebuttal

Market socialists are critics of capi-talism, but they also acknowledge thatunder certain conditions perfectly com-petitive markets have desirable effici-ency characteristics. An essential prem-ise of market socialism is the denial thatmarket structures under late capitalismresemble, in any meaningful way, per-fect competition. According to this viewfew competitive industries exist any-more, having been replaced by corpo-rate giants, cartels, and monopolies.Contemporary capitalism thus lacks thebeneficial efficiency characteristics ofcompetition, while retaining all of itsdefects. With careful planning marketsocialism can replicate the benefits oftruly competitive markets, correct forremaining problems regarding effi-ciency, and all the while avoid capital-ism’s pernicious distributional effects.The chief spokesman for this view inthe later 1930s was a Polish émigré toAmerica, Oskar Lange, whose two-partpaper appeared in the Review of Eco-nomic Studies in 1936−37, and was soonreprinted in a book of the same title,On the Economic Theory of Socialism(1938).

Lange’s first argument was directedagainst Mises. Lange agreed with Mises


that prices are necessary for rationalcalculation. Mises’ mistake was to thinkthat prices must be formed in markets.If one instead understands that the cor-rect definition of prices is “terms onwhich alternatives are offered,” and thattheir determination in markets is notessential but rather a peculiarity of aparticular institutional arrangement(capitalism), then Mises’ argument col-lapses. Accounting prices could be sup-plied by the Central Planning Board,and these could be taken by socialistmanagers as parameters in their deci-sion making. Rational calculation undersocialism is not “impossible” after all(Lange 1938, p. 60−62).

Lange’s next step was to demonstratehow a socialist commonwealth could bemade to yield the same results as a truecompetitive market system. In hismodel there exists a free market forboth consumer goods and labor, but(because of public ownership of themeans of production) no market fornon-labor productive resources likecapital. Because labor incomes wouldstill be market-determined, income in-equality would not be eliminated. Butbecause capital ownership is a principalsource of income disparities, its elimi-nation would serve to reduce inequality.Individual incomes would also be sup-plemented by receipt of some share ofthe “social dividend,” the income thathad previously gone to owners of capi-tal.

The sticking point for this variantof market socialism is the absence ofprofit maximizing firms and of a market(and hence of prices that reflect rela-tive scarcities) for non-labor productiveresources. Lange proposed that theCentral Planning Board provide provi-sional “prices” for all goods and factorsof production. Managers of socialistfirms would be instructed to choose, onthe basis of these “given” prices, the

combination of inputs that minimizedtheir costs and the level of outputthat maximized profits. Planners incharge of industries would likewise ex-pand or contract them as necessary,thereby replicating the beneficial ef-fects of free entry and exit under com-petition.

Lange’s proposal begged a key ques-tion: What if the Central PlanningBoard fails to choose prices that accu-rately reflect underlying relative scarci-ties? Here Lange suggested that plan-ners follow a “trial and error”procedure, one similar to that used inactual markets, adjusting prices up ordown in any factor or product marketsin which gluts or shortages existed.Through the trial and error method the“right” set of accounting prices will ulti-mately be found (Lange 1938, pp. 86−89).

Lange responded to Hayek’s concernsabout replacing entrepreneurs with cen-tral planners as follows:

the trial and error procedure would, or atleast could, work much better in a socialisteconomy than it does in a competitive mar-ket. For the Central Planning Board has amuch wider knowledge of what is going on inthe whole economic system than any privateentrepreneur can ever have, and, conse-quently, may be able to reach the right equi-librium prices by a much shorter series ofsuccessive trials than a competitive marketactually does. (Lange 1938, p. 89, emphasis inthe original)

What about the skewing of incentivesunder socialism? Acknowledging theimportance of the problem, Lange of-fered two responses. First, he deniedthat such agency questions are a propertopic for economists to study: “The dis-cussion of this argument belongs to thefield of sociology rather than of eco-nomic theory and must therefore bedispensed with here” (Lange 1935, p.109). Second, he insisted that the realproblem was one of bureaucracy. But


bureaucratization, he continued, is a ge-neric problem that afflicts both capital-ism and socialism. Because of the ab-sence of competition, the managers of abureaucratic modern capitalistic corpo-ration are as likely to be inefficient asare their counterparts under socialism.The separation of ownership from con-trol exacerbates this: the modern capi-talist corporation is increasingly run bya professional managerial class whosemembers care more about their ownwelfare than about running an efficientfirm. Bureaucracy is a problem of mod-ern life, not one that is unique to social-ism (Lange 1938, pp. 109−10, 120). Tosupport his case Lange cited AdolfBerle and Gardiner Means (1933), theclassic study of the separation of owner-ship from control in the modern corpo-ration and a forerunner of the modernprincipal-agent literature. Lange’s argu-ment underlines again the crucial im-portance for socialists of the claim thatold-style atomistic competition is rareunder late capitalism.

D. Answering Lange—Mises on Appraisement and the Entrepreneur

Mises never directly replied toLange. But it is clear from his laterwritings that he rejected Lange’s claimthat prices are nothing more than“terms on which alternatives are of-fered.” For Mises, prices are social phe-nomena “brought about by the interplayof the valuations of all individuals par-ticipating in the operation of the mar-ket” (Mises 1966, p. 331). They reflectthe plans and appraisements of millionsof acting individuals at a particular mo-ment in time.

Given its origin in the appraisementsof millions of people, the price struc-ture is constantly changing. Even so, itis an essential tool used by entrepre-neurs to make calculations about thehighest-valued use of scarce resources.

Crucially, such calculations are alwaysfuture-oriented.

In drafting their plans the entrepreneurs lookfirst at the prices of the immediate pastwhich are mistakenly called present prices.Of course, the entrepreneurs never makethese prices enter into their calculationswithout paying regard to anticipated changes.The prices of the immediate past are forthem only the starting point of deliberationsleading to forecasts of future prices. . . . Theprices of the past are for the entrepreneur,the shaper of future production, merely amental tool. (Mises 1966, pp. 336−37, empha-sis in the original)

Entrepreneurs must make decisionsabout resource use in a world in whichproduction takes time and in which theconstant evolution of human plans cre-ates a constantly changing structure ofprices. In such an environment errorsclearly are unavoidable. But they do notpersist, because every mistake made byone entrepreneur is simultaneously aprofit opportunity for another: “it is thecompetition of profit-seeking entrepre-neurs that does not tolerate the preser-vation of false prices of the factors ofproduction” (Mises 1966, pp. 337−38,emphasis in the original). Thus theever-changing structure of prices thatexists within a market system, the messygroping that appears so anarchic, endsup being a passably efficient systemfor revealing relative scarcities. Andparadoxically, though the price systemoperates through the self-interested ac-tions of thousands of individuals, its endresult is social cooperation: Paris getsfed.

For Mises, the entrepreneur is theessential actor of the piece, but equallyimportant is that his actions take placewithin a specific institutional frame-work. Absent a market system withwell-defined and enforced propertyrights, the entrepreneur would haveneither the necessary information northe proper incentives to perform his es-


sential function.4 Mises acknowledgedthat Lange’s attempt to introduce pricesand competition into a socialist frame-work showed some recognition of allthis (Mises 1966, p. 706). But withoutthe rest of the requisite institutionalsetting, such efforts must ultimatelyfail.

E. Answering Lange—Hayek on Computation and on Knowledge

Hayek initially responded to Lange ina book review (Hayek [1940] 1948),then elaborated on and extended his ar-gument in a series of articles (Hayek[1945, 1946] 1948; [1968a] 1978).

1. The Computation Problem and“Trial and Error”—Recall that Hayekhad in 1935 already provided argumentsagainst such approaches as “the mathe-matical solution” and, more importantlygiven Lange’s arguments, against thefeasibility of trial and error methods. Inhis review he wondered why Lange hadfailed to address his objections aboutthe latter, and why his opponent hadeven neglected to answer the obviouslyimportant question of how often priceswere to be adjusted under his proposedsystem. Hayek ([1940] 1948, p. 188)added that,

it is difficult to suppress the suspicion thatthis particular proposal has been born out ofan excessive preoccupation with problems ofthe pure theory of stationary equilibrium.

Hayek’s point was a simple one: staticequilibrium theory concentrates onend-points, on a system that has

achieved a state of rest. But the notionof a system moving toward some “final”end-point as determined by “given”data is radically at odds with the situ-ation in the real world, “where constantchange is the rule” (Hayek [1940] 1948,p. 188). Hayek was suggesting thatLange’s use of an equilibrium modelhad misled him into thinking that themovement toward some final equilib-rium set of accounting prices would bea one time adjustment, whereas in real-ity it would be a never-ending process.

Hayek’s arguments about the difficul-ties of coming up with the necessarydata have been challenged by each newgeneration of socialists, as first input-output analysis, then planometrics, thencomputable general equilibrium mod-els, then the advent of supercomputersall promised finally to provide an instru-ment that could replace the market’sprice adjustment mechanism. As a ma-ture Lange (1967, p. 158) provocativelyput it:

Were I to rewrite my essay today my taskwould be much simpler. My answer to Hayekand Robbins would be: so what’s the trouble?Let us put the simultaneous equations on anelectronic computer and we shall obtain thesolution in less than a second.

More recently, Allin Cottrell and W.Paul Cockshott (1993) and Steven Hor-witz (1996) have offered contrasting as-sessments of the feasibility of socialistcomputational proposals.

As a practical matter, though, fewconvincing examples of the successfulreplacement of markets exist. Ironi-cally, the practical effects of the com-puter revolution so far seem to havebeen to undermine the ability of totali-tarian states to restrict access to infor-mation, and to enable entrepreneurs toengage in the sort of “atomistic compe-tition” that socialists of the 1930s hadassumed had died out. It is also note-worthy that within the artificial intelli-

4 Joseph Salerno (1990) recently sparked a livelydebate with the provocative claim that the Mise-sian emphasis on appraisement and entrepreneur-ship differs from, and is more fundamental than,Hayek’s arguments about knowledge. LelandYeager (1994) and Kirzner (1997) contend that thetwo positions are better considered complemen-tary, while Boettke (forthcoming) argues that anydifferences of emphasis that may exist betweenMises and Hayek are due to their responding todifferent audiences.


gence literature the relationship be-tween markets and computers is becom-ing very nearly the opposite of thatimagined by Lange: rather than regard-ing markets as primitive forms of com-puters, efforts are aimed at makingcomputers replicate the allocational andadaptative characteristics of markets(Lavoie 1990, p. 76).

2. Hayek’s “Knowledge” Arguments—Hayek provided an example of the con-sequences of taking static theories tooseriously in a discussion of Lange’s costminimization rule. Hayek asked: howwill planners come to know what theminimum costs are ([1940] 1948, p.196)? His basic contention was that onlythrough the workings of a rivalrous mar-ket process are ever lower cost methodsof production discovered or created (cf.Hayek, [1946] 1948, pp. 96−97). Stan-dard equilibrium theory misleads by as-suming that an end-state is alreadyreached, so that cost minimizing inputcombinations are already known. Thisobscures the process by which theycome to be known, and may lead to theerroneous belief that one can dispensewith the very process (rivalrous marketcompetition) that generates the knowl-edge. More generally, the static theoryof perfect competition “starts from theassumption of a ‘given’ supply of scarcegoods. But which goods are scarcegoods, or which things are goods, andhow scarce or valuable they are—theseare precisely the things which competi-tion has to discover” (Hayek [1968a]1978, p. 181). In a phrase: Market com-petition constitutes a discovery proce-dure.

Lange had also argued that, becauseentrepreneurs have knowledge aboutonly a limited set of markets and prices,a Central Planning Board (which wouldhave access to more knowledge thanwould individual entrepreneurs) couldmake better capital allocation decisions.

In his 1937 article “Economics andKnowledge” Hayek noted that thoughstandard equilibrium theory assumesthat all agents have access to the same,objectively correct information, in real-ity there is a “division of knowledge.”Actually-existing knowledge is dis-persed; different individuals have ac-cess to different bits of it. Their actionsare based on their varying subjectivebeliefs, beliefs that include assumptionsabout future states of the world andabout other people’s beliefs and ac-tions. The “central question of all socialscience” ([1937] 1948, p. 54) is howsuch dispersed knowledge might be putto use, how society might coordinatethe knowledge that exists in many dif-ferent minds and places. Equilibriumtheory with its emphasis on end-statesassumes that the process of coordin-ation has already taken place. By doingso, it assumes away the most importantquestion.

A number of authors (Caldwell1988a; Kirzner 1988; Meghnad Desai1994) have argued that “Economics andKnowledge” was a seminal piece both inthe development of Hayek’s ideas andfor its implications for the calculationdebate. In his review, Hayek cited thearticle to show that Lange had againbeen misled by “equilibrium theory.”

As I have tried to show on another occasion,it is the main merit of real competition thatthrough it use is made of knowledge dividedamong many persons which, if it were to beused in a centrally directed economy, wouldhave all to enter the single plan. To assumethat all this knowledge would be automat-ically in the possession of the planningauthority seems to me to miss the main point.(Hayek [1940] 1948, p. 134)

Tracing out the implications of the “dis-persion of knowledge” would become an-other major theme in Hayek’s work. Inthe present context, three relatedstrands may be identified. First, freely


adjusting market prices that reflect rela-tive scarcities are profoundly importantwhen errors in perception exist, becausethey help market participants to bringtheir subjectively formed expectations inline with the actual state of the world.Next, much knowledge (particularly ofbusiness conditions) is localized, whatHayek came to call “knowledge of theparticular circumstances of time andplace” ([1945] 1948, p. 80). Finally, cer-tain knowledge is tacit; it is “knowledgehow” rather than “knowledge that”([1968b] 1978, p. 38). Localized andtacit knowledge is difficult (and may beimpossible) to pass on to others, even ifone wanted to.

Modern theorists were quick to pickup on Hayek’s insight that, in a world ofdispersed knowledge, prices convey in-formation. The same cannot be saidabout his writings on localized and tacitknowledge, the importance of which iswell captured in the following exampleprovided by Leland Yeager:

Often an entrepreneur makes business deci-sions partly on his intuition or feel for tech-nology, the attitudes and tastes of consumersand workers, sources of financing, and condi-tions in markets for inputs and for consumergoods and services—all in the future as wellas the present. The entrepreneur receives in-formation for judgments about such mattersby reading specialized and popular publica-tions, watching television and movies, experi-encing various services and products person-ally, chatting with innumerable people, andstrolling through town or the shopping mall.Much of what he thereby observes—orsenses—he could not express in explicitwords or numbers. A socialist system wouldlet much such entrepreneurial knowledge goto waste even if it emerged in the first place.(Yeager 1996, p. 138)

The idea that “the pure theory of sta-tionary equilibrium” is inadequate as atool for understanding the workings of amarket economy, and that it should bereplaced by a view of the market as acompetitive-entrepreneurial process for

the discovery and coordination ofknowledge, has become a central tenetof Austrian thought.5 Hayek’s argu-ments are probably most effective ifone considers centrally controlled so-viet-style economies, where price fixingis extensive and, as a result, the co-ordination of plans is hindered. But theyalso apply to many market socialist re-gimes. In Lange’s proposal, for example,prices were still set by central authori-ties; this is why Hayek insisted thatLange must reveal how often priceswould be adjusted under his “trial anderror” regime. And Hayek’s argumentsabout discovery, error correction, and thecoordination and communication of lo-calized and tacit knowledge apply when-ever entrepreneurs acting within an in-stitutional context of market competitionmake decisions that differ from those oftheir socialist manager counterparts.

Because of its connection to recentdebates on the economics of informa-tion, we will delay until the last sectionour discussion of Hayek’s response toLange on the incentives question. In-stead, we will next examine Hayek’s po-litical arguments against socialism. Hebegan developing the argument in arti-cles written in the late 1930s, but morefamous is the analysis contained in TheRoad to Serfdom ([1944] 1976).

III. The Road to Serfdom

A. Origins of the Book

The Road to Serfdom carries thededication, “To the Socialists of All

5 For more on “market process theory” see Kir-zner (1973, ch. 1; 1997); Ludwig Lachmann(1976); and Esteban Thomsen (1992). Hayek’s wasnot a blanket rejection of general equilibrium the-ory. He thought that even the static Walrasianmodel contained important insights about marketinterdependence, and he spent much of the later1930s in an unsuccessful attempt to develop a dy-namic intertemporal general equilibrium model ofa capital-using monetary economy, as Jack Birner(1994, pp. 2–5) discusses.


Parties.” This was not just a rhetori-cal swipe; in the 1930s most Britishintellectuals (Hayek’s presumed audi-ence) were sympathetic to socialism(Arthur Marwick 1964). The many dif-ferences that might ordinarily sepa-rate a Liberal-Labour centrist from amember of the Communist Party ofGreat Britain were intentionally down-played once the Popular Front, a unionof leftist groups against fascism,emerged in the middle 1930s. In theirown way even the Conservatives joinedin. In 1938 future Prime Minister (thenthe Conservative MP from Stockton-on-Tees) Harold Macmillan published TheMiddle Way (1938), in which extensivegovernment control of the economy wasextolled. Hayek had little sympathy forsuch views, dubbing them “the muddleof the middle.”

Hayek later reminisced that theoriginal impetus for the book was com-ments made by Lord Beveridge (Hayek1994, p. 102). But he was also fightingagainst the widely accepted view thatNational Socialism and other fascismswere a natural outgrowth of capitalism,and that only by adopting socialismcould the remaining western democra-cies avoid a similar fate. Karl Mann-heim, an academic who had fled Frank-furt in 1933 and who soon gained anappointment in the Department of Soci-ology at the LSE, was one of the moresophisticated proponents of this posi-tion.

Mannheim outlined the causal mech-anisms at work in a lecture beforehis British colleagues first publishedin 1937. His starting premise wasthat monopoly capitalism was the rootcause of widespread and sustained un-employment. In an age of mass democ-racy, orators skilled in the use of thelatest propaganda techniques can ma-nipulate public opinion. Demagoguesemerge who play on the collective

insecurity of the masses, providingscapegoats and offering escape intosymbols of past glories. There is a grad-ual breakdown of societal responsibility,and totalitarian forms of governmentstep in to fill the vacuum. The break-down is aided by capitalists, whoseallegiances are few and fleeting, andwho see new profit opportunities inevery change of regime (Mannheim1940, ch. 3).

Reflecting on the recent experienceof Germany, Mannheim concludedthat the nascent democracies of Mid-dle Europe were lost. He held outsome hope for England, but only if itwould give up liberal democracy andembrace a comprehensive system ofplanning. In the latter half of the booka variety of modern methods ofsocial control are outlined, all of whichcould be used to make the transi-tion from a liberal to a planned soci-ety. What implications did Mann-heim think such planning had for free-dom?

At the highest stage freedom can only existwhen it is secured by planning. It cannot con-sist in restricting the powers of the planner,but in a conception of planning which guar-antees the existence of essential forms offreedom through the plan itself. For every re-striction imposed by limited authoritieswould destroy the unity of the plan, so thatsociety would regress to the former stage ofcompetition and mutual control. (Mannheim1940, p. 378)

Mannheim’s book was received well. Thereviewer in Economica described it as“epoch-making,” adding that “it is de-voutly to be wished that the teaching ofthe book may, though various channels,filter down from the specialist readersand seep into the popular mind, espe-cially into the political mind” (F. Clarke1940, pp. 330−31).

Hayek’s political works of the late1930s and early 1940s may be read as


an attempt to turn arguments likeMannheim’s on their head.6

Rather than the only means of coun-teracting totalitarianism, Hayek arguedthat planning itself constituted a signifi-cant step along the road toward the to-talitarian state.

The main point is very simple. It is that com-prehensive economic planning, which is re-garded as necessary to organize economic ac-tivity on more rational and efficient lines,presupposes a much more complete agree-ment on the relative importance of the differ-ent social ends than actually exists, and thatin consequence, in order to be able to plan,the planning authority must impose uponthe people the detailed code of values thatis lacking. (Hayek [1939] 1997, p. 193; cf.[1944] 1976, p. 57)

******

In the end agreement that planning is neces-sary, together with the inability of the demo-cratic assembly to agree on a particular plan,must strengthen the demand that the govern-ment, or some single individual, should begiven powers to act on their own responsibil-ity. It becomes more and more the acceptedbelief that, if one wants to get things done,the responsible director of affairs must befreed from the fetters of democratic proce-dure (Hayek [1939] 1997, p. 205; cf. [1944]1976, pp. 62−63)

Because of the absence of a sharedcode of values, authoritarian govern-ment tends inevitably to expand beyondthe economic and into the political do-main, even under those forms of social-ism that may have started out as demo-cratic. Mannheim was wrong to thinkthat only under planning would free-dom persist. It was just the opposite:Only if democracy is allied with a freemarket system will freedom of choicebe permitted to exist.

Democratic government has worked success-fully where, and so long as, the functions ofgovernment were, by a widely acceptedcreed, restricted to fields where agreementamong a majority could be achieved by freediscussion; and it is the great merit of theliberal creed that it reduced the range of sub-jects on which agreement was necessary toone on which it was likely to exist in a societyof free men. It is now often said that democ-racy will not tolerate “capitalism.” If “capital-ism” means here a competitive system basedon free disposal over private property, it is farmore important to realize that only withinthis system is democracy possible. When itbecomes dominated by a collectivist creed,democracy will inevitably destroy itself.([1944] 1976, pp. 69−70; cf. [1939] 1997, pp.205−06)

B. Prediction or Warning?

Hayek’s book found a large popularaudience (in America after The Reader’sDigest came out with a condensed ver-sion in April 1945, and among later gen-erations in Eastern Europe whensamizdat copies circulated), but its re-ception within much of the Anglo-American academic community wasnegative. A common criticism focusedon Hayek’s apparent prediction thatplanning must necessarily and inevita-bly lead to authoritarianism; the lastsentence in the quotation above is thesort of passage that gave rise to thecriticism.

This criticism was repeatedly raised,for example, by Wootton in FreedomUnder Planning (1945, e.g., pp. 28, 36−37, 50).7 Wootton’s courteous book wasexplicitly written as a response to The

6 Hayek explicitly identifies Mannheim as an op-ponent in the second chapter of The Road to Serf-dom ([1944] 1976, p. 21); and also mentions himin his ([1941–44] 1952b, pp. 156, 166n) and([1970] 1978, p. 6).

7 Wootton was not unique; such diverse figuresas George Stigler (1988, pp. 140, 147) and PaulSamuelson came away with similar readings. Atone point Hayek sent Samuelson a stronglyworded letter objecting to the latter’s repetition ofthe inevitability thesis in the 11th edition of hisPrinciples text. In a letter dated January 2, 1981,Samuelson graciously apologized and promised totry to represent Hayek’s views more accurately inany future work. See the Samuelson file, Box 48,number 5, in the Hayek Archives, Hoover Institu-tion, Stanford, CA.


Road to Serfdom. It is also repre-sentative of its genre regarding ques-tions of “human nature.” In the firstnine chapters Wootton assumes that“planners are public-spirited peoplewho seek only to discover the commongood, and to do their best for it” (1945,p. 19). Only in the last chapter is thisassumption dropped; but even there sheinvokes such devices as a better in-formed electorate, more forthright po-litical parties, and oversight by bodiesof concerned citizens to accomplish themanifold goals of planning.

Hayek objected to the criticism, argu-ing that The Road to Serfdom wasmeant to be a warning, not an historicalprediction. He noted that the book’s in-troduction contains such caveats as “nodevelopment is inevitable” (p. 1); “thedanger is not immediate” (p. 2); and,most significantly, “Nor am I arguingthat these developments are inevitable.If they were, there would be no point inwriting this. They can be prevented ifpeople realize in time where their ef-forts may lead” (p. 4).

Certain of Hayek’s other writings,and in particular his methodological cri-tique of “historicism,” may also be men-tioned in his defense. Historicism in-cludes the claim that human historyconsists of “a necessary succession ofdefinite ‘stages’ or ‘phases,’ ‘systems’ or‘styles,’ following each other in histori-cal development.” But Hayek rejectedthis historicist endeavor “to find lawswhere in the nature of the case theycannot be found . . .” ([1941−44] 1952b,p. 128). Given his clear statements thatthere exist no immutable laws dictatinginevitable historical trends, he was sur-prised to find others claiming that hehad sought to demonstrate the exist-ence of such a trend.

One can see why the “prediction orwarning” issue is an important one. Ifone takes Hayek’s words as predictions

of inevitable trends, the events foretoldobviously did not come to pass in En-gland, the country Hayek had in mindwhen writing the book. If one takesthem as warnings, however, his later re-marks about the negative “psychologicaleffects” of postwar Labour rule in Brit-ain become more comprehensible, per-haps even apposite (Hayek [1944] 1976,pp. x−xvi).

A final point may help put the dis-pute into perspective: Neither Mann-heim’s nor Hayek’s books were uniquein their times. The western world wasturned upside down, and all manner ofintellectuals, from journalists like Wal-ter Lippmann (1937), to economists likeSchumpeter ([1942] 1950) and Karl Po-lanyi (1944), to philosophers like JamesBurnham (1941), to historians like Ed-ward H. Carr (1945), felt compelled toruminate on its past development, pre-sent dilemmas, and future prospects.

None of the other authors, though,quite shared Hayek’s fate. He had ahard time finding a publisher for thebook, when it finally appeared some ofhis critics’ reactions bordered on the li-belous (e.g., Herman Finer 1945), andHayek later was to claim that the book“went so far as to completely discreditme professionally” (1994, p. 103). Thisis one reason why the book should beread today. Readers who know it onlyby its reputation will be surprised tofind that it contains a number of recom-mendations for state intervention in theeconomy, as is evident in the following(admittedly qualified) passages:

There is no reason why in a society whichhas reached the general level of wealth whichours has attained the first kind of security [hehad earlier mentioned “security against se-vere physical privation, the certainty of agiven minimum of sustenance for all”] shouldnot be guaranteed to all without endangeringgeneral freedom. There are difficult ques-tions about the precise standard which shouldthus be assured; there is particularly the im-


portant question whether those who thus relyon the community should indefinitely enjoyall the same liberties as the rest. An incau-tious handling of these problems might wellcause serious and perhaps dangerous politicalproblems; but there can be no doubt thatsome minimum of food, shelter, and clothing,sufficient to preserve health and the capacityfor work, can be assured to everybody.

******

Nor is there any reason why the stateshould not assist the individuals in providingfor those common hazards of life againstwhich, because of their uncertainty, few indi-viduals can make adequate provision. Where,as in the case of sickness and accident, nei-ther the desire to avoid such calamities northe efforts to overcome their consequencesare as a rule weakened by the provision ofassistance—where, in short, we deal withgenuinely insurable risks—the case for thestate’s helping to organize a comprehensivesystem of social insurance is very strong.(Hayek [1944] 1976, pp. 120−21)

Hayek’s remarks are qualified, but itseems clear that, at least in 1944, he waswilling to countenance some level of“safety net” policy. In a new preface forthe book prepared in 1976, however,Hayek wrote that “I had not wholly freedmyself from all the current intervention-ist superstitions, and in consequence stillmade various concessions which I nowthink unwarranted” ([1944] 1976, p. xxi).

C. Liberty and Knowledge—Hayek’s Alternative Liberal Utopia

Hayek’s initial arguments promptedone socialist critic (Dickinson 1940) tochallenge him to describe his ideal lib-eral society in more detail. After read-ing The Road to Serfdom, Keynes alsochallenged Hayek to state clearly thecriteria he would use to distinguish ac-ceptable from pernicious governmentintervention (Jeremy Shearmur 1997).Though these tasks were begun in his1944 book, Hayek’s most refined re-sponse to his critics may be found in

The Constitution of Liberty (1960).8 And indeed, Hayek’s prolonged effortto explore the origins, nature, and sus-tainability of a liberal market order isprobably the most important legacy ofThe Road to Serfdom.

Hayek (1960, p. 11) began by defin-ing “liberty” as a condition “in whichcoercion of some by others is reducedas much as possible in society.” Thisproduces a dilemma, because the bestway to avoid coercion is to set up a co-ercive power that is strong enough toprevent it. Free society has met theproblem by defining a private sphere ofindividual activity, granting the state amonopoly on coercion, then constitu-tionally limiting the power of the stateto those instances where it is requiredto prevent coercion. The state’s coer-cive actions are constrained by the ruleof law: the laws it makes in protectionof the private sphere must be prospec-tive, known, certain, and equally en-forced (pp. 205−10). Hayek contraststhese with laws that seek specific out-comes within the private sphere, suchas certain redistributive patterns (p.232). What especially rankled him wasthe attempt to turn what Hayek viewedas legitimate “safety net” insuranceschemes into explicit instruments of re-distribution aimed at securing “socialjustice” (e.g., p. 289). In his discussionsof old age and health insurance policy,some of his dark warnings about inter-generational political strife make fortimely, if uncomfortable, reading today(e.g., pp. 295−98).

In his political writings Hayek fre-quently stated that he was no fan oflaissez-faire. By this he meant that a

8 Hayek’s other major contribution was the tril-ogy Law, Legislation and Liberty (1973–79),which contained his diagnosis of why liberal de-mocracies were being taken over by special inter-ests, and a proposed legislative reform aimed atstrengthening liberal constitutionalism.


market system must be embedded in aset of other institutions—a democraticpolity, with strong constitutional pro-tection of a private sphere of individualactivity, with enforced and exchange-able property rights—if it is to work.The problems of the Eastern Europeantransition have made these ideas seemalmost obvious, and they are ancientideas. But Hayek was stressing them be-fore they (re)gained popularity.

Hayek also linked liberty to his recur-ring theme, the problem of how tomake use of dispersed knowledge. Associety progresses the division of knowl-edge increases, as does our dependenceon the knowledge possessed by others:“When we reflect how much knowledgepossessed by other people is an essen-tial condition for the successful pursuitof our individual aims, the magnitude ofour ignorance of the circumstances onwhich the results of our action dependappears simply staggering” (p. 24). Themost successful societies are those inwhich each individual is able to put hisown local knowledge to its best uses.From this perspective, The Constitutionof Liberty describes the complex of in-stitutions and beliefs that promote thediscovery, transmission, and use ofknowledge so that individuals mighthave the best chance to have success inpursuing their own goals. Chief amongthe enabling conditions is liberty itself:

The rationale of securing to each individual aknown range within which he can decide onhis actions is to enable him to make the full-est use of his knowledge, especially of hisconcrete and often unique knowledge of theparticular circumstances of time and place.The law tells him what facts he may count onand thereby extends the range within whichhe can predict the consequences of his ac-tions. (pp. 156−57)

Hayek’s political philosophy is notuncontroversial. Critics have pointedout that he mixes a number of different

ethical and political philosophies to-gether, positions that may not cohereand all of which have been indepen-dently criticized; that in particular it isdifficult to square his Kantian ethicalideas about universalizability with hisHumean epistemological pessimism;and that the characteristics he requiresthat laws possess are not adequate toguarantee liberty (Arthur Diamond1980; Chandran Kukathas 1989; Gray1986, ch. 6). If one is judging his workagainst the standard of whether he pro-vided a finished political philosophy,Hayek did not succeed though Shear-mur (1996) contains a diagnosis of whatwould need to be done to build a coher-ent liberal political philosophy startingfrom Hayek’s foundations, and an ad-mittedly preliminary attempt to beginthat enterprise. It nonetheless remainsan impressive attempt to construct anintegrated system of social philosophy,one that blends insights from such di-verse fields as economics, political phi-losophy, ethics, jurisprudence, and in-tellectual history. And it is hard to denyHayek’s foundational contention that aliberal order allows individual knowl-edge to be better used than does social-ism.

IV. Hayek and Spontaneous Orders

A. The Critique of Rationalist Constructivism

Hayek’s final set of arguments againstsocialist planning begin from the prem-ise that the market system and certainother social institutions are examples ofspontaneously organized complex phe-nomena, spontaneous orders which gen-erate unintended beneficial conse-quences for those lucky enough to liveunder them. Now especially in mid-cen-tury, most of Hayek’s readers wouldhave thought it quaint to consider themarket system as a paradigmatic exam-


ple of a self-organizing system. Forthem it was more like a machine thathad broken down, something that re-quired radical repair, if not outright re-placement. Hayek labeled this opposingviewpoint “rationalist constructivism,”located its origins in the French ration-alist variant of Enlightenment thought,and used it as a kind of foil in manylater writings. Thus, according toHayek, rationalist constructivists be-lieved:

that human institutions will serve human pur-poses only if they have been deliberately de-signed for these purposes, often also that thefact that an institution exists is evidence of itshaving been created for a purpose, and al-ways that we should so redesign society andits institutions that all our actions will bewholly guided by known purposes. To mostpeople these propositions seem almost self-evident and to constitute an attitude aloneworthy of a thinking being. (1973, pp. 8−9;also see his [1964a] 1967; [1970] 1978)

Hayek’s earliest critique was aimed at“the engineering mentality” (1952b, ch.10) of the “men of science” ([1941] 1997;cf. Editor’s Introduction), but his targetwas later broadened to include propo-nents of rationalism, empiricism, positiv-ism and utilitarianism (1988, pp. 60−62).Hayek’s apparently creative historiogra-phy has been challenged by both Dia-mond (1980) and Gray (1988).

Following Carl Menger ([1883]1985,Book 3), Hayek argued that spontane-ous social orders are products of humanaction but not intentionally designed.Such institutions evolved gradually, andonly after they emerged were their ad-vantages recognized, first by certain ofthe Scholastics, then by various mem-bers of the Scottish Enlightenment, andthen again by Menger. Hayek’s fasci-nation with such orders began back inthe 1930s ([1933] 1991, pp. 17−34;Caldwell, 1988b, discusses the impor-tance of the piece), and indeed is evi-

dent in the question he broached in“Economics and Knowledge” about howhuman action gets coordinated even inthe absence of a central controllingauthority. In his early writings the co-ordinating role of freely adjusting mar-ket prices was highlighted. But soonHayek began including all sorts of prac-tices, norms, rules and other forms ofinstitutions as aiding social coordi-nation. Thus he would argue that suchinstitutions as language, the law, andmoney emerged because they contrib-uted to the ability of individuals to pur-sue their own goals; that our moralswent through a similar sort of culturalevolution; and that (as we shall soonsee) even the ordering of neural net-works within the human brain developsanalogously.

Hayek employed his account of cul-tural and institutional evolution to criti-cize the view that society can recon-struct institutions or moral codes to bemore rational. As always, knowledgequestions played a prominent role inthe argument. Hayek believed thatspontaneously emerging practices, normsand institutions not only allow humansto use knowledge better, they also per-mit knowledge gained in the past to bepreserved, because they are artifacts re-flecting the experimentation of manypeople over long periods of time:

Far from assuming that those who createdthe institutions were wiser than we are, theevolutionary view is based on the insight thatthe result of the experimentation of manygenerations may embody more experiencethan any one man possesses. (1960, p. 62)

As a result, attempts to radically alter orreconstruct these institutions are fraughtwith dangers; we simply do not haveenough knowledge about what they doand how they do it. Given our ignorance,only “the hubris of reason” would leadone to believe that we can rebuild soci-ety from the ground up (1973, p. 33).


“Reason properly used” understands thelimits of what reason can do; like DavidHume before him, Hayek sought to“whittle down the claims of reason bythe use of rational analysis” (1960, p.69). The implication for economic analy-sis was for Hayek straightforward: “whatwe can know in the field of economics isso much less than people aspire to”(1983, p. 258).

Hayek’s evolutionary arguments havebeen extensively criticized (particularlyhis reliance in later work on the idea of“group selection”), perhaps most effec-tively by those like Viktor Vanberg(1994, chs. 5−7 & 12) who believe thatthe design of constitutions is an essen-tial element in the improvement of lib-eral orders. And Hayek, after all, wasnot averse to making proposals aboutconstitutional design himself (1979, ch.17), in apparent violation of his ownstrictures concerning constructivism.But he was also concerned with howspecific instances of liberal market or-ders ever came to be created in the firstplace, particularly because they sit souneasily with, as he put it (1988, ch. 1),both our “instinct and reason.” This is aquestion that his critics have thus farleft unanswered.

In any event, one particular exemplarof Hayek’s treatment of spontaneous or-ders, his book on the foundations oftheoretical psychology, The Sensory Or-der (1952a), has not been much studiedby economists (exceptions include Wil-liam Butos and Roger Koppl 1993;Horowitz 1994; and Steve Fleetwood1995). The book does not deal even pe-ripherally with socialism. But properlyunderstood it provides a key for com-prehending the nature and extent ofHayek’s divergence from mainstreameconomics in the postwar period, and somay help to explain why modern infor-mation theorists writing about marketsocialism have had such a hard time un-

derstanding him. As such, a short di-gression follows.

B. A Digression—The Sensory Order

During the war Hayek began a majorproject, one with both historical andmethodological dimensions, on “theabuse of reason.” As he later explained(1952b, pp. 10−11), the larger projectwas never completed, but he did finishan extended piece on methodology enti-tled “Scientism and the Study of Soci-ety” (reprinted in Hayek, 1952b). Theessay contains a critique of the idea thatthe methods that had been so successfulin the natural sciences should also beapplied in the social sciences; “scien-tism” refers to what Hayek called the“slavish imitation” (1952b, p. 24) by so-cial scientists of (what often turned outto be a caricature of) such methods. Healso offered a positive methodologicalalternative to scientism. His startingpoint was to describe what he took tobe the central subject matter of the so-cial sciences: the acting individual.Hayek outlined some fundamentalpremises about the mind of such an in-dividual, and about the relationship be-tween beliefs and action. Among thepremises were: the structure of the hu-man mind is everywhere the same; ouractions are based on our “opinions”(presumably these include both percep-tions and beliefs); these opinions aresubjectively held (and as such, some arewrong); finally, perceptions and beliefsdiffer among individuals.

Turning from a description of thesubject matter of the social sciences toa discussion of how best to study it (thatis, from ontology to methodology),Hayek proposed a “compositive” meth-odology in which the social scientist’stask is to show how the actions of manysuch individuals come to constitutebroader, more complex social phenom-ena. In his examination of the issues,


Hayek reached a crucial conclusion:when studying complex phenomena,typically one cannot predict individualactions. Pattern predictions, or explana-tions of the principle underlying thecomposition of social phenomena, areusually the best one can do (Hayek1952b, chs. 3 & 4; cf. Caldwell 1989).

Hayek’s next step was to try to pro-vide a firmer basis for his statementsabout the nature of mind. In the sum-mer of 1945 he dug out a manuscript onpsychology that he had written duringhis student days in Vienna nearly aquarter of a century earlier. The essayformed the basis for his next major con-tribution, The Sensory Order (1952a).

It appears that Hayek viewed TheSensory Order as providing the physi-ological underpinning for the varioustheses he had advanced concerning therelationship between mind, knowledge,and human action in the “Scientism” es-say.9 Hayek began by distinguishing be-tween the physical order existing in theexternal world and our phenomenal ex-perience of it. Our sensations, our per-ception of the qualities of objects, ourwhole mental image of the world, startfrom stimuli received through our cen-tral nervous system. This central nerv-ous system constitutes the common“structure of the mind” shared by all in-dividuals that had been posited in “Sci-entism.”

The system provides a relational or-dering of stimuli: the sensory order.The qualitative differences in percep-tions and sensations that we experiencedepend on the specific pattern of neu-ron firings that a given stimulus pro-duces within various neural networks.The experiences and beliefs of individu-

als will differ according to the patternof neural firings that each develops.Thus though the structures of ourminds are the same, there is a physi-ological basis for the notion of a disper-sion of perceptions and experiencesand, ultimately, of knowledge. The sen-sory system that results as the individ-ual interacts with and receives feedbackfrom the environment is both self-orga-nized and adaptive, one that is con-stantly adjusting (strengthening certainpathways and connections, while dimin-ishing the strength of others) as newstimuli come in. The adaptations repre-sent “learning” by the individual.

C. The Significance of The Sensory Order

The Sensory Order clearly heralds thefull emergence of complex “spontane-ous orders” as a theme in Hayek’s work.But for our purposes, its import lies inviewing it as Hayek’s decisive step awaynot just from the study of economics,but from standard economic reasoning,as well.

To see this note that, first, The Sen-sory Order provides a naturalistic onto-logical grounding for a number of meth-odological prescriptions. Both therecourse to ontology and the specificmethodological prescriptions Hayek de-rived were diametrically opposed to thephilosophical world view not just ofmost mainstream economists of his day,but of psychologists as well.

In mid-century, positivism and instru-mentalism were the philosophicaldoctrines that most influenced themethodological self-consciousness ofpsychologists and economists. Positiv-ists eschewed talk of ontology or meta-physics, and measured the veracity oftheories not by their ontologicalgrounding but by their ability to allowscientists to control or predict phenom-ena. This was the reasoning that al-

9 Hayek links the two essays in a retrospectivepiece (1982a, p. 289), and in his letter to John Nef(cited earlier). Further, the “Scientism” essay isthe only work by Hayek cited in The Sensory Or-der.


lowed behaviorist psychologists to denyscientific status to any and all attemptsto “get beneath” observable behavior todiscover its origins in human conscious-ness. Likewise, both Paul Samuelson’srecourse to “revealed preference” andMilton Friedman’s “as if” methodology(that is, the claim that the realism ofa theory’s assumptions is unimportantcompared to its ability to predict) sharethe positivist emphasis on observablephenomena and prediction. (The meth-odological positions of both Samuelsonand Friedman are examined in detail inCaldwell [1982] 1994.)

This aversion to ontology was exactlythe sort of view that Hayek attacked inboth the “Scientism” essay and in TheSensory Order. Furthermore, by provid-ing a grounding for his methodologicalclaims in physiological psychology,Hayek believed that he was the one whowas truly being “scientific.” Hayek’spsychological work is now recognized asan early challenge to behaviorism (e.g.,Gerald Edelman 1987, ch. 1; DonaldHebb 1949 was another early dis-senter), and his ontological boldness ex-plains his attractiveness to scientific re-alists of today like Tony Lawson (1994)and Fleetwood (1995).

The specific methodological conclu-sions that Hayek drew also contradictedthe positivistic dogma of the time. Asnoted above, when considering complexphenomena, Hayek claimed that typi-cally the best that we can do is to ex-plain the principles by which they work.To the extent that prediction is possibleat all, only broad “pattern predictions”may emerge (1952b, ch. 8; cf. [1955]1967, [1964b] 1967). Such pessimisticclaims about the ability of economists topredict may sound a bit less controver-sial today. But when he was writing andfor a number of decades to come itwould separate him from nearly allmainstream economists, even those, like

his colleagues at the University of Chi-cago, whose views on policy were oftenvery similar to his.

Finally, The Sensory Order markedHayek’s further movement away fromthe constructs used by economists tounderstand the market order. For main-stream economists, the concepts of“equilibrium” and “rationality” are vir-tually coextensive with economic analy-sis. By the middle of the century Hayekbelieved that neither one shed muchlight on the most important questions.

Recall that in “Economics andKnowledge” Hayek had asked: Howdoes human action get coordinatedwhen knowledge is limited and dis-persed? In light of this question, he feltthat the “equilibrium theory” of his daywas of little use: in equilibrium, full co-ordination has already occurred. Stan-dard theory also assumed the presenceof rational agents who all share thesame, objectively correct information.The individual described by Hayek inThe Sensory Order had little in com-mon with this “rational (and fully in-formed) economic man” construct.

Hayek never explicitly tried tochange economics, he never argued thathis own description should replace “ra-tional economic man” within economicmodels. And certainly a central ques-tion is what difference Hayek’s variousdepartures all might make. One way tosee is briefly to examine the latest in-stallment of the market socialism de-bates.

V. Market Socialism and the Economicsof Information

A. A New Chapter in the History of the Debate over Socialism

There is renewed interest in marketsocialism, and emerging with it hasbeen a new stylized history of the de-bates, one that highlights the contribu-


tions of the economics of information(Bardhan and Roemer, eds. 1993, pp.3−9; Stiglitz 1994).

The first part of this history incorpo-rates the conventional story that hadbeen told about the socialist calculationdebate through about the 1970s, withAbram Bergson (1948) being the classicearly statement (see Lavoie 1985, how-ever, for a modern revisionist reply). Inthat account, the debate was opened byvon Mises, who declared the “impossi-bility” of rational calculation under so-cialism. The “mathematical solution” ofDickinson, which utilized the Paretianor Walrasian general equilibrium modelto show the formal similarities betweena socialist and capitalist regime, under-mined Mises’ claim of impossibility.Hayek’s contribution via his “complexityargument” was to question the practicalfeasibility of solving a general equil-brium system. Lange’s proposed “trialand error method” undercut Hayek’sclaim, and the actual feasibility of so-cialism then became viewed as an em-pirical matter. Later the emergence ofsupercomputers and of computable gen-eral equilibrium models suggested thatthe complexity problem might well besoluable, so that the actual estab-lishment of a viable socialist state mightonly be a matter of time.

But this was not to be. The poor per-formance of communist regimes sug-gested that the chief problem was notone of computation, but of “incentivecompatiblity.” In the new account theWalrasian general equilibrium model(or its more recent counterpart, the Ar-row-Debreu model) became the villainof the piece, and particularly its impli-cation that a purely competitive systemwill yield efficient market outcomes, al-beit once a stringent set of marginalconditions is met. If information isasymmetric, however, deviations from aPareto-efficient outcome can occur.

While economists in earlier periods mayoccasionally have hit upon specific in-stances of information problems (e.g.,Berle and Means 1933), the Arrow-De-breu framework posed an obstacle, di-recting attention away from questionsof information.

In contrast, the economics of infor-mation provides a systematic means ofidentifying and classifying the wide ar-ray of problems that informationalasymmetries can produce. For example,within a planned economy variousmonitoring problems may arise: produc-tion managers might decide to feathertheir own nests, to seek patronage, topractice favoritism, or in other ways todeviate from the plan; product quality,being multidimensional, is difficult tomonitor; and so on. Nor is it an easymatter for the government to designand enforce a system of constraints toprevent managers from behaving oppor-tunistically. A government that is in anyway responsive to an electorate will findit difficult to commit itself credibly tostringent actions, such as raising pricesfor consumer goods, shutting down pro-duction lines when product demand de-clines, firing well-liked but incompetent(or worse, unlucky) managers, or in-forming redundant workers that theymust be retrained or transferred. Butplant managers confronted with such“soft budget constraints” (so dubbed byKornai 1986) are less likely to take theirproduction goals seriously. Investmentdecisions also pose problems, because itis difficult to avoid either too much ortoo little risk-taking behavior, and with-out competing projects, it becomesproblematical even to assess whether aninvestment once undertaken is in facteconomic (Roemer 1993, pp. 91−92).

As these brief examples suggest, theeconomics of information provides apowerful set of tools for identifying andanalyzing the problems of a socialist


economy. According to the new history,current models greatly improve onthose of the past because they contain asophisticated understanding of agencyproblems. Indeed, if one accepts thisaccount, the evolution of the variouspositions in the socialist calculation de-bate becomes a paradigmatic exampleof how theoretical progress can occur,and has occurred, in economics.

Perhaps inevitably, historical ac-counts that emphasize progress tend to-ward Whiggism: the work of economistsin the past is but a prelude to develop-ments in the present. Stiglitz, for exam-ple, invokes the debates of the 1930s,but mostly as window dressing. Hisbook Whither Socialism? (1994) is bestread as a celebration of the economicsof information, one that uses the ques-tion of the viability of socialism as itschief application.10

Other accounts treat the argumentsof the 1930s more carefully (e.g., Bard-han and Roemer 1993, pp. 3−9). Be-cause of their reliance on a full informa-tion general equilibrium framework,Lange and other market socialists farepoorly because they (inevitably) ignoredthe question of incentives. Hayek makesfor a harder interpretive case. To besure, he did understand that the pricesystem is a mechanism for conveying in-formation, and modern theorists arequick to acknowledge this. Thus in hiscollection The Informational Role ofPrices (1989), Sanford Grossman quotesin three different articles the same pas-

sage from Hayek’s “The Use of Knowl-edge in Society” (1945) about the pricesystem as a communication mechanism.In a popular textbook, Hayek is cred-ited with having provided the initial in-sight later formalized by Leonid Hur-wicz about the low informationalrequirements of a decentralized pricesystem (Paul Milgrom and John Roberts1992, p. 85).

But this only raises a new set of ques-tions. Given the limited nature of histools, how was it that Hayek stumbledonto the insights that he did? And giventhat Hayek’s critique of Lange focusedon how changing relative prices conveyknowledge, why did he make so littleheadway toward developing an econom-ics of information? This last question ispointedly raised by Makowski and Os-troy who, noting that the field of“mechanism design” differentiates be-tween (i) the information/communica-tion requirements of mechanisms, and(ii) the incentive properties of mecha-nisms, are led to the following assess-ment of Hayek’s contribution:

Failing to emphasize incentives as he did andconcentrating on the problem of communica-tion, Hayek should have arrived at a conclu-sion much closer to market socialism than hedid. Of course it could be noted that whereasthe division between (i) and (ii) may be ana-lytically useful, it is clearly artificial—anymechanism is a composite of (i) and (ii). If, asit seems to us, Hayek did not recognize theneed to separate the two functions (equiva-lently, to separate communication of informa-tion from elicitation of information), his cri-tique of market socialism is “fuzzy.” (inBardhan and Roemer, eds. 1993, p. 82)

From this perspective, Hayek’s majorcontribution was to point out that in anenvironment of scarcity in which infor-mation is disaggregated or dispersed, theprice system is a low-cost mechanism forits aggregation and transmittal. Latertheorists formalized this insight, in thecourse of which it became clear that

10 To be fair to Stiglitz, the book was adaptedfrom his Wicksell lectures and was originally to beon the economics of information, which affectedits format. But even so, there is reason for histori-ans to be uneasy when in the bibliography of abook on the economics of socialism, Mises andRobbins, Dickinson and Dobb go uncited andHayek, Lange, Abba Lerner, and Fred M. Taylorget only one entry each (Hayek’s citation is to TheFatal Conceit). By way of contrast, there are 122articles or books written or co-authored by Stiglitzlisted in his bibliography.


Hayek (and his opponents) failed tograsp a second problem, that of oppor-tunistic behavior in situations of infor-mational asymmetries. Because Hayek(and the Austrians in general) failed tosee the importance of incentives, and ofthe necessity of developing mechanismsto overcome problems associated withasymmetric information, they could notmove the field forward. They missed thesalient fact that information problemsplague both market and socialist re-gimes, and as a result their analyses wereat best incomplete, and at times defi-cient. These informational issues are thevery problems that a new generation ofmore sophisticated models are able toaddress.

I will argue that though there is sometruth in this account, it also constitutesa paradigmatic example of Whig history,where earlier contributions that do notfit well with contemporary theory per-force are diminished or misunderstood.I will then try to show how such pres-ent-centered interpretations of the his-torical record may have significant con-sequences for a proper understandingof the issues.

B. The Incentives Question

First, is it true that Austrians had hadno appreciation of the question of in-centives? The evidence here is mixed.

Hayek frequently mentioned incen-tive issues in the final chapter of Collec-tivist Economic Planning (1935). Hepointed out that the “Russian experi-ment” provided a concrete example ofthe “obvious difficulty of making peoplefollow out the plan loyally” (p. 206). Heraised the question of whether privateownership of property is necessary formanagers to be properly motivated, andasked what criteria should be used toassess managerial decision making (pp.219−20, 231−32). Hayek was particu-larly concerned with the investment de-

cision, and the “tendency to prefer thesafe to the risky enterprise” that he be-lieved would prevail among socialistmanagers (p. 234).

Mises, too, recognized the problem.His 1920 paper contains a section enti-tled “Responsibility and Initiative inCommunal Concerns,” where he wrote:“It is now universally agreed that theexclusion of free initiative and individ-ual responsibility, on which the suc-cesses of private enterprise depend,constitutes the most serious menace tosocialist economic organization” (Misesin Hayek, ed. 1935, p. 116). Other earlywork that mentioned the incentiveproblems facing socialism (John Rae1891, ch. 11, section 4; William H. Mal-lock 1907) provides evidence for Mises’claim that this limitation of socialist sys-tems had long been recognized.

A concern for incentives is also im-plicit in some of the later Austrian writ-ings. For example, Boettke (1995) ar-gues that certain sections of The Roadto Serfdom anticipated some of the pub-lic choice literature on government fail-ure. Furthermore, the notion that cer-tain institutional arrangements (i.e.,free markets operating under a regimeof enforced and exchangeable privateproperty rights) spur and reward en-trepreneurial alertness is a direct recog-nition of how institutions affect the in-centives agents face. And as Hayekhimself wrote in a late piece,

We have come to understand that the marketand the price mechanism provide in thissense a sort of discovery procedure whichboth makes the utilization of more facts pos-sible than any other known system, and whichprovides the incentive for constant discoveryof new facts which improve adaptation to theever changing circumstances of the world inwhich we live. ([1976] 1978, p. 236)

The conclusion, no matter how conve-nient, that the Austrians ignored incen-tives questions must be rejected.


On the other hand, it is also clearthat the Austrians made no sustainedsystematic contribution to the nowenormous literature on the economicsof information. More interesting, it isalso evident that Hayek sometimesdropped the argument even when hecould have used it. Thus in his rebuttalto Lange he was willing to grant, “forthe purpose of this argument,” that so-cialist managers “will be as capable andas anxious to produce cheaply as the av-erage capitalist entrepreneur” ([1940]1948, p. 196). The real question for thehistorian is: Given Hayek’s unrelentingopposition to socialism, given his readi-ness to develop political and even evo-lutionary arguments against it, andgiven his evident knowledge of this lineof attack, why was it not pursued morevigorously?

A number of plausible explanationsexist. Recall that Lange had offeredtwo responses to Hayek’s claim that asocialist regime would be plagued by in-centive problems, namely: 1) the onlyreal agency problem is one of bureau-cracy, but this is a generic problem thatexists whenever corporations are largeand competition is absent, and 2) such“sociological” issues are not properlyincluded within the domain of econom-ics.

Hayek never bought the first argu-ment, denying Lange’s premise thatthe emergence of large corporationsnecessarily implied the end of compe-tition. In “The Meaning of Com-petition” (1946) Hayek argued thatmarket rivalry is the key to true compe-tition, and that such rivalry can existeven when the conditions for perfectcompetition do not hold. The assump-tions of the theory of perfect competi-tion misled people like Lange to con-clude that, because so few markets inthe real world satisfy the assumptions,they must be monopolistic. Once again

static equilibrium theory obscured thenature of reality.11

It may be, however, that Hayek atleast in part accepted Lange’s secondargument. Questions of “bureaucracy”had indeed traditionally been the do-main of sociologists (e.g., Max Weber1978, pp. 956−1005). Though the linebetween economics and fields like soci-ology was less clearly drawn whenWeber was writing before the FirstWorld War, by the 1930s under the dualinfluence of positivism and the profes-sionalization movement the disciplinarydivisions were hardening.

It was also an article of faith amongintellectuals living in the 1930s that hu-man behavior was supremely malleable.Time and motion studies, techniques ofmass propaganda, and other methods ofbehavior modification were being stud-ied and perfected everywhere; evenLenin had come grudgingly to admirethe Taylor System of Scientific Manage-ment (Merkle 1980, ch. 4). Who was tosay that an appropriately “educated” or“conditioned” manager might not some-day be “socially engineered,” one whoseactions would be indistinguishable fromone who was responding to the carrotand stick of pecuniary profits andlosses? And what special insight couldan economist bring to this question of“psychology”? Indeed, Hayek himselfoccasionally used the adjective “psycho-logical” when discussing issues of incen-tives (e.g., 1935, pp. 206, 232).

Had Hayek pressed his argument, hewould at a minimum have opened him-self up to the charge that he was speak-

11 Economists continue to disagree, of course,about the consequences of the fact that few realworld markets meet the assumptions of perfectcompetition. Discussions of “workable competi-tion” in the 1940s and of “countervailing powers”in the 1950s, and the development of “contestablemarket” theories in the 1970s and of the “imper-fect competition” and “strategic trade” theories inthe 1980s, all appear to share this common origin.


ing outside of his area of expertise. Thatvery claim was made at one point.Hayek’s occasional reference to “humannature” in The Road to Serfdom elicitedthe following admonition from EvanDurbin (1945, p. 359): “It is a pity thatProfessor Hayek makes no use of thelight that has been thrown upon thesematters by scientific inquiry in recentyears; and remains content with old-fashioned and unsupported, yet dog-matic, generalizations about human na-ture and emotional behaviour.” It is nosmall irony, given Durbin’s remarks,that Hayek would within a decade com-plete his book on the foundations oftheoretical psychology (Hayek 1952a).

Hayek would also have to contend withthe old socialist belief that the opportun-istic behavior that gives rise to agencyproblems is itself a product of capitalistsociety. Marx had long ago argued thatone’s consciousness was determined byone’s social condition, rather than the re-verse. While capitalism promoted greed,under socialism a new socialist man, onewilling to sacrifice his own comfort forthe greater good, was supposed toemerge.12 Though this utopian vision haslong been challenged by opponents,13

there is no way to falsify it, because itrefers to an unspecified point in time ina supposedly radically altered future.

Hayek knew that socialism faced in-centive problems, but he didn’t want tobe drawn into what promised to be anunproductive debate, so he did not pur-sue the question. The reasons behindhis action were complex and subtle, andcurrent historical treatments have notshed much light on them.

Modern theorists might concede thehistorical point. But they would still in-sist that the significance of the incen-tives question has less to do with its his-tory than with the adequacy of Austriananalysis. Because of their failure to pro-vide a systematic analysis of the prob-lems arising from asymmetric informa-tion, Austrians have missed certainlimitations of markets. As a result, theirconfidence is misplaced; their defenseof the market system is naive.

It is true that, except for the occa-sional aside (e.g., Hayek [1982b], inChiaki Nishiyama and Kurt Leube, eds.1984, p. 61), Mises and Hayek never en-gaged the economics of information lit-erature. A number of younger Austrianshave begun that task, however, amongthem Stephen Boehm (1989), Thomsen(1992), and several of the authors inBoettke and David Prychitko, eds.(1996). And surely it is uncontroversialthat the Austrians have much to gainfrom a continuation of their interactionwith this literature.

But the coin may be flipped: main-stream analysts have seldom engaged(and when they have, have not alwaysunderstood, as is evident in the ex-change between Hurwicz 1984 andKirzner 1984; compare Boehm 1989;Kirzner 1997) the Austrian position,and thus may also have something togain from a better understanding of theAustrian insights. For those who mightdoubt that this advice about the advan-

12 Even the usually hard-headed modern social-ists Bardhan and Roemer (1993, p. 8) exhibit occa-sional nostalgia for these earlier arguments, leav-ing the door open a crack for the emergence ofthe long-hoped-for socialist economic man(ager):“it may be the case that the culture of manage-ment in a market-socialist economy, with its Wel-tanschauung of egalitarianism, would be different,at least to some extent, from the culture of capital-ist management.”

13 Mises (1935, p. 119) made the point withcharacteristic flair: “All socialist systems . . . pro-ceed from the assumption that in a socialist societya conflict between the interests of the particularand the general could not possibly arise. . . . Theobvious objection that the individual is very littleconcerned whether he himself is diligent and en-thusiastic, and that it is of greater moment thateverybody else should be, is either completely ig-nored or is insufficiently dealt with by them. Theybelieve they can construct a socialist common-wealth on the basis of the Categorical Imperativealone.”


tages of better mutual comprehension isanything more than a bromide, let usget down to cases by examining howHayek might have responded to the lat-est iteration of the market socialism de-bate.

C. An Austrian View of the New Debate on Market Socialism

Information theorists differ overwhether some form of market socialism,one that is cognizant of incentive com-patibility problems and takes steps toresolve them, can survive or even pros-per. We may take the works of Stiglitzand Roemer as representative of the op-posing views.

Stiglitz, representing the mainstream,admits that market economies arethemselves plagued with agency prob-lems, particularly when it comes to theclassic problem of the separation ofmanagement from ownership in largecorporations. The economics of infor-mation provides both the glasses to seewhat the problems are and, as progressoccurs, the tools with which to fix them.The prospects for market socialism,however, are less bright. Because theylack both the market discipline pro-vided by competition and the innova-tion that a decentralized market econ-omy promotes, socialist regimesinvariably will be less successful thantheir market counterparts (Stiglitz1994, chs. 7−9).

Roemer demurs, believing that theeconomics of information can be usedto fix socialist regimes, too. Roemer isdaring in the sorts of innovations he iswilling to countenance. For example,because of the commitment credibilityproblems that governments face, hedoes not believe that state ownership ofthe means of production is necessary(or even desirable) under socialism:“Socialists should be eclectic in their at-titude toward property relations: there

may be many forms of ownership moreamenable to socialism’s goals than tradi-tional state ownership of the means ofproduction” (Roemer 1994, pp. 6−7). Inthe place of the competition that is pro-vided naturally within market econo-mies, he envisions foreign competition,or even separate free market sectors co-existing within socialist economies, asdoing the job.

Not all of the proposals put forwardby market socialists are as “radical” asRoemer’s in terms of the leeway theywould permit to markets; indeed, somemight question whether Roemer’s pro-posals even qualify as socialism. But alldo seek to face the problem of incen-tives squarely. The middle section of hisbook with Bardhan (1993), entitled “Mi-croincentives,” is devoted to the solu-tion of a variety of incentive compatibil-ity problems that afflict market socialistregimes. For economists who mightdismiss these proposals as utopian,Roemer and Bardhan (1994, p. 179)rightly point out that cynicism about re-form flies in the face of “much of mod-ern economic theory, which is devotedto designing institutions that will leadself-interested players to arrive at effi-cient outcomes . . .”

The specifics of the new proposalshave been criticized: N. Scott Arnold(1996) claims that monitoring problemswould persist even were Roemer’s re-forms enacted. But surely, from atheoretical standpoint Roemer andBardhan are right to insist that if theeconomics of information can helpmake mixed market economies workbetter, there is nothing in principle toprevent it from being employed to mod-ify and reform market socialist econo-mies. Unfortunately, because he takesthe Arrow-Debreu model as his foilrather than the more recent work of so-phisticated market socialists, Stiglitzdoes not even address this issue. In a


review, Roemer (1995) pointedly notesthat Stiglitz’ Whither Socialism? shouldreally be read as an attack on the con-ventional general equilibrium modelrather than on recent models of marketsocialism.14

It certainly seems plausible that if thechief difficulty facing market socialismis the presence of the sorts of incentiveproblems that the economics of infor-mation identifies, then it should be pos-sible to construct a model of an effi-cient market socialist system. In thissense, there is a similarity between theeconomics of information and itstheoretical predecessor. (This is ironic,given the frequent emphasis by Stiglitzand other proponents of the economicsof information approach on how it dif-fers from general equilibrium theory.)Within the general equilibrium modelthere was no way on theoretical groundsto distinguish a competitive free marketsystem from a market socialist one. In alike manner, there is no way within themodel provided by the economics of in-formation to distinguish between a freemarket system and a market socialistone in which all agency problems havebeen identified, and to the extent possi-ble, corrected.

It is here that the contributions ofthe Austrians may make a difference.This has been recognized by advocatesof the Austrian view like Boettke (forth-coming), who considers the Austriananalysis of socialism to be the most im-portant Austrian contribution to eco-nomics, as well as by detractors likeFikret Adaman and Pat Devine (1996),

who nonetheless view the Austrian chal-lenge as more fundamental than thoseprovided by neoclassical economics.Precisely because they were at timeswilling to grant, for the sake of argu-ment, Lange’s claim that questions ofincentives were not the concern ofeconomists, the Austrian argumentsconcentrate on areas that are differentfrom those emphasized by informationtheorists.15 They focus instead on con-cerns that lie, as it were, outside of themodel. The sorts of issues that theywould raise are as follows:

1. Knowledge is different from infor-mation. Hayek’s writings about the abil-ity of the price system to convey infor-mation in a world in which knowledge isdispersed may well have inspired othersto investigate the design of mechanismsfor resource allocation within such envi-ronments. But it should also be clearthat when he used the term “knowl-edge,” he was referring to somethingthat is different in kind from the con-cept of “information” as it is used bycurrent theorists.

For Hayek, knowledge is dispersed,as mechanism design theorists pickedup on. But some knowledge, especiallythat generated as a result of day to daycontact within specific local markets, isalso tacit. Tacit knowledge is not di-

14 It is strange, however, that Roemer (1995)should characterize Stiglitz’ book “An Anti-HayekManifesto”rather than, say, “An Anti-Walras Mani-festo.”Roemer bases his characterization on a fewremarks made by Stiglitz in one of his last chap-ters, in which Hayek’s The Fatal Conceit is brieflycritically discussed. It does not seem to this readerthat either Roemer or Stiglitz has fully compre-hended Hayek’s evolutionary arguments.

15 To be sure, some of the points mentioned be-low have been made by information theorists, in-cluding Stiglitz, whose remarks (1994, p. 134) onthe importance of the institutional framework inwhich competition is embedded could have beenwritten by Hayek. What is less clear is how suchinsights follow analytically from the models foundin the economics of information literature.

One referee commented that many of the Austrianinsights have been incorporated within other areas ineconomics, such as in the “bounded rationality” litera-ture. It may be noted, however, that Thomsen (1992,ch. 4) argues that Austrian market process theory dif-fers significantly from Herbert Simon’s early work onbounded rationality, and Esther-Mirjam Sent (1997)suggests that the more recent literature on boundedrationality may face historiographical problems of itsown.


rectly communicable. Theories thattreat “information” as if it was con-tained in little packets that can be “elic-ited” once the appropriate mechanismis “designed” misunderstand the funda-mental fact that no mechanism canelicit tacit knowledge. Furthermore,tacit knowledge is important: it affectsand informs the decisions of entrepre-neurs and ultimately gets reflected inthe prices that emerge in a competitiveprice system. In proposed systems inwhich entrepreneurial decision makingis eliminated, or replaced by managerialprocesses, this knowledge gets lost. Onestill has “prices” in such systems, butthey contain less “information.”

The mainstream often focuses on howmarkets “convey information.” ThoughHayek considered this role important,he was concerned additionally with thecreation, discovery, and conservation ofknowledge. Furthermore, the Austriannotion of “discovery” is quite differentfrom the neoclassical idea of “search.”A neoclassical economist like Stiglitz(1994, p. 8) can ask: Are “the expendi-tures on information acquisition too lit-tle, too much, or just right?” To answerthis question one must be able to com-pare the expected costs of additionalsearch against expected benefits. Con-trast this with Kirzner’s (1997, p. 62)characterization of entrepreneurial dis-covery:

For the Austrian approach imperfect informa-tion is seen as involving an element whichcannot be fitted at all into neoclassical mod-els, that of “sheer” (i.e., unknown) ignorance.. . . the discovery which reduces sheer igno-rance is necessarily accompanied by the ele-ment of surprise . . . one had not hithertorealized one’s ignorance. Entrepreneurial dis-covery is seen as gradually but systematicallypushing back the boundaries of sheer igno-rance . . . (emphasis in the original)

2. The market process is not wellcaptured by mainstream theories of

competition. Hayek, like modern infor-mation theorists, believed that thetheoretical tools of his day were inade-quate for understanding the limitationsof socialism, and he identified the unre-alistic treatment of knowledge as aprime defect. But his response was notto add a more sophisticated account ofinformation to the neoclassical model.Rather, he rejected the ideas of “given”data and of perfect competition, puttingin their place the notion of a dynamicmarket process in which rivalry amongparticipants ensures that knowledge iscontinually generated and discovered.In doing so, Hayek was actually harkingback to an earlier tradition, one whoseeclipse in the twentieth century is dem-onstrated, and lamented, by FrankMachovec (1995).16

The Austrian analysis of the marketprocess is not utopian; there is no guar-antee of coordination. But it does insistthat a system in which prices deter-mined in competitive markets are freeto adjust to reflect relative scarcities isone in which the coordination of agents’plans is least likely to be hindered. So-viet-style central planning ignored thisinsight completely, but the “trial anderror” solutions proposed by Lange andothers also misunderstand the point. Allsuch proposals underestimate the diffi-culty of playing “catch-up” with a sys-tem in which the acts of millions of in-dividual decision makers produce acontinuous process of adjustment.

Competition is a key institutional fea-ture within a market system. It isneeded to provide incentives, as theeconomics of information emphasizes.But the “discipline of the marketplace”also assists the process of discovery: of

16 Makowski and Ostroy (1993, p. 74) refer tocompetition as market rivalry, but it is less clearfrom their brief paper how what they call an “oldand seminal idea” links up to the analytics of theirPerfectly Competitive Equilibrium model.


errors, of new processes and products,of knowledge itself.

Competition never has been “per-fect,” nor need it be. The dual claim ofthe early market socialists (that atomis-tic competition was necessary for a mar-ket system to work, and that it had dis-appeared) was not true when they madeit, and subsequent history has done lit-tle to support a revised judgment. Inparticular, perfect competition is an in-appropriate benchmark for policy, onethat is bound to mislead us into think-ing that we can accomplish much morewith interventions than we really can.To the extent that Austrians accept anefficiency welfare norm, their emphasisis on the dynamic adaptive efficiency ofa system (that is, how fast it adapts tothe introduction of emerging or new in-formation) rather than on static alloca-tional efficiency, which focuses on theuse of existing, known information.

In the real world errors occur all thetime. Mainstream theorists typicallytreat errors as disequilibrium phenom-ena, so their models seldom directly in-corporate this ubiquitous aspect of real-ity. But as Thomsen (1992) and Kirzner(1997) emphasize, the actions of entre-preneurs in disequilibrium situationsconstitute a fundamental aspect of theAustrian characterization of the marketprocess. Like their real world counter-parts, the entrepreneurial decision mak-ers in the Austrian account also makeerrors all the time. These errors resultin profit opportunities, and the searchfor profits works to reveal and eliminatedivergent expectations. The mainstreamemphasis on equilibrium states ob-scures the fact that one of the chief ad-vantages of a market system is that itcontains a built-in mechanism, imper-fect to be sure, but constantly operat-ing, for the correction of errors. Froman Austrian perspective, mainstreameconomists tend to overestimate the in-

formational content of equilibriumprices, and to underestimate the infor-mational content of disequilibriumprices (Boettke 1996, pp. 189−90).

3. The rejection of homo economicus.Though homo economicus is a crucialassumption of neoclassical analysis,Hayek had doubts about its usefulness,and even ridiculed it on occasion (1960,pp. 60−61). Rather than continuing toemploy the construct, he sought to pro-vide in his psychological work a physi-ological foundation for perception and,ultimately, for knowledge formation.The mind of the individual described inThe Sensory Order is a complex adap-tive self-organizing neural order, andthis is what leads to differences in per-ceptions and beliefs among people, toour ability to adapt to new environ-ments, and to the possibility of thegrowth of our knowledge. This is a dif-ferent creature altogether from homoeconomicus.

One way to put the contrast is as fol-lows. In standard neoclassical analysis,one starts out with rational agents whoby assumption optimize over all therelevant margins, including those thathave to do with acquiring more infor-mation. In cases in which information isasymmetric, the chief goal is to designmechanisms that will either elicit infor-mation or provide an incentive struc-ture to prevent inefficient social out-comes. If the world were reallypopulated by agents resembling homoeconomicus, such approaches mightmake some sense.

For Hayek, though, the world doesnot look like this. Humans are purpose-ful but imperfect beings, they have lim-ited knowledge, they make mistakes,and whatever knowledge that exists isdispersed, fragmented, and often diffi-cult to communicate. Many of thethings that neoclassical economics as-sumes (at the most fundamental level,


the ability of individuals to engage inrational behavior, and to perform mar-ginal analysis) are, from the Austrianperspective, properly viewed as the re-sults of certain institutional arrange-ments rather than as assumptions ofthe analysis (Hayek 1979, pp. 74−77;Boettke, forthcoming). Austrians ask:what combination of institutions canbest assist imperfect individuals tomake better decisions and better use oftheir knowledge? By starting with ra-tional agents, standard analysis getsthings exactly backwards.

4. The role of institutions. Instead ofstudying the design of alternative allo-cation mechanisms (a project thatwould anyway smack of rationalist con-structivism to him), Hayek proposedthat we examine the role of various in-stitutions in assisting the creation, dis-covery, use, conveyance, and conserva-tion of knowledge. This led him to makesome proposals of his own. It was hisfocus on knowledge that led Hayek toendorse a system of free markets, oper-ating within a liberal democratic polity,a system with established, enforced andexchangeable property rights, all pro-tected by a strong constitution.

Though he made proposals, Hayekalso recognized that the conscious con-struction or imposition of social institu-tions is a tricky business. Many such in-stitutions are the product of a longprocess of evolutionary development;they are themselves examples of com-plex self-organized adaptive orders.They have histories, and they performfunctions that are not well-understoodby outside observers. Their emergencewas certainly not inevitable, nor is theircontinuance: this led Hayek to explorewhy certain institutional forms persist,even though they might not accord wellwith either our instincts or our reason.But he also saw that those societies inwhich the right combination of liberal

institutions happened to emerge haveboth prospered and permitted a consid-erable amount of individual liberty tobe exercised.

Attempts to alter such institutions, beit by piece-meal or by full-scale socialengineering, will often generate unin-tended, unanticipated, and unwelcomeconsequences. “Mechanism design” maybe feasible within firms that then facethe test of market competition. It ismuch less likely that one can success-fully mimic markets as a whole, or re-design entire societies. As a generalrule, those who hope to redesign insti-tutions are over-optimistic about theamount of knowledge that is available tothem. It is precisely because knowledgeis not readily available that evolutionaryprocesses typically do better than ourreason in the area of institutional de-sign. This argument, if true, shouldmake one pessimistic about the pros-pects for socialism. But it should alsogive pause to market reformers whohope overnight to install new institu-tions in countries formerly under com-munist rule. If such reforms are to haveany hope of working, a sensitivity to thepreviously existing institutional frame-work is critical. Bardhan and Roemer(1993, p. 16), Boettke (1994), and PeterMurrell (1995, pp. 175−77) offer perti-nent observations.

In the end Hayek concluded thatmany institutions, from our moral codesto the market system, are themselvesexamples of complex adaptive phenom-ena. In this regard, though he may haveshared the complaints of the informa-tion theorists about static equilibriummodels, Hayek’s basic observations aremore in line with recent developmentsin complexity theory and evolutionarybiology than they are with the later de-velopment of neoclassical theory. Thesimilarity between Hayek’s work andthese other areas of research has so far


only just begun to be explored (Lavoie1989; Birner, forthcoming; Mirowski,forthcoming; Ransom 1992, 1996).

VI. Conclusion

Though the Austrians knew about in-centives problem, they did not system-atically pursue them. For modern dayinformation theorists, this means that,no matter how prescient their conclu-sions may have been, Austrian analysisis of limited value. If the Austrians wereright, they were right for the wrong rea-sons, because their models ignored theinsights revealed by the economics ofinformation.

This final section suggests at a mini-mum that modern information theoristshave not fully understood why the Aus-trians chose not to pursue the incen-tives problem. As a doctrinal historian,I cannot help adding that such misun-derstanding is virtually inevitable in aprofession that, at least in the UnitedStates, has nearly systematically elimi-nated the study of history of economicthought from graduate curricula.

But a different point was also made.Getting the history right matters. Itmatters not only for historical accuracy,but because one can draw the wrongconclusions if one only has recourse toWhig history. If one reads one’s historybackwards from the present, where to-day’s theoretical analysis provides thesole possible benchmark against whichto assess a program, then only one con-clusion is possible about Austrian“analysis”: it is deficient. But if onegoes back and looks at the record, whatis immediately apparent is that the Aus-trians went another way. Their path re-ally was different. Furthermore, themainstream might have something tolearn from that alternative path.

The virtue of the economics of infor-mation is to permit economists to dis-

cuss questions of agency with analyticalprecision and rigor. Like all theories, itsanalytic rigor is purchased at the priceof an oversimplified picture of reality.This poses a danger if it leads one tothink that the only problems facing so-cialism are agency problems, problemsthat can be overcome once the requisiteincentive compatibility mechanisms areput into place.

Hayek was no opponent of theory; in-deed, he frequently defended it from itshistoricist detractors. But he also un-derstood the limitations of theory. Ahalf century ago well-intentioned social-ists demonstrated with simplisticmathematical models that market so-cialism could duplicate the workings ofa competitive market system, plus re-move its deficiencies. More elegantmodels are available today for correct-ing a system with informational prob-lems. Over 60 years ago Hayek warnedabout the dangers of an “excessive pre-occupation with the conditions of a hy-pothetical state of stationary equilib-rium” (1935, p. 226). It is not altogetheroutrageous to suggest that today hemight warn against an excessive preoc-cupation with questions of information.

REFERENCES

ADAMAN, FIKRET AND DEVINE, PAT. “The Eco-nomic Calculation Debate: Lessons for Social-ists,” Cambridge J. Econ., Sept. 1996, 20(5), pp.523−37.

ARNOLD, N. SCOTT. “The Monitoring Problem forMarket Socialist Firms,” in PETER J. BOETTKEAND DAVID L. PRYCHITKO, eds., 1996, pp. 41−58.

BARDHAN, PRANAB, AND ROEMER, JOHN E., eds.Market socialism. New York and Oxford: OxfordU. Press, 1993.

———. “On the Workability of Market Socialism,”J. Econ. Perspectives, Spring 1994, 8(2), pp.177−81.

BEAN, PHILIP AND WHYNES, DAVID, eds. BarbaraWootton—Social science and public policy: Es-says in her honour. London: Tavistock, 1986.

BERGSON, ABRAM. “Socialist Economics,” in aSurvey of contemporary economics. Ed.:HOWARD ELLIS. Homewood, IL: Irwin, 1948,pp. 412−48.


BERLE, ADOLF A. AND MEANS, GARDINER C. The modern corporation and private property.New York: Macmillan, 1933.

BIRNER, JACK. “Introduction: Hayek’s Grand Re-search Programme,” in JACK BIRNER ANDRUDY VAN ZIJP, eds.London: Routledge, 1994,pp. 1−21.

———. “The Surprising Place of Cognitive Psy-chology in the Work of F. A. Hayek,” Hist.Econ. Ideas, forthcoming.

BIRNER, JACK AND VAN ZIJP, RUDY, ed. Hayek,co-ordination and evolution: His legacy in phi-losophy, politics, economics and the history ofideas. London and New York: Routledge, 1994.

BOEHM, STEPHEN. “Hayek on Knowledge, Equi-librium and Prices,” Wirtschafts Politische Blät-ter, 1989, 36(2), pp. 201−13.

BOETTKE, PETER J. “The Reform Trap in Eco-nomics and Politics in the Former CommunistCountries,” Journal des Economistes et desEtudes Humaines, June/Sept. 1994, 5(2/3), pp.267−93.

———. “Hayek’s The Road to Serfdom Revisited:Government Failure in the Argument AgainstSocialism,” Eastern Econ. J., Winter 1995,21(1), pp. 7−26.

———. “Book Review: Joseph Stiglitz’s WhitherSocialism?” J. Econ. Lit., Mar. 1996, 34(1), pp.189−91.

———. “Economic Calculation: The AustrianContribution to Political Economy,” in Ad-vances in Austrian economics. Vol. 5. Green-wich, CT: JAI Press, forthcoming.

BOETTKE, PETER J. AND PRYCHITKO, DAVID L.,eds. Advances in Austrian economics. Vol 3.Greenwich, CT: JAI Press, 1996.

BURNHAM, JAMES. The managerial revolution:What is happening in the world right now. NewYork: John Day, 1941.

BUTOS, WILLIAM N. AND KOPPL, ROGER G.“Hayekian Expectations: Theory and EmpiricalApplications,” Constitutional Polit. Econ., Fall1993, 4(3), pp. 303−29.

CALDWELL, BRUCE J. Beyond positivism: Eco-nomic methodology in the twentieth century.London: Routledge, [1982] 1994.

———. “Hayek’s Transformation,” Hist. Polit.Econ., Winter 1988a, 20(4), pp. 513−41.

———. “Hayek’s ‘The Trend of Economic Think-ing’,” in The review of Austrian economics. Vol.2. Eds.: MURRAY N. ROTHBARD AND WALTERBLOCK. Washington, DC: Ludwig von Mises In-stitute, 1988b, pp. 175−78.

———. “La Méthodologie De Hayek: Description,Evaluation, et Interrogations,” in FriedrichHayek: Philosophie, economie et politique. Eds.:GILLES DOSTALER ET DIANE ETHIER. Paris:Economica, 1989, pp. 71−85.

CARR, EDWARD H. Nationalism and after. Lon-don: Macmillan, 1945.

CARTWRIGHT, NANCY ET AL. Otto Neurath: Phi-losophy between science and politics. Cam-bridge: Cambridge U. Press, 1996.

CHALOUPEK, GÜNTHER K. “The Austrian Debate

on Economic Calculation in a Socialist Society,”Hist. Polit. Econ., Winter 1990, 22(4), pp. 659−75.

CLARKE, F. “Book Review: Mannheim’s Man andSociety in an Age of Reconstruction,”Economica, N.S., Aug. 1940, 7(27), pp. 329−32.

COTTRELL, ALLIN AND COCKSHOTT, W. PAUL.“Calculation, Complexity and Planning: the So-cialist Calculation Debate Once Again,” Rev.Polit. Econ., 1993, 5(1), Pp. 73−112.

CRAVER, EARLENE. “The Emigration of the Aus-trian Economists,” Hist. Polit. Econ., Spring1986, 18(1), pp. 1−32.

DESAI, MEGHNAD. “Equilbrium, Expectationsand Knowledge,” in JACK BIRNER AND RUDYVAN ZIJP, eds. 1994, pp. 25−50.

DIAMOND, ARTHUR. “F. A. Hayek on Constructiv-ism and Ethics,” J. Libertarian Stud., Fall 1980,4(4), pp. 353−65.

DICKINSON, HENRY D. “Price Formation in a So-cialist Economy,” Econ. J., June 1933, 43(170),pp. 237−50.

———. “Review: Hayek’s Freedom and the Eco-nomic System,” Economica, N.S., Nov. 1940,7(28), pp. 435−37.

DOBB, MAURICE. “Economic Theory and theProblems of a Socialist Economy,” Econ. J.,Dec. 1933, 43(172), pp. 588−98.

DURBIN, ELIZABETH F. New Jerusalems: The La-bour Party and the economics of democratic so-cialism. London: Routledge, 1985.

DURBIN, EVAN. “Professor Hayek on EconomicPlanning and Political Liberty,” Econ. J., Dec.1945, 55(220), pp. 357−70.

EDELMAN, GERALD M. Neural Darwinism: Thetheory of neuronal group selection. NewYork:Basic Books, 1987.

FINER, HERMAN. The road to reaction. Chicago:Quadrangle Books, 1945.

FLEETWOOD, STEVE. Hayek’s political economy:The socio-economics of order. London: Rout-ledge, 1995.

GRAY, JOHN. Hayek on liberty. 2nd ed. Oxford:Blackwell, 1986.

———. “Hayek, the Scottish School, and Contem-porary Economics,” in The boundaries of eco-nomics. Eds.: GORDON C. WINSTON ANDRICHARD F. TEICHGRAEBER. Cambridge:Cambridge U. Press, 1988, pp. 53−70.

GROSSMAN, SANFORD J. The informational role ofprices. Cambridge, MA: MIT Press, 1989.

HAYEK, FRIEDRICH A. VON. “The Trend of Eco-nomic Thinking,” [1933]; reprinted in FRIED-RICH VON HAYEK 1991.

———, ed. Collectivist economic planning: Criti-cal studies on the possibilities of socialism. Lon-don: G. Routledge, 1935.

———. “Economics and Knowledge,” [1937]; re-printed in FRIEDRICH VON HAYEK, ed. 1948,pp. 33−56.

———. “Freedom and the Economic System,”[1939]; reprinted in FRIEDRICH VON HAYEK1997, pp. 189−211.

———. “Socialist Calculation: the Competitive


‘Solution’,” [1940]; reprinted in FRIEDRICHVON HAYEK, ed. 1948, pp. 181−208.

———. “Planning, Science, and Freedom,”[1941]; reprinted FRIEDRICH VON HAYEK1997, pp. 213−20.

———. The counter-revolution of science: Studieson the abuse of reason. Glencoe: IL: Free Press,[1941−44] 1952b.

———. The road to serfdom. Chicago: U. of Chi-cago Press, [1944] 1976.

———. “The Use of Knowledge in Society,”[1945]; reprinted in FRIEDRICH VON HAYEK,ed. 1948, pp. 77−91.

———. “The Meaning of Competition,” [1946];reprinted in FRIEDRICH VON HAYEK, ed. 1948,pp. 92−106.

———, ed. Individualism and economic order.Chicago: U. of Chicago Press, 1948.

———. Letter to John Nef, Dated November 6,1948, Located in Box 55, Number 1, in theHayek Archives, Hoover Institution, Stanford,CA.

———. The sensory order: An inquiry into thefoundations of theoretical psychology. Chicago:U. of Chicago Press, 1952a.

———, ed. Capitalism and the historians. Chi-cago: U. of Chicago Press, 1954.

———. “Degrees of Explanation,” [1955]; re-printed in FRIEDRICH VON HAYEK, ed. 1967,pp. 3−21.

———. The constitution of liberty. Chicago: U. ofChicago Press, 1960.

———. “Rules, Perception and Intelligibility,”[1962]; reprinted in FRIEDRICH VON HAYEK,ed. 1967, pp. 43−65.

———. “Kinds of Rationalism,” [1964a]; reprintedin FRIEDRICH VON HAYEK, ed. 1967, pp. 82−95.

———. “The Theory of Complex Phenomena,”[1964b]; reprinted in FRIEDRICH VON HAYEK,ed. 1967, pp. 22−42.

———, ed. Studies in philosophy, politics andeconomics. Chicago: U. of Chicago Press, 1967.

———. “Competition as a Discovery Procedure,”[1968a], in FRIEDRICH VON HAYEK, ed. 1978,pp. 179−90.

———. “The Primacy of the Abstract,” [1968b],in FRIEDRICH VON HAYEK, ed. 1978, pp. 35−49.

———. “The Errors of Constructivism,” [1970], inFRIEDRICH VON HAYEK, ed. 1978, pp. 3−22.

———. Law, legislation and liberty: A new state-ment of the liberal principles of justice and po-litical economy. Vols. 1−3. Chicago: U. of Chi-cago Press, 1973−1979.

———. “The New Confusion about ‘Planning’,”[1976], in FRIEDRICH VON HAYEK, ed. 1978,pp. 232−46.

———, ed. New studies in philosophy, politics,economics and the history of ideas. Chicago: U.of Chicago Press, 1978.

———. “The Sensory Order after 25 Years,” inCognition and the symbolic processes. Vol. 2.Eds.: WALTER WEIMER AND DAVID PALERMO.

Hillsdale, NJ: Lawrence Erlbaum Associates,1982a, pp. 287−93.

———. “Two Pages of Fiction: the Impossibilityof Socialist Calculation,” in The essence ofHayek. Eds.: CHIAKI NISHIYAMA AND KURT R.LEUBE. Stanford: Hoover Institution Press,[1982b] 1984a, pp. 53−61.

———. “Nobel Prize-winning Economist.” Tran-script of an Oral History Interview Conductedin 1978 under the Auspices of the Oral HistoryProgram, University Library, UCLA, Regents ofthe U. of California, copyright 1983.

———. Money, capital, and fluctuations: Early es-says. Ed.: ROY MCCLOUGHRY. Chicago: U. ofChicago Press, 1984b.

———.The collected works of F. A. Hayek. Vol. 1.The fatal conceit: The errors of socialism. Ed.:W. W. BARTLEY. London: Routledge, 1988.

———. The collected works of F. A. Hayek. Vol. 3.The trend of economic thinking: Essays on po-litical economists and economic history. Eds.:W. W. BARTLEY III AND STEPHEN KRESGE.Chicago: U. of Chicago Press, and London:Routledge, 1991.

———. The collected works of F. A. Hayek. Vol. 4.The fortunes of liberalism: Essays on Austrianeconomics and the ideal of freedom. Ed.: PETERG. KLEIN. Chicago: U. of Chicago Press, 1992.

———. Hayek on Hayek. Eds.: STEPHEN KRESGEAND LEIF WENAR. Chicago: U. of ChicagoPress and London: Routledge, 1994.

———. The collected works of F. A. Hayek. Vol. 9.Contra Keynes and Cambridge: Essays, corre-spondence. Ed.: BRUCE CALDWELL. London:Routledge, 1995.

———. The collected works of F. A. Hayek. Vol.10. Socialism and war: Essays, documents, re-views. Ed.: BRUCE CALDWELL. London: Rout-ledge, 1997.

HEBB, DONALD O. The organization of behavior:A neuropsychological theory. N.Y.: Wiley, 1949.

HEILBRONER, ROBERT. “Analysis and Vision inthe History of Modern Economic Thought,” J.Econ. Lit., Sept. 1990, 28(3), pp. 1097−1114.

HOFF, TRYGVE J. B. Economic calculation in thesocialist society. London: W. Hodge, 1949.

HORWITZ, STEVEN. “From The Sensory Order tothe Liberal Order: Mind, Economy and theState in the Thought of F. A. Hayek.” A paperpresented at the Mount Pélèrin meetings, Can-nes, France, 1994.

———. Money, Money Prices, and the SocialistCalculation Debate,” in PETER J. BOETTKEAND DAVID L. PRYCHITKO, eds., 1996, pp. 59−77.

HURWICZ, LEONID. “Economic Planning and theKnowledge Problem: A Comment,” Cato J., Fall1984, 4(2), pp. 419−25.

KIRZNER, ISRAEL M. Competition and en-trepreneurship. Chicago: U. Chicago Press,1973.

———. “Economic Planning and the KnowledgeProblem,” Cato J., Fall 1984, 4(2), pp. 407−18.


———. “The Economic Calculation Debate: Les-sons for Austrians,” The review of Austrian eco-nomics. Vol. 2. Eds.: MURRAY N. ROTHBARDAND WALTER BLOCK. Washington, DC: Lud-wig von Mises Institute, 1988, pp. 1−18.

———. “Entrepreneurial Discovery and theCompetitive Market Process: an Austrian Ap-proach,” J. Econ. Lit., Mar. 1997, 35(1), pp. 60−85.

KORNAI, JANgOS. “The Soft Budget Constraint,”Kyklos, 1986, 39(1), pp. 3−30.

———. “Market Socialism Revisited,” in PRAN-HAB K. BARDHAN AND JOHN E. ROEMER, eds.1993, pp. 42−68.

KUKATHAS, CHANDRAN. Hayek and modern lib-eralism. Oxford: Clarendon, 1989.

LACHMANN, LUDWIG M. “On the Central Con-cept of Austrian Economics: Market Process,”in The foundations of modern Austrian econom-ics. Ed.: EDWIN G. DOLAN. Kansas City: Sheedand Ward, 1976, pp. 126−32.

LANGE, OSKAR. On the economic theory of social-ism. Ed.: BENJAMIN E. LIPPINCOTT. Minne-apolis: U. of Minnesota Press, 1938, pp. 57−143.

———. “The Computer and the Market,” in So-cialism, capitalism and economic growth: Es-says presented to Maurice Dobb. Ed.: C. H.FEINSTEIN. Cambridge: Cambridge U. Press,1967, pp. 158−161.

LAVOIE, DON. Rivalry and central planning.Cambridge: Cambridge U. Press, 1985.

———. “Economic Chaos or Spontaneous Order?Implications for Political Economy of the NewView of Science,” Cato J., Winter 1989, 8(3),pp. 613−35.

———. “Computation, Incentives, and Discovery:the Cognitive Function of Markets in MarketSocialism,” Ann. Amer. Acad. Polit. Soc. Sci.,Jan. 1990, 507, pp. 72−79.

LAWSON, TONY. “Realism and Hayek: A Case ofContinuing Transformation,” in The economicsof F. A. Hayek. Vol. 2. Capitalism, socialism,and knowledge. Eds.: MARINA COLONNA,HARALD HAGEMANN, AND OMAR F.HAMOUDA. Aldershot: Elgar, 1994, pp. 131−59.

LIPPMAN, WALTER. An inquiry into the principlesof the good society. Boston: Little, Brown &Co., 1937.

MACHOVEC, FRANK M. Perfect competition andthe transformation of economics. London: Rout-ledge, 1995.

MACMILLAN, HAROLD. The middle way: A studyof the problem of economic and social progressin a free and democratic society. London:Macmillan, 1938.

MAKOWSKI, LOUIS AND OSTROY, JOSEPH. “Gen-eral Equilibrium and Market Socialism: Clarify-ing the Logic of Competitive Markets,” inPRANHAB K. BARDHAN AND JOHN E. ROEMER,eds. 1993, pp. 69−88.

MALLOCK, WILLIAM H. A critical examination ofsocialism. NY: Harper, 1907.

MANNHEIM, KARL. Man and society in an age ofreconstruction: Studies in modern social struc-ture. London: Kegan Paul, 1940.

MARWICK, ARTHUR. “Middle Opinion in the Thir-ties: Planning, Progress and Political ‘Agree-ment’,” English Hist. Rev., Apr. 1964, 79(311),pp. 285−98.

MENGER, CARL. Investigations into the method ofthe social sciences with special reference to eco-nomics. Translated by FRANCIS J. NOCK. NewYork: New York U. Press, [1883] 1985.

MERKLE, JUDITH A. Management and ideology:The legacy of the international scientific man-agement movement. Berkeley: U. of CaliforniaPress, 1980.

MIKLÒS-ILLÈS, AGNES. “Otto Neurath’s War Eco-nomics.” Manuscript, 1996.

MILGROM, PAUL AND ROBERTS, JOHN. Econom-ics, organization and management. EnglewoodCliffs, NJ: Prentice Hall, 1992.

MIROWSKI, PHILIP. “Machine Dreams: EconomicAgents as Cyborgs,” in The new economics andits history. Ed.: JOHN DAVIS. Durham, NC:Duke U. Press, forthcoming.

MISES, LUDWIG VON. “Economic Calculation inthe Socialist Commonwealth,” in FRIEDRICHVON HAYEK, ed., 1935, pp. 87−130.

———. Human action: A treatise on economics.3rd rev. ed. Chicago: H. Regnery Co., 1966.

———. Notes and recollections. South Holland,IL: Libertarian Press, 1978.

MURRELL, PETER. “The Transition According toCambridge, Mass.,” J. Econ. Lit., Mar. 1995,33(1), pp. 164−78.

NEURATH, OTTO. Empiricism and sociology. Eds:MARIE NEURATH AND ROBERT S. COHEN.Dordrecht, Holland: D. Reidel, 1973.

POLANYI, KARL. The great transformation. NewYork: Farrar & Rinehart, 1944.

RAE, JOHN. Contemporary socialism. 2nd ed. NewYork: C. Scribner’s Sons, 1891.

RANSOM, GREG. “Insuperable Limits to Reduc-tion in Biology.” Manuscript, Apr. 1992.

———. “The Significance of Myth and Misunder-standing in Social Science Narrative: OpeningAccess to Hayek’s ‘Copernican’ Revolution inEconomics.” Manuscript, June 1996.

ROEMER, JOHN E. “Can There Be Socialism afterCommunism?” in PRANHAB K. BARDHAN ANDJOHN ROEMER, eds. 1993, pp. 89−107.

———. A future for socialism. Cambridge, MA:Harvard U. Press, 1994.

———. “An Anti-Hayek Manifesto,” New Left Re-view, May-June 1995, 211, pp. 112−29.

SALERNO, JOSEPH. “Ludwig von Mises as SocialRationalist,” The review of Austrian economics.Vol. 4. Eds.: MURRAY N. ROTHBARD AND WAL-TER BLOCK. Dordrecht and London: KluwerAcademic, 1990, pp. 26−54.

SCHUMPETER, JOSEPH. Capitalism, socialism anddemocracy. 3rd ed. New York: Harper & Row,[1942] 1950.

SENT, ESTHER-MIRJAM. “Bounded Rationality onthe Rebound.” Manuscript, 1997.


SHEARMUR, JEREMY. Hayek and after: Hayekianliberalism as a research programme. London:Routledge, 1996.

———. “Hayek, Keynes and the State,” Hist.Econ. Rev., Winter 1997, 26, pp. 68−80.

STEELE, DAVID RAMSAY. From Marx to Mises:Post-capitalist society and the challenge of eco-nomic calculation. La Salle, Il: Open Court,1992.

STIGLER, GEORGE. Memoirs of an unregulatedeconomist. NY: Basic Books, 1988.

STIGLITZ, JOSEPH. Whither socialism? Cam-bridge, MA and London: MIT Press, 1994.

THOMSEN, ESTEBAN F. Prices and knowledge: Amarket process perspective. London and NewYork: Routledge, 1992.

VANBERG, VIKTOR. Rules and choice in economics.London: Routledge, 1994.

WEBB, SIDNEY AND WEBB, BEATRICE. Soviet

communism: A new civilization? London: Long-mans, Green, 1935.

WEBER, MAX. Economy and society: An outline ofinterpretive sociology. Eds.: GUENTHER ROTHAND CLAUS WITTICH. Berkeley: U. of Califor-nia Press, 1978.

WERSKEY, GARY. The visible college: A collectivebiography of British scientific socialists of the1930s. New York: Holt, Rinehart and Winston,1978.

WOOTTON, BARBARA. Freedom under planning.Chapel Hill, NC.: U. of North Carolina Press,1945.

YEAGER, LELAND B. “Mises and Hayek on Calcu-lation and Knowledge,” Rev. Austrian Econ.,1994, 7(2), pp. 93−109.

———. “Salerno on Calculation, Knowledge, andAppraisement: Rejoinder,” Rev. Austrian Econ.,1996, 9(1), pp. 137−39.


Caldwell: Hayek and Socialism Hayek and Socialism

Documents

Transcript of Caldwell: Hayek and Socialism Hayek and Socialism