ama gi a - Lancaster UniversityThe logo and name of ama-gi is taken from a cuneiform inscription...

ama-giThe Journal of the Hayek Society at the London School of Economics

2 Globalization Through Human Action or Human Design? BRUCE L. BENSON

4 Hayek�s Sensory Order GERALD R. STEELE

7 Philosophical Speculations adapted from Whither Socialism? JOSEPH E. STIGLITZ

9 Review of John Kay�s The Truth About Markets: Their Genius, Their Follies, Their Limitations DAVID HENDERSON

10 The Risk of Paternalism adapted from Philosophy, Policy and Social Value MSc thesis PETER ROSSI

13 Capitalism Works SIR GEOFFERY OWEN

15 Steel Not Getting it Right VLADIMIR L. ANDONOV

18 Reforming Governments PHILIP E. GRAVES

20 Returning Planning to the Market: An Agenda for Private Land Use Control MARK PENNINGTON

Note from the Editor

My decision to study in London this yearhas brought me considerable grief. ComeJune, I will have to leave.

My experience at the London School ofEconomics has been just as I imagined-herds of intelligent students constantlyseeking and absorbing life and knowledge.They just don’t know when to stop. Rightat this very moment, the library is probablyfilled to the brim with them. And as clichéas that image is, there’s more: the diversityis overwhelming. Half the students are fromoutside of the United Kingdom; studentscome from every imaginable place and bring

with them every imaginable idea on politics,economics, philosophy, etc. Often timesviews merge and all is swell but the besttimes are when the ideas clash- when aspark of an idea provokes us to think. TheHayek Society strikes me as a particularlypotent example of this creation process. Itprovides students with a forum to thinkaloud and figure themselves out. Theconcepts of capitalism, socialism, freemarkets and equality, as integral as theyare to our studies, require suchinvestigation. We are not here to poundlibertarian ideals into heads or to disparage

those who lean to the far left; we are justhere to talk and to encourage others to talk.Without this freedom to be an individualthe foundations of our society would meannothing.

In this issue of ama-gi, articles reachbackward to a scientific basis for economicphilosophies and forward to policyproposals on land use and government, allin the spirit of Hayekian thought to provideour peers with a current look at classicalliberalism. Enjoy.

Erica C. Yu

ama-giHayek Society

www.hayek.org.uk

London School of EconomicsHoughton Street, London WC2 2AE

Lent Term 2004 Volume 6, Issue 1

Editor-in-ChiefErica C. Yu

Associate EditorsHarry CherniakMichael Chen

The opinions expressed herein are expressly those of theauthors, and do not necessarily reflect the views of theHayek Society or the London School of Economics.

The logo and name of ama-gi is taken from a cuneiforminscription found in ancient Sumeria. It is the earliestwritten representation of the concept of �freedom.�

2 ama -gi The Journal of the Hayek Society at LSE

behavioral patterns that other individualsexpect a person to adopt and follow in thecontext of various interdependent activitiesand actions: that is, rules specifyobligations. The rules one individual isexpected to follow influence the choicesmade by other individuals: like prices, rulescoordinate and motivate interdependentbehavior.

Some rules are designed and imposed bystate rule-making authorities, but there areother rules too, that are actually much moreimportant determinants of behavior in manyaspects of human activity, includingcommercial activity. Indeed, a keydistinguishing characteristic of such rulesis that they are initiated by an individual’sdecision to behave in particular ways underparticular circumstances. Hayekemphasizes that adopting a behavioralpattern creates expectations for others whoobserve it and thus creates an obligationto live up to those expectations.Furthermore, when individuals who interactwith one another observe each other’sbehavioral patterns they often emulatethose that appear desirable so that suchbehavior and accompanying obligationsspread. In other words, customary normsevolve spontaneously from the bottom uprather than being intentionally imposed bya legislator, and they are voluntarilyaccepted rather than imposed, even thoughthere may never have been an explicitstatement declaring that they are relevant.For an obligation to achieve the status of a“customary law” it must be widelyrecognized and accepted by the individualsin the affected group or community.

Customary law tends to be quiteconservative in the sense that it guardsagainst mistakes, leading some to believethat customary law is very slow to change,and thus “inefficient” in a dynamicenvironment. Certainly, this process can bea very important mechanism for creatingnew custom, but there also are fastermechanisms for initiating change incustomary law.

If conditions change and a set ofindividuals decide that, for their purposes,behavior that was attractive in the past hasceased to be useful, they can voluntarilydevise a new contract (explicit or implicit)

stipulating any behavior that they wish.Through negotiation and contracting,existing custom can be quickly replaced bya new rule of obligation toward certain otherindividuals without prior consent of orsimultaneous recognition by everyone inthe group. Individuals entering intocontracts with these parties are informedof the contractual innovation, however,and/or others outside the contract observethe results of a new contractual stipulation,so if it provides a more desirable behaviorrule than older customs do, it can be rapidlyemulated. Contracting may actually be themost important source of new rules in adynamic system of customary rules, but asconditions change, the inadequacy ofexisting customary rules also can berevealed when a dispute arises. If theparties to a dispute turn to a third party forarbitration, as they do in cases withinmodern international commercialcommunities, for instance, new customaryrules can be and often are initiated by thearbitrator’s resolution to the disputes.Unlike public court precedent, suchresolutions only apply to the parties in thedispute, but if a rule is implied that appearsto be more effective at facilitatinginteractions than previously existingcustomary rules have been, the rule canspread rapidly through the interactingcommunity.

Does Enforcement of Law Require State-Backed Coercive Power?

When individuals specialize and trade,depersonalized exchange of simple goodsmay occur if the good’s characteristics areeasily observed and quality can bedetermined before a purchase (e.g., as inan informal urban market for fruits andvegetables), but for more complex difficult-to-measure goods, the seller often knowsmore about the product and its quality thanthe buyer. Under these circumstances, thecost of the exchange for the buyer will berelatively high if he cannot believe theclaims or promises of the seller; the buyerwill have to invest in obtaining moreinformation, bear the risk of making anundesirable purchase or forgo what mightbe an advantageous exchange. Under suchconditions people may make small riskypurchases in hopes of getting somethingthat improves their lives, but for largeexchanges and/or exchanges that involvecomplicated products where quality canonly be determined with use, trust orrecourse generally must be available inorder to make promises credible so thatvoluntary transactions can occur.

Repetitious dealings within small close-

Globalization ThroughHuman Action or Human Design?

IF GLOBALIZATION REFERS to the ever-expanding opportunities for voluntaryinteraction between individuals around theworld, then the result has to be beneficialto most of the world’s population. Theexceptions are likely to be those individualsand groups who collect rents because ofthe protectionist policies of theirgovernments (e.g., organized labor,agricultural sectors, state supportedreligions). Nonetheless, governmentofficials around the world seem to believethat the process of globalization requirestremendous amounts of human designthrough alphabet-soup (e.g., UN, EU,WTO, NAFTA) organizations of and/oragreements between their governments. Infact, however, these organizations andagreements appear to be creating the samekind of constraints on voluntary interactionthat nation states have implemented, asgovernments attempt to protect variousdomestic interests, thus raising the cost ofglobalization, if not preventing it altogether.

A frequently alleged justification for theheavy hand of government in theglobalization process is that individualsfrom different countries and cultures willnot enter into what could be mutuallybeneficial interactions without a commonunderstanding generated by homogenousrules, and without contract enforcement bythe coercive power of states. In fact,however, globalization does not requireglobally standardized law or state coercion.It is true that recognized rules and eithertrust or recourse are required for manyvoluntary interactions to occur, but thereare several options beyond state-made and/or enforced law that are likely to be superiorsources of incentives for beneficialglobalization.

Must States be the Source of Rules?Hayek explains that rational individuals

are not able to use conscious reason toevaluate every particular option in the arrayof alternatives that are available, becausethere are significant limits on abilities toreason and to absorb knowledge. Thismeans, among other things, that rationalindividuals often find it beneficial tovoluntarily develop and conform to rules.In this context, “rules” should be seen as

Bruce L. Benson

ama -gi The Journal of the Hayek Society at LSE 3

knit communities encourage thedevelopment of trust (a willingness to makeoneself vulnerable to another even in theabsence of external constraints), becauseincentives to live up to promises and makevalid claims arise in such personalizedexchange in order to continue benefitingover a long period. The scope of trust-based trade tends to be limited to smallgroups, however, and this limits thepotential benefits of specialization andtrade. Globalization might be thought of asthe “extension of the market” from the trust-based arrangements in a local town orvillage to the region, nation, and ultimatelyto the far corners of the world. In order forsuch large scale depersonalized exchangeto occur in extensive markets, however,individuals generally require recourse tosome third party enforcement mechanismbefore they will accept the claims andpromises made by unknown tradingpartners.

One way for extension of the market tooccur is for traders to develop reputationsfor fair dealing, given that informationabout such reputations spread. After all,people have strong incentives to avoiddealing with someone they believe may notbehave opportunitisticly, so if the spreadof information about fraud or contractualbreaches is sufficiently effective, aspontaneous boycott of opportunists canbe anticipated (recourse is available as thespread of information will producepunishment for misbehavior: the third partyis the trading community as a whole).Depersonalized business-to-business tradenetworks often form “contractual” groupsthat facilitate the spread of information.Potential contractual arrangements arenumerous, including the implicit contractsof ethnic or religious networks, variouscorporate structures such as indirect equityties through pyramidal ownershipstructures, direct equity ties, interlockingdirectorates, and various informal andformal commercial and trade associations.In order to maintain the potential forongoing trade and protect reputations,mediation or arbitration services can beproduced within the group, or externalmediation and/or arbitration specialists(individuals with extensive knowledge oftrade practices and usage within thedisputants’ area of commerce) can beemployed.

Organizations to facilitate the flow ofinformation about reputation and/or lowerthe cost of dispute resolution can begeographically extensive (e.g., internationaltrade associations), although they often arefunctionally focused. Perhaps an

institution that is both geographically andfunctionally extensive could economize onthe spread of information and on disputeresolution? In particular, nation-states withcoercive power certainly can providestrong sanctions against fraud, breach ofcontract, and other forms of non-cooperative behavior, creating incentivesfor people to live up to their promises andmake valid claims. Indeed, a frequently madeclaim is that in order to induce compliancewith contracts and/or acceptance ofarbitration the other party must be willingand able to seek enforcement by a coercivepower. For example, some observers seesanctions under the 1958 New YorkConvention on the Recognition andEnforcement of Arbitral Awards and othersimilar international agreements betweenstates to recognize and enforce arbitrationrulings (e.g., 1961 European Convention onInternational Commercial Arbitration) as themajor reason for widespread acceptance ofinternational arbitration. If this is the case,then why not simply substitute the stateand its legal system for customarycommercial law systems. After all, a state’slegal system, or a multi-state organization’s(e.g., the EU) legal arrangements canprovide recourse for trades between peoplewho are not members of informal or formaltrading communities with internalarbitration or mediation services. Suchpotential benefits are likely to beoutweighed by significant costs.

Globalization in the Face of IncreasingComplexity: the Benefits of Specializationin Law

The wide variety of activities andrelationships that exist mean that many rulesthat are effective for one type of transactionor one group may not be effective foranother. The diamond traders in New York,Amsterdam and elsewhere may prefer avery different set of rules and institutionsthan those adopted by international oiltraders, for instance. The products beingtraded are very different, of course,suggesting that very different contractualissues are likely to be relevant, and thetrading communities are also very different.Many diamond merchants share commonethnic and religious backgrounds, creatingan environment of mutual understanding(e.g., of common trade practices and usage)and trust, for instance, thus reducing theneed for highly technical and specificcontracts, while oil traders display muchgreater ethnic and religious diversity as wellas differences in motivations (a number ofoil producing states have nationalizedproduction, for instance, so political

considerations can have major impacts ondecision-making), possibly reducing thelevel of common understanding andundermining trust relationships, thusdictating much more specific and complexcontracts. Combining all of the rules fromeach group into a single legal system wouldcreate unnecessary costs for both groups.

The Opportunity costs of Legal Transfersthrough Designed Rules Backed byCoercion

Another potential (and probablyinevitable) problem arises when states areheavily engaged in law creation andenforcement. Coercively imposed andenforced rules can facilitate voluntaryinteraction, but they also can influence thedistribution of wealth, and the use of law totransfer wealth (e.g., taxes and subsidies,regulations that limit entry, raise prices,impose trade barriers) actually reduceswealth creation for several reasons. First,such transfers are often achieved bypreventing or raising the costs of voluntaryexchange, including limitations on entry andcompetition, thereby limiting the wealth thatcould be created through these activities.Second, the resources consumed in thepolitical competition for wealth transfers(lobbyists, political organizers, etc.) couldbe employed to produce new wealth ratherthan to seek transfers of existing wealth.Third, potential victims of the transferprocess have incentives to resist, sotransfer-avoidance costs also arise. Thesecosts can take the form of diverting some oftheir productive resources in order to investin political information and influence in orderto compete in the political process. Exit isanother option, however, whether bymoving to an alternative politicaljurisdiction, or by hiding economic activityand wealth (e.g., moving transactions“underground” into the informal sector).Therefore, in order to induce compliancewith discriminatory transfer rules, the rulemakers will generally have to rely on a largeenforcement bureaucracy, both to limit exitand to execute the rules. The resourcesdevoted to such enforcement also arediverted from potentially wealth increasingproduction activities- a fourth cost.Enforcement bureaucracies can developtheir own agendas and extract wealthtransfers (e.g., excessive budgets, bribes)for their own benefit, raising the cost ofenforcement even further. The fifthconsequence is likely to be even moresignificant than the other four, however.Faced with the probability of involuntarytransfers, potentially productiveindividuals’ property rights to their


resources, products and income flow areperceived to be relatively insecure, so theirincentives to invest in maintenance of andimprovements to their assets, and theirincentives to produce and innovate in orderto earn more income are undermined. Iftransfers are expected to be large, frequentand arbitrary, production will be low andwealth expansion (economic growth) will bevery slow if it occurs at all.

Perhaps state-made law will not producebiased rules for the purpose of transferringwealth (rules that will also undermineincentives to be productive), but it is clearthat a necessary prerequisite for such lawis strong barriers to exit for those whoexpect to lose wealth through transfers.Interjurisdictional competition can occurbetween legal systems attempting tomonopolize law making and enforcement,and to the degree that wealth can escapeone to move to another, the potential forusing law as a transfer mechanism islimited. And importantly, when competitioncomes from customary law produced andsupported by institutions which are notattempting to monopolize law or transfer

wealth, but simply to substitute for trust inorder to facilitate depersonalized voluntaryinteraction, individuals may be able toescape the jurisdictions of all of those whoseek monopolies in law for transferpurposes. Furthermore, the relativelylimited jurisdictions of some customarycommunities are not as constraining as theymight appear to be. The membership oftrading communities can differ, althoughconsiderable overlap may also occur, soindividuals may deal with other individualson some dimensions but not on alldimensions. Perhaps more to the point,however, it took privately produced andadjudicated medieval lex mercatoria toovercome the limitations of politicalboundaries and localized protectionismduring the medieval period, thus paving theway for the commercial revolution anddevelopment of international trade. Indeed,there is absolutely no reason to believe thatany particular national government ororganization of governments is of the idealsize to take full advantage of the economiesof standardization in law. However, sincecustomary law can be geographically

extensive and functionally decentralized(i.e., specialized), in contrast to the law ofgeographically defined states ororganizations of states which tends to befunctionally centralized and geographicallyconstrained, customary law can havedifferent sized jurisdictions for differentfunctions. A customary system ofpolycentric law would appear to be muchmore likely to generate efficient sized“geographic market areas” (jurisdictions)for the various legal communities involved— perhaps many smaller than most nations,with others encompassing many of today’spolitical jurisdictions (e.g., as internationalcommercial law does today). Globalizationis supported by such an overlappingsystem by allowing for specialization andencouraging voluntary interaction throughspontaneously evolving institutionalarrangements.

Bruce L. Benson is the DeVoe MooreDistinguished Research Professor atFlorida State University and a senior fellowat the Independent Institute.

When working in the laboratory of a brainanatomist during the winter of 1919-20,Friedrich Hayek lighted upon a crucialinsight: a sensory fibre cannot carry, nor anerve cell store, the characteristics of mentalphenomena. Although Hayek believed thathe held the answer to an important question,he was unable to explain precisely what thequestion was! Over thirty years later,Hayek’s Sensory Order identified andaddressed that question: ‘What is mind?’

The Sensory Order provides the rationalefor self-knowledge, social adaptation andsocial science generally. Without somedegree of uniformity of minds, there can beno meaningful social interaction: a humanis more sensitive to (the perceptions of)another human than to a rat or (less still) toa bat or (less still) to a gnat. Introspectionreveals (hypothetically, and given geneticand cultural similarities) what is thoughtby other minds; and that same kind ofconscious self-examination provides abasis for ameliorating purely instinctive (or

emotional) responses. It thereby admits asocial dimension that invites bothconditioned and considered reactions.

ConnectionismArtificial intelligence research is directed

by competing methods. With the orthodox‘symbol-processing’ approach, a series ofbinary decisions is made against set rules,and memory is stored at specific locations.The alternative paradigm—of which TheSensory Order is an early statement—isthat of ‘connectionism,’ where the wholebrain (potentially) is engaged by thevariable strength of inter-neural impulses.Memory and thought are indistinguishableneurological processes - particularconfigurations of an intricate neural network- that are constantly adapted to stimuli fromthe external world. The mind is not a storeof data that reflect (or are correlated with)characteristics of elements in the physicalworld. Rather, organisms live in a sensoryorder that is created by the central nervoussystem.

For simple reflex action, higher centresreceive simultaneous reports of bothstimulus and response. At the highestlevels of consciousness, responses tostimuli are modified by the influence of thewidest range of current and past impulses.Between the extremes of reflex and

FRIEDRICH HAYEK REGARDED The SensoryOrder as one of his “more importantcontributions to knowledge” (Hayek, 1994,138). Given the thirty years of its gestation,it is unsurprising that the exposition andconclusions are implicit in much of Hayek’sother work. The Sensory Order is importantfor establishing the limitations ofintellectual endeavour and explicitknowledge, and (thereby) the rationale fora dependency upon the tacit knowledgethat is embodied in cultural and institutionalforms.

Limits to UnderstandingMind is the black hole of human science:

no empirical evidence emanates from within.Behaviourism has no access to cognitivefunctions, so it cannot reveal motivation.Introspection gains access to mentalimages, not mental processes. Empathymay disclose motivation, but gains noinsight into relevant neurologicalstructures.

Hayek’s Sensory Order

Gerald R. Steele


conscious response, a continuous rangeof ‘engaged’ connections is hypothesizedwith no qualitative distinction afforded tothe most abstract processes of thought. Allexperience is shaped by memory andunderstanding; and whenever a newexperience is inconsistent with ‘theclassification based upon past linkages,’the classification must be revised: “[w]hilethere can thus be nothing in our mind whichis not the result of past linkages (eventhough, perhaps, acquired not by theindividual but by the species), theexperience that the classification based onpast linkages does not always work, i.e.,does not always lead to valid predictions,forces us to revise the classification” (TSO,168).

The differentiating responses of theneurophysiological system are determinedby linkages previously created within theorganizational structure of the centralnervous system: a system of connections“acquired in the course of development ofthe species and the individual by a kind of‘experience’ or ‘learning’” (TSO, 53). Pre-sensory linkages determine ‘the order ofthe apparatus of classification’; that is, theframework that determines all our‘conscious experience of qualitativeattributes of external events’. Pre-sensorylinkages—”relations of which we are notconsciously aware” (TSO, 142)—are thatpart of a priori knowledge that “is notlearnt by sensory experience, but is ratherimplicit in the means through which we canobtain such experience” (TSO, 167).

What Mind IsThe irresolvable conundrum is that the

neural order is a subsystem of the physicalorder, but that any understanding of therelationship between the sensory order(mind) and the physical order must derivefrom the neural order. Hayek’s speculationwas that a machine designed by the humanmind might yet be capable of ‘explaining’what the mind is incapable of explainingwithout its help: “such a machine wouldnot differ in principle from ... a calculatingmachine which enables us to solveproblems which have not been solvedbefore, and the result of whose operationswe cannot, strictly speaking, predictbeyond saying that they will be in accordwith the principles built into the machine”(TSO, 189).

This conundrum is a generalization ofGödel’s theorem, that in no consistentaxiomatized mathematical system can theproposition expressing its own consistencybe proved: “Gödel’s theorem is but a specialcase of a more general principle applying

to all conscious and particularly all rationalprocesses, namely the principle that amongtheir determinants there must always besome rules which cannot be stated or evenbe conscious” (Hayek, 1967, 62).

Now, Gödel’s theorem says nothing inrespect of machines (or, rather, algorithmsof artificial intelligence) that might exercise(non-computational) mathematical intuitionas competently as the finestmathematicians; and so the human mindmight be explained by means of some moreelaborate artifact (though unlike acalculating machine). The argument is that,deep within the mind, there may lie some‘unconscious unknowable algorithm’ thataffords it an ability to judge logicalconsistency: “the Gödel argumentdemonstrates that whatever understandingis, it is not a computational thing. Thisallows that natural selection could operatefor this general non-computational quality-a quality which could be applied to a wholerange of problems and not simply tomathematics … If mentality is a function ofbrain action, and we accept that brain actionis subject to the same laws of physics aseverything else, those physical laws mustallow for non-computational action”(Penrose, 1995, 26).

In regard to the perennial mind-bodyissues, Hayek describes dualism and thenotion of “mind ‘stuff’” as an ‘old habit’that derives from humankind’s “early studyof nature’; and he delivers the conclusionthat ‘an account of mental phenomenawhich avoids the conception of a distinctmental substance is ... the opposite ofmaterialistic, because it does not attributeto mind any property which we derive fromour acquaintance with matter. In beingcontent to regard mind as a peculiar orderof events, different from the order of eventswhich we encounter in the physical world,but determined by the same kind of forcesas those that rule in that world, it is indeedthe only theory which is not materialistic”(TSO, 177-8). From that description, TheSensory Order finds its place within thecategorizations of dual-attribute or central-state theories that identify brain processesas wholly physical but with non-materialproperties that cannot be reduced tomaterial ones.

The Adaptation of MindThe brain is a biologically evolving

instrument of an adaptive system: thesensory order of mind. The three keyprinciples of Darwinian selection arediversity (of component elements),interaction (with the environment to testadaptive fitness) and differential

amplification (successful variants arereproduced in relatively greater number).The potential fallibility of both the brainand the mind is a necessary characteristicto allow Darwinian selection process tooperate.

Knowledge is not a unitary entity thatexists to a greater or less extent in differentspecies. Knowledge is domain-specific:different genes direct the selectionprocess of intelligence to producedifferent adaptations (knowledge) indifferent species. So, it would bemeaningless to ask how close perceptionsare to the noumenal world: “[w]hichexternal events are recorded at all, andhow they will be recorded, will ... dependon the given structure of the organism asit has been shaped by the process ofevolution” (TSO, 108).

While no clear boundaries separatebiological, psychological and socialadaptation, there are obvious variationsin the pace of evolutionary change. In asocial context, it is by a process ofrelatively rapid adaptation that the“knowledge and intentions of differentmembers of society are supposed to comemore and more into agreement” (Hayek,1937, cited from Hayek, 1949, 45). Inattempting to model those patterns, afeasible task for social science is to seekempirically testable theories ofexpectation formation and learning.

Evolution cannot proceed withoutvariations and it is impossible to anticipatemore than their immediate impact uponorganic function. Herein lies the argumentfor liberal social systems that haveallowed experiment, adaptation andselection such that “practices which hadfirst been adopted for other reasons, oreven purely accidentally, were preservedbecause they enabled the group in whichthey had arisen to prevail over others”(Hayek, 1973, 9).

Scientific KnowledgeIndividuals’ behaviour is based upon

the assumption that their sensory orderis both safe and similar to that oforganisms with which they associate. Toremain safe, knowledge must becontinuously revised at two levels: (i) ‘wenot only establish new relations betweenthe data given within a fixed framework ofreference’ but (ii) ‘we are led to adjustthat framework itself’ (TSO, 169). In thismanner, a priori knowledge of one kind(pre-sensory linkages, which precede allconscious experience) becomesaugmented by a priori knowledge of asecond kind (i.e. objective knowledge or


‘science’), as classifications based upon“immediately given sensory qualities’ arereplaced by those ‘defined in terms ofconsciously experienced relations” (TSO,170).

Science seeks ever-greater consistencyin the account of events. There is aconscious search for new classes—defined in terms of interrelationshipsbetween events rather than in terms oftheir sensory properties—such thatgeneral propositions (about the behaviourof events) are universally true. Ultimately(and hypothetically), ‘[s]uch a completesystem of explanation would necessarilybe tautological, because all that could bepredicted by it would necessarily followfrom the definitions of the objects to whichit referred’ (TSO, 173). While thiscategorization of events, by characteristicsthat belong to the objective order, definesthe (natural) scientific method, progressby this ‘reductionist’ approach must notbe so rigidly applied as to inhibitunderstanding.

Complex PhenomenaHayek distinguishes between ‘the

relatively simple phenomena with whichthe natural sciences deal’ and ‘the morecomplex phenomena of life, of mind, andof society’ (Hayek, 1967, 25), that are lessaccessible than those of physical systems.Although classes of patterns identified bysocial science might allow for predictions,these must be predicated upon highlyspecific circumstances; that is, uponextensive empirical data. However,predictions are not the sole concern and,if the data are insufficient to allowpredictions, the theory—the knowledge ofthe pattern—is still useful. Where a theoryhas little empirical content, “hypotheticalpredictions’ may be possible; that is,‘predictions dependent on yet unknownevents” (Hayek, 1967, 29). Beyond that,“[t]he very insight that theory provides ...that almost any event in the course of aman’s life may have some effect on almostany of his future actions, makes itimpossible that we translate our theoreticalknowledge into predictions of specificevents” (Hayek, 1967, 34).

So, Hayek warns of the ‘pseudo-entities’ that comprise the Keynesianapproach: “the number of separatevariables which in any particular socialphenomenon will determine the result of agiven change will as a rule be far too largefor any human mind to master andmanipulate them effectively. Inconsequence our knowledge of theprinciple by which these phenomena will

be produced will rarely if ever enable usto predict the precise result of anyconcrete situation” (Hayek, 1952, 73-4,italics added).

ConclusionIn the years that followed publication

of The Sensory Order, Hayek’s work onthe “far-reaching philosophical problems”that derive from “the distinction betweenwhat we can say ‘within a system’ andwhat we can say ‘about a system’” (Hayek,1994, 29) proved ‘so excruciatinglydifficult’ that a long and unfinished paperwas abandoned after he found thatnobody he “tried it upon couldunderstand” (Hayek, 1982, 290). Even so,the immutable constraints upon theunderstanding of mental processes haveimportant philosophical and socialimplications: “[w]hile our theory leads usto deny any ultimate dualism of the forcesgoverning the realms of mind and that ofthe physical world respectively ... we shallnever be able to bridge the gap betweenphysical and mental phenomena; and forpractical purposes ... we shall permanentlyhave to be content with a dualistic view ofthe world” (TSO, 179).

Hayek saw a way forward in drawing adistinction where none exists: betweenforces that govern the mind and forces thatgovern the physical world. The socialrelevance of that conclusion is that itremoves the basis for mitigatingcircumstances that might excuse someparticular action. There is no “metaphysicalself which stands outside the chain ofcause and effect” (Hayek, 1967, 232). Inother words, an individual’s action isalways integrally linked to thecircumstances of that action by ouressential ignorance of the physicalconditions that would have given rise tosome different action. Such ignorancewould be removed only by anunderstanding of the derivation of thesensory order from the physical order, butthat is impossible. This explains theemphasis that Hayek places upon anindividual’s action and the responsibilityof that individual for his or her action. Ifindividuals were not held responsible fortheir action, nothing would be left forwhich an individual could be heldresponsible.

To hold an individual “responsible forthe consequences of an action” is anassertion neither of causation nor of fact,but “is rather of the nature of a conventionintroduced to make people observe certainrules” (Hayek, 1960, 74-5). Suchconventions are the representation of

commonly held attitudes and relationshipsthat, by their evolutionary adaptation,endure as one generation succeedsanother. This is a genuinely socialstructure that is separated from anyparticular set of individuals. In acceptingthe discipline of those conventions, anddespite personal ignorance, the individualis guided by a cultural inheritance thatfacilitates the achievement of the widestrange of individual purposes.

ReferencesHayek, F.A. (1937) ‘Economics and

knowledge’, Economica (N.S.), 3, 33-54Hayek, F.A. (1949) Individualism and

Economic Order (London & Henley:Routledge)

Hayek, F.A. (1952) The Sensory Order(London: Routledge & Kegan Paul)

Hayek, F.A. (1960) The Constitution ofLiberty (London and Henley: Routledge& Kegan Paul)

Hayek, F.A. (1963) ‘Rules, perceptionand intelligibility’, Proceedings of theBritish Academy, XLVIII, 321-44

Hayek, F.A. (1967) Studies inPhilosophy, Politics, and Economics(London & Henley: Routledge & KeganPaul)

Hayek, F.A. (1973) Law, Legislation andLiberty: A new statement of the liberalprinciples of justice and politicaleconomy: Vol. 1. Rules and Order(London & Henley: Routledge & KeganPaul)

Hayek, F.A. (1982) ‘The Sensory Orderafter 25 years’, in W.B. Weimer & D.S.Palermo (eds), Cognition and theSymbolic Processes, Vol. 2, 287-93(Hillsdale, NJ: Erlbaum)

Hayek, F.A. (1994) Hayek on Hayek: AnAutobiographical Dialogue, S. Kresge &L. Wenar (eds) (Chicago, IL: University ofChicago Press)

Penrose, R. (1995) ‘Must mathematicalphysics be reductionist?’, in J. Cornwall(ed.), Nature’s Imagination: The Frontiersof Scientific Vision, 13-26 (Oxford: OxfordUniversity Press)

Gerald R. Steele teaches economics atthe Lancaster University ManagementSchool. He is the author of The Economicsof Friedrich Hayek (The Macmillan PressLtd) and Keynes and Hayek: the MoneyEconomy (Routledge).


people who are excessively competitive. Yetthe competitive market environment mayencourage and bring out these aspects ofindividual’s personalities. If ruthlesslycompetitive people are successful, suchbehavior may be imitated. At the same timethose who are (excessively) cooperativemay be taken advantage of, derogated aspansies. Accordingly such behavior will bediscouraged.

Self-interested BehaviorOne of the reasons that we frown upon

people who are excessively competitive isour ambivalence toward self-interestedbehavior. Adam Smith may have been rightthat we can rely on self-interestedness tolead to the public interest being served moresurely than we can rely on benevolence.There are some special conditions in whichthe reliance on self-interested behaviorresults in economic efficiency.

Yet trust is essential in the world in whichwe live; to be earned, trust often requiresacting in a less than perfectly self-interested manner. Trust was essential forthe early development of capital markets.In “imperfect” markets certain types of self-interested behavior impair economicefficiency. Indeed we know that we mustprovide incentives that are often quitecostly to make self-interested individualsact in a trustworthy way (the wages oftrust).1

There is, in this, a certain irony:Capitalism, as it promotes self-interestedbehavior, may create an environment lessconducive to efficiency. Capitalismprospers best in an environment with apeculiar combination of self-interestedbehavior—enough to induce individuals tolook for profitable activities—and non-self-interested behavior, where one’s word isone’s honor, where social rather thaneconomic sanctions suffice to enforcecontracts.

The critique of capitalism, that itpromotes self-interested behavior, goesbeyond, of course, the self-interest ofcapitalism in itself. A long-standing tenetof civilizations, both East and West, is thatindividuals must care about others.Capitalism may encourage self-interestedness, but is that really desirable?

We view charity as a virtue,2 but does itremain a virtue when we compel charityupon others? There is something different,certainly for the giver, perhaps for thereceiver, about giving money to a poorperson voluntarily rather than beingcompelled to do so. By changing the locusof caring and responsibility from theindividual to the government—not only forthe needy, but for oneself, one’s parents,one’s children—we change society and wechange ourselves. Here again we see acertain irony: Attempts to improve societyby having the government undertake agreater role in redistribution, mayultimately—through their effects onindividuals and the nature of the socialcontract—have more ambiguousconsequences.

Endogeneity of Human NatureI have spoken of how the design of the

economic system may foster certaincharacteristics—self-interestedness andcompetitiveness. In modern vocabulary, wesay that certain aspects of human natureare endogenous to the system. Theconcern about what the economic systemdoes to the human spirit, while it hasdisappeared from modern economics, waspresent in Adam Smith’s writing:3

The understandings of the greater partof men are necessarily formed by theirordinary employments. The man whosewhole life is spent in performing a few simpleoperations... generally becomes as stupidand ignorant as it is possible for a humancreature to become. The torpor of his mindrenders him, not only incapable of relishingor bearing a part in any rationalconversation, but of conceiving anygenerous, noble, or tender sentiment, andconsequently of forming any just judgmentconcerning many even of the ordinaryduties of private life... His dexterity at hisown particular trade seems, in this manner,to be acquired at the expense of hisintellectual, social, and martial virtues. Butin every improved and civilized society thisis the state into which the laboring poor,that is, the great body of the people, mustnecessarily fall, unless government takessome pains to prevent it.

Yet, while traditional economic theory isclearly wrong in treating individuals asimmutable—”tastes” no less thantechnology were the primitives of themodel—we have no scientific basis onwhich to judge one set of moral values, oneset of personality types, as superior toothers. Thus, while Hayek may have beenright in stressing the moral dimension ofmarkets—the kind of consequences in

Philosophical Speculations adapted from Whither Socialism?

Joseph E. Stiglitz

THE DREAM OF a better world here onearth has been a central theme in thedevelopment of Western civilization sincethe Reformation. The nineteenth centurysaw some of these utopian visionstranslated into experiments of rather limitedsuccess. But the nineteenth century alsosaw the development of ideologies, whichreplaced the religious doctrines that hadso long held sway over humankind but wereheld with the same emotional fervor; indeedthe fervor was reinforced by the false sensethat the ideologies rested on scientificpremises. The sway that the Marxistideology had over the minds—andeventually the lives—of so many for morethan a century should give us pause: It issurely a sign of the importance of humanfallibility. It should make us cautious in theconfidence with which we hold our views,and cautious in our appeal to “science” tojustify our beliefs about the organizationof society. But, beyond that, we need toseek the deep-seated reasons for the depthand persistence of the appeal of thesedoctrines.

Shaping IndividualsThe popular success of the Marxist

ideology was partly the hope of a moreefficient economy, one that would bringmore goods to more people. But the successamong intellectuals went beyond that: Itwas partly based on the belief of the effectof the economic system on the nature ofthe individual. The concern was well-founded: One of the most damningcriticisms of the socialist experiment hasbeen what it did to the human spirit—thecynicism that developed among the young,the bureaucratic mentality, the lack ofinnovation.

CompetitionCompetition is important, not only

because of its ability to promote economicefficiency but also because of the zest thatit gives to life. Here we encounter one ofthe many ambivalences that characterizesour views about market economies:Competition is good, but we have ourdoubts about excessive competition. Weencourage cooperation within teams butcompetition among them. We frown upon


shaping human nature that I have justdescribed—he fails to provide us with asystematic approach for addressing theseissues (e.g., see his 1989 book).

The Evolutionary Approach4

Neoclassical economists have attemptedto defend capitalism on the narrow groundof economic efficiency... There are otherstrands of thought in economics that arguefor market processes, but decry theneoclassical defense as too narrow, andindeed even misguided. I have alreadynoted Hayek’s argument for the “moral”dimension. One long-standing tradition,which includes Hayek and his followers,and Alchian, has emphasized anevolutionary argument for markets.

There is a natural analogy betweencompetition among species andcompetition among individuals in themarketplace. Spencer and others extendedDarwin’s ideas on natural selection andsurvival of the fittest to a social context.Terms like “fittest” have an essentiallynormative overtone: The term “survival ofthe fittest” is obviously meant to conveymore than just the observation that thosewho survive are the ones who survive. Insome essential respect the “fittest” whosurvive are better than the less fit who donot. There is a teleological aspect ofevolutionary processes: Nineteenth-century views of progress were reflectedin the motion of constant, albeit slow,“improvement” in species and societiesresulting from evolutionary forces. Callouspolicies that entailed governments ignoringthe suffering of the poor were justified inthe name of social Darwinism.

Indeed it is only with hindsight that wemay be able to tell whether a particularpolicy had survival value. Certainly thehypothesis that governments constitute apositive evolutionary step is supported bythe observation that societies withingovernments have survived, and thosewithout have not. By the same token,certain types of government interventionin the market may, from an evolutionaryperspective, enable a society to survivebetter. While we might conclude from thedemise of socialism that it did not havesurvival value, we cannot conclude thattherefore market institutions are superior.By the same token, one can argue from thesucess of the East Asian countries, that“managed markets” and strong governmentintervention have survival value, incomparison with market institutions withmuch more limited governmentintervention.

there is more than one form of capitalism ,that the conduct, structure (organization)and performance of the Japaneseeconomy—both the private sector, and therelationships between government and theprivate sector—differ, for instance, inimportant ways from those of the UnitedStates. Evolutionary theory does not giveus much basis for choosing among theseinstitutions.

By the same token, those who appeal tothe evolutionary processes also claim toomuch: There is no reason to believe thatevolutionary processes have anyoptimality properties. Indeed there arestrong arguments suggesting thatevolutionary processes are far from optimal.Biologists stress the randomness of theprocess, the seeming redundancies sofrequently observed, the vestigial elementsthat often lead to problems (such as thehuman appendix).

Economists note that in the absence of acapital market, a species with strong long-run prospects simply cannot “borrow” tocarry it through a temporary change in theenvironment. A species—or a firm—withgreater long-run adaptability can be wipedout in the competitive struggle by onebetter-suited for the particular environment.Thus more competitive environments—environment in which competition is moreruthless so that any but the most efficientfirms, in that environment, are weededout—may in the long run actually be lessproductive...

The fundamental point is that there is noreason to believe that market economies“naturally” make the right trade-offs or that,in particular, market economies with moreruthless competition are more efficient thaneconomies in which competition is moregentle. Moreover, since whether a particulartrait (species) survives depends on theenvironment, which itself is endogenous,there is no reason to believe that the systemas a whole has any optimality properties.The system simply ensures that those whohave characteristics that are rewarded, inthe particular environment which thesystem has created, survive. Thus one canimagine a world in which there are two typesof individuals, bureaucrats and innovators.Bureaucrats make life difficult forinnovators, and innovators make lifedifficult for bureaucrats. There are multipleequilibria to such a society. Bureaucratsmay dominate; in that environmentinnovators do not prosper. The bureaucratscreate an environment that is favorable totheir own type. But, conversely, innovatorsmight dominate. They create anenvironment that is favorable to their type.

Though the economy, from different initialconditions, could evolve toward eitherequilibrium, one of these might be (undersome welfare criterion, such as long-runeconomic growth) clearly superior to theother. (It is only when the two societiescome in direct conflict or comparison witheach other that the disadvantages ofbureaucracy become revealed.)

Good mutations (new social institutions)may not survive on their own, for theyrequire accompanying changes in othersocial institutions. There is a coordinationfailure. Many changes have to occur at thesame time, and market processes may notbe able to provide the necessarycoordination. There is thus no presumptionthat evolutionary forces, left to themselves,have any desirable welfare properties.Moreover, if we take seriously theobservations made in the first part of thischapter, concerning the endogeneity ofpreferences, we have fundamentalproblems even ascertaining what areappropriate criteria for judging evolutionaryprocesses.

Of course, if evolutionary forces“naturally” led to desirable outcomes(whatever that might mean), then theeconomist’s task would be a simple one: toobserve and comment on the process. Butas economists, we are called upon toanalyze a variety of proposed changes inpolicies and institutions. As our tools ofanalysis have improved, we are in a betterposition to ask of any proposed change,what are its effects? In evolutionary termswe can ask, is it likely to survive? We areeven in a position of engaging in socialengineering, of asking can we designinstitutions or policy reforms that are likelyto be welfare improving, or, again inevolutionary terms, that are likely to havesurvival value?

The great socialist experiment is comingto an end: We have learned a lot from theseexperiments, but because they were hardlycontrolled experiments, what we learnedremains a subject of some dispute. Whilegovernment ownership is clearly nopanacea, there remains scope for furtherexperimentation. For instance, we need tostudy forms of economic organizationinvolving more worker participation andownership. Not too much should be readinto the failures of the worker-managed firmsin the former Yugoslavia, for these involvedpeculiar (and obviously unsatisfactory)arrangements with respect to the transferof property rights, as well as otherinstitutional details which, both ex ante andin hindsight, were not conducive tosuccess. To return to the theme with whichMore broadly, we now recognize that


I began these speculations, the question iswhether the insights of modern economictheory and the utopian ideals of thenineteenth century can be brought closertogether?

Endnotes1 This notion was emphasized in the

reputation models (including the efficiencywage theories) I discussed earlier.

2 We should note the extensive literaturearguing that certain forms of altruism havesurvival value in an evolutionary context.

(The critical issue in that context is, Whatis the relevant unit of analysis?)

3 Quoted by Heilbroner, in “Reflections:Economic Predictions,” New Yorker, July8, 1991.

4 This section borrows heavily fromStiglitz, J. E. Notes on evolutionaryeconomics: Imperfect capital markets,organizational design, and long-runefficiency. Paper presented at OsakaUniversity International Symposium on“Economic Analysis of Japanese Firms and

Osaka, Japan, November 9, 1992.

Whither Socialism? expands on theWicksell lectures presented at theStockholm School of Economics.

Joseph E. Stiglitz is a professor ofeconomics and finance at ColumbiaUniversity and has received the John BatesClark Award and the Nobel Prize ineconomics.

JOHN KAY’S NEW book is a truly ambitiousventure. It covers five related topics: thewealth of nations, and the factors thatdetermine prosperity; the role ofcompetitive businesses in furtheringinnovation and progress; recentdevelopments in mainstream economics,and the light thrown by these on the usesand limitations of markets; some leadingtrends in economic events and policy overthe past decade or so; and the lessons tobe drawn, in the light of all the above, forthe conduct and direction of economicpolicy today. Given this large agenda, it isnot surprising that the book is a long one.But thanks to the author’s gift for clear andreadable presentation, laced with well-chosen and well-informed examples, it isnot difficult to read. Kay has the ability topresent arguments and evidence in a readilyassimilable and often informal way, withoutoversimplifying or writing down. He hasdrawn to good effect on a wide range ofpertinent sources, by no means confinedto economics, as well as on his wideknowledge of economics and businesshistory. Along the way, he makes manyinteresting and perceptive points to whicha short review cannot do justice.

In relation to the first two of the five abovetopics, Kay’s main thesis is not new, thoughhe presents it in his own distinctive andoften highly illuminating way. In the earlierparts of the book he outlines the positive

role of markets and competitive businesses,and the ways in which they havecontributed to economic progress. Here ofcourse he is in good company. The casefor what he terms ‘disciplined pluralism,’and others might prefer to call economicfreedom, is widely accepted; nor wouldmost advocates of that case wish to quarrelwith him when he stresses that “marketeconomies are embedded in a social,political and cultural context.” But the bookis only partly concerned to restate the casefor a market economy and the need for suchan economy to go together with supportinginstitutions and behaviour. Its secondpurpose is to argue against excessive anduncritical reliance on markets and, morebroadly, on forms of conduct which arebased only on self-interest. Kay departsfrom current pro-market trends of thinking,in that he is preoccupied with theinadequacies and dangers of what he seesas ‘market fundamentalism.’ He believesthat to state correctly ‘the truth aboutmarkets’ it is necessary to expose anddiscredit some currently accepted untruths.It is this second line of argument thatconstitutes the main distinctivecontribution of the book.

In part, the argument is developed byoutlining ways in which markets mayperform badly or be subject to limitations.In this connection, Kay writes informativelyabout particular recent episodes of failure,

excess or fraud, such as the dot.com bubbleand the Enron affair. In a more systematicway, as the third of the above listed topics,he considers, in five successive chaptersof the book, how far recent developmentsin economic theory support or underminethe case for reliance on competitivemarkets. Despite a few queries andreservations, I found these chapters clearand instructive.

Unfortunately, however, the variousinformed, discerning and well presentedarguments that I have summarised up tonow are linked in the book with an extendedfrontal attack on something which Kayterms ‘the American business model’ (orABM for short). His preoccupation withthis doubtful construct greatly reduces theinterest and usefulness, and indeed theaccuracy, of what he has to say about theworld of today.

The main features of the ABM, as statedin the opening paragraph on the book’sdust-jacket, are four: “the unrestrainedpursuit of self-interest, marketfundamentalism, the minimal state and lowtaxation.” A number of questions ariseconcerning this allegedly influential andthreatening way of thinking.

First, what is the history of the ABM,and when did it become a phenomenon thatshould be taken seriously? Although atone point Kay traces its growing influenceback to the election victories of RonaldReagan and Margaret Thatcher, and in whatmay be a throwaway line he refers toCommodore Perry as having brought it toJapan in 1853, he believes that the decisiveevent was the collapse of communism. Inthat connection, he takes the surprisingpositions that “The right won the cold warand the left lost… The profits of GoldmanSachs and Coca Cola are the fruits ofvictory” (281), and that the BritishConservative Party lost its main raison

Review of John Kay’sThe Truth About Markets: Their Genius, Their Follies,Their Limitations

David Henderson

Markets: A New Microeconomic Paradigm.”


d’etre with the collapse of the Berlin Wall(324). But in fact the collapse of communismmade little difference to the profits ofbusiness enterprises, the terms on whichcontroversies about the role of markets andthe state are conducted in democraticcountries, or the positions taken in thosecountries by non-communist politicalparties.

Second, who are the exponents of theABM, the gurus whose ideas havesupposedly carried weight in recent years?Although Kay refers to a number ofAmerican persons and sources, none ofthese is of any serious consequence. Morefundamentally, he argues that theintellectual basis for the ABM is to be foundin Robert Nozick’s 1974 book, AnarchyState and Utopia , and in the first‘fundamental theorem of welfareeconomics’ that was established in the1950s for an economy that meets a numberof highly restrictive assumptions. But it isnot apparent that either of these lines ofthought, if indeed they play the role heassigns to them, has come to the fore inrecent years.

Third, who actually believes in the ABM?Clearly, no government is, or ever has been,

committed to it. Kay maintains (8) that it“remains the working hypothesis of mostbusiness people and consultants.” This isnot so, however: most large businessestoday, and the consultants who stand readyto advise them on the subject, are nowcommitted to the questionable doctrine of‘Corporate Social Responsibility’ (CSR).Contrary to what Kay believes (314), CSRis deeply collectivist in its assumptions andimplications.

While the ABM is little more than afigment of the author’s imagination, it isnot just a harmless eccentricity. Becauseof his fixation on a largely non-existentthreat, Kay’s treatment of recent historyand current issues of policy is flawed. Inparticular, there are two serious omissions.First, he does not refer to the ways in whichcollectivist ideas have maintained andattracted support, and influenced thecourse of economic policies, in recentyears. The gains made by collectivism area much more worrying phenomenon thanthe so-called ABM. Second, he does notgive the consideration that his titlepromises, and the issues deserve, toarguments for extending the scope ofcompetitive markets- for example, in the

provision of health and education services,in tackling the various currentmanifestations of over-regulation, in furtherliberalisation of cross-border trade andinvestment flows, and in reversing theerosion of freedom of contract that hasoccurred in the name of ‘equal opportunity,’‘anti-discrimination’ and ‘diversity.’ Inthese respects especially, the bookrepresents a missed opportunity; and takenas a whole, it does not offer a balancedtreatment of its subject matter.

John Kay, The Truth About Markets:Their Genius, Their Follies, TheirLimitations, London, Allen Lane (PenguinBooks), 2003.

This review was prepared for, and hasbeen published in, the Swedish journal,Axess. It will also appear, in the presentEnglish version, in Economic Affairs.

David Henderson is a visiting professorat Westminster Business School and aformer chief economist of the OECD.

The Risk of Paternalism adapted from Philosophy, Policy and Social Value MSc thesis

Peter Rossi

THE TRADITIONAL CASE for liberal1 risk-management via a market order relies onhuman nature enabling individuals tointeract and better themselves within suchan order. Although market-driven progresssince the industrial revolution arguablysupports this view, calls for regulation to“help” the market reach so-called sociallyoptimal ends are ever-present. The latestcome from the field of behavioral law andeconomics (BLE), which, by questioningthe rationality of neoclassical economicman (NEM), attempt to show whypaternalistic regulation is warranted. Aftera definitional detour, this essay will attemptto show why BLE may have unwittinglystrengthened the liberal case by paving theway for a revival of Austrian marketprocess theory, which explains theworkings of markets without recourse to

the unrealistic assumptions of NEM, andpreceded BLE in incorporatingpsychological insights into economics.

The behavioral paradigm“The task of [BLE]...is to explore the

implications of actual (not hypothesized)human behavior for the law.”2 In particular,“Behavioral Economics is the combinationof psychology and economics thatinvestigates what happens in markets inwhich some of the agents display humanlimitations and complications.”3 Jolls et al.4

group these limitations or biases into threecategories: bounded rationality, boundedwillpower and bounded self-interest.5 Anexample of bounded rationality is theavailability heuristic, where people confusefamiliarity with probability.6 The morememorable or notable an event, the more

likely people assume it to be, in contrast toactual statistics.7 It is these biases “thatcan justify the need for paternalistic policiesto help people to make better decisions andcome closer to behaving in their own selfinterest.”8 The policy recommendationshave different names, but they amount tothe same thing: paternalism.9 Arepresentative example is the proposal formandatory cooling-off periods becausehot-sales lead to poor decisions due toprojection bias.10 Thus, for our purposes,BLE should be seen as a challenge to liberalrisk-management because it suggests thatindividuals cannot cope in market settings.

The next section will offer an alternativeto the main target of BLE: NEM. ThisVulcan-like figure is not the only economicactor, for there exists the older, moreplausible Austrian account, which rendersthe BLE account of economics semi-otiose.

Austrian Economic manFor Hayek, the economic problem is not

the neoclassical problem of optimally co-ordinating given resources among givenends. Instead, “It is rather a problem of howto secure the best use of resources knownto any of the members of society, for endswhose relative importance only these


individuals know. ...it is a problem of theutilization of knowledge not given toanyone in its totality.”11 Knowledge forHayek was distinguished from informationby temporal and spatial specificity, whereason neoclassical and BLE accounts,knowledge is a commodity,indistinguishable from information.12 ForAustrian Economic man (AEM) then, themarket, via specialisation, economises onthe need for the excessive information andrationality demanded of NEM andencourages learning by doing. TheHayekian/Austrian market, co-ordinated bythe (unhindered) price-mechanism,transmits knowledge far more efficientlythan a central planner because the relevantknowledge cannot be representedstatistically. Hayek’s insight was eclipsedby the formalist revolution in economics,13

and so we are in the ironic situation whereBLE is criticising neoclassical economicsfor ignoring psychology, when Austrianswere originally criticised by neoclassicaleconomics for incorporating psychology.Since the Austrian market does not dependon the achievement of static equilibrium,“the deviations of real markets from themodel do not constitute rebuttals to[Hayek]. Indeed, deviations from the modelwere Hayek’s starting point.”14 The ironythat perhaps lies ahead for BLE academicsis that in comprehensively underminingNEM, they may be laying the groundworkfor a Hayekian/Austrian revival.15

A brief note must be made hereconcerning the social aspect of Austriantheory. Scholars such as Beck andHabermas critique liberalism for its“atomism”– its supposed disregard for therole of the social in individuals’ preferences.On this account, because preferences aresocially constructed, public spheres are offundamental import. Interestingly, Hayekrecognised this critique,16 but advocatedmarkets precisely because they facilitatedthe transmission of social knowledge moreeffectively than alternatives. In Austriantheory, preferences are endogenous toinstitutions, and therefore, a liberal order isneeded to foster creative development andeffective transmission of such knowledge.17

Central planning of social knowledge andpreferences is as ineffective as its economiccounterpart.18 Furthermore, the patchworkquilt liberal order would allow deliberativedemocrats to live amongst their ilk,although one wonders if they would becontent to leave non-deliberative othersalone.

BLE, a BLE development that ought to beembraced by liberals will be adumbrated.Liberals arguably need to defend someredistribution, however minimal, in order toprovide for defence and security. Therefore,liberals should adopt any BLE proposalsthat make liberal laws more effective. Forexample, rules can be reframed to reflectcognitive biases,19 and the replacement ofprobabilities with natural frequencies caneliminate many cognitive biases, such asover-confidence.20 Furthermore, liberalswelcome extensive private paternalism(such as Victorian philanthropy) and theliberal State can be defined as a paternalisticinstitution. This liberal acceptance holdsonly for as long as BLE does notrecommend new, illiberal laws, although itcannot resist this temptation.

...and against paternalism.The first of five distinct arguments

against the policy recommendations of BLEis a familiar one: democratic politicaldecision making tends to pervert the initialpolicy. Firstly, as Bastiat prescientlyrecognised in What Is Seen and What IsNot Seen, politics relies on the availabilityheuristic. Rent-seekers are able to drivelegislation anecdotally by pleading specialcircumstances.21 Jolls et al. expect‘availability entrepreneurs’ and behavioralbureaucrats to use behavioral knowledgeto enact “good” legislation: “Thus, publicchoice accounts can work productively withbehavioral accounts.”22 This is wishfulthinking in the extreme.23 Only specificforms of political reform can limit suchlegislation, and perhaps the most effectiveis to reduce the power available tolobbyists. The absence of such accessiblepower is a defining feature of the liberalState. Secondly, we need to ask thefollowing: where are such biases likely tohave worse consequences– in politics orin civil-society? Even if, counterfactually,power did not corrupt, the consequencesof biases in the legislative process are farmore serious than in normal, day-to-daymarket activity. The consequences of mysuffering from the availability heuristicwhilst betting on horses, for example, areinsignificant compared to theconsequences of politicians suffering fromthe same bias after the Love Canal incident.At the very least then, behavioral bias andpower do not mix. Therefore, decision-making needs to be compartmentalised, byprotection of several property, to limit theeffects of biases.24 Only reliance onomniscient politicians can avoid such aconclusion.

present regulatory straightjacket, functionsdespite behavioral biases. Such biases havenot prevented the attainment of ouruniquely high standard of living.Furthermore, there are three reasons whythe market reduces the effects of biases.Firstly, most, if not all traders cannot sufferfrom the endowment effect without losingtheir jobs.25 Good drives out bad, or badhires good, which amounts to the samething. Secondly, reputational non-publicregulation is provided by organisationssuch as the Underwriters Laboratory andvarious consumer groups.26 Such privateregulation is based on the sound premisethat firms make more money by not harmingcustomers. Thirdly, BLE argues againstprivate provision of behavioral consultingservices to reduce the effects of biases intwo ways.27 One, firms offering cooling-offperiods will be driven out of the market.Initially this may be true, but in the long-run it is impossible to rule out a feedbackmechanism whereby consumers learn bytheir mistakes.28 Furthermore, support forharm caused by hot-decisions in seriousmatters that result in destitution is properlythe role of welfare. It is far easier and moreefficient to leave the market alone and relyon PUW and PRW than intervene in themarket with a view to reducing the need forwelfare. Two, biases will prevent peoplerealising their mistakes. Again, the marketis a feedback process, so it is not clear whyprivate firms cannot provide behavioralconsulting to alleviate the impact of suchbiases. Perhaps only BLE academics andbehavioral bureaucrats are qualified toprovide such a service?

Thirdly, a crucial question to ask ofsociety is not solely “who decides?” but inwhat type of institutional framework aredecisions motivated and played out?

“Neither constitutional democracy nor amarket economy relies on decision makersto have superior wisdom or morality. ...Historically, it was—and is—arevolutionary concept, rejecting theoriesgoing back thousands of years which insistthat what matters is which persons andwhich doctrines rule, rather than thesystematic incentives and constraints thatcontrol whoever rules under whateverdoctrine. The American Constitution leftlittle room for philosopher-kings.”29

Similarly, Hayek describes theadvantages of a liberal institutionalframework as follows:

“Rules alone can unite an extendedorder...Neither all ends pursued, nor allmeans used, are known, or need to beknown to anybody, in order for them to beFor paternalism...

Before discussing the further failings of Secondly, it seems the market, even in its taken account of within a spontaneous

order. Such an order forms itself...”30

These accounts of liberal society are notcompatible with NEM who needs to fullycomprehend the system in which he acts.It is real-life deviation from this Godlystandard that BLE is exposing and in thisrespect, it is questionable what relevanceBLE has to reality. The constructivistic BLEapproach falls to what Hayek called the fatalconceit: the belief that order must bedesigned. It assumes that actors mustcomprehend their institutions, andsubsequently posits the “irrational”incomprehension of actors as sufficientcause to “rationalise” them throughpaternalism. Ecological rationality,31 asopposed to constructivism, recognises thatthe human mind economises by relying onthe subconscious and thus we are notalways able to articulate the reasons forour actions. The Hayekian market, then,economises on cognitive computations.32

Thus, information markets can supply moreaccurate information than articulatedpolls.33

Fourthly, BLE advocates reducingconsumer options.34 However, whilstliberals can accept the premise,35 it is notclear which choices should be prohibited.If, out of choice-set A through D, Johnprefers A, B, and Mary C, D, whosepreferences win? If individual preferencesare subjective, and assuming A through Dare lawful,36 the choice will inevitably comedown to politics, which is well describedas choosing for others. The problem thenis less reducing options, but who restrictswhose options?37 Does John choose todecrease his options by buying onlybranded goods because he acknowledgestheir reputation, or does the State prohibitnon-branded goods to increase safety, orsome such? If the latter, what happens topeople who cannot afford branded goods?Similarly, competition, which lowers pricesand increases efficiency, works byincreasing choice. For BLE to say here that“less is more” contradicts economichistory.38 Furthermore, BLE actuallystresses (unintentionally) the need forcompetition: “...cognitive biases offer butanother illustration of the basic Hayekianinsight regarding the importance ofdecentralization in social affairs: thepartitioning of responsibility ... functionsas an error-correction mechanism.”39

Finally, in a sceptical mood, one couldaccuse BLE of being the latest manifestationin the historical association betweenintellectuals and socialism. Since Plato’sphilosopher king, many intellectuals haveembraced the fatal conceit ofconstructivism.40 Goodin’s moral

collectivism epitomizes this philosophy:“state officials are probably better informedas regards questions of what are the bestmeans to people’s chosen ends. It may alsobe true that they are better informed evenas regards questions of what people’s endsreally are– or will be.”41 On the BLE account,there are two classes of people: thebehaviorally irrational majority, and therational minority (intellectuals andpoliticians). Hayek’s explanation of thisintellectual deceit needs no furthercomment:

...intelligent people will tend to overvalueintelligence, and to suppose that we mustowe all the advantages and opportunitiesthat our civilisation offers to deliberatedesign rather than to following traditionalrules, and likewise to suppose that we can,by exercising our reason, eliminate anyremaining undesired features by still moreintelligent reflection, and still moreappropriate design and ‘rationalcoordination’ of our undertakings.42

Therefore, although BLE poses asubstantial and welcome challenge toliberal risk-management, it fails mainlybecause it succeeds only in toppling NEM.Since NEM was originally created as amodelling tool, BLE has ironically pavedthe way for Austrian economics, whichdeals with the real economic problem: howdo knowledge-rich, information-poorindividuals coordinate for mutual gain? Theanswer, through a free-market, has beenavailable for hundreds of years. It is thisconclusion that BLE ought to be drawn to:“The only practical implication ofbehavioral economics is to strengthen thecase for private institutions”.43

Endnotes1 Liberal here means classical liberalism,

as distinct from Rawlsian or modernliberalism.

2 Jolls et al., p. 14.3 Mullainathan and Thaler, p. 1.4 A referenced list is provided on pp. 52 –

53.5 This paper will not deal with the self-

interest critique. Generally, BLE expressessurprise when participants act “fairly” andhence irrationality (by NEM standards) inthe ultimatum game (on experimentaleconomics see Smith, 2003, pp. 36 - 52).However, it is a ‘completemisrepresentation’ to equate (classical)economics with self-interest - economicsmerely says that selflessness is notnecessary for the market to work (Coleman,pp. 136 – 162).

6 Tversky and Kahneman, pp. 1127 – 1128.

reaching political consequences of thisheuristic can be found in the Love Canalincident (Sunstein, 2002, pp. 79 – 82).

8 Camerer et al., p. 1218.9 ‘Libertarian paternalism’: absence of

coercion (Thaler and Sunstein, p. 175);‘asymmetric paternalism’: imposes little/noharm on NEM but helps others (Camerer etal., p. 1212); ‘anti-anti-paternalism’:between paternalism and anti-paternalism(Jolls et al. pp. 46 - 47).

10 Camerer et al., pp. 1238 – 1247.11 Hayek, 1945, p. 520. Indeed, the

Hayekian economic problem has to besolved before the neoclassical problem(Kirzner, p. 50).

12 Boettke, 2002, p. 266. For example,Mullainathan and Thaler (p.5), believemarket trades occur due to theoverconfidence bias, “even in the absenceof true information.” Furthermore, neo-classical “search” theory is unrealistic, asKirzner (p. 50) implies: “It assumes thatthose whose information is incompleteknow how much information they lack, thatthey know the value to them of the missinginformation, and that they know preciselyhow (and at what cost) it is possible toobtain the missing information.”

13 Boettke, 1997.14 Ibid., p. 51.15 Nobelist Vernon Smith’s Hayekian

experimental economics researchprogramme is excluded from my critique ofBLE.

16 Hayek, 1967, pp. 313 – 317.17 Scientific progress shares the

characteristics of an Austrian marketprocess (Walstad, 2002). It would be absurdto argue that science progressed by themethods advocated by deliberativedemocrats for the socio-political sphere.

18 Pennington, 2003.19 Camerer et al., p. 1230.20 Gigerenzer, 2002; Cosmides and Tooby,

1996.21 Jolls et al., p. 37. Sunstein, 2002, pp. 78

– 98.22 Jolls et al., p. 37, 48.23 Johnston (p. 779) dissents:

“Technocrats can do little or nothing toreduce political barriers to economicallyefficient or socially desirable regulation.”

24 Barnett, pp. 138 – 142.25 Epstein, 2003, pp. 210 – 227.26 Blundell and Robinson, 2000.27 Camerer et al., pp. 1251 – 1254.28 A taste of the illiberal BLE tendency is

found in Jolls et al. (p. 49), where Stateintervention is needed to prevent peoplefalling asleep under sunlamps and burningthemselves! BLE also suffers fromeconomic illiteracy: Camerer et al. (pp. 1253/7 A dramatic illustration of the far-



4) believe extended warranties are profitablebecause they are costly to the consumer.However, consumers obviously value thepiece of mind and convenience ofwarranties more than their cost. Also, if BLEregulations stymie feedback learning,ceteris paribus, more errors will occur,leading to further regulation, etc.

29 Sowell, pp. 381 – 382.30 Cited in Smith, 2003, p. 23.31 There are two definitions of ecological

rationality, neither of which support BLErecommendations. The first is theHayekian, implied in 4.2. Smith (2003, pp. 7- 8) defines an ecologically rational orderas “an undesigned ecological system thatemerges out of cultural and biologicalevolutionary processes” and whichcontains “rules, norms and institutions ofour cultural and biological heritage that arecreated from human interactions but not bydeliberate human design.”

The second, as employed by Gigerenzerand Selten (p. 9) matches the structure ofbehavioral biases to the evolutionaryenvironment in which they evolved, so thatrationality is particular to time and place,not a universal rule of consistency orcoherence. Gigerenzer and Selten (p. 6)criticise the BLE authors referenced in 4.1for illustrating cognitive biases “withoutanalyzing the structure of environments.”

The work of Gigerenzer and Selten,Gigerenzer, and Cosmides and Tooby, showthat many cognitive biases disappear if theenvironment is included in the analysis, soCosmides and Tooby (p. 69) suggest we

give millions of years of evolution thebenefit of the doubt over a few hundredyears of normative theorizing. Hayek (1988)came to a similar conclusion in his “twoworlds” thesis, whereby different types ofrationality are needed in formal (societal)versus informal (familial) orders. In sum, theHayekian critique of constructivism isinseparable from the research of Gigerenzerand company.

32 For experimental economics evidencesee Smith, 2003, pp. 15 – 24. For example, incertain institutions, zero-intelligence actorsachieve efficient gains. In sum, “What welearn from such experiments is that anygroup of people can walk into a room, beincentivized with a well-defined privateeconomic environment, have the rules ofthe oral double auction explained to themfor the first time, and they can make a marketthat usually converges to a competitiveequilibrium, and is 100 percent efficient—they maximize the gains from exchange—within two or three repetitions of a tradingperiod. Yet knowledge is dispersed, withno participant informed of market supplyand demand, or even understanding whatthat means” (Smith, 1999, p. 198). Similarly,simple linear decision making models canoutperform multiple regression versions

(Gigerenzer, 2001, p. 45).33 Hanson. Smith, 2003, pp. 18 – 20.34 Camerer et al., p. 1247. Beaulier and

Caplan.35 The market works by reputation which

reduces the need for individuals to evaluateall options.

36 The liberal State cannot be neutralbetween all preferences (see Tomasi, 2001),but in minimising the need for the State totake sides, its relative amorality iseverybody’s second-best preference(Barnett, pp. 303 – 308).

37 Jolls et al., p. 49, argue that Statepaternalism can rely on informationcampaigns rather than coercion. They donot admit, however, that the State wouldhave to raise the money for such campaignsthrough additional coercive taxation.

38 Such rhetoric is succour to leftist media:“Regulation may help us make betterchoices” (Editorial, The Observer 17th

August 2003). However, this potentialabsurdity is captured by Stephen Pollard(The Times 26th August 2003): “I boughtsome lavatory paper the other day. What anightmare!...How is anyone supposed tocope with the chaos involved? Go into oneshop and it’s 27p for a roll. Go into anotherand it’s 32p, but that’s double strength....Something should be done. TheGovernment should legislate.”

39 Epstein p. 232.40 Sowell (pp. 331 – 368) offers discussion

and empirical support. See also Hayek,1998.

41 Quoted in Schmidtz and Goodin, p.74.42 Hayek, 1988, pp. 53 – 54.43 Epstein, p. 208.

Peter Rossi received his MSc from theLondon School of Economics in May 2003.

shareholder-based model of corporategovernance is flawed and that managersshould be answerable, as they are inGermany and Japan, to a wider set ofstakeholders, including employees—iswrong. The recent scandals have exposedfailings in the way US companies aremanaged, failings which are now attractinga flurry of reforms. But the scandals havenot negated the enormous strengths of theUS system as it has evolved over the last25 years.

The principal US advantage lies in thespeed with which new industries and

companies are created and old onesdismantled or shrunk. This advantagestems from four forces that first emerged inthe 1970s and that became more powerfulin the next two decades. The first was anew relationship between companies andtheir shareholders. The rise of investinginstitutions—principally pension fundsand mutual funds—as the dominant ownersof most publicly quoted companiesbrought to an end a long period in whichmanagers had been largely free from activeoversight by shareholders. Theseinstitutions, holding larger stakes incompanies than the private shareholdersthey replaced, had the power and incentiveto press poorly managed companies forbetter performance. At the same time, anarray of financial innovations made newsources of funding available to ambitiousentrepreneurs. The result was an increase

Capitalism Works

Sir Geoffery Owen

THE AMERICAN VERSION of capitalism hasbeen knocked off its pedestal in the lastthree years. After the stock market crashand the corporate scandals which followed,the triumphalism inspired by theproductivity surge in the second half ofthe 1990s has been silenced. What JohnKay has called the American businessmodel (Prospect, May 2003) is now widelycondemned as both immoral and inefficient.

What happened at Enron, WorldCom, etal was indeed a shaming episode in thehistory of US capitalism. Yet the conclusionwhich some critics have drawn—that the

in the number and size of hostile takeoversand in the supply of venture capital forstart-up firms. The second force wasinternational competition. With the entryinto world markets of low-costmanufacturers from Japan and other EastAsian countries, several US industries,especially those which had lived for yearsin a state of comfortable oligopoly, foundthemselves under pressure. Thirdly,deregulation transformed the structure ofindustries such as telecommunications,airlines and banking. As barriers to newentry came down, former monopolists suchas American Telephone and Telegraphcame under attack from start-ups. AT&Tsubsequently broke itself up into separatepieces, and each of these businesses isnow operating in highly competitivemarkets. Finally, advances in informationtechnology created huge opportunities fornew companies. These changes also had abroader impact on the way older companiesorganised themselves. The general trendwas for big companies to “deintegrate,”relying on outside suppliers for productsor services that used to be provided in-house.

The consequence of all this was to put asharp brake on managerial empire-building.Companies which had paid little attentionto shareholder value in earlier decades wereforced to use their assets more efficiently.During the 1960s, businessmen such asHarold Geneen at ITT had created large,diversified groups, claiming that theirsuperior managerial skills enabled them torun a variety of disparate activities—ranging, in ITT’s case, from hotels andinsurance to vehicle components andpaper-making. By the end of the 1970s, itwas clear that many of the businesseswould be better off on their own, or underdifferent ownership.

The number of mergers and acquisitionsin the US reached a historic peak in the late1980s; after a lull in the early 1990s, activityresumed at an even higher level in thesecond half of the decade. Many of thebiggest mergers took place in industrieswhich were going through a period ofturbulence as a result of regulatory ortechnological change. Although somedeals worked out badly, the resultingreallocation of resources almost certainlycontributed to the strength of the USeconomy during those years. Recentresearch by two American economists,Gregor Andrade and Erik Stafford,suggests that, taken overall, the mergersof the 1980s and 1990s served to increaseproductivity. Mergers can be an efficientmeans of dealing with excess capacity in

declining industries, while in risingindustries they allow well-run companiesto grow faster.

Despite the mega-mergers, the level ofconcentration in the US—the share ofindustrial output accounted for by thelargest corporations—has been stable ordeclining over the last 15 years. This is duepartly to spin-offs and divestments by largecompanies, and partly to the growth of newcompanies. In the chemical industry, forexample, new entrants, sometimes createdthrough management buyouts, have takenover activities that had previously been partof long-established groups such as DuPontand Dow. Much of the heavy end of theindustry—the production of low-valuecommodities such as petrochemicals andplastics—is now in the hands of companieswhich did not exist ten or 15 years ago.

The shareholder value revolution, linkedto innovation in financial markets, has madeUS industry more flexible and productive.To the extent that Britain moved in a USdirection during this period—and it did soto a much greater extent than otherEuropean countries—the results were alsobeneficial. Unwieldy conglomerates werebroken up, and companies were quicker toclose loss-making businesses.Deregulation allowed new entrants toprosper, Vodafone being a spectacularexample. And an expanding venture capitalindustry supported many business start-ups in new areas such as biotechnology.

Large-scale acquisitions are risky, but thatdoes not mean they should never beundertaken. BP bought two large Americancompanies, Amoco and Atlantic Richfield,in the late 1990s; the rationale for thesedeals seems to have been sound, and post-merger integration has been handled well.

Acquisitions can succeed or fail for avariety of reasons, most of which are hardto assess in advance. What the recordshows, however, and here the critics ofexcessive deal-making are on strongerground, is that the incidence of mistakestends to be greater during a raging bullmarket. In the late 1990s, euphoria over newtechnology drove prices of internet andtelecommunications companies—includingsome which had no early prospect of makinga profit—to absurd levels. A mood ofirrational exuberance took hold, affectingmanagers and shareholders.

In these circumstances, companieswhose shares are riding high are temptedto use those shares to buy othercompanies—an apparently easy way ofmaintaining the momentum of growth andfeeding the appetite of analysts andbankers, who are urging them to do more

deals. Whether stock options contributeddirectly to unwise acquisitions is notcertain, but there were, in any case, otherreasons for questioning the effectivenessof this form of remuneration. When theywere introduced in the 1980s, their purposewas to align the interests of managers withthose of shareholders. The subsequentescalation in top managers’ pay could bedefended on the grounds that companies,operating in a more competitiveenvironment than in the past, nowdepended to a greater degree onexceptionally talented managers, and thosemanagers were in short supply. (Tenure oftop executive posts was also becomingshorter and more precarious—hence thegenerous “golden parachutes” in the eventof forced departure.) By the second half ofthe 1990s, however, it was clear that somecompanies had been far too lavish with theirstock option awards. The stock marketboom allowed them to cash in their optionsat very high prices, even though theseprices might have owed little or nothing totheir own efforts. More seriously, asEdward Chancellor has shown (Prospect,June 2002), poorly designed optionschemes with short payout periods createdan incentive for unscrupulous managers toinflate reported profits by dubious or illegalmeans.

A reappraisal of executive remunerationis now under way, in Britain as well as theUS. There are moves to lengthen the periodbefore options can be exercised, or toreplace them, as Microsoft has done, withother ways of enabling managers to acquireand retain ordinary shares in the business.Companies may also be forced to disclosethe cost of stock options when they aregranted. One of the reasons why stockoptions were so popular in the US was theirfavourable tax and accounting treatment,which made them appear less expensive forthe employer than straight salary or otherforms of share-based remuneration; thatnow seems certain to change.

The need to rethink stock options, whileretaining the link between rewards andshareholder value, is one of several lessonsthat have come out of the stock marketboom and bust. Another is the need forauditing reform. While Enron was anextreme case, the practice of manipulatingquarterly earnings statements in order tokeep investors happy was by no meansconfined to “new economy” companies. Tosome extent these lapses may be self-correcting as investors and managers learnfrom their mistakes. But there is clearly acase for stronger regulation to ensure theintegrity of financial reporting, and to



prevent auditors from being “captured” bytheir clients, as Andersen was by Enron.

It is probably healthy, too, thatinstitutional investors are taking a strongerline over practices, such as over-generousexecutive rewards, which they regard asdamaging to the interests of shareholders.In Britain this greater activism may be partlydue to pressure from the government: NewLabour is now more interested in makingthe shareholder-based system work wellthan in shifting to the stakeholder model.There will always be room for argument, asin the recent ousting of Michael Green fromthe chairmanship of the Carlton/Granadagroup, over whether institutional investorsor boards of directors know best- and it is amistake to put too much faith in shareholderactivism as a cure for underperformance.Nevertheless, as owners, the biginstitutions have a responsibility to bevigilant and to intervene when things aregoing wrong.

A regulatory response to Enron wasnecessary, but it would be wrong tooverreact. The benefits of stricter rules oncorporate governance need to be setagainst the potential costs, not least indeterring companies from going public.After all, the US economy has come throughthe stock market crash and its recession indecent shape, with business productivitygrowing at 3.5 per cent a year since 2000,much faster than in the eurozone. Theevents of the last three years do notundermine the case for well-developed anddemanding financial markets, nor do theycancel out the gains that stem from theshareholder value revolution. Indeed, thebursting of the bubble should strengthenthe US system if, as seems likely, it helps tore-establish what shareholder value reallymeans.

The definition has to start from thepurpose of the corporation, which is to

create wealth. Its success in doing so isshown by its profitability—its ability toprovide goods and services at a price whichmore than covers the costs it has incurred.If the corporation achieves this in acompetitive market, there is a reasonablepresumption that it is using its resourcesefficiently, and thus contributing to socialwelfare.

Shareholders are interested in expectedas well as current profitability, and shareprices should reflect the market’s estimateof what future earnings will be; it is a signalto managers of how well they are doing.But from time to time the stock market canfall out of line with reality, and, as Harvardbusiness professor Michael Jensen hasremarked, an overvalued share price can beas dangerous to a company as anundervalued one: it can tempt managers todo foolish things to keep it up. Recentexperience (not just at Enron) has shownthat a focus on boosting the share price inthe short term can lead to the long-termdestruction of value on a massive scale.The proper task for managers is to createwealth over the long term; it is for them todecide what the appropriate time frameshould be and to explain to investors howtheir decisions contribute to that objective.

This is a different goal from that ofpromoting the survival and growth of thecompany. Most senior managers are loyalto their company and want it to prosper.But it may happen that the sector in whichthey are competing enters a period ofmaturity or decline. The appropriateresponse is to get smaller, to demerge orsell off parts of the business and to returncash to shareholders. In that sense theinterests of shareholders—and byextension those of society as a whole—take precedence over those of otherstakeholders, including senior managers,in the enterprise.

This does not imply a lack of concern foremployees and local communities. A loyalworkforce and a reputation for fair dealingare assets which should contribute to long-term value creation. But to give these otherconstituencies equal weight withshareholders, as urged by some advocatesof stakeholder or “Rhineland” capitalism,is a recipe for confused accountability—ineffect, giving managers the freedom to setand pursue their own objectives.

These disadvantages are less serious aslong as the market in which stakeholder-based companies are operating iscompetitive. The post-war success ofGermany and Japan owed more to the vigourof internal competition than to thosecountries’ distinctive corporategovernance arrangements. It would also bewrong to exaggerate the link between thecreation of shareholder value and the threatof hostile takeover; after all, Toyota hasdone far better for its shareholders in recentyears than any US carmaker and Hoechsttransformed itself without the threat oftakeover. Nevertheless, financial markets ofthe sort which have taken shape in the USover the past two decades are a powerfulinstrument for constraining the power ofincumbent management and promoting thegrowth of new enterprises. The absence ofsuch markets provides greater stability foremployees and a cosier life for managers,but it makes adjustment to change moredifficult. Those concerned about theproductivity gap between Britain and theUS, such as Gordon Brown, shouldembrace the pursuit of shareholder valueas part of the solution.

Sir Geoffrey Owen is a senior fellow atthe LSE’s Institute of Management and aformer editor of the Financial Times

ALTHOUGH FREE TRADE has been acontentious issue ever since its inceptionas a theoretical proposition in the lateeighteenth century, it “remain[s] one of the

most durable and robust propositions thateconomic analysis has to offer for theconduct of economic policy.”1

Unfortunately, adopting free trade policies

is far from a straightforward decision-making act, as the multidimensionality ofinterest groups involved inevitablycomplicates the process. This realization isprecisely what US President George W.Bush has painfully come to himself duringthe course of the past nearly two years.Rescinding on December 4, 2003 thesafeguard tariffs he had imposed on steelimports 21 months earlier did successfullybreak the deadlock of an impending $2.2billion worth of retaliatory sanctions by theEuropean Union but Bush’s determination

Vladimir L. Andonov

Steel Not Getting it Right

to make this move was not motivated bysound economic argumentation. Neitherwas this the case in March 2002, when hemade the decision to unilaterally introducehigher steel tariffs. In both cases, the USpresident failed to get the calculus of freetrade right. Even though Bush made thecorrect choice in the end, his faultyunderlying justification for doing so doesnot bode well for the future of tradeliberalisation.

Needless to say, disentangling thecomplex argumentation surrounding thesteel case can be a daunting endeavour.For analytical purposes, therefore, theprocess can be decomposed into two‘stages,’ which I designate as consecutivedecision nodes. The first one involves therun up to President Bush’s decision toimpose the steel safeguards. What were thePresident’s main objectives in making hischoice, and were they subsequently met?Similarly, the second decision node relatesto the rationale that drove Bush to hisDecember 4 announcement. It should berather obvious that the latter decision-making process strongly hinges upon whatthe outcome of the preceding node is. Whatis perhaps somewhat more nuanced todiscern, however, is that an erroneousinterpretation of the intra-stageenvironment can easily engendermisleading conclusions regarding therationalization of the final outcome. I willargue that precisely this qualification iswhat explains why the Bush administrationpresented the case for termination of thesteel tariffs in the light it did.

At the outset, what was the motivationfor imposing the steel safeguards in March2002? According to Gary Hufbauer and BenGoodrich from the Institute for InternationalEconomics (IIE) in Washington, DC,“President Bush’s decision to protect theUS steel industry with Section 201 tariffshad no economic justification.”2 Beforerelief was granted, they had calculated thatsafeguard tariffs would cost over $400,000annually per job saved in the steel industry.Moreover, they would result in net joblosses in the economy due to downstreamlayoffs. Instead, a more likely explanationfor the tariffs is that they “were driven notby an economic match between problemsand solutions but by two politicalmotivations”3: passing Trade PromotionAuthority in Congress, and buying thesteel industry’s support in congressionaland presidential elections.

In fact, trade was not so much the problemas the administration was led to believe,and trade protection, therefore, hardlyconstituted the correct move at the first

decision node. Most observers at that timefailed to recognize that minimills were – andcontinue to represent – the main forcecrowding out steel firms.4 The US steelindustry consists of high-cost integratedmills and minimills, almost evenly splittingthe market amongst themselves. Over halfthe decline in traditional integrated steelproduction is attributable to the rise of thedomestic minimills due to their superiorproductivity. Each employee in Nucor’sminimills, for instance, makes three timesas many tons of steel as each employee atUS Steel, the largest integrated steelproducer in the United States. To be sure,“both input and output mixes differbetween the two firms, but the differencesdo not cancel out the crude comparison ofannual tons per worker.”5

The two primary problems that the USsteel industry faces are legacy costs andpersistent overcapacity. Legacy costs,which include pension and health carebenefits, represent a much greater burdenfor integrated mills because of the massivepool of retired workers these companieshave. Persistent overcapacity reflects theproclivity of the government to slow downthe closure of high-cost steel mills; it is notjust a cyclical phenomenon. The result istoo many firms, operating on too small ascale.6 To alleviate both burdens, theindustry should emphasize measures thatreduce capacity permanently, throughconsolidation, and import tariffs arecertainly not the right instrument to achievethis. In fact, additional safeguards wouldonly further exacerbate the problem. Thelogic is straightforward: steel tariffs raisethe domestic price of steel, which in turnallows inefficient steel producers to makeprofits more easily and, hence, stay in themarket. From this process, known as marketdistortion, follows that least efficient firmsare not necessarily the first to close theirdoors.7

The President’s stated rationale for thetariffs was to give the industry breathingspace in which to restructure.8 Steellobbyists point out that the industry hasconsolidated in the past year. Yet, theynever adequately explain how thesafeguards hastened the pace ofconsolidation, and instead simply assert arelationship between the safeguards andconsolidation. Consolidation in the steelindustry only occurs when a steel-producing firm is performing very badly andsells itself to a stronger firm. In contrast,by raising the profitability of ailing firms,the tariffs may have raised the price thestronger firms had to pay to acquire them.9

“All of the consolidation that occurred in

the steel industry since 1997 happened inthe context of bankruptcy, mostly beforethe tariffs were introduced in March 2002.In this sense, bankruptcy is a prerequisiteto consolidation.”10 By increasing steelprices moderately, the Bush administrationmay actually have propped up weak steelfirms for a longer period, delayingconsolidation.

Thus, we arrive at the second decisionnode. I take the December 4, 2003 WhiteHouse Press Briefing to stand for theofficial position of the Bush administrationon why steel tariffs needed termination.Based on that, I derive that the rationalebehind rescinding the tariffs related neitherdirectly nor genuinely to the existing steelindustry context. Indeed, some of thejustifications seem outright misconstrued.This is exactly what I cautioned against:that perceptions of realities are of criticalweight when forming policy preferencesand making decisions. President Bush wasright about repealing the tariffs; yet, hisparticular motivation for doing so appearsgreatly unfounded for four reasons, eachof which requires elaboration.

First, US officials sought to play downthe significance of the EU retaliation threat,arguing that the move was merely aresponse to the recovery in the market forsteel.11 Robert Zoellick, US traderepresentative, fittingly captured thestandpoint of the administration: “Thisdecision was independent of that [the EUthreat].”12 Instead, Zoellick maintained thatthe conclusion had been based on astraightforward cost-benefit analysis.However, it is hard to comprehend how nottaking a possible economic and politicalimpact of sanctions of an amount of $2.2billion can be presented as legitimate cost-benefit analysis. Ian Rodgers, director ofUK Steel, seems to have discerned muchbetter what underlay Bush’s move: “I haveno doubt that the president has been forcedinto this decision by the knowledge that ifthe tariffs had remained in place he wouldhave faced EU sanctions on •2.4 billion ofUS exports.”13

Next, the Bush administration arguedthat “the tariffs […] clearly had worked;”14

that jobs had been saved, and that steelbusinesses had been given another chanceto compete. Admittedly, there are signs that“inefficient capacity has been cut,productivity has risen sharply, and pricesare now 15 percent to 30 percent higher thanin February 2002.”15 Nevertheless, tariffs“are not painless for America.”16 Whilehigher steel prices benefit steel producers,they hurt steel users, of which many aremanufacturing companies also struggling



against foreign competition. For this reason,the administration “has not reaped thebenefits it expected from the introductionof tariffs in March 2002.”17 The IIE calculatesthat the cost to steel users so far has beenabout $600 million in lost profits from higherprices and 26,000 lost jobs. That dwarfs thebenefit to American steel firms, which theIIE reckons has been only $240 million,mostly from a 3.3% rise in average steelprices, with some 5,000 jobs saved.18 If thetariffs ‘clearly had worked,’ one wouldexpect their welfare impact to be somewhatmore supportive of such an argument,which clearly is not the case. In short, “theconclusion is that the safeguards areunambiguously a drag on the USeconomy.”19

Third, the White house briefing seems tosuggest that administration officials playeddown the likelihood that they would face astrong backlash in the steel states becausethey had lifted the tariffs early. However,Zoellick’s claim that “…this is an industrythat wanted help for a number of years. Itgot help.”20 is not supported by theevidence. For one, the fact that Bushpromised to continue a monitoring andlicensing scheme to guard against the riskof future surges in steel importsdemonstrates that the President has madeonly a small concession to his domesticindustry.21 For two, the union representingUS steelworkers promptly demonstrated itsdiscontent by calling the termination oftariffs “clear evidence of capitulating toEuropean blackmail.”22 The union’s reactionlends further support to the proposition that“in some sectors of the steel industrydemands for protection have becomevirtually institutionalized, a part of theindustry’s ongoing business strategy.”23

Finally, the administration’s perspectiveon the existing environment of the steelsafeguards could additionally have beenrefracted through the lens of the latesteconomic recovery. Arguably, the decisionto lift the tariffs was aided by the recentstrong growth in the US economy anddeclines in the dollar, both of which havebolstered the steel industry.24 Moreover,there appear to be healthy signs that jobcreation is picking up. However, this canhardly conceal the fact that the decision issure to draw sharp criticism from labourunions and several of the Democraticpresidential candidates, who say theincipient recovery is largely bypassingmanufacturing workers in the industrialstates where steel remains a force.25 Inaddition, even the relief provided bycurrently higher world steel prices due toChina’s insatiable appetite for raw materials

cannot represent a permanent shift in thedynamics of the industry.26

All things considered, one couldconceivably suggest that President Bushnever really had a chance of getting thetariff thing right. Yet, to maintain such adeduction is to miss the central argumentdeveloped here. Bush could have muchmore easily ‘gotten it right,’ if he hadadmitted having seriously taken the EUthreat into consideration; if he had revealeda more realistic picture of the US steelindustry by agreeing that there is still a lotmore consolidation left to be done; and ifhe had demonstrated that he haddownplayed electoral considerations whenweighing the pros and cons of tradeliberalization. In this respect, EU tradecommissioner Pascal Lamy is misguided toassert that “Whatever the reasons were,we got the results we wanted.”27 Gettingthe causal relationship of the process offree trade decision-making right is just ascritical as arriving at the decisionsthemselves, if not even more so. If we getthat right, there could be a brighter futurefor the world trading community, especiallyin light of the current post-Cancúnstalemate in the WTO Doha Round. It ishigh time that we realize that thecontemporary case for free trade is “notthe old argument that free trade is optimalbecause markets are efficient.” Rather, “itis a sadder but wiser argument for free tradeas a rule of thumb in a world whosepoliticians are as imperfect as its markets.”28

Endnotes1 Douglas, Irving (1996) Against the tide:

an intellectual history of free trade.Princeton. p. 230.

2 Hufbauer, Gary Clyde, and Ben Goodrich(2003a) Steel Policy: The Good, the Bad,and the Ugly. International EconomicsPolicy Brief 03-1. Washington: Institute forInternational Economics.

3 Ibid.4 Ibid.5 Ibid.6 Cold Steel. The Economist. November

13, 2003.7 Hufbauer, Gary Clyde, and Ben Goodrich

(2002) Time for a Grand Bargain in Steel?International Economics Policy Brief 02-1.Washington: Institute for InternationalEconomics.

8 Cold Steel. The Economist. November13, 2003.

9 Ibid.10 Hufbauer, Gary Clyde, and Ben

Goodrich (2003b) Next Move in Steel:Revocation or Retaliation? International

Institute for International Economics.11 US to Dismantle Steel Tariffs and Avoid

Sanctions. Financial Times. December 5,2003.

12 White House Press Briefing. December4, 2003. Accessed at http://www.whitehouse.gov/news/releases/2003/12/20031204-11.html on December 6, 2003 at 02:59.

13 Lifting of US Steel Tariffs WelcomedBut Fears Remain about ‘Backdoor’Penalties. Financial Times, December 5,2003.

14 Bush Rescinds Tariffs on Steel Imports,Averting Trade War. New York Times,December 4, 2003. Accessed at http://www.nytimes.com/2003/12/04/national/04CND-STEE.html on December 5, 2003 at02:17.

15 Ibid.16 Cold Steel. The Economist. November

13, 2003.17 Sparks Fly over Steel. The Economist.

November 13, 2003.18 Hufbauer, Gary Clyde, and Ben

Goodrich (2003b) Next Move in Steel:Revocation or Retaliation? InternationalEconomics Policy Brief 03-10. Washington:Institute for International Economics.

19 Ibid, emphasis in original.20 White House Press Briefing. December

4, 2003.21 US to Dismantle Steel Tariffs and Avoid

Sanctions. Financial Times. December 5,2003.

22 Ibid.23 Goodman, J, et al (1996) Foreign

Investment and the Demand for Protectionin the United States. InternationalOrganization. Vol. 50, No. 4. p. 579.

24 US to Dismantle Steel Tariffs and AvoidSanctions. Financial Times. December 5,2003.

25 Bush Rescinds Tariffs on Steel Imports,Averting Trade War. New York Times,December 4, 2003.

26 Sparks Fly over Steel. The Economist.November 13, 2003.

27 EU Wins Battle over US as Bush Backsdown on Steel. EUObserver.com. December5, 2003. Accessed at http://www.euobserver.com/index.phtml?aid=13791 onDecember 6 2003, at 02:56.

28 Paul Krugman (1987). Quoted inDouglas, Irving (1996) Against the Tide:an Intellectual History of Free Trade.Princeton. p. 227.

Vladimir L. Andonov is currentlystudying at the London School ofEconomics and will be completing hisdegree in Economics and Political Scienceat Williams College.Economics Policy Brief 03-10. Washington:


elected they have incentives to take actionsthat favor them and their Party. Hence,projects that provide concentrated benefitsfor special interest groups but greater coststo the general tax-paying populace areenacted into law.

So how do we prevent special interestdominance? Politicians currently have noincentive to tell the truth in campaignspeeches; they can later behave differentlywithout consequences. They are not, inshort, “putting their money where theirmouth is!” They have, in the jargon ofeconomics, property rights in political lies.The incentive-correcting mechanismadvanced here reassigns those propertyrights from politicians to voters. Themechanism could be worded as follows:

“Any Party wishing to place a candidateon the ballot for an impending election, must(in addition to existing requirements)indicate the total spending, S, that it willincur over the four-year term of election. Ifthe elected officials of the Party spend morethan S, the Party is itself liable to pay anyamount exceeding S into a fund that will beused to retire the national debt, except underspecific circumstances to be discussedfurther below.”

The spending limit, S, is all the Partiesneed reveal. The elected Party can spendtheir budget as they wish, even lying in thespeeches prior to election about theirspending intentions, although that mightdamage their reputation, hence futureelection prospects. Moreover, they canallocate S among the four years any waythey want—after all, the future cannot bepredicted and flexibility is desirable.Importantly, however, it would be irrationalfor a Party to exceed S since it would eithergo bankrupt or be replaced at the nextelection (likely both). Since Parties aren’tirrational, it is exceedingly unlikely thatthere will ever in fact be any money in the“fund.” The timing of the announcementof S could be debated.

Might not too many inefficient projectsstill be included in overly large S’s offeredby the Parties vying for office? This remainsproblematic under a BBA, although growthin such projects might be discouraged.Under the proposed mechanism, however,political competition will tend to giveAmericans the overall level of governmentexpenditure they wish over time. A common

criticism of BBAs is that a BBA is just aruse to halt spending growth, when somepeople really want more, not less, spending.Under the proposed mechanism,Americans wanting bigger government canvote for the Party offering a larger S, thoughI suspect that the majority would prefer asmaller more efficient government.

Total spending percentages have grownregardless of which Party is elected. Thisis expected, given the faulty incentivescurrently facing politicians. Under themechanism, the Party ultimately winning thenational election would likely proposemodest expenditure cuts, say an initialrollback to 19.2% of GDP (19.7% was theactual figure for 1998). Political competitionmight force percentages lower in futureelections.

With overall governmental spendinglimited to S, focus would appropriately shiftto the efficiency and equity impacts of thatspending. This does not, of itself,necessarily imply that programs wouldbecome more efficient. People appear tohave an affinity for wasteful agriculturalpolicies, for example. Moreover,measurement of benefits and costs formany government programs is difficult;increases in efficiency might emergeslowly.

But, many inefficient programs arestraightforward to analyze (e.g., theagricultural policies). Efficiency gains fromreform could be combined with transfers tomake all farmers better off, if that weredeemed fair. If concern is that poor farmerswill be forced from family farms, a means-based test could be applied, with moreefficiency gains being returned to votersin the form of lower food prices.

What if the elected political Party canonly “try” to deliver its promises, therebeing cases in which it is unable to do so?If the elected Party is forced bycircumstances beyond its control to exceedtheir S limit, it might be widely viewed asunfair to impose liability on them. Threeimportant cases of this problem come tomind. First, a national disaster or a largewar in some future year might occur after aparty, promising to spend S, has beenelected. In such cases, a congressionalvote could be taken allowing a temporarysupplement to S that would not countagainst the Party’s limit, using the samesuper-majority rules as advocated under aBBA.

Funds can be moved among differentexpenditure categories, within the overallS, in the event of more minor disasters orwars. For example, resources could betransferred from social programs to defense

Philip E. Graves

Reforming Governments

WHAT IS WRONG with politics today? Whyare voters so turned off? Voters know that,despite campaign promises, politicianshave incentives to incur deficits sinceincreasing spending is politically desirablewhile increasing taxes is not. Additionally,politicians promote inefficient expendituresthat benefit them, notably special interestabuse.

The preceding has led to growth ingovernment. Considering the US, thepercent of GDP spent by all levels ofgovernment was 22.8 (1950s), 25.1 (1960s),28.2 (1970s), 30.6 (1980s), and 30.5 (1990-1998).

Large US deficits caused some toadvocate a Balanced Budget Amendment(hereafter BBA) amendment. BBAs requireannually balanced federal budgets, exceptin times of war or national emergency.Deficits may then be run if both the Houseand Senate vote to do so with a super-majority, details differing among BBAproposals. The presumption underlying theBBA is that it would reduce the growth ofgovernment.

Many economists are concerned that aBBA would worsen recessions. In arecession, some expenditure automaticallyincreases (e.g. food stamps, unemploymentinsurance), while tax revenue automaticallyfalls. Balanced budgets necessitate eitherspending cuts or increases in taxes,possibly deepening a recession.

Another objection to BBA proposalsstems from variation in voter desires aboutthe size of government: any BBA wouldhave the practical result of cutting spendinggrowth. Some voters argue that large levelsof government spending are necessary toreflect the diversity of opinion about whatshould be funded. While a voting majorityis likely to support smaller government,some feel that a large government doingmany things reduces the tyranny of themajority over the minority.

I present an alternative mechanism forachieving the benefits of a BBA, along withmany other benefits, without its drawbacks.The approach corrects a flaw that exists atall levels of government, foreign anddomestic, although the discussionemphasizes the U.S. federal government.

The well-known flaw is that regardlessof what the candidates of Parties vying forthe presidency say prior to election, once


should a small, unanticipated war break out.Or, conversely, a minor disaster mightrequire transfers from military accounts toemergency aid agencies.

Second, what about spending that isbeyond the control of politicians, beingbuilt into the system and dependent on thelevel of income? If a recession were to occurafter election, spending automaticallyincreases and tax revenue automaticallydecreases (the “built-in stabilizers” of earlymacroeconomic models). The mechanismproposed here is less pro-cyclical than theBBA with its focus on deficits. However,spending variations that occurautomatically in a recession (e.g.,unemployment insurance, food stamps)may indeed be desirable, could optionallynot be counted toward S. The critical thingto be guaranteed by the mechanism is thatexogenous politician-increased spendingthat ultimately violates S not be allowed.Another option is that major recessionscould be treated as with major disasters,leaving the Party in power responsible forminor fluctuations.

Finally, what if a candidate of Party A,promising to spend S, is elected to thepresidency, while another Party B controlsone or both of the House ofRepresentatives or Senate? The mechanismadvocated here might most easily be firstadopted in a parliamentary system, sincethe majority party appoints the PrimeMinister. However, should the executiveand legislative branches be split, themechanism need not be enforced. But, itwould quickly be seen to be irrational toelect a Congress controlled by a Party thatdiffered from the President’s Party; thevoters would be thwarting their own desiresto obtain the S that they themselvesprefer!Operationally, the auditingassociated with the mechanism is simple,with spending being actual dollar outflowsover the time until the next election. Futurepromises to spend are, on the whole,valueless if occurring further out than thepresent four-year period. The mechanism’sinformation requirements are actuallysimpler than those of a BBA that requiresannual numbers for both expenditures andrevenues. For projects that can only becompleted in a longer time frame, onlycurrent period expenditures would countagainst the current budgetary period, withfuture period expenditures included in theS of the Party running for re-election. If thatParty does not get re-elected, some suchprojects might be eliminated. Indeed, votersmight reject an incumbent Party preciselyto halt certain projects (e.g., “Star Wars”defense initiatives or an overly generous

welfare expansion). Should a newly electedParty wish to continue the projects of aprior administration (as expected withpopular projects), they must takeresponsibility for this in their S.

The political Parties might, especiallyinitially, be risk averse, running on a higherS than they really plan on spending. Theresulting surpluses would be notundesirable, and such liability fears shoulddiminish over time.

There is great political competition in thepresent form of government. This is good,and is critical to the benefits of shifting tothe proposed mechanism. Because of thatcompetition, candidates of Parties hopingto get, and remain, elected will wish toincorporate, within their fixed S, policiesseen by voters to be equitable and efficient.Little enforcement will be needed, becausethe mechanism is self-policing, withenforcement analogous to that of any othersocial contract. In cases of ambiguousauditing (likely to be rare), courts coulddecide whether the conditions of the“contract” had been violated.

One would increasingly expect transfersto the poor to involve means tests.Inefficient policies that are currentlyrationalized as “helping the poor” (but that,of course, help many special interest groupsof means) would be replaced by meanstests enabling Party candidates to run onsmaller S’s.

Elected representatives of the Party incontrol will be more likely to seek programsthat are efficient or otherwise appeal to amajority of voters under the proposal, sincethat will make their S more appealing. WithParties meaningfully constrained by theirspending promises, debate may turn, morethan at present, to issues of the regulatoryburden.

Everything argued here applies with equalforce for state and local governments whichspend two-thirds as much as the federalgovernment and where spending as apercent of GDP has doubled from 7% in1953 to about 14% today. Indeed, onemechanism by which the proposedmechanism might spread is for states toimplement it first.

The proposed mechanism is likely tohave several other benefits. First, it shouldallow real incomes to grow more rapidly ifpolitical competition results in the S of theelected party being a smaller percentage ofGDP, presuming a higher rate of return toprivate sector investment. Second, themechanism encourages privatization ofthings that should never have becomecentrally planned. Third, there will be moreincentive to find low-cost suppliers of

government goods. Fourth, the incentivefor agencies to spend heavily prior to theend of a fiscal year (“use it or lose it”)evaporates, since the Party could savethese resources either for futurecontingencies or for advertising that they“did what they said they would and camein under budget.” These effects areenhanced if Parties eliminate inefficientprojects in their competitive scrambling todeliver lower S’s.

Many of the longer-run benefitsstemming from the mechanism are a resultof political competition. That competitionhas been very intense, since so muchspecial interest largesse is involved.However, competition will remain highunder the proposed mechanism, only thatcompetition will be to best please the voterrather than special interest groups!

It should be noted, however, that evenspecial interest groups, collectively, mightbe better off under the proposedmechanism. Much lobbying expense is“defensive” in nature, undertaken to offsetlobbying efforts of broadly definedcompetitors. And it is likely that muchlobbying is undertaken for projects thathave, in fact, benefits greater than costs;such projects would likely be undertakenin any event. Under the mechanism, moreresources will flow into goods productionand less into political manipulation, raisingwelfare.

Indeed, the principle harm from theproposed mechanism is to the politicalParties themselves. Political contributionsby special interest groups currently makepolitical Parties much better off. If the likelyreductions in those contributions to thepolitical Parties are viewed as unfair, a fixedamount of funding could be provided bygovernment (perhaps set at some fractionof actual recent expenditure) to all Partiesreceiving more than some minimal percentof the popular vote. Having done this, therewould be no reason to allow any corporateor individual contributions, and that couldbe made illegal. Campaign finance reformwould, in any event, become much less ofan issue under the proposed mechanism.

The pressure of the media, focused toforce political parties to “put their moneywhere their mouth is,” should not be under-emphasized. If “we, the people” clamor forit, the Parties will eventually adoptvoluntarily or be forced by legislation toadopt the mechanism advocated.

Philip E. Graves is a professor ofeconomics at the University of Colorado atBoulder.


Returning Planning to the Market:An Agenda for Private Land Use Control

Mark Pennington

“THE ISSUE IS not whether one ought orought not to be for town planning, butwhether the measures to be used are tosupplement and assist the market or tosuspend it and put central direction in itsplace. The practical problems which policyraises here are of great complexity, and noperfect solution is to be expected.”

F.A. Hayek (1960, p.350) The Constitutionof Liberty

The British town and country planningsystem represents one of those time oldparadoxes of political economy. Politiciansare agreed that the system encapsulates therigid bureaucracy and special interestcapture that are the hallmarks of a plannedeconomy and yet few have any idea howto reform it.

The British Land Use Planning SystemThe British land use planning system is

one of the few remaining relics of the Atleegovernment’s post-war nationalisationprogramme. Whilst virtually all societies inthe Western world have instituted somesort of government land use control in thelast 100 years, none have gone so far as tointroduce anything approximating theprovisions of the 1947 Town and CountryPlanning Act. That act nationalised theright to develop land and in its keyessentials has remained virtuallyunchanged to the present day. Outside ofthe agricultural sector virtually all land usechanges are subject to the requirement toattain planning permission from a localplanning authority, which must itselfformulate local land use plans on the basisof national planning guidelines. Thephilosophy underlying the post-warplanning legislation was fundamentallyhostile to private property and marketprocesses, with the initial intent to replacethe operation of the private landdevelopment market with a state directedsystem, epitomised by the New Townsprogramme. As a consequence the planningsystem has mutated into a ‘mixed economy’model, where the majority of productionand investment decisions are made byprivate developers and landowners, butwhere the latter are subject to a detailed setof regulatory procedures.

Many of the procedures embodied in the British land use planning system are of the‘command and control’ variety which,rather than working with the grain of marketforces, attempt to suppress the operationof the price system. It is not much of anexaggeration to say that the practice ofdrawing up five or ten year plans forhousing and commercial land supply owesmore to Stalin than it does to Adam Smith.Planning decisions themselves, meanwhileare not the subject of a competitive biddingprocedure, but are imposed via political fiat.It is the inefficiency and waste that resultsfrom this chronic suppression of the pricesystem that I want to focus on in this article.

Planning without Prices: DistortedInformation

A well functioning price system is crucialto the efficient allocation of resources in acompetitive market economy. Changes inthe structure of relative prices for bothfinished products and for inputs of land,labour and capital enable individuals tomake calculations of economic value inorder to determine which combinations ofresources generate the highest value fromthe minimum set of inputs. Contrary to theconventional wisdom, the case for themarket economy does not rest on theconcept of a “perfect market” as taught inA-level economics. Such a perfect marketdoes not exist anywhere, any more than aperfect government or a perfect planningoffice. The key question is this: Doindividuals, looking at price signals and theinformation that is available and taking intoaccount their own preferences make bettermore informed decisions in general thanbureaucrats sitting in Whitehall and TownHalls? In nearly all situations people wouldagree that they do. In what sense is landuse different? If it is different, then Isuggest that we adopt land use policies thatrun with the grain of the market and notignore it all together.

Like many bastions of governmentregulation, the British planning system isfrequently defended on ‘market failure’grounds and in particular is said to beessential in counteracting the externalitiesand neighbourhood effects that areconsidered an integral part of the market in

land. At no point, however, do thedefenders of the planning system accountsatisfactorily for the way in which plannersare to ‘correct’ for such market‘imperfections.’ The values and associatedtrade-offs that are attached to the protectionof green-field sites, the location of newshopping developments, the environmentalimpact of different transport patterns, etc.are subject to precisely the sameuncertainty and imperfect knowledge asany other goods and services. Planners,however, are not subject to a signallingmechanism equivalent to the profit and lossaccount, which can indicate their successor failure in responding to the relevant trade-offs. Similarly, planners do not have accessto a set of relative prices for differentenvironmental goods in order to take theirdecisions. Two illustrations from thecontemporary British planning systemshould suffice to illustrate the problemsbrought about by the suppression ofcompetition and the price system in thisregard. Consider first the topical issue of‘sustainable urban form.’

The urban land use debate has beendominated in the recent past by aconfrontation between those highly criticalof the environmental effects of low-densitycommercial and residential development,and writers who are equally critical ofproposals to encourage high density,compact development patterns. The formerargue that higher density developmentsreduce the need for car based travel andlonger commuter or shopping trips, whereinpeople are able to access a wider range ofservices within a smaller surface area. Thelatter, by contrast, contend that higherdensities may actually increase car use.Although people may travel longerdistances in low-density areas, thefrequency of these visits (to a large hyper-market, for example) tends to be less, so itis not at all clear that discriminating againstsuch developments will do anything toreduce auto-based pollution. In practice,there would appear to be considerableuncertainty concerning the effect ofdifferent urban forms on transit patternsand related levels of pollution. Centralgovernment, spurred on by the Urban TaskForce and increasingly by professionalbodies such as the Royal Institute forChartered Surveyors, has issued guidelinesrequiring that local authorities adoptstrategies of high density, ‘brown-field’development to the complete exclusion ofother alternatives. Both government andthe professions it would seem have failedto learn the lessons of the last attempt tosecure ‘sustainable urban form’ as


manifested in the post-war New Townsprogramme. New Towns were supposed tolead the way to a world of minimum auto-use, but most commentators now agree,ended up increasing the overall amount oflong distance commuting. The questionmust be asked: Have the planners reallytaken into account all the costs and benefitsof different ways of planning our town orare they imposing their subjectivepreferences on the rest of mankind. Or, asFrederic Bastiat put it 150 years ago, “atwhat height above the rest of mankind doour rulers believe they stand?”

Given the massive uncertaintiesinvolved, the most appropriate way ofdealing with these issues may not be todeliberately plan for an ‘optimal’ urban form,but to permit a wider variety of experimentsin urban living. The latter may allow acompetitive discovery process to revealwhich particular ways of organising urbanareas work best from the subjective viewof the consumer as signalled by the relativewillingness to pay for different types ofdevelopment scheme.

The detrimental effects of suppressingthe price system are still more apparentwhen one turns to the supposed ‘jewel’ inthe crown of the British planning system–Green Belt policy. As a blanket ban ondevelopment covering some 14% ofEngland, Green Belts pay virtually noattention to the huge variations inenvironmental quality that occur within thedesignated land. The London Green Belt,for example, whilst including wooded hillsand chalk downs, also includes large tractsof land on the western and eastern urbanfringes, consisting of dis-used gravel pits,quarries, and low-grade farmland/horticultural developments. Whilst there isclearly a desire of citizens to preserveaesthetically attractive sites within easyreach of the city, it is equally the case thatpeople searching for affordable housingmight be prepared to see the relatively lessattractive parts of the Green Belt developedfor residential purposes. The problemshighlighted above stem primarily from asuspension of the signalling functionprovided by market prices. Without beingguided by a set of relative prices,highlighting variations in subjectiveenvironmental quality between differentsites, planners are unable to know how tochoose effectively between competinguses for land. The results are predictable.

Returning Planning to the Market: AProperty Rights Agenda for InstitutionalReform

is accurate then the most appropriate wayto cure the ills of the British planningsystem would be to make greater use ofcompetitive market processes so as toincrease the amount of experimentation, togenerate more price signals and to changethe incentives facing the various actorsinvolved. The question remains, of course,can a market driven system of land usecontrol be trusted to operate effectively inan area frequently associated with ‘marketfailure,’ and if so, what is the mostappropriate way of introducing such asystem. I think the answer to the first ofthese questions is a definite yes.

In considering this statement it isimportant to emphasise that the case for amarket system of land use regulation doesnot challenge the need for ‘planning’ assuch, but rather questions the legitimatesphere over which any particular ‘planningmodel’ should be extended. Marketprocesses themselves involve aconsiderable element of ‘planning’properly conceived. As the theory of thefirm teaches us, firms are ‘planningorganisations’ that emerge where there areefficiency gains to be made from replacingpurely exchange-based systems with ahierarchy of command in order to reducetransactions costs.

How much ‘planning’ there should be,(which firms should exist and how big theyshould become) however, is something thatcan only be discovered by a process ofopen competition between differentorganisational forms. In the specific caseof urban land use, it may well be the casethat there is a need for institutions that canconsciously plan the pattern of landdevelopment within a particular area inorder to internalise various external effects.What is at issue, however, is the existenceof a mechanism that can subject suchattempts at conscious planning to a processof competition in order to generate pricesignals indicating the success or failure ofdifferent planning experiments andproviding actors with an incentive to weighalternatives. In one of his few publishedstatements on land use planning Hayek putthe issue very well, “Most of what is validin the argument for town planning is, ineffect an argument for making the planningunit for some purposes larger than the usualsize of individually owned property. Someof the aims of planning could be achievedby a division of the content of propertyrights in such a way that certain decisionscould rest with the holder of the superiorright... Estate development in which thedeveloper retains some permanent controlover the use of individual plots provides at

least one such alternative to the exercise ofsuch control by political authority” (1960,351-352).

A Proposal for Private PlanningA promising way of moving towards a

new system of private planning that mightoffer some prospect of political successwould be to adopt a variant of a proposalmooted by Moscovitz and O’Toole (2000).Moscovitz and O’Toole argue for the localcommunity ownership of conservationeasements and restrictive covenantsthrough the creation of local recreation andamenity companies. In this proposal,property owners would continue to owntheir acreage on an individual basis, theywould be free to maintain land in its existinguse and they would also have the capacityto bring forward proposals for newdevelopment schemes–as at present.Development rights would, however, beheld collectively by all the other propertyowners encompassed by a recreation andamenity company.

Under this particular model, the statewould divest itself of development rightsthrough the establishment of recreation andamenity companies which would purchaserestrictive covenants from participatingproperty owners in a given geographicalcommunity, paying for these with the issueof shares in the new company. Thecompany board consisting of all propertyowners/shareholders in the area would thenbe responsible for decisions regarding theapproval of new development. In turn allprofits and losses attributable to suchdecisions would be shared out betweenmember property owners, in a mannerproportionate to the scale of their holdings.Individual landowners would no longer holdthe right to the full profits fromdevelopments on their acreage and thestate would no longer hold the right toapprove or reject development applications.Rather, development rights would becomea form of collective private property rightshared by the members of the recreationand amenity company at theneighbourhood/community level under theauspices of a unified management system.

The great virtue of this proposal is that itprovides a way of internalising externalitiesat the local scale by creating a regime ofproprietary ownership. The value of theproprietary communities’ assets would betied directly to the decisions regarding landmanagement made within its jurisdiction.In contrast to the current British planningsystem, where the state holds the right torefuse or grant planning permission, underthe Moscovitz and O’Toole plan, theIf the analysis I have presented thus far


management board of the proprietarycommunity would itself be responsible forsuch decisions. Where under the Britishplanning system all the profits from asuccessful planning application go toindividual landowners proposing newdevelopment—but where planningpermission may be extremely difficult toacquire for such things as new housing—under the proprietary model, allshareholders in the community wouldreceive a share of the profits becausedevelopment rights would be held by therecreation and amenity company.

The above model of proprietarygovernance could help to frame an incentivestructure that would discourage a ‘free forall’ either on behalf of developers or bynimbyist organisations, because propertyowners/shareholders in recreation andamenity companies would be able toconsider decisions in terms of the likelyeffect on community asset prices.Decisions to prevent any development inthe locality would be based on knowledgeof the opportunity cost of such decisions–that is, the foregone financial gains fromallowing new development to proceed.Likewise, decisions to allow inappropriatedevelopment and to lower the quality oflife within the locality would be taken atthe risk of lowering the value of companyasset values. Within this context, recreationand amenity companies would have toexhibit an entrepreneurial sensitivity tomarket forces. A primary implication of thisproposal is that all local services such asroad maintenance, the provision of parks,refuse collection, as well as land useplanning would be the responsibility of the

relevant proprietary organisation.Companies would, therefore, have tochoose the particular bundle of servicesthat they provide and the environmentalcharacteristics that they wish to preservewith regard to the attractiveness of suchdecisions to future residents and hence thelikely effect on asset prices. Under theseconditions, one would expect to witness agood deal of entrepreneurialexperimentation by recreation and amenityboards in an attempt to discover the mostdesirable mix of environmentalcharacteristics necessary to maintain acompetitive edge.

This might not happen overnight.Meanwhile, the whole planning system,insofar as it remains in place should bereformed on economically rational grounds,using the price mechanism to the maximumextent and only intervening when clearexternalities existed. When interventiontook place it should work with the grain ofthe price mechanism and not involvebureaucratic, centralised direction.

ConclusionThe proposals I have just sketched

would present a radical alternative to thecontinuation of government land useplanning and would clearly require someimaginative thinking if they were to bedeveloped successfully. I do not wish toclaim that these proposals will offer apanacea for the ills of the current planningsystem or that the ideas I have advancedwill necessarily represent a politicallyacceptable path to reform. Others who areperhaps a little more attuned to politicalsensibilities than I may be better placed to

address some of these issues. What I amconvinced of, however, is that anysuccessful programme of reform will needto build on at least some of the principles Ihave set out in this article. As the quotationwith which I commenced indicated, the verynature of land use problems means that anysolution will necessarily be imperfect. Isubmit, however, that a system of privateland use planning is likely to beconsiderably less imperfect than thebureaucratic regime in which we are miredtoday.

ReferencesDavies, S. (2002) Laissez Faire Urban

Planning, in Beito, D., Gordon, P.Tabarrok, A. (eds) (2002) The Voluntary

City, University of Michigan Press.Hayek, F.A. (1960 The Constitution of

Liberty, London: RoutledgeMoscovitz, K. O’Toole, R. (2000) New

Incentives for Rural Communities, Portland,Oregon: The Thoreau Institute.

Nelson, R. (2002) Privatizing theNeighbourhood, in Beito et al, op cit.

Pennington, M. (2001) Planning and thePolitical Market: Public Choice and thePolitics of Government Failure, London:Athlone Press/ Continuum International.

Pennington, M. (2002) Liberating theLand: The Case for Private Land UsePlanning, Hobart Paper 143, London:Institute of Economic Affairs.

Mark Pennington is a senior lecturer inpolitical economy at Queen Mary,University of London.

ama gi a - Lancaster UniversityThe logo and name of ama-gi is taken from a cuneiform inscription...

Documents

Transcript of ama gi a - Lancaster UniversityThe logo and name of ama-gi is taken from a cuneiform inscription...