OntheOptimalityoftheCompetitiveProcess: Kimura'sTheoremandMarketDynamics J. STANLEY METCALFE ESRC Centre for Research on Innovation and Competition and School of Economic Studies, University of Manchester, UK (stan.metcalfe@man.ac.uk) Synopsis: Thispaperconsiderstheoptimalitypropertiesofamarketeconomyintermsofthreepropositions thatevaluatetheoutcomesofandtheprocessofcompetitionbetweenapopulationof®rmsworkingwithina given economic environment. We show that when ®rms differ in more than one competitive characteristic thencompetitiondoesnotselectingeneralthemostef®cient®rmnordoesitalwaysresultinincreasesinthe averageef®ciencywithwhichresourcesareutilized.DrawinguponatheoremofKimura,however,weshow thatcompetitionhasthepropertyofmaximizingtherateofchangeoftheaverageselectivecharacteristicsin the population. We conclude that a more nuanced appraisal of the institutions of the competitive process is surelynecessary.Fromanevolutionarystandpoint,theoutcomesofcompetitionarealwayscontingentonthe nature of the selection environment and the characteristics of the whole population of ®rms that are being selected. Keywords: competition, evolution, ef®ciency, optimality, growth JELclassi®cation: O33, L1 Introduction The question of the optimality properties of an evolutionary process, if any exist, has longbeenamatterofkeeninteresttoevolutionarybiologists.Similarly,thequestionof the possible optimality properties of the market process has occupied the thinking of economistsatleastsincethetimeofAdamSmith.Indeed,theconnectionhemadeinthe Wealth of Nations between self-interest and benevolent market outcomes has been a central theme in economics ever since. The purpose of this paper is to bring together thesetwostrandsofthoughtbyaskingthequestion:`Whatdoeseconomiccompetition optimize?'. This is not a trivial issue. Any answer informs the appraisal of the institu- tionsofcapitalismandhasgreatsigni®cancefortheconductofpublicpolicyinregard of such matters as the ef®cacy of competition, the bene®ts from free trade, and wider processes of growth and development. However, modern economics has generally approached this question from a decid- edly non-evolutionary direction, that of the theory of market co-ordination in which the allocation of resources in equilibrium is said to be Pareto ef®cient. This body of analysis rests on two general propositions: that all agents are rational, and that all agents produce, exchange and consume within a framework of perfectly competitive marketsinwhichnoagenthasanyin¯uenceonthepricesthatco-ordinateplanstobuy Journal of Bioeconomics 4: 109±133, 2002 # 2002 Kluwer Academic Publishers. Printed in the Netherlands.