<p>Problems and Prospects for Clusters in Theory and Practice</p><p>&nbsp;</p><p>&nbsp;Philip Cooke, Centre for Advanced Studies, Cardiff University</p><p>&nbsp;</p><p>Introduction</p><p>We have seen that since Marshall, observers have spoken of certain agglomerations of specialised economic activity as, first ‘industrial districts’ and more recently ‘clusters.’ When they do this they are mainly speaking of a concrete phenomenon rather than an artifice. Usually they mean many firms in more or less one place, the firms competing but having at least preferential relations with some. These relations may range from ‘preferred supplier’ contractual status at one end of the spectrum of economic interaction, to high-trust, non-contractual collaboration, exchanging of favours and non-monetary transactions at the other. So a question that is of undoubted and abiding interest to economic geographers and others is why, in a world largely characterised by utilitarian, unsentimental buying and selling, do clusters exist? They ought to have disappeared with the medieval guilds, the source that Marshall, in effect, gave as the ultimate origin of industrial districts. Certainly, they ought to have been replaced by the modern business corporation with its advantages in resource and administrative capabilities that have exercised generations of industrial economists from Young (1928) to Coase (1937), Penrose (1959), Richardson (1972) and the modern dynamic capabilities authors like Teece &amp; Pisano (1994), Winter (2000) and Zollo &amp; Winter (2002).</p><p>......</p><br/>