C.P. CHANDRASEKHAR* External Vulnerability and Industrial Policy in the Era of Globalization In recent years, two tendencies have come to dominate discussions on the theory and practice of development through industrialization. The first is a successful challenge to the view which informed industrial policy in the 1950s and 1960s, that late industrializers must protect domestic economic space as part of a strategy of creating a viable indigenous industrial sector. The second is that in deciding the frontier which can be drawn 'between the province of centralized and decentralized decisions', or the extent to which the state must interfere in the functioning of actually existing markets, it is seen as best to give all the benefit of doubt about the expedient extent of either to the latter. Both of these tendencies are, it is argued, the understandable consequences of the process of 'globalization9. Underlying this shift in perspective is the presumption that the premises that underlay earlier thfnking on industrial policy are no more valid. Development policy during the post-Depression years of decolonization emphasised the constraints that international and domestic inequality set on the process of industrialization. It believed that, while the international technology shelf offered an opportunity to late entrants seeking to 'catch-up* with their developed counterparts, excessive integration with the world economy significantly disadvantaged late industrializers. The disadvantage stemmed from the very nature of late industrialisation, which by definition involves substantial sunk costs for entry into factory production. These costs not only involved those associated with the acquisition of technology and capital equipment, which were most often monopolised by firms which are at the leading edge of technology in specific industries, but also the investment in that most primitive of research required to identify a technology, negotiate its transfer and then unpackage it for successful commercial production. With so much that is 'implicit' in technology, mere access to blueprints is inadequate. Implicitness also increased the 'learning by doing' benefits associated with modern technology. This implied that the new entrant takes time to produce at costs competitive with incumbent producers operating even with technol- '*'Centre of Economic Studies and Planning, Jawaharlal Nehru University, New Delhi.