Privatisation and Deregulation T N Srinivasan ASHOK RUDRA's questioning (December 21, 1991) of the policy conclusions reached in my recent writings (August 3-10 and September 14, 1991) calls for a response. Before 1 turn to it, let me thank him for his kind remarks about me and assure him that his critical remarks will in no way affect our long personal and professional relations stretching back to the year 1954 when young Rudra joined the faculty of the Indian Statistical Institute (ISI) Calcutta and taught time-scries analysis to students, including me, in ISPs two-year professional training course in statistics. Let me take this oppor tunity to respond also to the comments of my friend and former ISI colleague, Deb Kumar Bose (February 22, 1992) and to A C Minocha (October 26, 1991). There is much with which 1 agree in the comments of my friend and former ISI colleague K S Parikh (February 29, 1992). It should be clear from what follows where we disagree, Rudra disputes that economic analysis has much to say on what activities should be in the public sector. I am surprised at his asser tion. At a purely analytical level, neoclassical welfare economics (either in its Pigouvian or the Arrow-Debreu version) makes a case for public intervention, either on the grounds that a competitive equilibrium can fail to be a Pareto Optimum under certain situations collectively described as 'market failures', or because the distributional outcome of a laissez-faire competitive equilibrium, although Pareto Optimal, is deemed socially unsatisfactory. Textbook examples of market failures include: externalities of various kinds that are not internalised by any private decision maker, the existence of 'public' goods, i e, goods that have the characteristics such as non-rivalry in use (e g, one person's use of a park does not reduce its availability for use by another person, at least until the number of users is not large enough to create congestion) and non-excludability of some users from others (e g, radio broadcasts), and of significant scale economies in the production of particular goods or services, the presence of which precludes a com petitive market structure. Also if certain markets do not exist (e g, markets for insurance against various risks or for various assets), the competitive equilibrium based on markets that do exist need not be a Pareto Optimum. Thus if the market structure is not complete in the sense of Arrow-Debreu, there is scope for government intervention.