There are a number of ways in which the consequences of market imperfections for the public sector investment decision can be modelled. In chapter 10 we discussed the approach associated with the name of Harberger. This approach is based on the assumption that market interest rates reflect producer and consumer rates of discount, and examines the consequences of tax distortions which drive a wedge between these two rates. One difficulty with a market-based approach is that the solution to the model requires knowledge of a full set of intertemporal capital markets: the project planner would need to know how interest rates would be affected by the project over its entire lifetime – a formidable information problem. A second difficulty is that, arguably, market rates of interest have no welfare significance for developing countries with highly imperfect capital markets. This view has been very influential in the discussion of investment policy in developing countries (see Marglin, 1963 and UNIDO, 1972). Both practical problems of implementation and doubts about the welfare significance of market-based interest rates have led to the widespread use of a model in which capital market imperfections are modelled by the assumption that savings are a fixed proportion of income. A model of this form was originally developed by Arrow (1966, 1969) to consider public sector investment decisions, and the fixed savings ratio assumption is used by both the major project appraisal manuals of UNIDO and Little and Mirrlees in their discussions of the social value of investment.