| Sumario: | I have attempted to provide an explanation and approximate analytical generalization of Bergson's (1973) results. I have shown that in estimating welfare losses due to price distortions it is important to work at the most disaggregate level possible and to take full account of cross-price elasticities between goods and services. Failure to do this will in general lead to errors which will often (but not necessarily always) be underestimates of true welfare losses. The results have been illustrated by application to crude data for London Transport's bus and underground services from 1970-75. Welfare losses due to price restraint over the period appeared to be small as a proportion of all expenditure, although they are greater as a proportion of expenditure on the services themselves. Those due to failures in 1976 fares to match optimum marginal social cost fares are much greater, especially if it is reckoned that peak services should be distinguished from off-peak services. Severe under-estimation of welfare losses results from aggregate calculations. The numerical calculations presented here are really only intended as illustrations of the analytical results. As propositions about London Transport's pricing policy they must be treated with reserve. There are many difficulties but I mention two. First, the assumption that the "base" prices are indeed the optimum prices is debatable, especially in the case of the price-restraint argument. Second, some of the price distortions are so large that the approximation involved may not be valid. To be exact one would need to know the forms of the demand functions, as Bergson did by assumption, and to work out explicit forms for the welfare losses. In summary this paper supports the use of the measure proposed by Bergson (1973) for welfare losses due to price distortion, attempts to throw some light on its properties, and supports the proposition that accurate measurement requires a disaggregate treatment. An illustrative...
|
|---|