Comments on the Public-Sector Discount Rate: Response to Jenkins

Abstract

Despite the length of his commentary, Jenkins (1980) fails to answer the criticisms of the Treasury Board's Guide outlined by Sumner (1980). Indeed, many of his apparent answers are addressed to questions which were never raised in the first place. His discussion of risk is a classic example of ignoratio elenchi. His convincing rebuttal of the argument that no allowance for risk need be incorporated in the appraisal of public sector projects is irrelevant. The point at issue was not whether such an allowance should be made, but that it should not be made twice. The Treasury Board's recommendations are therefore inconsistent; if its preferred procedure, of making an explicit allowance for risk, is to be retained, a matter on which, incidentally, an opinion was not expressed, then the discount rate derived from observed returns in the private sector is too high for the purpose. In similar fashion, Jenkins then argues that the marginal rate of return in the private sector is not measured by the return on the set of least profitable projects observed. It would be difficult to disagree with this observation, but it is equally difficult to see its relevance to the distinction that was being made between marginal and average rates of return. That distinction relates to the distribution of expected returns across all potential projects, and not to the distribution of actual returns around their expected yields. Certainly 'there will be a tendency for marginal expected rates of return on new investment and the average expected rates of return to converge'; the substantive question is whether the convergence is so rapid as to render the distinction between average and marginal rates of return unimportant. Notice that if convergence occurred so quickly, Jenkins would not be able to argue that his calculated average understated the marginal rate of return by overrepresenting comparatively unprofitable sectors. More significantly, his implicit judgement that convergence takes place rapidly is inconsistent not only with his own estimates of dispersion in the average private rates of return across industries, but also with the substantial lags in adjustment identified in studies of investment behaviour. Jenkins' remaining arguments reiterate the view that divergencies between private and social returns are attributable solely to taxation. His dismissal of imperfections in the competitive process as a contributory factor leaves as open questions both the adequacy of national indices of concentration as proxies for barriers to entry, and the source of dispersion in private rates of return. The proposition that 'real corporation income tax rates will increase as a result of an inflationary situation' is, yet again, not in dispute. What is in dispute is the assertion that the increase in inflation has more than offset the lower effective tax rates resulting from more generous write-off provisions. The assertion does not apply to the manufacturing and processing incentives. More importantly, its validity or otherwise cannot be established from a simple scrutiny of the historical record, because the benefits of investment credits or more liberal capital cost allowances are typically not extended to existing assets: what matters is the social rate of return on new investment, which depends on current tax provisions. That is why an alternative procedure was recommended, of calculating social rates of return on different assets at the prevailing tax structure, for a given net rate of return. This procedure would eliminate the confusing influence of irrelevant history. By abstracting from other determinants, it would also serve the valuable purpose of throwing into sharper relief the