ABSTRACT

3. These dimensions are discussed in Mrs Baster’s introduction.4. Even for countries at a high level of development in any sense, the use of national income as an indicator is being widely challenged, e.g. by Mishan [1967], on the grounds that the environmental costs are ignored.5. In an interesting paper Divatia and Bhatt [1969] put forward a different index of development potential, based on fundamental factor inputs such as capital and skills (though it is misleadingly described as a measure of the ‘pace of development’). Move­ments in such an index could foreshadow what the future pace of economic growth could be. The index for India, for example, is encouraging because it shows a rate of increase twice as fast as the real national income. But, of course, it does not follow that growth potential will be released, let alone that development will take place.6. This use of the national income had been developed by Colin Clark [1937], In fact the great spurt forward in national income statistics in the 1930s and 1940s was due largely to the unemployment problem, although also to the need to quantify alternative wartime policies.7. In addition, indirect taxes of various kinds on luxuries are relatively heavy, so such biases are particularly severe when market prices are used as weights.8. For example, is a journey to work really an end product, as national estimators assume (especially a journey on a metropolitan underground railway!)? Additional issues are now being posed in industrial countries by the failure of national income to allow for the costs of environmental destruction, i.e. to be a sufficiently ‘net’ concept in that sense.9. Every so often a researcher tries to draw conclusions about trends in per capita food consumption, which of course simply means revealing the implications of assump­tions made by official statisticians.10. There is an upward bias as well. The share of output covered by official statistics, and included in the national income, tends to rise, partly because a growing proportion of output passes through the hands of organized business, which is more adequately covered by official statistics, but also partly because of the general improvement in data collection.11. The U.N. Statistical Office’s ‘A Complementary system of Statistics of the Distribution of Income, Expenditure and Wealth’ is a useful starting-point.12. Various poverty lines in India, where there has been much work on this question, are discussed by Fonseca [1970],13. See papers by Abel-Smith, Bagley, Rein and Townsend in Townsend [1970],14. This problem was first recognized by Rowntree [1901] in his classic enquiry in York, leading him to distinguish between ‘primary’ and ‘secondary’ poverty-the latter referring to the poverty of those who could afford the nutritional minimum but do not in fact attain it.15. See, however, an interesting pair of articles by Minhas [1970] and Bardhan [1970], which show that even using the same criterion of poverty (one proposed in 1962 by a distinguished group of economists to the Planning Mission) very different con­clusions can be reached on trends in the proportions lying below the poverty line through using different sources of consumption data, different allowances for price changes and different interpolation procedures.16. Several indicators can be combined to give us an indicative profile of the pre­valence of poverty in a nation, such as the U.N. Research Institute for Social Develop­ment has been experimenting with in Geneva. In fact they have taken a step further and produced a tentative ‘development indicator’, a weighted average of various series. The Institute’s investigations of multiple associations are interesting and worth while, but we should not fall into the trap (as we could, although the Institute’s Director warns us against it) of treating this indicator as ‘normative’. It simply measures the extent to which a country has advanced along a path indicated by data from countries at different states of progress; see UNRISD [1969],17. See I.L.O. [1970], The point is made there that the measurement of unemployment depends very much on the dimension of the problem that concerns one-unemployment as a cause of personal frustration, low income or loss of output.18. The Pareto coefficient, on the other hand, which long had its advocates, is expressly limited to measuring distribution among higher incomes.

Registered in England & Wales No. 3099067
5 Howick Place | London | SW1P 1WG© 2023 Informa UK Limited