ABSTRACT

early as 1839, over 55 per cent (R. Williams, The Long Revolutiony London, 1961, p. 166; Indian Census Data).9 T. Balogh, Economic Journal, Dec. 1964, pp. 996-999. Few persons with field experience of underdeveloped agriculture could dissent from the con­clusions of Dr. Balogh’s writings on poor countries; but the Schultzian model is a serious (if perhaps misconceived) ally in his struggle against blind re­liance on market mechanisms, not a frivolous or malicious Chicago dogma'.10 T. W. Schultz, The Economic T est in L a tin Am erica; T T E , ch. 12.11 Nobody has ever found such a curve. The latest of many refutations is E. R. Dean's work on Malawi tobacco farmers (North-Holland, 1966).12 These arguments are not directed at Dr. Hopper’s empirical work. The use of production functions, specified like his, to derive shadow prices will be considered in Section VIII below.13 I am grateful to Professor Walters for showing that this is the proper for­mulation, and to Dr. MacCallum for formalising it and for showing that it is not, in general, equivalent to the more plausible one that I initially chose: equating the expected value of the marginal product of money in each use. What must be equated are the additions to expected value made by the final penny spent on each factor, and in each activity.14 H. H. Mann, R a in fa ll and Famine, Poona, about 1950.15 S. Naqvi, ‘ Coefficient of Variability of Monsoon Rainfall in India and Pakistan’, P akistan Geographical Review , Pt. IV, no. 2, 1949; O. H. K. Spate and A. K. Learmonth, India and P akistan , London, 1967, p. 47.16 These limitations would apply also to much of Central Africa, and with even more force to areas almost too arid for cultivation (e.g. in West Pa­kistan). They would not apply to equatorial rainforest regions such as Indonesia.17 A finding confirmed by Mrs. S. Das Gupta’s Ph. D. thesis (unpublished, but available in London School of Economics Library). She shows by a quadratic programming formulation that, of a number of possible crop combinations ‘efficient’ in the sense that higher expected profit implies higher variance, farmers almost always choose the lowest profit, lowest variance policy. A mixed strategy minimax approximates the policy of Jamaican fishermen in W. Davenport, Jam aican Fishing: a Game Theory Analysis, Yale, 1960.18 M. Kalecki, Theory o f Economic Dynamics, London, 1954, pp. 94-6.19 Dr. Hopper, in an unpublished paper for the Indian Intensive Agricultural Development Programme, showed that top-decile farmers in Indian villages around Delhi usually outperformed research stations. In his analysis, he incidentally showed inter-decile variations of output per acre far larger than the data in, for example, E. Heady, Economics o f A gricultural Poroduction and Resource Use, Prentice-Hall, 1964, ch. 7. Data are far too fragmentary as yet, but the following hypothesis is advanced here: the less developed an agri­cultural community, the greater is the inter-farm coefficient of variability of output in the normal year.20 See note 19 above.21 The whole PPS analogy implies, incorrectly, that each factor input is fixed in total but perfecdy transferable among uses. The analogy is used merely for convenience of exposition, and the arguments are equally relevant where either or both implications fail to hold.22 Either time preference (justifiable by the uncertainty of the future and technological improvements) or a finite planning horizon must be specified. Otherwise, however fast returns-to-scale and utility diminish, the fact that infinity is a long time generates a famous paradox: it always pays to cut consumption to the bare minimum now, so as to grow even faster ‘for ever’; but ‘for ever’ never comes.23 Balogh, loc. c it.24 D. Narain, Im pact o f Price M ovements on Areas under Selected Crops in India ,

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