POVERTY AND FAMILY SIZE 35 24 Data relating to households cross-classified by both size of landholdings and per capita expenditure show the existence of "poor'* among the landed as well as "non-poor^ among those who do not have much land. See, Tables with Motes on Some Aspects of Agriculture in India, Eleventh Round (1956-57), Rural, NSS Report No 140. 25 The mechanics of over-adjustment is discussed in S K Rao, "Reducing Growth Rate of Population through Decline in Mortality, A Note", Economic and Political Weekly, Vo\ IX, 1974, pp 1623-1628. Rao does not, however, hold that such over-adjustment need be the rule for all agrarian classes. In respect of workers, for example, he says (in his 1976 article referred to in footnote 9), "Moreover, the cultural practices tend to be such that children, as soon as they grow into adulthood, set up separate families denying any income to their parents, due to pressure of povery. All in all,it is fair to say that expected addition to family income associated with children in this class would be low." He concludes that the motives for producing more children are stronger among the rich than among the poor. 26 Mamdani is unaware of this. He assumes that the poor have large families and large numbers of children. The theory he provides for explaining high fertility among the poor is colourfully illustrated by quotations attributed to Punjabi farmers. Two examples: "A forest is not made of one tree AJat (farmer) is not made of one son*', and "A rich man has his machines, I have my children. It's that simple'*. The references are to his 1976 article referred to in footnote 2. 27 The arithmetical basis for the relationship between poverty and family size is the point of departure for the study: M V Nadkarni, "Overpopulation and the Rural Poor", Economic and Political Weekly, Vol XI, 1976, pp 1163-1172. Nadkarni dismisses the hypothesis that poor households are burdened by excessive population fin terms of labour utilization^). It is interesting (from our point of view) that he uses. size of landholding as the classificatory variable in his analysis.