330 THE GENERAL THEORY OF EMPLOYMENT bk. vi

made to the quantity carried over into later years. Theproceeds of this addition are added to the current in-comes of the farmers and are treated by them asincome; whereas the increased carry-over involves nodrain on the income-expenditure of other sections ofthe community but is financed out of savings. That isto say, the addition to the carry-over is an addition tocurrent investment. This conclusion is not invali-dated even if prices fall sharply. Similarly whenthere is a poor harvest, the carry-over is drawn uponfor current consumption, so that a corresponding partof the income-expenditure of the consumers creates nocurrent income for the farmers. That is to say, whatis taken from the carry-over involves a correspondingreduction in current investment. Thus, if investmentin other directions is taken to be constant, the differencein aggregate investment between a year in which thereis a substantial addition to the carry-over and a yearin which there is a substantial subtraction from it maybe large; and in a community where agriculture is thepredominant industry it will be overwhelmingly largecompared with any other usual cause of investmentfluctuations. Thus it is natural that we should findthe upward turning-point to be marked by bountifulharvests and the downward turning-point by deficientharvests. The further theory, that there are physicalcauses for a regular cycle of good and bad harvests, is,of course, a different matter with which we are notconcerned here.

More recently, the theory has been advanced thatit is bad harvests, not good harvests, which are good fortrade, either because bad harvests make the populationready to work for a smaller real reward or because theresulting redistribution of purchasing-power is held tobe favourable to consumption. Needless to say, it isnot these theories which I have in mind in the abovedescription of harvest phenomena as an explanation ofthe trade cycle.