CH. 18 THE GENERAL THEORY RE-STATED 251

(iii) When there is a change in employment, money-wages tend to change in the same direction as, but notin great disproportion to, the change in employment;i.e. moderate changes in employment are not associatedwith very great changes in money-wages. This is acondition of the stability of prices rather than ofemployment.

(iv) We may add a fourth condition, which pro-vides not so much for the stability of the system asfor the tendency of a fluctuation in one direction toreverse itself in due course; namely, that a rate of in-vestment, higher (or lower) than prevailed formerly,begins to react unfavourably (or favourably) on themarginal efficiency of capital if it is continued for aperiod which, measured in years, is not very large.

(i) Our first condition of stability, namely, that themultiplier, whilst greater than unity, is not very great,is highly plausible as a psychological characteristic ofhuman nature. As real income increases, both thepressure of present needs diminishes and the marginover the established standard of life is increased; andas real income diminishes the opposite is true. Thus itis natural—at any rate on the average of the community—that current consumption should be expanded whenemployment increases, but by less than the full in-crement of real income; and that it should be diminishedwhen employment diminishes, but by less than the fulldecrement of real income. Moreover, what is true ofthe average of individuals is likely to be also true ofgovernments, especially in an age when a progressiveincrease of unemployment will usually force the Stateto provide relief out of borrowed funds.

But whether or not this psychological law strikesthe reader as plausible a priori , it is certain that experi-ence would be extremely different from what it is if thelaw did not hold. For in that case an increase of in-vestment, however small, would set moving a cumulativeincrease of effective demand until a position of full