PpT HE modern theory of employment determination implicitly or explicitly assumes a close positive relationship between changes in physical output and changes in the volume of employment. If we construe employment very generally as referring to increased utilization of any resource (labor, capital, raw materials, entrepreneurship), then it is undoubtedly true in this general sense that to produce more requires more employment: it is impossible to produce more physical output without employing more resources of some kind. But modern employment theory specifically refers to labor employment. It is not quite so certain that employment in this more restricted sense is as closely related to physical output changes as employment in the more general sense. Yet most economists have carried over the equation, or rather, close positive correlation of output and employment in its general meaning to output and labor employment. It is usually conceded that the relationship needs correction for productivity changes (generally construed as 2 to 3 per cent annually), but that labor input and physical output are closely and positively related is seldom questioned. The responsibility for this assumed, and apparently common-sense, relationship rests largely with Keynes' who refused to use a physical production index and adopted the device of measuring output in terms of labor input. For example, he states that he will measure changes in current output by reference to the number of hours of labor paid for.2 That Keynes had some doubt about the precision of the outputlabor-employment relationship is revealed by the following: