An interesting analysis of productivity in Canada from 1931 to 1949 has recently been published by Mr. A. Maddison in the Economic Journal. Mr. Maddison admits, however, that his choice of dates, which was limited by the lack of official estimates of the employed labour force prior to 1931, is unfortunate in that a strong cyclical influence colours the results. To overcome this difficulty calculations have been made of the employed labour force prior to 1931 (see Table III), thus extending the period that can be covered back to 1926. In addition this paper makes use of revised and more complete data published by the Dominion Bureau of Statistics. No attempt is made to duplicate Maddison's excellent summary of the diverse factors contributing to productivity, or his brief look at inter-industry shifts, except for the very important shift of labour from agricultural to non-agricultural employment. But more emphasis is placed on aggregate movements during particular periods and on the comparison of productivity in Canada with that in the United States. In brief, the two papers are complementary, and should be so considered.
Productivity is, of course, an elusive concept when applied to operations of more than one industry, or even to more than one line of production. Is an automobile, for example, really worth 1,000 bushels of wheat in terms of economic value, and is the productivity of a farmer low because he can produce only 2,600 bushels of wheat in a year on 100 acres, whereas the average automobile worker turns out perhaps 10 cars a year, worth 10,000 bushels of wheat or more? Yet the only way in which products can be compared is through the exchange values attached to them in the market. Any attempt to evaluate productivity on a national scale must therefore involve the weighting of goods produced according to the prices at which they were sold. The most that can be done to place production figures on a real basis is to attempt to eliminate variation in prices; and thus determine the extent to which the total volume of production of all goods in the economy, weighted according to their price relationships at some specific point of time, has grown over a period of years. In the discussion that follows, therefore, the gross national product in 1935–9 dollars is used as output. It would perhaps be better to use national income, but this is difficult to deflate in view of the inclusion of indirect taxes in price indices.