Knowledge that Transforms

ONE of the most troublesome statistical I shortcomings at the outset of World War II was the lack of pertinent data on the behavior of the labor force under conditions of exaggerated demand. The exertions of the last war had been of much lesser magnitude; moreover, they had been inadequately and imperfectly recorded. There was little to serve as a guide in appraising manpower resources or in determining the most fruitful and feasible channeling of pressure and persuasion. The statistical record of this war will be much more complete. Not only are we better supplied with current information on the development of the nation's wartime labor force, but we also have become more fully aware of the necessity of attacking our manpower problems on an area basis. The geographical distribution of the stresses and strains accompanying the war effort has been far from uniform. An appreciation of the extent to which the population has met the increased demands, through mobility of workers and through the recruiting of persons who normally would not work, and of the practical limits of the expansibility of the labor force may be obtained in part by an analysis of some of the boom areas in which war production has been concentrated. In the spring of I944 the Bureau of the Census, in cooperation with the President's Committee for Congested Production Areas, conducted a census of ten boom areas, to secure data about the wartime population.' This census yielded information on the size of the labor force, the relative contributions of the various population groups to each area's labor supply, the part played by migrant workers, the length of the work-week, and other significant items. When compared with the benchmarks provided by the Decennial Census of Population four years previously, just before the beginning of the defense boom, these data furnish a graphic description of the dynamics of labor supply in wartime.2 Each of the designated areas was so defined as to include not only the center of industrial activity but also the surrounding territory in which many of the workers and their families live and in which many large new war plants have been built. Each area comprises a county or a group of contiguous counties, embracing a critical war production center. The areas and the counties which they comprise are as follows:

TWO surveys of the operating expenses of Canadian manufacturers have been made in the last twenty-five years. The fLrst covers all establishments reporting to the Dominion Bureau of Statistics for the Census of Manufactures in each of the five years I9I 7 to 192 I; the second, undertaken for the Royal Commission on Dominion-Provincial Relations in I938, covers in greater detail a limited but fairly representative group of manufacturers for each of the years I929, I933, and I936. The present paper describes the origin, methods, and results of the second or I938 survey, and provides a summary of the I9I7-2I study. In addition, it attempts to arrange the results of both surveys in comparable form.' Because of the difficulty of securing reliable figures by the method employed in the I938 survey, special attention is given to the representativeness of the results. Certain details of importance for the text will be found in appendices. Among the tables presented, Table 9 is of most general interest, while certain applications of the results will be found in Tables io and i i.

A FEW years ago Professor Alvin H. Hansen IAJL developed a model sequence in which the multiplier analysis and the acceleration principle were combined. This was done by making increments to the national 1 consist of three parts: (I) governmental deficit spending in which each of the periodic expenditures by government was assumed to be of equal size; (2) induced private consumption expenditure the size of each of the successive induced consumption expenditures being determined by applying the marginal propensity to consume to the preceding period's increment to income; (3) induced private investment the size of each of the successive expenditures being determined by applying the coefficient of magnification (the acceleration principle, or relation) to the difference (positive or negative) between the induced consumption expenditure of any given period and that of the preceding period. As is well known, without the inclusion of the third component, the curve of national rises at a decreasing rate and approaches, as a limiting increment, the product of the multiplier and the constant increment to governmental contributions.2 Inclusion of a given coefficient of magnification, in conjunction with the first two factors, yields (for most relevant values of the multiplier and the coefficient of magnification) a curve of national which fluctuates about this limiting increment 3 as Professor Paul A. Samuelson has shown.4 Samuelson's generalization of Professor Hansen's model sequence covers a wide range of values for the marginal propensity to consume and the coefficient of magnification. This generalization proceeds on two assumptions: (i) increments to governmental deficit spending are of equal value for each successive period of expenditure; (2) in any given case, the marginal propensity to consume (a) and the coefficient of magnification (03) do not change in value, over time. Under these assumptions, Samuelson shows that the successive increments to national can be written: Yt = It + aYt-i + a,q (Y,-lY t-2), where Yt is the increment to national at time t, and where It is the governmental contribution at time t.5 This, of course, follows because the second and third terms are the equivalents, respectively, of induced consumption and induced investment. Samuelson's formula is sufficiently general to be applicable to cases in which autonomous investment (I t) is permitted to fluctuate over time, but his analysis is confined to the case in which It is constant; and, to my knowledge, no attention has ever been given to the interactions of a constant multiplier, a constant coefficient of magnification, and autonomous investment expenditures which vary in size over time; nor has the equally important case of cyclically fluctuating multipliers and magnification coefficients -a case which I shall treat in a later article received any attention. Such investi1 In studies of this sort it is, I believe, usually implicit that an increment to national income is defined as national at time t minus national at time 0, where national at time 0 is either zero or comprised entirely of Hansen's basic-income-level consumption (see pp. 289-92 of his Fiscal Policy and Business Cycles, New York, I94), and where national for each of the two periods preceding the initial period (time O) is equal to that of time 0; in short, an increment to national income has been implicitly defined, in economic literature, in such a manner that the national of time 0 does not contain any element of autonomous investment, induced investment, or induced consumption. The terminology and the charts of this paper are based upon this definition for an increment to national income, 'where incorme at time 0 is comprised entirely of basicincome-level consumption. However, as long as the national for each of the two periods preceding the initial period (time 0) is equal to that of time 0, we should get the same general results where the national of time 0 included basic-income-level consumption, autonomous investment, and induced consumption; for, although total autonomous investment and total induced consumption would each be greater, in any given period, in this latter case than in the case where these two components were each equal to zero in the initial period, the in1cremenzts to autonomous investment and the increments to induced consumption would, in any given period, be identical in the two cases. 2 Professor Hansen has a graphic presentation of this point: Op. Cit., p. 272. 3 This, too, is graphically presented by Hansen: ibid., p. 277. 4 Interactions Between the Multiplier Analysis and the Principle of Acceleration, this REVIEW, XXJ (I939), PP. 75-78. 5 Being concerned with governmental deficit spending, rather than with autonomous investment, Samuelson's symbol for It is gt. It is used here to give the term a more general mnemonic connotation.

and absolute freedom to choose an occupation and job are allowed. We have never had that fine state of affairs. The crucial issue is, it seems to me, what are the minimum restraints on liberty and freedom of movement that are consistent with full employment? Possibly incentives to move offered by the government (e.g., payment of transportation to workers and tax concessions to capital), and a smoothly functioning employment exchange which gives precise knowledge as to opportunities, may be all that is required. In that case, we may still retain our essential liberties and have full employment. That brings us to the failure of the bill to define full employment precisely and, therefore, its failure to deal with certain aspects of the problem. For example, we should like to know what are the appropriate number of hours of work per week? Are 6o million jobs to be provided at 3o hours (a goal sought by many), or at the prewar level of 37-38 hours, or at 45 hours (present level) which may be assumed to be the choice of the masses of the workers, or some other number which will yield maximum output per man-hour or maximum total output? And if hours are (say) 30, what will the effect be on productivity and wage rates and, therefore, on jobs? Should the government, again, seek to find employment for all those who seek work, or should the government, through proper educational policies, social security programs, and the like, influence the numbers seeking work and thus perhaps solve the unemployment problem? Finally, let us not forget that economists themselves have not given Senators Murray or Wagner clear-cut answers regarding the appropriate policies which will yield the maximum number of jobs. High wages or low wages? A 40-hour week or a 30-hour week? Fewer taxes on investment or on consumption? Higher or lower tariffs? A reduction of taxes or a rise of spending? Less or more public investment?

NFLATION in China is a subject of considerable interest and serious concern to all Allied Nations, inasmuch as it might so affect the Chinese war economy as to delay considerably a large-scale counteroffensive from this vital base, thereby prolonging the war. Lack of reliable information has given rise to much speculation regarding the subject, such speculation being based largely on the superficial evidence of price movements. The conclusions thus drawn naturally have involved considerable misunderstanding of the Chinese economy. This situation is not conducive to full cooperation between China and its allies. It is essential, therefore, that an objective and factual picture of inflation within Free China should be presented.