Our civilization is properly called an urban civilization, for the forces which shape it gather in the cities and spread out from them. The cities integrate, and in many ways dominate, the entire society. An inquiry into the nature of the city, then, is also an inquiry into much that is important to the nature of civilization itself. The defects of cities are defects of the whole system of living.

There are different types of large cities, but the type which is most significant in this inquiry is the industrial-commercial metropolis. Resort cities, capital cities, shrine cities, and the like may be conspicuous, but they do not have the same kind of influence. It is the metropolis which expresses the essence of our system. All references to cities in this paper are to this type.

The most thorough studies of cities have been made in the United States, but in so far as other observations have been made they indicate that with caution the results may be extended to other large cities of the world. It has been shown that the large expanding city has a natural form which is repeated in all cities, and which is not only unplanned, but for the most part not responsive to efforts at control. There is a basic pattern which no city escapes; a spatial order of population groups and institutions and of functional parts of the economy, and a number of states of social organization and disorganization which are related to this order. This paper is principally concerned with the typical forms of disorganization, and the conditions under which they arise and some of the results of efforts to deal with them.

Early in the 1938 session of Parliament the Canadian Senate appointed a Special Committee to “enquire into and report upon the best means of relieving the country from its extremely serious railway condition and financial burden consequent thereto.” After hearing extensive evidence from the officials of both railways and of the labour unions, the Committee stated that through lack of time it was unable to analyse this material in a satisfactory manner and therefore recommended that it be reappointed early in the 1939 session. It is the purpose of this article briefly to review the evidence presented and in particular to suggest lines along which further investigation by the Committee might profitably be undertaken.

The Duff Commission believed that the policy of co-operation which it espoused would lead to annual savings of at least $35,000,000, though it developed that it had never secured, or been given, detailed information on which to base this estimate (1374). The actual joint net economies after five years' trial were only $1,772,000 per annum (121).

While this inquiry arose out of the temporary closing of a mill at Sherbrooke, attributed by the management to the fear of Japanese competition, the Dominion Government called for “an investigation much more searching and extensive than is possible at the present time either by officials of any of the Government departments or by the Tariff Board.” The Honourable W. F. A. Turgeon, a judge of the Court of Appeal of Saskatchewan, was appointed a Royal Commissioner to make this investigation as speedily as possible. The completion of the report was delayed by the Government's decision to requisition the services of the same commissioner for a further study of the wheat problem, which occupied several months and necessitated investigations in England, but after two years' work and the collection of nineteen thousand pages of evidence, the report was laid before the Minister of Finance on January 20, 1938. It is an important document, which will undoubtedly be quoted, especially as a source of factual data, in discussions of the textile industry for years to come.

Conforming with the terms of reference, the Commissioner sought information with respect to tariff protection, volume of production, investment, costs, profits, wages, salaries, and bonuses, and such other matters as might help the Government to arrive at conclusions “regarding the position of this and other branches of the textile industry in relation to British and foreign competition, and in particular, the extent to which the employer can reasonably and properly be expected to maintain employment over periods of temporary difficulty.”

This paper is an attempt to suggest methods of approach to the materials of Canadian social history. As such it is an appeal for a greater utilization of historical data in the development of sociological principles. It should be admitted at the outset that the point of view advanced is partly the result of a bias in favour of what might be described as the documentary as opposed to the interview method of research. Certain problems can be studied only by means of the interview. There are times, however, when a greater understanding could be gained from documentary records with much less expenditure of time and energy. Furthermore, the interview method, if it is to yield fruitful results, must be employed by those with a particular aptitude for “feeling out” information in this way. It is no accident that the most successful research students in sociology very often have been those with experience in journalism or those with a flare for journalism who found an outlet for their inclinations in sociological research. The penetrating insight of a Sinclair Lewis or Sherwood Anderson cannot be matched by the finest tools forged in the laboratory of the sociologist.

If for any national income and its given distribution we could estimate the amount of annual savings, we should be able to tell how much investment is necessary to make the aggregate receipts of entrepreneurs (plus the value of any additions to the inventories) equal to their outlay and thus avoid entrepreneurs' losses and shrinkage of output, and keep national income at a given level.

The relationship between national income and savings has been estimated in three ways: (1) by studying the relationship between incomes and savings of families, leaving the undistributed corporation incomes out of the picture; (2) by studying separately savings of wage-earners and of profit-receivers out of their respective incomes (distributed or not); (3) by confronting, through time, national income and investment. Only the first method will be discussed here.

If people adjusted their behaviour to changes in their income more quickly than they actually do, the family budget statistics could give an answer as to how a change in the total national income affects the national savings, apart from the undistributed profits of corporations. Actually, there is probably a time lag between income changes and changes in consumption and savings, which should be studied separately.

When people receive their money incomes they may employ them in any of three ways: (1) they may spend them on imported products, (2) they may spend them on services or goods produced locally, (3) they may save, accumulating bank deposits or securities. The way in which people in general and exporters in particular dispose of additions to their incomes, and the manner in which they economize when their incomes diminish, are important factors in determining the ultimate effect of an initial change in the level of incomes. If most of an increase of incomes is spent upon the purchase of goods produced within the country, it will induce re-employment and a further increase in incomes; if spent upon imports, its effect will be upon the foreign exchange situation rather than employment.

The Marginal Propensity to Import. The extent to which a community devotes any addition to its existing incomes to the purchase of goods from abroad has been called by Mr. Paish the “marginal propensity to import.” The high per capita trade of the Dominions ensures a fairly high propensity to import. But another cause contributes to the high propensity. Their imports are, in considerable degree, high grade manufactures and luxury articles: just the sort of things which will be more in demand if incomes rise and will be dispensed with if incomes fall. Moreover, machinery and capital goods rank high amongst their imports, and these will be in demand chiefly when incomes are rising.

Some attention must first be given to the nature of the determination of the general levels of incomes, production, prices, and employment. Of course these levels are in some way or other affected by most things under the sun and are, in fact, merely statistical aggregates composed of millions of individual activities and transactions each determined by its circumstances, and all interdependent. Even if one concedes that some grouping is necessary to bring this chaos within the realms of description and analysis, it may still be urged, particularly in Canada and the United States, that regions and industries must be dealt with separately, that we cannot lump together production of wheat, newsprint, and houses and get any results of significance. While recognizing the validity of this contention, it has seemed necessary to use a number of familiar aggregative concepts in this paper.