AND WORKERS ' PREFERENCES

The first set of hours of work estimates constructed for Canada and its regions for the 1880-1930 period is presented in this article. These estimates suggest a trend decline in hours of work, especially following First World War. In addition, these estimates suggest that the decline in hours of work came at no or little cost in terms of real weekly income. The trends uncovered for Canada are found to be similar to those revealed for the U.S. In effect, by the early twentieth century workers were realizing their long expressed preferences for a shorter workweek at no loss in real income. There has been a resurgence in the discussion and analysis of the length of the workday in America at the turn of the century (Atack and Bateman, 1992; Shells, 1990; Whapples, 1990). However, the Canadian case is both little known and little discussed, and for good reason. Not much is known about the evolution or extent of hours worked per week or day in Canada from Confederation (1 867) to 1930. This is in sharp contrast to the U.S. where careful estimates for hours worked were constructed long ago by Paul Douglas (1930) for the entire 1890-1926 time frame. His estimates are for manufacturing, construction, transpor-tation, mining, unskilled workers, and government employees. More recently his estimates have been extended back to 1880 (Atack and Bateman, 1992). In Canada, reference is typically made to census year weekly hours of work estimates quoted in Sylvia Ostry's and Mahmood Zaidi's textbook, Labour Econ-omics in Canada (1975, pp. 60-61). However, this provides estimates for the manufacturing sector alone and for only five years of the entire 18701930 period