Published online by Cambridge University Press: 10 November 2009
The principle of redistribution in all the provinces is not one man, one vote. But how far from equal population or equal number of voters for all constituencies do the provinces deviate?
If redistribution were based only on the population in the provinces, one would need a measure which would indicate the degree of inequality. There are several measures available. Hayward R. Alker, Jr., and Bruce N. Russett, after comparing different measures of inequality, concluded that the Gini index was an attractive way of summarizing the important parts of an accurate cumulative distribution and that it was sensitive to inequalities present in all parts of a value distribution. Because of these findings, it was decided that the Gini index would be used as a measure of inequality for provincial representation.
1 For details of the redistribution procedures in the provinces, see Qualter, Terence, The Election Process in Canada (Toronto, 1970), 105–11Google Scholar; Doucet, Philip, “Representation in the Nova Scotian House of Assembly (1867–1965),” unpublished MA thesis, Dalhousie University, 1965Google Scholar; Gordon Macniven, Hugh, “The Legislative Assemblies of the Canadian Provinces,” unpublished PHD thesis, University of Minnesota, 1960Google Scholar; Eager, Evelyn L., “The Government of Saskatchewan,” unpublished PHD thesis, University of Toronto, 1957, pp. 142–57Google Scholar; Copes, Parzival, “The Fishermen's Vote in Newfoundland,” this Journal, III, no 4 (Dec. 1970), 589–93Google Scholar; and John Anthony Long, “Maldistribution in Western Provincial Legislatures: The Case of Alberta,” ibid., ii, no 3 (Sept. 1969), 345–55.
2 On Measuring Inequality,” Behavioral Science, ix (1964), 217.
3 As space is limited in a research note the graphs have been eliminated, but an expanded version is available from the author.
4 This method was suggested by Professor Duff Spafford; calculations were carried out using a computer program written by Robert Lucas.
5 See Qualter, The Election Process, 28.
6 Ibid., 107. In fact, Alberta does not have available the total number of individuals in the provincial constituencies. See Long, “Mal-distribution,” 349, n 16.
7 Qualter, The Election Process, 102.
8 See ibid., Table 1:5, p. 29, for a comparison of the percentages of the total voting-age populations who are registered on official voters' lists.
9 New Brunswick the Report of the Chief Electoral Officer 1967 (n.p., n.d.), 54, 55, 61, and 62. For the Saskatchewan elections of 1934, 1938, 1944, and 1948 the relevant data for a few districts had to be estimated from the number voting and from the number of ballot papers sent out. For the 1911 election in Ontario five constituencies were not included because of incomplete data, and for the 1970 election in Nova Scotia the preliminary report was used.
10 Quaker, The Election Process, 106.
11 See Québec, , Journal des Débats, x, no 47, Dec. 19, 1970Google Scholar, Robert Bourassa, p. 2622; and the Globe and Mail, Toronto, Dec. 21, 1970, and March 9,1972.
12 I am indebted to Professor Duff Spafford for permitting me to use his data for the federal redistribution.
13 Qualter, The Election Process, 86 and 107.
14 It is likely that in the future Saskatchewan will have the lowest Gini index, as it has passed legislation providing for an independent boundaries commission and for a tolerance level of 15 per cent, which is the lowest in Canada. See Saskatoon Star-Phoenix, April 4, 1972, and April 19,1972.