¹ Pickersgill, J. W., ed., The Mackenzie King Record (Toronto, 1960), I, 569–70.Google Scholar

² Part of the Liberal success lay in Mowat's careful attention to problems of redistribution. He was easily able, as Sir John Willison put it, to “redistribute constituencies with Christian humility and partisan ingenuity.” See Willison, Sir John, Reminiscences: Political and Personal (Toronto, 1919), 96Google Scholar. A careful analysis of the results in each riding, however, shows that the Conservatives invariably won a majority in the towns whose electoral strength was lost in the heart of dominantly rural, and Liberal, constituencies.

³ For an account of the development of the Ontario party system see our “Ontario: A Three-Party System in Transition,” in Martin Robin, ed., Provincial Party Politics, forthcoming from Prentice-Hall of Canada.

⁴ In nine federal elections since 1945 the Liberals have been the leading party in Ontario on six occasions. For the federal electoral history of the province since Confederation see Beck, J. Murray, Pendulum of Power: Canada's Federal Elections (Scarborough, Ont., 1968)Google Scholar. For provincial results in the same period see our “Ontario: A Three-Party System in Transition.”

⁵ See, for example, Quo, F. Quei, “Split Ticket Voting in Alberta: ‘Amphibious Voting’ and Its Implications,” paper presented to the annual meeting of the Canadian Political Science Association, Calgary, 1968Google Scholar; and Courtney, John C. and Smith, David E., “Voting in a Provincial General Election and a Federal By-Election: A Constituency Study of Saskatoon City,” Canadian Journal of Economics and Political Science, XXXII, no. 3 (Aug. 1966), 338–53.CrossRefGoogle Scholar

⁶ Lipset, Seymour Martin, “Democracy in Alberta: Part I,” Canadian Forum, Nov. 1954, 176.Google Scholar

⁷ See Dawson, R. Macgregor, The Government of Canada, rev. Ward, Norman (4th ed., Toronto, 1963), 528Google Scholar. See also the examples given by Professor Dawson himself in the first edition of his standard work on Canadian government and politics at pp. 581–2. The development of this notion generally ignored the performance of the governing federal party within particular provinces.

⁸ Modern Democracies (London, 1921), I, 475.

⁹ Quoted in Glazebrook, G. P. de T., A History of Canadian Political Thought (Toronto, 1966), 191Google Scholar. A political commentator of the last century expressed much the same view of the federal and provincial elections of 1878 and 1879: “It was Reform votes that gave Ontario to Sir John A. Macdonald, last September, by a majority of 66 against 22. Protection being secured, Ontario Reformers came back to their party allegiance and sustained Mr. Mowat by 58 to 30.” “The Ontario Elections,” Canadian Monthly and National Review, Sept. 1879, 228.

¹⁰ “Canadian Liberal Democracy in 1955,” in Press and Party in Canada: Issues of Freedom (Toronto, 1955), 39–40.

¹¹ “Twenty Years as Prime Minister,” Canadian Forum, July 1946, 78.

¹² See Muller, Steven, “Federalism and the Party System in Canada,” in Wildavsky, Aaron, ed., American Federalism in Perspective (Boston, 1967), 155–61.Google Scholar

¹³ Howard Scarrow has challenged the idea of balance through an examination of a limited set of post-war Ontario data. He argues that the increase in Conservative strength in provincial elections is insufficient to sustain the claim that a significant number of electors deliberately seeks to counter the power of the federal government. See “Federal-Provincial Voting Patterns in Canada,” Canadian Journal of Economics and Political Science, XXVI, no. 2 (May 1960), 295.

¹⁴ See, for example, the data in Scarrow, Howard A., “Patterns of Voter Turnout in Canada,” Midwest Journal of Political Science, 5, no. 4 (Nov. 1961), 362.CrossRefGoogle Scholar

¹⁵ The data in Table I give federal and provincial election results in Ontario since 1919 with the number of votes cast for each party shown as a percentage of the total registered electorate in the province. The number not voting in any election thus appears as another party under the heading Abstentions. Before 1930 these data are not necessarily accurate because of uncertainties in the enumeration of voters and because of the fairly substantial number of acclamations in provincial elections. However, since we have not counted the registered electorate in cases where no poll was held the data we present represent the best estimate available of the effect which the level of turnout has on the pattern of voting.

¹⁶ The percentage of abstentions is, of course, the mirror image of the level of turnout.

¹⁷ The data in Tables ii and iii have been assembled by reaggregating results in individual polling subdivisions in the federal general elections of 1963 and 1968 and the Ontario general election of 1963 on the basis of the new constituency boundaries established for the 1967 provincial general election. The cases illustrated are therefore entirely based on the provincial constituency boundaries of the ridings named.

¹⁸ The federal equivalent on the 1966 distribution of seats is very roughly the riding of Hamilton East. For an account of the organization the federal Liberal member is able to muster, see the Hamilton, Spectator, Nov. 9, 1965.Google Scholar

¹⁹ The pattern is repeated, with varying intensity, in every constituency outside of the three regions of Western Ontario. We have chosen to summarize these data in Tables iv and V simply because the very much larger table which would be required to make the point more explicitly would occupy an undue amount of space, as would the inclusion of the data for the 1965 to 1967 set. The voting data for each of the regions are available from the authors.

²⁰ These data, organized on the basis of the 1967 provincial constituencies, are available from the authors for the federal elections of 1963 and 1968 and for the Ontario election of 1963.

²¹ There are altogether seventeen provincial ridings where the French-Canadian vote may be regarded as decisive – that is to say, where 25 per cent or more of the population is of French origin. In 1965 – an election unencumbered by a federal Liberal leader of French origin – the Liberals would have won fourteen of these and the ndp three. In the provincial election of 1967 the Liberals in fact only won five, the Conservatives eight, and the ndp four.

²² This has evidently been the case for a number of years. See, for example, the aggregate data assembled for the 1948 provincial and 1949 federal elections in Opferkuch, Paul R., “Southern Ontario Voting Patterns, 1945–1959,” unpublished ma dissertation, University of Wisconsin, 1963, 126–8.Google Scholar

²³ For a more extensive discussion of these problems see Wilson, John, “The Use of Aggregate Data in the Analysis of Canadian Electoral Behaviour,” paper presented to the Conference on Statistics of the Canadian Political Science Association, Ottawa, 1967.Google Scholar

²⁴ The data in Table vi have been extracted from material based on specific Ontario town-ships and urban census tracts which have the principal characteristic of being significantly French in ethnic origin. For this illustration we have been compelled to substitute the 1965 to 1967 set for 1963 because the extraordinarily high level of support drawn by Social Credit in the federal election of 1963 in heavily French areas complicates the federal-provincial comparison.

²⁵ Since 1930 there has always been at least one French-Canadian member of the Ontario cabinet – with the single exception of the Drew administration from 1943 to 1948 – and always from the Ottawa Valley region of the province, even when other candidates were available. There are now two French-Canadian cabinet ministers, the second coming – as might be expected – from northeastern Ontario.

²⁶ We have confined our analysis to areas of French and British concentration largely because hardly any 1961 urban census tracts have a sufficiently large proportion of their population concentrated in groups other than these two. A more elaborate investigation based on the reaggregation of census enumeration areas might overcome this difficulty. None the less, it is a curious fact that provincial constituencies which are known to have large immigrant populations (such as Toronto Dovercourt) invariably have a significantly smaller number of registered electors than the corresponding area in the immediately previous federal election, and yet the number always rises for the next federal election. Whether this is an expression of the lower salience which provincial elections may have for immigrant voters or whether it is the result of the application in previous years of section 90 of the Ontario Election Act (which permits the denial of the vote to a non-English-speaking elector who cannot be made, through an interpreter, to understand the rules of the election) is unknown.

²⁷ The census tracts illustrated in Table VII are all very heavily British in ethnic origin.

²⁸ The long discussion concerning the use of aggregate data in the analysis of electoral behaviour is summarized in Shively, W. Phillips, “‘Ecological’ Inference: The Use of Aggregate Data to Study Individuals,” American Political Science Review, 63, no. 4 (Dec. 1969), 1183–90.Google Scholar

²⁹ This is primarily because the characteristics which make people unavailable to vote on election day also make them unavailable for interview whenever approached. For example, single men are rarely at home, while older people are frequently confined to home through illness.

³⁰ The material which follows is based on information obtained from a combined total of 1,598 respondents to two waves of a survey conducted in June and July 1968 and again in November and December of the same year, under the direction of David Hoffman and Fred Schindeler. The sample was a specially designed probability sample of Ontario residents eligible to vote in the 1968 federal election. The study was made possible by a research grant from the Canada Council.

³¹ Abstainers at either level were considered to be (1) those who said they did not vote (26 per cent in 1967), (2) those who said that they did not know whether they had voted (3 per cent in 1967), and (3) those who, having said that they had voted, said that they did not know for which party's candidate they had voted (4 per cent in 1967). The number of cases shown for each election in Table VIII is less than the total number of respondents to the survey because we have excluded those who gave no answer to the question “did you vote?”

³² In the tables which follow only 1,384 cases are reported because we have excluded from consideration all respondents who were ineligible to vote in either or both of the 1965 and 1967 elections, as well as those who refused to report their voting behaviour.

³³ Analysis of aggregate data led Howard Scarrow to criticize the balance theory, but he concluded that the weakness of the provincial Liberals was largely explained by federal Liberal supporters switching to the ccf in provincial elections. See “Federal-Provincial Voting Patterns in Canada,” 295. Jean Laponce, using (like Scarrow) percentage shares of votes cast, also concluded in a recent paper that in the Ontario case the federal-provincial shift occurred between the Liberals and the ccf/ndp. See “The Conservative Party, Its Leftwing Allies and Its Provincial Reincarnations: The Feeding Habits of Small Parties,” part of a paper presented to a joint conference of the Canadian Political Science Association and the Société canadienne de science politique, Ottawa, 1968.

³⁴ See Rosenberg, Morris, The Logic of Survey Analysis (New York, 1968), 50–3.Google Scholar

³⁵ This finding – since the Franco-Ontarians in the sample were mainly in urban areas – tends to support our analysis of the aggregate data (see Table VI).

³⁶ See Blishen, Bernard R., “A Socio-Economic Index for Occupations in Canada,” Canadian Review of Sociology and Anthropology, 4, no. 1 (Feb. 1967), 41–53.CrossRefGoogle Scholar

³⁷ That the theory continues to support analyses of provincial electoral behaviour can be seen in two comments on the June 1969 victory of the ndp in Manitoba. In probing the possible consequences for a future federal election Anthony Westell wrote that “the voters often seem, perversely, to reverse nationally what they have done provincially – seeking perhaps an instinctive balance of power.” Toronto Daily Star, June 27, 1969. In the August 1969 number of Maclean's an editorial asked the question: “Does a provincial victory have any federal overtones in a nation whose electors seem instinctively to assert a balance of power between the provinces and Ottawa?”

³⁸ It is possible that some respondents acquired their attitudes to the balance theory after the two sets of elections and therefore would not have been affected in their voting by it at the time. More plausible, in the light of our findings, is the conclusion that most respondents have never acquired an opinion on the question which was reflected in their voting behaviour at any time.

³⁹ As a check upon the quality of responses to the statement a “reversal” of the interview statement was included in the so-called “piggy-back questionnaire” which was left by interviewers for respondents to complete and return at their leisure. Since a large number of respondents returned this questionnaire it is possible to compare the answers of a very substantial sub-set of the sample on the two similar items and to establish their degree of consistency. The remainder of this aspect of the analysis is based on the responses only of these people.

⁴⁰ This pattern of change may help to explain the fact that pre-election surveys in Ontario invariably overestimate the polling-day strength of the provincial Liberal party. In its issue for January 7, 1967, the Toronto Daily Star reported a specially commissioned Gallup Poll of the province, taken the previous December, which showed the Liberals with 40 per cent of the committed vote, the Conservatives with 35 per cent, and the ndp with 25 per cent. See also a series of articles by Peter Regenstreif in the Toronto Daily Star, Sept. 18, 19, 21, 23, 24, and 26, 1963, where he reported a study of the Ontario electorate completed some two weeks before the 1963 provincial election and found the Liberals with 39 per cent of the committed vote, the Tories with 48 per cent, and the ndp with 13 per cent. In February of 1967 a survey carried out on behalf of the Ontario New Democratic Party found that the Liberals were leading the Conservatives by two percentage points with 38 per cent of the committed vote, while the ndp trailed with only 26 per cent. The fact is that there are more Liberals than Conservatives in the province. When respondents in the ndp survey were asked to identify with one of the provincial parties 30 per cent said they were Liberals, 26 per cent were Conservatives, 19 per cent were New Democrats, and the rest were either independents or uncertain of their party identification. This 1967 study was based on an area probability sample of 1,043 electors in the four standard regions of the province.

⁴¹ The measure used in classifying respondents in the 1968 survey as federally oriented or provincially oriented was a five-item index. Ten per cent of respondents were considered highly “federally oriented” and six per cent were considered highly “provincially oriented” with the rest being distributed in between the two extremes. Since this scale does not correlate with the Blishen scale it can be assumed that we are not simply reporting the status differences noticed earlier. Higher status federal Liberals who were highly provincially oriented were even more disposed to vote Conservative in provincial elections than higher status federal Liberals who were highly federally oriented.

⁴² The malaise in the party may be seen in the frequent changes in leadership since the end of the Second World War. After Hepburn (who had become House leader during the 1944 session) was personally defeated in Elgin in 1945 Farquhar Oliver was chosen as provincial leader. Since then there have been six further changes – in sharp contrast to the ccf/ndp which has had only two leaders in Ontario since the office was established in the party in 1942. The constant shuffling at the top can hardly contribute to a strong image for the Liberals in the province.

⁴³ While only 13 per cent of ndp identifiers and 30 per cent of Conservative identifiers made such vague or mildly hostile remarks about their party, no fewer than 43 per cent of all Liberal identifiers did so. These results were obtained in the 1967 Ontario ndp survey referred to in note 40.

⁴⁴ The most striking of all these departures must surely have been that of N. W. Rowell, who resigned the provincial leadership of the party in 1917 to join Borden's Unionist cabinet, taking a good part of the organizational structure he had built up in Ontario with him. See Peter Regenstreif, “The Liberal Party of Canada: A Political Analysis,” unpublished ph d dissertation, Cornell University, 1963, 129. On this point see also Silcox, Peter, “Some Problems of Opposition Leadership in Ontario Provincial Politics,” paper presented to the annual meeting of the Canadian Political Science Association, Calgary, 1968.Google Scholar

⁴⁵ See Regenstreif, “The Liberal Party of Canada,” 246–51. The presence of eight federal cabinet ministers at the convention which chose Harry Nixon as leader in 1943 has a more recent parallel in the federal influence which was evident at the Ontario party's 1964 leadership convention. The eventual winner, Andrew Thompson, had to work extremely hard to escape the charge that he was the federal party's candidate. Even before the war the national party often put its own interests first. In 1930, for example, Mackenzie King was reluctant to see Mitchell Hepburn contest the Ontario leadership because he feared the outcome of a federal by-election in Elgin West – the riding for which Hepburn sat in the House of Commons. See McKenty, Neil, Mitch Hepburn (Toronto, 1967), 34.Google Scholar

⁴⁶ There is some evidence that these circumstances are changing. In a provincial by-election in Middlesex South in September 1969 – where the ndp won a striking victory over the Conservatives – the party not only overwhelmed its opposition in the working-class east end of London, but frequently pushed the Liberals into third place in the dominantly rural townships which take up roughly two-fifths of the constituency's population. For an analysis of the by-election based on an extensive survey of voter opinion, see Surich, Joachim E., “Political Socialization in Ontario: The Case of Middlesex South,” unpublished ma dissertation, University of Waterloo, 1970.Google Scholar