Voting Behaviour and the Ethnic-Religious Variable: A Study of A Federal Election in Hamilton, Ontario

Abstract

Although much interest has been manifested lately in voting behaviour in Canadian elections no recent study attempts to segregate the religious and the ethnic variables. Moreover, when the religious-ethnic variable is considered as a single component, other differences such as socio-economic status, education, median age, and sex-ratio of the groups concerned are commonly slighted. This paper summarizes an attempt to examine the variables of religious affiliation and ethnicity separately, while at the same time controlling for other relevant variables. The data used in this study were collected in the “North End” district of the city of Hamilton, Ontario, in May and June, 1962, immediately prior to the federal election. The district selected is a working class area inhabited by many English- and some French-speaking Canadians and by immigrants from many European countries. It is relatively homogeneous with regard to income and style of life. Geographically, the area's boundaries are those of census tract 14. It is bordered on the north and west by Lake Ontario and on the south and east by a railway. The area is thus somewhat segregated and distinct from the rest of the city. To make the sample as fully representative as possible the probability sample method was employed. In this method every unit in the sample has an equal opportunity of being selected to be interviewed. It was carried out as follows: (1) All households located in the area to be studied were numbered in the order they are listed in the 1961 Vernon City Directory, i.e., by streets in alphabetical order. There were 2208 households listed. (2) 400 random numbers, without duplications, were selected, all of which fall between 0001 and 2208. (3) Households bearing these numbers (in our numbering of the directory) represent the sample. (4) Case numbers from 1 to 400 were assigned arbitrarily. In households where both resided, male respondents were to be interviewed in households designated by even numbers, and female respondents in households designated by odd numbers. (5) The sample is of households, not of families or buildings or persons.