Document Type


Publication Date



Political Science


Although exit polling has not been used to study Canadian elections before, such polls have methodological features that make them a potentially useful complement to data collected through more conventional designs. This paper reports on an experiment with exit polling in one constituency in the 2003 Ontario provincial election. Using student volunteers, a research team at Wilfrid Laurier University conducted an exit poll in the bellwether constituency of Kitchener Centre to assess the feasibility of mounting this kind of study on a broader scale. The experiment was successful in a number of respects. It produced a sample of 653 voters that broadly reflected the partisan character of the constituency, and which can hence be used to shed light on patterns of vote-switching and voter motivations in that constituency. It also yielded insights about best practices in mounting an exit poll in the Ontario context, as well as about the potential for using wireless communication devices to transmit respondent data from the field. The researchers conclude that exit polling on a limited basis (selected constituencies) is feasible, but the costs and logistics associated with this methodology make a province-wide or country-wide study unsupportable at present.


This article was originally published in Canadian Journal of Political Science, 39(4): 919-933. (c) 2006 Cambridge University Press