Referendum Voting as Political Choice : the case of Quebec
By: CLARKE, Harold D.; KORNBERG, Allan; STEWART, Marianne C
.
Material type: 
ArticlePublisher: Cambridge : Cambridge University Press, April 2004British Journal of Political Science 34, 2, p. 345-355Abstract: In an article published in this journal, Nadeau, Martin and Blais argiue that perceptions of the costs and benefits of alternative outcomes and general orientations to risk interact to affect voters' decisions in referendums on fundamental political questions such as Quebec sovereignty. We use Nadeau et al.'s data to demonstrate that their interaction-effects mdel is overly complex and suffers from serious multicollinearity difficulties. A simpler main-effects model has virtually identical explanatory power and removes anomalous findings. We also argue that their model is too simple because it omits variables such as party identification, feelings about party leaders and government performance evaluations that voters use as heuristic devices to help them make decisions when stakes are high and information about the costs and benefits of referendum outcomes is low. We analyse a dataset that includes these variables and demonstrate that they have strong effects in a model of referendum voting that controls for perceived costs and benefits of alternative referendum outcomes and several other variables. Additionally, differences in the magnitudes of the perceived costs and perceived benefits of alternative referendum outcomes are not statistically significant. Tjis latter finding contradicts widely cited experimental results in behavioral economics and related 'asmmetry' hypotheses concerning the presumed status quo bias in major referendums
In an article published in this journal, Nadeau, Martin and Blais argiue that perceptions of the costs and benefits of alternative outcomes and general orientations to risk interact to affect voters' decisions in referendums on fundamental political questions such as Quebec sovereignty. We use Nadeau et al.'s data to demonstrate that their interaction-effects mdel is overly complex and suffers from serious multicollinearity difficulties. A simpler main-effects model has virtually identical explanatory power and removes anomalous findings. We also argue that their model is too simple because it omits variables such as party identification, feelings about party leaders and government performance evaluations that voters use as heuristic devices to help them make decisions when stakes are high and information about the costs and benefits of referendum outcomes is low. We analyse a dataset that includes these variables and demonstrate that they have strong effects in a model of referendum voting that controls for perceived costs and benefits of alternative referendum outcomes and several other variables. Additionally, differences in the magnitudes of the perceived costs and perceived benefits of alternative referendum outcomes are not statistically significant. Tjis latter finding contradicts widely cited experimental results in behavioral economics and related 'asmmetry' hypotheses concerning the presumed status quo bias in major referendums
Click on an image to view it in the image viewer
There are no comments for this item.