ABSTRACT
The description of regional dominance of a party has led to some early chef d’oeuvre in electoral studies. The democratic South depicted by V.O. Key (1949) and the conservative Western France by A. Siegfried (1913) are examples of such monuments in political science. They invite us to take into consideration the problem of scale in the study of party dominance. As G. White (1973) remarked, party dominance can be defined at least at three levels, from the very local one (that of the district or constituency especially for electoral systems with low district magnitude) to the regional and national levels. These different levels seem in fact quite disconnected. A nationally dominant party can rely on only weak bases at the local level (for instance the Gaullist party in France in the 1960s) and a country that districts are all dominated may have a competitive pattern of political competition at the national level, depending on the distribution of local dominance among parties. This is especially true in federal contexts such as Canada, for example.
