Although Duverger is traditionally seen as synonymous with the institution-list approach to party systems, this article shows that he believed social pressures were the driving force behind the multiplication of parties. Electoral institutions are important, but only because they determine the extent to which social forces are translated into political parties. Although the literature has finally come to realize that social and institutional forces interact to shape party systems, scholars still do not seem to fully understood the implications of Duvergers theory. This article shows that existing research employs flawed statistical specifications, makes inferential errors, and does not calculate desired quantities of interest. Using a new data set that includes elections since 1946, the authors reexamine Duvergers theory and find that modern tests largely bear out his expectations when properly specified and interpreted. This analysis also extends current research by specifically estimating the mechanical and strategic modifying effects of electoral institutions.
