Abstract
I propose an alternative index to operationalize the variablenumber and size of partiesin a party system. To support it, I present a critical overview of the two most common indices, Laakso and Taagepera'seffective number of parties andKesselman and Wildgen'shyperfractionalization, showing the causes of their weakness. Then I explain the computational logic of my alternative, “number of parties,” and compare it with the other two, using hypothetical cases. After that, I contrast the Laakso-Taagepera index with mine, using data from actual elections between 1945 and 1981. I conclude that my index outperforms the other two as an operationalization of the variable number and size of parties.

This publication has 10 references indexed in Scilit: