Old theses, new data: A methodological analysis

The author revisits the themes proposed by Jairo Nicolau in a previous article (Dados, vol. 47, n(o)1, 2004) concerning coalition votes in Brazil. Using election tables from the Superior Electoral Court and tables amended by Nicolau, the author evaluates whether the data studied by Nicolau support the thesis that he significantly modified the original Electoral Court data and thereby invalidated current hypotheses on variations in party votes from 1946 to 1962 in relation to the degree of nationalization in the Brazilian party system and fractionation rates, both during the same period. The author concludes that Nicolau's theses are invalid, including the use he makes of the concept of "effective number of parties".