Party loyalty and Irish voters - EEC referendum as a case study

Abstract

In their paper Tom Garvin and Anthony Parker set out to describe the geographical variation of the voting pattern in the EEC Referendum, and to analyse the relationship between party loyalties and the Referendum vote. The variates used are the percentages in the 1972 Referendum, and the percentages of first preferences cast for the major parties in the same constituencies in the 1969 General Election. However, a certain amount of confusion has been created through using percentages of the total poll, rather than of the total valid poll. There are good reasons for preferring the latter measure. Although the proportion of spoiled votes is of the order of 1 per cent, it cannot be related to any known political commitment, and as only valid votes are used in determining a result, the use of the total poll introduces a source of error into the effective variables. Because of. this latter effect, the correlation coefficients given in Table 1 of their text do not show up the really simple pattern which exists.