Suchergebnisse

TWICE BEFORE, FROM 1966 AND 1975, THE WRITER HAS ANALYSED UNDERLYING VOTING SWINGS IN A NEW ZEALAND GENERAL ELECTION, USING A LINEAR REGRESSION MODEL1. THE ABSENCE OF ELECTORAL BOUNDARY CHANGES IN THE 1981 ELECTION ENABLES THE MODEL TO BE USED AGAIN FOR THAT YEAR WITH ONLY MINOR VARIATIONS, AND THIS THIS NOTE REPLICATES THE EARLIER ANALYSES REFERRED TO. THE MODEL DESCRIBED IN THE 1977 ARTICLE GIVES RISE TO THREE EQUATIONS TO BE ESTIMATED, SUCH AS L*=OL + (LL-OL) L + (NL-OL) N + (SL-OL)S WHERE FOR THE REPRESENTATIVE ELECTORATE, L* DENOTES THE PROPORTION OF LABOUR VOTES IN 1981; L, N AND S RESPECTIVELY DENOTE PROPORTIONS FOR LABOUR, NATIONAL AND SOCIAL CREDIT IN 1978; AND THE PAIRS OF LOWER-CASE LETTERS DENOTE PERCENTAGE VOTING SWINGS AS SHOWN BELOW. O REFERS TO VOTES CAST FOR MINOR PARTIES (MAINLY VALUES), INFORMAL VOTES AND NON-VOTES (SO FAR AS THESE CAN BE ASSESSED).