ABSTRACT
This study uses econometric methods for the analysis of nonstationary, cointegrated variables to specify and analyze an error-correction model of interrelationships among Labour vote intentions, best prime minister judgements and party identification since the beginning of 1992. These three series have varied over this time period, and analyses reveal that both best prime minister judgements and party identification have had important influences on the short- and long-term dynamics of Labour-vote intentions. The analyses also show that subjective economic evaluations have exerted direct and indirect effects on Labour support.
