PRESIDENTIAL APPROVAL, PARTISANSHIP AND THE ECONOMY: EVIDENCE FROM THE 1984 CONTINUOUS MONITORING SURVEY IN THE US

This paper examines several key debates in the literature onthe effects of economic performance on political support, usingthe Continuous Monitoring Survey of the 1984 CPS American nationalelection study. Box-Jenkins time-series analyses show that theelectorate operates in terms of a reward-punishment model whichis asymmetrical and sociotropic in its effects. In the modelsdeveloped party identification and economic evaluations Granger–causepresidential approval, the effects of the former being strongerthan those of the latter. Partisanship and economic evaluationsform two independent streams of influence on presidential approval,casting doubt on some revisionist interpretations in the votingliterature, i.e., that party identification is simply a runningtally of voters' judgments about party performance on economicand other issues. In addition, the results suggest that popularityfunctions which omit party identification are theoreticallymisspecified.     

Contextualizing the Interpretation of Reliability Data

What is the relationship between the context and substantive meaning of a measure and the assessment of the reliability of the measure? Does the interpretation of a reliability coefficient depend on the nature of the measure? How does this approach to reliability relate to alternative assessments?