Search

Within the field of public opinion, there is a growing awareness that the public as a whole does not make decisions in accord with strict notions of rationality. The public, like individuals, is subject to the inuence of partisanship and inattention. Distinguishing between these effects has important normative consequences, yet it is difficult to disentangle their empirical effects in aggregate behavior. Both partisanship and inattention can lead people to be slow to update their opinions in light of new information. In this study, I use an important and underutilized dataset, the German Socio-Economic Panel, to isolate the effects of partisanship and inattention on aggregate opinion. Because this is a longitudinal study (1984-2009) with questions about partisanship, economic evaluations, and several policy evaluations, it is possible, through dynamic factor models, to identify when individuals exhibit partisanship and inattention and trace this impact to the aggregate level. [ABSTRACT FROM AUTHOR]

In order to better understand information processing at the aggregate level, I a) describe a model of information processing of aggregate opinion that is consistent with the individual-level models, b) show how this information processing model works in the context of the public's evaluation of the economy and c) compare aggregate opinion about the economy to a benchmark of strict \rational" decision-makers represented by the opinions of professional forecasters compiled in the Survey of Professional Fore- casters. [ABSTRACT FROM AUTHOR]

In this paper, we use the public's approval of the president in managing the economy as a context for testing therole of partisanship in aggregate opinion. Drawing on Bartels' (2002) notion of biased partisan updating andGerber and Green's (1999) model of unbiased Bayesian updating, we find consistent (although not uniform) evidence of partisan bias in aggregate opinion. These findings in support of partisan bias lead us to consider how partisanship can be incorporated into models of aggregate public opinion. Drawing on the work of Mankiw and Reis (2002,2006) and Carroll (2003), we develop a model of economic approval that takes seriously the role of partisanship--namely, the influence of partisanship is likely to create a public that respondsmore slowly and heterogenously to changes in the economic environment than one that updates in a --rational" and --unbiased" way. ..PAT.-Unpublished Manuscript [ABSTRACT FROM AUTHOR]

Abstract Although much of the systematic, over-time research on the presidency has focused on presidential approval and how it varies in response to economic conditions and world events, there is growing interest in the impact that the public’s personal evaluations of the president can have on presidential approval. In an age where ‘image is everything,’ it is important to understand how these personal evaluations of the president vary over time and to assess their impact on presidential approval. In this paper, I develop a measure that can track the public’s personal approval of the president. I begin with an examination of the cross-sectional studies on character assessment to build a measurement strategy that will yield a clear and consistent measure of the president’s personal approval. I then describe the measurement technique used to calculate this measure. With the measure in hand, I provide a set of analyses that explain the changes in the personal approval measure over time and examine the impact of the measure on the president’s job performance approval rating. [ABSTRACT FROM AUTHOR]

In the procedure for merging time series outlined here, I describe a means to create a latent factor measured from different pieces of time series that are observed, even if some of the series have missing data. The Kalman filter and state-space methods provide the structure to model this problem, and the EM algorithm is particularly well-suited to the maximization task, since it is designed to handle missing data. ..PAT.-Unpublished Manuscript [ABSTRACT FROM AUTHOR]

The focus of this paper is on the role of political knowledge and partisanship in the connections citizens make between changes in the environment and their political evaluations. Although this question could be studied in a variety of settings, presidential approval is perhaps the most useful. ..PAT.-Conference Proceeding [ABSTRACT FROM AUTHOR]