Suchergebnisse

Recently, a number of prominent Republican elites have argued that the economic plight of African Americans is attributable to undocumented immigration to the United States. Have these arguments concerning the link between black economic well-being and undocumented immigration become commonplace in the rhetoric of Republican elites, and if so, does exposure to these appeals impact black vote choice? Employing data from over forty years of congressional speeches, the campaign speeches and public addresses of President Donald Trump, televised campaign advertisements from the Wisconsin/Wesleyan Advertising Projects, and a survey experiment embedded in the 2016 Cooperative Congressional Election Study, we find that Republican elected officials have increasingly made substantive appeals to blacks on the issue of immigration reform, that exposure to this type of substantive appeal leads blacks to more strongly support a fictional Republican candidate, and that this support is moderated by a respondent's level of linked fate. These findings challenge existing scholarship that Republican elites ignore the concerns of the black community and suggest that Republicans may be using the issue of immigration to drive a wedge in the Democratic electoral coalition by targeting the Democratic Party's strongest constituency.

AbstractContemporary dictionary-based approaches to sentiment analysis exhibit serious validity problems when applied to specialized vocabularies, but human-coded dictionaries for such applications are often labor-intensive and inefficient to develop. We demonstrate the validity of "minimally-supervised" approaches for the creation of a sentiment dictionary from a corpus of text drawn from a specialized vocabulary. We demonstrate the validity of this approach in estimating sentiment from texts in a large-scale benchmarking dataset recently introduced in computational linguistics, and demonstrate the improvements in accuracy of our approach over well-known standard (nonspecialized) sentiment dictionaries. Finally, we show the usefulness of our approach in an application to the specialized language used in US federal appellate court decisions.

Although racial bias in the law is widely recognized, it remains unclear how these biases are in entrenched in the language of the law, judicial opinions. In this article, we build on recent research introducing an approach to measuring the presence of implicit racial bias in large-scale corpora. Utilizing an original dataset of more than one million appellate court opinions from US state and federal courts, we estimate word embeddings for the more than 400,000 most common words found in legal opinions. In a series of analyses, we find strong and consistent evidence of implicit racial bias, as African-American names are more frequently associated with unpleasant or negative concepts, whereas European-American names are more frequently associated with pleasant or positive concepts. The results have stark implications for work on the neutrality of the legal system as well as for our understanding of the entrenchment of bias through the law.

Past research has shown that issues vary significantly in their salience across citizens, explaining key outcomes in political behavior. Yet it remains unclear how individual-level differences in issue salience affect the measurement of latent constructs in public opinion, namely political ideology. In this paper, we test whether scaling approaches that fail to incorporate individual-level differences in issue salience could understate the predictive power of ideology in public opinion research. To systematically examine this assertion, we employ a series of latent variable models which incorporate both issue importance and issue position. We compare the results of these different and diverse scaling approaches to two survey data sets, investigating the implications of accounting for issue salience in constructing latent measures of ideology. Ultimately, we find that accounting for issue importance adds little information to a more basic approach that uses only issue positions, suggesting ideological signals for measurement models reside most prominently in the issue positions of individuals rather than the importance of those issues to the individual.

We examine a problem that is confronted frequently by political science researchers seeking to model longitudinal data: what to do when one suspects a lag between the realization of a regressor and its effect on the outcome variable, but one has no theoretical reason to suspect a particular lag length. We examine the theoretical challenges posed by atheoretic lags, review existing methods for atheoretic lag analysis—most notably distributed lag specifications—and their shortcomings, and present an alternative approach for atheoretic lag analysis based on Bayesian model averaging (BMA). We demonstrate the use and utility of our approach with two examples: the litigant signal model in American politics and modernization theory in political economy. Our examples show the increasing difficulty of analyzing models with atheoretic lags as the set of possible specifications increases, and demonstrate the effectiveness of BMA for the modal type of specification in time-series cross-sectional applications.

AbstractThe Institutional Grammar (IG) is used to analyse the syntactic structure of statements constituting institutions (e.g., policies, regulations, and norms) that indicate behavioural constraints and parameterize features of institutionally governed domains. Policy and administration scholars have made considerable progress in methodologically developing the IG, offering increasingly clear guidelines for IG‐based coding, identifying unique considerations for applying the IG to different types of institutions, and expanding its syntactic scope. However, while validated as a robust institutional analysis approach, the resource and time commitment associated with its application has precipitated concerns over whether the IG might ever enjoy widespread use. Needed now in the methodological development of the IG are reliable and accessible (i.e., open source) approaches that reduce the costs associated with its application. We propose an automated approach leveraging computational text analysis and natural language processing. We then present results from an evaluation in the context of food system regulations.