Predictors of recidivism among halfway house residents

This study concerns client success or failure while residing in a halfway house for adult offenders. As with some other studies, failure or recidivism is defined in this analysis as behavior that results in removal from the facility and subsequent incarceration. This quantitative case study of one treatment facility in Tennessee utilized case records on the population of clients from 1981 through 1987 (N = 75). A number of hypotheses were developed and tested with both bh/ariate and multivariate methods. Results indicated that, at least in this halfway house, those residents who had experienced two or more prison terms and who had most recently been convicted for a violent crime were most likely to recidivate while in the facility. Other substantively significant factors were alcohol use, religious attendance, and the experience of mental health counseling. Programmatic implications of these results are discussed.     

Toward a New Critical Theory with a Cosmopolitan Intent

In this article I want to outline an argument for a New Critical Theory with a cosmopolitan intent. Its main purpose is to undermine one of the most powerful beliefs of our time concerning society and politics. This belief is the notion that “modern society” and “modern politics” are to be understood as society and politics organized around the nation‐state, equating society with the national imagination of society. There are two aspects to this body of beliefs: what I call the “national perspective” (or “national gaze”) of social actors, and the “methodological nationalism” of scientific observers. The distinction between these two perspectives is important because there is no logical co‐implication between them, only an interconnected genesis and history.     

Testing Ricardian equivalence under uncertainty

This paper uses experimental methods to analyze Ricardian equivalence when the probability of debt retirement is less than one. The results suggest that the presence of outstanding debt and the probability of debt retirement have a strong influence on savings behavior. When the probability of debt retirement is low, consumption by the current generation increases, as predicted by Keynesian theory. However, as the probability of debt retirement increases, bequests rise to offset the future generation's expected repayment liability, and deficit spending becomes much less expansionary, as predicted by Ricardian theory. In general, the average bequest is significantly larger when an outstanding debt is passed on to the next generation than when no debt exists, regardless of the probability of debt retirement. However, as long as there is some uncertainty about debt repayment, the presence of debt always stimulates some additional consumption, so that strong variants of Ricardian equivalence are not found.     

Encouraging Political Defection: The Role of Personal Discussion Networks in Partisan Desertions to the Opposition Party and Perot Votes in 1992

Drawing on data from a unique study of the 1992 American presidential election, this article demonstrates that personal discussion networks influence voting behavior, independent of candidate evaluations and partisanship. These social networks encouraged two different kinds of defections from otherwise-expected behavior. People were more likely to vote for Perot if their personal discussants supported him and to convert preferences for him into a Perot vote on election day. Partisans also were more likely to defect to the other major party if their discussion network failed to fully support the candidate of their own party. These results withstood controls for candidate evaluations and partisanship as well as for selective exposure to discussants and selective perception of their preferences. They show the importance of adding social context to personal attitudes, interests, and partisanship in explaining voting behavior.