Political Research Quarterly

This paper argues that existing research on divided government has ig nored the intergovernmental dimensions and implications of this phe nomenon. We address the intergovernmental case by exploring the historical rise and micro-level foundations of divided outcomes among same-state U.S. senator and governor pairs. The electoral results for senator-governor pairs lie at the intersection of state and national-level divided government, and are subject to the same historical and electoral forces that have shaped these other divided outcomes. We trace the histor ical rise of split outcomes in these contests, outline the theoretical and policy implications of these results, and offer an analysis of individual- level voting behavior intended to compare two competing explanations for divided outcomes: the "simple party loyalty model" and our own "com parative informational relevance model."

This research addresses two questions of current interest in American po litics : Does divided government reduce accountability? Do voters reward or punish incumbent governors and/or their parties on the basis of state economic performance? A multivariate pooled cross-sectional time-series analysis of aggregate election outcomes from 1972-91 in forty-three states reveals that the effect of state unemployment on the incumbent guber natorial party's vote percentage is greatly enhanced when the governor's party controls both chambers of the state legislature. This evidence is con sistent with the claim made by responsible party theorists that voters are more able and/or willing to reward or punish a political party when that party's responsibility for government performance is complete.

This paper disaggregates change among Democratic and Republican elites in the 1980s. Using data about the views of delegates to the 1980, 1984 and 1988 national party conventions, I examine the relative contributions of conversion and replacement coupled with the preference for candidates among party activists. The analysis of the effects of preferences for candi dates and the ideological positions of the delegates revealed that regardless of whether the activists are veterans or newcomers, support for particular candidates plays an important role in the shaping of the parties' aggregate ideological orientation.

In the following analysis, we investigate two important questions: (1) how closely does media coverage of the economy reflect real changes in eco nomic conditions? and (2) to what extent does economic coverage of the economy exert an independent effect on economic evaluations? We then use this information to explore Republican claims that media coverage of the economy hurt the Bush reelection campaign. Consistent with previous research, we find that, overall, the media tend to follow negative economic conditions more closely than positive economic conditions. In addition, news coverage appears to be strongly related to aggregate public evaluations of the economy, even after controlling for real economic conditions. Final ly, we also find that news coverage of the economy was significantly differ ent during 1982, 1991, and 1992 than during other years under study. During these years, coverage of the economy was more negative than would have been expected on the basis of economic conditions alone. The implications of these findings, particularly with respect to Republican claims of media bias, are explored.

Social science has provided poor guidance for educational policymakers interested in improving student performance. While education research is a large and active field of scholarship, much of the literature has concen trated on influences beyond the control of policymakers. This paper ex plores the influences of realistic policy options. Several policies are identified that meet the criteria of a positive relationship with student per formance and controllability by policymakers.

James Tong develops a rational choice model for collective violence by peasant outlaws in pre-modem Ming Dynasty China. I argue that Tong's model can, with some alterations, be used to explain sudden, widespread outbursts of unorganized, collective violence by workers-phenomena which are anomalous vis-à-vis the "expanded logic of collective action." I apply Tong's model to the Great Strikes of 1877. The spark that lighted this prairie fire was a spontane ous nationwide strike by railroad workers. Consistent with Tong's model, rail road workers struck when their wages were reduced below subsistence levels. Because this subsistence crisis coincided with coercive crises at the local and state level and the slow mobilization by the United States Army-which reduced the likelihood of arrest, injury or death in action-these strikes quickly deep ened. They ignited an explosion of concurrent disturbances; sympathy strikes, protest marches and rallies, riots, and general strikes.

Research on litigation outcomes suggests that success is related to the level of resources available to the parties in conflict. For more developed socie ties, the "haves" appear to win more often. Research on the Philippine Supreme Court finds that the "have nots" appear to have a significant advantage, particularly for some issue areas. While these studies find inter national variations in the effects of resources on litigation outcomes, in tranational variation has not been explored. This analysis investigates the relationship between regional bias, resource inequalities, and litigation outcomes before the Philippine Supreme Court for certain economic is sues. The results indicate that there is a greater likelihood of success for individuals who challenge corporations or the government, and reside in the least developed region of the country.

This research examines the impact of changes in the presidential appoint ment process on case processing in the regional offices of regulatory boards using the National Labor Relations Board (NLRB) as a case in point. While controlling for regionally determined factors, this study investigates whether changes in the presidential appointment process influence clien tele strategies and the decision-making routines of regional offices. The results show that changes in the presidential appointment process, group strength, and unemployment differentially influence case-filtering and clientele filing behavior in NLRB regional offices generally, and particularly in those offices where clientele are weak or caseloads are low. Only changes in the presidential appointment process had a significant influ ence on both staff and clientele behavior

This study builds on previous work that has shown congressional roll-call indexes to be both reliable and valid indicators of House members' ideolo gy. Using a longitudinal set of data comprised of Americans for Democratic Action (ADA) ratings, Americans for Conservative Action (ACA) scores, Congressional Quarterly's Conservative Coalition (CC) ratings and Nation al Journal's (NJ) scores, we examine the stability of the indexes using a con firmatory factor model. The results confirm the high level of validity and reliability found previously and provide evidence of an equally high level of stability. Congressional roll-call indexes, though based on a sampling of votes taken in the House of Representatives, perform quite well with regard to these three measurement characteristics.

Participation in organized violent collective action against a government carries significant penalties should one be apprehended. Further, because such actions generally pursue collective goods, the participants will receive that good (if the action is successful) regardless of whether they participate. The free-rider hypothesis suggests that rational people will forego partici pation in large "N" collective action, unless they receive side payments of some kind. Yet, large numbers of people have periodically engaged in the type of behavior the free-rider hypothesis suggests they would not. This essay examines the solutions to this apparent paradox that have been pro posed in the literature and asks whether there has been any progress or cumulation of knowledge. Using Lakatosian criteria to evaluate this research program, the essay contends that there has indeed been progress. Further, I argue that future efforts should be invested in strategic models of collective action and that more empirical work needs to be done to flesh out the utility of recently proposed solutions.