Abstract: This study analyzes determinants of union election outcomes at the level of the work unit. Within a theoretical framework of utility maximization, voting behavior is modeled as a function of the social psychology of groups, the economic and sociopolitical environment, NLRB procedures, and the extent of union organization of the industry. Utilizing NLRB certification-election records for 1979, the author finds a negative relationship between unit size and union victories in units of fewer than 65 workers, but no relationship in larger units. Also negatively related to union victories are delays between petition and election dates, elections held in southern states having right-to-work laws, and elections involving the Teamsters. In contrast, workers are more likely to vote for representation as unemployment levels and the proportion of consent elections rise and as the rate of unionization in their industry rises to 35 percent.

Abstract: Competing theories about differences between American blacks and whites in rates of political participation and joining of voluntary organizations are tested using data from a 1967 national sample survey. Prior literature offers four theories that seek to explain these differences: ethnic community theory, compensatory theory, isolation theory, and cultural inhibition theory. A typology of political orientations is developed, based on the underlying dimensions of political trust and political efficacy, and it is argued that the competing theories deal with the expected behavior of blacks with each of the four orientations thus defined, The theories are tested by computing expected mean participation rates for blacks of each orientational type and comparing these to the observed mean rates. The comparisons reveal whether participation rates for blacks with each orientation differ from those to be expected on the basis of their social class and demographic characteristics. The results give strong support to the ethnic community theory and clear counterevidence to the compensatory theory of black participation; the isolation and cultural inhibition theories receive virtually no support. Some broader theoretical implications of these results are suggested.

Abstract: This paper probes the process by which the foreign policy orientation of weak nations comes to reflect the preferences of more powerful nations. Two general conceptions of the nature of this process are identified. The most common view, that of the bargaining model, regards the policymaking process of weak states as relatively autonomous though influenced by reward/punishment actions of a more powerful nation which condition the weaker partner. By contrast, the dependency model stresses the long-term character of the influence and the indirect path by which it occurs. It regards the decisionmaking process as imbedded in a social/political structure which is itself distorted by the dependency relationship. Both a cross-sectional and a longitudinal analysis relying upon UN voting data and measures of the relations between the United States and 88 less-developed nations indicate that the explanatory power of the bargaining model is relatively limited and that the dependency model is a more appropriate conception. Though cross-sectionally, both reward behavior (various forms of aid) and dependency-indicating transactions (e.g. treaties, trade, arms sales, IGO memberships, consultations, etc.) exhibit correlations with voting behavior, those of the latter are generally considerably stronger. Further, longitudinal analysis exposes much greater stability in voting behavior over time—and much less correlation with aid-giving—than one would expect if bargaining were present. This stable pro-American behavior is precisely that which would be predicted by a theory resting upon long-term distortions implicit in an enduring and penetrating structural relationship.

Abstract: Most recent studies of congressional elections have depicted Senate elections as hardfought contests in which both candidates are familiar to voters. Evidence from state-level surveys indicates, however, that although some Senate elections are hard-fought, a substantial proportion are low-key races in which one candidate is unable to garner the resources necessary to mount a real challenge. In low-key elections, voters are much less familiar with the challenger than with the incumbent, and defections favor the incumbent. In hardfought races, both candidates are well-known, and the incumbent has no advantage in defections. Constituency-based explanations for why some Senate races are more competitive than others are questioned; an explanation focusing on decisions of the candidates is suggested. Until recently, research on voting behavior in congressional elections focused almost exclusively on House races. Occasionally, Senate elections have been mentioned as points of contrast; more often they have not been considered. Within the last several years, attention to Senate elections has increased. The 1978 Center for Political Studies National Election Study contained for the first time a battery of questions on voting for Senate candidates, leading to a number of articles comparing voting patterns in House and Senate contests. The tendency in this new literature is to portray Senate elections as hard-fought, highly publicized races, which, unlike House races, are highly salient to the electorate. This new literature also claims that the vast majority of voters acquire some information about both candidates in Senate elections and that this leads to voting patterns quite unlike those in House elections. This article disputes this newly emergent wisdom about Senate elec

Abstract: left-right locations do not have much political meaning for British voters (Butler and Stokes, 1969), even though comparatively large numbers of them will assign themselves an abstract left-right location when asked to do so (Inglehart and Klingemann, 1976; Barnes and Kaase, 1979). Norwegians have never been asked to locate themselves on a purely left-right scale, although at the Norwegian electoral study of 1977 they were asked to rate themselves on a Conservative-Radical dimension, and almost 85 percent of the respondents did so. Our approach to assigning left-right locations to the voters in our samples did not involve such direct questioning. Instead, we employed the method already used by Valen (1973) in his analysis of the Norwegian prereferendum survey data. Respondents at the British and Norwegian postreferendum surveys were asked three questions, posed in Likertscale format, relating to conventional elite-level interpretations of the left-right dimension. TWo of the questions concerned government control of the economy and the third concerned equality of income.8 The responses to the three questions intercorrelated more strongly in Britain than in Norway, but the three items formed a single cluster for each country when subjected, along with other items, to factor analysis. Not IThe questions asked respondents to react on a five-point agree-disagree scale to the following statements: (1) "Unless the government controls private business, industrialists will have too much influence"; (2) "It is easier to keep full employment in Britain [Norway] if the government has control over private business"; and (3) "There should be far greater equality of wage and salary levels in Britain [Norway]." These questions were first used in Norway and translated into their closest renditions in English for the 1975 British survey. For Britain, in order to maximize case numbers from our comparatively small sample, missing data were coded at the midpoint of the response categories. For Norway, we only included respondents who replied to all three questions; this excluded, at most, three percent of the Norwegian respondents. This content downloaded from 207.46.13.60 on Thu, 21 Apr 2016 07:36:06 UTC All use subject to http://about.jstor.org/terms REFERENDUM VOTING BEHAVIOR 49 surprisingly, the two items that explicitly referred to government control loaded more heavily than the item relating to income equality, but all three items reflect a common factor. We constructed a simple additive scale for each country based on replies to the three questions. The scales are identical in range (from 3 to 15) and, we presume, identical in meaning. Given the nature of the questions on which the scales are based, we interpret them to be a version of Downs' tentative formulation of the left-right dimension, reducing "all political questions to their bearing upon one crucial issue: how much government intervention in the economy should there be?" (Downs, 1957, p. 116). Moreover, the items faithfully reflect, at least partially, issues on which left and right, as conventionally perceived at the elite level in both countries, differ in routine political debate and action.9 The distributions of our British and Norwegian respondents on this contrived left-right dimension appear in Figure 1. Of course, there, is considerable overlap between left-right locations and affiliation with parties conventionally considered leftist or rightist. In both countries thie contribution of people affiliated with the Labor Party (and, in the case of Norway, the Socialist People's Party) is greatest at the extreme left and diminishes gradually as one moves right, while the contribution of Conservative Party supporters is just the reverse. On the other hand, the patterns for the British Liberal Party, and for the Norwegian Christian, Center, and Liberal parties, are almost invariant across the spectrum. There is only a slight peaking at the center, with some modest falloff at the extremities. For those voters, left-right locations are essentially irrelevant to their partisan attachments. That does not necessarily mean, however, that their left-right locations are also irrelevant to their attitudes toward the EC. The Relevance of Left-Right Attitudes The policy content of our left-right scale is theoretically related to expected attitudes toward the EC. Although leftist political movements have historically emphasized internationalist themes, such a posture raises serious concrete problems for a leftist government in power. To the extent that a government emphasizes central control over the economy, there is a risk that the obligations of membership in international organi90ne caveat, cited by Downs, is that "the parties designated as right-wing extremists in the real world are for fascist control of the economy rather than free markets" (Downs, 1957, p. 116). This ambiguity is at work in connection with British attitudes toward the EC as well, inasmuch as certain groups regularly regarded as extreme rightist were opposed to British membership. The left-right dimension implicitly applied to those groups has little to do with government intervention in the economy but rather refers to such matters as racial attitudes or attitudes toward foreigners. This content downloaded from 207.46.13.60 on Thu, 21 Apr 2016 07:36:06 UTC All use subject to http://about.jstor.org/terms I / //~~~~/ g e_--I ! v i z~~~~~~~~~~~~~~~~~~~~~

Abstract: This book covers a variety of topics on newspaper and television coverage including the media and voting in presidential elections, coverage of the 1980 debates, the primaries, polls, economics and energy issues, print and broadcast coverage as compared with the 1976 and 1972 elections, and the influence of TV treatment of the campaign on voting behavior.

Abstract: This book deals with the problems that arise in the analysis of electoral history -- the sources for which are fragmentary, or biased, or plagued with discrepancies. The contributors discuss ways of uncovering new, previously untapped data, detecting vote fraud by using computer-readable sources, and using new statistical techniques for working with quantitative data. Outside of its specialist subject, it is a much needed overview for the historian. '...an informative introduction to the use of electoral statistics...' -- Canadian Journal of History, Vol 18 No 3, December 1983

Abstract: This paper analyzes voting behavior for President, Senate, and House att the state level for the period 1914-80. The major finding of this study is that since World War II, there has been a pronounced trend toward 6'particularization" in voting for federal offices. This trend consists of more than the erosion of party attachments atmong the electorate: it also includes the dissolution of other traditional electoral ties such as presidential coattails, the midterm congressional swing, and incumbency. As a result of this trend, the vote in federal elections is largely determined by factors which alre unique to the specific setting in which each election occurs and by the ability of candidates to convert these factors to their political advantage through mass media campaigns. Peter S. Tuckel is an Assistant Professor of Sociology alt Hunter College, The City University of New York. Felipe Tejera is a doctoral student in sociology at New York University. The auithors wish to thank D)r. Richard Matisel of the Department of Sociology. The Graduate School of Arts and Sciences, New York University. for his invaluable aissistance in all phases of this project. Public Opinion Quarterly Vol. 47:230-246 ? 19S3 hy the Trustees of Columbi.a University Published by Elsevier Science Publishing Co.. Inc. CX)33-362X/M31)0047-23(1S2.50) This content downloaded from 157.55.39.104 on Mon, 20 Jun 2016 06:29:42 UTC All use subject to http://about.jstor.org/terms AMERICAN VOTING BEHAVIOR, 1914-1980 231 In this article, we use the term "particularization" to describe this more general process in which each election becomes an isolated event determined by the unique set of circumstances surrounding it. Specifically, this article aims to document the increasing trend toward particularization in voting in federal elections from 1914 to the present. The discussion is in five sections: (1) the data base, (2) the decline in communality in voting for federal offices during the same election year, (3) the decline in the effect of incumbency on the outcomes of U.S. Senate races, (4) an examination of the impact of campaign expenditures on senatorial voting, and (5) a brief overview of the findings.

Abstract: Previous papers on the voting activities of the FASB (Newman [1981a; 1981b] and Selto and Grove [1982] have examined whether some members or coalitions of members of the FASB wielded voting power in excess of their votes. Each of these studies was conducted without the benefit of a general theory of voting on the FASB-a theory which would predict (explain) members' voting behavior(s). Instead, each provided descriptive evidence on the relative ability of different voting indices to ascertain the existence of voting coalitions from prior voting behavior. To illustrate, Newman [1981] suggested that such voting power indices as the Shapley-Shubik or the Banzhaf indices could describe how former "Big Eight" members of the FASB appeared to control the outcomes of the FASB. Selto and Grove (SG) [1982] extended Newman's work to confirm that while former "Big Eight" members apparently did not control the FASB on SFAS Nos. 26-44, three other coalitions seemed likely. SG tested the descriptive power of the Shapley-Shubik and the Banzhaf indices for SFAS Nos. 26-44 for these coalitions. For two of the three the best descriptive index was the Shapley-Shubik index. However, the best descriptor for one of the coalitions was the nominal voting power or simply the proportion of votes held. Since we have no theory to support these results, we attempted to

Abstract: the controversy. Stages one and two both concern &dquo;gatekeeping&dquo; activities. The first stage involves the grant or denial of certiorari and writs of appeal and is the subject of an extensive literature (e.g., Schubert, 1959; Provine, 1980; Teger and Kosinski, 1980; Baum, 1977; Ulmer, 1978; Ulmer et al., 1972; Tanenhaus et al., 1963). The second stage, which is the focus of our analysis, markedly differs from the first. Whereas stage one enables the Court to control its own docket by refusing to accept writs of certiorari and appeal, stage two involves cases that the Court has accepted for review. In this regard stage two is akin to stage three: both produce formal decisions complete with an opinion of the Court. Stage two differs from stage three, however, in that the former, as mentioned, does not address the substantive merits of the controversy. Because stage two does not resolve controversies on their merits, it serves as a second &dquo;gate&dquo; that litigants must pass through in order to secure resolution of their substantive contentions. Not only is this second &dquo;gate&dquo; distinguishable from the first stage by the presence of formal decisions, it also articulates the policy that is to guide the federal courts in their determination of what sorts of issues and what types of litigants may have their day in federal court. Whereas stage one enables the Court to control access only to its own docket, stage two establishes the policies that are to govern access to the lower federal courts as well as to the Supreme Court itself. Few empirical studies have considered stage two, notwithstanding its importance to an understanding of Supreme Court policy making. In one such analysis, we attempted to ascertain the factors that motivated each of

Abstract: F OR SOME DECADES now, political scientists have investigated the influence of various motivational, behavioral, and institutional variables on voting behavior. More recently, the phenomenon of ticket-splitting has captured the attention not only of political scientists but of political practitioners and journalists as well. Ticket-splitting refers to the behavior of individual voters; it generally is taken to mean the act of casting votes for candidates representing more than one political party on a single ballot. The study of split-ticket voting, however, has been confined to the use of indirect measures, primarily survey and aggregate data, because of the unavailability of actual ballots. This study, however, utilizes actual ballots.

Abstract: .jOLITICAL SCIENTISTS have devoted great energies to understanding individual choices in presidential elections.' Studies, usually involving survey methods, have taught us a great deal about how people who report their choices reach their decisions. In contrast, we know little about respondents who report themselves undecided, except as indecisiveness has been analysed in the context of nonvoting (Hadley 1978; Wolfinger and Rosenstone 1980). This treatment is not surprising as earlier American evidence suggested that individuals who report themselves undecided in the middle of an intense political campaign are unlikely to vote (Campbell, Converse, Miller, and Stokes 1960). With the focus on who is going to vote for which candidate and why, undecided respondents become relegated to "missing data." In the small but significant number of democracies with compulsory voting (Crewe 1981: 240), undecided respondents deserve serious attention. Since in mandatory voting systems undecided individuals eventually must cast their ballots, undecided respondents do become voters. This is recognized by pollsters and campaign strategists who devote great time and energy attempting to uncover the political motivations and preferences of the undecided. Campaign consultants assume that understanding undecided respondents is a key to winning elections. Political scientists concerned with electoral choice in obligatory voting systems have not matched the political consultants' concern with undecided respondents. This oversight may be explained in part by the methodological difficulty in dealing with subjects who respond "I don't know" to the quintessential question of most psephological research. The purpose of this paper is to examine the voting behavior and motivations of voters who report themselves undecided in preelection days surveys. We propose to show how undecided respondents can be studied and to demonstrate that a focus on the undecideds enhances the

Abstract: "Delegates Fresh from the People": American Congressional and Legislative Behavior Systematic quantitative research in congressional and state legislative history has the same virtues as similar research in other areas: the ability to examine the behavior of large masses of individuals with great precision and to uncover insights and relationships hitherto unknown or only partially understood. History of this kind attempts to refine crude generalizations and impressionistic hunches, to clarify old and contentious problems, and to open up new ideas and problems. If political history is only "surface history," as researchers of the Annales school have suggested, that surface still needs to be clearly marked and understood for other purposes.1 Most students of legislative history would reject the Annales stereotype, however. Andrew Jackson once referred to members of a Democratic national convention as "delegates fresh from the people." The same imagery with the same meaning reflects one of the prime motivations stimulating recent work in congressional-legislative voting behavior by historians. Most research has been guided by the idea that "the study of Congress can be much more than the study of a political institution and its functions; it can also be a study of the political life of the entire society.2

Abstract: The purpose of this study was to replicate earlier research into the development of student political attitudes by examining the patterns of attitudes which emerged at the third, eighth, and twelfth grades, including comparisons by gender and race. Their attitudes were conceptualized as having 3 dimensions–-attitudes towards political institutions, public political participation, and school participation. A 48-item attitude instrument was used to collect data in two large surveys. Overall results support earlier research which indicated that student political attitudes are shaped by cognitive capacity interacting with social-political environmental factors. Subtest data reveal details about the development of political attitudes, including a self-other dichotomy in some attitude dimensions. Differences by race and gender provide new insights into how social factors affect political attitudes.

Abstract: This study examines the nature of comparative political candidate evaluation. While the traditional view advocates a unitary voter decision‐making process, with one candidate compared to another using a general standard criterion for candidate acceptability, this research looked at the possibility of multiple criteria unique to each candidate as part of the evaluative process. The results of the study supported the idea of unique criteria being employed in the evaluation of each individual candidate, suggesting a more complex model of voting behavior than the traditional view heretofore has advanced

Abstract: Earlier expectations and assertions to the contrary notwithstanding, ethnic politics is alive and well in urban America. In this study of the voting behavior of four ethnic groups in a recent election in Philadelphia, Professor Featherman finds that candidates received considerably larger support from voters belonging to their respective ethnic groups than they did from the rest of the city's voters.

Abstract: AUTHOR Masterson, John T.; Biggers, J. Thompson TITLE Emotion Eliciting Qualities of Television Campaign Advertising as a Predictor of Voting BehaviOr. PUB DATE Apr 83 NOTE 17p.; Paper presented at the Annual Meetings of the Central States Speech Association (Lincoln, NE, April 7-9, 1983), and the International Communication Association (Dallas, TX, May 26-30, 1983). PUB TYPE Reports Research/Technical (143) Speeches/Conference Papers (150)

Abstract: India constitutes a rich and varied laboratory for the exploration of voting behavior. An enormous reservoir of data has accumulated not only through seven national elections spanning 28 years, but also through numerous elections held for the 21 state legislative assemblies. With a few instructive exceptions, however, inquiry into Indian elections has been limited and has proceeded with greater concern for describing electoral results and the political micro-climates of constituencies than for the systematic accumulation and testing of knowledge.1 India, like practically all of the non-Western world, has been largely excluded from consideration in comparative election studies-studies that have sprung primarily from the American case but which have expanded to encompass Canada and most of the states of Western Europe during the past two decades.2 The present investigation is a preliminary effort to widen these restricted horizons by exploring the underlying dimensions of Indian electoral processes beyond the mere description of turnout and party vote statistics that constitute most of the commentary on this extraordinary political phenomenon. Ultimately we would expect the Indian case to become fully integrated into the wider comparative analysis of electoral process and behavior. It is commonplace to note that India is the world's largest democracy. Its voter turnout rate is approximately the same as the United States while the number of votes cast is over twice as large. India, however, is neither a twoparty nor a multi-party system. Rather, it has been a one party dominant system with the Indian National Congress standing at its institutional core, although the organizational scaffolding of that core has suffered atrophy over the past decade.3 Interaction between a complex pattern of multi-party

Abstract: The development in the British parliament, during the latter half of the nineteenth century, of highly cohesive legislative parties is a leading thread in the complex of events marking Britain's passage from an aristocratic to a democratic politics. Since the 1870s, journalists and scholars have attempted to account for the marked increase in the frequency with which MPs voted with their parties, and a number of plausible hypothesis have been advanced. There has not, however, been a systematic exposition of the kinds of factors which might, in theory, have been responsible for the change, nor much in the way of testing those ideas which have been suggested. We argue that most of the explanations in the literature cannot explain the earliest increases in party cohesion -- in the 1860s and 1870s. Sometimes, this is simply because the factors to which the explanation refers are not operative until a later date. In other cases, we devise tests of the hypotheses and find them wanting. In particular, we find no support in the 1870s for the idea, associated with Mosei Ostrogonski, that the new local party associations which developed after the second Reform Act effectively pressured MPs to support their party's leadership. Our theoretical discussion of the determinants of party cohesion leads us to investigate, as an alternative source of legislative change, the effect of electoral voting behavior (or, more properly, anticipations of such behavior) on legislative voting behavior. The bulk of the thesis is devoted to this task, and proceeds as follows: First, extensive use is made of a peculiarly detailed form of electoral documentation available in the double-member districts of pre-1885 Britain to study electoral behavior in the 1841-1880 period. This study reveals clear and marked changes in British electoral behavior in the 1860s and 1870s which have not hitherto been documented in the literature. An expected utility maximization model of the decision problem faced by electors in the double-member districts is developed and used to interpret these behavioral findings. We argue that voters became more party-oriented in the 1860s and 1870s, voting more on the basis of their preferences between the two great parties -- the Liberals and Conservatives -- and less on the basis of their attitudes toward the individual candidates. This shift in the basis of electoral choice, we argue, with electors becoming less responsive to the issue positions adopted by MPs, meant that the electoral benefits to an MP of dissent were smaller relative to the sanctions available to party leaders. Hence, we expect a decline in the influence of constituents over the voting behavior of their MPs (and a concomitant increase in party voting.) A number of approaches to the measurement of the influence of constituents over their MPs' voting behavior are taken, and the findings, on the whole, support the hypothesis.

Abstract: This article examines the structure and stability of public opinion between the 1972 and 1976 presidential elections. The data consist of a four-wave panel study of 800 residents in Dearborn, Michigan; interviews were conducted in the spring of 1974, 1975, and 1976 and in the fall of 1976 immediately after the presidential election. The paper focuses on the structure of political attitudes, the stability of attitudes, and the impact of attitudes on the presidential votes. A central thesis is that public opinion can be disaggregated into two clusters-the Disintegration Cluster and the Alignment Cluster-one of which disrupts and one of which reinforces partisan voting. On the basis of correlations between the seven factors making up the clusters, we anticipate neither a rapid realignment nor a disintegration of the party system, but instability, with election outcomes affected by attitudinal factors salient to a specific campaign. We find signs of increased crystallization of political attitudes compared to the 1956-1958-1960 national panel. Finally, our voting studies indicate that the Disintegration Cluster had a significant effect on the 1972 election, but that issues in general had little impact in 1976, once controls were introduced for party identification and candidate image. Frank Wayman and Ronald Stockton are Associate Professors of Political Science at the University of Michigan-Dearborn. The two are co-equal authors. The study was funded by the Rackham School of the University of Michigan. The authors are grateful to Fran Featherston for computer assistance. The anonymous reviewers for Public Opinion Quarterly made several useful suggestions. Public Opinion Quarterly Vol. 47-329-346 ? 1983 by the Trustees of Columbia University Published by Elsevier Science Publishing Co., Inc 0033-362X/83/0047-329/$2.50 This content downloaded from 207.46.13.51 on Mon, 01 Aug 2016 04:31:30 UTC All use subject to http://about.jstor.org/terms 330 FRANK WHELON WAYMAN AND RONALD R. STOCKTON existing partisan alignment and on party stability? Third, which dimensions show signs of stability or instability? Fourth, what impact do issues have on presidential voting? In this study, the authors benefited greatly from a wealth of previous research which is both empirically impressive and theoretically insightful. Trends and patterns, well known to most scholars (and hence only summarized here), show that the old social welfare issues of the 1930s, 1940s, and 1950s are declining in saliency; that identification with the parties has weakened but not broken; that party voting has declined; that political apathy and political ideology both increased from the early 1960s to the early 1970s; and perhaps more critically, that issues have emerged which are volatile, strongly held, and persistent. These issues are not overtly economic in nature; they include such social or moral topics as race, crime, abortion, drug use, and the family. Furthermore, such issues do not always correspond to socioeconomic or partisan lines and hence have served to disrupt the parties and to reduce vote predictability (Campbell, et al., 1960; Nie, et al., 1979; Weisberg and Rusk, 1970; Miller and Levitin, 1976; Converse, et al., 1969; and A. Miller, et al., 1976). Several questions which emerged from these studies will be addressed in this paper. The first set deals with exactly how many new issue dimensions there are, how and if they are interrelated, and whether they are linked to the previous social welfare dimension. There is reason to believe that various of these new issues are linked to each other. Gun control and crime, for example, seem logically linked, as do busing and other attitudes about the relation of the state to one's children. Other questions, such as abortion and gun control, seem to have no logical linkage. The question of whether the new issues are linked in some coherent manner or are merely ad hoc is a critical one. We are assuming that if they are merely ad hoc they can be isolated in the political system and handled one at a time without producing any change or disruption. However, if they are linked together, they can restructure the present partisan alignment or disrupt it on an on-going basis (Campbell, 1966). The realignmentdisruption potential is also linked to the question of whether the old issues are still relevant or are merely artifacts of the past. If old issues have faded greatly in salience, and if new issues arise that appeal to the voters in a way that cuts across the cleavage structure of the old party system, realignment of the party system becomes possible.2 2 Sundquist (1973) provides an excellent theoretical discussion of these matters. His analysis of old and new issues is the best we have seen. See Butler and Stokes (1971) for a classic discussion of the earliest realignment to be documented with public opinion findings-the British realignment between 1910 and 1918. This content downloaded from 207.46.13.51 on Mon, 01 Aug 2016 04:31:30 UTC All use subject to http://about.jstor.org/terms STABILITY OF POLITICAL AlTITUDES 331 Weisberg and Rusk (1970) contributed a pioneering effort that demonstrated the relevance of such cross-cutting dimensions to recent presidential election campaigns.3 Subsequently, Miller and Levitin (1976:267-72) have demonstrated the utility of indices of cross-cutting social concerns in analyzing the 1972 and 1976 presidential elections. Building upon these efforts, we have attempted to further clarify the structure of cross-cutting issues in the 1970s and the impact of those issues on the 1972 and 1976 campaigns. The Structure of Attitudes We asked respondents dozens of questions tapping political and social attitudes and issue positions. Twenty-five of these items were linked together in a series of seven factors (Table 1). These seven factors in turn were linked to each other in two main clusters.4 The first cluster we call the Alignment Cluster because it is associated with and reinforces the persistent partisan loyalties of the electorate. Its components are party identification and social welfare liberalism. The second cluster we call the Disintegration Cluster because it encompasses attitudes which destabilize the party system. It is made up of law and order, an Inglehart-type value system (Inglehart, 1971), racial prejudice, racial fears, and lifestyle. The Alignment Cluster and the Disintegration Cluster are almost completely unrelated to each other. They are "nearly decomposable," in the language of cognitive psychologists (Simon, 1969). Simply stated, the fact that a person is liberal on the Alignment Cluster does not tell us whether he will be a liberal or conservative on the Disintegration Cluster. His mind can be split, for our purposes, into two distinct parts, because he does not himself make any connection between attitudes in the Disintegration Cluster and attitudes in the Alignment Cluster. But while the two clusters are independent of each other we find that within each cluster the various factors are associated in an integrated way. This is a pattern which deserves elaboration. First, however, let us look at the nature of the seven factors and their component parts: (1) The first factor, racial prejudice, has measures of attitudes toward blacks associated with it: a thermometer question 3 See Rabinowitz (1973), which illuminated many of the technical issues in such analysis. 4 The factor analysis reported in Table 1 is a varimax rotation of a correlation matrix based on pairwise deletion of missing cases, using OSIRIS iII. The analysis is based on the 1974 wave, chosen because it was halfway between the two presidential elections. A cluster is defined as a set of variables, the correlations among which are all higher than the correlations between any member of the cluster and any variable not in the cluster. This content downloaded from 207.46.13.51 on Mon, 01 Aug 2016 04:31:30 UTC All use subject to http://about.jstor.org/terms 332 FRANK WHELON WAYMAN AND RONALD R. STOCKTON Table 1. Factor Analysis of Key Questionnaire Items (Varimax Rotation)

Abstract: Sufficient impressionistic evidence exists in the previous literature to warrant a detailed consideration of the voting behavior of working-class and minority groups in relation to radical third parties. The present paper is concerned with analyzing some of the sources of support for radical third parties in Illinois from 1880 to 1924. We find that a coalition between a rural proletariat and the urban working class appeared to be a potentially viable source of support for the Socialist Party of America, since the groups involved were both numerically large and potentially crucial in determining the outcome of any election. Yet the Socialists were never able to capture more than a very modest fraction of the total vote, and even among their potential backers, the electoral enthusiasm for the Socialists was fractionalized and unstable.

Abstract: This study examines the effect of majoring in political science on students' political knowledge, attitudes, and participation. Survey research was carried out in 1979 and 1982, which compared a cohort of full-time undergraduates in their freshmen and senior years, controlling for political science as an independent variable. The findings suggest that the study of political science clearly increases political cognition and possibly political efficacy and liberalism. Although increases in political participation exist, they were not directly attributable to the major. Finally, there was little evidence to support the “self-selection” hypothesis, which argues that political science does not make people more politically aware but rather that those more politically aware major in political science.

Abstract: JOHN ANDERSON'S third-party candidacy for the presidency in 1980 has re-kindled interest in the behavior of the American voter when faced with more than two candidates. When more than two candidates are present, voters are forced from their normal pattern of straight preference voting to a more complex decision-making process. This process employs both the voters' preference ranking of the candidates and the voters' perception of each candidate's chance of winning. Furthermore, in multi-candidate races voters must contend not only with who they would like to see elected, but also with who they would not like to see elected. With three or more candidates present the voter cannot simply vote for his most preferred candidate particularly if this action in any way increases the likelihood of his least preferred candidate being elected.