The political transition in Spain provides a rare opportunity to monitor popular attitudes toward alternative regimes. Through the analysis of national surveys conducted in 1978, 1979–80, and 1984, we first establish that the Spanish public distinguishes not only between successive governments—the Franquist and the center-right and socialist governments of the post-Franco period—but also between Francoism and democracy as political systems. Second, we show that during the post-Franco era the criteria of legitimacy have begun to shift from formal political to social democratic values. These analytical results are achieved by comparing standard with less orthodox measures of political legitimacy and performance, and by revising conventional theories of system support. Third, we estimate the determinants of support for and opposition to the two regimes. The Franquist system remains more polarizing than does the democratic system; the constituencies of the democratic regime are considerably broader and more heterogeneous. However, while the new democratic state is comparatively inclusive and autonomous, low rates of political participation and changes in traditional socialist ideology have made the institutional bases of legitimacy ambiguous.

By-elections in parliamentary democracies are remarkable for the almost invariable loss of support experienced by governments relative to the preceding general election. There has been, however, surprisingly little research into this phenomenon. Using Britain as a test case, this article goes some way towards filling this gap. It identifies and tests three political explanations of governments' by-election vote losses, which it labels the turnout, referendum and personal vote models. No simple conclusion emerges, since different results are obtained depending on whether the models are treated as separate entities or their interplay is taken into account by integrating them into a single causal model. This latter strategy is argued to involve a superior conceptualization of the processes at work in bringing about governments' midterm vote losses.

Machiavelli's most famous political work, The Prince, was a masterful act of political deception. I argue that Machiavelli's intention was a republican one: to undo Lorenzo de Medici by giving him advice that would jeopardize his power, hasten his overthrow, and allow for the resurgence of the Florentine republic. This interpretation returns The Prince to its specific historical context. It considers Machiavelli's advice to Lorenzo on where to reside, how to behave, and whom to arm in light of the political reality of sixteenth-century Florence. Evidence external to The Prince, including Machiavelli's other writings and his own political biography, confirms his anti-Medicean sentiments, his republican convictions, and his proclivity for deception. Understanding The Prince as an act of political deception continues a tradition of reading Machiavelli as a radical republican. Moreover, it overcomes the difficulties of previous republican interpretations, and provides new insight into the strategic perspective and Renaissance artistry Machiavelli employed as a theoretician.

It is often argued that internationally recognized human rights are common to all cultural traditions and adaptable to a great variety of social structures and political regimes. Such arguments confuse human rights with human dignity. All societies possess conceptions of human dignity, but the conception of human dignity underlying international human rights standards requires a particular type of “liberal” regime. This conclusion is reached through a comparison of the social structures of ideal type liberal, minimal, traditional, communist, corporatist and developmental regimes and their impact on autonomy, equality, privacy, social conflict, and the definition of societal membership.

There is a glaring and potentially important discontinuity surrounding discussions of U.S. defense expenditure policy making. On the one hand, a growing body of empirically-based research questions whether the U.S. reacts in any significant fashion to the military expenditures of the Soviet Union. On the other hand, defense policy makers routinely justify defense increases as a response to similar increases by the Soviet Union. The discontinuity is resolved in the context of a multistep model of the defense expenditure policy-making process that incorporates a new estimate of Soviet defense spending and mass public opinion. Once formulated, the model provides answers to the following two questions: (1) Does the U.S., insofar as defense spending is concerned, react to the military expenditures of the Soviet Union? (2) If so, what is the magnitude of the reaction? The answers indicate that not only does the U.S. react to estimated Soviet defense spending, but that the reaction is directly responsible for a very substantial portion of the post-1978 increases in U.S. military expenditures.

How can protectionism and “free” trade succeed one another? Our answer focuses on the changing balance of private actors' political demands. These actors acquire interests in tariff policies because their assets are spatially concentrated, and trade in these assets is subject to various limitations. Actors in regions experiencing no new investment in an established industry (“old” regions) have interests that sometimes differ from those in regions where there is new investment. We show that old regions have no reason to be involved in tariff politics at business cycle peaks; during troughs, whether a state becomes more or less protectionist depends, ceteris paribus, on the relative political strength of old import-competing and old exporting interests. If old import-competing industries outweigh the old exporters, then protection will tend to increase at the trough and decrease at the peak of a business cycle; the opposite result occurs when old exporters are more influential.

By faithful representation we mean the delegation of decision making to a relatively small committee that, using a weighted voting rule, will for each pair of alternatives make sincere choices identical to those that would be made by the society as a whole, and with the same vote margins. We show that for any society, no matter how large, faithful representation is possible by a committee with no more than m(m − 1)/2 members, where m is the number of alternatives. We also show that for any society, no matter how nonideological the bulk of its electorate, social preferences can be faithfully represented by a committee whose members all have singlepeaked or single-troughed preferences. Thus, all societies can be faithfully represented by a committee whose members see the world in unidimensional terms—that is, representatives can share a coherent ideological perspective even though the electorates they represent lack such a perspective. We further show that the usual mechanisms of proportional representation and the modified form of proportional representation recently proposed by Chamberlin and Courant (1983) do not guarantee faithful representation, and we discuss mechanisms that may provide faithful representation, even in a context in which new alternatives can arise.

We present a model of party competition that produces more realistic patterns of results than those often emphasized in the literature. Reversing Downs (1957), we assume that parties win elections in order to formulate policies, rather than formulate policies in order to win elections. Voters are modeled first as having perfect information about candidate positions, and then under conditions of uncertainty. In simulation experiments we show that policy motivation and voter uncertainty can bring about persistent and predictable party differences in sequential majority rule elections. As the degree of voter certainty decreases, parties diverge towards their optima, whereas increases in voter certainty draw parties towards cycles in which party positions vary, but predictable issue stances are maintained on the average.

How do political issues arise, and come to affect political party politics? We develop a theory and model of issue evolution, illustrating both by examining the dynamic evolution of the issue of racial desegregation. Our modeling concerns two central problems: (1) the structure of the evolution—a pattern of dynamic causality between the early policy cues from professional politicians, in Congress in the case at hand, and later mass response, and (2) the sequence of changes in elite behavior, changes in mass perceptions of party issue stances, changes in mass affect toward the parties, and changes in party identifications among citizens. We suggest that the causal process developed for the racial case is quite general for other times, other nations, and other issues. The theory of issue evolution is developed as a general statement of the organic connection between elite and mass behavior, a working model of the dynamics of American politics across time and issues.

War, postwar demobilization, and economic depression are national crises that ultimately test the state's capacity to respond simultaneously to internal and external challenges. This analysis probes the nexus between crises and domestic violence, investigating how this relationship is mediated by the influence of two variables: the severity of crisis and the presence or absence of government accommodation. Box-Tiao impact assessment models are used to estimate the separate and combined effects of American involvements in wars (the Spanish-American War, World Wars I and II, and the Korean and Vietnam Wars), their postwar periods, and the 1930s depression on economic, social, and political forms of American violence from 1890 to 1970. After establishing historical evidence for the role of national accommodation, I demonstrate that strong, positive associations between severe crises and domestic violence are to be found during the tenure of nonaccommodating administrations. Accommodating governments are associated with either negative or historically weak linkages between severe crises and domestic violence. Overall, the evidence underscores the benefit of using broad theoretical perspectives for understanding the linkages between international and domestic conflict.

An ethical defense of the welfare state may, as Robert E. Goodin has argued, adumbrate the notion that responsibility to provide for “needy strangers” flows from the same source as family responsibility. But this view of the moral responsibility for protecting vulnerable people is open to challenge. In this Controversy, Jeffrey Obler takes issue on this point, and Goodin replies.

In this research note I seek to determine whether a significantly predicting social background model for analyzing the votes of Supreme Court justices is time-bound. I argue that an affirmative result poses serious questions for past uses of such models, none of which has controlled for the possibility that time is a confounding variable. A model that significantly predicted the votes of the justices in the Court's 1903–1968 terms was constructed. Analysis with this model for two periods—from 1903 to 1935, and from 1936 to 1968—established that the model was not timeneutral. Appropriate theoretical implications are drawn.

We propose measuring group support for political parties by means of multivariate techniques that have become standard in other areas of political behavior. This approach yields improved insights into the marginal difference made by membership in each group and into the nature of a party's support coalition. As an example of this approach, we analyze the Democratic coalition since 1952. Our results differ from those of previous studies in a number of ways. Most significantly, differences with respect to the strength and timing of partisan changes lend support to Carmines and Stimson's conclusion that a realignment centering on race occurred in the mid-1960s. Our findings also indicate that the Democratic party is no longer so dependent on a few groups, as it was in the 1950s, but is now almost equally dependent on six groups.