The relationship between electoral systems and corruption in a large sample of contemporary democratic nations is analysed in this article. Whereas previous studies have shown that closed-list proportional representation is associated with greater (perceived) corruption than open-list PR, it is demonstrated here that this relationship fails to hold once district magnitude is considered. The theory underlying this study draws on work on ‘the personal vote’ that suggests that the incentives to amass resources – and perhaps even to do so illegally – increase with district magnitude in open-list settings but decrease in closed-list contexts. Extending this insight, it is shown that political corruption gets more (less) severe as district magnitude increases under open-list PR (closed-list PR) systems. In addition, once district magnitude exceeds a certain threshold – the estimates here are that this is as low as fifteen – corruption is greater under open lists than closed lists. Only at small district magnitudes (below fifteen) is closed-list PR associated with more corruption, as conventionally held. These results hold for alternative measures of corruption, for different sets of countries analysed, for different measures of district magnitude and regardless of whether the political system is presidential or parliamentary, and of the number of parties.

Using an objective measure of corruption in public works contracting, corroborating evidence is also presented from Italian electoral districts. In Italy's open-list environment in the period prior to 1994, larger districts were more susceptible to corruption than smaller ones.

What impact does the negotiation stage prior to the adoption of international agreements have on the subsequent implementation stage? We address this question by examining the linkages between decision making on European Union directives and any subsequent infringements and delays in national transposition. We formulate a preference-based explanation of failures to comply, which focuses on states' incentives to deviate and the amount of discretion granted to states. This is compared with state-based explanations that focus on country-specific characteristics. Infringements are more likely when states disagree with the content of directives and the directives provide them with little discretion. Granting discretion to member states, however, tends to lead to longer delays in transposition. We find no evidence of country-specific effects.

Although a rich body of research has explored the sources of party polarization in the US House of Representatives, it has focused primarily on the House since the late 1970s. Drawing on a dataset of historical election outcomes, legislative voting and survey data, this article takes an alternative approach that examines both the US Senate and the House in their broader historical contexts. It is argued that the unusually bipartisan era of the 1950s created a set of circumstances that enabled congressional parties to remain relatively unpolarized throughout the 1960s and early 1970s. Although the national parties became more ideologically distinct in the mid-1960s, congressional parties lagged behind. As a result, a group of moderate legislators emerged who were cross-pressured between their national parties and their constituencies. Only when natural patterns of electoral loss and retirement replaced these legislators did congressional party polarization re-emerge.

Political scientists have long recognized that the number of parties in a country influences the way that interests are represented in that country. One explanation for the number of parties in a system relies on the idea of strategic voting, i.e. voters may not want to ‘waste a vote’ by voting for a third party. However, work in this area does not address the role of an important factor that may affect party systems through strategic voting: information. Without polls, how could voters know which parties were likely to win, and hence how to vote strategically? Using an agent-based model, this article assesses the role that information plays in shaping the party system through strategic voting. The results of this model demonstrate that, contrary to Duverger's Law, more than two parties may emerge in single-member plurality systems, even when all voters are strategic.

We examine ‘heuristic’ and ‘systematic’ candidate-appraisal strategies within a presidential election context. Controlling for political knowledge, we determine whether individual differences in the capacity for ideological thought condition voters' reliance on the major determinants of candidate choice, increasing reliance on policy considerations and decreasing reliance on the heuristic cue of party identification and on perceptions of candidate character when ideological capacity is high, and exerting the opposite effect – decreasing the role of issues and increasing the role of party identification and candidate qualities – when such capacity is low. Using American National Election Studies data from the 1984–2000 period, we find that ideological thinking consistently heightens voters' reliance on issues and decreases their reliance on candidate cues, but only among voters who report being concerned about the outcome of the election. In contrast, the effect of partisanship is stable across levels of ideological thinking and concern about the campaign. We discuss the cognitive processes by which ideological thinking regulates political choice, and assert its centrality in the political decision-making process.

Political science has not come to terms with the idea of electoral mandates. The discipline's view is a hodgepodge of competing claims. In this article we review the empirical issues about mandates asking whether or not mandates occur and with what effect. We observe evidence of mandates as social constructions, as dialogues in the Washington community and in the press which serves it.We find that these dialogues accurately reflect election results – consensus emerges from actual sweeping election victories and not from mere strategic attempts to claim policy mandates. We find that Congress is highly responsive to the consensus interpretation. It engages in bursts of unusual policy activity which run for some months and then cease. And we find that these have a substantial policy legacy, that the changes that occur in these bursts produce some of the most important movements in American public policy.

This review presents a critical account of the most powerful critique of liberal political thought to have emerged in recent years: a critique it calls the ‘politics of compulsion’. Drawing on the work of a wide range of critics of contemporary liberalism, this article contends that although those who advance this critique are divided in many ways they are nonetheless held together by a series of powerful descriptive and normative challenges to liberal political philosophy as it has developed since the publication of John Rawls's Political Liberalism. The article further demonstrates that most of these challenges centre on the place of coercive power in modern political life and suggests that, although these challenges should not undermine liberals' commitment to their central normative claims, they do nonetheless provide an essential rejoinder to some of liberalism's more complacent assumptions.

Structural limitations in models of representative democracy have enhanced the space for other mechanisms of legitimacy in the European Union, including participatory models in which organized civil society interests are significant players. To some observers, such actors are likely only to aggravate already problematic input legitimacy. A range of less hostile approaches also prevail, from a neutral standpoint through to those sharing the perspective of EU policy practice where such actors are seen as a complementary mechanism of democratic input. Whilst concerns about the impact of asymmetries of power between different types of organized civil society interests arise as potential issues in any democratic setting, a particularly vigorous neo-pluralist regime, in which EU institutions actively create and develop as well as empower citizen interest groups, effectively mitigates these asymmetries in an EU context, although it can give rise to paradoxical tensions of elitism.

In an article published in this Journal, we have examined whether American welfare recipients possess beliefs and attitudes that differ from mainstream public opinion. Although our findings cast doubt on the existence of a widespread ‘culture of dependence’, we do show that programme beneficiaries exhibit self-interested support for government policies that provide assistance to needy segments of the population. However, our analysis – like most others on this topic – assumes that welfare participation affects opinions and not vice versa. This is problematic because there are theoretical reasons to believe that influences flow in the opposite direction, as well. Therefore, some important questions remain to be answered: is it really the case that welfare participation, itself, shapes the outlook of programme beneficiaries? Or do political attitudes affect reliance on public assistance in the first place?

In this Research Note, we examine the underlying structure of the relationship between welfare participation and attitudes towards government activity. Our analysis uses the 1992 Center for Political Studies (CPS) National Election Study, the same dataset employed in the earlier article. But, we test a statistical model that allows for reciprocal influences between welfare and public opinion. The empirical results reinforce the earlier conclusion that welfare participation has an impact on mass attitudes. Conversely, political attitudes have no effect on whether citizens use public aid. Mirroring our earlier work, we find that people who rely on governmental assistance take rational, self-interested positions on the public policy issues which operate to their own direct benefit.

In this research note we have set ourselves the following three principal objectives. First, we show that the well-known concept of consensus democracy, which covers various forms of the division of power, involves analytical problems. Confusion may arise when relating consensus democracy to government action, because the institutions subsumed under the broad concept of consensus democracy, such as executive power-sharing, the multiparty system and federalism, are likely to have different and contradictory effects on the size of government. In this vein, we provide considerable evidence that different aspects of consensus democracy have contradictory effects on government size. In doing so, we endorse the view that it is only variance in the type of democracy (majoritarian versus consensus democracy) that causes systematic differences in government action. Secondly, in scrutinizing the contradictory effects of various aspects of consensus democracy on government size, we distinguish and operationalize the three different analytical views of Crepaz, Lijphart and Tsebelis on how political institutions may be distinguished with regard to their veto nature. Thirdly, we try to close a gap in understanding comparative politics, by quantifying and comparing the veto potential of direct democracy. International comparative investigations of the effects of direct democracy on public policy are hardly possible. The Swiss cantons present themselves as a suitable alternative source of evidence, given that they vary considerably with respect to their plebiscitary elements.

To understand why consensus democracies have a contradictory effect on government size, it is crucial to bear in mind that there are two separate dimensions of the majoritarian–consensual contrast.

What motivates citizens to vote or abstain from voting? What determines vote choice? Models of turnout predominantly focus on a series of factors that affect perceptions of the benefits and costs of voting. The economy's influence on those perceptions has gone largely undeveloped. Models of vote choice routinely account for economic influences while ignoring turnout. This article presents a comprehensive analysis of the role of the economy as mobilizer and demobilizer. I argue that the economy has an impact on turnout decisions and that its impact is asymmetric. All else equal, the incumbent party is able to mobilize more supporters, relative to the non-incumbent party, in good economic times and more supporters are demobilized in adverse economic times. However, Democrats and Republicans are not sensitive to the same elements of economic performance. This is in part a function of the way in which the economy influences perceptions of the closeness of the impending election and of the candidates and their parties. Analysis of pooled data from ANES surveys for the twelve presidential elections from 1956 to 2000 confirms the validity of these arguments.

Comparative scholars have disagreed for some time now as to whether democratic institutions in a divided society are more likely to remain stable if those institutions are premised on a concern for inclusion or on a concern for moderation. But since the empirical evidence marshalled by such scholars is often open to interpretative dispute, neither side has been able to prove its case conclusively. In order to help move this stability debate forward, this article demonstrates how inclusion and moderation can be recast as co-requirements of an underlying principle of political equality. To this end, it offers a deliberative democratic account of political equality, expressed in terms of requirements of publicity and reciprocity, that enables us to see how inclusion and moderation might be reconciled. Moreover, it shows how this deliberative reconciliation may itself provide for a more effective form of institutional stability than can be achieved under either of the two main contending comparative approaches.

Americans not only bowl less today than they did fifty years ago, but also some bowl more than others. This is one of the major and simple messages of Robert Putnam's influential study of social capital in America. Using a variety of data sources, Putnam documents a significant variation in the states' levels of social capital, while arguing for specific general causes of the decline of social capital across the United States. Here, we evaluate the power of Putnam's theory in explaining state-level variation of stocks of social capital. We find that the strongest determinants of social capital levels are basic social and economic differences between states, such as education, church membership, farming and unemployment. Controlling for these determinants, we also find no evidence for a much-debated link between diversity and social capital.

Since the publication of Putnam's book, a growing quantitative literature on social capital has contributed to a much more nuanced and theoretically precise understanding of the link between social capital and the quality of American democracy. Pamela Paxton, as well as Dora Costa and Matthew Kahn, have re-examined Putnam's finding of the aggregate decline in social capital in the United States since the 1960s. Putnam's claim that higher levels of social capital improve the functioning of democracy on the state level has been examined systematically by, among others, Stephen Knack and Tom W. Rice. On a methodological level, Eric M. Uslaner has argued for a need to disaggregate different concepts of trust, and focus on generalized social trust and its effect on making democracy more effective.

Whatever its limitations, the academic literature has not been short in supplying adjectives to describe and characterize the political economy of New Labour. Yet until Ben Clift and Jim Tomlinson's provocative and imaginative intervention in the debate published in this journal, ‘Keynesian’ was not one of them. In a way it is odd that New Labour's political economy should not previously have been labelled Keynesian – it has, after all, been called almost everything else and, as Clift and Tomlinson show well, there is a certain intuitive appeal to this designation.

In this brief response, however, I seek to demonstrate that the label ‘Keynesian’ is ultimately inappropriate, the resulting characterization misplaced. Despite their pretensions to normative neutrality, I suggest, Clift and Tomlinson's self-evident desire to label New Labour's political economy ‘Keynesian’ leads them to mischaracterize both the literature they critique and the putative object of that literature's and their own analytical attentions. I argue that whilst they do largely succeed in showing that New Labour has not ‘decisively repudiated’ Keynesianism, they do not show that it has actively embraced it. Indeed, even if one were to accept every aspect of their analysis, at best they show that the decision to cede operational independence to the Bank of England has bought consecutive New Labour governments since 1997 the fiscal policy discretion to respond – if, as and when the need arises (or is seen to arise) – to recessionary pressures in a Keynesian fashion. In short, they make a far stronger case for suggesting that New Labour's policies since 1997 are not incompatible with future Keynesianism than they do for suggesting that those policies are themselves Keynesian in intent or content.

In a recent article, Steven Lukes offers a thought-provoking reflection on the apparently growing resort to torture by liberal democracies today. Professor Lukes aptly asks whether ‘torture is just another case of dirty hands in politics?’ – that is, the idea that in order to do the right thing or achieve the best public outcome in the circumstances one cannot avoid committing a wrong, such as deceit or cruelty. His answer is that torture differs from other cases of dirty hands in that it cannot be made ‘liberal-democratically accountable’, thus raising the question of ‘how should it be addressed in liberal democracies?’ Here he suggests we can learn from the sociologist Emile Durkheim and, especially, his conception of modern societies as being held together by a ‘religion of individualism’. Allowing state officials to violate basic rights of the individual thus profanes against the religion and threatens the ‘moral disintegration’ of society itself. Lukes argues that Durkheim's account helps us see both the folly in ‘certain ways of thinking and talking about torture’, such as construing it as a ‘lesser evil’ given the competing claims of risk and security, and the damage that liberal democracies inflict on themselves, in responding to the threat of terrorism, by so readily compromising their own principles.

There is much that is appealing in Professor Lukes's remarks. I also share his concern about the current erosion of civil liberties in liberal democracies. I wish, however, to raise some doubts about three key aspects of his argument: that torture is not susceptible to liberal-democratic accountability; that the ‘Durkheimian argument’ effectively precludes the practice of torture; and that we cannot rightly speak of torture as a ‘lesser evil’.

Do elites with scientific expertise hold similar policy preferences in technical policy areas irrespective of their ideological and national background? It is expected that elite beliefs will exhibit a higher level of ideological constraint than mass beliefs do, yet we do not know much about the role of ideology and national context in shaping more specific policy preferences. In this research note, we report the findings of an analysis of the influence of ideology and national context on the policy beliefs of scientific elites in the member countries of the European Union and in the United States.

Recent research on epistemic communities, or ‘networks of knowledge-based experts’, suggests that scientific elites will achieve transnational convergence on policy positions, particularly over complex and technical policy questions: ‘to the extent that epistemic communities make some of the world's problems more amenable to human reason and intervention, they can curb some of the international system's anarchic tendencies, temper some of the excesses of a purely state-centric order, and perhaps even help bring about a better international order’. Nuclear security issues and environmental issues are policy areas identified as amenable to an epistemic communities approach. Earlier research is divided over the degree to which ideology influences scientists' policy beliefs. But there have been no systematic cross-national analyses of the influence of scientific training in suppressing ‘state-centric’ and ideological concerns in the determination of policy preferences.

With a substantive focus on scientific elites' beliefs and preferences on security and energy policy issues, our aim is to extend this research to a comparative setting and provide a preliminary understanding of the extent to which knowledge and scientific training provide insulation from ideology, patriotism and ‘state-centric’ concerns.

The Editors and publishers apologize for the lack of distinction in Figures 2 and 3 printed on pages 255–6 of this volume of the Journal, and offer the following as replacements:

I am really grateful to Geoffrey Brahm Levey for his critical comments, which push me to try to make clear, first, why I claim that torture is not susceptible of democratic accountability in liberal democracies; secondly, where I believe the residue of deep insight in the Durkheimian argument lies; and thirdly, why I think we cannot rightly speak, or even whisper, of torture as a lesser evil.

Torture and democratic accountability. Levey argues that torture can be rendered collectively accountable and that it is not different from other cases of political dirty hands on the democratic front. It could be subject to judicial oversight, for example, by ‘torture warrants’ in the manner of Dershowitz, or by the government or relevant minister being held accountable after the fact. He adds that states do not conceal torture because it is ‘inherently anti-democratic or unpopular’; on the contrary, today it appears ‘only too popular and open to democratic endorsement’.

But I do not mean to equate democracy and popularity. Democracy is not simply majority rule; majorities can tyrannize over minorities and individuals, who need protection on democratic grounds. ‘Democracy’ names an ideal and we could either, like Schumpeter, revise the ideal in a ‘realistic’ direction or, like Dahl, while retaining the ideal as background, characterize real political systems as approximations to it, using other terms, such as ‘polyarchy’. Either way, one of the things which makes a modern state more rather than less democratic is the extent to which it incorporates institutional arrangements and mechanisms that secure a range of basic rights, protecting all, especially the most vulnerable, from arbitrary abuses of power.

Colin Hay's combative response is a welcome engagement with our ideas. Let us first set out where the battle lines are drawn. The most important points in Hay's shot across our bows relate to accommodating the notion of complexity; defining Keynesianism; invoking Labour's past; the relationship between fiscal and monetary policy; and the evidential basis for our argument.

The first point involves admitting complexity when characterizing a political economy. In essence, we agree with Hay about the need to admit complexity into the explanation and characterization of New Labour's political economy, but he does not recognize in our account that this has been successfully achieved, nor do we see it in how he characterizes the relationship between New Labour and neo-liberalism. This is in part a disagreement about how to operationalize the insight about the need to admit greater complexity into the analysis of the political economy of New Labour's macroeconomic policy making. For example, our account of complexity accepts that, under a particular set of monetary policy conditions, which are not in any straightforward sense Keynesian, it is nevertheless legitimate and useful to consider the ‘Keynesianness’ of the fiscal policy regime. Hay, by contrast, sees this as a dubious manœuvre, no doubt fuelled by the unspecified ‘normative bias’ he detects in our work and refers to enigmatically but repeatedly.

Similarly, we recognize that Hay has identified the need to admit complexity into analysis. However, in our reading of his work there remains a sense that, in the final analysis, New Labour's is a neo-liberal political economy.