Relationships between public preferences and government policies are at the heart of both theories of democratic representation and everyday politics itself. The theoretical literature is rich with instances in which exactly this supposition this made - that is, where attention is directed not just at symbolic or descriptive representation in government, but at policy representation (see Dahl 1971; Pitkin 1967). Indeed, it is not clear how to seriously address the existence of representation in government, or a lack thereof, without concerning ourselves directly with policy decisions and actions (Bardach 1981). It is similarly difficult to imagine an election campaign without issues and policy, and politicians clamoring to represent the public will. The representation of opinion in policy is clearly important, both theoretically and practically.

Equally significant is the extent to which public preferences are informed and responsive to changes in both real-world circumstances and public policy (see, e.g., Lippmann 1925; Schumpeter 1950; Wahlke 1971). The representation of public preferences actually requires that the public not only has meaningful preferences but that it notices and responds to what policymakers do. Without such responsiveness, policymakers would have little incentive to represent what the public wants in policy - there would be no real benefit for doing so, and no real cost for not doing so. Moreover, without public responsiveness to policy, expressed public preferences would contain little meaningful information. There not only would be a limited basis for holding politicians accountable; expressed preferences would be of little use even to those politicians motivated to represent the public for other reasons. We need a responsive public - effective representation depends on it. Thus, policy representation and public responsiveness not only are fundamental characteristics of democratic government; the extent to which they coexist is a critical indicator of democratic performance.

Building on Wlezien's (1995, 1996) 'thermostatic' model of opinion-policy relationships, and drawing together directly comparable bodies of data on public preferences and government spending, this project explores the extent and nature of both public responsiveness and policy representation in the UK, US and Canada.

Our comparative focus allows us to introduce characteristics of policy domains and countries as mediators of opinion-policy relationships. We begin with salience. Some policy domains are more important than others. More people care about them and they, thus, are more likely to pay attention to politicians' behavior in the areas. Politicians, meanwhile, are more likely to pay attention to public opinion in the areas. In effect, we expect a certain symmetry between public responsiveness to policy and policy responsiveness to opinion across policy domains.

This is our starting point, and presumes that institutions are not distorting. We argue that they can be, however. For the thermostatic model to work, we not only need a certain level of accurate media coverage and political competition, we need government institutions that make it easy for the public to become informed about what policymakers do and give an incentive to policymakers to represent public opinion. We argue that the division of powers among government institutions has big effects: (1) that vertical division of powers, or decentralization, makes it more difficult for the public to gauge and react to individual governments' levels of policy, and thus dampens public responsiveness; and, (2) that the horizontal concentration of powers, or parliamentary government, makes politicians less responsive to changes in public opinion.

Results speak to the structure of relationships between citizen preferences and government policies across countries, political institutions, and policy domains. And the story that emerges is one in which representative democratic government seems to function surprisingly well. Simply put, the thermostatic model works in each of the three countries. We observe that the public adjusts its relative preferences for spending in response to spending itself - there is negative feedback. We also observe that spending itself follows changes in preferences over time - there is representation. There are important differences in the details, however, and it also appears that democracy works better in some domains and countries than others. We argue that the differences are understandable given differences in the nature of issues and government institutions themselves; indeed, they are perfectly in keeping with our expectations. Our conclusions are thus partly descriptive, partly prescriptive. Coupled with the literatures on democratic theory, comparative political institutions, and policy representation, this work has much to tell us about present and possible future degrees of democracy.