Our book is mostly descriptive, not so much giving theories as laying out some facts about public opinion; briefly:

1. The pattern of income and voting varies geographically and demographically in the population. In particular, income predicts Republican voting more strongly in red states and among frequent church attenders, while income is only weakly associated with voting patterns in blue states and among non-attenders.

2. Or, to put it another way, the differences in political attitudes between red and blue America are concentrated among the upper middle class and do not show up so strongly among lower-income voters.

3. These patterns have largely arisen in the past thirty years, during which voting has become more issue-based.

We connected these facts to misconceptions about voting in the news media. In particular, we point out that the major media are concentrated in places such as New York, Maryland, Virginia, and California where individual income is less associated with conservative attitudes and Republican voting.

What I think will be more interest to the Talking Points Memo discussion, though, is not these bare facts but the questions of how these voting patterns arose and what they mean for American politics. We offer some speculations in the later chapters of the book, but, as we say in academia, further research is needed. Some key issues include aspects of political polarization among political officeholders and among ordinary Americans; the extent to which the electoral system favors political moderation; the sources of the political changes that have moved West Virginia from the Democratic to the Republican column and Vermont in the opposite direction; and the ways in which asymmetric bases of support for the two parties influence policy.

Our reason for writing this book was to get some basic facts out there--some facts which have not always been apparent even to well-informed journalists, because understanding them required us to analyze survey and election data more elaborately than is usually done--and to get people thinking about how Republican and Democratic voters differ. I welcome discussion of these issues.