This book brings to completion an ambitious project by Erikson, Wright, and McIver. The book integrates, refines, and elaborates their earlier work on partisanship and ideology across the states. This work presents numerous analyses that make a significant contribution to our understanding of state politics. It is a work that anyone interested in differences across the states should read carefully.

The project was apparently a reaction to the "determinants of state public policy" literature of the 1960s and 1970s, which concluded that socioeconomic conditions are the primary source of variations in state public policy. To assess the neglected role of public opinion, Erikson, Wright, and McIver assembled and merged the national CBS/New York Times polls from 1976-88 to create separate samples for each state. All these surveys asked respondents about their political ideology and their partisan identification.

This merged data set is used to construct a distribution for each state of partisan identification and ideology. These distributions are also converted into mean scores on ideology and partisanship for each state. The authors then create an index of liberalism of state policies and assess the relationship between ideology and public policy across the states. They find a strong relationship between ideology and the liberalism of public policy during the early 1980s. They also find that ideology (or opinion liberalism) has a much stronger relationship to policy liberalism than socioeconomic variables.

The conclusion is an important one for state politics. They find that the policies adopted across the states are strongly associated with the dominant opinions of the public across the states. The broad political process in the states does work to reflect and translate the opinions of the public into policy.

They also address several other important topics about political institutions within the states. They assess the nature of political divisions within the states. They compute the mean ideology by party within each state and are able to assess the extent of mass polarization by state (p. 41). They also draw upon a previous analysis by Uslaner and Weber of state party elites to characterize the extent of liberalism by Democratic and Republican elites and the differences between the two sets of elites (p. 131). Their findings are of fundamental importance to state politics. We have lacked systematic information on the extent of political polarization and division within the states. Erikson, Wright, and McIver find that Republican masses and elites are more conservative than Democratic masses and elites in every state but that the degree of distance between them varies across the states. It has long been acknowledged that we need to have some sense of this polarization to understand party politics in the states. The efforts of Erikson, Wright, and McIver have finally given us our first comprehensive sense of these divisions at both mass and elite levels.

They also track changes in partisanship and ideology over time, using data sets from prior eras and find interesting shifts. One of the more interesting discussions revolves around changes in the role of fundamentalism. They argue that when issues revolved largely around economic opportunity, fundamentalism in the South translated into liberalism. As cultural issues came to dominate the national political agenda, fundamentalism produced a conservative reaction and pushed the South away from liberalism. In the 1930s, fundamentalism was positively associated with liberalism, but it now has a negative association with liberalism.

While the book makes very valuable contributions in charting ideology and partisanship differences across states, it falters when the analysis moves to using the data to assess the role of the political process. From my perspective, students of state politics have often been too willing to accept and use inadequate data and too willing to infer dynamics from static data. That unfortunately happens here.

With regard to the former problem, in moving to assess the impact of the political process on the connection between opinions and public policy, several problems emerge (pp. 113-49). Governors are excluded because of "measurement considerations" (p. 126). The percentage of the legislature held by Democrats is used as an indicator of Democratic legislative strength, even though we are all aware (and their data document) that the South is the highest on Democratic control, but the lowest on liberalism. Since there is a clear equivalence problem across states as to the meaning of party labels, it is unclear why such an indicator would be included in the analysis. In general, it is not clear that their broad, cross-section data on masses' and elites' attitudes are appropriate or valid indicators of the political process within states. We know nothing, for example, of the actual positions taken by gubernatorial and legislative wings of parties, which party controlled government, the negotiations between parties, and the effects of particular historical situations in each state. These broad validity questions are largely ignored.. Gathering additional information would, of course, be a huge undertaking, and their efforts and accomplishments in this project are already considerable and commendable. The point is that it might have been more appropriate not to take such data beyond representing broad state differences to presume the data capture the political process. The authors' belief in their own data goes so far that they explain differences in their results from prior case-study time-series studies by arguing that prior analysts probably confused what they were finding (p. 118). Such a conclusion needs more substantiation. How the political process translates opinion into policy is clearly not a simple matter, and it becomes too simple here.

The second general problem that plagues the analysis of state politics has been that cross-section data are presumed to capture dynamics. This book contains discussions about the Downsian party model, the dynamics of party position taking, the interaction of mass reactions and elite adjustments, and the effects of policy enactment across time. While these theoretical approaches may or may not be of value, the data contained in this study do not capture such dynamics. We can only understand the dynamics of party position taking and elite reactions by examining actual histories within states. The movement to causal diagrams and discussions of dynamics is too uncritical here.

Despite these limitations, it is important to stress that this book makes some very significant contributions to our study of state politics. The authors are to be commended for undertaking this project, for the range of data they have assembled, and for contributing to our understanding of state differences. Finally, it should be noted that this is one of the better written books to appear lately. The writing is clear, concise and communicates very well. Graphs are used well. It is a model of clarity of presentation and writing.