WARREN E. MILLER AND J. MERRILL SHANKS, (Cambridge, MA: Harvard University Press, 1996), xxviii + 640 pp. $39.96 cloth (ISBN 0-674-60840-2) $19.95 paper (ISBN 0-674-69841-0).
This is an important book by two important political scientists. Miller is of the founding generation of behavioral studies and Shanks is among the first-born products of those pioneers, and here they consolidate more than four decades of their individual and collaborative research and reflections on the difficult terrain of electoral studies. The product is a bold encounter with both the insights and the pitfalls of the behavioral approach in which they reach new plateaus from where our collective vision should become more clear--along the way, exposing some intellectual and methodological flaws that continue to cast confusion among us.
The first third of the book reproduces, and elaborates on, the important work produced by Miller, and Miller and Shanks, from the early 1980s to the mid-1990s on the generational cleavages among the American electorate. The conclusions drawn from this work offer significant steps forward in our understanding of the dynamic elements of political behavior: for one of the rare times in the often anti-septic, formulaic accounts of electoral choice that dominate the literature, we encounter a creative intersection of historical analysis, individual-level behavior and multi-dimensional thinking that is free of the shaky, but mostly unrevealed assumptions of the statistical methods pro tempore.
The division of the electorate into cohorts defined by their time of entry into the active polity, building on the work of Karl Mannheim, Kurt Lewin, and the Michigan school among others (including respectful asides to Paul Lazersfeld and the Columbia school of the 1940s), allow us to nest subsequent behavioral interpretations in a sensible context of individual and collective responses to historical forces, mediated by the more proximate settings and events to which the entire electorate is exposed at particular moments. Beyond age, per se, cohort segments of the electorate can be distinguished in their passage through time by the continuing influence of forces that accompanied their transition into adulthood, and which interact with short-term forces to produce the critical, realigning, deviating, and simply aberrant elections that have become part of our collective lore. Thus, the pre-New Deal and New Deal generations, now rapidly disappearing from...