Citizens, Contexts, and Politics
Robert Huckfeldt and John Sprague
Politics is about winners and losers, influence and coercion, exchange and bargaining, coalitions and factions, conflict and compromise. All these topics involve individuals and groups tied together in complex relationships that defy easy disaggregation and reaggregation. Yet, when we address the topic of citizen politics in the mass, the temptation appears overwhelming to shift the level of understanding and analysis to that of independent individuals -- individuals abstracted from time, place, and setting. Indeed, opinions, interests, preferences, attitudes, beliefs, and values are readily defined with respect to individually defined circumstance. Rich people are Republicans, black people are Democrats, educated people participate more, and so on, but such an analysis frequently lacks the capacity to reconstruct a compelling account of political life. Contextual analysis provides one antidote to this common analytic disjuncture between individuals and politics -- to the gap in our understanding between micro and macro analyses of political life. Contextual theories of politics are built on an assertion of behavioral interdependence: the actions of individual citizens are to be understood as the intersection between individually defined circumstance and the circumstances of surrounding individuals. The distinguishing irreducible element of a contextual analysis is that, in addition to measures of individual properties and preferences, the political behavior of individuals is characterized as contingent on the environment. Measurements on the environment, as well as theoretical arguments based on the environment, occupy fundamental positions in the logical structure underlying theories of individual political behavior that appeal to explanatory contextual hypotheses.
A number of consequences follow from this thesis. Contextual theories of politics are inherently multi-level -- they require cross-level inference -- and hence have consequences for the ways in which politics is conceived at multiple levels of analysis and meaning. First, this means that the political choices of individuals are best and most fully understood in relationship to the surrounding environment. But, second, it also means that politics in the mass is not simply an additive consequence of individually discrete interests and impulses. Rather, mass politics is understood as the end product of these intersections between groups and individuals within a particular time period and a particular place.
This paper presents a comprehensive view of contextual analysis. We elaborate the idea of behavior in context as an explanatory concept, as well as the intellectual roots that give rise to multi-level contextual analysis. Contextual analysis is construed here as a line of attack upon the more general problem of cross-level inference, and we argue that ecological fallacies arise only when a contextual effect is present. Finally, attention is given to alternative mechanisms of interdependence -- to alternative micro theories of contextual influence.
The modern intellectual roots of contextual analysis can be located in the work of Emile Durkheim, Herbert Tingsten, V.O. Key, and the early election studies of several Columbia University sociologists -- most notably, but not exclusively, Berelson, Lazarsfeld, and McPhee ( 1954), followed in the same tradition by Ennis ( 1962), Segal and Meyer ( 1974), and others. These efforts articulated several themes that continue to provide a focus for contextual theories: behavioral interdependence, multiple levels of observation, and problems of cross-level inference. These themes were set in a somewhat broader context by Harold Lasswell ( 1966 originally published 1939), and Lasswell's relevance to the theoretical aspects of contextual analysis is repeatedly touched on by Eulau ( 1986) in a work that includes systematic development of contextual analysis.
The general topic of contextual analysis is vast and spreads across many fields. This review is selective and focuses on the usefulness of context in the analysis of political behavior. In sociology and particularly in the sociology of education there has been extensive work,