American politics, it is argued, is representative of the forces of American public opinion. That representation emerges over time as politicians and other policy actors sense change in national mood, and respond to it. Public mood is considered to be a vague and general set of orientations toward public policy. Not support or opposition for particular policies, or particular variations in particular policies, it is considered so general that its effects are seen in each of the domestic policy debates of a particular period. Systematic movement of such moods produces political eras, times when the public is willing to go along with particular kinds of policy changes, but to restrict others. The 1960s, for example, is widely considered a time when the public was tolerant toward liberal sorts of policy experimentation across the board, while the (early) 1980s is the reverse case. These ideas have wide, almost universal, currency in political commentary, so much so that many would think their truth too obvious for research. But in political science there is no role at all for moods and eras in prevailing conceptions of public opinion. There it is usually thought that attitudes toward policy particulars arise from the policy debate itself, are specific reactions to specific questions. Or in political psychological approaches they arise because they meet needs of the opinion holder. If moods and eras exist, then because information of this sort is abundant and cheap, it is likely that policy actors would sense them. It is likely also that acutal policy choices would be affected, that for example, those who sensed supportive moods for their positions would take stronger positions. This would produce a kind of over-time representation of a quite direct and meaningful sort. Changes in public preference would lead to changes in policy. This is a conception of "representation" that is a good fit to democratic theory, but quite different from our usual focus on the concept in geographical terms - - districts and representatives. "Mood" as a purely mass phenomenon has been successfully measured in the past by this investigator. The key is the analysis of policy preference "marginals" - - the percent who chose one or another response to a question - - over time and issues. The measure uses the leverage of very large numbers of issues and time points to overcome the inherent shortcomings of the marginals as data and the biases arising from highly selective measurement and reporting. A methodology is developed for measuring the central tendency of time series of such preferences. This project moves on to the next tasks at hand for the study of mood representation, (1) exploring whether mood might have more than one dimension (such as "economic" and "social" components - - widely believed to be the case but without supporting evidence), (2) developing measures of public policy outcomes on a left-right scale over time, in order to (3) examine the statistical link, if indeed it exists, between changes over time in public preferences and changes in public policy.
