The dynamic structure of congressional elections

We address three questions. First we ask to what degree elections to the U.S. House of Representatives may be understood to reflect the influence of current public opinion? This requires us to entertain models of House elections with incumbency effects as a dynamic system. We ask, second, how do the effects of exogenous influences of all sorts work their way into ultimate House composition? Third, we ask if some portion of the effects attributed to incumbency might not be simply these dynamics, reasserted effects of previous elections, in addition to the well-known advantages individual incumbents accrue? We estimate pooled models of state-level aggregate voting for the House. We add Policy Mood to the usual set of explanations as a means to specify the central focus of democratic theories of elections. We focus on how incumbency translates contemporaneous effects of the explanatory variables into dynamic effects and on how this translation differentially affects the various inputs to the system.