Representation in Congress provides a theory of dyadic policy representation intended to account for when belief sharing, delegate, responsible party, trustee, and 'party elite led' models of representational linkage arise on specific policy issues. The book also presents empirical tests of most of the fundamental predictions for when such alternative models appear, and it presents tests of novel implications of the theory about other aspects of legislative behavior. Some of the latter tests resolve contradictory findings in the relevant, existing literature - such as whether and how electoral marginality affects representation, whether roll call vote extremism affects the re-election of incumbents, and what in fact is the representational behavior of switched seat legislators. All of the empirical tests provide evidence for the theory. Indeed, the full set of empirical tests provides evidence for the causal effects anticipated by the theory and much of the causal process behind those effects.