Despite a voluminous literature on the subject of representative–constituency relationships, one question central to that relationship remains underdeveloped. It is: what does an elected representative see when he or she sees a constituency? And, as a natural follow-up, what consequences do these perceptions have for his or her behavior? The key problem is that of perception. And the key assumption is that the constituency a representative reacts to is the constituency he or she sees. The corollary assumption is that the rest of us cannot understand the representative–constituency relationship until we can see the constituency through the eyes of the representative. These ideas are not new. Their importance has been widely acknowledged and frequently repeated ever since. But despite the acceptance and reiteration of Dexter's insights, we still have not developed much coherent knowledge about the perceptions members of Congress have of their constituencies.

A major reason for this neglect is that most of our research on the representative–constituency linkage gets conducted at the wrong end of that linkage. Our interest in the constituency relations of U.S. senators and representatives has typically been a derivative interest, pursued for the light it sheds on some behavior – like roll call voting – in Washington.