Measuring Representation at the Arizona Legislature

Political scientists have long recognized the importance of representation as being critical to the American form of government. This paper begins by establishing the theoretical basis for representational studies in political science. It then turns to the empirical question of measuring the degree to which legislators represent their constituents. While there have been comprehensive studies of representation among members of Congress, several practical difficulties have prevented similar studies at the state legislature level. Underlying measures of representation are the preferences of constituents and the representational behavior of legislators. This paper proposes two metrics, a modified Partisan Voting Index and the American Conservative Union State Legislative Ratings, to compare the district-level preferences with roll call voting behavior (a form of substantive representation) of the corresponding legislators. This methodology is then tested against data collected from the Arizona legislature, specifically incorporating election returns from 2012, 2014, and 2016, and votes taken during the 2017 and 2018 sessions of the legislature. The findings suggest a strong relationship between the partisan and ideological leaning of a state legislative district and how conservative or liberal a legislator’s voting record is, even when controlling for political party. Two special cases are also examined in the context of this data. First, the three legislative districts with a split delegation are analyzed and determined to have a moderating effect on representative behavior. Second, five mid-term vacancies are examined along with the process for filling a vacancy. The paper concludes with a discussion of how this line of research can better inform legislators and help constituents hold their elected officials to account.