Rydberg - Research
  • Home
    • Dissertation
    Posters
    Working Papers

    Abstract: Representation in the American States
    The dissertation examines the effect of institutional design on the quality of representation provided by elected officials to their constituents. I use the American states as a testing ground for the effect of the initiative process. There are two primary contributions of this dissertation. First, the formal model considers how voters will select candidates given that voters can alter public policy through a mechanism that bypasses their elected officials. In addition to a series of tests indicating that voters are less concerned about the policy aspects of candidates in direct democracy states, I even find that voters are more sensitive to the personal attractiveness of candidates in initiative states.

    Second, the statistical estimator used to evaluate representation was derived from the formal model. I use an estimation procedure that simultaneously predicts the ideally representative public policy as well as factors that increase the variance around that estimate. It is through this variance around the ideal policy that the quality of representation is measured. This method is uniquely able to directly compare the effects of initiative process with other institutions, such as state legislative professionalism and term limits, as well as incorporate control variables. Final empirical evaluation confirms that the initiative leads to public policy that is more representative of public opinion.


              State Politics & Policy             Society for Political Methodology
       


    Direct Democracy & Representation: Institutionally Induced Substitution Between Symbolic and Substantive Representation
    Direct democracy allows constituents to directly implement legislation, and thereby eliminate the legislature's monopoly over public policy. This paper extends the canonical model of the initiative's effect on legislative responsiveness in order to incorporate the initial selection of legislators by voters. Given that the initiative provides an alternative route to policy-making that bypasses the legislature (satisfies policy demands through extra-legislative means), the considerations used in selecting candidates are shown to be altered. To the degree that the initiative separates candidate selection from policy outcomes, voters will be more inclined to vote for candidates based upon their personal (or non-policy) characteristics.
    This substitution effect explains three explicit puzzles in the literature. First, economic voting (voters punishing or rewarding legislators for economic performance) is diminished in initiative states. Second, voters in initiative states are considerably less likely to "Vote Correctly" in the sense of selecting the candidate with the most similar policy portfolio. Third, there is superior descriptive representation in initiative states in terms of gender.
    New empirical tests find that voters in initiative states are more sensitive to how attractive candidates look, but less sensitive to how competent they look. Finally, the selection of more honest candidates is demonstrated by lower levels of corruption in initiative states.




    Measuring Representation: Direct Democracy and Policy Responsiveness in the American States (Under Review)
    Institutional design is critical moderate the representational relationship. Previous studies largely validate the link between direct democracy and policy responsiveness by elected officials. Generally concurring with Matsusaka's (2001) critique that modeling responsiveness through interactions with public opinion measures has limited validity; this paper reexamines the theoretical data generating process of these findings and presents an alternative empirical specification for evaluation. This updated method provides a better fit to the underlying theoretical process, as well as permits multivariate controls. I demonstrate the validity of this alternative model with replication of previously published data. Two of the three replications indicate that the initiative results in superior policy representation, and one finding affirms the original null results. All models demonstrate strong confirmation of the validity of the method. A new test with 32 years of data demonstrates greater policy convergence to the median voter as a result of the initiative process.





    Supreme Court Nominations
    Studies of confirmation politics, drawing on the influential research of Cameron, Cover, and Segal (1990), have supported the proposition that it is the combination of a Senator’s ideological proximity to a Supreme Court nominee and that nominee’s qualifications that jointly determines their vote. However, more recent journalistic and scholarly research consistently asserts that political battles over the confirmation of Supreme Court Justices have become increasingly ideological over the past twenty years. Although this changing dynamic is well documented (Epstein et. al. 2006), extant explanations focus on the underspecified importance of the failed confirmation of Robert Bork in driving the observed changes in Senators’ behavior. Drawing upon literature from the Electoral Connection tradition, we specify a concrete institutional mechanism that can account for this ‘Bork Effect.’ Our theory posits that variation in state electoral institutions (particularly the competitiveness of elections) conditions the effect of ideology on an individual Senator’s vote. Specifically we hypothesize that as elections become less competitive, Senators will increasingly be freed from state level electoral constraints. Such freedom allows them to eschew a nominee’s qualifications—assumed to be important to their constituencies—in favor of their own ideological preferences. Expecting that electoral competitiveness, as measured by margin of victory in the previous election, conditions the effect of ideology, we estimate a probit model using just over 2000 Senate votes covering the confirmation of Justice Arthur Goldberg through the most recent addition to the Court, Samuel Alito. By controlling for the interaction between safeness of a Senator’s seat and ideological distance, we can more accurately consider the role that ideology plays in confirmation politics over time, ceteris paribus. Our evidence suggests that the “Bork Effect” can be justifiably considered an artifact of the changing competitiveness of elections.