Home Law 5. Diversity, Party Identification, and Political Legitimacy
Chapter
Licensed
Unlicensed Requires Authentication

5. Diversity, Party Identification, and Political Legitimacy

View more publications by New York University Press
Diversifying the Courts
This chapter is in the book Diversifying the Courts
735Diversity, Party Identification, and Political LegitimacyDescriptive representation (in which an institution’s demographic makeup mirrors that of the population at large) is thought to increase the level of diffuse support (long-term support) for an institution among formerly under-represented groups. Descriptive/substantive represen-tation (in which a group’s representative shares that group’s political views and race/gender) is thought to increase the level of specific sup-port (short-term support). To account for both of these possible paths toward greater legitimacy levels for the courts, I designed a set of experi-ments aimed at measuring legitimacy. In the descriptive representation experiments, the experimental manipulation involves the aggregate per-centage of Blacks or women on the bench versus their presence in the population. In the descriptive/substantive representation experiment, the experimental manipulation is the race of the judge; it is designed to measure support for a legal decision based on the race of the judge.Almost all prior empirical studies examining the legitimacy of the American government and its institutions have relied exclusively on mass public opinion surveys.1 I, instead, use an experimental approach, which for this type of research is optimal. First, experiments are recognized as the optimal research method for establishing causation, as opposed to studies utilizing mass public opinion, which can only demonstrate correla-tion (McDermott 2002, 334–335; Brewer 2000, 3–16). This causal link re-sults from the random assignment of participants into distinct treatment groups and control groups, leaving no differences between the groups but for the treatment; any difference between the two groups is then pre-sumed to be caused by the treatment (Iyengar and Kinder 1987). In my ex-periments, the treatments are varying percentages of Blacks and women on the bench (descriptive representation) and varying combinations of the races of the judges and participants (substantive representation).The second reason I chose an experimental design was that I could control the information the participants possessed about female and
© 2023 New York University Press, New York, USA

735Diversity, Party Identification, and Political LegitimacyDescriptive representation (in which an institution’s demographic makeup mirrors that of the population at large) is thought to increase the level of diffuse support (long-term support) for an institution among formerly under-represented groups. Descriptive/substantive represen-tation (in which a group’s representative shares that group’s political views and race/gender) is thought to increase the level of specific sup-port (short-term support). To account for both of these possible paths toward greater legitimacy levels for the courts, I designed a set of experi-ments aimed at measuring legitimacy. In the descriptive representation experiments, the experimental manipulation involves the aggregate per-centage of Blacks or women on the bench versus their presence in the population. In the descriptive/substantive representation experiment, the experimental manipulation is the race of the judge; it is designed to measure support for a legal decision based on the race of the judge.Almost all prior empirical studies examining the legitimacy of the American government and its institutions have relied exclusively on mass public opinion surveys.1 I, instead, use an experimental approach, which for this type of research is optimal. First, experiments are recognized as the optimal research method for establishing causation, as opposed to studies utilizing mass public opinion, which can only demonstrate correla-tion (McDermott 2002, 334–335; Brewer 2000, 3–16). This causal link re-sults from the random assignment of participants into distinct treatment groups and control groups, leaving no differences between the groups but for the treatment; any difference between the two groups is then pre-sumed to be caused by the treatment (Iyengar and Kinder 1987). In my ex-periments, the treatments are varying percentages of Blacks and women on the bench (descriptive representation) and varying combinations of the races of the judges and participants (substantive representation).The second reason I chose an experimental design was that I could control the information the participants possessed about female and
© 2023 New York University Press, New York, USA
Downloaded on 9.10.2025 from https://www.degruyterbrill.com/document/doi/10.18574/nyu/9781479818754.003.0009/html?srsltid=AfmBOopWzcrJgjxiw_8ZC3m3g7ehAtr59qoo3ezjXWKbjA5dg0QNSkD1
Scroll to top button