This project develops measures of representation that may be affected by redistricting.
- We reviewed existing literature to map possible connections between redistricting and various concepts of political representation. Early in the project, we developed some initial research questions--these research questions are evolving as the project evolves.
-
We have developed an original dataset on legislator behavior based on all speeches in the congressional record from 2007 to 2020. This includes approximately one million speech units, which we are coding for different kinds of legislator behavior, including partisan rhetoric and attention to one's district, and other behaviors that may be affected by district characteristics.
-
We have merged our new measures of legislator behavior and representation with other datasets, including data on district partisanship, demographics, and community connectivity developed in other parts of the UW2020 project, as well as external datasets on legislators' demographics, voting records, and reëlection margins. These combined data will allow statistical models to relate legislator behavior to district characteristics. In the future, we hope that they will also allow us to predict changes in the quality of representation under hypothetical redistricting scenarios.
These data on congressional behavior may also serve as a template for similar analyses of state legislator behavior.