During the Spring 2021 semester, members of the LPEL lab coded bills introduced in the Connecticut, Minnesota, and Georgia State Senates. Bills were downloaded from openstates.org on April 1, 2021. The coding therefore does not include bills introduced after this date.Â
All members of the LPEL lab coded a common set of 400 randomly selected bills and resolutions and discussed criteria for coding. After discussion, coding instructions were developed for coding bills in six content areas: Voting and Elections, Education and Schools, Healthcare, Abortion and Contraception, Policing, and Taxes. Bills and resolutions were coded into categories based on the titles of the bills. Disagreements in the coding of the 400 randomly selected bills were adjudicated through discussion. Lab members then coded the remainder of the bills individually with a 20% overlap to measure intercoder reliability. Intercoder reliability was high (Krippendorff’s alpha > 0.80) for four of the six categories: Education and Schools, Healthcare, Taxes, and Voting and Elections.
Intercoder reliability was lower for the Policing and Abortion and Contraception categories, due in part to the low frequency of bills in these categories. The analyses in the blog posts only focus on State Senate bills and resolutions and do not include State House bills or resolutions. In some states, state senators are allowed to cosponsor State House bills and resolutions. Since State House bills were not coded, they are not included in any calculations of bills (co)sponsored by State Senators in the blog posts.
