Supreme Court Codebook | Codebook Appendices | Certs Denied Dataset

Supreme Court Dataset Codebook

Harold J. Spaeth's and the Policy Agendas Project's Supreme Court Dataset: A Comparison

If you are currently researching or are pondering a research project regarding the United States Supreme Court, you have most likely encountered the US Supreme Court Databases that were developed by Harold J. Spaeth. If not, they may be accessed at here.

The datasets that have been developed by Harold J. Spaeth, and the dataset that you will find here at the Policy Agendas Project are quite different and both have their strengths and weaknesses regarding the types of research in which they can be best utilized. The Spaeth databases have classified the US Supreme Court cases since 1953 based on the legal principles that were examined in each case. The Policy Agendas Project has classified the same cases based on the policy implications of each case. As a result of these methodological differences, it is very important for the researcher to determine the type of research project he or she is interested in conducting and then deciding which dataset will best meet their needs.

The Spaeth databases and the Policy Agendas Project datasets should not be used interchangeably, and they should not be assumed to generate similar data. Because of the differences in how cases are classified in their respective datasets, the results that a researcher may get from each dataset when charting the behavior of the Supreme Court over time can be quite different. That being said, a researcher who plans to study questions regarding the Supreme Court and policy will likely find the Policy Agendas Project dataset to be the most helpful, while a researcher who plans to study questions regarding the Supreme Court and legal issues will likely find the datasets that have been developed by Harold J. Spaeth to be best suited for their needs.

Cross-coding Analysis

We have conducted an analysis of how closely Spaeth's coding system aligns with our own. All cases in the dataset were cross-coded. For example, if a case fell into Spaeth 221 - Desegration of Schools and PAP 201 - Ethnic Minority and Racial Group Discrimination, the case would be coded in the row and column that corresponded to the appropriate categories. Next, reliabilities between each of the datasets were calculated based on row and column percentages.

We found large disparities between the two coding systems. However, these differences were largely based on the significantly larger number of categories in the Spaeth dataset (261 vs. 225). After consolidating a number of Spaeth codes, the row and column percentages gained greater parity. See below graphical analysis.

Users may download the documentation in Excel format. The spreadsheet is divided into three separate worksheets. The "Raw Numbers" worksheet identifies the number of cases that fall into both the Agendas topics and Spaeth's topics. The "Column %" breaks down the percentage of cases from the Spaeth dataset that fall into the Agendas code along the horizontal axis. The "Row %" worksheet shows the percentage of cases from the Agendas dataset that fall into the Spaeth code along the vertical axis.

Thank you for your interest in the Policy Agendas Project’s US Supreme Court dataset. We hope that you find it to be useful in answering your questions about the US Supreme Court and policy issues.

Written by:

Paul Bryce Counts Jr. , fmr. Undergraduate Fellow, CAPP