tPP Includes…

Cases are identified through a variety of sources and then assigned to a team of two coders. These two individuals code each case for over 50 variables and then compare their answers. Any discrepancies are negotiated and decided upon using the Codebook. The cases are then verified by a third coder, who checks to make sure all of the variables are correct and that there are no inconsistencies. Finally, the case is validated by one of the tPP auditors. 

As of September 1st, 2020 tPP’s dataset includes 2,586 completed cases! These cases have been researched by two coders, verified by a third coder, and are in the process of undergoing a final audit. Each case is cross-referenced with a series of court documents, newspaper accounts, and other sources.

We also have over 2,000 cases that are in the process of being coded or have been excluded. These cases include:

  • 691 cases in the process of being investigated and coded by 12 teams
  • 578 cases coded to completion with sentencing details pending
  • 532 cases identified for likely inclusion but not yet investigated
  • 215 cases from a single mass indictment coded and awaiting demographic information
  • 136 cases identified, investigated, and coded for exclusion

Counting it all up, there are 4,738 total cases in the tPP universe so far.

…..We’ve also set aside more than 50 documents and archives (over 2,000 pages of reporting) to scrape for additional cases.

Whenever possible, tPP aims to add and code cases based on primary sources and government documents. Our dataset includes information derived from a variety of such sources including:

tPP researchers have already completed  the cross-referencing, assimilation, & re-coding of cases located in the following databases & reports: