The Data
Majority Black Detroit's 36th district court has one of the busiest eviction dockets in the country.
Majority Black Detroit's 36th district court has one of the busiest eviction dockets in the country.
The Eviction Machine houses Detroit's first comprehensive, real-time process dataset. We track cases from start to finish – no similar data exists elsewhere in the state or for the most part, anywhere in the US. Gathering this data enables us to shed light upon Detroit's harmful eviction crisis so that we can sharpen our tools to stop it.
The Eviction Machine houses Detroit's first comprehensive, real-time process dataset. We track cases from start to finish – no similar data exists elsewhere in the state or for the most part, anywhere in the US. Gathering this data enables us to shed light upon Detroit's harmful eviction crisis so that we can sharpen our tools to stop it.
We retrieve case records from the 36th District Court’s online Register of Actions (ROA) case inquiry system and partner with Data Driven Detroit (D3) to translate case-level information into a quantitative dataset. We link eviction records to property assessor and Certificate of Compliance data from the City of Detroit Open Data Portal.
We retrieve case records from the 36th District Court’s online Register of Actions (ROA) case inquiry system and partner with Data Driven Detroit (D3) to translate case-level information into a quantitative dataset. We link eviction records to property assessor and Certificate of Compliance data from the City of Detroit Open Data Portal.
We define residential properties based on assessor class and use codes and exclude cases where the defendant name indicates an LLC or other business. To include mixed-used multi-unit buildings that may be coded commercially, we use a dataset of multi-family properties compiled by students at the University of Michigan Dearborn.
We define residential properties based on assessor class and use codes and exclude cases where the defendant name indicates an LLC or other business. To include mixed-used multi-unit buildings that may be coded commercially, we use a dataset of multi-family properties compiled by students at the University of Michigan Dearborn.
We analyze outcomes (e.g., judgments, orders of eviction) by case, not individual defendants.
We analyze outcomes (e.g., judgments, orders of eviction) by case, not individual defendants.
We include cases brought against the same tenant and/or address on different dates.
We include cases brought against the same tenant and/or address on different dates.
To ensure the highest possible data quality, we confirm random samples of case records manually.
To ensure the highest possible data quality, we confirm random samples of case records manually.
We cannot account for court-recorded errors or records entered after the case record is retrieved.
We cannot account for court-recorded errors or records entered after the case record is retrieved.