Pretrial RATs include many inputs about a person’s demographic information to determine risk.
Many view pretrial RATs as objective and technological ways to ensure fairness. By using these demographic data points, however, they often reinforce existing biases and assumptions baked into the criminal legal system.
Common demographic variables include substance use, age, age at first arrest, residential stability, and employment. Many also include questions about mental health and education level.
Using these factors solidifies the idea that being young or experiencing housing instability, substance abuse, or unemployment makes a person more dangerous.
Many of the most commonly used tools utilize more than one such variable, as seen in the chart below and detailed in our subsections about substance use, age, livelihood, housing, and mental health.
Our researchers identified over 315 jurisdictions with a pretrial risk assessment tool in use, and over 300 jurisdictions that listed variables in use.
Over 60% of jurisdictions in our database use at least one variable associated with age, housing, health, or employment.
| Substance Use | Age | Housing | Employment | Mental Health | Education | |
| Number of jurisdictions using this variable | 221 | 210 | 203 | 198 | 78 | 62 |
| Percent of jurisdictions using this variable n=307 | 72.0% | 68.4% | 66.1% | 64.5% | 25.4% | 20.2% |
Source: MAP Research database: includes 307 jurisdictions with known variables, 316 jurisdictions with tools overall
For some of the most widely used tools, demographic and socioeconomic data make up 15% to 67% of the information used to assign risk level.
These heavily weighted demographic and socioeconomic items are often proxies for race and class.
Pretrial RATs rely heavily on these factors, but they all define and weigh each factor differently based on how they define risk. Jurisdictions do not always explain why the tool they use includes particular variables and what kind of risk they are meant to predict.
One study found that few of these commonly used factors are actually strong predictors of pretrial failure.1Kristin Bechtel, Christopher Lowenkamp, and Alex Holsinger: Identifying the Predictors of Pretrial Failure: A Meta-Analysis Factors such as residency and alcohol use were found to actually be insignificant or to predict outcomes in the wrong direction with regard to pretrial failures.
These demographic and socioeconomic factors link to histories of criminalizing youth, poverty, mental health challenges and, implicitly, race.
