Risk assessments are typically comprised of a handful or more factors that are then assigned weights to come up with a risk score.  Assessments, if validated, are only validated to predict risk, but are not validated to set bail and conditions of release.  Thus, a risk assessment is but one more piece of information a judge can use to set the bail and conditions of release.  Most risk assessments are intuitive—they are based on prior criminality and failures to appear.

Risk assessments, when proper safeguards are not involved, may have a negative impact on the rights of defendants.

There is also the question of whether interview of defendants are necessary for the assessments to be successful, and there are several assessments that do not rely on interviews.  Conducting interviews and putting together reports can be resource intensive.
In addition, there has been some particularly harsh criticism by the Attorney General of the United States and others for the use of demographic factors in these assessments that then institutionalize discrimination against minority populations.

Further, to the extent risk assessments ignore required consideration of statutory factors, they may not be employed as the only or sole mechanism to decide release.

Finally, when proper safeguards are in place, a risk assessment can give judges one more view of a defendant that they may not have had and aide in informing a bail decision.

• Risk assessments typically use a numerical risk score.
• Most risk assessments do not distinguish between risk of flight versus risk of a new crime.
• Risk assessments are not validated to set bail—a higher score does not necessarily mean that more supervision, a higher bond, GPS monitoring.