Document Type

Book Chapter

Publication Date



Various studies of the administration of justice during riots establish that persons arrested during riots generally have been detained for substantially longer periods than persons arrested during nonriot situations. This "extra" period of pretrial detention has been the product of quite different practices in different cities, and these practices have been based on several different administrative policies including, but not limited to, preventive detention.

In several cities, extra-lengthy detention resulted from a general judicial policy of setting extraordinarily high bail for the duration of the riot. A prime example was Detroit, where 70% of the bail initially set was in excess of $5,000, and few people gained their release until the judges reexamined their bail decisions and began to release persons on their own recognizance after the disorder had substantially subsided.

Other cities required somewhat lower bonds, but at amounts still extraordinary in light of punitive use of nonriot practices. In Baltimore, for example, high bail bonds for curfew violators were originally set at $500. Although available evidence indicates that these high bonds were set for the very purpose of preventing persons from obtaining their release, courts were not always concerned primarily with preventive detention-that is, detaining persons because of the fear that they would return to the streets and commit crimes following their release. Very frequently, the judges indicated a desire to use pretrial detention in a basically punitive fashion. Judges in several cities suggested that a policy of high bail (in effect, a policy of nonrelease) would tend to deter other persons from violating the law. They felt that the "word" would get around-"if you are arrested, you will automatically spend a few days in jail under very uncomfortable circumstances, so you had better play it safe and stay off of the streets."

In a few cities, extra-lengthy detention was probably an unintended product of following normal bail practices. Judges in these cities set bail at normal levels, but bondsmen were simply unavailable or unwilling to write bonds until matters "cooled off." As a result, many persons arrested in these communities often stayed in jail almost as long as persons arrested in communities following high bail policy.

In other areas, extra-lengthy detention was as much the product of administrative difficulties at the police and prosecutor level as any judicial policy. In Detroit, the prosecutor sought to check the record of each arrested person before any individual bail recommendation would be made. Under this approach, even if local courts were willing to release arrested persons on their own recognizance where they were charged with a minor offense and had a "good" record, those persons might be detained two or three days -before their police and employment record could be checked and such information could be conveyed to the court.

Extra-detention based on administrative difficulties or the punitive use of high bail raises numerous problems that are worthy of treatment in a separate conference on that subject alone. Our primary concern here is with preventive detention, and I mention detention based on these other grounds only to put our problem in its proper perspective, and to indicate the need for extensive control of bail practices during riots irrespective of one's position on preventive detention.

With respect to preventive detention, I hope to make a single point; even if one assumes that preventive detention in the normal situation is both unwise and unconstitutional, it does not necessarily follow that preventive detention should also be rejected in the riot situation. The pragmatic and legal considerations relating to preventive detention during riots are sufficiently different so that a quite distinct, and perhaps stronger, case can be made for at least limited preventive detention during such disorders.


Research indicated this is an orphan work. If you are the rightsholder and want to restrict access, contact