men it appears, were engaging in their own limited program of
preventive detention.
The appearance record of those for whom bonds were writ-
ten has been exemplary. Mr. Lewis reports only one non-appear-
ance of those for whom he wrote bonds.
Cash Bonds. Little more is to be said about the problems in
obtaining release on cash bond beyond the information in the In-
terim Report (p. 89). There it was pointed out that either cash or
a certified check was required; that from Friday evening through
Monday morning the banks were closed, inhibiting the ability of
defendants or their relatives to obtain the necessary cash from
their bank accounts, if they had bank accounts. Finally, it ap-
pears to be police procedure to sequester arrestees' personal prop-
erty at the station house; hence defendants who might have had
enough cash in their possession to post bond immediately were
detained until an attorney or a relative could, amid the difficul-
ties of the disturbance, go to the station house, reclaim the de-
fendant's wallet, return to court and post the necessary deposit.
Bond Reviews. An essential part of the policy of restricted
release recommended for the judges of the Court of General Ses-
sions was provision for prompt review of money bonds. 28
As it began to appear that the disorder was waning, Chief
Judge Greene designated two judges of the court to hear all bond
review motions beginning on Monday, April 8. (The procedure
provided under the Bail Reform Act is that the judge who origi-
nally sets the conditions of release, unless unavailable, must hear
28 Such review, 18 U.S.C. Sec. 3146(d), is usually referred to as "24-hour review"
because of the requirement of a lapse of 24 hours after the initial conditions of
release have been set before a petition for review may be filed.
49
the 24-hour review petition; attorneys of course could and some
did resort to the original judge, but the special procedure provided
additional judicial manpower to speed disposition of motions
made by defendants who had been remanded to custody.) As a
matter of fact, a number of informal bond review hearings on
oral motion were held, sometimes in chambers, during the course
of the weekend. By such informal review one defendant (of the
413 whom release information is available) seems to have ob-
tained release on personal recognizance from the D.C. Jail on
Saturday, April 6, and five on Sunday, April 7 (Appendix C).
Judges who reported in interviews reviewing bonds either on
motion or at preliminary hearings stated almost uniformly that,
with order being restored and more information available on com-
munity ties, they felt that they could be, and were, more lenient
than during the riot weekend. It has been impossible to obtain
reliable figures on the bond review proceedings themselves. The
Interim Report (p. 91) shows that 85 bond review petitions were
formally filed; others were made orally before the two designated
judges. As the Report goes on:
"Student observers of 40 bail review hearings on Wed-
nesday and Thursday, April 10-11, reported that in 17 cases
petitions were denied; 1 1 were released on personal recogni-
zance or in third-party custody; 7 had bonds reduced or 10%
deposit allowed; several were passed for further verification."
The records of the D.C. Jail show that of the 413 out of
449 prisoners for whom information was available, 96 were re-
leased on personal recognizance from April 8 through April 1 1 :
22 on April 8, 40 on April 9, 24 on April 10 and 10 on April 1 1
(Appendix D). It is unknown how many of these releases were
the result of formal bond review hearings, because preliminary
hearings began to be held on those same days and the judges used
those hearings as a bond review device. A number of defendants
in custody on money bond may have obtained release on personal
recognizance at their preliminary hearings. Moreover, as noted,
some lawyers did not report to the two judges expressly desig-
nated to hear bond reviews but presented bond review motions
to the judge who had set the initial conditions of release. Hence,
statistics of bond actions by the two designated judges, even if
available, would be an unreliable indicator of bond review results.
Transcripts of only a handful of bail hearings before the two
judges could be obtained of three hearings before Judge Murphy
and two before Judge Hyde on Monday, April 8. In two of the
50
three hearings before Judge Murphy, the Judge insisted upon a
third party custodian before consenting to change money bond.
As he said, "My general rule, across the board, except to unusual
circumstances, is on looters that I have to have a third party cus-
todian." Hence in one case where a relative was in court, the de-
fendant was released; in the other, where the defendant's mother
was ill and could not appear, change in conditions was denied.
(The defendant eventually obtained release by posting cash bond
on April 11.) In the third case the Judge refused to change condi-
tions of release because the defendant was on parole for a homi-
cide conviction.
Judge Hyde released two defendants in the third party cus-
tody of relatives. The proceedings show some of the difficulties
still present in obtaining release. In the first case the attorney re-
ported to the court that despite having visited the Police Depart-
ment earlier, he had been unable to obtain any verification of the
defendant's statement that he had no prior criminal record.
Then, after ordering release, the judge advised the defendant's
counsel of the difficulties that might be involved in effecting re-
lease after issuance of the order changing conditions:
"THE COURT: When this can be processed I don't know.
The clerical facilities of the Court are absolutely inun-
dated. What had to be done, of course, is that this
notification of this action will be communicated to the
Clerk's office. We will make a notation on the copy of
the application that you have filed. We will make up a
release or at least we will first, what has to be done is
that the what they call 4 come-up' order has to be made
out just to have him brought up here from the Jail.
And then after he is brought up he has to appear before
us to take his oath on his personal recognizance. And
then we have to sign a release order in order for him to
be finally released. . . .
... I hope that this can be processed so that he can be
out of there within the next day or so. ...
"[COUNSEL]: Is there anything that counsel can do?
"THE COURT: No. Unfortunately, there isn't anything
you can do. My clerk here will have to see that the
thing gets over to the Clerk's Office, and we will sign
the necessary papers. We can do everything we have to
51
do right here within a few minutes. But how long it
will take to process it, I don't know. . . ."
The judge made a similar comment to the attorney in the
next case before him-and, the attorney reported, because of the
administrative delays that the judge predicted, the defendant,
whose third party custody order had been signed on Monday, was
not brought back to court and released until Wednesday.
In any event, by the end of the day on April 10, out of 449
defendants remanded to custody, no more than 172 (and possibly
as few as 136) remained in custody. 29 By April 25 this had been
reduced to 67 (Interim Report, p. 91) and by July 26, 1968, to
16.
These 16 cases were the subject of a study by the American
Civil Liberties Union released on September 10, 1968. The rea-
sons for the continued incarceration of these 1 6 people illustrate
some of the inequities in a system whether adopted for emer-
gencies or in effect in normal times that relies on money bail. A
number of them had been neglected by their attorneys; one had
never had an attorney appointed for him. Vigorous representa-
tion and the pursuit of bail review procedures might have freed
these prisoners. In one case the Bail Agency on May 16 had
not had time to check into the case. In one a defendant, in jail
on a $500, 10% order had remained in custody for over four
months for want of $50; and in one case a defendant with strong
community ties, for whom a $1,000, 10% order had been set,
was denied personal recognizance by a District Court judge who
felt that his inability to raise $ 1 00 in cash showed him to be a
poor risk.
Statistical Summary 30
How Many Defendants Went to Jail Between April 5 and
April 10, 1,616 riot defendants were processed through the Court
of General Sessions. The cases against 235 were dismissed, nolle
pressed or no papered (or in a few cases the defendant was ac-
quitted) at the first appearance, the defendant "walking out." In
41 cases the defendant was adjudged guilty and sentenced forth-
with. Most curfew violators were released on personal recogni-
29 The difference, 36, is the number of prisoners about whom release date informa-
tion is unavailable.
3"The complete figures are in Appendices A and D.
52
zance. Personal recognizance or third party custody release was
ordered in 493 non-curfew cases (43.2%) and money bond (surety
or percentage deposit) was set in 596 non-curfew cases (52.2%).
Of the 1,616 defendants, 449 were actually remanded to
jail. This is 28% of all processed, 33% of all who had bail hear-
ings, and 75% of all who had money bond set.
Subtracting the 449 who went to jail from the 604 (includ-
ing curfew defendants) for whom money bond was set leaves
155, or 26% of all defendants for whom money bond was set
who obtained immediate release because they or their friends or
relatives had funds available and found a bondsman or deposited
money with the court. Adding them to the 664 released on non-
financial conditions leaves 819 (61% of all who appeared before
a magistrate and had bail set) who were released without having
been in custody longer than the time between their arrest and
their hearing.
How Long Did They Stay in Jail. The D.C. Jail records
(Appendix D) contained information about release, indicating
the length of time individuals spent in jail between the date of
commitment and the date of release, for 413 of the 449 persons
who were sent to jail after hearing. (There is no way of comput-
ing accurately how long defendants stayed in jail before a hear-
ing. As Part I suggests, it was not inconsiderable for many de-
fendants.) There was often, however, a time lag between hearing
and commitment. An unknown number of persons may have
spent a day or more in the cell-block awaiting transportation to
jail.
More than half of the 413 prisoners were released after
spending less than three days in jail. 1 5.7% (64) were released
the same day they were committed. 19.7% (89) spent one day in
jail. 1 5.2% (63) spent two days in jail. 1 1 .4% (47) spent three
days in jail, a total of 62.0% of all defendants, or 260 defendants.
Another 24.1% (100) spent from four to ten days in jail,
and 12% (50) spent more than ten days in jail.
How Did They Get Out of Jail. One hundred twenty-four
prisoners (29.9% of 413) posted a percentage bond with the
court. Ninety-seven (23.5%?) made surety bond. One hundred
twenty-nine (30.7%) were released on their personal recogni-
zance. Finally, 63 (15%) were released when their cases were
dismissed, nolle pressed, no-papered, or in a few instances, after
serving a thirty or sixty day sentence.
53
Evaluation of Bail Setting
The Court of General Sessions, like other courts which have
had to deal with civil disorder situations, thought it was necessary
to adopt a restrictive bail policy in the emergency. With part of
the city in flames, the action of the court was understandable in
the circumstances. The judges heard reports, relayed by police
officers and by the United States Attorney, that people being re-
leased were returning to the scene of the disorder. They felt that
this could not be tolerated and that something had to be done.
Hence they set conditions of release of defendants pending trial
based not only on the likelihood of the conditions set to assure
the return of the defendant for trial, but also on the likelihood
of the conditions to prevent the defendant from returning to and
participating further in the riot. Where the conditions set were
money bail, this was expected to lead to the jailing of the defend-
ant for at least a day or more, effectively isolating him from the
disorder.
In the following pages this policy of restricted release and
its implementation is measured first against applicable legal stand-
ards, and then against its results. What benefits, if any, did the
community or the criminal justice system gain from the policy?
What harm, if any, did it do? What lessons, if any, does it teach?
1 . Conformity of the Policy to Law
The Bail Reform Act, in effect in the District of Columbia
last April, provides that conditions of release are to be set in or-
der to ensure that the defendant will return for trial. Various
legal arguments have been offered in support of the policy of set-
ting conditions (including money bail expected largely to result
in incarceration) designed to prevent future misconduct. Con-
formity with legal standards and constitutional principles is an
important measure of the actions of courts and judges. The re-
sults of research into the legal issues raised by the April bail pol-
icy are set forth in Appendix J.
To summarize the issues: It is said that the Bail Reform
Act provision that the court shall take into account "the cir-
cumstances of the offense" supports restrictions on release for
the purpose of preventing future criminal conduct, because the
"circumstances" included the general context of civil disorder.
Under the Act, however, the circumstances of the offense, like
the other items the court is to take into account, bear only upon
the likelihood of the defendant to reappear for trial. Moreover,
54
comparison of the sections of the Act providing for bail after
trial and in capital cases, which expressly allow the court to con-
sider the defendant's potential dangerousness, with the sections
providing for bail before trial in non-capital cases which make
no such provision, and the legislative history of the Act, indicate
that Congress considered and rejected providing in the Bail Re-
form Act for conditions designed to prevent future criminal
conduct.
In any event, the question whether the Act can presently
be interpreted as allowing conditions of release to be set for any
purpose except "appearance of the person for trial" has been
laid to rest by the Court of Appeals in the District of Columbia.
On April 17, 1969, that court held:
"The Bail Reform Act specifies mandatorily that condi-
tions of pretrial release be set for defendants accused of
noncapital offenses. When imposing these conditions, the
sole concern of the judicial officer charged with this duty
is in establishing the minimal conditions which will 'reason-
ably assure the appearance of the person for trial. . . .' The
structure of the Act and its legislative history make it clear
that in noncapital cases pretrial detention cannot be pre-
mised upon an assessment of danger to the public should
the accused be released. " 30a
Supreme Court decisions interpreting the Eighth Amend-
ment and lower federal court decisions both before and since
enactment of the Bail Reform Act have uniformly held, with
one exception, that the only function of bail or release condi-
tions is to insure the return of the defendant for trial. The
exception is where the conduct of a defendant, clearly evidenced
to the trial judge during the course of a trial, indicates that the
orderly progress of the trial and the administration of justice
will be impeded by his continued release. The cases are limited
to those involving tampering with or intimidation of witnesses,
but the existence of this exception suggests that others, includ-
ing the use of bail to keep offenders detained during civil dis-
orders, might be within the constitutional limits. The only test
of the court's policy was an action filed against the judges during
the riot weekend by the American Civil Liberties Union claiming
3Q *United States v. Leathers, U.S. Ct. App. D.C. Cir, Nos. 22,816, 22,818 (April 17,
1969); slip opinion 3-4,
55
violation of the Bail Reform Act. The action was dismissed by
the District Court and the dismissal was not appealed. The dis-
missal is not a precedent on the merits of the claim; it was based
on the ground that the prompt action of the Court of General
Sessions to provide 24-hour review of bail conditions made any
intervention by the District Court unnecessary.
The weight of existing statute and judicial precedent is
against using bail conditions as a means of detaining defendants
to prevent them from endangering the community. The common
law grows, however, by changing to meet conditions not yet pro-
vided for. The succeeding sections examine the evidence for and
against the court's policy of restrictive release.
2. Need for and Utility of a Policy of Restricted Release
The hearsay that triggered adoption of the policy of restrict-
ing release for the purpose of preventing further participation in
the riot, that people released at court were returning to the riot,
announced as a rumor by Judge Edgerton, was repeated in many
forms. One judge expanded the few into "many." One Assistant
U.S. Attorney stated that "a number of defendants" let out ear-
lier had been arrested again for looting (p. 33, supra). Judge Daly
pinned down the rearrests to "five" of "the first twenty-five
people they released on personal recognizance on Friday." None
of this was ever substantiated. A check of those persons who ap-
peared in court on Friday, April 5, shows that only one was re-
arrested that same day. The defendant, after release, scuffled
with a policeman outside of court, was arrested again and charged
with disorderly conduct.
The principal justification given for a policy of restriction
on release during a disorder is that it must be assumed that sus-
pected looters are likely to return immediately to the scene of
the looting to continue their former lawlessness. 31 This has been
offered in justification of the policy of the Court of General Ses-
sions in April. This speculation has prompted the response of a
commentator reviewing release policies in Detroit that it
"ignores the fact that prior to release there has been an ar-
rest with its attendant formalities, a period of detention,
and an appearance before a judge who undoubtedly im-
pressed upon the accused the possible consequences of rein-
31 See, e.g., Comment, The Administration of Justice in the Wake of the Detroit
Civil Disorder of July 1967, 66 Mich. L. Rev. 1542, 1549, 1550, 1565 (1968).
56
volvement in the lawlessness. Such exposure to the criminal
system may be sufficiently sobering for the defendant to
lead him to conclude that for him the party is over. This
would seem to be partially true in regard to those persons
who became involved in the first place only because of the
excitement and novelty of the disorder." 32
The only available empirical basis for assessing the necessity
or utility of a policy of restricted release is the rate of re-arrest of
rioters.
Of the 6,230 civil disorder arrests reported by the Police De-
partment from April 4-15, 193 persons were arrested twice for
any combination of offenses, including two curfew arrests (Ap-
pendix E). This is 3.1% of all persons arrested during the civil
disturbance. Out of 1,604 non-curfew arrests on April 4-7, 46
(2.9%) were rearrested either then or later, their first arrest hav-
ing been for a felony or U.S. misdemeanor. Only 21 were arrested
twice during the April 4-7 period ( 1 .4% of the 1 ,604 arrests) and
1 5 of these were arrested the second time only for curfew viola-
tions, leaving six arrested twice for serious offenses from April
4-7. While, as we have seen, some persons guilty of serious mis-
conduct were charged only with curfew, there remain something
less than 2 1 persons demonstrated as having made a second con-
tribution to disorders. Of these 2 1 , nine had been released on
personal recognizance, seven had money bond set but apparently
made it at court, one appeared in court but there is no record of
what bond was set and five do not appear in the court records at
all.
A low rearrest figure, of course, does not demonstrate that
many persons did not return to the disorder who escaped arrest.
If, as the Washington Post estimated, there were 20,000 rioters,
the chance of being arrested was no more than one in three.
There conceivably were others who were arrested only once but
participated in unlawful activity again. Moreover, a low rate of
rearrest may evidence the perspicacity of the judges in picking
out, for restrictive release conditions, precisely those who would
have been likely to return to the riot.
The figures: total number of rioters, total arrested, total
brought to court and number arrested twice, do demonstrate,
however, how limited is the opportunity of a court even in a ma-
jor disorder to restrain those posing a danger to the community.
32 /d at 1565-66; see also id. at 1574.
57
The 1 ,340 persons for whom the court had the opportunity to
set bail from April 5-10 is only 6.7% of the estimated rioting
population. Further, to restrain the 2 1 arrested twice between
Thursday, April 4 and Sunday, April 7, the "proven" dangerous,
by a blanket policy of preventive detention (which the court did
not adopt) would have required the detention of all the 406 per-
sons who were released at court on non-financial conditions on
April 5, 6 and 7 (Appendix A).
Had the court been faced with more than a handful of insti-
gators of disorder, breakers, arsonists or violent rioters, this fact
would have supported the argument that restricted release was
necessary. But such "hard core" offenders were not before the
court.
Few of the looters were actual breakers. Early in the disor-
der, Judge Edgerton observed: "We are getting a lot of, appar-
ently, hangers-on in this thing; we haven't had many actual perpe-
trators." When the United States Attorney made his argument
for a money bond policy, Judge Edgerton asked if the prosecu-
tion was going to show that the defendant was actually breaking.
When the United States Attorney said no, but that the prosecu-
tion would prove that the defendant was inside or coming out of
a store, Judge Edgerton remarked that "that fits more nearly the
pattern." Only three of 415 transcribed bail hearings (Appendix
G) indicate on the record that the defendant was actually break-
ing into a store; most defendants were seen inside or coming out
of stores already broken into. The transcripts of preliminary
hearings of 158 persons (35% of all defendants who had prelimi-
nary hearings) show only five defendants against whom there was
evidence of breaking. In the aftermath, judges of the United
States District Court thought that most of the cases bound over
to the grand jury for indictment were not cases involving serious
criminal acts. Judges admitted in open court that although they
were following a policy against release they found most defend-
ants to have been "respectable," "law-abiding" people before
their arrest in the disorder, and stated in interviews that most of
the riot defendants seemed to be more stable and reliable than
the normal defendant. The Bail Agency recommended a higher
percentage of riot defendants for non-financial release than it
does for the usual defendant population; and one of the reasons
bondsmen refused to write bonds during the emergency was that
so few of the defendants were "prior customers."
It is apparent that virtually no instigators, fire-setters and
breakers were caught. According to the Washington Post ac-
58
count, 33 arrests were made block by block as order was restored
to that particular area. The dangerous participants had by then
moved on or ended their activity. The hangers-on, people who
joined the crowds in the street to "get something" and failed to
avoid arrest (perhaps because of their relative inexperience at
lawlessness) were caught in the dragnet. If the "hard core" were
deterred and restrained from further criminal activity, it was not
by preventive detention but by the police, National Guard and
Army.
Of the 1 ,340 persons before the court for bond hearings
only 29 were charged with felonies arguably indicating a danger-
ous potential: three for arson, eight for assault on a police offi-
cer, six for assault with a gun or knife or other dangerous weapon,