CASELOAD HIGHLIGHTS
E
XAMINING
T
HE
W
ORK
OF
S
TATE
C
OURTS
A Profile of Hung Juries
National Center for State Courts • Brian J. Ostrom, Project Director • Nicole L. Mott, Neal B. Kauder, Brian J. Ostrom, and Paula L. Hannaford-Agor, Authors May 2003
Volume 9 • Number 1
State Felony Dispositions in 30 Counties, 1996-1998
Jury trials are held in open
court, but jury deliberations
can be mysterious, especially
when the jury “hangs.” A
jury “hangs,” or is “dead-
locked,” if jurors cannot
agree upon a verdict: the
result is mistrial.
1
Hung
juries threaten court and
public interests in several
ways. First, they challenge
perceptions about the cred-
ibility and fairness of the
legal process. Beyond issues
of basic justice, jury trials are
expensive and require consid-
erable time and resources
from the judge, court staff,
legal counsel, litigants, and
witnesses. Finally, hung
juries raise questions about
juror abilities to handle com-
plex legal and evidentiary
issues. This issue of Case-
load Highlights profiles the
circumstances surrounding
hung juries.
The National Center for State
Courts (NCSC) recently
completed a national exami-
nation of hung juries, the
first such cross-jurisdictional
study in nearly forty years.
The background of the NCSC
study and methods used are
discussed to the left.
The following figure shows
the manner in which felony
cases were resolved in 30
state courts between 1996
and 1998. Approximately
70 percent of the 1.2 million
cases examined ended in a
guilty plea, about 24 percent
were dismissed or disposed
as an “other” disposition
(e.g., diversion to an alterna-
tive court program such as a
drug court or transfers to a
federal court), and less than
3 percent were resolved by
jury trial. Of the 33,000
cases disposed by a jury
trial, about 1,600, or 6 per-
cent, “hung.” Regardless of
whether resolved by bench
or jury, about three-quarters
of trials end in conviction.
Study Background & Methods
The results presented in this Caseload Highlights are derived
from a 2002 National Center for State Courts’ study covering
32 counties across 14 states and all federal circuit courts in
two separate phases. The project was funded by the National
Institute of Justice with the first results presented in Are Hung
Juries a Problem?
2
The purpose of the study was to conduct
a comprehensive analysis of the factors that relate to hung
juries and determine how often juries hang and why.
In the first stage of the project, NCSC compiled aggregate
civil and criminal data from 1980 through 1997 for all 12
federal circuit courts from the Administrative Office of the
U.S. Courts. In addition, NCSC assembled wide-ranging data
from the state courts and prosecutors’ offices supplemented
with information from annual reports. These data covered 30
courts between 1996 and 1998 on state felony criminal trials.
In the second stage of the study, researchers conducted an in-
depth review of four sites on nearly 400 felony trials. In this
phase, NCSC collected detailed data on felony jury trials that
hung or ended in a verdict. NCSC surveyed jurors, judges,
and attorneys to examine case characteristics and jury behav-
ior across dispositions (hung or verdict) and compare hung
jury rates across jurisdictions using comparable definitions.
2
Paula L. Hannaford-Agor et al., Are Hung Juries a Problem? National Center for State
Courts, at http://www.ncsconline.org/WC/Publications/Res_Juries_HungJuriesPub.pdf.
Total Felony Dispositions: 1,198,338
Convicted
Acquitted
Hung
Other Mistrials
71%
19%
6%
4%
Jury Trial: 32,753
Bench Trial: 51,535
Dismissed: 123,561
Other: 168,120
Guilty Pleas: 822,369
1
A judge may also declare a mistrial due to
prejudicial error or an extradordinary event.
■ How Often Do Hung Juries Occur?
Past reports are scarce, but
suggest considerable variation
in rates of hung juries. In
1966, Kalven and Zeisel re-
ported an average criminal
hung jury rate of 5.5 percent
nationwide of all jury trials.
3
The only other empirical study
of hung jury rates was con-
ducted in 10 California coun-
ties in 1975 by the Planning
and Management Consulting
Corporation.
4
They reported a
12.2 percent hung jury rate,
with significant variations over
Compared to the state rate of
6.2 percent, the federal hung
jury trial rate for criminal
cases averages 2.5 percent.
The federal Administrative
Office of the Courts’ central-
ized data system makes it
possible to examine federal
rates over time. The trend
lines below show fairly
stable rates from 1980 to
1997 for both civil and
criminal trials.
time and between localities.
The first phase of the cur-
rent NCSC study confirms
these earlier findings. The
average hung jury rate across
all 30 sites was 6.2 percent,
with slightly higher rates
ranging from 8 percent to
14.8 percent in 5 of the 6
California counties. The
figure below displays the
individual jurisdictional rates
and shows the variation
across counties.
State Hung Jury Rates (Average 1996-1998)
It should be noted that the
federal D.C. Circuit Court
reports a much higher hung
jury rate of 9.5 percent (a
clear exception to the other
11 federal districts). One
potential explanation for this
difference is the demographic
characteristics of the court’s
jurisdiction; the D.C. court is
the only federal circuit court
that comprises a single,
highly urbanized city. All
other federal circuits encom-
pass multiple states.
Federal Hung Jury Rates (1980-1997)
A Profile of Hung Juries, continued
3
Harry Kalven & Hans Zeisel, THE AMERICAN JURY 453 (1966).
4
Planning & Management Consulting Corporation, Empirical Study of Frequency of Occurrence
Cases Effects and Amount of Time Consumed by Hung Juries, 4-30 to 4-37 (1975).
0% 3% 6% 9% 12% 15
%
Pierce, WA
Oakland, MI
Pinellas, FL
Fulton, GA
Hennepin, MN
Macomb, MI
Philadelphia, PA
Nassau, NY
Monroe, NY
St. Louis, MO
San Francisco, CA
Suffolk, NY
Bronx, NY
Pima, AZ
Queens, NY
Erie, NY
Westchester, NY
Kings, NY
Dallas, TX
Middlesex, NJ
Wayne, MI
Harris, TX
Fresno, CA
Santa Clara, CA
Riverside, CA
New York, NY
Travis, TX
Alameda, CA
Shelby, TN
Los Angeles, CA
0%
1%
2%
3%
4%
19971992198819841980
Criminal
Tot al
Civil
Note: Differences exist in the way courts calculated hung jury rates.
Most courts do not systemati-
cally track hung jury out-
comes in their automated
case management systems.
Many courts do not consider
a hung jury to be a final
disposition, but rather an
intermediate case status. To
overcome difficulties of com-
parisons, NCSC obtained
detailed data on 382 felony
jury trials from Los Angeles,
CA; Maricopa, AZ; Bronx,
NY; and the District of Co-
lumbia.
5
Los Angeles and
D.C. participated in the study
due to concerns over high
hung jury rates, Maricopa’s
participation allowed NCSC
to evaluate recent trial inno-
vations, and the Bronx courts
were solicited, in part, due
to the high volume of trials.
Thus, these four sites were
not intended to be representa-
tive of courts nationwide.
Definitional differences are
one reason hung jury rates
vary. That is, a broader defi-
nition captures higher rates,
while a more restrictive one
A Closer Look at Four Large Courts
yields lower rates. To make a
fair comparison, in-depth
data from the four sites were
examined using three defini-
tional “levels”—determined
if the jury hung on:
1. Any count against the
defendant
2. The most serious count
against the defendant
3. All counts against the
defendant
As seen below, the first defini-
tion yielded the highest rates–
almost 13 percent of the juries
hung on at least one count.
The most restrictive defini-
tion, where the jury hangs on
all counts, occurred in 7.5
percent of the cases. Juries
hung on the most serious
charge facing the defendant in
about 10 percent of the cases.
Hung jury rates varied across
the jurisdictions regardless of
which definition was used.
Rates of hung juries were
highest in the D.C. Superior
Court at about 22.3 percent,
followed closely by Los An-
geles County at 19.5 percent.
Rates were lowest in the
Bronx Supreme Court where
just three of 97 jury trials
ended in a hung jury (3.1 per-
cent). In these three cases,
the jury hung on all counts
faced by the defendant, so the
hung jury rate was identical
across all three definitions.
Hung Jury Rates in Four Sites
5
The courts and corresponding dates for data
collection were as follows: the Central
Division, Criminal, of the Los Angeles County
Superior Court, CA, June–October 2000; the
Maricopa County Superior Court (Phoenix),
AZ, November 2000–October 2001; the Bronx
County Supreme Court, NY, February–August
2001; and the Superior Court of the District of
Columbia, April–August 2001.
Bronx
Maricopa
Los Angeles
District of
Columbia
3.1%
3.1%
3.1%
3.3%
5.1%
7.7%
11.7%
16.2%
19.5%
12.8%
16%
22.3%
All Sites
7.7%
10.1%
13.2%
■ What Characterizes
Hung Juries?
To better understand the
circumstances and character-
istics of hung juries, numer-
ous social and legal factors
were analyzed across the
four selected sites. NCSC
examined demographic fac-
tors such as juror race and
general juror attitudes on
such issues as police, courts,
and crime in the community.
Specific attitudes relating to
the case at hand—law, evi-
dentiary issues, and interper-
sonal small-group dynam-
ics—were also analyzed and
found to be especially criti-
cal in predicting the likeli-
hood of a hung jury.
The legal fairness, eviden-
tiary, and jury dynamic
issues are discussed on
the following two pages.
■ Hung on all counts
■ Hung on most serious count
■ Hung on any count
Legal Fairness
Both popular culture and
recent commentators paint a
vivid picture of a hung jury
characterized by a holdout
juror engaged in race-based
jury nullification. Nullifica-
tion occurs if one or more
jurors intentionally disregard
the law or vote to acquit a
defendant despite the evi-
dence. The adjacent table
shows that although specific
juror intentions (e.g., holding
out due to racial issues or
general distrust of police) are
difficult to ascertain, the
diversity of the jury’s racial
makeup did not predict
whether the jury would hang.
Even though it was not pos-
sible to determine defini-
tively if individual jurors
consciously or deliberately
refused to agree on a verdict
due to perceptions that the
legally correct outcome was
unfair, jurors who hung be-
lieved the outcome was less
fair than did jurors who
reached a verdict. Specific
attitudes of legal fairness
varied, yet there were no
significant differences be-
tween hung and verdict ju-
ries regardless of demo-
graphic factors or general
perceptions of trust of the
courts or police.
6
Evidentiary Issues
Do cases with complex evi-
dence or legal instructions
make it more difficult for
jurors to reach agreement?
Previous research suggested
that closeness of evidence
plays a significant role in
hung juries.
7
The bar chart
to the right shows further
support that evidence was
most ambiguous in hung juries
(i.e. did not favor one side).
Juries that hung on at least
one count noted that the case
was more complex and diffi-
cult than did juries that
reached a verdict (see table
below). Yet interestingly,
Jury Demographics and Attitudes
Percent of Juries that Hang,
by Ambiguity of Evidence as
Assessed by Jurors
7
Supra note 2.
6
The average ratings by hung juries in the
Bronx are to be read with caution, as only
three juries hung in the Bronx.
Does the Perceived Complexity of Evidence Relate to Hung Jury Rates?
Complexity of Case
Verdict Juries Hung Juries
Juror Responses
How complex was this trial? 3.57 4.13
Judge Responses
How complex was the evidence presented at trial? 2.53 2.52
How complex was the law? 2.82 3.20
Attorney Responses
How complex was this trial?
Prosecutor 3.20 3.39
Defense 3.18 3.34
The numbers represent values on a 7-point scale (1= least complex to 7 = most complex).
* The difference between the two outcomes was statistically significant.
LA Maricopa Bronx DC
Hung Verdict Hung Verdict Hung Verdict Hung Verdict
Race/Ethnic Diversity of Jury
a
.73 .71 .35 .36 .70 .63 .51 .53
Juror Attitudes
b
Trust in Police 5.0 5.3 5.6 5.7 4.8 4.6 4.7 4.9
Trust in Courts 5.3 5.6 5.7 5.7 4.8 5.2 5.2 5.3
Fairness of Legally Correct Outcome 4.5 5.8 4.2 5.6 3.9 5.3 4.8 5.2
a
Diversity ranges from 0 = least diverse (e.g., all-white jury) to 1 = most diverse (e.g., one juror of each racial/ethnic category). Categories include:
Black/African American, White/Caucasian, White/Hispanic, Nonwhite/Hispanic, Native American, Asian/Pacific Islander, and other.
b
On a scale of 1 to 7, 7= a great deal or very fair.
*
Most Ambiguous
Somewhat Ambiguous
Less Ambiguous
Least Ambiguous
0%
10%
13%
24%
neither judges nor attorneys
found hung jury cases to be
more complex than verdict
cases, suggesting that judges
and lawyers may consistently
underestimate the level of
case complexity as compared
to the jurors’ perspective.
Generally, judges and attor-
neys rated the jurors’ com-
prehension of legal and evi-
dentiary issues high. How-
ever, once the jury hung,
judges and prosecutors ex-
pressed concern of jurors’
abilities. Of course, attor-
neys and judges provided
the ratings after the jury
declared a verdict, so per-
haps some reflection was
done in hindsight.
A Closer Look at Four Large Courts, continued
In general, jurors tend to per-
sist in their early opinions
about defendant guilt or inno-
cence. More often than not,
the jurors’ first vote in delib-
erations is the same as the
jury’s final group decision.
The graphic below shows that
when jurors first leaned to-
ward acquittal or conviction,
their final decision was likely
to be consistent with their
earlier opinion. However, if
on the first vote jurors were
more evenly split, the case
was more likely to hang.
Although serving on a jury
sometimes bonds jurors to-
gether on a personal level,
interpersonal dynamics may
also cause contention. For
instance, hung juries were
more likely to report that one
or two jurors had dominated
the discussion in delibera-
tions. Group dynamics also
shape the way jurors conduct
Interpersonal Jury Dynamics
deliberations. Typically,
jurors focus either on secur-
ing a verdict in deliberations
or on assessing the evidence.
The former orientation is
more characteristic of hung
juries than verdict juries.
Declaring a vote to other
jurors solidifies the opinion
as compared to deliberations
in which jurors delay a for-
mal vote and focus their
discussion on the evidence.
Jury Decisions in Four Districts – From First Vote to Final Vote
Summing Up: What Factors Relate to Hung Jury Rates?
Factor Related? How?
Demographics & Jurisdictions
Locality demographics yes Evidence of urban rates higher, such as for DC Federal Circuit Court.
Racially diverse jury? no Differences in racial makeup by jurisdiction, but no difference for hung/verdict juries.
State (more likely) vs. Federal (less likely) yes Federal court rates lower than state courts, possibly due to case mix and attorney resources.
Complexity
Number of charges yes More charges increase likelihood that jury will hang on at least one. Yet juries with fewer
charges to consider more likely to hang on all charges.
Quantity of evidence no
Length of trial no
Level of case complexity yes Hung cases were more complex for jurors, but not according to judges/attorneys.
Evidence
Ambiguity of evidence yes Close-call cases more likely to hang.
Police credibility yes Hung juries disagreed more often on believability of police testimony.
Defendant believability yes Hung juries thought defendant was less believable.
Attorney skills yes Judge rated defense attorney skills as more favorable in hung jury cases.
Deliberations
Timing of the first vote yes Juries were more likely to hang if they voted early in deliberations.
Structure of deliberations yes Verdict-driven more likely to hang as compared with evidence-driven discussions.
Interpersonal dynamics of jury yes Hung juries report more conflict and more unreasonable jurors.
Jurors’ perception of unfairness of law yes Hung juries thought legally correct outcome and law were unfair.
Acquitted
Hung Convicted
49%
21%
4%4%
8%
15%
7%
36%
28%
28%
2%
6%
16%
53%
22%
■ Strongly leaning towards acquittal
■ Slightly leaning towards acquittal
■ Evenly split
■ Slightly leaning towards conviction
■ Strongly leaning towards conviction
Final Vote
First Vote
Although mandating certain
procedures for conducting
deliberations may be viewed
as an invasion of the jury’s
province, providing jurors
with guidance to structure the
process might prevent some
group and individual juror
behaviors that relate to jury
deadlock. For example, the
court may suggest methods
for resolving conflicts effec-
tively and conducting ballots.
49% of juries acquitting on final
vote were strongly leaning
towards acquittal on first vote
Non Profit Org.
U. S. Postage
PAID
Richmond, VA
Permit No. 750
Points of view expressed herein are those
of the authors and do not necessarily repre-
sent the official position or policies of the
Bureau of Justice Statistics.
C
ASELOAD
H
IGHLIGHTS
National Center for State Courts
300 Newport Avenue
Williamsburg, VA 23185-4147
Research Division
800/616-6109
NCSC President NCSC Vice President
Roger K. Warren Research Division
Victor E. Flango
Court Statistics Project Staff
Brian J. Ostrom, Director
Fred L. Cheesman, II, Senior Court Research Associate
Neal B. Kauder, Consultant, VisualResearch, Inc.
Robert C. LaFountain, Court Management Consultant
Nicole L. Mott, Court Research Associate
Shauna M. Strickland, Court Research Analyst
Brenda K. Uekert, Senior Court Research Associate
Brenda G. Otto, Program Specialist
Conclusion
In existence since 1975, the CSP is administered by the
National Center for State Courts, with generous support by
the Bureau of Justice Statistics. The CSP receives general
policy direction from the Conference of State Court Ad-
ministrators through its Court Statistics Project Advisory
Committee. Those wishing a more comprehensive review
and analysis of the business of state trial and appellate courts
are invited to read the CSP’s latest publication, Examining
the Work of State Courts, 2002.
BJS
There will always be cases
in which juries are unable
to reach consensus. The
current work has provided
valuable data on how often
jurors hang and whether,
as some commentators con-
tend, jurors hang due to
illegitimate reasons. We
know that, on average, state
court jurisdictions have
hung jury rates of about
6.2 percent. The data also
show that rates vary consid-
erably due to definitional
standards and from jurisdic-
tion to jurisdiction, from a
low of 3 percent to a high
of 22 percent.
The data from this project
suggest that hung jury
rates may be reduced by
paying attention to evi-
dence strength and charging
decisions, focusing on fair-
ness of the law as applied
in individual contexts, and
determining how courts
may guide jurors in
deliberations.
A summary table of what
characterizes hung juries is
displayed on the previous
page. In addition, the cur-
rent study proposes the fol-
lowing goals and recommen-
dations to courts, especially
for jurisdictions with high
hung jury rates:
Evidentiary Issues
Facilitate learning process for
jurors by providing note-
books, plain-English instruc-
tions, copies of instructions
for jurors and allowing jurors
to take notes and ask ques-
tions during evidence presen-
tation/testimony.
Legal Fairness
Prosecution should evaluate
charging decisions for cases
with ambiguous evidence and
prepare to offer corroborating
evidence in trials relying
heavily on police testimony.
Group Dynamics
Offer guidance on selecting
presiding juror, conducting
ballots, and resolving con-
flicts, as many jurors lack
previous experience in small-
group problem solving similar
to deliberations.
These goals improve the
likelihood that juror dead-
lock will only take place in
cases in which the evidence
is so closely matched that
reasonable members of a
jury can disagree about a
case outcome. While this
may not be the preferred
result, it helps preserve the
traditional role of the jury in
the American justice system.
The Court Statistics Project (CSP)
