*1 Bryant 575 PEOPLE v BRYANT (Calendar 2). 6, Argued Docket No. 141741. December 2011 No. Decided denied, June 2012. US_; Certiorari Ct 133 S 664. by jury Court, Ramon L. was convicted a in the Kent Circuit Soet, J., first-degree conduct, H. David criminal sexual armed robbery, possession marijuana. selected, jury and After the was sworn, timely objection but it before was defendant had amade to composition ultimately the racial of the venire. The court denied objection. conviction, Following argu- appealed, his defendant ing part deprived in he had been of his Sixth Amendment right impartial jury an to drawn from a fair cross section of community only because there was one African-American in the people. Appeals, PJ., venire 45of The Court of and SMOLENSK!, J., concurring part dissenting part), J. in and (Borrello, Sawyer, part unpublished curiam, opinion per in an affirmed issued (Docket 16,2004 241442), March No. and remanded the case purpose conducting trial evidentiary court for the sole hearing regarding challenge remand, defendant’s to the venire. On court, Kolenda, J., evidentiary trial Dennis C. conducted hearings computer and programming found that as the of a result error, disproportionately larger a questionnaires number of and, hence, persons residing zip summonses were sent to codes populations with disproportionately smaller African-American and questionnaires and, hence, fewer persons were sent to summonses residing zip larger populations. codes with African-American The right trial court held Sixth Amendment anto impartial jury was not violated because African-Americans were underrepresented in the venire from which defendant’s County’s jury-selection process, selected and that Kent at the trial, systematically time of defendant’s did not exclude African- appealed. Appeals, Americans. Defendant The Court of Jansen, EJ., JJ., and reversed and a remanded for Borrello Stephens, impartial jury, new trial concluding before an that defendant had established violation of the Sixth Amendment’s fair-cross-section (2010). requirement. App Supreme Mich Court granted prosecution’s application appeal. leave 489 Mich (2011). 491 Mich joined by opinion by Justice YOUNG Justice Chief In an Zahra, Supreme Court Markman and Justices Mary Beth Kelly,
held:
determining
distinct
When
*2
juries
group
are selected is fair and
in
from which
venires
persons
the
of
in
in relation to the number
those
reasonable
jury pools
composition
community,
of
and
must examine the
courts
using
data
time
the most reliable
available.
venires over
the United States Constitution
1. The Sixth Amendment of
by
impartial
right
guarantees a
the
to be tried
defendant
community.
the
Under Duren
a fair cross section of
drawn from
prima
Missouri,
(1979),
facie case of
US
to establish
requirement, a defendant must
the fair-cross-section
violation of
(1)
alleged
group
to
excluded is a distinct
that the
have been
show
(2)
community,
group
group
representation
the
of this
in
that
the
juries
not
and
were selected was
fair
in venires
which
persons in the
in relation to the number of those
reasonable
(3)
community,
awas
result of
and
that this
jury-selection process.
systematic
group
of the
in the
exclusion
Duren,
respect
prong
African-Americans
2. With
to the first
of
community
group
purpose
the
deter-
are a distinct
in the
for
of
mining
was a violation of the Sixth Amendment’s
whether there
requirement.
fair-cross-section
Duren,
prong
respect
no
3.
to the second
of
individual
With
be
of
to
whether the
method
used exclusive
others
test
should
group
representation
and
of a distinct
in a venire was fair
proffered
parties
that
have
sufficient
reasonable. Provided
the
case-by-case approach
evidence,
employ a
that
courts should
all
considers the results of
the tests.
applying the
to determine
4. WHhen
relevant statistical tests
reasonable,
representation
must
was fair and
courts
composition jury pools
of
and venires over time—not
examine the
just
composition
the
of
individual
venire. The
the
considering
only
by
Appeals
tests
Court of
erred
the results of
composition
Those results
the
of defendant’s venire.
examined
misleading
exaggerated
the
set.
because of
incorrect data
were
Evaluating
representation
group
of a
in venires
5.
the
distinct
requires using
input
data available to
over time
the most reliable
case,
appropriate
it
to
into the
In this
was
evaluate
relevant tests.
using
composition
statistical estimates. Those estimates
venire
January
percent
4.17
issued from
showed that
of the summonses
African-Americans,
through
while the
March
were sent
community
jury-age
in
was
population of African-Americans
the
Thus,
percent.
underrepresented
8.25
African-Americans were
composition
jury pools
during
period.
venires
absolute-disparity
group’s
6. The
test measures a distinct
underrepresentation by subtracting
percentile representation
group
jury pools
percentile represen-
or venires from the
group
population
tation of that
overall
of the relevant
community.
generally required
disparity
have
Courts
an absolute
percent
representation
than
more
hold
distinct
reasonable,
group
absolute-disparity
was not fair and
but the
test
has often been criticized
it
because
is difficult for a defendant to
showing
group
population
make this
if the distinct
has a small
community
Courts
look
should
at the results of the test and
examine how far above or
are
below
threshold
results
when
determining whether
the defendant has established that
Appeals
fair and
reasonable.
Court
by disregarding
absolute-disparity
erred
the results of the
test
population
County.
because
small African-American
in Kent
comparative-disparity
7. The
test measures the decreased likeli-
underrepresented group
hood that members of an
will
called
be
by dividing
It
service.
is calculated
the results of the absolute-
disparity
percentage
of the distinct
in the overall
*3
population
community.
of the
The test has been criticized because it
alleged underrepresentation,
particularly
invites distortion
population
group
the
when
of the distinct
is small. The Court of
Appeals
by effectively establishing
bright-line
erred
favoring
a
rule
comparative-disparity
population
the
test when the
the
of
distinct
group
examining
Properly
composition
is small.
the
of venires over a
period,
comparative disparity
three-month
the
in this case was 49.45
percent.
comparative disparities
Other courts have found
above 50
percent permissible,
Appeals
and the
the
cases
Court of
cited for the
proposition
percent comparative disparity
that a 30 or 40
has been
representa-
found sufficient to
unfair
demonstrate
and unreasonable
distinguishable.
tion are
Defendant failed to establish that African-
representation
American
was not fair and
under
reasonable
the
comparative-disparity test.
probability
The
8.
standard-deviation
calculates the
underrepresentation
group
the
of the distinct
result
was the
of
chance, comparing
group
the actual distribution of the distinct
proportional
within the data set to the
and
distribution
thus
measuring
likely
vary
the
extent which
observed result is
expected
accepted
from an
result. No court has
a standard-
analysis
deviation
alone as determinative in a Sixth Amendment
challenge
jury-selection system
degree
to a
because whether the
of
underrepresentation
is the
of
result
chance does not inform the
Mich 575
representation
fair and
level of
was
whether the
determination
to be
test was not entitled
reasonable. The standard-deviation
any
given
weight.
disparity-of-risk
the likelihood that
the
test measures
9. The
representation
jury pool
group’s
in
and its
a
the
difference between
significant
community
in a
risk that the
population in the
will result
by comparing
fairly represent
group
the chance that
jury
the
will not
group
jury
if
of a distinct
will include members
a defendant’s
popula-
pool
its
group’s representation
the
is consistent with
in
community
a
in
the chance that
defendant’s
will
tion
the
with
alleged underrepresen-
group given the
include
of the same
members
percent
Disparities
should be deemed
of risk that exceed 50
tation.
measuring a defendant’s
because when
unfair
unreasonable
percent
injuries,
disparity
probabilistic
of 50
or lower shows
a risk
that,
not, removing
underrepresentation
likely than
the
would
more
jury.
composition
of
defendant’s
Defendant’s
not have altered
disparity
only
percent.
was
24.39
of risk
tests,
failed to
10. Given the results of
statistical
defendant
prong of
that the
show under the second
Duren
fair and
the venires at issue was not
reason-
African-Americans
able,
prima
a
failed
a
facie case of
Sixth
and he
to establish
violation.
Amendment fair-cross-section
Duren,
systematic
prong
11.
exclusion of
Under
third
jury-selection
particular
inherent
is one
is
computer programming
process
error
inherent
used.
was
systematic
jury-selection process
used and resulted in the
during
period at
of African-Americans in venires
exclusion
upon
fact that
error was unintentional and fixed
its
issue. The
discovery
systematic
to whether
exclusion oc-
was immaterial
Remand),
(After
App
People v
Mich
curred. In
Hubbard
(1996),
by holding
Appeals
erred
that the threshold
the Court
prong
underrepresentation
when the
under
second
is lower
benign
the result
circumstances less
underrepre-
if a
can
than random selection. Even
defendant
show
systematic,
was
the defendant must show that the
sentation that
any underrepresentation
not fair and reasonable.
extent
was
to the
that it
inconsistent
Hubbard must be overruled
extent
holding
with
this case.
convictions and sentences rein-
Reversed
*4
stated.
opinion,
concurring,
joined
majority
the
Justice
Makkman,
reasonably
governing
applied
the
the Duren test
Sixth
which
separately
requirement, but wrote
Amendment’s fair-cross-section
concerning
questions
raise
test and the constitutional
to
Significant uncertainty
standard toward which it is directed.
exists
regarding how
relevant in
to ascertain which statistical tests are
claims,
assessing
interpret
how to
the results of
fair-cross-section
tests,
conflicting
to
such statistical
and how reconcile
results
among
statistical
such
tests.
dissenting,
the
Justice
concluded that
Court of
Cavanagh,
by
Appeals
clearly
holding
did not
err
that defendant was
unique
a
the
entitled to
new trial under
of the case.
facts
always
Although
analyzed
courts have
of the
this area
law
precision,
longer
with
defendant was entitled to relief
if
even
representa-
time frame had been
evaluate
used to
whether the
County
tion
was fair
of African-Americans in Kent
venires
by
approach
Appeals
reasonable. The
taken
the Court
of
Hubbard,
may give
under which a court
the defendant
underrepresentation
jury-selection
benefit
the doubt on
if
of
process
nonbenign influence,
bears
mark of a
should
abe
determining
relevant consideration in
whether unfair and un-
Applying
has been
reasonable
shown.
case,
approach
Appeals
to the facts of the
the Court of
did not
by
clearly
holding
err
that defendant was entitled to new trial.
Cavanagh
judgment
pass
Justice
would further have declined to
disparity-of-risk
on
the merits of
the benefit
without
of
briefing
argument.
full
and oral
joined by
dissenting,
Justice
Justice Hathaway,
Marilyn
Kelly,
agreed
majority
with the
had
that defendant
satisfied
first and
prongs
third
of the
for a
Duren test
violation of
fair-cross-section
Amendment,
disagreed
the Sixth
but
had failed
defendant
to
prong
Appeals
meet the second
of the test. The Court of
did not err
disparity
composition
specific
when it considered the
in the
of
evaluating
prong
venire when
second
the Duren
multiple
considering
test but considered
venires over time when
prong
by
authority
test.
the third
the Duren
The
cited
majority
proposition
requires
prong
for the
that Duren's second
multiple
compelling
consideration
venires over time was not
by
authority.
opinion
majority
was contradicted
other
also
Appeals’
miseharacterized the Court
treatment
comparative-disparity
Appeals properly
test. The Court of
exam-
it,
presented
ined the results
the statistical tests
hut found
helpful
of one
results
test more
than the
of the other
results
majority
by
tests under the facts of
case. The
erred
also
disparity-of-risk
importing the
test into this
its use was
case when
by
parties
not briefed
or considered
the lower courts.
Further,
overruled;
partially
Hubbard should not have been
in-
stead,
courts should he able
consider the
for the
reason
*5
When whether juries fair and venires from which were selected was reasonable community, persons in the courts relation to the number of those composition jury pools and venires over time must examine the available, using just composition the most reliable data venire; may the individual defendant’s statistical estimates he available; they provide if reliable data the United used most Supreme specified preferred method for States Court has not measuring group in a of a distinct venire reasonable, was fair and and no individual method should be used others; parties proffered provided that have exclusive of evidence, case-by-case apply approach sufficient courts should (US VI). Const, Am that considers the results of all tests Opinion of the Court — — — 3. Constitutional Law Juries Venires Fair Cross Section of the — — Community Groups Distinct African-Americans. guarantees The Sixth Amendment of the United States Constitution right impartial jury a defendant the tried be drawn from community; a fair cross section of the African-Americans are a community purpose determining distinct for the requirement whether there was a violation of the fair-cross-section (US VI). Const, Am Schuette, Bill Attorney General, Bursch, John J. *6 General, Solicitor A. Forsyth, Prosecuting William At- torney, McMorrow, and K. Timothy Assistant Prosecut- ing Attorney, for the people.
Arthur James Rubiner for defendant.
Amici Curiae: Schuette, General,
Bill Attorney Bursch, John J. General, Bandstra, Solicitor Legal Richard A. Chief Counsel, Restuccia, and B. Eric Deputy Solicitor Gen- eral, for the Attorney General.
Bradley R. Hall for Criminal Attorneys Defense Michigan.
ZAHRA, J. This
presents
question
case
whether
right
defendant was denied his Sixth Amendment
an
to
impartial
jury drawn from a fair cross section of the
community. A
claim
fair-cross-section
under the Sixth
Amendment
requires
defendant to make a
facie
prima
case as
set forth
the United States
Court in
Supreme
Namely,
Duren v Missouri.1
a defendant must show:
(1)
group alleged
that the
to be excluded is a ‘distinctive’
(2)
group
community;
representation
in the
that the
of this
group
juries
in venires from which
are
fair
selected is not
in
persons
reasonable
relation to the number of such
(1979).
Missouri,
357;
Duren v
439 US
99 S Ct
satisfied drawn from a fair cross right impartial of his community, and defendant new granted section erred because Appeals trial. We conclude that Court that the prong defendant failed to show under second in venires from representation of African-Americans not fair and reasonable in juries which were selected was African-Americans in the com- relation to the number of evaluating munity. Appeals The Court erred prong significant ways. second two First, Appeals wrongly the Court of relied on mis- leading by considering repre- data representation only sentation of African-Americans defendant’s ve- addressing representation nire when was fair explicitly requires reasonable. Duren courts of a in ve- consider distinct inadequate only nires. The use of this sample evaluating defendant’s venire caused the tests the de- *7 gree any underrepresentation of to skewed and produce exaggerated results. Second, the of our decision in Appeals misapplied Court 3 Smith, In held that an
People Smith. we evaluation of case-by-case the second courts to a prong requires employ approach that considers all the relevant statistical tests data a evaluating regarding representation for the of group using any distinct one individual method without Contrary the to this the holding, exclusive of others. Court Appeals effectively adopted bright-line of a rule favor of 2 Id. at 364. (2000). Smith, People v Mich 615 NW2d Opinion op the Court
the in all comparative-disparity test instances which the is that all population group distinct small. Given the shortcomings, requires relevant tests have Smith a comprehensive degree courts to take view of the of elevating without one over the test of a Nonetheless, Appeals, using others. the Court skewed test, result from the comparative-disparity elevated this precisely above others situation in which its use distorting degree underrep- is most of criticized— when the population group resentation of the distinct is small. that all applying
We hold when the relevant tests for data, exam- evaluating representation a court must composition pools ine the of or over venires time using the most reliable data available to determine representation distinct fair group of a is and Having reasonable.4 considered results these tests using set, most reliable data which included or composition jury pools venires three- over month we period, conclude defendant failed to show that the of African-Americans was not fair and reasonable. Accordingly, we reverse the judgment of Court of and Appeals reinstate defen- and dant’s convictions sentences.5 “venire,” “jury pool,” “jury “array” panel,” The terms and are (9th ed) interchangeably. Dictionary sometimes used Black’s See Law “ (defining panel persons jury duty [a] “venire” as selected for — among jurors array; jury whom are chosen. to he Also termed panel; jury pool”). this, Because of our references to “venire” are to the jurors group potential petit in the courtroom from which a defendant’s “jury pool” are selected and our references are to people appear jury duty day. particular summoned on a engages unnecessary opinion believes that dissent error correcting. reasons, disagree addressing published For obvious we opinion misapplied Appeals principles, Court of constitutional Supreme precedent, precedent United States Court our unneces sary. *8 Mich 575 Opinion of the Court PROCEDURAL HISTORY
I. FACTS AND criminal first-degree defendant juryA convicted robbery, 750.520b(l)(e), armed conduct, MCL sexual MCL marijuana, 750.529, and possession MCL 333.7403(2)(d). she at- testified that when The victim defendant, he a put from buy crack cocaine tempted He then money. her her head and demanded gun to him, taking her car oral sex on her to perform ordered until not let her leave her that he would telling and keys defendant, he police apprehended she did so. When in his marijuana possession. had sworn, selected, but before it was jury was After composi- to the racial timely objection made a defendant it noted that The trial court jury tion of his venire. Latino in defen- and one one African-American observed ruling to reserve its venire, but decided 45-person dant’s hearing following day.6 objection until on regard clerk hearing, jury At the testified and venires. jury pools composing for procedure clerk, Secretary of State jury to the According of Kent all the residents the court a list of provides and have a valid years age at least 18 County who are From that or valid state identification. driver’s license residents to list, randomly selects computer program then ran- program jury questionnaires. be sent the names using to be summoned domly jurors selects jury questionnaire to the responded of those who service opted or out disqualified had not been age.7 because of record, exchange between defense the court went off the Before the indi trial showed confusion about counsel and the court as an African-American member that the trial court had identified
vidual actually step-father. of defendant’s venire (addressing grounds disqualification MCL 600.1307a See service). exemption *9 Opinion the Court of The clerk of jury Secretary testified that the State database does not include the race of the individuals listed and computer program that the does account for race selecting jurors. jury when date For the defendant’s was selected, Januaiy 28,2002, only 132 of the 182 who people had the randomly by computer been selected and program jury jury issued summonses for service. the appeared By one inspection, only clerk’s visual was African-American. Of the appearing, computer program randomly 132 the selected for people 45 defendant’s venire. The clerk also to the voluntary submitted trial court results of taken some surveys actually appearing of those duty given days on in 2002.8 January
Defendant, relying voluntary on the results of the surveys, that the of argued disparity African-Americans for jury compared to African- appearing duty American of that population county showed jury-selection current did method not include fair community. cross section The trial ulti- court mately venire, denied rul- challenge defendant’s his ing jury-selection system that because the was race neutral, of African-Americans was a voluntary function failure of indi- those participate. viduals to 8 results, “Jury The which were contained a document entitled Representation Compilation,” Community Survey on Janu reflect that ary 7, appearing responded, of 169 of those 160 with 2 individuals they
indicating indicating they that were and 2 that African-American multiracial; 9, January potential jurors appearing were on the 77 3 of African-American; responded, indicating none or she with that he was on 14, January potential jurors appearing responded, of the 140 2 with they indicating indicating they that were African-American and 2 that multiracial; January 22, potential .jurors appearing were on the 18 responded, indicating with none he or she was and that African-American multiracial; and, indicating January he or she was on 52 of the potential jurors appearing responded, indicating with 1 he or she was African-American. 491Mich 575
Opinion the Court sentencing, Following defendant his conviction majority Appeals appealed. affirmed Court of part, the trial court for the case to remanded but hearing regarding evidentiary claim that a fair cross section not reflect venire did his community.9Addressing representation (African-Americans) was fair the distinct majority prong, the second under Duren’s reasonable had not shown that defendant concluded under the and reasonable not fair majority Nonetheless, the tests.10 relevant statistical approach applied v Hubbard set forth Remand), (After found to “the defendant was which *10 underrepresentation where shown substantial have ‘non-benign’ disparity circumstances; resulted from underrepresentation did not occur as is, that where approach, this of random chance.”12Under the result majority had satisfied assumed that defendant possi prong the evidence indicated the because second bility underrepresentation not the result that the was of random selection.13 prong, prosecution
Regarding
admitted
the third
jury-selection process disproportionately se-
that the
zip
jurors
result,
certain
codes.14As
lected
majority
from
to the trial court for an
remanded the case
“present
evidentiary hearing in which defendant could
County
system
jury selection
that the Kent
evidence
9
Bryant, unpublished
opinion per
People
curiam of the Court of
(Docket
241442)
I).
(Bryant
Appeals,
March
No.
issued
10 Id. at 2-4.
Remand),
459, 477-478,
(After
App
People
217 Mich
v Hubbard
(1996).
14 Id. Opinion Court in systematic resulted exclusion of African-Americans causing to be substantially underrepresented defendant’s venire.”15 remand,
On the trial court16 held several hearings testimony and heard from the manager, court’s case clerk, a County Board, member of the Kent Jury and two statistical experts. testimony, From this trial court found that a computer programming error responsible of African- Americans in venires from June 2001 to August 2002.17
The trial court found that Kent County, an effort money spent fees, save on software in April switched 2001 from using a vendor’s for summoning software jurors to software developed its information technol- ogy department. Rather than drawing the entire 456,435 database18 of names that the Michigan Secre- tary of State had provided for County, Kent the new computer program had an erroneous setting using only 118,169 of program those names. The selected ran- Appeals rejected Id. at 5. remaining The Court of issues appeal. Judge majority on Id. at 5-7. concurred with the on Borrello issues, regard these but dissented with to defendant’s fair-cross-section claim because he believed that the evidence established sufficient under- computer excluding and that the zip having error codes larger minority populations systematic constituted exclusion of African- J., Americans from concurring part the venire. Id. at 2 (Borrello, dissenting part). reassigned Judge This case was Dennis Kolenda on remand *11 Judge Soet, presided trial, because David who had over defendant’s had retired. 17 frequency prospective jurors zip with which from certain codes jury questionnaires prompted investigation, were sent an which resulted discovery programming 2002, of the error June four months after defendant’s trial. 18 summary jury testimony, As noted in the clerk’s this database people included the Michigan names and addresses of shown to have a Michigan personal driver’s license or identification card with an address County. in Kent Mich 575
Opinion of the Court 118,169 be sent names would domly list who 118,169 individuals Because the jury questionnaires. codes, zip from certain came disproportionately selected sent to disproportionately jury questionnaires were in a disproportionately This resulted zip those codes.19 going zip to codes jury questionnaires number of larger dispro- and populations African-American with smaller going zip codes questionnaires fewer portionately African-American larger populations.20 with selected, jury was For the week that defendant’s The court people court for service. summoned 28, January 183 of the 293 for specifically summoned Of the 183 picked. people when defendant’s was summoned, randomly and 45 of them were appeared noted, the court used in defendant’s venire. As placed and race of those voluntary surveys identify gender jurors All 132 who jury duty. potential appearing January voluntary to the appeared responded on African-American survey, specifying with one individual specifying and one individual multiracial. First, at the experts hearings.
Two statistical testified for defen- expert Dr. Chidi Chidi testified as a statistical analyzed voluntary surveys potential dant. He Rely- from 2001 to 2004. jurors appeared completed who surveys, Dr. Chidi ing voluntary on the results of and comparative- concluded that the standard-deviation 19The trial court found that the there was no evidence that zip anything other than the of certain codes evidence, however, that reflects “result of a random draw.” There is some Secretary original grouped that the from the of State the names database analysis by zip discrepancy This does not affect our because we code. underrepresentation was inherent conclude in either event that the systematic jury system thus constituted a exclusion within the meaning prong. of Duren’s third County following by again hiring Kent corrected the error the month changing computer program it used. outside vendor *12 Opinion of the Court tests disparity proved there had. been systematic exclusion of juiy duty. African-Americans from The trial court, however, rejected testimony, Dr. Chidi’s finding that Dr. Chidi personal showed bias and a failure to under- stand basic statistics because he had analyzed only those opted individuals who voluntary answer the survey after appearing duty.21
Given its disapproval of testimony, Dr. Chidi’s pursuant 706,22 to MRE the trial court selected Dr. Paul Stephenson expert. as its Using data from the 2000 Census, Dr. Stephenson conducted analysis his with the assumption that the population of African-Americans old enough jurors to serve as percent constituted 8.25 County. Kent records,
From court Dr. Stephenson identified the number jurors summoned from each zip code for each month January from 2002 through March 2002. Dr. Stephenson then used those records and the census data for racial population in zip each code to estimate that, as a result bias, of the zip-code only 163 of the (4.17 3,898 summonses percent) sent out January through March 2002 went to If African-Americans. 8.25 percent of the summonses sent during out that period gone had to African-Americans, then 322 African- Americans would have been sent them.
Considering only
venire,
Dr. Stephenson
calculated that the
disparity23
absolute
percent
was 6.03
testimony
Defendant summarizes Dr. Chidi’s
in his brief and asserts
any meaningful analysis
without
wrongly rejected
that the trial court
testimony.
suggest
Our review of the record does not
that the trial court’s
rejection
testimony
2.613(C).
error,
of his
amounted to clear
MCR
Accordingly,
testimony
we will not consider
analysis.
Dr. Chidi’s
in our
permits
appoint
expert
MRE 706
a court to
witness on its own
motion.
absolute-disparity
portion
test measures the
of the overall
population
of a
subtracting
distinct
that has been excluded
Opinion the Court percent. 73.1 How- disparity24 comparative and the these the results of ever, disregarded Stephenson Dr. that because of small tests, report in his explaining County, in Kent African-Americans population *13 identify whether the test could not absolute-disparity He statistically significant. was underrepresentation changes representation that small explained further the result of the distorting the effect of in the venire had test. comparative-disparity standard- considered the Dr. also Stephenson test,25 the use of this test because rejected but deviation ,”26 . . Dr. is not valid. approximation “the normal to the however, applied analogous a test Stephenson, distri- test, calculating the binomial standard-deviation pro- the venire-selection to determine whether bution calculation, Dr. Stephenson From this cess was valid.27 find insufficient evidence to concluded that there was significantly underrepre- African-Americans were if there had because even sented defendant’s venire out, sent no bias in how the summonses were been venires would have randomly selected percent 10.477 group jury pools percentile representation or venires from the of that group population percentile representation of the of that in the overall 111(B)(2)(a) opinion. community. part See of this relevant comparative-disparity test measures the decreased likelihood underrepresented group will be called for service that members of an by dividing absolute-disparity test and is calculated the result of population percentage in the overall of the distinct 111(B)(2)(b) opinion. community. part See of this degree probability measures the that the The standard-deviation part of random chance. See could be the result 111(B)(2)(c) opinion. of this approximation of a The standard-deviation test uses a normal sample Stephenson size variable. Dr. indicated that binomial random large enough given proportion the test of African- was not community. Americans in the analogous distribution. This test used the “exact” binomial Opinion of the Court had In Stephenson’s one or no African-Americans. Dr. view, this likelihood was sufficient for the disparity African-American to be in- statistically significant, but this conclusion was related to the small sample examining just size when defendant’s venire.
Examining larger three-month sample, Dr. Stephenson performed further using calculations the bi- nomial results to find that there was essentially “no chance” that the reduced numbers of African-Americans pools January between and March 2002 occurred Further, as a result of random chance. a venire selected during zip-code problem time the occurred ap- proximately four times likely more to contain no more than one African-American than if problem had not present. been He concluded if the estimates matched practice, systematic actual “a did bias exist in the selection of individuals summoned for jury duly... inevita- [that] bly led to the under representation” of African-Americans *14 in the jury pools January through March 2002.
In a written opinion, trial court ruled that defen- dant was not entitled to a new trial because he had failed to satisfy Duren’’s second and third prongs. Addressing whether the representation of African-Americans was fair reasonable, and the trial court reasoned that there was no proof of actual underrepresentation in group of indi- viduals that the computer program identified and to jury questionnaires whom were sent because the Secre- tary of State identify database does not race.28In the trial view, court’s comparing an many estimate of how African- Americans were questionnaires sent and many how would only individuals, group The trial court considered this not the resulting pools, pools because the were affected considerations for responsible disparities which the court was not such as racial in whether questionnaire delivered, response rates, disqualifications, was hard ships, people appear. and who failed to 491 MICH575
Opinion of the Court computer pro- absent the questionnaires have been sent data is required flaw was not sufficient because hard gram under Smith.29 there no
The trial court also concluded that prong exclusion under Duren’s third be- systematic the defective com- cause there was no evidence that Rather, setting any simply randomly had bias. it puter jurors reduced the number of individuals whom were Therefore, selected from. the end result —that these taken from certain disproportionately individuals were in the court’s zip jury- codes —was not inherent selection processes. appeal, Appeals
On defendant’s second
the Court of
opinion
concluded in an authored
that defendant had
fair-
established
violation of
Sixth Amendment’s
requirement
cross-section
reversed
remanded
panel
case for a new trial.30 The
referred to each of the
tests
used to measure
generally
reasonable,
a distinct
is fair and
purportedly follow-
Smith,31
ing
case-by-case
set forth in
approach
29 Smith,
holding
First, relying on Dr. Stephenson’s calculations for only venire, defendant’s the panel stated that absolute disparity percent. was 6.03 Although acknowl edging such a result does not indicate substantial underrepresentation, panel declined to find the absolute-disparity test controlling because it viewed it as an ineffective measure of acceptable disparity in circumstances, one, like this in which the group question makes up a small percentage of the total population.32
Next, the panel addressed the comparative-disparity acknowledged difficulties in applying this a group test to up makes a small percentage of the population.33 Nonetheless, the panel decided that comparative-disparity test was the most appropriate to measure the underrepresentation in cases in which the percentage of the distinct the population is low.34 on Dr. Relying Stephenson’s calculations only venire, panel stated that the compara- tive disparity was 73.1 percent, which it viewed as a significant disparity and “sufficient to demonstrate that the representation African-Americans the venire for defendant’s trial was unfair and unreasonable.”35
32Id. at 269.
33Id. at 269-270. point, Id. at panel 270-271. On this relied on United States v (CA Rogers, 8, 1996), 73 F3d which concluded that “the comparative disparity provides meaningful calculation a more measure of systematic impact group: vis-a-vis the ‘distinctive’ it calculates the representation injury pools of African Americans relative to the African- community American[s] [in the] rather than relative to the entire population.” II, App panel 289 Mich at 271. The concluded that the 73.1 percent comparative disparity was sufficient to demonstrate an unfair and unreasonable substantially higher because it was percent than the 30 or 40 that has been deemed sufficient in other cases. Id. at 271-272. *16 Mich 575 Opinion the Court briefly addressed the standard- addition, panel In Stephenson Dr. It concluded that because deviation test. because the appropriate was not that the test testified has valid and no court was not approximation normal as determinative analysis the standard-deviation accepted challenge, it had little value here.36 type in this Duren, panel Addressing prong the third from caused underrepresentation was held panel The of African-Americans.37 systematic exclusion in this case was underrepresentation that the concluded in County jury-selection process inherent in the Kent error resulted over- programming a computer which minority from codes with small jurors zip selection of codes jurors zip and underselection of populations Further, the evidence minority large populations. with occurred over a showed that this Therefore, because defen- of time.38 significant period facie case for a fair-cross- prima dant established claim under the Sixth Amendment section rebut, panel failed to reversed and prosecution for a trial.39 remanded new Court, sought appeal The leave to prosecution granted.40 which we
36Id. at 272-273.
37Id. at 274.
38Id. at 273-275.
39Id. at 275-276. III). (2011) Bryant, {Bryant People Mich Our order stated v part: (1) among parties include the issues to be briefed: shall
whether,
evaluating
group has been
whether a distinctive
Missouri,
sufficiently underrepresented
II. STANDARD OF REVIEW Whether defendant denied his Sixth Amendment right impartial drawn from a fair cross community section is a question constitutional that we review de novo.41We findings review the factual error, of a trial court for clear which “if the exists reviewing court is left with a definite and firm convic- tion that the trial court made a mistake.”42
III. ANALYSIS A. FAIR-CROSS-SECTION JURISPRUDENCE The Sixth Amendment of the United States Consti guarantees tution a defendant right to be tried by an impartial jury from a fair drawn cross section of the community.43 The United Supreme States Court recog- (2) prospective jurors; whether a defendant’s claim of such under- always supported by data, must be hard or whether permissible and, so, statistical estimates are if under what circum- (3) stances; any underrepresentation of African- venire, County jury pools
Americans in the defendant’s
or in Kent
systematic
between 2001 and
was the result of
exclusion
prong
of Duren.
under the third
[Jd.]
Armstrong,
(2011).
v
281, 289;
See
Mich
Assistance of [US VI.] Counsel for his defence. Am Although only provides the text of the Sixth Amendment in reference to a jury right impartial jury,” Supreme “the to... the United States Court Mich 575 Opinion Court Taylor v Louisi- guarantee nized the fair-cross-section ana.,44 successfully challenged In the defendant Taylor, scheme which women jury-selection Louisiana’s they service unless jury would not be considered willingness of their to serve.45 filed a written declaration district, in jury percent which 53 For female, 1,800 individuals population nearly only fill in a period year, drawn to venires held that practice were female.46The Court Louisiana’s women, systematically eliminated a “numerous and group, jury pool, denying distinct” from the the defen- drawn from a fair cross section right dant his community of the in violation of the Sixth Amend- ment.47 Duren,
In
forth
Supreme
the United States
Court set
right
has ascribed to that
must be drawn from sources
community
reflecting a fair cross section of the
in order to effectuate the
jury: “guard[ing] against
arbitrary power [by
purpose of a
the exercise of
making]
judgment
community
hedge
available the commonsense
as a
against
prosecutor
preference
the overzealous or mistaken
and in
to the
perhaps
response
judge.”
professional or
overconditioned or biased
of a
Louisiana,
522, 530;
692;
(1975),
Taylor v
419 US
95 S Ct
a more substantive designed framework evaluate challenges. Specifically, fair-cross-section to make a prima facie case of a violation of the Sixth Amend- requirement, ment’s fair-cross-section a defendant must show:
(1) group alleged that the to be excluded is “distinc- (2) group community; tive” in the that the group juries this in venires from which are selected is not fair and reasonable in relation to the number of such (3) persons community; in the underrepre- systematic group sentation is due to exclusion of the in the jury-selection process.[48] successfully defendant Duren argued of women venires violated requirement. fair-cross-section Regarding the first
prong, there dispute was no that women awere distinct group community.49 The defendant established the second prong “by [his] statistical presentation,” showing that while women were 54 percent county’s population, women were only percent 26.7 persons summoned service and percent 14.5 of veniremembers during approximately nine-month period.50 The Court concluded gross “[s]uch discrepancy between the percentage of women in jury venires the percentage of women in the community requires the conclusion that women fairly were not represented in the source from which petit juries were ,”51 drawn . ..
48Duren,
49Id. at Id. 364. Court, naming calculation, applied Id. at 366. The without its absolute-disparity by comparing percent the difference between the age population percentage of the distinct in the and the group appearing distinct in venires. *19 Mich 575 491
598 Opinion of the Court concluded that the third the Court Regarding prong, systematic of the was a result jury-selection process. in the group exclusion statistics, that the evidence Specifically, automatically exempted women scheme selection that a their and evidence upon request, service in venire every weekly had occurred large discrepancy “that the cause of the year almost a established is, systematic inherent underrepresentation was —that utilized.”52 jury-selection process in the particular THE DUREN TEST B. APPLICATION OF A DISTINCT GROUP IS 1. WHETHER ALLEGED TO HAVE BEEN EXCLUDED African-Americans, dispute There is no excluded, in the alleged group to be are distinct group determining community purposes for the Amendment’s fair-cross- there is a violation of the Sixth defendant satisfied Accordingly, requirement.53 section prong. Duren’s first IS FAIR AND REASONABLE
2. WHETHER REPRESENTATION defendant to show that prong requires The second of this venires from which “representation selected not fair and reasonable relation juries are community[.]”54 persons to the number such Smith,55 the United States Su- recognized As we a method or test that Court has not identified preme [52] Id.
States v
2009);
[55] Smith,
But in order to properly consider the
results
tests,
relevant
we
answer
questions
posed
must
the
we
in our grant
identify
order to
data
what
into
input
the tests.
we
the
Specifically,
parties
asked
to brief
the
evaluating
prong,
may
second
“courts
choose
only
to examine
the composition of the defen-
dant’s particular jury venire, or whether courts must
always examine the
of
composition
broader
or
pools
arrays
prospective jurors”
of
and “whether a defen-
of
dant’s claim such underrepresentation
always
must
by
data,
be supported
hard
or whether statistical esti-
permissible
mates are
. . . .”59
We
hold
when applying the relevant statistical
tests, a court must examine
composition
pools and venires over time using the most reliable data
56
Berghuis,
(acknowledging
See
329
US at
that no
decision
specified
proper
Court
underrep
has
method or methods
which
appropriately
taking
position
resentation is
measured and
no
on the
used). Additionally,
method or methods that
be
should
the United States
Supreme
underrep
has
Court
not identified a
for what
threshold
level of
Maskeny,
resentation is
fair and reasonable. United States v
609 F2d
(CA 1980).
183,
57Smith,
58Id.
III,
available reading compels of Duren Our reasonable. Specifically, that the sec- Duren sets forth conclusion. [the “representation prong ond is used to evaluate juries distinct] se- are which venires again plural .”60The used the “venires” lected ... Court defendant’s evidence under when it evaluated the “discrepancy pointing prong, out the between second per- percentage women venires centage community addition, . . .”61In in the . women requirement referred back to Court second-prong underrepresentation must occur over introducing on third time prong, its discussion when stating, necessary petitioner “[I]t *21 underrepresentation women, generally show that the of systematic venire, due to their exclusion and on his was jury-selection process.”62Therefore, when in the ering consid- reasonable, is fair composition requires a of court to evaluate the Duren significant period just time rather over a than venires venire.63 defendant’s individual 60 added). Duren, (emphasis at 439 US 364 61 added). particular, Id. Duren considered the (emphasis In at 366 nearly reasoning year part concluding a a of venires used for as its by presen prong that the was satisfied defendant’s statistical second 362-363, tation. Id. at 365-366. 62 added). Id. (emphasis 366 at 63 (CA 9, 1985) Miller, (stating v 1219, See United States F2d in 771 1228 prong appears a discussion of Duren’a second “[i]t that to that us the Duren Supreme setting up plural of in a clear Court’s use test is requirement a the fair cannot be indication that violation of cross-section single jury”); United States premised upon proof underrepresentation of in a (CA Allen, 1096, 1103 6, 1998) (stating in a discussion Duren’s v 160 F3d of however, “[a]ppellants, prong second that must show more than that their Rosans, unrepresentative”); People De App v particular panel 27 Cal (1994) (“The Duren 611, 621; Rptr prong requires 4th 32 Cal 2d 680 second cognizable venires showing underrepresented a is juries selected, panel the defendant’s which are not on from which People 601 v Opinion op the Court of
Consequently,
wrongly
the Court Appeals
consid-
ered the
of the
from a
results
tests
data set that
only
included
defendant’s
Relying solely
venire.
on
of
venire
composition
resulted
mislead-
selected.”);
Verdugo-Munoz, unpublished
v
United States
of the
order
Arizona,
12,
United States
Court for
District
the District of
entered October
*
(Docket
CR-03-1161-PHX-SEB),
2571608, 2;
2005
No.
2005 WL
US
2005
* (“[B]ecause
23448,
Supreme
plural
5
Dist LEXIS
use of
Court’s
describing
Duren,
prong
proffer
the second
of
a defendant must
evidence
underrepresentation
multiple venires.”);
that the
has occurred in
cf. United
(CA
561,
Williams,
5, 2001).
addition,
States v
264 F3d
568
In
an abundance
supports
applying
prong,
of
caselaw
when
Duren’%second
look to
courts
See,
degree
underrepresentation
e.g.,
over time.
United States v
(CA
792,
10, 2006); Weaver,
238,
Orange,
447 F3d
267 F3d at
(CA
Royal,
1, 5, 10-11
1, 1999);
Borg,
States
United
v
174 F3d
v
Thomas
(CA
1147,
9, 1998);
Rioux,
648,
F3d
United States v
97 F3d
657-658
(CA
(D
2, 1996);
Fabian,
970,
Supp
Minn, 2009);
Francis v
669 F
2d
People Washington,
(Colo, 2007);
Bell,
179 P3d
162-164
502, 526-527;
(1989).
Rptr 1;
Cal 3d
262 Cal
Dr. estimated that 4.17 of the summonses issued were sent to African-Americans January through March 2002. Given that the jury-age population census data reflects that the community percent, African-Americans is 8.25 it underrepresented. is clear that African-Americanswere pertinent question underrep- The then is whether this composition jury pools resentation in the and venires during this time was fair and nonetheless reasonable.
a. ABSOLUTE-DISPARITYTEST
absolute-disparity
widelyapplied
The
is most
jurisdictions
majority
test and
used
representation
group
evaluate whether the
of a distinct
group’s
was fair and reasonable.66This test
measures
underrepresentation by subtracting
percentile rep-
group
jury pools
resentation of that
or venires from
percentile representation
group
of that
in the
population
community.67
overall
of the relevant
absolute-disparity
permits
test is useful
it
because
straightforward
per-
and undistorted measure of the
Stephenson’s
We note that Dr.
estimate is more relevant
than the
voluntary survey
determining
results of the
body
potential
jurors
underrepresented
African-American
as a whole was
actually
it
because
looked at who was chosen to receive summonses
appear
given day.
rather than who decided to
for service on a
(D
2007)
Delgado Dennehy,
Mass,
Supp
See
503 F
2d
425-426
cases).
(collecting
Royal,
6-7,
See
centage
group
of more
disparity
an absolute
generally required
have
to indicate
percent
than 10
fair and
was not
reasonable.69
the distinct
*24
test, however,
is often criti-
absolute-disparity
The
difficult,
impossible,
if
makes it
not
cized because it
if
showing
to make
the distinct
a defendant
example,
For
community.70
in the
population
has a small
population
if the
African-American
percent
even
8.25
entirely
pools
here had been
excluded
the abso-
period analyzed,
for the three-month
venires
only
falling
8.25 percent,
would have been
disparity
lute
applied
determine
generally
below the threshold
is
the
fair and reasonable.71
68
7;
Note, Re-justifying
also
the
cross section
See id. at
see
fair
Equal representation
requirement:
the American
enfranchisement
(2007).
1568,
jury,
L
1596
116 Yale J
criminal
69
(CA
311,
7, 1995);
Ashley,
See
States v
54 F3d
313-314
United
Supreme
Maskeny,
Although
F2d
the
States
Court has
609
at 190.
United
test,
performed
absolute-disparity
the
calcula
it
same
not endorsed
Duren,
364-366;
disparity
at
in Duren.
US
see
tion to evaluate the
833, 860;
People Burgener,
Rptr
129 Cal
2d
62 P3d
v
29 Cal 4th
also
(2003).
70Smith,
elaborated on a
African
defendants would
juries, despite
petit
having
their
the 50%
African Americans on
If,
hand,
population
disparity.
is
on
other
absolute
50%African American and venires are 0%African
overall
American, then
drop
petit juror
having
the
from
an African American
would
odds
impossibility.
near-certainty
abso-
to total
cannot
fact
radically
distinguish
disparity
between these
lute
test
it
not
scenarios indicates that
does measure defendants’
different
[Commentary, Jury poker:
probabilistic injuries.
analysis
A statistical
requirement,
L
Ohio St J Crim
cross-section
fair
(2011).]
533, 545
71See,
Thomas,
(addressing
disparity
e.g.,
b. COMPARATIVE-DISPARITYTEST
Some courts have used the comparative-disparity
test, which measures “the decreased likelihood that
members of an underrepresented
group will be called
for
.
jury service . .
.”74It
by
is calculated
dividing the
(CA 1982)
9,
(addressing
disparities
percent,
percent,
absolute
of 2.8
7.7
(CA
percent);
Clifford,
150,
8, 1981)
and 4.7
United States
640 F2d
155
(addressing
disparity
percent).
an absolute
of 7.2
72
percent
percent
percent.
8.25
minus 4.17
is 4.08
result the of the population the overall group the distinct test is community.75 comparative-disparity it because invites distortion and is criticized widely used particularly when underrepresentation, alleged of the is small.76 group of the distinct the population of the result disfavoring after Appeals, The Court of of percentage the absolute-disparity the test because low, community was group the distinct in the relevant bright-line favoring rule the effectively established test, Supreme absolute-disparity United States the Id. Unlike the practice. applied comparative-disparity test has the Court never Thomas, Smith, (disfavoring at at see also 159 F3d 463 Mich any exaggerates comparative-disparity effect “it the the test because deviation”); example, assuming Royal, at For that the accord 174 F3d 8-9. excluded, person population the distinct was one and though jury comparative disparity percent without is 100 even result of community.” clearly ‘fair cross that member “would form a section’ 24 (CA 1,1984). put Hafen, As F2d one commentator United States v it, underrepresentation in a situation in which “[a] test that finds maximal virtually jury composition are unaffected cannot chances of measuring generally.” Note, proposal good apply A under- be a one to wheel, J 1913, 1928 composition YaleL (1994). problem comparative- with the Another commentator described disparity follows: test as disparity disparity comparative Yet lacks the absolute [population] of the total has been test’s awareness of what fraction example, tampered all African Americans are with. For when venires, highest possible comparative is the result absent score, figure disparity But is unless one also 100%. useless many are in the overall accounts for how African Americans American, population majority population. If African the total would reduce the odds then observed certainty drawing juryperson from near an African American hand, If, impossibility. African Americans on the other total comprise just population, 0.1% of the then likelihood total significantly drawing de- would not have an African American *26 Thus, commentators, among prominent despite support its clined. test, test, disparity disparity comparative like the absolute the injuries generated by probabilistic simply not measure the does [Jury poker, J L at 8 Ohio St Crim fair cross-section violations. 545-546.] Opinion of the Court comparative-disparity test when the of population the group distinct directly is small. This holding contradicts case-by-case the approach set forth Smith.77 Again, the comparative-disparity test is defective particularly claim when the involves a small of population a distinct group any because it distorts the extent of underrepre Thus, it sentation. does not follow to elevate the comparative-disparity test while disregarding the oth ers tests precisely the circumstance comparative-disparity test is most criticized apt to produce distorted results.
The Court of Appeals further erred it when consid- percent ered the 73.1 result of the comparative- disparity test only defendant’s venire. Using the proper Stephenson’s data from Dr. three-month exami- venires, nation of the comparative disparity was 49.45 percent.78 The United States Courts for the Appeals First, Third, Ninth, and Tenth Circuits have each found percent percent. Americans in defendant’s Mich at Smith African-Americansin the under the test and no to Justice result from the exclusive of the others.” measure must mischaracterized the Court of absolute-disparity continued, “[T]he comparative-disparity ment of a opinion Smith, population apply requires. belies this claim. In African-American absolute-disparity bright-line dissent, Kelly’s second Justice [463] 204. Mich at other, which some test comparative-disparity prong. requires rule in favor of distinct Court This ultimately By using relevant to measure the venire ....” the Court population figure yields particular, result of 4.08 approach clearly Appeals’ opinion regarding in cases in which the comparative-disparity test, stated, Appeals “no individual community Marilyn test to evaluate defendant’s claim Bryant II, is small. The Court of Appeals comparative-disparity did panel, test most percent Kelly just is low.”Id. ultimately is Mich [289] after a method should contrary comparative-disparity divided claims that we have a result of 49.45 opposite discussion the Thus, appropriate percentage App the establish used of African- Smith, test when at 270. It Appeals’ contrary only of what be used 8.25 “We *27 491 Mich 575
608
Opinion of the Court
percent.79
above 50
comparative disparities
permissible
for
Appeals
the Court
Moreover, the cases cited
has been deemed
or
percent
that 30
40
proposition
rep-
and unreasonable
unfair
to demonstrate
sufficient
readily distinguishable.
resentation are
percent comparative
v
a 30.96
Rogers,
In United States
Circuit
significant by
Eighth
deemed
disparity was
in dicta as the
was made
this determination
but
panel,
Eighth
precedent
Circuit
by earlier
was bound
panel
Ac-
system under review.80
regarding
the particular
convic-
the defendant’s
had
affirm
cordingly,
panel
Circuit has not
Thus,
Eighth
given
tions.81
or found it deter-
adopted
comparative-disparity
any weight
Rogers
do
case,
we
not afford
any
minative
jurispru-
it
an outlier in fair-cross-section
and view as
v
the Court
Beyer,
which
Additionally,
dence.
Ramseur
comparative disparity
a 40
cited,
percent
also
Appeals
minority
population
was deemed “borderline.”82
disparity
35.9
the absolute
percent,
that case was
far
Thus,
minority population
percent.
was 14.1
percent compara-
at hand. A 40
larger than in the case
case
a
persuasive
tive
is not
baseline
disparity
results
test distorts the
comparative-disparity
because the
involving
populations.83
small
Given
cases
79
upheld
Orange,
(noting that
the court had
See
It is type determinative in this of this test alone as result nothing challenge to do effect has because test prong. is, That the evaluation of second with degree underrepresentation is statisti- cally significant chance does not and not result representation is fair and inform whether level appropri- Instead, a result is more such reasonable.88 ately equal-protection context as an considered determining discrimination aid in whether intentional part perhaps third or as a of the evaluation the exists reality simply prong.89 function of what Duren This given actually randomness of a the test measures—the disparity.90 disparity, not the extent of inap- Stephenson it Further, Dr. concluded that was propriate apply the standard-deviation test approximation was not valid. because the normal case apply did, determine that however, He a related underrepresentation January the extent of through of random March 2002 was not result garner Nonetheless, all we from the result is chance. just not a random that —the *29 mere occurrence. The fact the 88As one commentator stated: test], question by [T]he [the answered standard-deviation one, interesting appropriate fair is not the one for a while an composition analysis. probability a that the of cross-section The community
jury by arose random selection from the is not wheel drawing directly jury a of of a related to the defendant’s chances composition. [Measuring underrepresentation, 103 L J certain Yale 1928.] at 89 (setting Jefferson, F2d at forth that “in the context See 962 1189 of claims, larger the the number of standard devia racial discrimination tions, product likely the is the more the observed result discrimination chance”). rather than 90 Jury poker, L 550. See 8 Ohio St J Crim at 611 Opinion the Court random was not result of chance does not establish Thus, was not that it fair reasonable. we afford the of this weight.91 result test no
d. DISPARITY-OF-RISKTEST is Another test sometimes discussed is the This disparity-of-risk test.92 test measures “the likeli- that the between group’s hood difference a representa- tion in the jury pool population its the commu- result in a nity significant jury will risk that will not fairly represent group.”93 It does so comparing (before the chance that a defendant’s or without dire)94 voir will include a distinct if group members of group’s representation in the jury pool is consis- tent with its in the population community with the 91 Marilyn Kelly’s Justice dissent viewsthis treatment of the standard- Appeals. test as with deviation inconsistent our of the Court criticism nothing Yet she does not contest that the standard-deviation test has measuring do with fair whether is and reasonable. Thus, merely is not it that the standard-deviation has flaws like tests; other it is that it irrelevant to is the consideration of the second Therefore, prong. tests, logically unlike other it cannot our inform evaluation. 92Although discussed, occasionally appears applied it that no court has it. 93 556, 566-567; Arriaga, Commonwealth v 438 Mass 781 NE2d (2003); Measuring underrepresentation, (proposing L see Yale J 1913 test). disparity-of-risk the use of analysis jury-selection system, The focuses on effects of the peremptory effects or for-cause strikes because effects of these equal-protection strikes on are resolved under analysis. Kentucky, 1712; See Batson v 476 US S Ct L 2d 90 Ed (1986). strategic light decisions voir dire no on shed Thus, representation in venires fair and at the reasonable outset. analysis given probability drawing considers the number from a randomly jurors drawing potential distinct when at a time. *30 Mich 575 Opinion the Court members of will include a defendant’s
chance that underrepresenta- given particular group the same alleged.95 tion employs the Jury n 25. This test poker, 8 St J Crim L at 537 See Ohio comparison. necessary probabilities The for theorem to obtain binomial percentage expresses the difference as a theorem in this situation binomial group expected of a distinct distribution be the normal between what would jury pool assuming the same juries in the is 12-person group 12-person community a actual distribution of in the and the as underrepresented pool. group in the
juries assuming
that the distinct
using
following
computed
available at
a binomial calculator
results were
chttp://
University Department
Statistics
Texas A&M
(accessed
www.stat.tamu.edu/~west/applets/binomialdemo.html>
June
drawn,
jurors
“p” representing
2012),
representing
number of
“n”
with
choosing juror
group in
the distinct
one
probability
of success
jurors
group on
representing
possible
from that
drawing, “x”
number of
(x)”
probability
number
jury,
representing the
of that
and “Prob
expected
probabilities
number of
resulting.
The results show
drawing
fully
12-person jury
group
if the
were
members of the distinct
= 0.0825)
expected
probabilities
number of
representative (p
for an
given
degree
12-personjury
the known
in a
members of the distinct
= 0.0417):
(p
underrepresentation in
case
*31
Opinion of the Court
Although
new,
this test is not
primary
the
reason for
its disfavor is because it has yet
garner approval
any
given
court.96 But
of uniformity
absence
for
what
apply,
tests to
we will consider it among other
measures of underrepresentation.
Its purpose
esti-
—to
mate the
probability
underrepresentation
actual
on a
jury
consistent with the United States Supreme
—is
protect
Court’s aims to
a
right to an impar-
jury
tial
and a fair trial by
jury
means of a
drawn from
fair
a
cross section of the community.97 Moreover, con-
sidering this test
is consistent with
holding
Smith’s
“[p]rovided
parties
proffer sufficient evi-
dence, courts should consider the results of all the tests
in determining whether
representation was fair and
Thus,
reasonable.”98
it is relevant to consider the extent
to which a defendant’s
a representative
chances of
by
were altered
pool
measuring the diminished
randomly
likelihood that a
12-person jury
drawn
a given
includes
number from a
example
results,
12-person jury
To take an
from these
a
drawn from a
pool proportionate
population
to the actual
of African-Americans
in Kent
(8.25
County
percent)
expected
would he
to have one African-American
percent
time,
12-person jury
38.4
pool
while a
drawn from a
containing
percent
expected
4.17
African-Americans would be
to have
percent
one African-American
31.32
of the time. For a more detailed
Jury poker,
description
of the binomial theorem see
mathematical
8 Ohio
St J Crim L at 537 n 25.
96
Arriaga,
Green,
566;
See
United States v
Supp
438 Mass at
389 F
2d
(D
Mass, 2005),
States,
grounds by
In re United
overruled on other
(CA
(D
2007).
1, 2005); Delgado,
Mass,
Supp
F3d
F
2d at 425
Florida,
See Williams v
78, 100;
399 US
90 S Ct
Opinion of the Court case, considering In this when group.99 distinct jury would 12-person that a defendant’s likelihood disparity of risk was no African-Americans contain percent.100 24.39 absolute-disparity
Unlike test, courts have not considered comparative-disparity disparity under which the appropriate threshold deemed fair and reasonable. We believe risk should be percent.101 drawn at 50 the normative line should be exceed 50 should is, percent That of risk that disparities logical deemed unfair and unreasonable. This is be measuring a defendant’s normative line because when percent of 50 or probabilistic injuries, disparity risk *32 not, the that, likely removing more than lower shows Re-justifying requirement, L cross section 116 Yale J at See fair already group (stating is a small of what “appreciably impact possibility’ a the defendant’s ‘fair of does not jury”). representative disparity and the actual risk We consider the between the ideal risk randomly 12-person jury having no African-Americans on a selected Thus, largest disparity. represents it where the under- because it is the expectations particular representation most affected the of a result. See Jury poker, J L at 540 n 28. 8 Ohio St Crim any randomly jury pool exactly 12-person In drawn drawn from a County’s population proportionate a to Kent African-American as whole (8.25 percent), expect a can no African-Americans on the defendant percent called the “ideal because it 35.59 of time. This is risk” probability particular underrepresen- measures the of a result without However, randomly drawing disproportionate jury tation. when from the (4.17 percent African-American), pool that occurred in this case probability 12-person containing a no African-Americans rises to percent. risk” it 59.98 This is called “actual because measures particular given underrepresentation. probability of a result the actual rate, expect percent representation With a 4.17 a defendant would 12-personjury percent on the time. have no African-Americans 59.98 test, thus, disparity-of-risk calculates the difference between the (35.39 (59.98percent), resulting percent) ideal risk and the actual risk disparity percent. risk of 24.39 threshold). percent (proposing See id. at 541-542 a 50 Opinion of the Court underrepresentation would not have altered the com- position of a jury.102 defen- Consequently, dant has failed to representation show that African-Americans was not fair and reasonable under the disparity-of-risk test.103 tests,
Given the results of the foregoing defendant has failed to show of African- Americans in the venires at issue was fair and reason- Instead, able. test, results the absolute-disparity test, comparative-disparity and disparity-of-risk test all support the opposite conclusion: the representation of African-Americans fair and reasonable. Accord- ingly, we conclude that defendant did not make out a prima facie case for his Sixth Amendment fair-cross- section claims. Notwithstanding our conclusion on this issue, determinative we will address third prong order to consider the argument that a defendant who systematic shows exclusion under the third prong is entitled to make a lesser showing under the second prong.
3. WHETHER UNDERREPRESENTATION RESULTS
FROM SYSTEMATIC EXCLUSION
The third Duren prong requires a defendant to show
that “this underrepresentation is due to systematic exclu-
sion of the
the jury-selection process.”104A
systematic exclusion is one that is “inherent in
par-
*33
102
See id.
(stating
“parallel[s]
commonplace
that such a
legal
line
rule
that
they
claimants are entitled to no relief
likely
when
fail to show it is more
they
wronged”).
than not that
have been
We also note that defendant’s risk
disparity
roughly
percent
percent
even falls below the 37
threshold
proposed by the
Measuring
underrep
author who first
introduced this test.
resentation, L
adopt
Yale
J at
percent
1936-1937. We do not
the 37
threshold because
doing
there is no normative rationale for
so.
103Although
dissenting justices question
disparity-
our use of the
test, they notably
of-risk
critique
make no substantive
of the test itself.
104 Duren,
ticular the practice concluded that Court Supreme United States nearly weekly venire every women excluding system- that was year constituted atic.106 program- that a computer here shows
The evidence used to ran- computer in the software ming error Secretary of from the jurors domly potential select in Kent eligible jurors State database of names from database of names County truncated that 118,169. The smaller list 453,414 eligible jurors randomly potential jurors. select names used to was included more list, however, disproportionately This codes and fewer from other zip certain individuals codes on underrepresented zip codes. The zip of African- concentrations higher had whole error, Thus, the which computer program Americans. exclusion, one systematic was the cause of was computer program, “inherent” in the which that was process utilized” jury-selection “the particular was jurors for service. potential select analysis purpose It for the is irrelevant not intentional and was cor- error was computer discovery prong because under third upon rected its in- demonstrates “systematic disproportion itself in jury the defendant’s interest chosen fringement of Thus, the fact community a fair cross section.”107 unintentional, and that it error was computer discovery, is immaterial to whether upon fixed its defen- at the time occurring exclusion was systematic Accordingly, we conclude dant’s was selected. 105Id. at 366. Id. at 366. added). (emphasis n Id. at 368 26 *34 Opinion of the Court by
defendant satisfied the third that the prong showing was exclusion systematic.108 Hubbard,
In panel of our Court of ad- Appeals dressed a fair-cross-section claim held that underrepresentation is threshold lower when the “the result underrepresentation is of circumstances less benign case, than random selection. . . In that .”109 on produced “[t]he evidence remand that the reveal[ed] juror process employed by County allocation Kalamazoo July before 1992—and not random selection —caused the underrepresentation.”110 The concluded that panel “given causation, the lack benign . . . the level of [absolute disparity disparity percent per- of 3.4 to 4.1 underrepresentation cent] constituted substantial un- der the Sixth Amendment.”111
In lowering the threshold in of the second prong circumstances which the level disparity result of nonbenign circumstances, panel the Hubbard erroneously assumed underrepresentation that the con- by the templated second Duren prong depends part on the reason for the underrepresentation. The reason for the underrepresentation is the third basis prong, only and the in the prong issue second is whether degree underrepresentation In acceptable. presented systematic Because defendant direct evidence of exclu sion, may we need address whether alone statistics establish that systematic was the result of a exclusion inherent jury-selection process. 109 Hubbard, minority App population 217 Mich in Hubbard at 480. The percent. panel only test, absolute-disparity was 7.4 considered but flawed, Osorio, largely relying Supp found the test United States on 801 F (D Conn, 1992), holding disparity 978-979 for its such a level of resulting nonbenign prong. circumstances satisfied the second Duren 110 Hubbard, App 217 Mich at 480. Hubbard, Id. at 481. Although panel given not addressed minority population in Hubbard was 7.4 percent, compara ranged disparity percent percent. tive from 44.6 to 55.4 491Mich Opinion op the Court requires dis- of three satisfaction words, Duren
other arbitrarily gives approach a de- prongs. An tinct prong the second the doubt on benefit of fendant the analysis. three-part can if a defendant Even vitiates the *35 systematic, a defen- that was show any underrepresen- the extent show that dant must Moreover,it would fair and reasonable. was not tation a certain level to conclude be inconsistent underrepresentation fair otherwise be that would suddenly systematic exclusion is absent reasonable underrep- of the the cause because fair and reasonable nonbenign. resentation is adopting Additionally, for rationale Hubbard’s approach is belied States v Osorio112 forth United set approach. Specifically, by case-by-case Hubbard our applying the absolute- about concerns articulated disparity minority in which test in a situation relatively population Smith, however, in- small. tests to be to limit the statistical courts not structs turning justification Thus, the considered.113 by case-by-case approach our diminished Osorio is evaluating result, As a because the relevant tests. all improperly approach second conflates the the Hubbard prongs its in Duren and because as forth and third set case-by-case unnecessary light of our rationale is reject approach, Hubbard to it and overrule we opinion. with this it is inconsistent extent that IV CONCLUSION presented defendant was the issue whether This case impartial right to an Amendment denied his Sixth 112Osorio, Supp F 966. concurring opinion’s Smith, disapproved endorsement of In we unnecessary to Hubbard, it was to reach the issue because hut declined Smith, n Mich at 205 1. the case. resolve Concurring Opinion Markman, J. drawn from a fair cross community. section of the Because we conclude that defendant did not establish of African-Americans was not fair and reasonable under prong second test, Duren we reverse the judgment of the Court of Appeals and reinstate defendant’s convictions and sen- tences. Mary
Young, C.J., JJ., Kelly, and Markman Beth J. ZAHRA, concurred with join J. I (concurring). majority
MARKMAN, opinion, reasonably which applies governing the Sixth Amendment’s “fair cross section” require- ment, as articulated by the United Supreme States Missouri, Court Duren v 357, 364; 439 US 99 S Ct (1979). L58 Ed 2d 579 I write separately only I because have questions concerning both Duren’s test and the constitutional standard toward which this test *36 is directed.
The Sixth guarantees Amendment criminal defen- dants the right “by a trial an impartial jury . ...” In Taylor Louisiana, 522, 526; 419 US 692; 95 S Ct 42 L (1975), Ed 2d 690 the Supreme Court determined that “the presence of a fair cross section community of the venires, on panels, or lists from which petit juries are drawn is essential to the fulfillment” of this constitu- tional guarantee. The “fair cross section” requirement is satisfied as long as “distinctive” groups are reason- ably represented on the jury venire; however, it does not entitle a jury defendant to a whose composition is proportional to that group’s presence within the com- munity from which the venire is Taylor chosen. As emphasized: any particular
Defendants are not entitled ato of composition, jury wheels, names, but the pools panels, of or 491 Mich by Opinion Concurring Markman, J. systemati- juries are drawn must
venires from which
community and
cally
groups in the
exclude distinctive
reasonably
at
representative
[Id.
thereof.
thereby fail to be
omitted).]
(citations
it
unneces-
analysis,
section”
Under the “fair cross
the lack of “reasonable
a defendant to show that
sary for
of
is the result
“distinctive”
representation”
would be
jury-selection system,
in the
as
discrimination
of the Four-
Equal
under the
Protection Clause
required
Partida,
See,
e.g.,
teenth Amendment.
Castaneda
(1977). Rather,
482, 494;
Ct
(1) group alleged that the to be excluded is a “distinctive” (2) group community; in the representation that the of this group juries in venires from which are selected is not fair and reasonable in persons relation to the number of such (3) community; systematic due to group jury- exclusion of the in the process. selection [Id. at 364.] The dispositive question in this case concerns the second part i.e., of Duren’s whether the represen- test — tation of African-Americans in venires from which juries were selected in County Kent during the period which defendant was tried and convicted is “fair and reasonable in relation to the number of such persons the community.” Id.
To determine whether representation is “fair and reasonable” test, under the Duren courts have applied yet more In tests. People Smith, 463 Mich NW2d 1 (2000), this Court discussed three statistical tests that have been used to measure whether repre- sentation of a “distinctive” in the jury pool is “fair and reasonable”: the “absolute disparity” test, the “comparative disparity” test, and the “standard devia- tion” test. Recognizing that all three tests are imperfect and susceptible criticisms, Smith held:
We thus approaches consider all these to measuring representation whether reasonable, was fair and and con- clude that no individual method should be used exclusive of Accordingly, the others. adopt case-by-case we approach. parties Provided proffer evidence, that the sufficient courts should consider the results of all the determining tests in was fair and reasonable. [Id. at 204.]
After a decision on habeas corpus by the review United States Court of Appeals for the Sixth Circuit asserting that Smith constituted an “unreasonable” application of “clearly established law,” federal Smith v *38 491 MICH Opinion by Concurring Markman, J. (CA 2008), the 326, 329, 543 F3d
Berghuis, unanimously reversed Court Supreme United States Duren nor any Circuit, that “neither holding the Sixth test the method or specifies this Court decision of other of dis- representation the must use to measure courts Smith, US Berghuis groups injury pools.” tinctive (2010). L Ed 2d 249 1382, 1393; 176 314, 329; 130 S Ct Supreme the imperfect,” test is Noting “[e]ach or today on the method “to take sides Court declined appropriately methods which measured.” Id. at 329-330. law, majority opinion I the join of the this state
Given
application
in a reasoned
engages
it
because
Court
Supreme
United States
decisions of the
relevant
approach
outlined
Consistently
this Court.
with
Smith,
majority opinion
considers
in our decision
parties
which the
have
of all three tests for
the results
determining
repre-
proffered evidence
African-Americans,
the “distinctive”
sentation
County
in Kent
“fair and reasonable”
question,
considers
majority opinion
Specifically,
venires.
“comparative
and the
disparity”
results of the “absolute
test,
of an additional
tests, as well as those
disparity”
test,1 and concludes
of risk”
“disparity
finding
to warrant
results are insufficient
during
in the venires
African-American
“fair and reasonable.”
was not
period
the relevant
facie
prima
establish a
Thus, defendant has failed to
“fair cross section”
Sixth Amendment’s
violation of the
Appeals’ opinion
Although the Court
requirement.
also, in my
of this Court
dissenting opinions
and the
majority
analyze
the “standard
opinion
the results of
does not
only expert
the trial court found
whom
deviation” test because
appropri
credible,
Stephenson,
that the test was “not
Dr. Paul
testified
present circumstances.
ate” in the
Concurring Opinion by Markman,
J.
Duren,
apply
reflect reasonable efforts to
judgment,
“comparative
the results from the
only
their use of
ascertain a “fair cross section” viola-
disparity” test to
tion,
multiple
to use data from
their decisions
time,
fully
and their decisions not
venires over
disparity”
consider the results of the “absolute
Smith,
See
prefer
majority’s analysis.
cause me to
(“[N]o
In
I
particular,
am concerned about the statistical
part
tests used to determine whether Duren’s second
has been satisfied. The limitations of these tests have
noted, see,
widely
e.g.,
been
559 US at
Berghuis,
603-611;
637-638,
ante at
at
and need
post
not be
when,
say
here,
revisited here. It suffices to
as
comprise only
members of the “distinctive”
a
relatively
community’s
small
percentage
jury-
eligible population,
arguably
one test
makes it difficult
satisfy
requisite showing
for defendant to ever
“underrepresentation,”
arguably exagger-
another
“underrepresentation,”
ap-
ates this
and the third
to be
it
pears
generally disfavored because
does
anything
constitute an
measure of
obvi-
appropriate
ously relevant
to a determination
the level of
Mich 575
J.
Concurring
by
Opinion
Markman,
was “fair and reason-
representation on
venires
deficiencies,
In
do the results of
light
able.”
of these
how
tests,
individually
either considered
or collec-
these
tively, usefully
illuminate whether
“fair and reasonable”? How do the bench and bar draw
meaningful
legal
application
conclusion from
statistical data?
ratio-
these tests to
available
How
nal,
flexible,
are the
thresholds that
how
statistical
courts in
by
distinguishing
have been established
some
“underrepresentations”
compatible
between
are
section,”
with a “fair cross
and those that are not? To
extent,
any, may
what
if
these thresholds be raised or
lowered, as a
degree
any
function of
nature or the
Duren,
“systematic
part
exclusion” under the third
parts
analyzed entirely discretely?
or must these
be
To
extent, if any,
what
should these thresholds be raised or
Amendment
lowered
Fourteenth
considerations of
intention,
discriminatory purpose or
or are those con-
siderations simply irrelevant
to the “fair and reason-
analysis
able”
under the Sixth Amendment? To what
extent are these thresholds emblematic of what
predicted
dissent Duren
would become a mere “con-
Duren,
stitutional numbers game,”
However, an even more fundamental why tion is also raised certain statistical tests here— Supreme and not others? The United States Court has acknowledged that the three tests described in Smith imperfect, Berghuis, are each 559 US at and has today declined “to take sides on the method or methods by appropriately which mea- sured,” id. at Doubtless, there is no end to 329-330. might statistical tests which a court seek to compare populations various-sized of “distinctive” groups community within a and their Equally on doubtless, venires. as evidenced in this very case, can tests be devised that will tend both sustaining repudiating finding toward a of “un- derrepresentation.” “disparity Is the new of risk” test genuinely adducing valid means the existence of dissenting violation, Sixth Amendment or are the justices improves correct it “neither nor *41 491Mich 575 Opinion by
Concurring Markman, J. area of the Post at 633. What are the clarifies this law”? this Court can discern which tests standards which identifying Amendment viola- are relevant in Sixth And is the relevance of the fact that some tions? what regarding might point in one direction second tests point opposite part, might in and others Duren suggest asking direction? Does this that these tests are answering questions, different or that one test is and asking answering wrong question? do
and How judges right question test the tests to ensure being results, differ in their how are asked? Wdientests answering these results to be reconciledin the ultimate May compare question? the court constitutional degree or extent to different contrast which tests distinguishing acceptable deviate from thresholds disparity? unacceptable levels of Is the court statistical simply among free to choose at its discretion such conflicting tests? If there is some actual decision- making selecting among conflicting tests, standard anything it? If has what is such standard do with determining which test better identifies “fair and rea- groups sonable” “distinctive” ve- tautology, nires, wit, then not a Catch-22 is this choosing among identify tests best absence of representation, “fair and reasonable” a court must employ “fair and the test that best identifies reason- representation? multiple applied, if able” What tests are produce split case, as in the instant and these results of plaintiff 2-1 or 3-1 or 7-6 in favor of the or “majority requires defendant? Is there some rule” that that we resolve conflicts favor outcome of the majority applied? so, of statistical tests If does this not all-important render initial court’s determination of going considered, which tests are to be and how that is “majority And if the rule” to be determined? does not distinguish among apply, conflicting do courts tests how Concurring Opinion by Markman, J. *42 in determining which of these will be dispositive concluding that the Sixth Amendment has or has not been breached?2
These and related questions
meaning
concern the
Smith’s
directive
courts must “consider all.
. .
approaches to measuring
representation
(em-
Smith,
fair and reasonable . . .
Mich
at 204
added). Indeed,
phasis
very different conceptions of this
obligation are reflected in the
and
majority
dissenting
opinions.
argues
Justice MARILYN KELLY
that the Court
of Appeals
“properly
below
considered the results of all
tests
[including
absolute-disparity test], but decided
the comparative disparity test was ‘the
ap-
most
propriate test to measure underrepresentation
in this
”
case,’
post at
quoting People Bryant,
Mich
App 260, 271;
(2010),
certainly
by
concept
“underrepre-
the Duren
implied
sentation”)3 and then
some statistical
apply
equivalent
I
it”
once articulated
of the “I know it when see
Supreme
United States
Court Justice Potter
former
Ohio,
obscenity
in the realm of
law. Jacobellis v
Stewart
(1964)
184, 197;
S Ct
That
even
could
answer each of
which
to
questions,
aforementioned
knew
tests
results,
to their
give legal import
“consider”
how
be difficult to
Duren because the
apply
it still would
ultimate
it
directed
constitutional standard
which
I
remains unclear.
know what the constitutional stan-
dard in “fair cross section”
not. It is not an
cases is
*43
under
equal-protection
any “underrep-
standard
which
*
Duren,
J.,
(Rehnquist,
dissenting)
n
See also
used as a
or
substitute
votes.
Stewart,
recognize
As did Justice
I also
that both this
and the
Court
Supreme
quite possibly
are
faced here
United States
Court
with “the task
Jacobellis,
trying
may
at
to define what
be indefinable.”
378 US
(Stewart, J., concurring).
by
Concurring Opinion Markman,
J.
resulting
resentation”
from intentional or purposeful
in the jury
prohibited by
discrimination
venire is
Taylor,
526-528; Duren,
Constitution. See
because there is also
proportion-
a venire must come to
to measure how close
proportionality.
or indeed even how one measures
ality,
are
nondiscriminatory, deviations
“systemic,” but
What
Constitution,
“sys-
and
under
what
acceptable
are not?
temic,”
nondiscriminatory,
deviations
but
and the
lack of clear external standards
Given this
available,
divergent
statistical measurements
wealth
her
private
ensure that his or
own
judge
how can
is “fair and reasonable”
concerning
sensibilities
what
not come to
over
prevail
of the venire do
make-up
and that
what is
the Constitution
statistical
required
selected, and
tests do not come to be
standards
not come to be
evaluating
adopted,
their results
merely
Focusing
tend to match those sensibilities?
exclusively on the merits of the various tests obscures
trees,
clear sense of
the forest for the
for without some
guarantee
“impartial
the constitutional
of an
what
difficult,
it
jury” requires
stage,
prove
at the venire
will
uniformity
analyses
if
to achieve
in the
impossible,
This,
turn,
of different venires.
composition
re-
making judicial
incurs the risk of
determinations
juries” fraught
partiality
with
garding “impartial
credence to
gamesmanship,
lending
mathematical
while
about a “constitutional
Rehnquist’s
Justice
concerns
J.,
Duren,
In the trial and have appellate case, reviewed this and the whether the venire question closely here was “fair and reasonable” has divided them I negative response. in favor of a While have no doubt question these addressed the judges that each of has reasonably,” and in accordance “fairly this case Duren, there understanding with his or her own best way seems to be little in the of a coherent constitutional exclu- distinguishes “systematic standard that between *45 People Bryant v Dissenting Opinion by Cavanagh, J. that violate the and “systematic sions” Constitution not, exclusions” that do much a clear less statistical giving method for effect to this constitutional standard. result, a decision-making And I believe that our in as uncomfortably judicial this realm resembles a Ror- test, in judge ultimately schach which the to required is in determining look inward what is “fair and reason- able,” rather comprehensible than outward to a consti- majority tutional rule of law. In and joining opinion, what I to be despite concerning believe confusion as- test, pects the Duren I sought my have to the best of is understanding required by of what the United States Court Supreme give to reasonable to this test meaning guarantees to the Constitution. I (dissenting). J. be- CAVANAGH, respectfully dissent cause I do not think that the Court of Appeals clearly by erred concluding that defendant is entitled to newa trial under the unique presented facts this case.
I agree with Justice have KELLY courts MARILYN always applied Missouri, Duren v 357; 439 US 99 S 664; L (1979), Ct 58 Ed 2d precision with and that reasonable disagree regarding minds can the proper application Regardless Duren. of the debate raised Kelly case, however, I agree with Justice relief, defendant is nevertheless to if entitled even evaluating broader time frame for Duren's second prong is considered. explained People Smith, 216,
As I
463 Mich
(2000)
222;
(CAVANAGH, J.,
showing none particularly are analysis methods of well-suited ahead” Duren’s case,” “glance I courts should believe evidence of third and consider a defendant’s prong Mich at Smith, 463 222. Under systematic exclusion. *46 process appears if the approach, jury-selection this that group, to exclude a distinctive likely systematically is, the mark of a non- jury-selection process the bears a the benign influence, may give a court defendant 218, of on Id. at underrepresentation. benefit the doubt case, 222-224. this to the facts this approach of Applying of agree with KELLY’S conclusion the Court Justice I clearly holding err that defendant is Appeals by did not entitled a new trial. to hastily adopt “dispar-
The decision to the majority’s when, of has also me as ity given pause, risk” test observes, was aptly Justice KELLY the test not addressed courts, argued or this by lower not briefed to Court, concedes, and, not been majority as has Thus, in country. endorsed court because the by any of the merits test and disparity-of-risk substantive presented were not to majority’s percent threshold1 Court, pass I decline to on the merits of judgment briefing the test at this time without the benefit of full and oral argument. I dissent.
Accordingly, respectfully I J. concur with the (dissenting). Marilyn Kelly, majority defendant satisfied the first and third of for a prongs the test fair-cross-section violation 1 Indeed, argued by from the fact that the issue was not raised or aside case, parties given majority’s in this conclusion that defendant’s by disparity proposed risk below even the threshold the author first falls introducing test, disparity question of risk ante at 615 n I see necessary adopt higher it threshold in this case. People Dissenting Opinion by Marilyn Kelly, J. ,1 Sixth However, Amendment under Duren v I Missouri disagree defendant failed to meet the second prong Accordingly, of Duren. I respectfully dissent from the majority’s the judgment decision reverse of the Court Appeals.
The Court Appeals and, used law existing part, most it applied majority The im- properly. opinion Worse, error putes where there is none. it sua sponte introduces a “disparity accepted any risk” test not in country. analysis court majority set has forth today improves neither nor clarifies this area of the law.2
The majority concludes that the Court of Appeals in erred ways two when evaluating the second prong First, Duren. it considered the representation African-Americans only venire, multiple Second, venires over time. it our misapplied by “effectively decision Smith3 adopting] bright-line rule favor of the comparative-disparity *47 in all test instances in which the the population of distinct is group small.”4 With regard the first contention, majority the does not make a persuasive case that the Court of Appeals erred. And the second allegation of error is based on inaccurate statement of what the of Appeals Court did.
I with generally agree that, the majority under Du- ren, courts must consider the composition of venires 1 Missouri, Duren v 357, 364; 664; 439 US 99 L S Ct 58 Ed 2d 579 (1979). 2 altogether unnecessary. jury-selection It is also process The error in the August at issue here occurred between June 2001 2002 as a result of a computer Accordingly, programming raising error. the number cases appellate likely issue that remain pipeline majority’s in the is low. The today dressed-up correction, pure simple. decision is error 3 Smith, (2000). Mich 463 NW2d 1 615 4 Ante at 582-583. 491 Mich Kelly, J. Opinion by Marilyn Dissenting specific the considered Appeals The Court
over time. venire when of defendant’s composition the disparity considered And it prong. the Duren evaluating second the third applying time when venires over multiple this was erro- persuaded am not I prong.5 Duren neous. relies on under- authority majority the the
Some
if a
For
distinctive
example,
mines its conclusion.
particular
defendant’s
in a
underrepresented
not
Amendment claim.
venire,
cognizable
no
Sixth
there is
in a
underrepresentation
Thus,
group’s
a distinctive
component
necessary
venire is a
particular
Accordingly,
composi-
claim.
of a
Amendment
Sixth
venire must be exam-
particular
tion of a defendant’s
analysis.
Duren
Yet the
at some
point
ined
effectively ignores it.6
majority’s approach
do not conclu-
majority
cases
cites
Other
composition of multiple
demonstrate that
sively
be
under
second
venires over time must
considered
contrary,
than the third. To
prong
Duren
rather
propo-
cases seem to stand for the uncontroversial
these
sition that
defendant must show
Duren
time to
multiple
satisfy
venires over
Indeed,
majority’s quotations
one of the
generally.7
(2010)
273-275;
App 260,
People Bryant,
Mich
796 NW2d
prong using
(applying
third
data and
for venires over
Duren
statistics
period).
a three-month
majority
approach
into
that its
takes defendant’s venire
asserts
by “including
it in the data set of
used to calculate
account
venires
disparity
underrepresentation.”
n
degree
601 63. But
in the
Ante at
particular
composition
shown
order to
of a defendant’s
venire must be
harmed because a constitutional
demonstrate
the defendant was
particular
actually
Accordingly,
disparity
in that
occurred.
violation
lumped
independently,
simply
in with
must be
venire
considered
*48
concerning other venires.
statistics
7
(CA
(“[A]
Miller,
1219, 1228
9, 1985)
See, e.g.,
v
771 F2d
United States
premised
requirement
upon
cannot be
the
cross-section
violation
fair
of
People Bryant
635
v
Dissenting Opinion by Marilyn Kelly, J.
s.8
supports
analysi Moreover,
Duren
cases not cited
by
majority
the
contradict
its conclusion and suggest
that multiple venires over time are relevant
to the
question
systematic
of
exclusion rather than the ques
tion
underrepresentation.9
of
proof
underrepresentation
added);
single jury.”) (emphasis
of
United
(CA
Allen,
1096,
6, 1998) (“Appellants
States v
F3d
160
1103
have satisfied
prong
test,
they
the
[of]
first
the Duren
hut
have not
other
satisfied
two.”)
added).
(emphasis
majority
“compels”
asserts that
itself
Duren
analysis
majority’s
However,
prong.
of the second
at
Ante
599-600.
quotations
Allen,
authority
coupled
from Miller and
with the
that I cite in
9,
applied
precision.
footnote
demonstrate that courts have not
with
Duren
Thus,
majority’s analysis
foregone
is far from a
conclusion. It is also
telling
majority
that
identifies few cases decided since
that
1979
significance
plural
attribute such
to the
of
use
“venires” in Duren’s
second-prong analysis.
of the
discussion
The reliance in
on
these cases
plural
by
use of
subsequent
is also undercut
statements in
caselaw from
Supreme
See, e.g.,
Illinois,
474,
the United States
Court.
Holland v
493 US
(1990) (“It
478, 480;
803;
long
110 S
Ed
Ct
107 L
2d
has
905
been established
groups
that
jury
racial
cannot be excluded from the venire from which a
is
selected____[A]
by
requirement
imposed
fair-cross-section venire
is
Amendment[.]”)
added).
(emphasis
Sixth
8
600,
Duren,
quoting
(“Finally,
See
at
ante
636 Opinion by Marilyn Dissenting Kelly, J. majority the that I with agree do Consequently, “evaluating that proposition for the Duren stands is fair of a distinct whether composition venire over evaluating requires reasonable evaluating sys I Rather, time.”10 believe pat a looking for requires exclusion occurred tematic underrepresentation over time. This makes tern of in a jury- a flaw because to show constitutional sense the a defendant must show system, selection unrepresentative leads venires. consistently system system’s of the Indeed, consistency it is the general in proves venires that produce representative failure to 11 systematic exclusion. using err the Thus, Appeals of did not the Court venire —73.1 comparative disparity defendant’s Duren12 prong the second of applying percent —when in the disparity it the appropriately And evaluated applying of over time when composition racial venires argument single under-representation in a venire. This is inferred from the (CA 1982) 8, Jones, 1265, merit.”); v United States 687 F2d 1269 without (“Even met, assuming requirements two been there is no that the first have procedure.... systematic of in the No evidence exclusion selection composition regarding the of other venires the evidence was introduced district.”). 10Ante at 602. 11 (“[T]hat Duren, just large discrepancy at occurred not 439 US year occasionally weekly nearly every period for a of but venire manifestly indicates that the cause of .”). systematic .. . composition disparity Other courts also examine the between racial composition community and defendant’s venire when of the See, Hester, prong. e.g., applying SW3d Duren State second 2010) by calculating (Tenn, prong (applying the Duren
42-44 second composition county’s population disparity between racial venire); Bowman, makeup at 467-468 racial NC (same). ¶ 31; (same); Holland, A2d at 2009 Me at 237-238 event, opinion, any explained later I conclude In for reasons defendant established a Sixth Amendment fair-cross-section viola- percent. comparative disparity if I the three-month of 49.4 tion even use Dissenting Opinion by Marilyn Kelly, J. prong. majority’s assignment the third Duren The first error therefore without merit. majority’s Ap- second criticism of the Court peals’ opinion supposedly is that it established “a bright-line favoring comparative-disparity rule disregarded test” and the results of the other tests.13 Respectfully, majority I believe that the misreads or Appeals’ opinion. mischaracterizes the Court *50 Appeals The Court of did consider the both results of absolute-disparity test and the standard-deviation unhelpful resolving test, but found both appeal.14 majority The error, determines that this was why explanation repeats but offers little this is so. It that Smith mandated that courts consider the results of determining all the tests when whether the second prong Appeals of Duren is met. But the Court of specifically recognized requires that Smith an such approach analyze and did of results each test. It simply comparative- found the results of one test—the disparity helpful. test—most It does not follow that ultimately Appeals because the Court of settled one on meaningful, test as most it relied on that test without considering Nothing supports majority’s the others. sweeping Appeals assertion that the Court of estab- bright-line comparative- lished a rule in favor of the disparity test.
13Ante at 606-607.
“
Bryant,
App
See
(concluding
289 Mich
at 269
‘the absolute
disparity
acceptable disparity’
test is an ineffective measure of
because of
percentage
the low
of
eligible
African-Americans who were
to vote in
County”
Kent
“declining]
and for that reason
to find the absolute-
disparity
case”),
controlling
quoting People
(After
test
in this
v Hubbard
Remand),
(1996)
459, 477;
(citation
App
217 Mich
I my analysis dispels believe that majority’s findings of error Appeals Court of and demon- strates that the Court Appeals correctly analyzed majority’s case. The sole remaining reversing basis the Court of Appeals’ judgment is its disagreement with that court’s reliance on (After Hubbard Re- mand.j.18 I disagree Hubbard should be partially overruled. Hubbard,
Under may a court consider the reason for a systematic exclusion when deciding whether repre- sentation of the distinctive group was fair and reason- If a jury-selection able. process systematically excludes distinctive on the basis of nonbenign factors, a may give court a defendant the benefit of the doubt on underrepresentation.19 Hubbard borrowed this ap- proach from United States v Biaggi20 and United States v Osorio.21Although Biaggi and Osorio are not binding Court, on this they persuasive. Moreover, are contrary 17 propriety us, properly Because the of this test is not before I decline majority’s indulge critique invitation to in a substantive of it. See ante at n 37 103. 18 Hubbard, App 217 Mich 459. 478, Id. at 481. (CA 1990). Biaggi, United States v 2, 909 F2d Osorio, (D 1992). United States v Conn, Supp F *52 491 MICH575 640 by Marilyn Opinion Dissenting Kelly, J. “vitiates majority’s approach that this to the conclusion endorsed analysis,”22 other courts have the three-part and third Duren merges the second analysis that in this fashion.23 prongs as Justice Hubbard, I would follow Accordingly, in concurring opinion advocated his CAVANAGH benefit of Thus, I defendant “the give Smith.24 ahead” “glance underrepresentation” on doubt Smith, the defendant Unlike prong.25 the third Duren exclu- systematic here has established defendant error that caused computer occurred because the sion pro- the jury-selection “inherent” in the exclusion was Hubbard, “did result As the exclusion cess used. instead, selection, but, resulted ‘benign’ random pro- juror inherent allocation from a defect . . . .”26 cess systematic exclu- defendant established
Because “the showing underrepresentation occurred and sion has close,”271 conclude that defendant established Amendment’s fair- facie violation of the Sixth prima identifies prosecution requirement. cross-section no state interest that was advanced significant African- excluded systematically selection process 22 Ante at 618. 23 (D (“[T]he 1995) Rioux, 1558, 1566 Conn, Supp States v United 930 F test, of the Duren prongs unfair second and third exclusion, inextricably.”); Commonwealth systematic are intertwined (2003) (“Evidence
Arriaga,
of a
438 Mass
Finally, sight we must not lose of the fact right at issue here —the to right jury a trial —is the cornerstone of the justice system.28 American right- adjudged by be a jury of one’s peers precious part a right.29 The majority’s careless decision imposes new, wholly unnecessary restriction right by on this creating error where there is none and new law that no party has advocated.
For reasons, these I believe that the of Appeals Court correctly reversed defendant’s convictions re- manded for a I new trial. affirm would its judgment. Marilyn J., J.
Hathaway,
Kelly,
concurred with
suffrage
people’s
“Just as
ensures the
ultimate control
in the
legislative
branches,
and executive
trial is
meant
ensure their
judiciary.” Blakely Washington,
296, 306;
control in the
542 US
124 S Ct
(2004).
2531;
L159 Ed 2d 403
29 “Providing
right
peers
an accused with the
to be tried
of his
gave
safeguard against
him an
corrupt
inestimable
or overzealous
prosecutor
against
compliant, biased,
judge.”
or eccentric
Dun
(1968).
Louisiana,
1444;
can v
391 US
S88 Ct