HEINONLINE
Citation:
Gennaro F. Vito; Thomas J . Keil, Capital Sentencing in
Kentucky: An Analysis of the Factors Influencing
Decision Making in the Post-Gregg Period, 79). Crim.
L. & Criminology 483 (1988)
Provided by:
Available Through: University of Louisville Brandeis School of Law Library
Content downloaded/printed from HeinOnline
Wed Nov 21 14:19:19 2018
-- Your use of this HeinOnline PDF indicates your
acceptance of HeinOnline's Terms and Conditions
of the license agreement available at
https://heinonline.org/HOL/License
-- The search text of this PDF is generated from
uncorrected OCR text.
-- To obtain permission to use this article beyond the scope
of your HeinOnline license, please use:
Copyright Information
io] Use QR Code reader to send PDF
to your smartphone or tablet device
0091-4169/88/7902-483
Tue Journat. or Crimtwa Law & GrimtnoLocy Vol. 79, No. 2
Copyright © 1988 by Northwestern University, School of Law Printed in U.S.A.
CAPITAL SENTENCING IN KENTUCKY:
AN ANALYSIS OF THE FACTORS
INFLUENCING DECISION MAKING
IN THE POST-GREGG PERIOD*
GENNARO F. VITO, PH.D.**
THOMAS J. KEIL, PH.D.***
I. BackGROUND AND PuRPOSE
The reimposition of capital punishment by the United States
Supreme Court in 1976 has resulted in the execution of ninety pris-
oners, as of September 1, 1987.! The use of the death penalty con-
tinues despite research evidence from several states which has
consistently demonstrated that offenders who have white victims in
general, and black offenders who kill whites in particular, are more
likely to face a “death qualified” jury and receive a death sentence.”
For this reason, the growing use of capital punishment and its sen-
* The authors wish to express their appreciation to the following persons who
offered important suggestions in the development of this Article: David Baldus of the
University of Iowa, Kevin McNally of the Kentucky Department of Public Advocacy,
Leslie Bienin of the New Jersey Department of the Public Advocate, C. Bruce Trauber
and Penny Warren of the Kentucky Office of the Attorney General, and Wayne Usui of
the University of Louisville. We also express our thanks to John Sternberg who
diligently and carefully collected data, to Betty Lou Vaughn and her staff in the
Kentucky Corrections Cabinet, to Pat Sims of the Kentucky Administrative Office of the
Courts, and to Dr. Deborah Wilson and her staff in the Research and Evaluation Unit of
the Kentucky Corrections Cabinet.
The authors would also like to express their appreciation to the following University
of Louisville Urban Studies Center staff members who provided consultation and
support services to the project: Patricia Hardyman, Jack Ellis, Dr. Jerry Hutchins,
Vernon Smith, Nancy Roseberry, Lisa Johnson, Monica Bowles, and Elizabeth Jones.
Points of view and opinions expressed in this article are those of the authors and do
not necessarily represent the official position of the University of Louisville or the
Commonwealth of Kentucky.
** Associate Professor, School of Justice Administration, University of Louisville.
Ph.D., The Ohio State University, 1978.
*** Professor and Chair, Department of Sociology, University of Louisville. Ph.D.,
Temple University, 1972.
1 NAACP Legal Defense Fund, “Death Row, U.S.A.,” September 20, 1987.
2 See generally H. Bepau, DEATH Is DirFERENT 182 (1987).
483
484 VITO AND KEIL [Vol. 79
tencing system is a central concern in policy research.3
The research findings raise significant constitutional issues. Ra-
cial discrimination in capital sentencing has long been a central con-
cern of the Court. For example, in Furman v. Georgia,* which
temporarily outlawed executions, several Justices cited racial dis-
crimination and the arbitrary use of capital punishment.® In Gregg v.
Georgia,® the Court approved Georgia’s “‘guided discretion” statute
which outlined several procedures designed to prevent the abuse of
discretion and to exclude ‘“‘extra legal’’ factors, such as race, from
the capital sentence process.? Yet, post-Gregg studies clearly and
consistently demonstrate that the use of such statutorily approved
guidelines ‘‘may operate to promote, rather than to curb, arbitrari-
ness and discrimination early in the handling of potentially capital
cases.”
In fact, research conducted by Baldus® became the central issue
in the recent landmark decision McCleskey v. Kemp, in which the
defendant, Warren McCleskey, was a black Georgia inmate sen-
tenced to death for the murder of a white police officer during an
armed robbery.!! The Baldus research introduced statistical proof
demonstrating that the Georgia capital sentencing process was ad-
ministered in an discriminatory, arbitrary, and capricious manner.!2
Justice Powell, writing for the five Justice majority, assumed the va-
lidity of the Baldus study but considered this evidence insufficient to
demonstrate either unconstitutional discrimination or arbitrary and
capricious sentencing.!3 He considered the research an indication
of a “discrepancy” correlating with race but noted that it failed to
3 See, eg., Cheatwood, Capital Punishment and Corrections: Is There an Impending Crisis,
31 Crime & DeLing, 461 (1985).
4 408 U.S. 238 (1972).
5 See id. at 241, 249-50, 251 (Douglas, J., concurring); id. at 364-65 (Marshall, J.,
concurring).
6 428 U.S. 153 (1976).
7 Id. at 197-98. The Georgia statute provides for a bifurcated trial, an outline of the
aggravating and mitigating circumstances surrounding the offense and the offender for
consideration by the jury during the penalty phase of the trial, and an automatic appeal
to the state supreme court. Ga. Cope ANN. § 27-2537 (1983).
8 Bowers, The Pervasiveness of Arbitrariness and Discrimination Under Post-Furman Capital
Statutes, 74 J. Crim. L. & Criminotocy 1067, 1071 (1983).
9 Baldus, Pulaski & Woodworth, Comparative Review of Death Sentences: An Empirical
Study of the Georgia Experience, 74 J. Crim. L. & CriminoLocy 661 (1983) [hereinafter
Baldus].
10 107 S. Ct. 1756 (1987).
11 Jd. at 1761-62.
12 Jd. at 1763-64.
13 Id. at 1777-78.
1988] CAPITAL SENTENCING IN KENTUCKY 485
demonstrate “major systemic defects.” !4
Despite McCleskey, the impact of race upon the capital sentenc-
ing process is still a profound issue. If the Baldus study is replicated
in other states, either more evidence of racial discrimination will be
uncovered or a state system free of such problems will be revealed.
Perhaps, a refinement in the eligibility criteria of capital sentencing
can produce a more racially equitable result.
This Article focuses upon the capital sentencing process in
Kentucky, and examines the central questions of whether the death
penalty is applied in an arbitrary or discriminatory fashion and
whether such a system can operate in a manner consistent with the
requirements of the United States Constitution, especially the due
process clause of the fourteenth amendment. This Article extends
the post-Gregg analysis of the factors influencing the capital sentenc-
ing decision to another location and adds to the literature on this
crucial subject.
Il. Lirerature Review
Using landmark Supreme Court decisions as benchmarks, stud-
ies that have examined the capital sentencing process were clearly
divided into different frames of reference. The research of the pre-
Furman period clearly revealed a pattern of discriminatory sentenc-
ing by race of the offender. This research typically focused upon the
race of the death row inmate, examining the result of the capital
sentencing process.
A. PRE-FURMAN RESEARCH
In one study, Guy Johnson investigated 220 cases from Rich-
mond, Virginia occurring between 1930 and 1939 and 330 murders
in five North Carolina counties between 1930 and 1940.15 He dis-
covered that only eight out of 141 black offender/black victim cases
received a life sentence (5.6%), and none received a death sen-
tence.!6 However, in twenty-two cases involving blacks who killed
whites, seven defendants received life sentences (31.8%), and six
were targeted for execution (27.2%).!7 Similarly, Garfinkel ex-
amined racial patterns in the charge, indictment and conviction
stages of the capital sentencing process in ten North Carolina coun-
44 Td. at 1777.
15 Johnson, The Negro and Crime, 217 ANNats 93 (1941).
16 Id. at 99.
17 Id.
486 VITO AND KEIL [Vol. 79
ties over a ten year period.'8 He also determined that blacks who
killed whites were more likely to be charged, indicted, and convicted
of first degree murder. His conclusion was that such offenses were
considered the most “heinously murderous” of the possible of-
fender/victim combinations.'° Elmer Johnson studied a North Car-
olina database of 660 offenders sentenced to death since 1909 and
reported that blacks clearly accounted for the majority of death row
inmates (approximately seventy-four percent).2° This research base
demonstrates that blacks who killed whites were specifically targeted
for capital punishment.
Race was also a factor in the decision to commute death
sentences. In an examination of 439 Pennsylvania death row in-
mates the years 1914 to 1958, Wolfgang, Kelly, and Nolde reported
a statistically significant difference between the proportion of blacks
(eleven percent) and whites (twenty percent) who had their
sentences commuted.?! They also reported the execution of ninety-
four percent of the black as compared to eighty-three percent of the
white Pennsylvania death row convicts.2?
In addition, racial differences influenced the death penalty deci-
sion in crimes other than murder. The race of the offender was a
significant factor in the capital sentencing of rapists. For example,
Wolfgang and Riedel examined over 3,000 rape convictions ob-
tained in 230 counties in eleven southern states over a twenty year
period.23 They discovered that blacks who raped whites were eight-
een times more likely to receive a death sentence than any other
offender/victim combination.?4 The authors concluded that the ra-
cial factor between the defendant and the victim resulted in a “‘pat-
terned, systemic and customary imposition of the death penalty.”’5
Finally, using the first 204 homicides reported to the police in
Philadelphia in 1970 as a basis, Zimring, Eigen, and O’Malley re-
ported that blacks who killed whites were most likely to receive
either a life or death sentence.26 In fact, blacks with white victims
18 Garfinkel, Research Note on Inter and Intra-racial Homicides, 27 Soc. Forces 369
(1949).
19 Id. at 376.
20 Johnson, Selective Factors in Capital Punishment, 36 Soc. Forces 165 (1957).
21 Wolfgang, Kelly & Nolde, Comparisons of the Executed and the Commuted Among Admis-
sions to Death Row, 53 J. Crim. L. & Criminotocy & Potice Sct. 301, 306 (1962).
22 Id.
23 Wolfgang & Riedel, Race, Judicial Discretion and the Death Penalty, 407 ANNats 119
(1973).
24 Id. at 307.
25 Id. at 419.
26 Zimring, Eigen & O'Malley, Punishing Homicides in Philadelphia: Perspectives on the
Death Penalty, 43 U. Cut. L. Rev. 227, 232 (1976).
1988] CAPITAL SENTENCING IN KENTUCKY 487
accounted for all three of the death sentences administered to this
group.?7 The authors viewed this outcome as a “paradox” since the
black offenders with white victims accounted for only twenty percent
of the total group examined.?°
B. POST-FURMAN AND GREGG RESEARCH
This pattern of racial bias also characterized the research dur-
ing the period following the Furman and Gregg decisions. With these
two decisions, the Court intended to remove arbitrariness and dis-
crimination from the death sentencing process.2® Here, the re-
search revealed that the race of the victim continued to play a
significant role, and that the offender/victim racial combination also
determined if the sentence was to be death. This research examined
the factors influencing the imposition of a death sentence at differ-
ent stages of the capital sentencing process. The United States
Supreme Court, in Gregg, approved of this process as a possible
method of curtailing discrimination.°°
Riedel was the first to examine the results of post-Furman capi-
tal sentencing decisions.3! He examined data from the NAACP
Legal Defense Fund and the National Prisoner Statistics on 493 of-
fenders from twenty-eight states under a sentence of death in 1971
and compared them to 142 death row inmates in six states sen-
tenced after Furman, between June 29, 1972 and August, 1975.32
He reported that a greater proportion of nonwhites were sentenced
to death, sixty-two percent versus fifty-three percent following
Furman.3? Although the interaction between the race of the offender
and that of the victim failed to yield a statistically significant result,
Riedel noted that eighty-seven percent of the homicide victims were
white.2+ He concluded that the disproportionality in the death sen-
tencing in white victim cases had continued despite the intentions of
the Supreme Court.35
Another early post-Furman study reached an almost identical
conclusion. Lewis interviewed eighty-three Florida death row in-
27 Id. at 233.
28 Id.
29 Gregg, 428 U.S. at 197, Furman, 408 U.S. at 241.
30 Gregg, 428 U.S. at 197-98. For an outline of the Georgia capital sentencing system
see supra note 7. °
31 Riedel, Discrimination in the Imposition of the Death Penalty: A Comparison of the Charac-
teristics of Offenders Sentenced Pre-Furman and Post-Furman, 49 Temp. L.Q. 261 (1976).
32 Id. at 275.
33 Id. at 276.
34 Id. at 282.
35 Id.
488 VITO AND KEIL [Vol. 79
mates in 1977 and reported that the majority of offenders (55.4%)
and victims (92.4%) were white.2 However, using arrest data as a
basis, he also determined that the probability of obtaining a death
sentence was significantly higher for blacks who killed whites than
for any other offender/victim racial combination.37 He concluded
that the fact that although the number of Florida death row blacks
was substantially reduced after 1972, this did not indicate Florida
had resolved its pre-Furman problems.38
Similarly, Bowers and Pierce conducted a study of persons in-
dicted for first degree murder in Florida, Georgia, Ohio, and Texas
during the post-Furman period.*® They determined that black de-
fendants who killed whites had the highest probability of receiving a
death sentence in all four states.*° This study examined all stages of
the capital sentencing process and considered the effect of such per-
tinent variables as the felony circumstances surrounding the mur-
der. Yet, the differential impact of the offender/victim racial
variable was still apparent.t! In another Florida-based study, Arkin
investigated the results of 350 murder cases from Dade County dur-
ing the period 1973 to 1976.42 He detected “‘selectivity and arbi-
trariness” at every stage of the sentencing process.** Black killers of
whites were for times more often convicted of first degree murder
than blacks who killed blacks.44 This combination accounted for
half of the death sentences levied.#5
Similarly, Zeisel examined the records of 114 Florida offenders
who reached death row after 1972 and before September of 1977.46
He determined that the death row murderers who killed whites out-
numbered those who killed blacks by a ratio of thirty-one to one.47
Zeisel also contended that after they were presented with evidence
of discrimination, prosecutors tried to cover such discrimination
36 Lewis, Life on Death Row: A Post Furman Profile of Florida’s Condemned, in THE
SuPREME Court AND THE CRIMINAL Process 948 (P. Lewis & K. Peoples, eds. 1978).
37 Id.
38 Id.
39 Bowers & Pierce, Arbitrariness and Discrimination Under Post Furman Capital Statutes, 26
Crime & DeLing, 563 (1980).
40 Id. at 577.
41 Id. at 629.
42 Arkin, Discrimination and Arbitrariness in Capital Punishment: An Analysis of Post Furman
Murder Cases in Dade County, Florida, 1973-1976, 33 Stan. L. Rev. 75 (1980).
43 Id. at 77.
44 Id. at 87.
45 Id.
46 Zeisel, Race Bias in the Administration of the Death Penalty: The Florida Experience, 95
Harv. L. Rev. 456 (1981).
47 Id. at 460.
1988] CAPITAL SENTENCING IN KENTUCKY 489
through “cosmetic” changes, which lead to a “more damaging
proposition” that “their [prosecutors] discretionary power to deter-
mine the death row population is not within any legal
boundaries.’”’48
Using data from 637 homicide indictments in twenty counties
from 1976 and 1977, Radelet also studied the Florida capital sen-
tencing process.*9 He made a distinction between primary homi-
cides, such as acts of passion between family members, friends, or
acquaintances, and nonprimary homicides, such as instrumental, oc-
curring during another felony, or among strangers.5° Log linear
analysis revealed a tendency to indict defendants accused of killing
whites for first degree murder, although offenders whose victims
were black were indicted for a lesser charge.5! Among both white
and black defendants, offenders who killed whites had a twelve per-
cent higher probability of receiving a death sentence.5? The major
unexplained finding was that nonprimary homicides with white vic-
tims were more likely to result in an indictment for first degree mur-
der than nonprimary homicides with black victims.53 The race of
the victim seemed to be a stronger factor in the prosecutorial deci-
sion than the nature of the homicide.
Radelet and Pierce reviewed 1382 Florida homicide cases from
a proportionate random sample of twenty-one counties and com-
pared the charges made at arrest with those eventually levied by the
prosecutor.5* Among 174 cases in which charges between the po-
lice and prosecutor differed, eighty-two were downgraded, with the
felony level lowered, and ninety-two were upgraded by the prosecu-
tor.55 Statistical analysis revealed that cases in which a black killed a
white were most likely to be upgraded, followed in order by cases in
which whites killed whites, blacks killed blacks and whites killed
blacks.5® The authors also determined that upgraded cases were
about twice as likely to result in a death sentence than were “‘consist-
ently classified” cases.57 Furthermore, the majority of the defend-
ants who refused to negotiate a plea, in this case sixty-nine percent,
48 Id, at 468.
49 Radelet, Racial Characteristics and the Imposition of the Death Penalty, 46 Am. Soc. Rev.
918 (1981).
50 Id. at 921.
51 Id. at 923.
52 Id. at 924.
53 Id. at 925.
54 Radelet & Pierce, Race and Prosecutorial Discretion in Homicide Cases, 19 Law & Soc.
Rev. 587, 595 (1985).
55 Id. at 598.
56 Id. at 604-05.
57 Id. at 612.
490 VITO AND KEIL [Vol. 79
received a death sentence.5®
In a South Carolina study, Paternoster also uncovered a “race
of the victim” capital sentencing effect.59 The data set included
1805 homicide acts committed between June 8, 1977, the effective
date of the state’s death penalty law, and December 31, 1981. Re-
garding prosecutorial discretion, Paternoster reported that black of-
fenders with white victims were over forty times more likely to have
the death penalty requested than the black killers who did not cross
racial lines.! Overall, rural prosecutors were more likely to seek
the death penalty.6? At every level of aggravation, the probability of
a death sentence was higher in white victim cases.®3 Accounting for
a number of independent variables through the use of the logit re-
gression statistical procedure, Paternoster revealed that prosecutors
were most likely to seek the death penalty in white victim cases.5+
He also determined that killers of whites in less aggravating circum-
stances were more likely to face a death penalty recommendation
than were the killers of blacks in more aggravating circumstances.®
The study concluded that the murder of a white, in and of itself, was
treated as a more serious offense.®°
Gross and Mauro provided more evidence of discrimination in
capital sentencing in their study of all homicides reported to the po-
lice in Arkansas, Florida, Georgia, Illinois, Mississippi, North Caro-
lina, Oklahoma, and Virginia.6? Again, blacks who killed whites
were several times more likely to receive a death sentence, even
when felony circumstances were taken into account.®® Logit regres-
sion analysis uncovered substantial levels of racial discrimination.
Killers of whites were often times more likely, four times more often
in Illinois, five times more often in Florida, and seven times more
likely in Georgia, to receive a death sentence.®? This pattern held in
Georgia even after a statutorily mandated “proportionality review”
58 Id. at 611.
59 Paternoster, Race of the Victim and Location of the Crime: The Decision to Seek the Death
Penalty in South Carolina, 74 J. Crm. L. & Criminotocy 754 (1983).
60 Id, at 762-63.
61 Id. at 766.
62 Id, at 780.
63 Id.
64 Id. at 776-77.
65 Paternoster, Proseculorial Discretion in Requesting the Death Penalty: A Case of Victim
Based Racial Discrimination, 18 Law & Soc. Rev. 437 (1984).
66 Id.
67 Gross & Mauro, Patterns of Death: An Analysis of Racial Disparities in Capital Sentencing,
37 Stanrorp L. Rev. 27, 66 (1985).
68 Id, at 56.
69 Id. at 66.
1988] CAPITAL SENTENCING IN KENTUCKY 491
took place.”° The rate of affirmation of the death sentence in white
victim cases versus that of black victim cases was nine to one.7!
Other studies of different states reached similar conclusions
concerning the use of the death penalty. In Louisiana, Smith re-
viewed 504 murder cases and discovered that white victim cases
were twice as likely to end in a death sentence.72 For murders oc-
curring between 1977 and 1984 in Texas, Henderson and Taylor
reported that killers of whites were five times more likely to receive
a death sentence than the killers of blacks.7? The evidence of dis-
crimination became even greater with the inclusion of the of-
fender/victim racial composite variable. Over an eleven year period
in Texas with 300 death sentences, blacks who killed whites were
most likely to arrive on death row.74
In the study reviewed in McCleskey v. Kemp,7> Baldus, Pulaski,
and Woodworth reviewed the Georgia capital sentencing process.76
The Georgia statute called for the use of a proportionality review in
order to determine whether individual death sentences were exces-
sively severe in comparison to the sentence imposed in factually in-
distinguishable cases.”7_ The Baldus research task was to determine
the best method of making such a determination. The authors uti-
lized two data sets: 130 pre-Furman cases, with twenty resulting in
death sentences; and 594 defendants tried and sentenced since
Furman.’® The research considered the effect of over 200 different
independent variables designed to account for aggravating and miti-
gating circumstances.” Their statistical regression model revealed
that, among cases with statutory aggravating factors, the death sen-
tencing rate did not exceed sixty-two percent.8° Thus, prosecutors
and jurors used considerable discretion in selecting defendants for
70 Id. at 85-86. Proportionality review is designed to determine whether the several
trial courts in the state are imposing the death penalty for roughly the same sort of
murder, or whether some courts have become execution-prone or otherwise aberrant in
their capital sentencing practices. The United States Supreme Court has held that this
procedure is not required. Pulley v. Harris, 465 U.S. 37, 44 (1984). See H. BeDau, supra
note 2. 2
71 Gross & Mauro, supra note 67, at 85.
72 Smith, Patterns of Discrimination in Assessment of the Death Penalty: The Case of Louisiana,
15 J. Crim. Just. 279, 283 (1987).
78 Witherspoon & Senderling, Racism, Even in Death, Dallas Times Herald, Nov. 17,
1985, at 4, col. 28-A.
74 Id.
75 107 S. Ct. 1756 (1987).
76 Baldus, Pulaski & Woodworth, supra note 9.
77 Ga. Code Ann. § 27-2537(c)(3) (1983).
78 Baldus, Pulaski & Woodworth, supra note 9, at 680.
79 Id. at n.81.
80 Jd. at 699.
492 VITO AND KEIL [Vol. 79
execution, resulting in a very low rate of capital sentencing in black
victim cases.8! They concluded that the level of aggravation must
be “substantially great” for prosecutors to seek the death penalty,8?
that juries appeared to tolerate greater levels of aggravation in black
victim cases,®4 and that “Georgia’s death sentencing system has con-
tinued to impose the type of inconsistent, arbitrary death sentences
that the U.S. Supreme Court condemned in Furman v. Gergia.”8+
These post-Furman and Gregg studies demonstrated that the
capital sentencing process in a number of states and at different
stages was characterized by racial discrimination. Thus, an exami-
nation of the nature of the capital sentencing processes in other
states is warranted. These findings served as the backdrop for the
present study of the Kentucky capital sentencing process.
Til. CaprraL PUNISHMENT IN KENTUCKY
Following the last execution of a Kentucky inmate,®> Kentucky
governors routinely granted inmates a stay of execution and the leg-
islature considered abolishing capital punishment altogether.8®
When the Furman decision was announced, the sentences of twenty-
three Kentucky death row inmates were commuted to life
imprisonment.87
After the Gregg decision, a Georgia-like capital sentencing stat-
ute was enacted by the Kentucky legislature.88 This statute estab-
lished several guidelines for the administration of the death penalty
to effectively curb the abuse of discretionary power.®° In Kentucky,
one of the targets for the imposition of capital punishment are mur-
der cases which involve another felony.®° The aim is to reserve capi-
tal punishment for the most severe murder cases, such as killings
which are accompanied by at least one of five felonies, including ar-
son, rape, robbery, sodomy and burglary, listed in the statute.9!
Other aggravating circumstances listed in the statute are: a substan-
81 Jd. at 707-08.
82 Id, at 710.
83 Id.
84 Id, at 730.
85 On March 2, 1962, Kelly Moss was electrocuted at the Kentucky State Penitentiary
at Eddyville. W. Bowers, Lecat Homicioe: Data as PUNISHMENT IN AMERICA, 1864-
1982 445 (1984).
86 Lecistative RESEaRcH Commission No. 218, Caprrat PuntsHMENT 9 (1985).
87 Vito & Wilson, Back From the Dead: Tracking the Progress of Kentucky's Furman Com-
muted Death Row Population, 5 Just. Q, 101, 103 (1988).
88 Ky. Rev. Stat. Ann. § 507.020 (Mitchie/Bobbs-Merrill 1985).
89 See id.
90 See id.
91 Ky, Rev. Stat. Ann, § 582.025 (Mitchie/Bobbs-Merrill 1985).
1988] CAPITAL SENTENCING IN KENTUCKY 493
tial history of serious assaultive convictions; prior conviction for a
capital offense; the use of a destructive device which creates a great
risk of death to more than one person in a public place; offenses
committed for the profit; murders involving multiple victims; and
murders of ‘“‘special categories” of victims, such as prison employ-
ees, state or local officials, police officers, sheriffs or deputy sheriffs,
killed during the lawful performance of their duties.9? The prosecu-
tor is the key actor in the capital sentencing process, for if the prose-
cutor does not seek the death penalty, the process halts.
Once the prosecutor files a motion to proceed with capital
cases, the process begins. The next step is the formation of a
“death qualified” jury. The potential jurors are told that the death
penalty is possible in this case. The jury consists of individuals who
do not demonstrate an unwillingness to sentence a defendant to
death, regardless of the circumstances of the case. When this pro-
cess is complete, the trial can begin.
Specifically, the Kentucky statute provides for a bifurcated trial
in death penalty cases: a guilt and a penalty phase.% If the defend-
ant is found guilty, the court resumes the trial and conducts a pre-
sentencing hearing. At this stage, the jury must follow specific
directions provided in the statute.25 The jury must determine that
one of seven “aggravating” circumstances is present in order to sen-
tence the defendant to death.°6 Similarly, the jury can consider one
of eight “mitigating” circumstances and impose a lesser sentence.97
The alternatives include life imprisonment, life without benefit of
parole for twenty-five years or, if the jury finds the defendant guilty
of a lesser offense such as manslaughter, a sentence of years defined
by statute.°8 Unlike states such as Alabama, Florida, Indiana,°9 the
Kentucky judges are not presently permitted to raise penalties lev-
92 Id.
93 The Supreme Court has held that in order for a juror to be legitimately excluded
the juror must state that he would automatically vote against the death penalty regard-
less of the evidence at trial or that this position would make it impossible for him to
make an impartial decision concerning the guilt of the defendant. See Buchanan v. Ken-
tucky, 107 S. Ct. 2906 (1987); Lockhart v. McCree, 476 U.S. 162 (1986); Wainwright v.
Witt, 469 U.S. 412 (1985); Witherspoon v. Illinois, 391 U.S. 510 (1968).
94 Ky. Rev. Stat. ANN. § 532.025 (Mitchie/Bobbs-Merrill 1985).
95 Id, at § 532.075
96 Id. at § 532.025
97 Id.
98 Id.
99 See Ata. Cove § 13A-5-47(e) (1982); Fra. Svar. ANN. § 921.141(2) & (3) (West
Supp. 1985); Inv. Cop Ann. § 35-50-2- 9(e) (West 1978 & Supp. 1984); Radelet, Re-
jecting the Jury: The Imposition of the Death Penalty in Florida, 18 U.C. Davis L. Rev. 1409,
1412 (1985); Tabak, The Death of Fairness: The Arbitrary and Caprious Imposition of the Death
Penalty in the 1980's, 14 N.Y.U. Rev. L. & Soc. CuancE 797, 820-821 (1986).
494 VITO AND KEIL [Vol. 79
ied by the jury.!°° The judge cannot overrule the jury and impose a
death sentence.
The statute also provides for an automatic review of any capital
case by the Kentucky Supreme Court to determine whether the pen-
alty was due to “the influence of passion, prejudice, or any other
arbitrary factor” and to ensure that the “evidence supports the
jury’s or judge’s findings.”!1 The review also directs the court to
determine that the death penalty is not ‘excessive or disproportion-
ate to the penalty imposed in similar cases, considering both the
crime and the defendant.”!0? The courts determined that this bifur-
cated sentencing process could apply only to persons indicted for
murder in Kentucky for offenses committed after December 22,
1976,108
This Article is as an evaluation of the effectiveness of the capital
sentencing process in Kentucky. The aim of the research was to as-
sess the ability of the policies listed in the statute to provide “super”
due process in capital cases. Basically, the process works if it effec-
tively curbs both intentional and unintentional arbitrariness and
discrimination.104
IV. METHODOLOGY
A. RESEARCH QUESTIONS
In light of the research findings of the previous studies of capi-
tal sentencing systems,!°5 it was essential to focus on certain aspects
of the process in Kentucky. The first research question considers
the prosecutorial decision to seek the death penalty and attempts to
ascertain those factors associated with this decision. The second
question focuses upon the jury’s decision to sentence the defendant
to death. It is essential to consider such stages in the process be-
cause the decisions at each level are dependent upon decisions
made at the preceding level.
B. RESEARCH DESIGN
The universe of cases consisted of all persons, with complete
data, indicted for murder in Kentucky between December 22, 1976,
100 Ky. Rev. Stat. ANN. § 532.025 (Mitchie/Bobbs-Merrill 1988).
101 jd, at § 532.075.
102 Jd.
103 See Hudson v. Commonwealth, 597 S.W. 2d 610 (Ky. 1978).
104 See Hubbard, Reasonable Levels of Arbitrariness in Death Sentencing Patterns: A Tragic
Perspective on Capital Punishment, 18 U.C. Davis L. Rev. 1113 (1985).
105 See supra notes 15-84 and accompanying text.
1988] CAPITAL SENTENCING IN KENTUCKY 495
the effective date of the statute, and October 1, 1986, who were con-
victed and sentenced to prison under a death or lesser sentence.
This universe comprised 864 cases. However, not all of these indi-
viduals were candidates for the death penalty. In order to proceed
capital, the prosecutor must determine that a least one of the previ-
ously cited statutory aggravating circumstances must be present.106
Following this determination, the universe dropped to 557 cases.
The multivariate statistical techniques used in this study require that
each case be complete and contain no missing data.!©7 For this rea-
son, the universe was once again reduced to a sample of 458 cases.
Finally, the cases studied consisted of individuals indicted for mur-
der, with at least one aggravating circumstance, with complete data
on the variables included in this analysis. Three cases in which the
death sentence was imposed by a judge and one case in which the
victim was Asian were excluded from the analysis.
Among these cases, there were 104 death qualified juries and
thirty-five death sentences. Thus, the final research sample con-
sisted of persons indicted for murder in a case with at least one stat-
utorily prescribed aggravating circumstance and sentenced to
prison, as long as a death sentence was a possibility. The universe
did not include individuals indicted for murder who had the charges
dropped or who were acquitted at trial.
Data on the offense and the offender were collected from insti-
tutional files compiled and maintained by the Kentucky Corrections
Cabinet. The presentence investigation report (PSI) was the main
source of information for the study. The PSI is a vital tool in the
sentencing process. It contains a volume of information about the
offense and the criminal, including a social history of the accused.
In its final form, the data collection form addressed approximately
ninety-seven variables on the characteristics of the offender, offense,
and victim(s). Here, the conscious attempt was to replicate the
methodology of the Baldus study.!°8 Professor Baldus made his
data collection codebook available and it served as a model for the
present study.109
A list maintained by the Department of Public Advocacy was
106 See Ky. Rev. Stat. ANN. § 532.025 (Mitchie/Bobbs-Merrill 1985).
107 See S. FINEBERG, THE ANaLysis Or Cross Crassiriep Catecoricat Data 142
(1980).
108 Baldus, Pulaski & Woodworth, supra note 9.
109 jd. The data collection procedure was as follows: a student collected data on the
first two records on the data collection schedule involving demographic data on the
offender while the professor compiled data on the circumstances of the offense, the vic-
tim(s) and made determinations about the presence of aggravating and mitigating cir-
cumstances in the case.
496 VITO AND KEIL [Vol. 79
used to determine if a death qualified jury was used in the case. The
present study conservatively underestimates the actual number of
cases in which the prosecution sought the death penalty, because
this list either did not contain all of the offenders who were sen-
tenced to prison over this time period or did not have this informa-
tion on every case. Another computerized list compiled by the
Kentucky Administrative Office of the Courts served as a basis to
verify that the offender had been indicted for murder.
The findings of this study are limited by two factors. First, com-
plete data on the race of the victim was not obtained for all cases,
except for cases of offenders on death row. Second, there are
problems with the classification of aggravating and mitigating cir-
cumstances by a person without formal legal training. However,
before the data collection began, the researchers consulted with the
Kentucky Offices of Attorney General and Public Advocacy.
V. RESEARCH FINDINGS
The analysis begins with a review of the frequencies of different
independent variables used in the study. These variables give some
initial indication of the importance.of the racial variable, a break-
down of the overall presence of aggravating and mitigating circum-
stances, and the background of victims for the universe of cases.
Table 1 lists the presence of aggravating circumstances, as dic-
tated by Kentucky law, within the research sample. The top aggra-
vating circumstance present was a substantial history of violent
offenses. A majority of the offenders had a “‘substantial history of
violent offenses,” as indicated by a prior conviction for a violent of-
fense. The next highest category was homicide concurrent with rob-
bery in the first degree (thirty-one percent of the cases), followed by
burglary (12.7%) and cases involving multiple victims (12.2%).
Table 2 contains a list of the independent variables considered
in the analysis. The effect of several indicators of the heinousness of
the crime along with an estimate of the effect of the racial composite
variable BLACK KILLS WHITE (BKW) was constructed and esti-
mated. Of course, most of the independent variables reflect the ag-
gravating circumstances present in Kentucky law. These specific
variables were CONCUR, MDEATH, SILENCE, FEMALE VICTIM
and KMAGG. CONCUR indicates whether the accused had been
charged with a homicide committed in conjunction with one or
more of the following offenses: Arson I, Robbery I, Rape I, Sodomy
I, or Burglary I. MDEATH signifies whether the accused was being
tried for more than one homicide. SILENCE demonstrates that the
1988] CAPITAL SENTENCING IN KENTUCKY 497
TABLE 1
PRESENCE OF AGGRAVATING CIRCUMSTANCES LISTED IN THE
KENTUCKY STATUTE IN THE UNIVERSE OF KENTUCKY MURDER
CASES (N = 458): 12/22/76 To 10/01/86
Aggravating Circumstance Yes = Pet.
1. Prior conviction for a capital offense: 17 3.7
2. Substantial history of assaultive convictions: 353 771
3. Homicide committted in conjuction with:
a. ArsonI 6 1.3
b. Robbery I 142 31.0
c. Rapel 22 48
d. Sodomy I 8 1.7
e. Burglary I 58 12.7
4. Did the offense involve the use of a destructive
device or weapon which created a great risk of
death to more than one person in a public place? 0 0.0
5. Was the offense committed for the purpose of
receiving money or any other thing of monetary
value or for other profit? 17 3.7
6. Prison employee victim: 3 0.7
7. Did the offender’s act of killing result in multiple
deaths? 5612.2
8. State/local official or law enforcement victim: 8 1.7
accused killed the victim in order to prevent the victim from testify-
ing against the offender. While this factor is not an element under
the Kentucky statute, it is a legal factor in several other states and it
reflects a rational, instrumental motivation for the homicide.!!° Pre-
vious studies!!! have suggested that killers of females (FEMALE
VICTIM) are more likely to be targeted for a death sentence. Ken-
tucky multiple aggravating circumstances (KMAGG) indicates
whether one or more of the statutory elements were present in the
case.!!2 KMAGG is particularly important to the analysis since it has
been suggested that one possible way to avoid discrimination in
death sentencing is to focus on those cases which have more than
one aggravating circumstance. Such an approach “would limit the
class of death eligible defendants to those that society has already
shown they feel are most opprobrious, regardless of the race of the
defendant or the victim.”!!3 The inclusion of KMAGG permits a
110 Baldus, Pulaski & Woodworth, supra note 9, at 693.
111 See Paternoster, supra note 59, at 781; Smith, supra note 72, at 283.
112 See Ky, Rev. Stat. ANN. § 532.025 (Mitchie/Bobbs-Merrill 1985).
113 Note, EIGHTH AND FOURTEENTH AMENDMENTS — THE DEATH PENALTY Survives, 78
J. Crm. L. & Criminotocy 1080, 1113 (1988)(authored by Anderson Bynam).
498 VITO AND KEIL [Vol. 79
TABLE 2
FREQUENCY DISTRIBUTION OF INDEPENDENT VARIABLES
Independent Variable NU Pet.
CONCUR
Yes 188 41.0
No 270 59.0
MDEATH
Yes 56 12.2
No 402 87.8
SILENCE
Yes 72 15.7
No 386 84.3
KMAGG
Yes 182 39.7
No 276 60.3
FEMALE VICTIM
Yes 141 30.8
No 317 69.2
direct test of this hypothesis in Kentucky.
Together, these variables give some indication of the different
factors present in murder cases. These are the legal factors, out-
lined by statute, which have been highlighted as the variables which
should legitimately guide the discretionary power of decision mak-
ers in the capital sentencing process. Severity of the offense, such as
the more heinous murders, and murders committed during the
course of another felony, are all highlighted as the types of homi-
cides which should draw the death penalty. However, research has
determined that extralegal factors, particularly race, impinge upon
capital sentencing decision making.
Table 3 gives us some indication of the pattern present in the
dependent variables of this study. This initial pattern is remarkably
consistent with the findings of the previously cited studies on this
subject. The data reveal that a higher proportion of black offenders
who murdered whites go on to receive a death qualified jury
(47.7%) than is the case among other racial combinations. None of
the fourteen white offenders with black victims made it to the level
of a death qualified jury. Of the 140 black offenders eligible for a
death qualified jury, 33.5% had white victims. Of the thirty-three
blacks tried for a capital offense, 63.6% had white victims. Finally,
of the eight blacks who received a death sentence approximately
87.5% had murdered whites. Blacks who killed blacks, like whites
1988] CAPITAL SENTENCING IN KENTUCKY 499
TABLE 3
RACIAL COMBINATIONS BETWEEN OFFENDERS AND VICTIMS IN THE
RESEARCH SAMPLE OF KENTUCKY MURDER CASES (N = 458):
12/22/76 To 10/1/86*
PROBABILITY OF PROBABILITY OF
NOF DEATH QUALIFIED DEATH
OFFENDER/VICTIM INTERACTION CASES ‘JURY SENTENCE
Black Kills Black 93 12 (12.9%) 1 (8.3%)
Black Kills White 47 21 (44.7%) 7 (83.3%)
White Kills White** 304 71 (23.3%) 27 (38.0%)
White Kills Black 14 0 (0%) 0 (0%)
TOTALS 458 104 35
* The figures in parentheses are row percentages which reflect the probabilities of,
first, facing a death qualified jury and then receiving a death sentence.
** Two of the death sentences in this group were overturned on appeal. A case in which
a white raped and murdered an Asian was excluded from the analysis. In this case, the
offender was tried before a death qualified jury, found guilty and was sentenced to life in
prison.
who killed blacks, had a very slight chance of being sentenced to
death. Only 12.9% of this subgroup faced a death qualified jury and
only 8.3% received a death sentence. This type of univariate analy-
sis does not provide a satisfactory explanation of how and why such
an apparently discriminatory pattern is present.
As Kleck has indicated, blacks who kill whites may be more
likely to be punished more severely for homicides due to legally rel-
evant factors related to such offenses.!1* Thus, there may be legiti-
mate reasons, as indicated by aggravating circumstances, why blacks
who kill whites are more likely to be singled out for and sentenced
to death in Kentucky. In order to approach this question, it is neces-
sary to conduct a multivariate analysis which includes legally rele-
vant factors and indicators of the race of the victim and the race of
the offender. The issue addressed in this Article is the extent to
which the overrepresentation of blacks who kill whites in the capital
sentencing system is a function of their involvement in objectively
more serious crimes or the extent to which such a result is due to
the extralegal factor of race.
A. THE PROSECUTORIAL DECISION TO SEEK THE DEATH PENALTY
While prosecutors in Kentucky may not seek the death penalty
unless at least one statutory aggravating circumstance is present, the
114 Kleck, Racial Discrimination in Criminal Sentencing: A Critical Evaluation of the Evidence
with Additional Evidence on the Death Penalty, 46 Am. Soc. Rev. 783 (1981).
500 VITO AND KEIL [Vol. 79
law gives prosecutors considerable latitude in determining who is
prosecuted for a capital offense. By virtue of their discretionary au-
thority, prosecutors may consider extralegal factors in deciding who
will be charged with a capital crime. The specific extralegal factors
that are addressed in the Article are the race of the offender, race of
the victim and the sex of the victim.
TABLE 4
Locir REGRESSION MODEL: FACTORS ASSOCIATED WITH THE
FORMATION OF A DEATH QUALIFIED JURY
Parameter Coefficient Standard Error z-Value
DQJURY 032 -126 25
DQJURY BY CONCUR 271 084 3.23**
DQJURY BY MDEATH 324 .091 3.58**
DQJURY BY SILENCE 263 079 3.33**
DQJURY BY BKW -252 .093 2.7188
DQJURY BY KMAGG 181 079 2.30**
DQJURY BY FEMALE VICTIM 19 068 1,75**
* Coefficient greater than one and one half its standard error.
** Coefficient greater than twice its standard error.
LIKELIHOOD RATIO CHI SQUARE = 43.584 DF = 50 P > .10
MEASURES OF ASSOCIATION
ENTROPY = .114
CONCENTRATION = .143
* Significant at .05 level or beyond, 2 tailed test.
The dependent variables are DQJURY, a dichotomous variable
indicating whether or not the prosecutor has decided to qualify a
jury to impose a death sentence in a particular trial, and LORD, a
dichotomous variable indicating whether a convicted murderer actu-
ally received a death sentence from the jury. There was complete
data for 432 murder cases that met the legal qualifications for being
charged with a capital offense. In Table 4 is an examination of the
effect of CONCUR, KMAGG, MDEATH, SILENCE, FEMALE VIC-
TIM and BKW on DQJURY by logit regression. Due to the pres-
ence of sampling zeros, the degrees of freedom for the overall
model were calculated using the formula and method suggested by
Fienberg,!!5 All of the predictors measuring the seriousness of the
homicide were statistically significant and their signs were in the ex-
pected direction. They were all positive, indicating that prosecutors
were most likely to proceed capital in those cases which met the
115 Fineberg, supra note 107, at 142.
1988] CAPITAL SENTENCING IN KENTUCKY 501
legal requirements for an accused to be charged with a capital
crime: the accused had more than one aggravating circumstance,
there was more than one victim; and the killer committed the crime
to prevent testimony from being introduced by the victim. In addi-
tion, with female victims there was a greater likelihood of being
charged with a capital crime. Even with these variables in the equa-
tion, however, the effect of BKW continued to be statistically signifi-
cant and positive. Thus, blacks who kill whites have a higher than
average probability of being brought before a death qualified jury.
This conclusion holds even though, as Table 5 demonstrates, BKW
cases are correlated with the several indicators of seriousness.
TABLE 5
PEARSON CORRELATIONS AMONG THE PREDICTORS OF DQJURY
AND LORD
FEMALE
CONCUR MDEATH SILENCE BKW KMAGG _ VICTIM
CONCUR 1.000
MDEATH 014 1.0000
SILENCE -420* .077* 1.0000
BKW .201* —.060* .071* 1.0000
KMAGG .583* -.267* 311* .078* 1.000
FEMALE
VICTIM 040 —.004 .089* 008 —:001 1,000
* Significant at or beyond the .05 level.
In the next stage of the analysis, the effect of the various predic-
tor variables upon the receipt of a capital sentence among those
who were tried before a death qualified jury was examined. In this
equation, the only statistically significant predictor is SILENCE.
This finding indicates that juries regard crimes where this motiva-
tion is present to be especially serious. Therefore, among those de-
fendants who face a death qualified jury, there is no evidence that
blacks who have white victims, compared to other killers, are more
likely to receive a death sentence when the seriousness of the homi-
cide is considered.
The absence of a significant effect for BKW in the LORD equa-
tion means that this variable does not directly determine which of
the eligible defendants will be sentenced to die. Rather than react-
ing to the combination of the race of the victim and the race of the
accused in imposing a sentence, Kentucky juries may react to the
objective heinousness of the murder cases brought before them.
502 VITO AND KEIL (Vol. 79
TABLE 6 ;
Logit REGRESSION MODEL: FACTORS ASSOCIATED WITH THE
RECEIPT OF A DEATH SENTENCE (LORD)
Parameter Coefficient Standard Error —z-Value
LORD 189 «175, 1.083
LORD BY CONCUR —.168 167 —1,005
LORD BY MDEATH 167 131 1.276
LORD BY SILENCE 353 127 2.793*
LORD BY BKW .010 .140 068
LORD BY KMAGG 197 148 1.330
LORD BY FEMALE VICTIM 166 133 1.462
LIKELIHOOD RATIO CHI SQUARE = 37.012 DF = 50 P > .10
MEASURES OF ASSOCIATION
CONCENTRATION = .114
ENTROPY = .143
* Coefficient is greater than twice its standard error.
B. CONCLUSIONS
However, as demonstrated in Table 4, BKW does influence the
decision of Kentucky prosecutors to proceed with a capital prosecu-
tion. Here, the pattern of effects demonstrated by the race of the
victim-race of the offender combination indicates that, controlling
for differences in the objective heinousness of the offense, prosecu-
tors are more likely to seek the death penalty when a black kills a
white than in other homicide cases. Table 3 demonstrates that,
without controlling the seriousness of the offense, a pattern in which
blacks who kill blacks have a relatively low probability of facing a
death qualified jury (12.9%) and a slight chance of receiving a death
sentence (8.3%) results.
Racial differences may not be the only factors which influence
this prosecutorial decision. Such a determination may also reflect
an instrumental orientation on the part of prosecutors. It is possible
that prosecutors are more likely to proceed capital when a black kills
a white because they believe that it is easier to obtain a conviction in
such cases. Prosecutors may be responding to public pressure to
take a hard line in such cases. It also appears that a reliance upon
the use of multiple aggravating circumstances as a requirement to
proceed capital offers no guarantee that racial discrimination will be
prevented or eliminated. Blacks who kill whites are still more likely
to be the target of a death qualified jury under such restrictions.
1988] CAPITAL SENTENCING IN KENTUCKY 503
Regardless of the underlying motive, the fact remains that racial
discrimination was present in the Kentucky capital sentencing pro-
cess during the post-Gregg period. In Kentucky, blacks who kill
whites have a generally greater risk of arriving on death row than
other murderers in part because they are more likely to face a jury
empowered to consider a death sentence. This result is not exclu-
sively because these murders are objectively more heinous than
others. When factors indicative of heinousness were considered,
the BKW variable continued to have a statistically significant effect
on the probability of facing a death qualified jury.
The finding that the race of the victim-race of the offender com-
posite variable does not have a direct effect upon the final stage of
the process, on who actually receives a death sentence, does not
mean that race does not structure the probability of receiving a capi-
tal sentence. Race is a crucial factor in the “gatekeeping” first stage
of seeking the death penalty. Once a person faces a death qualified
jury, factors other than race produce the final disposition. This find-
ing does not mitigate the evidence of racial effects, because the com-
bination of the race of the victim and the race of the offender has
significant consequences in the determination of who faces a death
qualified jury. The racial element is not eliminated in the latter
stages of the capital sentencing process. In this fashion, Kentucky’s
capital sentencing system has produced the same flawed, discrimina-
tory result which characterizes all of the capital sentencing systems
evaluated during the post-Gregg period to date. In Kentucky, blacks
who kill whites are more likely to face a death qualified jury than
other combinations of victim/offender race, even when the objective
nature of the homicide is taken into account.
