Layout 1

RACE, PROSECUTORS, AND JURIES: THE DEATH PENALTY

TENNESSEE

OHN M. SCHEB II, WILLIAM LYONS, AND KRISTIN A. WAGERS

The behavior of prosecutors and juries with respect to the decision to impose the death penal-

ty is examined. Using data on 968 first-degree murder convictions in Tennessee from 1977

to 2006, models of prosecutorial and jury behavior are constructed with a focus on the role

of the defendant’s race and the victim’s race. Once significant controls are introduced, nei-

ther the race of the defendant nor that of the victim is a significant predictor of prosecutors’

decisions to seek the death penalty. Nor is the race of the victim a significant predictor of the

jury’s decision to impose the death penalty. The defendant’s race is a significant, albeit weak,

predictor of jury behavior, but only because white defendants are more likely to be sentenced

to death. While a number of other variables are significantly related to both prosecutorial and

jury behavior, much of the variance in these decisions remains unexplained.

n Furman v. Georgia (1972), the United States Supreme Court held that Georgia’s

death penalty statute violated the Eighth Amendment’s ban on “cruel and unusu-

al punishments.” Of central concern to the Court was the perceived inadequacy of

procedures in capital cases and, in particular, the unfettered discretion vested in trial

juries regarding the imposition of the ultimate punishment. The resulting national

moratorium on the death penalty lasted until the Court approved Georgia’s revised

death penalty statute in

Gregg v. Georgia (1976). The revised statute adopted new

procedures designed to render the imposition of the death penalty less arbitrary and

capricious. Most notable of these was the “bifurcated trial,” in which the guilt/inno-

cence phase of the trial is separated from the penalty phase.

In the wake of

Furman and Gregg, a majority of states enacted new capital-pun-

ishment laws modeled after the new Georgia statute. Today, thirty-seven states and

the federal government have death penalty statutes. While these laws vary somewhat

procedurally, all are based on the bifurcated-trial model upheld in

Gregg v. Georgia.

All of these statutes are predicated on the notion that structuring the jury’s discretion

in death penalty cases reduces the potential for arbitrariness and discrimination to a

constitutionally acceptable level.

In the American system of criminal justice, pr

osecutors ar

e vested with enor

mous discr

etion. Subject to minimal pr

ocedural checks thr

ough grand jur

y pr

oceed

ings, pr

eliminar

y hearings, or both, pr

osecutors decide whether to bring criminal

char

ges and which char

ges to bring. In jurisdictions that have the death penalty

osecutors must decide whether to seek the death penalty in cases wher

e defendants

e char

ged with first-degr

ee mur

der

. Befor

e such trials begin, prosecutors must for-

mally notify trial cour

ts and defendants that they intend to seek the death penalty

upon conviction.

egg v

. Geor

gia

and subsequent judicial decisions have done little

THE JUSTICE SYSTEM JOURNAL, VOL. 29, NUMBER 3 (2008)

RACE, PROSECUTORS, AND JURIES: THE DEATH PENALTY IN TENNESSEE 339

if anything to reduce prosecutorial discretion to seek the death penalty. We seek

here to determine the case characteristics that influence prosecutors to seek the death

penalty and those that lead juries to impose it.

RACE AND THE DEATH PENALTY

One of the long-standing concerns about the administration of the death penalty has

been the potential for racial discrimination. Many have documented racial inequal-

ity in the administration of the death penalty (see Mangum, 1940; Garfinkel, 1949;

Johnson, 1957). However, Kleck (1981) criticized these early studies for failing to

utilize adequate controls. Justice William O. Douglas expressed concern for racial bias

in his concurring opinion in

Furman, where he characterized the old Georgia capital-

punishment statute as “pregnant with discrimination” (408 U.S. at 257). Supporters

of the death penalty were hopeful that the structured discretion model established by

the new post-

Furman statutes would lead to race-neutral application of the death

penalty. However, social-science research in this area has cast doubt on this aspira-

tion. While there is little evidence of racial discrimination in terms of the race of

defendants, there is substantial evidence of disparity with regard to the race of victims

(e.g., Baldus, Pulaski, and Woodworth, 1983, 1986; Lempert, 1983; Radelet and

Pierce, 1985; Smith, 1987; Ekland-Olson, 1988; Vito and Keil, 1988).

Summarizing the thrust of research in this area, Radelet and Borg (2000:47)

conclude that “the death penalty is between three and four times more likely to be

imposed in cases in which the victim is white rather than black.” There is also some

evidence of an interaction between the race of defendants and that of victims. Some

studies have found that the greatest disparity in the rate at which the death penalty

is sought or imposed exists between cases where black defendants are alleged to have

killed white victims and those where black defendants are alleged to have killed black

victims (see, e.g., Baldus, Pulaski, and Wordworth, 1983, 1986; Paternoster, 1983,

1984). To the extent that prosecutors and juries reflect dominant attitudes of their

communities, such disparities would seem quite plausible. On the other hand, some

researchers argue that apparent racial differences in the administration of the death

penalty can be explained in terms of nonracial case characteristics (see, e.g.,

Heilbrun, Foster, and Golden, 1989; Klein and Rolph, 1991).

osecutorial Behavior

. Pr

osecutors possess the discretion to “formulate charges that

deter

mine whether or not the death penalty is permitted if a conviction is obtained”

and to pr

opose a sentence of life in exchange for a guilty plea in cases wher

e the

defendant is eligible for the death penalty (Zeisel, 1981:466). Usually elected of

cials, pr

osecutors must be attentive to their constituency’

s r

eaction to crime, which

may influence the pr

osecutor

’

s decision to seek the death penalty

, accept a guilty plea

for a lesser sentence, or agr

ee to a plea bar

gain in exchange for testimony against

another defendant (Bowers, 1984). Bowers (1984:339, 345) suggests that elected

osecutors ar

e mor

e likely to seek the death penalty, given the legal requirements to

340 THE JUSTICE SYSTEM JOURNAL

do so, when there is public outcry for capital punishment in a given case or when

there is pressure from the police department. Therefore, he asserts that the unfettered

discretion exercised by prosecutors may be a vehicle of “arbitrariness and discrimina-

tion” in the early stages of capital cases (Bowers 1984:340). In a study of death penal-

ty cases in North Carolina, Nakell and Hardy (1987) found that a prosecutor’s deci-

sion to seek the death penalty largely depended on which prosecutor was assigned to

any given case, thus buttressing previously expressed concerns about arbitrariness and

the potential for discrimination.

Focusing specifically on race as a determinant of prosecutorial behavior,

Paternoster analyzed 300 homicide cases and found that prosecutors were four-and-a-

half times more likely to seek the death penalty when black defendants have white

victims. This would seem to suggest, as does other research, that “black

offender/white victim homicides are treated as more aggravated killings, and black

offender/black victim homicides are treated as less aggravated deaths” (Paternoster,

1984: 453). In another analysis, Paternoster found that prosecutors were forty times

more likely to seek the death penalty when black defendants were accused of killing

white victims than when black defendants were accused of killing other African-

Americans (Paternoster, 1983: 766). Similarly, after examining data from five South

Carolina counties, Johnson (2003) found the race of the victim to be a significant

predictor of prosecutors’ decision to file a notice to seek the death penalty.

Jury Behavior. Both prosecutors and jurors rely on their own socially conditioned

views in the criminal justice system. Bowers (1984:338-39) found jurors “embody

community sentiments” by having their own preconceived notions about members of

society, and as a result, “jurors have difficulty replacing their socially conditioned views

of victims and offenders with strictly legal considerations, especially for the crimes they

find most shocking and abhorrent”; this allows “extralegal” factors to intrude on the

interpretation of aggravating and mitigating factors presented at sentencing.

Gross and Mauro (1989:112) found that for jurors to sentence a defendant to

death, they must be “particularly horrified by the crime and perhaps frightened by the

defendant. As the jurors review the circumstances of the crime, they will inevitably

consider the personal characteristics of the defendant, and the defendant’s race might

carry some weight.” They argue that the one aspect of a trial that influences a juror’s

decision to sentence a defendant to death is the ability to identify with the victim,

rather than seeing the victim as a stranger

. In mo

der

n society

, where social classes

tend to be segr

egated and jurors tend to be Caucasian, “jurors are more likely to be

hor

rified by the killing of a white than of a black, and more likely to act against the

killer of a white than the killer of a black,” which is “a natural pr

oduct of the patterns

of inter

racial relations in society” (Gross and Mauro, 1989:113).

In ter

ms of the race of the defendant, racial stereotypes become relevant in the

sentencing phase of capital jur

y trials. In many states, juries are “allowed or required

to deter

mine whether the defendant will be dangerous in the future, even in states

where future dangerousness is not formally part of the sentencing process” (Johnson

RACE, PROSECUTORS, AND JURIES: THE DEATH PENALTY IN TENNESSEE 341

2003:136), and, according to a 1990 National Opinion Research Center survey, a

majority of whites feel African-Americans are more “prone to violence than whites”

(Johnson 2003:136). Given these survey results, Johnson (2003:137) suggests that

many Caucasian jurors think blacks are more likely to be involved in violent behav-

ior, and as a result, these jurors are more likely to impose a death sentence. Johnson

also believes that mitigating factors, specifically good character, are less likely to be

given significant weight in the sentencing phase for blacks because jurors “attribute

fewer positive traits to people of color, see them as less intelligent, less hard-working

and less good” than whites. Therefore, there is support for the conclusion that “jurors

are more likely to perceive the presence of aggravating factors in black defendant

cases than they do in white defendant cases with equal evidence of aggravation, and

they are less likely to give weight to mitigating factors” (Johnson, 2003:137).

THE TENNESSEE STUDY

Clearly, the two loci of potential bias in the imposition of death sentences are 1) the

prosecutor’s decision to seek the death penalty, which triggers the special death penal-

ty procedures, and 2) the jury’s decision to sentence a convicted murderer to death.

Prior research has tended to focus on one or the other, but it is important to examine

both together. Our study focuses on Tennessee, where in fall 2006 the Administrative

Office of the Courts released data on 986 first-degree murder convictions dating from

1977, when T

ennessee reformed its death penalty law, to early 2006. The data were

taken from reports submitted by trial judges across the state in compliance with

Tennessee Supreme Court Rule 12, which requires judges to complete detailed reports

on cases in which defendants are convicted of first-degree murder.

While the Tennessee data set is far from perfect, it is a useful tool with which

to analyze prosecutorial and jury behavior. The data set is quite comprehensive in

terms of the variables included, allowing us to utilize a variety of controls.

Specifically, we have data on demographic characteristics of perpetrators and victims.

We also have information about the nature and location of the homicide, the means

by which death was inflicted, and the perpetrator’s motive, relationship to the victim,

and criminal history. The data set also contains the nature of the evidence before the

jury, including whether the defendant confessed to the homicide. Finally, it also

includes records of the statutory aggravating and mitigating factors found by juries.

The Rule 12 database has been criticized as being incomplete because trial

judges were not always careful or diligent in filling out questionnaires. Indeed, it has

been suggested that as many as 20 percent of first-degree murder cases may be miss-

ing from the data set (see

State of Tennessee v. Godsey [2001], Birch. J., dissenting).

At the present time, we have no way of assessing how extensive are the missing data;

for present purposes, we assume any missing data to be randomly distributed across

variables and cases.

Death Sentences and Race in Tennessee: A First Look. Ninety-eight percent of

defendants in these cases were either black (45.5 percent) or white (52.5 percent).

342 THE JUSTICE SYSTEM JOURNAL

The 2 percent who were Hispanic, Asian, Native American, or “other” have been

excluded from the analysis. Similarly, 97.5 percent of victims were either black (33.6

percent) or white (63.9 percent); the remaining 2.5 percent have been excluded.

Looking first at the race of defendants, we note that 19.2 percent of white defendants

are sentenced to death as compared with only 11.5 percent of black defendants (see

Table 1). Clearly, there is a racial discrepancy, but in the direction opposite from

what many might expect. However, with respect to the race of victims, the opposite

is true: 18.5 percent of cases involving white victims led to the death penalty, where-

as only 9.8 percent of cases involving black victims resulted in death sentences.

Prosecutors were actually more likely to seek the death penalty in cases involving

white defendants, but they were also more likely to seek the death penalty when vic-

tims were white. Juries were more likely to impose the death penalty both in cases

involving white defendants and in cases involving white victims.

If we intersect the race of the victim with the race of the defendant (see Table

2), we find that black defendants accused of killing black victims receive the death

penalty about 10 percent of the time, while black defendants accused of killing white

persons are sentenced to death 15 percent of the time. Similarly, white defendants

accused of murdering blacks get the death penalty about 12 percent of the time, while

whites accused of killing other whites are sentenced to death 19 percent of the time.

However, when we separate the prosecutorial function from the jury function, we see

that the disparity in terms of the victim’s race exists in the decisions of prosecutors to

seek the death penalty, not in juries’ decisions to impose it. The data show that blacks

accused of killing black victims were just about as likely—actually, slightly more like-

ly—to be sentenced to death by juries as blacks who killed whites. Similarly, whites

who killed blacks were slightly more likely to be sentenced to death by juries than

whites who killed whites. On the other hand, when blacks are accused of killing

blacks, prosecutors seek the death penalty only about 25 percent of the time; when

blacks are accused of killing whites, the rate jumps to 39 percent. Likewise, when

whites are accused of killing blacks, prosecutors seek death only about 24 percent of

Table 1

Prosecutorial and Jury Behavior by Race of Defendant and Victim

Defendant Victim All

Black White Black White

Sentenced to death 11.5% 19.2% 9.8% 18.5% 15.8%

(% of all cases)

Prosecutor seeks death penalty 30.3 39.4 24.7 39.9 35.3

(% of all cases)

Death sentence imposed by jury 37.3 48.1 39.7 46.0 44.2

(% of cases where death penalty is sought)

RACE, PROSECUTORS, AND JURIES: THE DEATH PENALTY IN TENNESSEE 343

the time; when whites are charged with killing other whites, the rate increases to

nearly 40 percent. Thus, it would appear that, if there is racial discrimination with

respect to victims, it occurs more in the prosecutorial office than at the jury stage.

This significant difference between prosecutors and juries would make sense given the

enormous discretion vested in prosecutors, as contrasted with the much more struc-

tured decision making of juries under current death penalty law.

Introducing Controls. Before concluding that Tennessee prosecutors engage in racial

discrimination with respect to the administration of capital punishment, we need to

look at other factors that might explain the apparent disparities noted above.

Therefore, we constructed more complex statistical models incorporating the data on

defendant characteristics, victim characteristics, crime characteristics, and case char-

acteristics as controls with the goal of generating the most parsimonious model, that

is, one that maximizes explanatory power with a minimal number of additional vari-

ables.

We wished to see whether the racial variables remain significant when incor-

porated into such a model. We find that, once appropriate controls are introduced,

including the place, method, motive of the homicide, the nature of the evidence,

whether there are multiple victims, and the defendant’s criminal history,

the race of

the defendant and the race of the victim are not significant predictors of a death penalty out-

come

(see Table 3).

With the exception of the defendant’s race and the victim’s race, all variables in

the model are significant at the .05 level, which means that the probability of the

observed relationship occurring randomly is less than 5 percent (see Table 3). Among

Table 2

Prosecutorial and Jury Behavior by Race of Defendant by Race of Victim

Black Def. Black Def. White Def. White Def.

/Black /White /White /White

Victim Victim Victim Victim All

Sentenced to death 9.7% 15.0% 11.8% 19.0% 15.8%

(% of all cases)

Prosecutor seeks death penalty 24.7 39.0 23.5 39.7 35.3

(% of all cases)

Death sentence imposed by jury 39.4 38.5 50.0 47.2 44.2

(% of cases where d.p. is sought)

Details about the statistical tests performed, and statistical results, are available from the authors on request.

Specifically, the other independent variables are 1) whether the defendant confessed to murder; 2) whether

the defendant had thr

ee or mor

e prior felonies; 3) whether the defendant made incriminating statements; 4)

whether the metho

d of killing was dr

owning; 5) whether the method of killing was throat slashing; 6) whether

the murder was committed for pecuniary or other gain; 7) whether the murder was committed for sexual or other

gratification; 8) whether the mur

der was committed to escape appr

ehension or punishment; 9) whether ther

were three or more victims of the murder; 10) whether the victim was killed at place of business or employment;

and 11) whether the victim was killed in a field, woods, or rural area.

344 THE JUSTICE SYSTEM JOURNAL

the significant predictors, all of the effects are in the expected direction; that is, all these

factors increase the probability that a death sentence will be imposed. A nontechnical

way of gauging the overall explanator

y power of the model is to look at the improve-

ment in the ability to “pr

edict” correctly the number of cases in which the defendant

eceives the death penalty. In the absence of the model, we are able to correctly predict

individual cases 84 per

cent of the time simply on the basis that we know that 84 per-

cent of defendants did r

eceive the death penalty; that is, if we pr

edict that no defen-

dants receive the death penalty, we are correct 84 percent of the time. The model

improves our ability to predict correctly to 87 percent, a marginal improvement at best.

But what of the prosecutor’s decision to seek the death penalty? Here we should

note that the rate at which prosecutors seek the death penalty varies by county from

6 percent to 100 percent. Of course, at the extremes are counties with very small

Table 3

Multivariate Model: Whether Death Penalty is Imposed (n=968)

A priori Predicted

Observed Not Sentenced to Death Sentenced to Death Percent Correct

Not sentenced to death 815 0 100.0

Sentenced to death 153 0 .0

Overall 968 0 84.2

Model Predicted

Observed Not Sentenced to Death Sentenced to Death Percent Correct

Not sentenced to death 792 23 97.2

Sentenced to death 100 53 34.6

Overall 892 77 87.3

Table 4

Multivariate Model: Prosecutors’ Decisions to Seek the Death Penalty (n=968)

A priori Predicted

Observed Does not seek DP Seeks DP Percent Correct

Does not seek DP 626 0 100.0

Seeks DP 342 0 .0

Overall 968 0 64.7

Model Predicted

Observed Does not seek DP Seeks DP Percent Correct

Does not seek DP 574 52 91.7

Seeks DP 213 129 37.7

Overall 787 181 72.6

RACE, PROSECUTORS, AND JURIES: THE DEATH PENALTY IN TENNESSEE 345

numbers of cases. But looking at the counties with the most cases, one still sees

tremendous variation. In Shelby County, which contains Memphis, prosecutors

sought the death penalty 52 percent of the time, while in Davidson County

(Nashville) prosecutors pursued capital punishment only 12 per

cent of the time. To

some extent, this variance may be attributed to varying prosecutorial philosophies,

but it likely also represents political factors that vary from county to county.

When we take into account the other controls used in the overall model, the

defendant’s race and the victim’s race become nonsignificant (see Table 4). This

model performs better in terms of predictiveness. In the absence of the model, we are

able to make correct predictions about 65 percent of the time, but the model

impr

oves the rate of correct predictions to roughly 73 percent.

The results are similar for a jury’s decision to impose the death penalty (see

Table 5). When we examine only the subsample of cases (35 percent) where prose-

cutors actually sought capital punishment, we find, as we did with the model of prose-

cutor behavior, that the race of the victim is an insignificant predictor of jury behav-

ior once the other variables are taken into account. The race of the defendant is sig-

nificant, although the relationship is weak, but the effect is opposite of what some

might expect in that being an African-American defendant actually reduces the

chances of being sentenced to death. The model performs reasonably well, improv-

ing our ability to correctly predict who will get the death penalty from 56 percent to

75 percent. Thus, it would seem that juries are more predictable than prosecutors.

CONCLUSION

Few topics have generated as much concern as has the death penalty. Even if one

accepts the legitimacy of capital punishment, any systematic bias in its application

erodes claims that it has a role in furthering justice. Despite conventional wisdom to

Table 5

Multivariate Model: Jury Imposition of Death Penalty (n=342)

A priori Predicted

Observed Does not impose DP Imposes DP Percent Correct

Does not impose DP 191 0 100.0

Imposes DP 151 0 .0

Overall 342 0 55.8

Model Predicted

Observed Does not impose DP Imposes DP Percent Correct

Does not impose DP 156 35 81.7

Imposes DP 49 102 67.5

Overall 205 137 75.4

346 THE JUSTICE SYSTEM JOURNAL

the contrary, our study of a large population of first-degree murder convictions in a

southern state produced no evidence that race is a significant factor in either prose-

cutors’ decisions to seek the death penalty or juries’ decisions to impose it at sentenc-

ing. This is in no way an argument for the legitimacy of the death penalty. Indeed,

our data suggest that, despite reforms, application of the death penalty in Tennessee

remains quite unpredictable, and this in itself may well be cause for concern.

jsj

REFERENCES

Baldus, D. C., C. A. Pulaski, Jr., and G. Woodworth (1986). “Arbitrariness and Discrimination

in the Administration of the Death Penalty: A Challenge to State Supreme Courts,” 15

Stetson Law Review 133.

___

(1983). “Comparative Review of Death Sentences: An Empirical Study of the Georgia

Experience,” 74

Journal of Criminal Law and Criminology 661.

Bowers, W. (1984).

Legal Homicide: Death as Punishment in America, 1864-1982. Boston:

Northeastern University Press.

Ekland-Olson, S. (1988). “Structured Discretion, Racial Bias, and the Death Penalty: The

First Decade After

Furman in Texas,” 69 Social Science Quarterly 853.

Garfinkel, H. (1949). “Research Note on Inter- and Intra-Racial Homicides,” 27

Social Forces

369.

Gross, S. R., and R. Mauro (1989).

Death and Discrimination: Racial Disparities in Capital

Sentencing.

Boston: Northeastern University Press.

Heilbrun, A. B., Jr., A. Foster, and J. Golden (1989). “The Death Sentence in Georgia, 1974-

1987: Criminal Justice or Racial Injustice?” 16 Criminal Justice and Behavior 139.

Johnson, E. (1957). “Selective Forces in Capital Punishment,” 36

Social Forces 165.

Johnson, S. (2003). “Race and Capital Punishment.” In S. P. Garvey (ed.),

Beyond Repair?:

America’s Death Penalty

. Durham, NC: Duke University Press.

Kleck, G. (1981). “Racial Discrimination in Criminal Sentencing: A Critical Evaluation of

the Evidence with Additional Evidence on the Death Penalty,” 46 American Sociological

Review

783.

Klein, S. P., and J. E. Rolph (1991). “Relationship of Offender and Victim Race to Death

Penalty Sentences in California,” 32

Jurimetrics Journal 33.

Lemper

t, R. (1983). “Capital Punishment in the ‘80s: Reflections on the Symposium,” 74

Journal of Criminal Law and Criminology 1101.

Mangum, C. (1940).

The Legal Status of the Negro. Chapel Hill: University of North Carolina

ess.

Nakel, B., and K. Hardy (1987). The Arbitrariness of the Death Penalty. Philadelphia: Temple

University Press.

Pater

noster

, R. (1984). “Pr

osecutorial Discretion in Requesting the Death Penalty: A Case of

Victim-Based Discrimination,” 18

Law and Society Review 437.

RACE, PROSECUTORS, AND JURIES: THE DEATH PENALTY IN TENNESSEE 347

____

(1983). “Race of Victim and Location of Crime: The Decision to Seek the Death Penalty

in South Carolina,” 74

Journal of Criminal Law and Criminology 754.

Radelet, M., and M. Borg (2000). “The Changing Nature of Death Penalty Debates,” 26

Annual Review of Sociology 43.

Radelet, M. L., and G. L. Pierce (1985). “Race and Prosecutorial Discretion in Homicide

Cases,” 19 Law and Society Review 587.

Smith, M. (1987). “Patterns of Discrimination in Assessments of the Death Penalty: The Case

of Louisiana,” 15 Journal of Criminal Justice 279.

Vito, G. F., and T. J. Keil (1988). “Capital Sentencing in Kentucky: An Analysis of the Factors

Influencing Decision Making in the Post-Gregg Period,” 79 Journal of Criminal Law and

Criminology

483.

Zeisel, H. (1981). “Race Bias in the Administration of the Death Penalty: The Florida

Experience,” 95 Harvard Law Review 456.

CASES CITED

Gregg v. Georgia, 428 U.S. 153 (1976).

Furman v. Georgia, 408 U.S. 238 (1972).

State of Tennessee v. Godsey, 60 S.W.3d 759 (2001).