A TASTE FOR PUNISHMENT: Black and White Americans' Views on the Death Penalty and the War on Drugs

STATE OF THE ART

A TASTE FOR PUNISHMENT

Black and White Americans’ Views on the

Death Penalty and the War on Drugs

Lawrence D. Bobo

Department of Sociology and Department of African and

African American Studies, Harvard University

Devon Johnson

W. E. B. Du Bois Institute for African and African American Research,

Harvard University

Abstract

It is commonly accepted that Black and White Americans hold divergent views about the

criminal justice system. Furthermore, many accept the view that U.S. public opinion is

unflinchingly punitive where issues of criminal justice policy are concerned, with this

punitiveness among White Americans deriving to a significant degree from anti-Black

prejudice. Using a series of survey-based experiments and large, nationally representative

samples of White and African American respondents, we subject the questions of

Black-White polarization, unyielding punitiveness, and the influence of racial prejudice to

close scrutiny. Our results, first, confirm large Black-White differences in opinion with

Blacks consistently less punitive than Whites. These differences are substantially a result

of beliefs about the extent of racial bias in the criminal justice system. Second, the

framing experiments suggest that responses to the death penalty are very different than

responses to drug-related crimes like crack or powder cocaine use, with the former

exhibiting far less malleability than the latter. Third, racial prejudice is a consistently large

influence on White public opinion and a weaker, but sometimes important influence

among Blacks as well. Implications for discourse on race and crime are also discussed.

Keywords: Death Penalty, Cocaine, Crack, Public Opinion, Crime, Prejudice

INTRODUCTION

At the same time that our criminal justice policies have taken on an aggressive and

deeply punitive tenor, they have also fallen with special severity on minority com-

munities, particularly on African Americans+ These changes are sufficiently large

that many scholars now write of a qualitatively distinct policy-regime governing

Du Bois Review, 1:1 (2004) 151–180.

DOI: 10.10170S1742058X04040081

151

criminal justice issues in the U+S+ today ~Beckett and Western, 2001; Garland 2001;

Miller 1996; Wacquant 2001!+ To wit, after a century of relative stability, incarcera-

tion rates in the U+S+ are at an all time high and greatly exceed, by a factor of

anywhere from five times to ten times, those in all other Western industrial nations

~Garland 2001!+ Total annual expenditures on jails and prisons now approach fifty

billion dollars+ In some large states, like California, such spending now exceeds that

on higher education ~Chambliss 1999!+ Crime is now such a large preoccupation of

what government does at all levels that issues of crime and control occupy a place in

the national political discourse unheard of prior to the mid-1960s ~Beckett 1997;

Parenti 1999!+ With respect to African Americans, the years from 1980 to 2000

witnessed a near tripling of the rate of Black incarceration, increasing the Black-

White disparity in the odds of incarceration to an extreme of eight to one ~Blumstein

2001!+ Although African Americans constitute only 12% of the U+S+ population, they

currently make-up about half of those in jail or in prison, and constitute more than

half of those on death row ~Bureau of Justice Statistics 2002!+

The social underpinnings and effects of the new criminal justice policy regime

are the subject of far greater controversy+ A number of factors doubtlessly contribute

to the current state of criminal justice policy, ranging from the nature and actual rate

of crime, the larger economic and political contexts in which crime and criminal

justice policy formation occur, and prevailing ideas about the causes and appropriate

responses, both practically and morally, to violations of societal rules+ A major factor

mentioned in many lay and scholarly accounts, and the one that is the center of our

attention, is the putatively staunch “law and order” appetite of public opinion in the

U+S+, and the possibility that such outlooks are heavily rooted in anti-Black racial

prejudice+

Indeed, a number of scholars see in changing U+S+ criminal justice policy a delib-

erate, if loosely coordinated, effort to re-assert control and domination over African

Americans+ The elevation of so-called social disorder and “lawlessness” to matters of

national political discourse was first undertaken by Republican presidential nominee

Barry Goldwater in his 1964 campaign+ This was a time when the struggle over Black

civil rights was widely recognized as “the nation’s most important problem” ~Schuman

et al+, 1997!+ This often racialized law and order message was subsequently honed to

precision by Richard Nixon in 1968, just as the Black Power movement grew in po-

tency ~Beckett 1997!+ Parenti ~1999! argues that the turn toward harsh criminal justice

policies became necessary to maintain class and racial hierarchies in the wake of Black

political mobilization, labor law changes, and re-structuring global capitalism+ More

specifically, the reintroduction of the harshest form of punishment—the death penalty—

has been credited by some analysts to a decidedly racial politics+ Accordingly, “Whites

from all classes successfully demand enhanced criminal punishments in areas with the

largest minority populations” ~ Jacobs and Carmichael, 2002, p+ 111!+ Indeed, these

researchers find strong and consistent support for a racial threat hypothesis which holds

that the death penalty is most likely to exist in states with large African American pop-

ulations+ Net of other economic, crime related, and ideological indicators, Jacobs and

Carmichael find that the effect of the size of the Black population in their analyses of

state-level data from 1971 to 1991 is so large and durable that:

+++ our most important results concern race+ Findings that document a close

relationship between the presence of African Americans in a state and the pres-

ence of the death penalty may not be surprising, but they show how important

the politics of racial division and the racial foundations of punishment continue

to be in the United States ~ Jacobs and Carmichael, 2002, p+ 128!+

Lawrence D. Bobo and Devon Johnson

152

DU BOIS REVIEW: SOCIAL SCIENCE RESEARCH ON RACE 1:1, 2004

Casting an even broader theoretical net, sociologist Loic Wacquant characterizes

current rates of Black incarceration and changes in policing practices and policy as

tantamount to a new “peculiar institution,” or fourth stage of American racial oppres-

sion aptly termed the “carceral state” ~Wacquant 2001!+

We wish to subject three aspects of these notions about the popular underpin-

nings of the new racialized “law and order” regime to careful empirical scrutiny+

First, we take as problematic the assumption of an unflinching punitive “law and

order” tilt of U+S+ public opinion on crime+ A close reading suggests that there is a

measure of ambiguity in the extant research literature on the degree of flexibility and

on the underlying sources of attitudes on criminal justice policy issues+ This is

particularly true as regards the degree of support for policies such as the death

penalty and some aspects of the now more than decade old “war on drugs+”

Second, we hope to increase our leverage on the magnitude and sources of

Black-White differences in attitudes on criminal justice policy+ Many studies and

polls have pointed to substantial racial differences in opinion about criminal justice+

The differences are so large and recurrent as to begin to raise basic questions about

the democratic legitimacy of the criminal justice system, at least in the eyes of

African Americans ~Blumstein 2001; Hagan and Albonetti, 1982; Hawkins 2001;

Kennedy 2001!+ Ye t, few studies have had large enough samples of African American

respondents to shed much light on these differences+ Fewer still have combined large

samples of both Blacks and Whites with systematic measures of other theoretically

relevant variables such as the general salience of, or perceived vulnerability to crime,

perceived racial bias in the criminal justice system, and racial prejudice+ We are in a

position to do both+

These two concerns have a natural point of intersection+ Namely, third, we seek

to pinpoint just how substantial a role individual level anti-Black racial prejudice

plays in the contemporary “taste for punishment+” Some analysts suggest that anti-

Black prejudice is a critical source of staunchly punitive criminal justice views among

many White Americans+ Such a factor may serve to sharply differentiate the well-

springs of Whites’ opinion on crime issues from those observed among Blacks+ And

by the same token, Blacks’ policy views on crime may be driven more by the practical

exigency of controlling crime in the neighborhood rather than by an array of general

sociopolitical attitudes and beliefs ~Cohn and Barkan, forthcoming; Meares and Kahan,

1998; Young 1991!+

In pursuing answers to these queries, we also seek to move beyond simple

correlational analyses in understanding the contours of public opinion on criminal

justice policy+ Specifically, we report on a series of survey-based experiments that

manipulate cues about bias in the criminal justice system, both racial- and non-racial,

given to respondents ~Schuman and Bobo, 1988!+

Doing so allows us to simulta-

neously test important theoretical claims about the underlying foundations of public

opinion on criminal justice policy ~Hurwitz and Peffley, 1997!, and to simulate or

mimic major features of larger political discourse on these issues ~Kinder and Sand-

ers, 1996!+ In contrast to purely correlational analyses, this strategy gives us greater

purchase on the malleability of opinion and a capacity to pinpoint racial differences

in patterns of response+

BACKGROUND

Criminal justice policies range from policing practices and policies, to prosecutorial

and judicial authority, to the functioning of the jails, prison, probation, and parole

A Taste for Punishment

DU BOIS REVIEW: SOCIAL SCIENCE RESEARCH ON RACE 1:1, 2004 153

monitoring systems+ Two major thrusts of criminal justice policy over the last two

decades have garnered significant attention+ First, the death penalty was re-introduced

in a number of states+

A second policy, introduced as part of the war on drugs, was

the imposition of particularly severe sentences for involvement with certain types of

drugs ~i+e+, crack cocaine!+ We focus on public support for these two policies+ The

literature on criminal justice issues and public opinion is voluminous and in many

ways points to exceedingly complex, if not contradictory, patterns of results ~for

recent reviews see Cullen et al+, 2000 and Roberts and Stalans, 2000!+ One pivotal

source of debate and one robust empirical finding lead us to several hypotheses

outlined below+ The debate involves a clear point of conflicting results and claims:

namely, whether mass opinion on crime is relatively malleable and open to suasion

based on information or, in contrast, is relatively fixed and emotionally-based+ The

empirical result involves a matter of relative consensus: namely, the often large

Black–White differences on views about criminal justice+

The Malleable Public?

Some research suggests that the American public is more flexible in its thinking

about criminal justice policy issues than is often assumed+ First, when allowed more

qualified or nuanced options, many respondents endorse less punitive policy out-

comes+ Thus, studies that use vignettes to compare global and specific attitudes or

allow respondents to choose from a variety of sentences suggest that support for

punitive policies is “mushy” and that information can change opinion+ Vignette

studies demonstrate that support for harsh policies declines when respondents have

more information about the criminal or about the circumstances of the crime ~Apple-

gate et al+, 1996; Cullen et al+, 2000; Zamble and Kalm, 1990!+ Studies that offer

respondents a choice between various sentencing options also suggest that policy

preferences are flexible+ For instance, several academic studies find that support for

the death penalty drops significantly when life without parole ~and especially life

without parole plus restitution! is offered as a specific alternative to capital punish-

ment ~Bowers 1993; McGarrell and Sandys, 1996; Moon et al+, 2000; Sandys and

McGarrell, 1995!+ A number of Gallup Polls since 1985 report similar findings

~ Jones 2001!+ Taken together, these studies suggest that public opinion on criminal

justice issues is not unalterably set on the most punitive position+

To be sure, there is much evidence of a punitive tilt in public opinion on crime+

Ye t, as Cullen et al+ ~2000! stressed: “This punitiveness is not fixed on a single point

but is ‘mushy+’ Even when expressing punitive opinions, people tend to be flexible

enough to consider a range of sentencing options, including sanctions that are less

harsh than those they may have favored either at first thought or when provided with

only minimal information on which to base their views+++They moderate punitive-

ness when less stringent interventions have utility for victims, the community, and

offenders” ~Cullen et al+, 2000, p+ 8!+ If the public is more open and malleable, we

hypothesize that framing criminal justice policies as unfair or biased should, all else

equal, move public opinion in a less punitive direction ~Hypothesis 1!+

Second, the demographic correlates of attitudes on criminal justice views suggest

potential malleability+ Political scientist Shmuel Lock ~1999! has recently stressed

the importance of education and the openness it brings to new information as

influences on public opinion about law and order issues+ To the extent that better

educated people express less punitive points of view, and to the extent those who are

better informed express less punitive points of view, there may be grounds to expect

even greater movement toward a less punitive tilt in the future, or at least substantial

Lawrence D. Bobo and Devon Johnson

154

DU BOIS REVIEW: SOCIAL SCIENCE RESEARCH ON RACE 1:1, 2004

openness to elite leadership in that direction+ Accordingly, we hypothesize that bias

framing effects should interact with education such that it is the better educated who

should show the greatest sensitivity to bias framing ~Hypothesis 2!+

The general logic of these patterns is to suggest that public opinion on criminal

justice is importantly reasoned and tempered rather than unreasoned and obdurately

harsh+ If so, it is legitimate to hypothesize that framing particular criminal justice

policies as unfair or biased should substantially move public opinion+ To some degree,

of course, a test of these hypotheses rests on the ability to successfully depict a policy

or outcome as inequitable+ For the purposes of the tests reported below, we rely on

actual levels of disparity or inequality in framing tests of support for criminal justice

policies+ That is, our experiments focus on providing respondents with “the facts”

and are not intended to be a search for particularly appealing political rhetoric+ We

believe this is the appropriate starting point for our research given our theoretical

interest in the nature of public opinion ~as opposed to an advocacy agenda where the

goal might appropriately be to find the rhetoric or strategy that works best!+

The Rigid Public?

In contrast, other research indicates that public opinion on criminal justice matters

may be much more fixed and inflexible than the studies reviewed above suggest+

Previous research indicates that punitive attitudes are largely rooted in certain fairly

stable underlying social psychological predispositions+ Central among these predis-

positions are sociopolitical ideology and racial prejudice ~Barkan and Cohn, 1994;

Cohn et al+, 1991; Ellsworth and Gross, 1994; Ellsworth and Ross, 1983; Johnson

2001, Tyler and Boeckmann, 1997; Tyler and Weber, 1982!+ Given the sources of

these attitudes, and the fact that political and social values are developed early in life

and are rather stable, it is likely that views on the criminal justice system are rather

rigid and resistant to change+ As Tom Tyler explained: “Since social values represent

long-term political orientations, they reflect a stable influence on public opinion and

are unlikely to change in reaction to contemporary public events+ Hence, the strong

influence they have over punitiveness suggests that current levels of public support

for punitiveness are not simply the result of highly visible events like the Polly Klaas

kidnapping+ Instead, they develop from underlying social values that are stable and

that will shape public views for the near future” ~Tyler and Boeckmann, 1997,

p+ 199!+ Indeed, studies have found that information does little to shift opinion on the

death penalty ~Bohm et al+, 1991; Ellsworth and Ross, 1983; Sarat and Vidmar,

1976!+ Even Lock ~1999!, who had hoped to find substantial effects of information

levels, produced at best mixed results in terms of reducing punitiveness+ And Johnson

~2001!, using General Social Survey data for the period of 1974–1994, has recently

shown that indicators of racial prejudice strongly predict Whites’ levels of support

for the death penalty and their desire for harsher courts+

To the extent that harsh criminal justice policy views are based in stable social

psychological orientations such as political ideology and racial prejudice, we should

find that bias framing has little effect on policy views ~Hypothesis 3!+ That is, cueing

a form of inequity or bias should not greatly influence levels of support for criminal

justice policies because the roots of such views have a moral and normative founda-

tion unlikely to be sensitive to information about the problematic aspects of policy

implementation or consequences+ In addition, we should find consistent effects of

demographic background characteristics such as the respondent’s level of education

~Hypothesis 3A!+

A Taste for Punishment

DU BOIS REVIEW: SOCIAL SCIENCE RESEARCH ON RACE 1:1, 2004 155

Black/White Differences on Crime and the Role of Prejudice

Numerous studies identify substantial differences between Blacks and Whites on

issues of criminal justice+ These differences emerge in views of police behavior and

practices, the functioning of prosecutors and the courts, and a range of policy views

on how to respond to crime and punish criminals+ The general pattern of these

differences shows Blacks to be more skeptical of the equal application of law ~Hagan

and Albonetti, 1982; Henderson et al+, 1997; Hurwitz and Peffley, 2001; Wortley

et al+, 1997!, more critical of the police and their tactics ~Huang and Vaughn, 1996;

Peek et al+, 1981; Smith and Hawkins, 1973; Taylor et al+, 2001; Tuch and Weitzer,

1997!, less supportive of harsh punishments ~Blumstein and Cohen, 1980; Bohm

1991; Combs and Comer, 1984; Miller et al+, 1986; Secret and Johnson, 1989!, and

more supportive of rehabilitative strategies ~Gerber and Engelhardt-Greer, 1996;

McCorkle 1993!+ Thus, all else equal, we hypothesize that African Americans should

express less punitive views than Whites and show greater responsiveness to bias

framing cues than Whites ~Hypothesis 4!+

More important than this baseline expectation of race-based differences, how-

ever, is the goal to increase our leverage on understanding the sources of Black-

White differences+ Previous research points to three major possibilities regarding

Black-White differences in criminal justice views+ First, Blacks are more likely to be

the victims of crime than Whites and to live in communities with higher levels of

crime and disorder ~Blumstein 2001; Hawkins 2001; Kennedy 2001!+ As a result,

African Americans usually express higher levels of fear and concern about crime than

do Whites+ On this basis it is plausible to expect that the heightened general salience

of crime in the lives of African Americans may encourage greater punitiveness in the

criminal justice policy views of Blacks than is likely to be found among Whites+ As

legal scholars Tracey Meares and Don Kahan ~1998! argue, residents of inner city

Black communities may be more ready for, and see themselves more in need of,

draconian crime control measures than do Whites+ They suggest that for many

Blacks it is far more important to fight crime, regain “control of the streets,” and

assert a strong position about the norms that should govern youth activity than it is

to scrupulously preserve traditional civil liberties protections+ Hence, for example,

Meares and Kahan argue that Blacks will support many harsh law enforcement

strategies such as tough curfews, loitering, and other anti-gang ordinances that

might appear as unwarranted state intrusions on individual freedom to many sub-

urban Whites+ Accordingly, they maintain Blacks may have a strong taste for expand-

ing police powers and punishing crime:

What the civil libertarian critique ignores is the grounding of these laws in the

efforts of inner-city African Americans to shape their own norms+ Far from

opposing curfew and gang-loitering laws, the African American residents of the

inner city have in fact supported much of the impetus for the resurgence of them

~Meares and Kahan, 1998, p+ 830!+

At the same time, many studies of White public opinion have found that the general

salience of crime does little to influence Whites’ views on criminal justice policy+ As

a result, we hypothesize that the salience of crime should matter more for the policy

views of African Americans than it does for Whites ~Hypothesis 5!+

Second, beliefs about the fundamental fairness of the criminal justice system may

help account for different policy views in this domain between Blacks and Whites+

Much prior research suggests that African Americans are far more likely to believe

Lawrence D. Bobo and Devon Johnson

156

DU BOIS REVIEW: SOCIAL SCIENCE RESEARCH ON RACE 1:1, 2004

that the criminal justice system is riddled with systematic bias on the basis of race

than are Whites+ To the extent this is so, perceptions of bias in the functioning of the

criminal justice system may account for much of the Black-White difference in

support for specific harsh punishments or other new anti-crime initiatives ~Hypoth-

esis 6!+ Both the overall level of support for a particular criminal justice policy and

reactions to experimental manipulations that frame issues of punishment in terms of

potentially racially differential effects are likely to be contingent on the degree of

perceived racial bias in the criminal justice system as a whole+ Thus, while we expect

perceptions of bias to influence the views of Blacks and Whites, these effects should

be more pronounced among African Americans+

Third, a handful of studies suggest that anti-Black racial prejudice may be a key

factor differentiating the crime policy views of Blacks and Whites ~Cohn et al+, 1991;

Hurwitz and Peffley, 1997; Johnson 2001!+ Issues of race and crime have a long

history of interconnection in the U+S+ In particular, ideologies of anti-Black inferi-

ority in the U+S+ have consistently depicted Blacks as more violent and prone to

crime ~Du Bois 1996 @1899#; Fredrickson 1971!+ As such a central component not

only of American race ideology, but also of American jurisprudence, we would

naturally expect those Whites scoring higher in racial prejudice to support an array

of criminal justice policies that fall with greater severity on African Americans+

It is

entirely possible that the views of African Americans may also be influenced by

anti-Black prejudice+ Situated amidst a larger culture that puts forward many nega-

tive images and ideas about African Americans and on the basis of actually experi-

encing more adverse living conditions, some African Americans may come to accept

elements of a larger anti-Black perspective or ideology ~Adams 1978; Clark 1965;

Cross 1991; Grier and Cobbs, 1968!+ Social psychological research suggests impor-

tant asymmetries in the views of members of dominant or privileged groups as

compared to subordinate or minority groups ~Sachdev and Bourhis, 1991; Tajfel

1981, especially p+ 323!+ As a minority group, Blacks are not immune to a social

environment where large fractions of the majority ~dominant! social group hold

negative beliefs about Blacks and where the culture has long projected debasing

images of Blacks+ Immersion in such a setting should influence thinking with the

minority group as well ~Sidanius and Pratto, 1999!, though in the context of a long

history of struggle, consciousness raising, and resistance, many Blacks should be able

to stave-off the influence of the prevailing culture of stigmatization and resentment

~ Jackman 1994!+ This situation leads us to hypothesize that racial prejudice should

increase punitiveness among Blacks and Whites, though overall levels of prejudice

among Blacks should be lower and the impact of prejudice on punitiveness should be

weaker than that observed among Whites ~Hypothesis 7!+

DATA AND MEASURES

The data for this paper come from the 2000 National African American Election Study

~NAAES! and from the 2001 Race, Crime and Public Opinion Study ~RCPOS!+ Each

survey is a representative national sample of White and Black Americans living in U+S+

households+ The NAAES was conducted in the fall of 2000 and included 831 African

Americans and 724 Whites+ The Race, Crime and Public Opinion Study was con-

ducted by the authors in the summer of 2001, and included 978 White and 1,010 Black

respondents+ Data for both studies was collected by Knowledge Networks using new

Internet-based survey technology+ See the Appendix for detailed information on sur-

vey methodology and sample characteristics for the NAAES and the RCPOS+

A Taste for Punishment

DU BOIS REVIEW: SOCIAL SCIENCE RESEARCH ON RACE 1:1, 2004 157

The analyses below include several demographic and background variables+ Race

is treated as a dummy variable labeled “White,” and is coded 1 for White, 0 for Black+

Education is a nine-category variable, ranging from “less than high school” ~coded 1!

to “doctorate degree” ~coded 9!+ Political conservatism is a five category variable,

ranging from “very liberal” ~coded 1! to “very conservative” ~coded 5!+

To measure perceived racial bias in the criminal justice system, respondents were

asked, “In general, do you believe the criminal justice system ‘is biased in favor of

blacks’ ~coded 1!, ‘gives blacks fair treatment’ ~coded 2!, or ‘is biased against blacks?’”

~coded 3!+ Salience of crime is an index created by averaging the responses to three

items—perceptions of the crime problem in one’s neighborhood, fear of having one’s

house burglarized, and fear of being robbed—and ranges from low ~coded 1! to high

~coded 4!+ The alpha for the Crime Salience Index is +74+ Finally, as an indicator of

prejudice we use the Racial Resentment Scale, created by averaging responses to the

following six items:

• Irish, Italian, Jewish and many other minorities overcame prejudice and

worked their way up+ Blacks should do the same without any special favors+

• Over the past few years, Blacks have gotten less than they deserve+

• Government officials usually pay less attention to a request or complaint

from a Black person than from a white person+

• Most Blacks who receive money from welfare programs could get along

without it if they tried+

• It’s really a matter of some people not trying hard enough; if Blacks would

only try harder, they could be just as well off as Whites+

• Generations of slavery and discrimination have created conditions that make

it difficult for Blacks to work their way out of the lower class+

Response options for these items ranged from “strongly disagree” ~coded 1! to

“strongly agree” ~coded 5!~some items were reversed scored!+ Those who answered,

“don’t know,” or refused, were coded at the midpoint ~3!+ Scores on the Racial

Resentment Scale range from 0 ~low racial resentment! to 5 ~high racial resentment!+

The alpha for the Racial Resentment Scale is +77+

The variables and coding for analyses using the NAAES data differ slightly from

those above+ In these analyses, perceived racial bias is measured with a different item;

respondents were asked to indicate which statement is most true: a! “Generally, the

American legal system treats all groups fairly” or b! “The American legal system is

unfair to Blacks+” Respondents who answered that the system is fair to all groups were

coded 1, those who answered that the system is unfair to Blacks were coded 3, and

those who answered “both,” “neither” or “don’t know” were coded at the midpoint ~2!+

These analyses also include a measure of common fate identity: “Do you think what

happens generally to Black people in this country will have something to do with what

happens in your life?” Response options included “a lot” ~coded 4!, “some” ~coded 3!,

“not very much” ~coded 2!, and “none” or “don’t know” ~both coded 1!+

ANALYSIS AND RESULTS

Experiment One: Death Row Demography

In order to gauge the potential malleability of public opinion on criminal justice

issues we began with a focus on the most severe form of punishment the state can

Lawrence D. Bobo and Devon Johnson

158

DU BOIS REVIEW: SOCIAL SCIENCE RESEARCH ON RACE 1:1, 2004

apply to criminals, namely the death penalty+ Specifically, we call our first experiment

the “Death Row Demography” experiment+ It focuses on the heavy overrepresenta-

tion of Blacks on death row as compared to their numbers in the general population+

We sought to determine if, by drawing attention to this racial disproportionality

alone, public opinion might shift in a less punitive direction+ Any number of social

science critiques of the criminal justice system point to just such racial disparities as

evidence of a problem of serious racial bias ~Cole 1999; Kennedy 2001; Miller 1996;

Parenti 1999; Sidanius and Pratto, 1999; Tonry 1995!+ Still, we conceived of this

manipulation as but a weak test of a racial cue effect+ Committing the grave offense

of homicide is essentially a pre-requisite to death row status; hence, even if the

question cues a substantial racial inequity, it may have only a modest mitigating effect

on death penalty support+ In order to enhance the contrast, we employed a three-way

split-ballot with one-third of the sample receiving a baseline or standard death

penalty question with no racial cue, one-third receiving the positive racial cue of

Black overrepresentation on death row, and one-third receiving a negative racial cue

of disproportionate Black commission of homicides+ We assume the latter condition,

highlighting as it does the commission of a deadly violent crime, should elicit the

strongest level of support for the death penalty+ And the disproportionate involve-

ment of Blacks in the commission of violent offenses is precisely the sort of pattern

that more conservative social science analysts tend to emphasize ~Thernstrom and

Thernstrom, 1997!+ Exact question wording and basic distributions by race and

experimental ballot are shown in Table 1+

Does explicitly mentioning the overrepresentation of Blacks on death row decrease

support for the death penalty? The results of Table 1 show that for neither the Black

nor the White respondents did our experimental manipulation influence the level of

death penalty support+ Better than 80% of Whites favor ~combining responses of

strongly or somewhat! the death penalty across all three conditions and, if anything,

including a racial cue of any kind slightly ~insignificantly so! reduces death penalty

support+ Somewhat better than 50% of Blacks support the death penalty ~combining

responses of strongly or somewhat!, with a negligible rise in the percentage support-

ing the death penalty when the question mentioned the percentage of Blacks com-

mitting homicides+

The absence of a strong main effect of the racial disproportionality cue does not

mean that no important experimental effects exist+ As Kinder and Sanders ~1996!

stressed, when public policy questions are framed in importantly different ways they

may well elicit different resonances in the mass public+ If so, the absence of a main

experimental effect can mask important underlying differences in patterns of corre-

lation with respondent background characteristics and social psychological outlooks+

The analyses reported in Table 2 begin to address this possibility+

The results of

Model 1 confirm the absence of an experimental effect+ However, there is a substan-

tial and highly significant Black-White difference+ Whites are substantially more

supportive of the death penalty than are Blacks+ As Model 2 shows, however, there is

no significant interaction between experimental ballot and respondent race+ White

and Black respondents were equally immune to this experimental manipulation+

Model 3 reports results for only the White respondents+ Ultimately, our interest

is to determine whether the racial cue manipulation interacts with respondent back-

ground characteristics or key psychological measures like political ideology+ That is,

even if the overall distribution did not change, the better educated as compared to

the poorly educated, and the political liberal as compared to the political conserva-

tive, may well have moved in response to the experimental manipulation+

As model

4 shows, however, none of the interactions are statistically significant+ We should

A Taste for Punishment

DU BOIS REVIEW: SOCIAL SCIENCE RESEARCH ON RACE 1:1, 2004 159

note that independent tests for each possible interaction also failed to reach conven-

tional criteria of significance+

Models 5 and 6 report results separately for the Black respondents+ In this case

we are interested not only in the possibility of differences based in background

characteristics and in political ideology, but also based in perceptions of racial bias in

the criminal justice system and levels of group identification+ That is, Blacks who see

the criminal justice system as systematically biased on the basis of race or who

generally express high levels of common fate identification may be especially sensi-

tive to the manipulation of racial cues in the experiment+ Although Blacks scoring

higher in common fate identity are indeed generally less likely to support the death

penalty, there is no interaction with the experimental manipulation+ However, per-

ceptions of racial bias in the criminal justice system do not reduce support for the

death penalty ~in the context of the common fate identity measure!+ The absence of

a main experimental effect among either Blacks or Whites and the general absence

of interactions with other factors may mean that for most Americans the application

Table 1. Death row demography experiment

Ballot A

Basic Question

Ballot B

Blacks on Death

Row Cue

Ballot C

Black Offenders

Cue

Blacks are about

12% of the U+S+

population, but

they are almost

half ~43%! of

those currently

on death row+

Blacks are about

12% of the U+S+

population, but

they were half of

the homicide

offenders last

year+

Do you favor or

oppose the

death penalty

for persons

convicted of

murder?

Do you favor or

oppose the

death penalty

for persons

convicted of

murder?

Do you favor or

oppose the

death penalty

for persons

convicted of

murder?

White Respondents

Strongly favor 49% 52% 55%

Somewhat favor 37 30 26

Somewhat oppose 7 10 14

Strongly oppose 8 8 8

Chi square 5 11+95 ns

~p 5 +06!

~244!~237!~232!

Black Respondents

Strongly favor 18 18 14

Somewhat favor 35 35 41

Somewhat oppose 22 24 22

Strongly oppose 24 24 24

Chi square 5 2+57 ns ~190!~189!~200!

Source: NAAES, 2000+

Lawrence D. Bobo and Devon Johnson

160

DU BOIS REVIEW: SOCIAL SCIENCE RESEARCH ON RACE 1:1, 2004

of the ultimate penalty for the ultimate crime operates as a fairly strong general

principle+ Minimally, the results suggest that the mere racial inequity cue we devel-

oped here does little to move significant numbers of respondents and does not lead to

a sufficient re-framing of the issue so as to tap into distinctive points of resonance in

the mass public+

Experiment Two: Murder Victim Race Bias

We now shift attention to one of the remaining signs of clear racial bias in the

application of the death penalty: People who murder Whites, whether they be Black

Table 2. OLS regression analysis of the death row demography experiment

Model 1

Pooled

Model 2

Pooled

Model 3

Whites

Model 4

Whites

Model 5

Blacks

Model 6

Blacks

Constant 2+55*** 2+56*** 3+53*** 3+49*** 2+65*** 3+03***

~+15!~+16!~+21!~+29!~+32!~+57!

Demographic Variables

White +74*** +73*** ————

~+06!~+10!

Education +01 2+01 2+07** 2+02 +09** +04

~+02!~+02!~+02!~+04!~+03!~+06!

Experimental Ballot

Blacks on death row ballot 2+04 2+06 2+02 +30 2+07 2+61

~+07!~+11!~+09!~+39!~+11!~+74!

Black offenders ballot 2+05 2+05 2+06 2+42 2+04 2+64

~+07!~+10!~+09!~+43!~+11!~+75!

Social Psychological Variables

Political conservatism +07* +07* +05 +001 +09 +02

~+03!~+03!~+04!~+07!~+05!~+10!

CJS biased against Blacks ————2+05 +05

~+08!~+14!

Common fate identity ————2+07* 2+18**

~+04!~+07!

Interactions

White * Death row ballot — +03————

~+14!

White * Offenders ballot — 2+004————

~+14!

Education * Death row ballot — — — 2+08 — +11

~+06!~+08!

Education * Offenders ballot — — — 2+06 — +04

~+06!~+08!

Conservatism * Death row ballot — — — +003 — +12

~+10!~+13!

Conservatism * Offenders ballot — — — +19 — +08

~+11!~+14!

CJS bias * Death row ballot —————2+24

~+19!

CJS bias * Offenders ballot —————2+07

~+19!

Common fate * Death row ballot —————+17

~+09!

Common fate * Offenders ballot —————+16

~+09!

Adjusted R Square +15 +15 +01 +01 +05 +06

N 1208 1208 670 670 523 523

Source: NAAES, 2000+

Note: Models also include controls for age, gender, and region+

*p,+05, ** p , +01, *** p , +001

A Taste for Punishment

DU BOIS REVIEW: SOCIAL SCIENCE RESEARCH ON RACE 1:1, 2004 161

or White, are more likely to receive a death sentence than those who murder

non-Whites+

Since in this case there is clearer evidence of actual racial bias ~not just

racial disparity!, we view it as a stronger test of the potential influence of a racial cue

on death penalty support+ Everything else equal, we expect that once respondents are

informed that the killer of a White person is more likely to receive the maximum

punishment than the killer of a Black person, the level of support for the death

penalty should decline+ Full question wording and the results of the experiment are

shown in Table 3+

Here we do find a statistically discernible effect of the experimental manipulation

among Blacks, but not among Whites+ Support for the death penalty among Blacks

falls from 63% in the basic question condition to 52% in the racial bias cue condition+

There is a very slight, but insignificant trend in the same direction among Whites,

where death penalty support falls from 77% to 75%+ Given the slight trend in the same

direction for both Whites and Blacks, the three-way interaction of race, experimental

ballot, and response does not attain conventional criteria for statistical significance+

The Murder Victim Race experiment data were collected as part of our Race, Crime,

and Public Opinion Study+ As a result, we have more extensive measures in these mod-

els including a measure of criminal justice bias against Blacks, a multiple item index of

the perceived salience of crime to the respondent, and a multiple-item measure of ra-

cial prejudice ~the racial resentment scale!+ Table 4 reports results including these mea-

sures in models pooled for Black and White respondents and separately by race+

The results for Whites show that those with less education, political conserva-

tives, the racially resentful, and those who see little racial bias in the criminal justice

are more supportive of the death penalty+ None of the tests for interactions between

the racial bias ballot and any of the background or social psychological measures met

conventional criteria for statistical discernability among the White respondents+

This indicates that with or without an explicit racial bias cue, two racially explicit sets

Table 3. Murder victim race experiment

@At present, someone who murders a white person is much more

likely to be sentenced to death than someone who murders a

black person+# Do you favor or oppose the death penalty for

persons convicted of murder?

Ballot A

Ballot B

Racial Bias Cue

White Respondents

Strongly favor 41% 40%

Mostly favor 36 35

Mostly oppose 14 17

Strongly oppose 9 9

Chi square 5 1+15 ns ~389!~386!

Black Respondents

Strongly favor 25 21

Mostly favor 38 31

Mostly oppose 20 28

Strongly oppose 18 20

Chi square 5 8+59* ~332!~374!

Source: Race, Crime and Public Opinion Study, 2001+

*p,+05

Lawrence D. Bobo and Devon Johnson

162

DU BOIS REVIEW: SOCIAL SCIENCE RESEARCH ON RACE 1:1, 2004

of attitudes—resentment and perceived racial bias—contribute to how Whites feel

about the death penalty+ Whatever else these results may show, emphasizing the

persistent bias in application of the death penalty to those who murder Whites does

nothing to enhance or reduce how many or even which Whites support the death

penalty+ Among Blacks, we also find that those who perceive racial bias in the

criminal justice system and those who are low in racial resentment are least likely to

support the death penalty+

Two further points are worthy of note here+ First, racial resentment has much

stronger effects on death penalty attitudes among Whites than among Blacks ~see

Table 10 below!+ Second, for neither Blacks nor Whites does the perceived salience

of crime matter for their views on the death penalty+ That is, concrete apprehensions

about criminal victimization do not appear to play an important role in how Amer-

icans feel about the death penalty+

Experiment Three: Innocent on Death Row

Since we had not produced large changes in death penalty support, at least among

White respondents, with either a racial disproportionality cue or a racial bias cue, we

developed a strong test of seemingly implacable death penalty support by testing a

race-neutral, but, we believe, powerful challenge to the death penalty: Namely, the

risk of executing innocent people+ We also wanted to put matters in a more realistic

Table 4. OLS regression analysis of the murder victim race bias experiment

Model 1

Pooled

Model 2

Pooled

Model 3

Whites

Model 4

Whites

Model 5

Blacks

Model 6

Blacks

Constant 2+52*** 2+56*** 2+10*** 2+36*** 3+11*** 3+19***

~+27!~+27!~+34!~+36!~+43!~+47!

Demographic Variables

White +14* +07————

~+07!~+09!

Education 2+04** 2+04** 2+05* 2+07** 2+04 2+05

~+02!~+02!~+02!~+03!~+03!~+04!

Experimental Ballot

Racial bias ballot 2+12* 2+19** 2+07 2+64* 2+18* 2+35

~+05!~+07!~+07!~+29!~+08!~+37!

Social Psychological Variables

Political conservatism +06* +06* +09* +04 +01 +01

~+03!~+03!~+04!~+05!~+05!~+07!

CJS biased against Blacks 2+20*** 2+20*** 2+15* 2+15* 2+20* 2+20*

~+06!~+06!~+07!~+07!~+10!~+10!

Salience of crime index +02 +02 +06 +07 2+03 2+02

~+04!~+04!~+05!~+05!~+06!~+06!

Racial resentment +29*** +29*** +36*** +36*** +20*** +20***

~+04!~+04!~+05!~+05!~+06!~+06!

Interactions

White * Racial bias ballot — +13————

~+10!

Education * Racial bias ballot — — — +05 — +02

~+04!~+05!

Conservatism * Racial bias ballot +12 +04

~+08!~+11!

Adjusted R Square +14 +14 +14 +15 +04 +03

N 1414 1414 735 735 678 678

Source: Race, Crime, and Public Opinion Study, 2001+

Note: Models also include controls for age, gender, and region+

*p,+05, ** p , +01, *** p , +001

A Taste for Punishment

DU BOIS REVIEW: SOCIAL SCIENCE RESEARCH ON RACE 1:1, 2004 163

political context so we focused on a potential gubernatorial candidate taking a stand

on the death penalty+ If the candidate’s stated position was framed in terms of the real

risk of executing innocent people, would support for the death penalty finally waver?

The short answer, unexpectedly, is not much+

Specifically, we asked respondents a hypothetical question about a candidate for

governor+ In the baseline condition the question notes only that the candidate has

announced her0his opposition to the death penalty “because there was too great a

risk that innocent people might be executed” and asks whether the respondent would

be more likely to vote for or against such a candidate+ In the “informed” condition,

the question opens with the statement: “Since 1976, seventy-nine people convicted

and sentenced to death were later found to be innocent and have now been released

from prison+” We regard this as the strongest of the framing cues considered here

because it essentially places before the respondent the possibility of the state execut-

ing an innocent person+ We expected to produce substantial movement in the direc-

tion of opposition to the death penalty; that is, toward greater support for the

candidate opposed to the death penalty+ However, Table 5 shows a shift of 7%

~borderline significance! among White respondents toward the “more likely to vote

for” response in the “informed” condition+ A shift in the same direction, but of only

2%, occurs among Black respondents+

Model 2 in Table 6 shows that there is no significant interaction of race and exper-

imental ballot+ Among the White respondents we find that political conservatism and

racial resentment enhance support for the death penalty ~reduce support for the candi-

date who has announced opposition to the death penalty!+ Perceptions of racial bias in

the criminal justice system had the opposite effect+ Among Blacks we find that racial

resentment increases support for the death penalty and that political ideology interacts

with the experimental ballot ~Model 6!+ The interaction shows that in the “informed”

condition, the impact of ideology among Blacks is reduced to a negligible size+

Table 5. Innocent on death row experiment

@Since 1976, seventy-nine people convicted and sentenced to death

were later found to be innocent and have now been released from

prison+# If a candidate for Governor of your state said that he or

she opposed the death penalty because there was too great a risk

that innocent people might be executed, would that make you more

likely to vote for this candidate, less likely to vote for this candi-

date, or not influence your vote?

Ballot A

Ballot B

Informed

White Respondents

More likely to vote for 17% 24%

Would not influence vote 49 46

Less likely to vote for 34 31

Chi square 5 5+62 ns ~p 5 +06!~401!~374!

Black Respondents

More likely to vote for 37 39

Would not influence vote 55 53

Less likely to vote for 9 8

Chi square 5 +46 ns ~373!~334!

Source: Race, Crime and Public Opinion Study, 2001+

Lawrence D. Bobo and Devon Johnson

164

DU BOIS REVIEW: SOCIAL SCIENCE RESEARCH ON RACE 1:1, 2004

One possible interpretation of the small or non-existent experimental effects in

the Death Row Demography, Murder Victim Bias, and Innocent on Death Row ex-

periments is that the information we provided in each of the bias cues is already known

to respondents+ Since the information is not “new” it does not provide a basis for shift-

ing opinions+ However plausible, we think this an unlikely explanation+ In general, we

know that the public does not have high levels of knowledge on many political issues

~Delli Carpini and Keeter, 1997!+ And although media coverage of death penalty cases

has been high in some selected places ~e+g+, in the state of Illinois following former

Governor Ryan’s decision to announce a death penalty moratorium!, it is far from con-

stituting a steady top news item nationwide+ To check on this possibility, however, we

examined whether poorly educated respondents showed any unusual sensitivity to the

experimental manipulations+ The poorly educated are generally less interested in pol-

itics, less informed on political matters, and less politically sophisticated in their think-

ing ~Kinder 1998; Zaller 1992!+ Hence, if political information or knowledge is key to

the general absence of effects, this should be the segment of the population in which

our experimental manipulations are most likely to introduce previously unknown facts

and to show the greatest impact on respondents+

To test this possibility we examined experimental results for the two lowest

education groups: those with less than a high school diploma level of schooling and

those who had only graduated from high school+ As Table 7 shows, whether we

looked at the results for Blacks and Whites combined, or just among the White

Table 6. OLS regression analysis of the innocent on death row experiment

Model 1

Pooled

Model 2

Pooled

Model 3

Whites

Model 4

Whites

Model 5

Blacks

Model 6

Blacks

Constant 1+95*** 1+91*** 1+86*** 1+96*** 2+05*** 1+69***

~+40!~+29!~+40!~+43!~+42!~+46!

Demographic Variables

White +27*** +35*** ————

~+07!~+08!

Education +02 +02 +02 +01 +01 +02

~+02!~+02!~+02!~+03!~+02!~+03!

Experimental Ballot

Innocent ballot 2+11* +02 2+19** 2+41 2+02 +70*

~+05!~+05!~+08!~+35!~+08!~+35!

Social Psychological Variables

Political conservatism +17*** +17*** +20*** +19** +13** +24***

~+03!~+03!~+05!~+06!~+05!~+07!

CJS biased against Blacks 2+19** 2+19** 2+22** 2+22** 2+01 +003

~+06!~+06!~+08!~+08!~+10!~+10!

Salience of crime index 2+001 2+002 +02 +02 2+04 2+04

~+04!~+04!~+06!~+06!~+05!~+05!

Racial resentment +28*** +27*** +36*** +36*** +16** +15*

~+04!~+04!~+06!~+06!~+06!~+06!

Interactions

White * Innocent ballot — 2+17————

~+08!

Education * Innocent ballot +04 2+03

~+05!~+05!

Conservatism * Innocent ballot — — — +02 — 2+22*

~+09!~+10!

Adjusted R Square +16 +16 +14 +14 +01 +02

N 1416 1416 736 736 679 679

Source: Race, Crime, and Public Opinion Study, 2001+

Note: Models also include controls for age, gender, and region+

*p,+05, ** p , +01, *** p , +001

A Taste for Punishment

DU BOIS REVIEW: SOCIAL SCIENCE RESEARCH ON RACE 1:1, 2004 165

respondents, there is no sign that important shifts occurred among the most poorly

educated in response to the experimental framing cues+ This, of course, is only an

indirect test of the possibility that information is the key to the relative absence of

framing effects, but it is strongly suggestive that the reasons for the non-findings lie

elsewhere, and are not an artifact of giving people information they already possessed+

Experiment Four: Crack versus Powder Cocaine

To this point we have focused on punishment for those who have committed the

most serious violent offense of homicide+ Generally, the results suggest only a small

degree of malleability of public opinion as a function of the framing of the issues+ As

we moved from framing cues of simple racial inequity, to more clear-cut racial bias,

to potentially irrevocable injustice, we find, surprisingly, only mild evidence of a

capacity to move the distribution or the underlying wellsprings of public opinion+

But these results may tell us more about how the public thinks about only the most

violent crime of homicide, not how reactions will play out in other important domains

of criminal justice+ Indeed, much of the debate on racial bias in the criminal justice

system focuses on the effects of the “war on drugs+” In particular, enormous attention

has centered on the sharp differential in sentences applied to those caught using

crack cocaine versus those caught using powder cocaine+

We sought to determine public support for the more severe sanctioning of crack

cocaine+ Under one condition, respondents were asked a baseline question on sup-

port for the dramatically more severe punishment of crack versus powder cocaine+ In

the Racial Bias cue condition the question opens by pointing out that “Most of those

convicted for crack cocaine use are Blacks and most of those convicted for powder

cocaine use are Whites+” The basic distribution of results is reported in Table 8+

Here there is strong evidence of a racial bias framing effect+ The level of support for

Table 7. Low education and means on the death penalty experiments

Pooled

Less than

HS Diploma

Pooled

HS Diploma

Only

Whites

Less than

HS Diploma

Whites

HS Diploma

Only

Death Row Demography

Ballot A 2+09 2+08 1+61 1+68

~61!~147!~29!~84!

Ballot B ~Blacks on Death Row! 2+71 1+90 1+74 1+54

~58!~137!~21!~77!

Ballot C ~Black Homicide Offenders! 1+98 2+29 1+33 1+81

~67!~149!~+31!~75!

F Score 2+40 ns 1+31 ns

Murder Victim Race Bias

Ballot A 2+91 3+13 3+22 3+31

~129!~215!~+57!~121!

Ballot B ~Racial Bias Cue! 2+85 2+81 3+35 3+07

~110!~274!~30!~136!

F Score +73 ns +56 ns

Innocent on Death Row

Ballot A 2+91 2+94 3+36 3+28

~122!~244!~50!~121!

Ballot B ~Informed! 2+61 2+86 2+94 3+08

~120!~243!~37!~130!

F Score 2+73 ns +00 ns

Source: NAAES, 2000 and Race, Crime, and Public Opinion Study, 2001+

Lawrence D. Bobo and Devon Johnson

166

DU BOIS REVIEW: SOCIAL SCIENCE RESEARCH ON RACE 1:1, 2004

the stiffer punishment of crack vs+ powder cocaine is cut in half among both White

and Black respondents in the racial bias condition+ This level falls from 44% approv-

ing of the differential ~combining responses of strongly and somewhat! to 22%

among Whites ~with most of this coming out of the strongly approve category!, and

from 32% to 16% among Blacks+ Most of this shift moves into the “strongly disap-

prove” category among both Whites and Blacks+

The magnitude of the racial bias effect is similar for Blacks and Whites, resulting

in an insignificant coefficient for test of the interaction of ballot with race, as shown

in Model 2 of Table 9+ Among Whites a number of variables contribute to views on

the crack vs+ powder issue+ In general, Whites who are poorly educated, politically

conservative, who deny the existence of racial bias in the criminal justice system, and

are racially resentful are more supportive of the differential punishment+ However, a

significant interaction with political ideology exists+ Conservatives are more likely to

support the differential than liberals in the baseline condition, but ideology has no

effect in the racial bias condition+

Among Blacks, those who deny the existence of racial bias, who worry most

about crime, and who harbor racial resentments, are the most likely to support the

crack vs+ powder sentencing differential+ Here too, however, important interactions

exist+ There is a strong interaction with perceived bias in the criminal justice system+

There is a very large negative effect of perceived racial bias in the criminal justice

system in the racial bias condition of the experiment ~more than twice the size of the

effect in the baseline condition!+

The crack versus powder cocaine experiment is the only case where we find a

significant impact of the salience of crime on policy views+ It is particularly telling

that this effect emerges only among the African American respondents+ It points to

the frictions within the Black community over the issue of drug-related crime+ On

Table 8. Crack versus powder cocaine experiment

@Most of those convicted for crack cocaine use are blacks and most

of those convicted for powder cocaine use are whites+# Under Fed-

eral law, the punishment for possession and distribution of crack

cocaine is 100 times more severe than for powdered cocaine+ Do

you approve or disapprove of much harsher punishments for crack

cocaine as compared to powdered cocaine?

Ballot A

Ballot B

Racial Bias Cue

White Respondents

Strongly approve 22% 9%

Somewhat approve 22 13

Somewhat disapprove 33 35

Strongly disapprove 23 44

Chi square 5 55+62*** ~360!~389!

Black Respondents

Strongly approve 19 11

Somewhat approve 13 5

Somewhat disapprove 22 20

Strongly disapprove 45 65

Chi square 5 25+61*** ~351!~462!

Source: Race, Crime and Public Opinion Study, 2001+

*** p , +001

A Taste for Punishment

DU BOIS REVIEW: SOCIAL SCIENCE RESEARCH ON RACE 1:1, 2004 167

the one hand, the general doubts about the fairness of the criminal justice system

greatly reduce the willingness of Blacks to support the sharply differential treatment

of a type of drug used more commonly in Black communities+ On the other hand,

fighting drug use and attendant crime is clearly a priority for those Blacks most

apprehensive about criminal victimization+

A Focus on Black-White Differences

Three of the four experiments came from the Race, Crime, and Public Opinion

Study which contained measures of perceived criminal justice system bias, crime

salience, and racial prejudice+ Each of these factors may be implicated in the Black–

White differences in levels of punitiveness observed in the above analyses+ Since we

did not find significant interactions between respondent race and any of the experi-

mental ballots, we pooled the data across experimental conditions and focused atten-

tion on Black–White differences in the effects of these measures+ ~See Table 10+!

Three consistent patterns emerge regarding Black–White differences in punitive-

ness+ First, across the three experiments, the dummy variable for race always produces

the single largest effect in the models+ These effects range in size from about a third of

a point to almost two-thirds of point, recalling that each dependent measure isa1to4

scale, with standard deviations of approximately 1+0+ Second, introducing the mea-

Table 9. OLS regression analysis of the crack versus powder cocaine experiment

Model 1

Pooled

Model 2

Pooled

Model 3

Whites

Model 4

Whites

Model 5

Blacks

Model 6

Blacks

Constant 1+82*** 1+78*** 1+69*** 1+53** 2+12*** 1+33**

~+27!~+27!~+37!~+40!~+39!~+50!

Demographic Variables

White 2+03 +03————

~+07!~+09!

Education 2+06*** 2+06*** 2+07** 2+10*** 2+04 2+04

~+02!~+02!~+02!~+03!~+02!~+03!

Experimental Ballot

Racial bias ballot 2+50*** 2+45*** 2+56*** 2+09 2+43*** 1+04

~+05!~+07!~+07!~+32!~+07!~+61!

Social Psychological Variables

Political conservatism +07* +07* +11** +24*** +04 +10

~+03!~+03!~+04!~+06!~+05!~+07!

CJS biased against Blacks 2+25*** 2+25*** 2+18* 2+18** 2+38*** 2+17

~+06!~+06!~+07!~+07!~+09!~+12!

Salience of crime index +10** +10** +06 +05 +14** +14**

~+04!~+04!~+06!~+06!~+05!~+05!

Racial resentment +23*** +23*** +18** +16** +29*** +29***

~+04!~+04!~+06!~+06!~+06!~+06!

Interactions

White * Racial bias ballot — 2+11————

~+10!

Education * Racial bias ballot — — — +06 — 2+01

~+04!~+05!

Conservatism * Racial bias ballot — — — 2+24** — 2+11

~+08!~+10!

CJS biased * Racial bias ballot —————2+39*

~+17!

Adjusted R Square +15 +15 +15 +16 +12 +12

N 1495 1495 715 715 779 779

Source: Race, Crime, and Public Opinion Study, 2001+

Note: Models also include controls for age, gender, and region+

*p,+05, ** p , +01, *** p , +001

Lawrence D. Bobo and Devon Johnson

168

DU BOIS REVIEW: SOCIAL SCIENCE RESEARCH ON RACE 1:1, 2004

Table 10. OLS regression analysis of the impact of race, prejudice, and crime-related beliefs

Murder Victim Race Bias Innocent on Death Row Crack vs. Powder Cocaine

Model 1 Model 2 Model 3 Model 1 Model 2 Model 3 Model 1 Model 2 Model 3

Constant 2+61*** 2+50*** 2+86*** 2+01*** 1+89*** 1+82*** 1+60*** 1+74*** 1+70***

~+14!~+27!~+37!~+15!~+29!~+40!~+14!~+27!~+36!

White +52*** +15*** 2+65 +63*** +27*** +08 +34*** +01 +31

~+06!~+04!~+47!~+06!~+07!~+50!~+06!~+07!~+48!

CJS biased against Blacks 2+20*** 2+21* 2+18** 2+01 2+29*** 2+39***

~+06!~+09!~+06!~+10!~+06!~+09!

Salience of crime index +02 +02 2+003 2+003 +11** +13**

~+04!~+04!~+04!~+06!~+04!~+05!

Racial resentment +28*** +19*** +28*** +16** +20*** +29***

~+04!~+06!~+04!~+06!~+04!~+06!

White * Bias 2+04 2+21 +13

~+12!~+13!~+12!

White * Salience +08 +05 2+06

~+08!~+08!~+08!

White * Racial resentment +17* +21** 2+17*

~+08!~+09!~+08!

Adjusted R Square +08 +13 +14 +12 +14 +17 +05 +08 +10

N 1414 1414 1414 1416 1416 1416 1495 1495 1495

Source: Race, Crime, and Public Opinion Study, 2001+

Note: Models include controls for education, age, gender, region, and political ideology+

*p,+05, ** p , +01, *** p , +001

DU BOIS REVIEW: SOCIAL SCIENCE RESEARCH ON RACE 1:1, 2004 169

sures of perceived criminal justice system bias and racial resentment significantly re-

duce the size of the Black-White difference in each model+ Indeed, introducing these

two measures cuts in more than half the size of the race dummy variable in the Murder

Victim Race and Innocent on Death Row experiments, and reduces the effect to in-

significance in the Crack vs+ Powder Cocaine experiments+

Third, we find significant interactions between respondent race and racial resent-

ment for each of the experiments+ The patterning is very straightforward for the Mur-

der Victim Race and Innocent on Death Row experiments: The impact of racial

resentment on support for the death penalty is about twice as large for Whites as it is for

Blacks+ The results are more complicated in the case of the Crack versus Power Cocaine

experiment where the sign of the interaction coefficient actually reverses direction+

Remarkably, racial resentment has a stronger effect on Blacks’ support for the crack

versus powder sentencing differential than it does among Whites+ Three observations

seem important to understanding this effect: ~1! The crack versus powder sentencing

case is the only occasion where we find a significant influence of the salience of crime+

~2! Omitting the interaction of respondent race and racial resentment, we do find a

significant interaction of perceived criminal justice bias and race ~b 52+30, p , +001,

modelnot shown!+ This suggests thatperceived racialbias substantiallyreduced Whites’

support for the sentencing differential, but had essentially no effect among Blacks+ ~3!

Careful examination of patterns among the Black respondents indicated that the effect

of racial resentment was largest among those Blacks aged sixty-six or older+ Taken to-

gether, we speculate that these patterns indicate that a segment of the Black population

that is older, probably more fearful of crime, and came of age prior to the civil rights

movement ~and thus more likely to come under the cultural sway of anti-black cultural

ideas! very much wants to control youth crime associated with the trade in drugs+

DISCUSSION AND CONCLUSION

Over the past two decades criminal justice policy in the U+S+ became sharply more

punitive+ Judicial discretion in sentencing has been curtailed and “mandatory mini-

mum sentences” introduced+ We’ve passed three-strikes laws+ We increasingly try ju-

veniles in adult court+ We pressed for more police on the streets and “zero-tolerance”

style policies are lauded as reducing crime+ The result of these and other changes has

been a rapid increase in the number of people incarcerated or under some form of

supervision by the criminal justice system, and in social expenditures on jails and pris-

ons+ These changes have taken a heavily disproportionate toll on Black communities,

especially low-income Black communities+ Among the most widely discussed and de-

bated aspects of this change in criminal justice policy regimes has been the reinstitu-

tion of the death penalty, perhaps the greatest symbolic instantiation of a shift from

rehabilitation to retribution, and the greater severity in punishment attached to the

use of crack cocaine as compared to powder cocaine, as an element of the nation’s “war

on drugs,” as well as broader efforts to stem the tide of urban street violence+

To varying degrees, many discussions of this transformation of policy regimes

make assumptions about the tenor, malleability, and sources of public opinion on

criminal justice policy+ Using survey-based experiments we sought to explicitly test

these ideas+ We believe our results support three broad conclusions+

First, on the subject of the basic malleability of public opinion on criminal justice

issues, at least with respect to the death penalty, our results largely point in the direc-

tion of the relative fixity of opinion+ In none of three death penalty related experiments

did our framing experiments shift White opinion by as much as ten percentage points+

In addition, there were no consistent indicators of particular types of respondents, ei-

Lawrence D. Bobo and Devon Johnson

170

DU BOIS REVIEW: SOCIAL SCIENCE RESEARCH ON RACE 1:1, 2004

ther on the basis of background characteristics such as level of education or social psy-

chological attributes such as political ideology, who were especially susceptible to move

on the basis of framing cues of racial inequality or racial bias+ Indeed, that even cueing

the risk of executing a wrongly convicted person does little to change opinion suggests

a special rootedness to this domain of criminal justice thinking+

Second, a racial bias frame has more effect in the domain of drug-use policy than

in regard to capital punishment+ This may be read to mean that people react very

differently to violent crimes than to those criminal acts more apt to be regarded as

self-inflicted or victimless+ Thus, the racial bias cue in the crack versus powder

cocaine experiment produced a full twenty percentage change among both White

and Black respondents+ If our interpretation is right, then these results suggest that

attempts to re-frame some specific “war on drugs” reforms as racially biased may

produce important changes in public opinion+

At least two issues require more careful examination in order to tease out the

effects of abhorrence for violent crime versus other policy features as influences on

public opinion+ For instance, it is important to better specify whether the involve-

ment with violence is what so sharply differentiates the magnitude of experimental

effects in capital punishment experiments from the crack versus powder cocaine

experiment+ It is entirely possible that homicide is viewed as such a serious crime that

no amount of effort to provide cues of racial inequity or racial bias in the application

of the law overrides a desire for punishment where the crime of murder is concerned+

Focus group discussions we held as part of our Race, Crime, and Public Opinion

project are strongly suggestive in this regard+ Two illustrations make the point+ Par-

ticipants in separate White and Black focus group discussions were almost uniform,

for example ~and in strong contrast to our original expectations!, in their expressed

willingness to see teenagers as young as fourteen and fifteen tried as adults if they were

accused of committing murder+ After the moderator of the White focus group had

gone around the room on this subject and heard everyone voice support for trying

juvenile killers as adults, one of the White male participants said: “Does everybody

realize that we think it is all right to try a child younger for murder than it is to try them

for shoplifting? It just struck me that that is what we have decided+” No one in the

group seriously re-examined their opinions in the wake of this pointed observation+

Furthermore, in the Black focus group one of the participants who’s own son had been

arrested for what she regarded as a trivial possession of marijuana charge and who be-

moaned the stultifying presence of police officers and metal detectors in her son’s school,

nonetheless took a very harsh position with regard to the death penalty even though

some other members of the group argued against her+ In the course of a discussion on

the death penalty the focus group moderator said: “Should we be using it more ag-

gressively?” Despite an immediate answer of “no” from one of the Black male partici-

pants, the woman whose son had the marijuana arrest and attended a barricaded school

said: “Yes we should! People have been sitting on death row for years+ It is time to

electrocute them+ Make room for the new ones! ++++I understand that who are we to

judge? But, hey, you do the crime, you do the time+ Forget the time+ You killed that

person @and# if it’s that severe, they should go+ And I would pull the switch+”

In addition, the crack versus powder experiment also focuses on a type of racial

bias that may be especially troubling to many respondents: that is, a racial bias built

into the design or formulation of a policy+ Differential outcomes by race in the effects

of a policy may occur in other circumstances without such a clear-cut prior basis to

have expected sharply differential outcomes for Blacks and Whites+

Third, the most consistent predictor of criminal justice policy attitudes is, in fact,

a form of racial prejudice+ While racial resentment does not ever explain a large share

A Taste for Punishment

DU BOIS REVIEW: SOCIAL SCIENCE RESEARCH ON RACE 1:1, 2004 171

of the variation in any of the attitudes we have measured, it is the most consistently

influential of the variables outside of race classification itself+ This pattern has at least

two implications+ It further buttresses the concern that some of the major elements of

public support for punitive criminal justice policies are heavily tinged with racial an-

imus and thus quite likely to be resistant to change based on suasion and information-

based appeals+ What is more, this pattern reinforces the claim by criminological conflict

theorists ~ Jacobs and Carmichael, 2002; Miller 1996; Parenti 1999; Turk 1969! that

one major function of the criminal justice system is the regulation and control of mar-

ginalized social groups such as African Americans+

Corresponding author

: Professor Lawrence C. Bobo, Acting Chair and Director, Department of

African and African American Studies, Barker Center 233, Harvard University, Cambridge, MA

02138. E-mail: [email protected]

NOTES

1+ Survey-based experiments combine the capacity to generalize to important populations

characteristic of sample surveys with the experimental control that allows stronger con-

clusions as to cause and effect ~see Schuman and Bobo, 1988, especially pp+ 273–274!+ The

use of experiments embedded within surveys has a long history in the social sciences,

though until the last decade or so such experiments were aimed mainly at methodological

research on, for example, question order and response format effects ~Schuman and

Presser, 1981! rather than substantive hypothesis testing+ Kinder and Palfrey ~1993!

provide an excellent general introduction to the use of experiments in studies of public

opinion+ This approach has been used to test theories of attitudes on affirmative action

and related race-targeted policies ~Bobo and Kluegel, 1993; Kinder and Sanders, 1996;

Sidanius et al+, 2000; Sniderman and Carmines, 1997!, social welfare policy ~Peffley et al+,

1997; Smith 1987!, residential composition preferences ~Charles 2000 and 2001; Emerson

et al+, 2001!, and political campaign messages ~Mendelberg 2001; Reeves 1997; Valentino

et al+, 2002!+

2+ Since 1976, thirty-eight states have reinstated capital punishment+ Only six of these states

have not held an execution+ Reinstating the death penalty is not a hollow gesture+ As Jacobs

and Carmichael explain: “From 1991 to 1995, only 2 of the 35 death-penalty states did not

impose this sentence+ The mean number of death sentences in these three five-year periods

in death penalty states was 30+7 ~after 1980 and the reversal of the Furman decision, this

mean rises to 39+4!+ If states allow the death sentence, their courts use it+ The legality of this

punishment is not just a symbolic issue” ~ Jacobs and Carmichael, 2002, p+ 116!+

3+ Distinguished jurist A+ Leon Higginbotham convincingly argued that a foundation stone

of American law has been the “precept of black inferiority+” He writes: “But the truth was

that our nation was founded explicitly, prospered implicitly, and still often lives uneasily

on the precept of black inferiority and white superiority+ Indeed, that precept helped to

legitimize slavery in America and served to justify the segregation of African Americans in

this nation long after slavery had been abolished+ To this day, the premise of black

inferiority and white superiority remains an essential element of the ‘American identity,’

mesmerized as we still are by race and color” ~Higginbotham 1996, p+ 9!+

4+ We employ this measure despite a long history of controversy surrounding its meaning

and that of the closely allied notion of “symbolic racism” ~see Sears et al+, 2000 for a

review!+ Here it is mainly important to emphasize that we understand the measure as

tapping a set of group-level or collective sentiments that have much in common with

theories of group conflict and group position ~Blumer 1958; Bobo 1999+! Theoretically,

Eliot Smith has persuasively argued that: “there is reason to identify symbolic racism with

group-based, emotion-linked appraisals” ~Smith 1993, p+ 308! and that such appraisals

may not flow from discrete individual exigencies or experiences but do very much flow

from “perceivers’ views of their group interests’ ~Smith 1993, p+ 309, emphasis in origi-

nal!+ Or, as Michael Hughes put it on the basis of a careful analysis of National Election

Study data: “In short, symbolic racism may, in itself, constitute self-oriented group inter-

est in the sense of status defense++++The critical issue, however, is also not moral resent-

Lawrence D. Bobo and Devon Johnson

172

DU BOIS REVIEW: SOCIAL SCIENCE RESEARCH ON RACE 1:1, 2004

ment and irrational antagonism: The issue represented by symbolic racism is status and

power and whites’ fear of losing them” ~Hughes 1997, p+ 74!+

5+ For each experiment, dependent variables are a one to four scale, with higher scores

indicating more punitive attitudes+

6+ White respondents to the NAAES were not asked the criminal justice racial bias question

or the common fate identity question+

7+ A 1990 report by the U+S+ General Accounting Office ~GAO! synthesized the results from

twenty-eight studies of death penalty sentencing since 1976+ The report concluded that

“In 82 percent of the studies, race of victim was found to influence the likelihood of being

charged with capital murder or receiving the death penalty, i+e+, those who murdered

whites were found to be more likely to be sentenced to death than those who murdered

blacks+ This finding was remarkably consistent across data sets, states, data collection

methods, and analytic techniques+ The finding held for high, medium and low quality

studies” ~U+S+ General Accounting Office, Death Penalty Sentencing: Research Indicates

Pattern of Racial Disparities, February 1990!+

8+ Federal legislation in 1986 and 1988 established higher penalties for the possession and

distribution of crack cocaine as compared to powder cocaine+ These provisions treat one

gram of crack as the equivalent of 100 grams of powder cocaine+ Thus, it takes only five

grams of crack, but 500 grams of powder cocaine to trigger a five-year mandatory mini-

mum prison sentence for possession with intent to distribute+ Similarly, it takes fifty grams

of crack compared to 5,000 grams of powder cocaine to trigger a ten-year mandatory

minimum sentence for trafficking ~United States Sentencing Commission, Special Report

to Congress: Cocaine and Federal Sentencing Policy, February 1995!+

REFERENCES

Adams, B+ D+ ~1978!+ Inferiorization and self-esteem+ Social Psychology Quarterly+ 41: 442–53+

Applegate, Brandon, et al+ ~1996!+ Assessing public support for three-strikes-and-you’re-out

laws: Global versus specific attitudes+ Crime and Delinquency+ 42~4!: 517–534+

Barkan, Steven E+ and Steven F+ Cohn ~1994!+ Racial prejudice and support for the death

penalty by whites+ Journal of Research in Crime and Delinquency, 31: 202–209+

Beckett, Katherine ~1997!+ Making Crime Pay: Law and Order in Contemporary American Politics+

New York: Oxford University Press+

Beckett, Katherine and Bruce Western ~2001!+ Governing social marginality: Welfare, incar-

ceration, and the transformation of state policy+ In Garland ~Ed+!, Mass Imprisonment: Social

Causes and Consequences+ Thousand Oaks, CA: Sage Publications+

Blumer, Herbert ~1958!+ Race prejudice as a sense of group position+ Pacific Sociological Review,

1: 3–7+

Blumstein, Alfred ~2001!+ Race and criminal justice+ In Smelser, Wilson, and Mitchell ~Eds+!,

America Becoming: Racial Trends and Their Consequences, Volume II, pp+ 21–31+ Washington,

DC: National Academy Press+

Blumstein, Alfred and Jacqueline Cohen ~1980!+ Sentencing of convicted offenders: An analy-

sis of the public’s view+ Law and Society Review+ 14: 223–261+

Bobo, Lawrence D+ ~1999!+ Prejudice as group position: Microfoundations of a sociological

approach to racism and race relations+ Journal of Social Issues, 55: 445–472+

Bobo, Lawrence D+ and James R+ Kluegel ~1993!+ Opposition to race-targeting: Self-interest,

stratification ideology, or racial attitudes? American Sociological Review, 58: 443–464+

Bohm, Robert M+ ~1991!+ American death penalty opinion, 1936–1986: A critical examination

of the Gallup polls+ In Bohm ~Ed+!, The Death Penalty in America: Current Research+ Cincin-

nati, OH: Anderson+

Bohm, Robert M+, Louise J+ Clark, and Adrian F+ Aveni ~1991!+ Knowledge and death penalty

opinion: A test of the Marshall hypotheses+ Journal of Research in Crime and Delinquency, 28:

360–387+

Bowers, William ~1993!+ Capital punishment and contemporary values: People’s misgivings

and the court’s misperceptions+ Law and Society Review, 27: 157–175+

Bureau of Justice Statistics ~2002!+ Correctional Populations in the United States 1998—Statistical

Tables+ ~Publication No+ NCJ 192929!+ Washington, D+C+

Chambliss, William J+ ~1999!+ Power, Politics, and Crime+ Boulder, CO: Westview+

Charles, Camille Z+ ~2000!+ Neighborhood racial-composition preferences: Evidence from a

multiethnic metropolis+ Social Problems, 47: 379–407+

Charles, Camille Z+ ~2001!+ Processes of racial residential segregation+ In O’Connor, Bobo,

A Taste for Punishment

DU BOIS REVIEW: SOCIAL SCIENCE RESEARCH ON RACE 1:1, 2004 173

and Tilly ~Eds+!, Urban Inequality: Evidence From Four Cities, pp+ 217–271+ New York:

Russell Sage Foundation+

Clark, Kenneth B+ ~1965!+ Dark Ghetto+ New York: Harper & Row+

Cohn, Steven F+ and Steven E+ Barkan ~ forthcoming!+ Racial prejudice and public attitudes

about the punishment of criminals+ In Browning ~Ed+!, For the Common Good: A Critical

Examination of Law and Social Control+ Durham, NC: Carolina Academic Press+

Cohn, Steven, Steven E+ Barkan, and William A+ Halteman ~1991!+ Punitive attitudes toward

criminals: Racial consensus or racial conflict? Social Problems, 38: 287–296+

Cole, David ~1999!+ No Equal Justice: Race and Class in the American Criminal Justice System+

New York: The New Press+

Combs, Michael W+ and John C+ Comer ~1984!+ Race and capital punishment: A longitudinal

analysis+ Phylon, 43: 350–359+

Cross, William E+ ~1991!+ Shades of Black: Diversity in African-American Identity+ Philadelphia,

PA: Temple University Press+

Cullen, Francis T+, Bonnie S+ Fisher, and Brandon K+ Applegate ~2000!+ Public opinion about

punishment and corrections+ Crime and Justice: A Review of Research, 27: 1–79+

Delli Carpini, Michael X+ and Scott Keeter ~1997!+ What Americans Know About Politics and

Why it Matters+ New Haven, CT: Yale University Press+

Du Bois, W+ E+ B+ ~1996!@1899#+ The Philadelphia Negro: A Social Study+ Philadelphia, PA:

University of Pennsylvania Press+

Ellsworth, Phoebe C+ and Samuel R+ Gross ~1994!+ Hardening of the attitudes: Americans’

views on the death penalty+ Journal of Social Issues, 50: 19–52+

Ellsworth, Phoebe C+ and Lee Ross ~1983!+ Public opinion and capital punishment: A close ex-

amination of the views of abolitionists and retentionists+ Crime and Delinquency, 29: 116–169+

Emerson, Michael O+, George Yancey, and Karen J+ Chai ~2001!+ Does race matter in residen-

tial segregation?: Exploring the preferences of white Americans+ American Sociological Review,

66: 922–935+

Fredrickson, George M+ ~1971!+ The Black Image in the White Mind: The Debate on Afro-

American Character and Destiny, 1817–1914+ New York: Harper and Row+

Garland, David ~2001!+ Mass Imprisonment: Social Causes and Consequences+ Thousand Oaks,

CA: Sage Publications+

Gerber, Jurg and Simone Engelhardt-Greer ~1996!+ Just and painful: Attitudes toward sen-

tencing criminals+ In Flanagan and Longmire ~Eds+!, Americans View Crime and Justice: A

National Opinion Survey+ Thousand Oaks, CA: Sage Publications+

Grier, William H+ and Price M+ Cobbs ~1968!+ Black Rage+ New York: Basic Books+

Hagan, John and C+ Albonetti ~1982!+ Race, class, and the perception of criminal injustice in

America+ American Journal of Sociology, 88: 329–355+

Hawkins, Darnell F+ ~2001!+ Commentary on Randall Kennedy’s overview of the justice sys-

tem+ In Smelser, Wilson, and Mitchell ~Eds+!, America Becoming: Racial Trends and Their

Consequences, Volume II, pp+ 32–51+ Washington, DC: National Academy Press+

Henderson, Martha L+ et al+ ~1997!+ The impact of race on perceptions of criminal injustice+

Journal of Criminal Justice, 25: 447–462+

Higginbotham, A+ Leon+ ~1996!+ Shades of Freedom: Racial Politics and Presumptions of the

American Legal Process+ New York: Oxford University Press+

Huang, W+ and Vaughn, M+ ~1996!+ Support and confidence: Public attitudes toward the

police+ In Flanagan and Longmire ~Eds+!, Americans View Crime and Justice: A National

Survey of Public Opinion+ Thousand Oaks, CA: Sage+

Hughes, Michael ~1997!+ Symbolic racism, old-fashioned racism, and whites’ opposition to

affirmative action+ In Tuch and Martin ~Eds+!, Racial Attitudes in the 1990s: Continuity and

Change, pp+ 45–75+ Westport, CT+: Praeger+

Hurwitz, John and Mark Peffley ~1997!+ Public perceptions of race and crime: The role of

racial stereotypes+ American Journal of Political Science, 41: 375–401+

Hurwitz, Jon and Mark Peffley ~2001!+ Racial polarization on criminal justice issues: Sources

and political consequences of fairness judgments+ Paper presented at the meeting of the Amer-

ican Political Science Association in San Francisco, CA, August 30 to September 2, 2001+

Jackman, Mary R+ ~1994!+ The Velvet Glove: Paternalism and Conflict in Gender, Class and Race

Relations+ Berkeley, CA: University of California Press+

Jacobs, David and Jason Carmichael ~2002!+ The political sociology of the death penalty: A

pooled time-series analysis+ American Sociological Review, 67:109–131+

Johnson, Devon ~2001!+ Punitive attitudes on crime: Economic insecurity, racial prejudice, or

both? Sociological Focus, 24: 33–54+

Lawrence D. Bobo and Devon Johnson

174

DU BOIS REVIEW: SOCIAL SCIENCE RESEARCH ON RACE 1:1, 2004

Jones, Jeffrey M+ ~2001!+ Two-thirds of Americans support the death penalty+ Gallup Poll

Release March 2, 2001+

Kennedy, Randall ~2001!+ Racial trends in the administration of criminal justice+ In Smelser,

Wilson, and Mitchell ~Eds+!, America Becoming: Racial Trends and Their Consequences, Volume

II+ Washington, DC: National Academy Press+

Kinder, Donald R+ ~1998!+ Opinion and action in the realm of politics+ In Gilbert, Fiske, and

Lindzey ~Eds+!, The Handbook of Social Psychology ~4

Ed+!, pp+ 778–867+ London: Oxford

University Press+

Kinder, Donald R+ and Thomas Palfrey ~1993!+ On behalf of an experimental political science+

In Palfrey ~Ed+!, Experimental Foundations of Political Science+ Ann Arbor, MI: University of

Michigan Press+

Kinder, Donald R+ and Lynn M+ Sanders ~1996!+ Divided by Color: Racial Politics and Democratic

Ideals+ Chicago: University of Chicago Press+

Lock, Shmuel ~1999!+ Crime, Public Opinion, and Civil Liberties: The Tolerant Public+ Westport,

CT: Praeger+

McCorkle, Richard C+ ~1993!+ Research note: Punish and rehabilitate? Public attitudes toward

six common crimes+ Crime and Delinquency, 39: 240–252+

McGarrell, Edmund F+ and Marla Sandys ~1996!+ The misperception of public opinion toward

capital punishment: Examining the spuriousness explanation of death penalty support+

American Behavioral Scientist, 39: 500–513+

Meares, Tracey and Dan Kahan ~1998!+ Laws and ~norms of! order in the inner city+ Law and

Society Review, 32: 805+

Mendelberg, Tali ~2001!+ The Race Card: Campaign Strategy, Implicit Messages, and the Norm of

Equality+ Princeton, NJ: Princeton University Press+

Miller, Jerome G+ ~1996!+ Search and Destroy: African-American Males in the Criminal Justice

System+ New York: Cambridge University Press+

Miller, J+L+, Peter H+ Rossi, and Jon E+ Simpson ~1986!+ Perceptions of justice: Race and gender

differences in judgments of appropriate prison sentences+ Law and Society Review, 20: 313–334+

Moon, Melissa M+ et al+ ~2000!+ Putting kids to death: Specifying public support for juvenile

capital punishment+ Justice Quarterly, 17: 663–684+

Parenti, Christian ~1999!+ Lockdown America: Police and Prison in the Age of Crisis+ New York:

Verso+

Peek, Charles W+, George D+ Lowe, and Jon P+ Alston+ ~1981!+ Race and attitudes toward local

police: Another look+ Journal of Black Studies, 11: 361–374+

Peffley, Mark, John Hurwitz, and Paul M+ Sniderman ~1997!+ Racial stereotypes and whites’

political views of blacks in the context of welfare and crime+ American Journal of Political

Science, 41: 30–60+

Reeves, Keith ~1997!+ Voting Hopes or Fears? New York: Oxford University Press+

Roberts, Julian V+ and Loretta J+ Stalans ~2000!+ Public Opinion, Crime, and Criminal Justice+

Boulder, CO: Westview+

Sachdev, Itesh and R+Y+ Bourhis ~1991!+ Power and status differentials in minority and major-

ity group relations+ European Journal of Social Psychology, 21: 1–24+

Sandys, Marla and Edmund F+ McGarrell ~1995!+ Attitudes toward capital punishment: pref-

erence for the penalty or mere acceptance? Journal of Research in Crime and Delinquency, 32:

191–213+

Sarat, A+ and N+ Vidmar ~1976!+ Public opinion, the death penalty, and the eighth amendment:

Testing the Marshall hypothesis+ Wisconsin Law Review, 171–206+

Schuman, Howard and Lawrence Bobo ~1988!+ Survey-based experiments on white racial

attitudes toward residential segregation+ American Journal of Sociology, 94: 273–299+

Schuman, Howard and Stanley Presser ~1981!+ Questions and Answers in Attitude Surveys+ New

York: Academic Press+

Schuman, Howard, Charlotte Steeh, Lawrence Bobo, and Maria Krysan ~1997!+ Racial Attitudes

in America: Trends and Interpretations ~Revised Ed+!+ Cambridge, MA: Harvard University Press+

Sears, David O+, Jim Sidanius, and Lawrence Bobo ~2000!+ Racialized Politics: The Debate About

Racism in America+ Chicago: University of Chicago Press+

Secret, Philip E+ and James S+ Johnson ~1989!+ Racial differences in attitudes toward crime

control+ Journal of Criminal Justice, 17: 361–375+

Sidanius, Jim and Felicia Pratto ~1999!+ Social Dominance+ New York: Cambridge University

Press+

Sidanius, Jim, Pam Singh, John J+ Hetts, and Chris Federico ~2000!+ It’s not affirmative action,

it’s the blacks: The continuing relevance of race in American politics+ In Sears, Sidanius, and

A Taste for Punishment

DU BOIS REVIEW: SOCIAL SCIENCE RESEARCH ON RACE 1:1, 2004 175

Bobo ~Eds+!, Racialized Politics: The Debate About Racism in America+ Chicago: University of

Chicago Press+

Smith, Eliot R+ ~1993!+ Social identity and social emotions: Toward new conceptualizations of

prejudice+ In Mackie and Hamilton ~Eds+!, Affect, Cognition, and Stereotyping: Interactive in

Group Perception, pp+ 295–315+ New York: Academic Press+

Smith, To m W + ~ 1987!+ That which we call welfare by any other name would smell sweeter: An

analysis of the impact of question wording on response patterns+ Public Opinion Quarterly,

51: 75–83+

Smith, Paul E+ and Richard O+ Hawkins ~1973!+ Victimization, types of citizen–police con-

tacts, and attitudes toward the police+ Law and Society Review, 8: 135–152+

Sniderman, Paul M+ and Edward Carmines ~1997!+ Reaching Beyond Race+ Cambridge, MA:

Harvard University Press+

Tajfel, Henri ~1981!+ Human Groups and Social Categories+ Cambridge: Cambridge University

Press+

Taylor, Terrance J+, K+B+ Turner, Finn-Aage Esbensen, and L+ Thomas Winfree, Jr+ ~2001!+

Coppin’ an attitude: Attitudinal differences among juveniles toward police+ Journal of Crim-

inal Justice, 29: 295–305+

Thernstrom, Stephan and Abigail Thernstrom ~1997!+ America in Black and White+ New York:

Simon and Schuster+

Tonry, Michael ~1995!+ Malign Neglect: Race, Crime and Punishment in America+ New York:

Oxford University Press+

Tuch, Steven A+ and Ronald Weitzer ~1997!+ Trends: Racial differences in attitudes toward the

police+ Public Opinion Quarterly, 61: 642–663+

Turk, Austin ~1969!+ Criminality and Legal Order+ Chicago: Rand McNally+

Tyler, To m R + and Robert J+ Boeckmann ~1997!+ Three strikes and you are out, but why?: The

psychology of public support for punishing rule breakers+ Law and Society Review, 31: 237–207+

Tyler, To m R + and Renee Weber ~1982!+ Support for the death penalty: Instrumental response

to crime, or symbolic attitude? Law and Society Review, 17: 21–45+

Valentino, Nicholas A+, Vincent L+ Hutchings, and Ismail K+ White ~2002!+ Cues that matter:

How political ads prime racial attitudes during campaigns+ American Political Science Review,

96: 75–90+

Wacquant, Loic ~2001!+ Deadly symbiosis: When ghetto and prison meet and mesh+ Punish-

ment and Society, 3: 95–134+

Wortley, Scot, John Hagan, and Ross Macmillan ~1997!+ Just des~s!erts? The racial polariza-

tion of perceptions of criminal injustice+ Law and Society Review, 31: 637–676+

Young, Robert L+ ~1991!+ Race, conceptions of crime and justice, and support for the death

penalty+ Social Psychology Quarterly, 54: 67–75+

Zaller, John R+ ~1992!+ The Nature and Origin of Mass Opinion+ Cambridge: University of

Cambridge Press+

Zamble, Edward and Kerry Lee Kalm ~1990!+ General and specific measures of public atti-

tudes toward sentencing+ Canadian Journal of Behavioral Science, 22: 327–337+

APPENDIX

Knowledge Networks’ Methodology

Knowledge Networks has created an innovative research design that allows survey

questionnaires to be administered over the Internet to a representative sample of

U+S+ households+ Knowledge Networks employs a Random Digit Dialing ~RDD!

telephone methodology using a complex sample stratification design that takes into

account the known probabilities of selection associated with geographical location,

the number of phone lines and people in each household, and whether or not the

phone number is listed, to provide a representative and valid sample of the U+S+

population+ Once a Knowledge Networks household is selected, members are con-

tacted first by an express delivery mailing and then by telephone for enrollment in

the Knowledge Networks panel+ Every participating Knowledge Networks house-

hold receives free hardware, free Internet access, free e-mail accounts and ongoing

technical support+ Participants receive a short multimedia survey about once a week+

Lawrence D. Bobo and Devon Johnson

176

DU BOIS REVIEW: SOCIAL SCIENCE RESEARCH ON RACE 1:1, 2004

Surveys are delivered by e-mail on the same standardized hardware, through a

television set ~Knowledge Networks Methodology Report!+

Knowledge Networks maintains a panel of approximately 40,000 members at

any given time+ The average Knowledge Network respondent remains a panel mem-

ber for three years, and generally receives three surveys a month+ Approximately 20

percent of these surveys are fielded for academic or government research, another

20–30 percent are internal surveys ~to build up profile databases, etc+!, and the rest

are commercially sponsored surveys+ Panel members are not assigned a survey on the

same topic in a three-month period+ Due to the mode of delivery, respondents

answer questions at their convenience; some respondents complete a survey in one

sitting while others answer questions over a matter of hours or days+ For this reason,

calculating the average duration of surveys is not meaningful+

The benefits of this new Internet-based survey design include the ability to use

multimedia presentations, which are often more engaging than a traditional tele-

phone interview+ Moreover, the use of standardized hardware ensures a consistency

across interviews; and the lack of an interviewer may reduce response bias that is

often associated with social desirability pressures in a traditional interview setting+

The Knowledge Networks Panel closely resembles the demographic character-

istics of the U+S+ population+ Tables A1 and A2 show that the demographic charac-

Table A1. Comparison of Knowledge Networks’ Panel as of January 2001 with

population counts from the 1998 Current Population Survey for the top twenty-five

metropolitan areas

U.S. Census

1998 Estimate of

Total Population

Knowledge

Networks Panel,

Adults

New York-Northern New Jersey-Long Island 7+45% 6+75%

Los Angeles-Riverside-Orange County 5+84 6+01

Chicago-Gary-Kenosha 6+26 3+81

Washington-Baltimore 2+70 2+92

San Francisco-Oakland-San Jose 5+52 2+92

Philadelphia-Wilmington-Atlantic City 2+22 2+15

Boston-Worcester-Lawrence 2+08 2+31

Detroit-Ann Arbor-Flint 2+02 1+82

Dallas-Fort Worth 1+78 2+30

Houston-Galveston-Brazoria 1+63 1+70

Atlanta 1+39 1+58

Miami-Fort Lauderdale 1+35 1+53

Seattle-Tacoma-Bremerton 1+27 1+29

Phoenix-Mesa 1+08 1+17

Cleveland-Akron 1+08 1+10

Minneapolis-St+ Paul 1+05 1+76

San Diego 1+03 1+16

St+ Louis 0+95 1+06

Denver-Boulder-Greeley 0+88 1+14

Pittsburgh 0+87 0+48

Tampa-St+ Petersburg-Clearwater 0+83 1+24

Portland-Salem 0+80 1+04

Cincinnati-Hamilton 0+72 0+82

Kansas City 0+64 0+63

Sacramento-Yolo 0+62 0+98

Source: Knowledge Networks Panel Demographic Characteristics, January 2001+

A Taste for Punishment

DU BOIS REVIEW: SOCIAL SCIENCE RESEARCH ON RACE 1:1, 2004 177

Table A2. Comparison of the demographics characteristics of the adult U+S+ population,

the Knowledge Networks Panel, and the 2001 Race, Crime, and Public Opinion Study

Adult U.S.

Population

Estimates,

CPS 1998

Knowledge

Networks Panel,

January 2001

2001 Race,

Crime,

and Public

Opinion Study

(raw data)

2001 Race,

Crime,

and Public

Opinion Study

(weighted)

Gender

Male 48+9% 49+3% 45+246+5

Female 51+150+754+853+5

Age

18–24 13+112+57+010+4

25–34 18+622+320+523+2

35–44 22+126+123+121+3

45–54 18+220+520+818+6

55–64 11+710+014+213+8

651 16+38+614+512+7

Race

White 83+379+849+249+2

Black0African American 11+910+850+850+8

American Indian0Alaska native 0+82+1——

Asian0Pacific Islander 3+93+1——

Other — 4+3——

Hispanic Ethnicity

Hispanic 10+59+1——

Non-Hispanic 89+590+9——

Employment Status

In the labor force 65+770+371+971+2

Working full time 54+961+757+356+2

Working part time 10+78+614+615+0

Not in the labor force 34+329+727+028+9

Marital Status

Married 57+861+754+153+3

Not married 42+238+345+946+7

Housing Ownership

Own 70+973+071+069+7

Rent0other 29+124+229+030+3

Level of Education

Less than HS diploma 16+89+512+416+7

HS diploma or equivalent 32+923+026+233+1

Some college 19+731+230+325+0

Associate Degree 7+26+35+64+0

Bachelor’s Degree or beyond 23+331+025+621+1

Household Income

Under $10,000 7+83+57+39+2

$10,000–24,999 19+812+117+118+6

$25,000–49,999 30+532+834+634+2

$50,000–74,999 19+526+422+220+7

$75,000 or more 22+225+118+817+3

Census Region

Northeast 19+017+518+618+9

Midwest 23+226+926+922+6

South 35+432+939+244+2

West 22+422+715+314+3

Source: Knowledge Networks Panel Demographic Characteristics, January 2001 and Race, Crime, and

Public Opinion Study, 2001+

Lawrence D. Bobo and Devon Johnson

178

DU BOIS REVIEW: SOCIAL SCIENCE RESEARCH ON RACE 1:1, 2004

teristics of the Knowledge Networks Panel mirror current population data from the

U+S+ Census+ Table A1 presents a comparison of the Knowledge Networks Panel

with Current Population Survey estimates of the population in the top twenty-five

metropolitan areas+ A comparison of the demographic characteristics of the Knowl-

edge Networks Panel and estimates from the U+S+ Census’ Current Population

Survey of the U+S+ adult population is shown in Table A2+ The Knowledge Networks

Panel slightly under-represents those with incomes under $25,000 and less than a

high school diploma, while slightly over-representing those whose incomes exceed

$75,000 and the highly educated+ Census methodology for the Current Population

Survey allocates survey participants into one of four racial categories+ In contrast, the

Knowledge Networks method retains an “other” category for uncertain situations+

To reiterate, in all other demographic categories, the characteristics of the panel

members are very similar to the national population+

2000 National African American Election Study

The 2000 NAAES ~National African American Election Study! was collected in two

waves, including a pre-election survey of 831 African Americans ~conducted between

October 18

and 29

! and a post-election survey of 605 Blacks and 724 Whites

~conducted between December 1

and 15

!+ The pre-election survey of African

Americans had a within-panel response rate of 56%+ The post-election survey had a

within-panel response rate of 78% for Whites, and a re-interview rate of 75% for

blacks+ Under the most stringent assumptions ~taking into account the panel recruit-

ment response rate, the household connection rate and the within-panel completion

rate! the overall response rate was 21% for Blacks and 31% for Whites+ Lawrence D+

Bobo and Michael C+ Dawson were the co-principal investigators for the NAAES+

2001 Race, Crime, and Public Opinion Study

The Race, Crime and Public Opinion Study ~RCPOS! was conducted by the

authors in the summer of 2001+ Due to the length of the questionnaire, the data

were collected in two waves+ The first wave of the survey was conducted

between June 4

and June 20

, with a within-panel response rate of 72% for

African Americans and 61% for Whites+ These respondents were contacted for the

second wave of the survey between August 7

and August 20

+ The second wave

included 725 Blacks and 783 Whites, with a re-interview response rate of 72% and

80%, respectively+ Under the most stringent assumptions ~taking into account the

panel recruitment response rate, the household connection rate and the within-

panel completion rate!, the overall response rate was 22% for Whites and 25% for

Blacks+

As Table A2 shows, the demographic characteristics of the weighted sample

~Column 4! generally match those of the U+S+ population as a whole ~Column 1!+ A

demographic breakdown of the White and Black respondents by wave is shown in

Table A3+ Overall, the demographic characteristics of the respondents who com-

pleted Wave 2 closely match those of the respondents who completed Wave 1,

indicating that there was no systematic bias in the cases that were lost to attrition+

Ballot Randomization

We performed a series of tests for effective experimental ballot randomization+ We

compared distributions across race, gender, age, education, household income, region,

A Taste for Punishment

DU BOIS REVIEW: SOCIAL SCIENCE RESEARCH ON RACE 1:1, 2004 179

and political ideology+ In all, we performed twenty-three separate tests, twenty-one

of which yielded no statistically significant differences in respondent characteristics

across experimental ballot+

Table A3. Characteristics of the 2001 Race, Crime, and Public Opinion Study

Whites

Wave 1

~N 5 978!

Whites

Wave 2

~N 5 783!

Whites lost

to attrition

~N 5 195!

Blacks

Wave 1

~N 5 1010!

Blacks

Wave 2

~N 5 725!

Blacks lost

to attrition

~N 5 285!

Gender

Male 50% 52% 44% 41% 41% 39%

Female 50 48 56 59 59 61

N ~978!~783!~195!~1002!~725!~277!

Age

18–24 6 6 9 8 7 10

25–34 15 15 17 26 24 30

35–44 19 19 19 27 28 26

45–54 17 17 17 24 25 22

55–64 19 18 24 10 10 7

651 23 26 14 6 6 4

N ~978!~783!~195!~1001!~725!~276!

Education

Less than HS diploma 8 7 9 17 18 15

HS diploma 36 36 37 17 18 14

Some college0AA 28 27 29 44 42 48

BA or more 29 29 26 23 22 23

N ~976!~781!~195!~1001!~724!~227!

Household Income

Less than $24,999 19 20 15 30 30 30

$25,000–39,999 21 22 19 22 23 19

$40,000–59,999 26 25 28 23 21 28

$60,000 or more 34 33 38 25 26 23

N ~978!~783!~195!~1010!~725!~285!

Employment Status

Working FT0PT 61 60 66 75 75 76

Unemployed 3 3 2 5 4 7

Retired0Homemaker 33 34 28 14 15 10

Student 3 3 4 7 6 9

N ~971!~777!~194!~998!~722!~276!

Homeownership

Rent0other 19 19 19 40 39 44

Homeowner 81 81 81 60 61 56

N ~804!~643!~161!~900!~659!~241!

Urban/Rural

Non-metro 18 17 20 4 4 3

Metro 82 83 80 96 96 97

N ~978!~783!~195!~1002!~725!~277!

Region of Residence

Non-south 64 64 63 58 55 64

South 36 36 37 42 45 36

N ~978!~783!~195!~1002!~725!~277!

Political Ideology

Liberal 18 18 15 28 28 28

Moderate0DK 55 54 57 62 62 60

Conservative 28 28 28 11 10 11

N ~959!~768!~191!~805!~575!~230!

Source: Race, Crime, and Public Opinion Study, 2001+

Lawrence D. Bobo and Devon Johnson

180

DU BOIS REVIEW: SOCIAL SCIENCE RESEARCH ON RACE 1:1, 2004