CONFIDENCE IN THE DEATH PENAI,TY AND SUPPORT FOR ITS USE: EXPLORING THE VALUE-EXPRESSIVE DIMENSION OF DEATH PENAI,TY ATTITUDES*

Transcription

1 CONFIDENCE IN THE DEATH PENAI,TY AND SUPPORT FOR ITS USE: EXPLORING THE VALUE-EXPRESSIVE DIMENSION OF DEATH PENAI,TY ATTITUDES* SCOTT VOLLUM** DENNIS 1~ LONGMIRE*** JACQUELINE BUFFINGTON-VOLLUM... Sam Houston State University Recent media and political attention has raised public awareness of a number of issues surrounding the death penalty. Questions regarding innocence, fair trials, and equitable access to counsel and the appellate process are ubiquitous in coverage of the death penalty. Adequate information about public attitudes toward the death penalty in light of these issues is currently lacking. In 2002, as part of the annum Texas Crime Poll, questions were asked about confidence in the administration of the death penalty, support for the death penalty, and support for a moratorium. The results indicate that, although a majority of respondents support the death penalty, a substantial proportion lack confidence in its use and support a moratorium on executions. Of those lacking confidence and those supporting a moratorium, strong majorities maintain support for the death penalty (68% and 73%, respectively). These findings suggest that death penalty attitudes may be largely value expressive. * The authors would like to acknowledge Jiletta Kubena and Robert Morris for their invaluable assistance in this research. We are also grateful for the insightful and constructive comments and suggestions of the anonymous reviewers at Justice Quarterly. Finally, we'd like to thank Justice Quarterly's diligent editor, Donna Bishop, for her patience and persevering assistance in the final preparation of this article. An earlier version of this article was presented at the annual meetings of the American Society of Criminology in Chicago, November Address all correspondence to Scott Vollum, College of Criminal Justice, Sam Houston State University, Huntsville, TX ; svetlum@cox.net. ** Scott Vollum is a doctoral student in criminal justice at Sam Houston State University and a part-time instructor at Old Dominion University. His research and teaching interests include animal cruelty, criminological theory, capital punishment, research methods, and institutional violence. He is currently researching prison violence as well as continuing research on death penalty attitudes.... Dennis Longmire is a professor of criminal justice at Sam Houston State University who is internationally recognized for his work in the death penalty abolition movement. He received his M.A. and Ph.D. in criminal justice from the University of Maryland. He also consults and testifies as an expert witness in capital cases in Texas.... Jacqueline Buffington-Vollum is a doctoral student in forensic clinical psychology at Sam Houston State University and a clinical intern at Eastern Virginia Medical School. Her clinical and research interests include risk assessment and the assessment of psychopathy, the psychology of institutional confinement, and the role of psychological expertise in capital cases. JUSTICE QUARTERLY, Volume 21 No. 3, September Academy of Criminal Justice Sciences

2 522 CONFIDENCE IN THE DEATH PENALTY Recent media and political attention has raised public awareness of troubling questions surrounding the death penalty. Questions regarding innocence, fair trials, and equitable access to legal representation and the appellate process are ubiquitous in the news media's reporting on death penalty cases. Particular attention has recently focused on evidence that innocent individuals have been executed or are awaiting execution (Liebman, Fagan, & West, 2000). In Illinois, such issues have led to a moratorium on executions and mass clemency hearings for those on death row. Moreover, the U.S. Supreme Court recently examined the constitutionality of the death penalty in several cases. Issues examined include execution of the mentally retarded (Atkins v. Virginia, 2002), the sentencing process (Kelly v. South Carolina, 2002; Ring v. Arizona, 2002), racial discrimination in jury selection (Miller-El v. CockreU, 2003) and adequacy of legal representation (Bell v. Cone, 2002; Mickens v. Taylor, 2002; Wiggins v. Smith, 2003). During both the 2001 and 2003 legislative sessions in Texas-- a state that is notorious as the staunchest defender of capital punishmentmseveral bills proposing a moratorium on executions were introduced. Although none passed, a bill that would have imposed a moratorium was thoroughly considered and received considerable legislative support during both sessions, eventually making it out of committee and being placed on the intent calendar of the senate during the 2001 session. A vote on the issue, however, was never taken. These events potentially signal a changing tide in policy makers' perceptions of the death penalty. Whether, and to what degree, issues regarding the death penalty have entered the public consciousness in Texas is not yet completely clear. The problems inherent in the current death penalty system in Texas have been well documented, and it has been strongly suggested that the system is in urgent need of repair (Texas Defender Service, 2000, 2003). Failure to provide competent legal counsel, racial bias in jury selection, use of faulty scientific testimony about DNA and future dangerousness, and intentional misconduct by both prosecuting attorneys and witnesses for the prosecution are some of the most troubling issues recently documented (Texas Defender Service, 2000, 2003). Has the public lost confidence in the system as a result of these revelations? And have these revelations eroded public support for the death penalty? These questions are addressed here. Historically the majority of Texans have reported support for the death penalty (Longmire & Vollum, 2001; Longmire, multiple years). However, support may be contingent on a number of

3 VOLLUM, LONGMIRE, AND BUFFINGTON-VOLLUM 523 factors. Commonly referred to as the "Marshall hypotheses" (because of their basis in Supreme Court Justice Thurgood Marshall's concurring opinion in Furman v. Georgia) 1 it has been argued that knowledge of how the death penalty system actually operates reduces support for its use (see Bohm, Clark, & Aveni, 1991; Longmire, 1996; Sarat & Vidmar, 1976). In a national study, Longmire (1996) found such a "contingency" effect with respect to knowledge of several factors, including disproportionate sentencing based on race and financial status, the likelihood of innocent people receiving the death penalty, availability of a "life without parole" sentencing option, and mental retardation of offenders. In each instance, support for the death penalty decreased among respondents who were made aware of these features (Longmire, 1996). Beginning with the landmark decision in Furman v. Georgia (1972), public opinion has played a central role in death penalty jurisprudence and policy (Bohm et al., 1991; Bowers, 1993; Ellsworth & Ross, 1983; Jones, 1994; Sandys & McGarrell, 1994; Smith & Wright, 1992; Whitehead, Btankenship, & Wright, 1999). In his opinion in Furman, Justice Brennan stated: "A severe punishment must not be unacceptable to contemporary society. Rejection by society, of course, is a strong indication that a severe punishment does not comport with human dignity" (Furman v. Georgia, 1972, p. 277). Justice Marshall agreed, asserting: "Where a punishment is not excessive and serves a valid legislative purpose, it still may be invalid if popular sentiment abhors it" (Furman v. Georgia, 1972, p. 332). In Weems v. United States (1910), the Supreme Court suggested that the meaning of "cruel and unusual punishment" evolves "as public opinion becomes enlightened by a humane justice" (p. 378). Public opinion has historically served and continues to serve as a marker of "the evolving standards of decency that mark the progress of a maturing society" (Trop v. Dulles, 1958, p. 109). 2 It is no surprise, then, that since Furman public support has consistently been relied on as an indicator of the legitimacy of capital punishment in the United States. In spite of the apparent importance of public opinion in Eighth Amendment jurisprudence, the Supreme Court, since Furman, has generally rejected social scientific measures of public opinion as benchmarks of public sentiment toward the death penalty or 1 Specifically, Marshall hypothesized that full information on the realities of the death penalty "would almost surely convince the average citizen that the death penalty was unwise" (Furman v. Georgia, 1972, p. 363). 2 See, for example, the recent Supreme Court decision in Atkins v. Virginia (2002).

4 524 CONFIDENCE IN THE DEATH PENALTY general penal policy. 3 The Court has opted instead to rely heavily on the presence or absence of legislation regarding the death penalty to gauge public sentiment. For example, the Court has refused to consider public opinion poll data as a benchmark in determining whether the execution of juvenile (Stanford v. Kentucky, 1989) and mentally retarded (Penry v. Lynaugh, 1989) offenders constitutes cruel and unusual punishment. Dismissing public opinion poll data suggesting otherwise, the Court reasoned that the lack of statutes prohibiting the execution of such offenders evidenced a consensus that the public did not view the death penalw for such offenders as cruel and unusual punishment. However, in the recent Atkins decision (Atkins v. Virginia, 2002) public opinion polls entered the Supreme Court's jurisprudence. Although the decision did not turn on this evidence, the dissenting opinion of Chief Justice Rehnquist revolved around a thorough review of public opinion polls. The source of the public consensus on which the majority based its decision, however, was state legislation. Regardless of the source, perceptions about public opinion continue to be a powerful force in both jurisprudence and general policy regarding the death penalty. In fact, some argue that it is these perceptions of public opinion that are most salient to judicial and public policy decisions (Lifton & Mitchell, 2000), and public opinion polls play no small role in shaping these perceptions. PRIOR RESEARCH Numerous studies have examined public support for the death penalty. Following record lows in the 1960s, public support for the death penalty increased steadily in the post-furman decades (Ellsworth & Gross, 1994; Gross, 1998; Longmire, 1996; Smith & Wright, 1992) reaching an all-time high in the late 1980s and early 1990s (Bohm, et al., 1991; Ellsworth & Gross, 1994). Consistently, in the post-furman era a majority of the public continued to support, and thus lent credence to, use of the death penalty as a legitimate penal sanction. Generally, in the 1980s and early 1990s, more than 70% of people surveyed indicated support when asked whether they favor or oppose the death penalty for the crime of murder (Ellsworth & Gross, 1994; Smith & Wright, 1992). In Texas, the leading state in the use of capital punishment, death Although the Court has manifestly refused to use public opinion polls as bases for their death penalty decisions, the latent role such polls play is another matter. There is little doubt that Supreme Court justices remain aware of such opinion polls and that the results impact their death penalty decisions in many cases (see Bowers, 1993 for a discussion).

5 VOLLUM, LONGMIRE, AND BUFFINGTON-VOLLUM 525 penalty support has remained robust, with over 80% of citizens endorsing it in the late 1990s and early 2000s (Longmire & Vollum, 2001; Longmire & Vollum, 2003). However, single-item measures of public attitudes, such as those resulting in the figures cited above, may be misleading (Durham, Elrod, & Kinkade, 1996; Harris, 1986; Jones, 1994; Longn~re, 1996). Longmire (1996), for example, suggests: "Support for the death penalty discerned through the use of single-item questions may overstate the actual levels of public support" (p. 100). However, research conducted by Durham et al. (1996) indicates that, depending on the nature and circumstances of the crime, support for the death penalty may actually be higher than that indicated by a global, single-item measure. In their research, respondents were presented with case vignettes describing homicide scenarios and then asked to indicate what they believed to be an appropriate punishment. Interestingly, as many as 93.2% of respondents supported the death penalty in response to one of the scenarios (Durham at al., 1996, p. 720). This suggests that a high proportion of individuals consider the death penalty a legitimate punishment in some cases. Clearly, an accurate representation of support for the death penalty cannot be obtained by simply calculating the proportion of citizens responding "yes" to the question, "Do you support the death penalty?" Numerous studies have determined that support for the death penalty varies based on personal and demographic characteristics 4 (Barkan & Cohn, 1994; Bohm, 1991; Borg, 1998; Cochran, Boots, & Heide, 2003; Curtis, 1991; Durham et al., 1996; Gross, 1998; Harris, 1986; Keil & Vito, 1991; Longmire, 1996; Mignon & Holmes, 1999; Whitehead et al., 1999), characteristics of the crime (Durham et al., 1996), characteristics of the offender (Cochran et al., 2003; Durham et al., 1996; Ellsworth & Gross, 1994; Moon, Wright, Cullen, & Pealer, 2000; Sandys & McGarrell, 1995; Skovron, Scott, & Cullen, 1989), and type of victim (Applegate, Wright, Dunaway, Cullen, & Wooldredge, 1993; Durham et al., 1996). People also vary greatly in the reasons for supporting capital punishment (Bohm, 1992; Gross, 1998; Harris, 1986; Sandys & McGarrell, 1995). Furthermore, much as Marshall predicted in Furman, several studies have found that support for the death penalty is contingent on the degree to which the public is informed about the realities of how the death penalty is administered and what alternatives are available (Bohm et al., 1991; Bowers, Vandiver, & Dugan, 1994; Fox, Radetet, & Bonsteel, 4 See Bohm (1991) and Longmire (1996) for a discussion of the relationship between such factors and death penalty attitudes.

6 526 CONFIDENCE IN THE DEATH PENALTY ; Sandys & McGarrell, 1994; Sarat & Vidmar, 1976; Vidmar & Dittenhoffer, 1981). Several studies have found support for Marshall's hypothesis. One of the earliest was conducted by Sarat and Vidmar (1976), who used a classical experimental design. After information about the death penalty was presented, support among those in the experimental group declined to less than 50% (from 62% to 42%). The control group experienced no change (Sarat & Vidmar, 1976). Vidmar and Dittenhoffer (1981), replicating the Sarat and Vidmar (1976) study in Canada, reaffirmed their findings, reporting: "Opinion in the experimental subjects changed from approximately 33 percent opposing capital punishment to approximately 71 percent opposing it" (p. 53). In a similar and more recent analysis, Bohm et al. (1991) found significant increases in opposition to the death penalty as knowledge of its application and administration was introduced. However, subjects whose support for the death penalty was based on a desire for retribution s were less affected by such knowledge. Several studies have also found that support for the death penalty diminishes when people are given information about sentencing options. Bowers et al. (1994) and Sandys and McGarrell (1994) report that the percentage of citizens favoring the death penalty decreased when people were presented with the option of life without parole. This was true among both samples of citizens and samples of legislators, although support declined more substantially among citizens. Longmire (1996), analyzing "attitudinal constancy," concluded that "73% of the people are inconstant in attitudes toward this sanction" and that "attitudes about this sanction vary with information about biases associated with its administration" (p. 107). Specifically, Longmire (1996) found that attitudes about the death penalty were affected by information on its deterrent effect, the cost of applying it, its disproportionate application to minorities and the poor, the age and mental status of the offender, and the likelihood of wrongful convictions. Bowers et al. (1994) go so far as to suggest: "Public support for capital punishment is an illusion that has become a selfperpetuating political myth" (p. 142). It is clear that support for the death penalty and public confidence in its application and administration is much more complex than that which can be 5 Retribution, in the context of Marshall's hypothesis and Bohm et al.'s related research, refers to what Finckenauer (1988) terms "retribution as revenge" in contrast to "retribution as just deserts." The distinction between these two forms of retribution is critical to the present study. It is the former revenge-oriented retribution that we refer to when discussing "retribution."

7 VOLLUM, LONGMIRE, AND BUFFINGTON-VOLLUM 527 determined by a single, simple question regarding whether one supports or opposes the death penalty. When people are "confronted with the particulars of crimes and defendants, with responsibility for its application, and with information about the realities of capital punishment" (Bowers et al., 1994, p. 144), public support for the death penalty erodes. Nevertheless, the death penalty persists as legal and criminal justice policy and continues to be based on questionable judgments regarding public support. Although "evolving standards of decency" indeed may be changing, policy makers and judges have yet to heed Justice Marshall's observations by considering informed public opinion and public confidence regarding the death penalty as currently administered. PRESENT STUDY The purpose of this study is to gauge public confidence in the death penalty in Texas and to examine that confidence in light of the numerous problems with its implementation that have become apparent in recent years. Moreover, it is our objective is to obtain a deeper understanding of public support for the death penalty by considering potential underlying bases and dimensions of such support. Support for a moratorium is also examined as further indication of public concern about the efficacy of the current death penalty system. The complexity of public attitudes is considered in the context of Marshall's hypotheses that revolve around the contention that support for the death penalty diminishes as information about how it is administered becomes known. We also extend our focus to a more neglected portion of Marshall's hypotheses. As Bohm et al. (1991) point out, Marshall also suggested that, to the degree that retribution is the basis for support of the death penalty, knowledge about how it is administered will have little effect on attitudes. This statement is of primary interest to us as we assess public support for the death penalty, public confidence in the death penalty system, and their interrelationship. We contend that attitudes about and support for the death penalty are more complex than Marshall's hypotheses might suggest. Although we hypothesize along with Justice Marshall that support for the death penalty will diminish as knowledge about the process increases, we also suspect that this may not hold true for a significant proportion of the death-penalty-supporting population. We hypothesize that a significant number of people support capital punishment regardless of practical or technical considerations. In essence, there are two classes of support: (1) support that is "elastic" in the sense that, as Marshall argued, it will wane as

8 528 CONFIDENCE IN THE DEATH PENALTY knowledge about the process increases ("soft" or "rational/practical" support), and (2) support that does not waver regardless of knowledge about the process ("hard" or "value-expressive "~ support). By '~alue-expressive" we mean that support for the death penalty is primarily a matter of principle, based in an individual's value system (see Katz, 1960). This may be related to culture, personal experiences, socialization, or many other factors, and, as suggested by Bohm et al.'s (1991) analysis, may revolve around the degree to which revenge-oriented retribution is the objective that supporters seek. Regardless, such attitudes appear to reflect underlying value systems and differ significantly from those of "soft" supporters whose views are more sensitive to how the death penalty is administered. These dimensions of attitudes seem to present an important avenue of inquiry. Understanding each group may lead to better understanding public attitudes about the death penalty. Examining confidence in the death penalty and how it is related to support for the death penalty will shed considerable light on those attitudes. METHODOLOGY Sample and Data Sources Data for this study were obtained as part of the Texas Crime Poll, a mail survey conducted by the Survey Research Program in the College of Criminal Justice at Sam Houston State University during the summer of Questions on attitudes about the death penalty--including confidence in the death penalty system and support for a moratorium--were included. A random sample of 3,114 postal addresses of individuals throughout the state of Texas was purchased from a commercial sample selection firm, Survey Sampling, Inc. 7 The sampling frame consisted of all names and addresses listed in white-page telephone directories throughout the stated Surveys were sent to all 3,114 selected individuals. In an attempt to maximize survey response, a combination of initial and follow-up mailings were used, following, in part, Dilman's (1978) total design method of survey research. Approximately 2 weeks before mailing the questionnaires, a 6We borrow this term from Katz's (1960) conceptualization of the valueexpressive dimension of attitudes. Katz (1960) claims that such attitudes "have the function of giving positive expression to [an individual's] central values and to the type of person he conceives himself to be" (p. 173). Information about Survey Sampling, Inc. can be obtained at surveysampling.com. 8A more detailed description of the sampling procedures can be found at http Y/

9 VOLLUM, LONGMIRE, AND BUFFINGTON-VOLLUM 529 postcard, written in both English and Spanish, was mailed to all prospective respondents telling them about the project and informing them that a questionnaire would soon arrive. They were offered the option of receiving a questionnaire in Spanish. Two weeks later, questionnaires were mailed with a cover letter explaining the purpose and importance of the project, indicating the value of each person's response, and,thanking them in advance for participating. A postage-paid return envelope was included. Again, a note in Spanish offered the option of requesting questionnaires printed in Spanish by checking a box on the questionnaire and returning it in the postage-paid envelope. Two primary follow-up mailings were also sent. First, approximately 2 weeks after the questionnaires were mailed, postcards were sent to all subjects who had not yet responded. The postcards restated the importance of the project and provided a telephone number for subjects to call if they had not received a questionnaire or wanted another copy. Finally, approximately 2 weeks following the mailing of the postcards, nonrespondents were sent a second questionnaire, a letter urging cooperation, and a postage-paid envelope. Questionnaires were numbered, but subjects were assured that their responses would remain confidential. In spite of our best efforts to maximize response, only 821 completed surveys were received, yielding a return rate of 27%. 9 Ninety-one surveys were returned by the post office undelivered due 9 Concerns over this low response rate and suggestions of anonymous reviewers led us to attempt phone interviews with a random sample of 400 nonrespondents. In spite of three attempts to reach nonrespondents, we were able to complete only 38 interviews. Although this procedure did not yield enough additional subjects to permit comparison to our original 821 respondents, our failure to obtain interviews did shed important light on our sample. The table below illustrates the response results from our attempts to reach nonrespondents. Results from Attempts to Contact Nonrespondents (N = 400) Result # % Complete Interview Refused Interview No Contact Made Disconnected or Nonworking Phone Number Subject Moved Subject is Deceased Wrong Phone Number Note that a substantial proportion of the sample either had disconnected or nonworking phone numbers, had moved, or were deceased (in all 43.3% were not located at the number corresponding to the address in the sampling frame). This indicates to us that a sizable proportion of the original postal sample may not have been reached. Those subjects who were not reached with the postal survey should not have been counted as nonrespondents in the calculation of the response rate because they never had the opportunity to respond or refuse to respond. Conservatively, we estimate that our "real" response rate was mere in the range of 37% to 43%.

10 530 CONFIDENCE IN THE DEATH PENALTY to bad or unknown addresses, returned uncompleted because the respondent was either disabled or deceased, or returned after the closing date established for data collection. Overall, we surmise that the low response rate is due to several factors. First, the survey was rather lengthy. The items included in our analyses were part of a larger survey composed of 179 items related to a variety of crime and criminal justice issues. Second, the survey was mailed in mid-summer when people are more likely to be away on vacation. Third, there were weaknesses in the sampling frame that contributed to the tow response. We learned of these weaknesses through a subsequent attempt to administer telephone interviews to a random sample of 400 subjects who had not responded to the mail survey. This procedure yielded several enlightening findings. Specifically, a high proportion of disconnected numbers, bad addresses, and deceased subjects was detected (43.3% of the sample). If we extrapolate back to our response rate based on these findings, we can conservatively conclude that a more accurate response rate is in the range of 37 to 43% (see footnote 9 for more detail on the results). A final issue regarding our sample is the disproportionate representation of some demographic groups. The Appendix presents the demographic characteristics of our sample and compares them to characteristics of the Texas population. Males, Whites, older persons, the more affluent, and the more highly educated are over represented, t Although this sample bias is a potential methodological weakness, it represents an oversampling of those segments of the population known to be more likely to support the death penalty and to oppose policies limiting its use. As the reader will see, some of our findings are somewhat astounding in light of these considerations. Variables and Measurement Support for the death penalty. In an attempt to establish a base indicator of support for the death penalty, a single-item response to the question "Do you support the death penalty for the crime of murder?" was obtained. The response provides a simple dichotomous measure ("yes" or "no") of general support for the death penalty. lo The completed follow-up phone interviews of nonrespondents reveal similar demographic breakdowns and the same levels of disproportionate representation. However, we caution the reader that these results may be tenuous given the very low number of completed interviews (N = 38).

11 VOLLUM, LONGMIRE, AND BUFFINGTON-VOLLUM 531 Confidence in the death penalty system. Respondents were asked to respond to several items about their confidence in the death penalty system in Texas. They were asked to indicate whether they have "a lot of confidence" (scored 4), "some confidence" (3), "little confidence" (2), or "no confidence" (1) in five particular aspects of the death penalty system. Specifically, they were asked: "How much confidence do you have that the death penalty system in Texas protects innocent people from being executed; 2. is being imposed fairly on poor people; 3. is being imposed fairly on members of minority groups; 4. insures that people charged with capital murder in Texas receive competent legal representation during their trial; and 5. insures that people convicted of capital crimes in Texas have acceptable levels of access to the appeal process?" Furthermore, in an effort to ascertain the general level of confidence Texans have in the death penalty system, responses to these items were used to construct an index. For each case, the sum of the scores on each of the five items was calculated and divided by a count of valid (not missing) responses. For example, if a respondent indicated "no confidence" (confidence score of 1) on three of the items, "some confidence" (confidence score of 2) on one item, and refused to answer or left blank the other, the score would be calculated as: Index of confidence in death penalty score = ( ) / 4 = 1.25 Index values range from 1 to 4 with higher values indicating high levels of overall confidence in the death penalty system and lower values indicating low levels of confidence. (The example given would indicate that the respondent had little confidence in the death penalty system in Texas.) The five confidence items were also examined in other ways to measure more specific aspects of confidence in the death penalty. An index measuring confidence centered on class-based issues was calculated based on the items related to the death penalty being imposed fairly on minorities and the poor. Similarly, the two items pertaining to criminal justice process concerns--adequate legal representation and access to appeals--were used to calculate an index of confidence in the process. Finally, the innocence item stands alone as a third issue and is its own confidence "index." This breakdown allows us to compare across dimensions of confidence in the death penalty.

12 532 CONFIDENCE IN THE DEATH PENALTY Finally, several variables were constructed in an effort to identify respondents who lacked confidence (reported "little" or "no" confidence) in one or more aspects of the death penalty system and those who lacked confidence in all of the five aspects. Isolating these individuals allows us to more deeply analyze support for the death penalty by looking at the degree to which those lacking confidence in the death penalty system nevertheless continue to support it. Support for a moratorium. Respondents were also asked: "Do you think there should be a 'moratorium' on Texas's executions (a period during which there are no people executed in Texas) until the system can be studied to make sure the system protects innocent people from being executed; 2. is being imposed fairly on poor people; 3. is being imposed fairly on members of minority groups; 4. insures that people charged with capital murder in Texas receive competent legal representation during their trial; and 5. insures that people convicted of capital crimes in Texas have acceptable levels of access to the appeal process?" Responses were coded "yes" = 1, "no" = 0, and a dichotomous variable identifying those who supported a moratorium for at least one of the five items was created. This supplements the similar variables for confidence in the death penalty. Given such a variable, we are able to ascertain the degree to which those who believe that a moratorium is needed maintain support for the death penalty in principle. In addition, the sum of the scores on each of the five items was calculated and divided by a count of valid (not missing) responses resulting in an "Index of support for a death penalty moratorium." For example, if a respondent indicated support ("yes") for a moratorium on three items, no support ("no") on one item, and left the other blank, the score would be calculated as: Index of support for a death penalty moratorium score = ( )/4 =.75 The index values range from 0 to 1, with higher values indicating more support and lower values indicating lower levels of support. The example given above indicates a respondent with a relatively high level of support for a moratorium. In the same manner as the more finite breakdowns of confidence in the death penalty and the associated indices, three moratorium support indices were also constructed: one for class-based issues, one based

13 VOLLUM, LONGMIRE, AND BUFFINGTON-VOLLUM 533 on criminal justice policy issues, and one based on innocence concerns. Drawing such distinctions allows for comparison of moratorium support across differing issues and the identification of potential moratorium support domains. Sociodemographic variables. The survey included multiple items that measure respondent characteristics, including age, race/ethnicity, gender, level of education, household income, political affiliation, and religious affiliation. The coding and distribution of these variables is shown in the Appendix. FINDINGS A strong majority of Texans in our sample (82%) support the death penalty. If we examine Texas Crime Poll data for prior years, we see that this represents the continuation of a slight upward trend that began about 5 years ago. Still, the level of current support for the death penalty is well below those recorded in the early 1990s, when support peaked at 93%. (Figure I displays the general trend in death penalty support as reported by the annual Texas Crime Poll over the last 25 years.) Surprisingly, in spite of the overwhelming support for the death penalty, 64% of respondents supported a moratorium and 48% indicated little or no confidence in the death penalty system for at least one of the five reasons (items) listed in the survey. Table 1 reports the proportion of respondents indicating little or no confidence in the administration of the death penalty with respect to each of the five concerns, the mean level of confidence in the death penalty, and the proportion supporting a moratorium based on each concern. More than one-third of subjects reported little or no confidence that the death penalty is "imposed fairly on poor people" or "imposed fairly on minorities" (36% and 35%, respectively). The greatest level of confidence was expressed with respect to access to the appeals process (x = 3.13). A majority of respondents support a moratorium in order to protect innocent people from being executed (55%), to ensure competent legal representation during trial (55%), and to ensure acceptable levels of access to appeals (52%). Interestingly, although respondents expressed much lesser confidence that capital punishment was administered fairly in terms of race or class, a majority were unwilling to support a moratorium on either of these grounds. Table I also presents Spearman correlations indicating relationships among death penalty support, confidence in the death penalty, and moratorium support. Predictably, there were positive correlations between respondents' confidence in the death

14 Figure 1. Percentage Supporting Death Penalty for Crime of Murder J 0% I I t ---T"" ~... ] "T [ ----'" I - - q... T... T... T... f... b~ C~ 0 100%... C~ 75% - 5O% 25%

15 Table 1. Confidence in the Death Penalty and Support for a Moratorium in Texas, Texas Crime Poll Spearman Correlations * Moratorium Confidence Confidence Support Percent Mean Percent x x x Little/No Confidence Supporting Moratorium Death Penalty Death Penalty Item Confidence Level Moratorium Support Support Support...protects innocent people from being executed is being imposed fairly on poor people is imposed fairly on members of minority groups ensures competent legal representation during their trials ensures acceptable levels of access to the appeal process Z j1 On at least one of five items On all items *All reported correlations are significant at p <.01

16 536 CONFIDENCE IN THE DEATH PENALTY penalty and support for this form of punishment and negative correlations between confidence in the death penalty and support for a moratorium. Moreover, individuals who supported the death penalty were less likely to support a moratorium (moderately negative correlations), although these two factors exhibited the weakest correlations among the three variables. The strongest correlations among all these factors were for the item regarding the protection of innocent people from execution. Tables 2 and 3 present mean index values for the measures of confidence in the death penalty system and support for a moratorium, respectively. For each, ANOVA was used to identify significant differences in index scores across demographic groups. Significant differences between Black and White respondents were detected for index scores on both confidence in the death penalty system and support for a moratorium with Whites reporting much more confidence in the death penalty system in Texas (2.98 vs for Blacks) and substantially less support for a moratorium (.47 vs..83 for Blacks). Given what is known about race differentials in support for the death penalty, in general, this is not surprising. Similarly, males, with an index score of.45, were significantly less supportive of a moratorium than females (.63) and exhibited more confidence in the Texas death penalty system (3.02) than females (2.66). Protestants and Catholics also differed significantly from one another on both measures. Catholics indicated less confidence in the administration of the death penalty (2.79) than Protestants (2.94) and were much more likely to support a moratorium (.61 vs..49 for Protestants). An even more striking contrast exists between Republicans and Democrats. Republicans were significantly less likely to support a moratorium (.37) than Democrats (.72) and had significantly more confidence in the death penalty (3.16 vs for Democrats). There appeared to be distinct class-related differences as well. Those with low household income (less than $15,000) reported less confidence in the death penalty system and more support for a moratorium (2.67 and.69, respectively) than those with high household income (more than $60,000). The latter group scored highest on the confidence measure (3.03) and lowest among the household income categories on the index of support for a moratorium (.41). Finally, those with less than a high school education were significantly more likely to support a moratorium than all other education groups as well.

17 VOLLUM, LONGMIRE, AND BUFFINGTON-VOLLUM 537 Table 2. Indices of Confidence in Texas' Death Penalty System and Comparison of Demographic Groups Mean Index-- CJ Class- All Innocence Process Based Items Concerns Concerns Concerns Total Sample Race~thnicity ~ White 2.98* 3.10 * 2.86 ~ Black 1.96" L97" " Hispanic 2.53 '~ 2.66* * Gender * Male Female Education Less than High School High School Degree College Degree Graduate School Religion ~ Protestant 2.94** Catholic Other Age ~ Under Over Household Income ~ Less than $15, $15,000 - $30, $30,000 - $60, More than $60, ~ 3.13 " 2.91 ~ ~ 3.06"* Political Party t Republican 3.16 ~ 3.25 * 3.04 ~ 3.23 * Democrat Independent , Significant mean differences at p <.05. ~ Post-hoc test (Bonferroni) reveals significant difference from all other groups at p <.05. ~* Post-hoc test (Bonferonni) reveals significant difference between Protestant and Catholic at p <.05. ~*Post-hoc test (Bonferonni) reveals significant difference from the two lowest income groups (Less than $15,000 and $15,000-$30,000) at p <.05. The overall mean death penalty confidence and moratorium support measures were further analyzed in terms of their underlying components, i.e., confidence and moratorium support related to: (a) criminal justice process concerns, (b) class-based concerns, and (c) innocence concerns. Results indicate that respondents had the least confidence that the death penalty is being

18 538 CONFIDENCE IN THE DEATH PENALTY Table 3. Indices of Support for a Death Penalty Moratorium and Comparison of Demographic Groups Mean Index-- All CJ Process Class-Based Innocence Items Concerns Concerns Concerns Total Sample Race/Ethnicity ~ White.47 ~.49*.44*.48 ~ Black Hispanic Gender t Male Female Education* Less than High School.69 ~*, *,74 ~ High School Degree ,54 College Degree Graduate School Religion t Protestant.49 *~ *~ Catholic.61"**.61.59***.65*~ Other Age Under 26.58*.52 ~.6if Over Household Income* Less than $15, $15,000 - $30, $30,000 - $60, More than $60, Political Party ~ Republican.37*.39*.33*.40* Democrat Independent * Significant mean differences at p <.05. Post-hoc test (Bonferonni) reveals significant difference from all other groups at p <.05. * Post-hoc test (Bonferenni) reveals significant difference between those with less than a high school education and those with a high school or college education at p <.05. "**Post-hoc test (Bonferonni) reveals significant difference between Protestant and Catholic at p < Post-hoc test (Bonferonni) reveals significant difference from the two lowest income groups (Less than $15,000 and $15,000 - $30,000) at p <.05. imposed ~hirly on the poor and minorities (i.e., class-based concerns, 2.76) and greatest confidence in defendants' access to competent legal representation and the appeals process (i.e., criminal justice concerns, 3.00). Nevertheless, respondents were less likely to support a moratorium based on inequitable

19 VOLLUM, LONGMIRE, AND BUFFINGTON-VOLLUM 539 administration by class (.48), compared to criminal justice process and innocence concerns (.54 and.55, respectively). Compared to other racial/ethnic groups, Blacks expressed the lowest levels of confidence in access to counsel and appeals (i.e., criminal justice issues, 1.97) in the equitable administration of the death penalty by class and race (1.97), and in protection of the innocent from execution (1.91). Hispanics had slightly more confidence in criminal justice issues and innocence concerns (2.66 and 2.64, respectively) than in class- and race-equity (2.36) and, not surprisingly, Whites expressed the highest level of confidence in all aspects of the system (3.10 for criminal justice issues, 2.66 for class-based issues, and 2.98 for innocence concerns). Table 4 presents an assessment of the intersection between confidence in the death penalty and support for the death penalty, as well as between support for a moratorium and support for the death penalty. Although there is clearly a correlation between confidence in the death penalty and general support for the death penalty (Spearman's Rho =.408, p <.01) as well as support for a moratorium and support for the death penalty (Spearman's Rho = -.296, p <.01), there is a substantial subgroup of respondents who maintain their support for the death penalty in spite of exhibiting a lack of confidence in the death penalty and support for a moratorium on executions. Of those who indicated little or no confidence in the death penalty system in response to one of the five items presented, 68% maintained support for the death penalty. Moreover, 42% of those who indicated little or no confidence in all five aspects of the death penalty system nevertheless continued to support it. Similarly, 73% of those supporting a moratorium for at least one of the five items and 66% of those supporting a moratorium across the board maintained Table 4. Interrelationship Between Support for the Death Penalty and Confidence in the Death Penalty System and Support for a Moratorium Correlation with DP Support (Spearman's Rho, p <.01) For At Least One Item Of These, % that Support DP For All Five Items Of These, % that Support DP Mean Confidence in DP Index Score Mean Support for Moratorium Index Score Little or No Confidence Support for Moratorium % (N) % (N) 48 (383) 64 (510) 68 (261) 73 (374) 11 (88) 37 (297) 42 (37) 66 (195)

20 540 CONFIDENCE IN THE DEATH PENALTY support for the death penalty. This would appear to suggest that multiple dimensions of support for the death penalty exist--that support for the death penalty in principle and in practice may be very different. At the least, we can conclude that for some individuals, support for the death penalty is based in values not shaken by their lack of confidence in the fairness and accuracy of its actual application or administration. Table 5 presents an examination of respondents who report little or no confidence in the death penalty system on at least one of the five items, yet maintain support for the punishment ("hard" Table 5. Demographic Composition of Those Lacking Confidence but Supporting Death Penalty and Those Lacking Confidence and Not Supporting Death Penalty Those Reporting Little or No Confidence on at Least One Item Do Not Support ("Hard" Supporters) ~2 the Death Penalty Support the Death % Penalty % (Sig. Level) Total Sample Race/Ethnicity White Black (p <.01) Hispanic Gender Male Female (p <.01) Education Less than High School High School Degree College Degree (p =.051) Graduate School Religion Protestant Catholic (p =.104) Other Age Under (p <.05) Over Household Income Less than $15, $15,000 - $30, $30,000 - $60, (p =.846) More than $60, Political Party Republican Democrat (p <.01) Independent

21 VOLLUM, LONGMIRE, AND BUFFINGTON-VOLLUM 541 or "value-expressive" supporters). The most glaring differences exist with respect to political party ()C = , p <.01), race ()C = , p <.01), gender (~ = 7.494, p <.05) ' and age (~2 = , p <.05). More than 80% of Republicans can be classified as "hard" supporters, in contrast to only 58% of Democrats. The greatest crossgroup differences, however, are found with respect to race, with Whites much more likely to maintain support for the death penalty in spite of lacking confidence (73%) than Blacks (39%). Men were more likely to remain "hard" supporters than women: 75% of men maintained support for the death penalty in spite of a lack of confidence, in contrast to 62% of women. Finally, older respondents were more likely to be "hard" supporters than younger ones. DISCUSSION AND CONCLUSION Although a strong majority of Texans continue to support the death penalty, many are concerned about how it is administered and practiced. We have shown here that a significant proportion of respondents lack confidence in the death penalty system and that a majority supports a moratorium in order to ensure that the death penalty is being applied fairly and with due care and integrity. To put it simply, most have concerns about how the death penalty is applied in Texas and would like to see these concerns--such as the potential execution of innocent individuals, lack of adequate legal counsel, and limited access to appeals--addressed and alleviated. As moratorium legislation proposed during the 2003 legislative session in Texas remained tabled as the session came to an end, these findings are especially worthy of state lawmaker attention. Given lack of confidence and high levels of concern about the death penalty system in Texas, it is perhaps surprising that such a strong majority of citizens continue to support the death penalty as a form of punishment. In this study, we have attempted to provide some insight into this phenomenon. We have found that a significant proportion of respondents continue to support the death penalty although they lack confidence in its application. Notably, more than two-thirds of respondents who reported a lack of confidence in at least one aspect of the administration of the death penalty and more than 40% of those who lacked confidence in five different aspects of its administration nevertheless maintained support for the death penalty. We propose that such support for the death penalty is "value-expressive," meaning that support is not simply a function of faith in its practice or administration as a form of punishment, but often reflects more deeply held beliefs and values. It is clear that, in contrast to Justice Marshall's suggestion that knowledge about the application of the death penalty will

22 542 CONFIDENCE IN THE DEATH PENALTY cause people to reconsider their support for it, many supporters of the death penalty maintain their support as a matter of principle regardless of the facts pertaining to its practice. This is not to suggest that we disagree with Marshall's hypothesis that there is a correlation between knowledge about the application of the death penalty and support for the death penalty. Indeed, our findings support such a correlation. Overall, however, it appears that support for the death penalty is not a unidimensional phenomenon. Rather, our fmdings suggest that, although there are those whose support wanes as concerns about the application of the death penalty enter their consciousness, there are also those who continue to support the death penalty in principle, if not in practice. It is important to acknowledge such value-expressive death penalty support, and attempt to gain deeper insights into the social-psychological dynamics underlying such views. That said, there are a number of methodological limitations of this research that should be considered when drawing conclusions from the current study. Most significantly, the sample is admittedly unrepresentative of the general population: It overrepresents the views of White, middle-upper class, educated men. However, we argue that this skew actually increases the substantive significance of certain of our findings. Those overrepresented in our sample are the very groups most likely to support the death penalty (see, for example, Cochran et al., 1996, and Bohm, 2003). It is thus all the more astounding that a majority of our sample supports a moratorium and that many report little or no confidence in the death penalty system. Again, these results only bolster our contention that much death penalty support is value expressive and that many supporters make the distinction between the principle and the practice of the death penalty. At a more practical level, these groups tend to consist of the most active constituents--and leaders--in state and federal government. Therefore, the level of reported support for a moratorium has particularly meaningful practical significance. It is noteworthy, however, that support for a moratorium is at its lowest when considered in the context of class-based issues, despite the fact that the lowest level of confidence is reported for these same issues. Whether or not this is an artifact of sample bias is not known. Further research is necessary to tease out and further understand this finding. Other important limitations of this study include the inability to fully explore racial and ethnic differences, or to examine attitudinal dimensions and variation within racial minority groups. There are too few Black and Hispanic respondents to

23 VOLLUM, LONGMIRE, AND BUFFtNGTON-VOLLUM 543 support these analyses. Although some distinctions are noted with relative confidence, more complex analyses of differences by race in support for a moratorium, confidence in the death penalty, and value-expressive death penalty support are needed. Finally, the sample is limited to a single (and many would say unique) state. We suggest that greater attention to different regions and more adequate representations of various demographic groups are important directions for future studies. Much more research is necessary to gain an understanding of the multidimensional nature of attitudes surrounding the death penalty. Most research has focused on support as a single, onedimensional construct, ignoring potentially different social and psychological bases for such support. Future research should focus on developing better attitudinal measures and exploring more fully the different meanings and rationales associated with attitudes toward the death penalty and its application. Predictive models might be used to assess the overall and relative impact of various factors on death penalty attitudes. One particularly salient issue derives from the Bohm et al. (1991) study examining Marshall's suggestion that when (revenge-oriented) retribution provides the foundation for death penalty support, knowledge about the administration of the death penalty will have little effect on attitudes. This needs to be further examined as a potentially integral factor in our notion of "value-expressive" support for the death penalty. REFERENCES Applegate, B. K., Wright, J. P., Dunaway, R. G., Cullen, F. T., & Wooldredge, J. D. (1993). Victim-offender race and support for capital punishment: A factorial design approach. American Journal of Criminal Justice, 18, Barkan, S. E., & Cohn, S. F. (1994). Racial prejudice and support for the death penalty by Whites. Journal of Research in Crime and Delinquency, 31, Bohm, R. M. (1991). American death penalty opinion, : A critical examination of the Gallup polls. In R. M. Bohm (Ed.), The death penalty in America: Current research (pp ). Cincinnati, OH: Anderson. Bohm, R. M. (1992). Retribution and capital punishment: Toward a better understanding of death penalty opinion. Journal of Criminal Justice, 20, Bohm, R. M. (2003). Deathquest H: An introduction to the theory and practice of capital punishment in the United States (2nd ed.). Cincinnati, OH: Anderson. Bohm, R. M., Clark, L. J., & Aveni, A. A. F. (1991). Knowledge and death penalty opinion: A test of the Marshall hypothesis. Journal of Research in Crime and Delinquency, 28, Borg, M. J. (1998). Vicarious homicide victimization and support for capital punishment: A test of Black's theory of law. Criminology, 36, Bowers, W. (1993). Capital punishment and contemporary values: People's misgivings and the Court's misperceptions. Law & Society Review, 27, Bowers, W. J., Vandiver, M., & Dugan, P. H. (1994). A new look at public opinion on capital punishment: What citizens and legislators prefer. American Journal of Criminal Law, 22, Cochran, J. K., Boots, D. P., & Heide, K. M. (2003). Attribution styles and attitudes toward capital punishment for juveniles, the mentally incompetent, and the mentally retarded. Justice Quarterly, 20,

24 544 CONFIDENCE IN THE DEATH PENALTY Curtis, M. S. (1991). Attitude toward the death penalty as it relates to political party affiliation, religious belief, and faith in people. Free Inquiry in Creative Sociology, 19, Dilman, D. A. (1978). Mail and telephone surveys: The total design method. New York: John Wiley. Durham, A. M., Elrod, H. P., & Kinkade, P. T. (1996). Public support for the death penalty: Beyond Gallup. Justice Quarterly, 13, Ellsworth, P. C, & Gross, S. R. (1994). Hardening of the attitudes: Americans' views on the death penalty. Journal of Social Issues, 50, Ellsworth, P. C., & Ross, L. (1983). Public opinion and capital punishment: A close examination of the views of abolitionists and retentionists. Crime & Delinquency, 29, Finckenauer, J. O. (1988). Public support for the death penalty: Retribution as just deserts or retribution as revenge? Justice Quarterly, 5, Fox, J. A., Radelet, M. L., & Bonsteel, J. L. ( ). Death penalty opinion in the post-furman years. Review of Law and Social Change, 18, Gross, S. R. (1998). American public opinion on the death penalty: It's getting personal. Cornell Law Review, 83, Harris, P. W. (1986). Over-simplification and error in public opinion surveys on capital punishment. Justice Quarterly, 3, Jones, P. R. (1994). It's not what you ask, it's the way that you ask it: Question form and public opinion on the death penalty. The Prison Journal, 73, Katz, D. (1960). The functional approach to the study of attitudes. Public Opinion Quarterly, 24, Keil, T. J., & Vito, G. F. (1991). Fear of crime and attitudes toward capital punishment: A structural equations model Justice Quarterly, 8, Liebman, J. S., Fagan, J., & West, V. (2000). A broken system: Error rates in capital cases, [Online]. Available at: dpstudy/ Lifton, R. J., & Mitchell, G. (2000). Who owns death? Capital punishment, the American conscience, and the end of executions. New York: William Morrow. Longmire, D. R. (multiple years). Texas crime poll. Huntsville, TX: Survey Research Program, College of Criminal Justice, Sam Houston State University. Longmire, D. R. (1996). Americans' attitudes about the ultimate weapon: Capital punishment. In T. J. Flanagan, & D. R. Longmire (Eds.), Americans view crime and justice: A national public opinion survey (pp ). Thousand Oaks, CA: Sage Publications. Longmire, D. R., & Vollum, S. (2001). Confidence in the death penalty system in Texas and support for a death penalty moratorium: Special legislative report (2000 Texas crime poll). Huntsville, TX: College of Criminal Justice, Sam Houston State University. Longmire, D. R., & Vollum, S. (2003). Confidence in the death penalty system in Texas and support for a death penalty moratorium: Special legislative report ( Texas crime poll). Hun sville, TX: College of Criminal Justice, Sam Houston State University. Mignon, S. I., & Holmes, W. M. (1999). Police recruits' attitudes toward the death penalty. Criminal Justice Policy Review, 10, Moon, M. M., Wright, J. P., Cullen, F. T., & Pealer, J. A. (2000). Putting kids to death: Specifying public support for juvenile capital punishment, Justice Quarterly, 17, Sandys, M., & McGarrell, E. F. (1994). Attitudes toward capital punishment among Indiana legislators: Diminished support in light of alternative sentencing options. Justice Quarterly, 11, Sandys, M., & McGarrell, E. F. (1995). Attitudes toward capital punishment: Preference for the penalty or mere acceptance. Journal of Research in Crime and Delinquency, 32, Sarat, A., & Vidmar, N. (1976). Public opinion, the death penalty, and the Eighth Amendment: Testing the Marshall hypothesis. Wisconsin Law Review, 1, Skovron, S. E., Scott, J. E., & Cullen, F. T. (1989). The death penalty for juveniles: An assessment of public support. Crime & Delinquency, 35, Smith, M. D., & Wright, J. (1992). Capital punishment and public opinion in the post-furman era: Trends and analyses. Sociological Spectrum, 12, Texas Defender Service. (2000). A state of denial: Texas justice and the death penalty. Austin, TX: Author. Texas Defender Service. (2003). Lethal indifference: The fatal combination of incompetent attorneys and unaccountable courts. Austin, TX: Author. Vidmar, N., & Dittenhoffer, T. (1981). Informed public opinion and death penalty attitudes. Canadian Journal of Criminology, 23,