USF Libraries
USF Digital Collections

Race, religion, and attitudes toward capital punishment

MISSING IMAGE

Material Information

Title:
Race, religion, and attitudes toward capital punishment a test of attribution theory
Physical Description:
Book
Language:
English
Creator:
Schlaupitz, Sheila M
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla.
Publication Date:

Subjects

Subjects / Keywords:
death penalty
fundamentalism
protestantism
public opinion
religiosity
Dissertations, Academic -- Criminology -- Masters -- USF   ( lcsh )
Genre:
government publication (state, provincial, terriorial, dependent)   ( marcgt )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Summary:
ABSTRACT: The present study attempts to provide a more thorough understanding of public attitudes toward capital punishment. Two theories are tested toward this end, utilizing a random sample taken from the jury pool in Hillsborough County, Florida. First, An indirect test of reference group theory demonstrates the degree to which faith group, religiosity, and race affect public attitudes toward the death penalty. Surprisingly, there is little, albeit mixed, support for reference group theory as it attempts to explain attitudes toward capital punishment. Using the relationships between religion and capital punishment attitudes that we presumed would emerge from the test of reference group theory, we developed a processual model to more accurately describe how the relationships between religion and capital punishment operates. A test of attribution theory provided the vehicle through which this processual model could be tested. Although it seemed that the model was dependent on significant findings between the relationships in reference group theory, this was not the case. Indeed, we found many relationships between certain dimensions of religion and capital punishment attitudes that were consistent with previous research. Moreover, this piece of research is among the first to examine the effect of possessing a more progressive theological emphasis on death penalty attitudes. It addresses important theoretical and empirical questions regarding the direct and indirect relationships between religion and capital punishment attitudes.
Thesis:
Thesis (M.A.)--University of South Florida, 2003.
Bibliography:
Includes bibliographical references.
System Details:
System requirements: World Wide Web browser and PDF reader.
System Details:
Mode of access: World Wide Web.
Statement of Responsibility:
by Sheila M. Schlaupitz.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 181 pages.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001461871
oclc - 54938284
notis - AJQ2283
usfldc doi - E14-SFE0000228
usfldc handle - e14.228
System ID:
SFS0024924:00001


This item is only available as the following downloads:


Full Text

PAGE 1

Race, Religion, And Attitudes Toward Capital Punishment: A Test Of Attribution Theory by Sheila M. Schlaupitz A thesis submitted in partial fulfillment of the requirements for the degree of Master of Arts Department of Criminology College of Arts and Sciences University of South Florida Major Professor: Christine S. Sellers, Ph.D. M. Dwayne Smith, Ph.D. David Myers, Ph.D. Date of Approval: December 16, 2003 Keywords: death penalty, fundamentalism, pr otestantism, public opinion, religiosity Copyright 2004 Sheila M. Schlaupitz

PAGE 2

i Table of Contents List of Tables iii Abstract iv Chapter 1: Introduction 1 Chapter 2: Research on Attitudes Toward Capital Punishment 12 General Trends Over Time 12 Demographic Variation in Public Attitudes Toward Capital Punishment 13 Motivations for Attitudes Toward the Death Penalty 19 Changes in the Question Wording of the Dependent Variable 26 Chapter 3: Religion, Race, and Support for Capital Punishments 29 Attribution Theory 33 Chapter 4: Methodology 40 Sample and Procedure 40 Measurement of Variables 42 Capital Punishment Attitudes 42 Measures of Religion 45 Religious Conservativism 45 Religiosity 47 Race 48 Measures from Attribution Theory 49 Control Variables 63 Analytic Strategy 64 Chapter 5: Results: The Effects of Race and Religi on on Death Penalty Attitudes 66 Chapter 6: Results: Attribution Theory 80 Chapter 7: Discussion and Conclusion 90 The Race-Religion-Death Penalty Attitudes Relationship 90 Theoretical and Policy Implications 93 Future Research 94 References 96 Appendices 101 Appendix A: Jury Pool Survey 102 Appendix B: Informed Consent Form 123 Appendix C: Factor Analysis Results for the Overall Death Penalty Scale 124 Appendix D: Factor Analysis Results for the Religiosity Scale 125 Appendix E: Factor Analysis Results for the Attribution Theory Variables 126 Appendix F: Factor Analysis Results for the Control Variable Scales 128

PAGE 3

ii Appendix G: OLS Regression of Control Variables on Death Penalty Attitudes 129 Appendix H: OLS Regression of Control Variables on Death Penalty Attitudes Among Whites 145 Appendix I: OLS Regression of Control Variables on Death Penalty Attitudes Among Non-whites 161

PAGE 4

iii List of Tables Table 1 Levels of Support for and Opposition to Capital Punishment by Demographic Variables, Averaged Across 21 Gallup Polls, 1953-1986 14 Table 2 Percentage Distribution of Capital Punishment Attitudes by Specific Offenders and Offense Types 44 Table 3 Correlation Matrix of Death Penalty Scale Variables, Religion Variables, the Race Variable, and Control Variables 50 Table 4 Descriptions of the Death Penalty Scale Variables, the Religion Variable, the Race Variable, and the Control Variab les 55 Table 5 Correlation Matrix of Variables in the Attribution Theory Analysis 61 Table 6 Descriptions of the Variables in the Test of Attribution Theory 65 Table 7 Analysis of Variance of Death Penalty Attitudes by Religious Conservativism, Measured by Faith Group, Religious Tradition, and Membership in a Conservative Protestant Denomination 67 Table 8 OLS Regression of Death Penalty Attitudes by Religious Conservativism (Measured by Religious Tradition and Membership in a Conservative Protestant Denomination) and Religiosity 70 Table 9 Analysis of Variance of Death Pena lty Attitudes by Race 73 Table 10 OLS Regression of Death Penalty Attitudes by Religious Conservativism (Measured by Religious Tradition and Membership in a Conservative Protestant Denomination) and Religiosity Among Whites 75 Table 11 OLS Regression of Death Penalty Attitudes by Religious Conservativism (Measured by Religious Tradition and Membership in a Conservative Protestant Denomination) and Religiosity Among Non-whites 79 Table 12 OLS Regression of Biblical Literalness by Theological Emphasis Measured by a Forgiving Theological Emphasis Scale and a Punitive Theological Emphasis Scale 81 Table 13 OLS Regression of Attribution Style by Biblical Literalness 83 Table 14 OLS Regression of Punishment Philosophy by Attribution Style 85 Table 15 OLS Regression of Death Penalty Attitudes by Punishment Ph ilosophy 87

PAGE 5

iv Race, Religion, and Attitudes Toward Capital Punishment: A Test of Attribution Theory Sheila M. Schlaupitz ABSTRACT The present study attempts to provide a more thorough understanding of public attitudes toward capital punishment. Two theories are tested toward this end, utilizing a random sample taken from the jury pool in Hillsborough County, Florida. First, An indirect test of reference group theory demonstrates the degree to which faith group, religiosity, and race affect public attitudes toward th e death penalty. Surprisingly, there is little, albeit mixed, support for reference group theory as it attemp ts to explain attitudes toward capital punishment. Using the relationships between re ligion and capital punishment attitudes that we presumed would emerge from the test of reference group theo ry, we developed a pro cessual model to more accurately describe how the relationships between religion and capital punishment operates. A test of attribution theory provided the vehicle through which this processual model could be tested. Although it seemed that the model was depende nt on significant findings between the relationships in reference group theory, this was not the case. Indeed, we found many relationships between certain dimensions of religion and capital puni shment attitudes that were consis tent with previous research. Moreover, this piece of research is among the first to examine the effect of possessing a more progressive theological emphasis on death penalty attitudes. It addresses important theoretical and empirical questions regarding the direct and indirect relationships between religion and capital punishment attitudes.

PAGE 6

1 CHAPTER 1 INTRODUCTION Attitudes toward capital punishment have been wide ly examined for decades, and to some extent, public opinion has always influenced debates surrounding the death penalty. In the decades following the 1930s, the implications of public opinion on the death penalty have progressed to include establishing evolving standards of decency under the Eight Amendment ( Trop v. Dulles 1958, p. 101), as well as developing conditions under which the death penalty is ( Gregg v. Georgia 1976; Tison v. Arizona 1987; Penry v. Lynaugh 1989; Stanford v. Kentucky and Wilkins v. Missouri 1989) and is not ( Furman v. Georgia 1972; Coker v. Georgia, 1977; Lockett v. Ohio 1978; Enmund v. Florida 1982; Ford v. Wainwright 1986; Thompson v. Oklahoma 1988) constitutional. The leading United States Supreme Court cases have maintained that public opinion is a legitimate justifica tion for capital punishment; however, the implication remains that if public sentiment shifts to condemn the death penalty, capital punishment may then be ruled unconstitutional. In fact, public opinion may be a necessary circumstance for the abolition of capital punishment, as recent case law has reflected the Courts reluctance to place limitations on capital punishment under any circumstances when the public supports it. Early applications of the Eighth Amendment to death penalty decisions permitted capital punishment while prohibiting torture a nd barbaric means of executions (P aternoster, 1991; Vilo and Morris, 1997). New York introduced the electric chair in 1890 as a humane alternative to the previously primitive methods that had been used (Coyne and Ent zeroth, 1994). Since then, st ates have continued to search for more humane methods of execution based on scientific evidence and eyewitness accounts that extant methods (hanging, firing squad, electrocution, lethal gas, and lethal injection) cause great suffering and slow deaths, thus amounting to cruel and unusual punishment. The Supreme Court, however, has rejected Eighth Amendment claims that any of these methods are cruel and unusual, and all five means of execution are still used in the United States (Coyne a nd Entzeroth, 1994).

PAGE 7

2 Despite the Supreme Courts refusal to apply Eighth Amendment claims to methods of execution, it has endorsed the principle that the Cruel and Unus ual Punishment Clause should not be interpreted statically (Weems v. U.S. 1910), but rather in light of evolving standards of decency that mark the progress of a maturing society ( Trop v. Dulles 1958, p. 101). Interestingly, though, Trops punishment of three years of hard labor, forfeiture of all pay and allowances, and a dishonorable discharge from the U.S. Army (and consequently, denationalization) was not considered disproportionate to the crime of wartime desertion, which was, in fact, punish able by death. Rather, it was asse ssed against a principle of civilized treatment guaranteed by the Eighth Amendment and subsequently barred as a punishment due to its cruel and unusual nature. The Court further ruled that capital punishment could not be declared unconstitutional in a time when it was so widely accepted ( Trop v. Dulles 1958, p. 99). In the late 1960s and early 1970s, the Supreme Court, in light of the Trop decision, clarified and refined the implementation of capital punishment, even beyond the scope of Eighth Amendment challenges. For example, in Witherspoon v. Illinois (1968), the Supreme Court held that potential jurors reservations about capital punishment were insufficient grounds for automatically excluding them from sitting on a capital jury. The Court relied heavily on the belief that jurors reflect the conscience of the community on the ultimate question of life and death (p. 519). This link between the community and the penal system is an important indicat or of the evolving standards of decency that mark the progress of a maturing society. Conversely, it was only recently that the Supreme Court held that the refusal to ask jurors if they would automatically impose a death sentence, regardless of the facts of the case, was inconsistent with the Due Process Clause of the Fourteenth Amendment ( Morgan v. Illinois 1992). The issue of the role of juror discretion in capital cases emerged again in 1971 in the consolidated cases of Crampton v. Ohio and McGautha v. California Although this was a Fourteenth Amendment case to which Trop was not directly applied, it was an important pre-cursor to the landmark case of Furman v. Georgia because it addressed the potentia l arbitrary and capricious senten cing of death row inmates as a consequence of unrestricted juror discretion in capital sentencing proceedings. The Supreme Court rejected the claims of McGautha and Crampton, thereby uphol ding the constitutionality of standardless juries, as well as a single capital proceeding to make both guilt and sentencing determinati ons, the latter of which Crampton raised. The Court reasoned, in part, that de spite many challenges to standardless juries that had

PAGE 8

3 been presented to state and federal courts, no court had upheld such a challenge Because McGautha did not address the Eight Amendment question of the c onstitutionality of capital punishment per se, the de facto moratorium which had begun on June 2, 1967 after the execution of Luis Jose Monge in Colorados gas chamber, remained in effect (Coyne and Entzeroth, 1994). Within a month of the McGautha decision, however, the Supreme Court agreed to hear a group of cases rega rding whether the imposition and administration of capital punishment violated the Cruel and Unusual Punishment Clause of the Eighth Amendment, thereby also viola ting the Fourteenth Amendment ( Robinson v. California 1962). These cases were consolidated under Furman v. Georgia (1972). In a 5 to 4 per curiam decision, the Supreme Court held that the current imposition and administration of capital punishment were cruel and unus ual due to their arbitrary and capricious nature, which, in part, resulted from unguided jury discretion. Thus, the death penalty statutes of 40 jurisdictions, as well as various federal statutory provisions, were effectively voided, and 629 inmates had their death sentences commuted (Death Penalty Information Center, 2001a; Coyne and Entzeroth, 1994). The Court stopped short of declaring capital puni shment, per se, unconstitutional, th ereby freeing states to rewrite their statutes to comport with the standards set forth in Furman and essentially establishing a new system of capital punishment under which a better judgment on wh ether it could be constitutionally administered could be made. Over the next four years, 35 of the initial 39 states whose statutes had been invalidated rewrote them to meet the Furman standards. Ten states created mandatory death sentences for prescribed crimes, reasoning that this policy removed all juror disc retion. Mandatory capital punishment schemes for individuals convicted of first degree or felony murder ( Woodson v. North Carolina 1976), individuals convicted under a Louisiana statute which imposed an automatic death sentence for all first degree murders, but limited that category of offenders to five types of murder ( Roberts v. Louisiana 1976), and later for inmates who committed murder while serving a sentence of life without parole ( Sumner v. Shuman, 1987), were declared unconstitutional. Other states established sentencing guidelines for the judge and jurors, which included the consideration of aggravating and mitigating circumstances in making a determination of life or death. Three forms of these guided discretion statutes were upheld by the Supreme Court in Gregg v. Georgia

PAGE 9

4 (1976), Jurek v. Texas (1976), and Proffitt v. Florida (1976). Georgias procedural safeguards included the specification of ten statutory aggravating circumstances, one of which must be established beyond a reasonable doubt for a death sentence to be imposed; no requirement that a m itigating circumstance be found for a recommendation of mercy; and an automatic appeal of all death sentences, which included a proportionality review. The statut ory provision that was upheld in Jurek v. Texas enumerated five deatheligible offenses, one of which must be proven beyond a reasonable doubt for a death sentence to be sought. Further and contrary to most statutory revisi ons, Texas established three questions for jurors to answer, upon which a sentencing determination would be made: 1. whether the conduct of the defendant that caused the death of the deceased was committed and with the reasonable expectation that the death of the deceased or another would result; 2. whether there is a possibility that the defendant will commit criminal acts of violence that would constitute a continuing threat to society; and 3. if raised by the evidence, whether the conduc t of the defendant in killing the deceased was unreasonable in response to the provo cation, if any, by the deceased ( Jurek v. Texas 1976, p. 269). The Supreme Court held that the enumerated offenses constituted the finding of an aggravated factor, and the three questions presented to the jury focused on mitigating circumstances in a sufficient enough manner for the statute to be constitutional. Finally, Floridas revised death pena lty statute was similar to Georgias, except that in Florida, the jurys verdict is determined by a majority vote, and it is only advisorythe trial judge makes the final sentenci ng determination. In essence, four procedural reforms (consideration of aggravating and mitigating circumstances, bifurcated trial, automatic appellate review of convictions a nd death sentences, and propor tionality review), coupled with the Supreme Courts explicit rejection that capital punishment, per se, is unconstitutional, reinstated capital punishment in the United States. The years following the Gregg decision (1977 to the present) have been characterized by attempts by the Supreme Court to clarify the conditions under wh ich capital punishment is and is not constitutional. In comporting with Trops requirement that the Eighth Amendment be interpreted in light of evolving standards of decency, the Supreme Court has consiste ntly relied on indices of public opinion in making its determinations.

PAGE 10

5 Within two years of Gregg the Supreme Court ruled in Coker v. Georgia (1977) that capital punishment was disproportionate to the crime of ra ping an adult woman, and therefore, unconstitutional under such circumstances. Justice White emphasized that only 3 of the 35 states that had rewritten their preFurman statutes had included rape as a death-eligible crime, and 2 of the 3 statutes were invalidated because the death penalty was mandatory. (When these states rewrote their statutes a second time, neither included rape as a death-eligible offense, thus leaving Georgia the sole state to authorize capital punishment for the crime of raping an adult woman.) Further, Justice White noted that juries in Georgia had given death sentences to only 6 of 63 convicted rapists since 1973. Thus, Americans evolving standards of decency, as reflected by legislatures and juries, seem ed to suggest that death was an impermissible punishment for rape. Five years later, the Supreme Court employed a similar justification of the role of public opinion in determining death penalty standards in Enmund v. Florida (1982). Enmund had driven a getaway car following a robbery-murder, and he, like Coker, appealed his death sentence on the grounds that it was disproportionate to his involvement in the crime. Consistent with Coker the Supreme Court overruled Enmunds death sentence, and again cited public opinion as a basis for its decision. Specifically, it indicated that only 8 of the 36 stat es with capital punishment statutes permitted the imposition of the death penalty for defendants who only participated in robberies that turned into murders. Moreover, it pointed out that only 6 of 362 offenders who had been executed since 1954 were put to death for participating in a felony during which an accomplice committed a murder. Again, these statistics were representative enough indicators of public support for the death penalty for the Supreme Court to use them in its justification for overturning the death sentence of Earl Enmund. In decisions such as Coker v. Georgia and Enmund v. Florida the Supreme Court demonstrated its ability to exercise discretion when determining the circumstances under whic h offenders were deatheligible. However, even when the Court implemente d procedural safeguards against the death penalty, it still cited public opinion as a legitimat e justification for its decisions. The cautionary approach to death penalty issues was short-lived, as the Supreme Courtarguably at the urging of Justice Rehnquistbeg an to stress the importance of expeditious executions. Indeed, in 1985, the majority ruled that every error in the capital sentencing process did not warrant judicial

PAGE 11

6 intervention and that the state had a legitimate interest in speeding up the appeals process for death penalty cases ( Zant v. Stephens 1983). Similarly, in Herrerra v. Collins (1993), the Supreme Court held that new evidence of innocence, without the pr esence of other constitutional viola tions, was not a basis for federal courts to order a new trial. These cases, as well as other death penalty-related cases, reflected the publics overwhelming support for capital punishment throughout the mid 1980s and early 1990s (Bohm, 1991). In 1989, the Supreme Court determined that capital punishment was not a cruel and unusual punishment for the mentally retarded ( Penry v. Lynaugh ), nor was it cruel and unusual for juveniles 16 and 17 years old ( Stanford v. Kentucky and Wilkins v. Missouri ). Although public opinion was implicit in the Supreme Courts trend toward increasing the imposition of capital punishment during the 1980s, it also remained an explicit determinant of its decisions. For example, Justice Scalia, writing for the majority in Stanford v. Kentucky pointed out that there was not public consensus against the execution of 16 and 17 year olds. Of the 35 states with capital punishment statutes, 15 did not impose the death penalty for 16 year olds, while 12 refused to permit it for 17 year olds. Thus Americas evolving standards of decency did not conflict with the execution of 16 and 17 year olds. The U.S. re-affirmed its endorsement of the execution of juveniles when, in the process of ratifying the International Covenant on Civil and Political Rights in 1992, which prohibits the death penalty for i ndividuals who committed their crimes when they were under 18, it reserved the right to execute juveniles. (The U.S. is the only country with an outstanding reservation to this Covenant, and te n countries have since filed objections to it) (Death Penalty Information Center, 2001b). Given the U.S. Supreme Courts rulings over th e past two decades, which have broadened the category of eligible defendants who may be put to death, weakened safeguards against unfair death sentences, and sped up the appeals process, it is reasonable to assume that the number and rate of executions will increase. In fact, several U.S. Supreme Court decisions in the mid to late 1990s have demonstrated little, if any, constitutional sensitivity with respect to death row inmates Sixth ( Strickler v. Greene 1999), Eighth ( Harris v. Alabama 1995; ODell v. Netherland 1997; Buchanan v. Angelone 1998), and Fourteenth Amendment rights. Interestingly and somewhat surprisingly, the turn of the 21 st century marked significant changes in public opinion toward capital punishment and several ma jor developments in related research, challenges

PAGE 12

7 against, and legislation regarding the administrati on of capital punishment. The February, 2000 Gallup Poll illustrated that support for capital punish ment had reached a 19-year low. This apparent and dramatic shift in public opinion was validated by several other polls, which showed that most Americans support a moratorium on capital punishment until issues of fair ness can be resolved; almost 90 percent support the provision of access to DNA evidence in capital trials; and an overwhelming majority of citizens (83%) support providing qualified and experienced attorneys to capital defendants (Death Penalty Information Center, 2000). Although there were no prominent death penalty cases decided by the Supreme Court in 2000, politicians nationwide appear to have responded to the publics demand that the current system of capital punishment be improved. Notably, the Republi can and pro-death penalty Governor of Illinois, George Ryan, issued a moratorium on executions in Illinois. The Florida Supreme Court struck down a bill passed by the state legislature that attempted to seve rely cut the appeals process in capital cases. Also around this time, the Innocence Protection Act was introduced in the U.S. Congress, allowing for the DNA testing of all inmates. This concern over wrongful executions seems to be widespreadon May 11, 2000, the establishment of the National Committee to Prevent Wrongful Executions was announced. Immediately thereafter, on June 12, 2000, Professor Liebman, a law professor at Columbia University released a significant study of death penalty cases, revealing serious mistakes in two-thirds of all capital trials. Concerns over the execution of the ment ally retarded and juveniles and racial and regional biases were also reflected (e.g., in the American Bar Associations urge for a moratorium, U.S. Department of Justice study, the U.S. Supreme Court granting certiorari to John Paul Penry, Clintons six-month reprieve to Juan Raul Garza, and the case of Saldano v. Texas 2000) (Death Penalty Information Center, 2000). The long-term rise in executions since Gregg combined with the potential increase of juvenile executions, the heightened scrutiny of errors in capital trials and the execution of innocents (Radelet, 1992; Liebman, 2000) and the blatant systematic biases in the administration of the death penalty ( McCleskey v. Kemp, 1987) may lead to a shift in public opinion back toward opposition. Although the events of 2000 reflect a year of dramatic changes and not a trend, the developments in the beginning months of 2001seem to suggest that the shift in public opinion is not id iosyncratic. Indeed, several states have proposed legislation barring executions either for special o ffender populations like the mentally retarded or

PAGE 13

8 altogether, and many states have developed avenues through which inmates sentenced to die may gain access to DNA evidence to prove their innocence (see the Death Penalty In formation Centers website for more details). Furthermore, the Supreme Court heard the case of Shafer v. South Carolina in which it held that when future dangerousness is at issue in a capital case, juries must be informed whether a life sentence carries the possibility of parole. Of greater magnitude, the Supreme Court also re-evaluated the case of John Paul Penry and determined that one supplementa l jury instruction regarding mitigating evidence was not compliant with its earlier mandate in Penry I requiring that the jury be able to consider and give effect to mitigating evidence in making a sentencing determination. This potential shift in death penalty sentiment has important implications for the standards and practices involved in death sentences, since state legislators, as well as the Supreme Court, use public opinion as a basis for their decisions. Thus, it is important to c ontinuously measure public attitudes toward capital punishment, in terms of the bases and circum stances for both support for and opposition to the death penalty. Public opinion, particularly toward capital punishment, is signi ficant because the public acts as a pressure group on legislatures. This influence has a direct impact on the creation of law, which, in turn, reflects the nature of American society. This imp act of public attitudes thus underscores the importance of understanding the process through which these attitudes are shaped. How is public opinion, especially on life and death issues, formed? One of the more powerful influences on the development of attitudes toward any given issue is the reference group. Reference groups are, generally membership groups to which individuals belong (Cochran, Chamlin, Beeghley, Harnden, and Black well, 1996). Reference groups include family, friendship networks, work groups, or voluntary associa tions such as religion or faith groups. In fact, religion constitutes a unique social structure through which attitudes toward capital punishment may be analyzed. Specifically, unlike other reference groups, membership in a faith group provides individuals with distinct moral messages to which they should subscribe. The extent to which individuals use their religion as a point of reference, moreover, is largely a function of the salience their religion holds for them. This may be evidenced by their identification with Biblical and other moral messages delivered by a religious official. Regarding death penalty attitudes, religion may be a basis for arguments from either side, as the Bible contains justifications both for and against capital punishment.

PAGE 14

9 Confounding the effects of faith group membership and religious salience on death penalty attitudes is the variable of race or ethnicity. For example, Blacks and Hispanicseven those members of fundamentalist faith groupstend to adopt a more pr ogressive theological emphasis than their white counterparts. That is, their faith groups identify with more moderate or liberal social and political beliefs. Thus, the effects of faith group and other religious factors on public attit udes toward the death penalty may also vary by racial or ethnic group. Although reference group theory addresses why we should anticipate the influence of faith group and race on attitudes toward capital punishment, it does not address how these groups function to shape arguments surrounding the death penalty. In other words, what processes motivate individuals within and across different reference groups to support or oppose capital punishment? Attribution theory (Heider, 1958) provides one answer to this question. Specifica lly, it posits that people occupy one of two different attribution styles with regard to placing blame and/or causation for deviant and criminal behavior. The first is a dispositional attribution style, which is characterized by the belief that crime is chosen. In other words, offenders commit crimes as a result of selfish, but rational, motivations. A situational attribution style, on the other hand, is characterized by the belief that crime is caused by environmental forces (Heider, 1958; Young, 1991; Grasmick and McGill, 1994). Although a ttribution theory was initially a psychological theory of motivation, it has since been applied to criminology. The concepts of situational and dispositional attribution styles refl ect the criminological debate between the classical and positivist schools of thought (Cullen, Clark, Cullen, and Mathers, 1985). Specifically, the dispositional position attributes blame to the individual offender, t hus legitimizing punishment, while the situational attribution style looks into the environment for social cause s of crime, thereby emphasizing refo rmation and re-int egrative efforts, rather than punitive measures (Grasmick and McGill, 1994). These attribution styles end to reflect individuals attitudes toward other issues, as well, including support fo r formal sanctions that are imposed by the criminal justice system on offenders. Specifically, those with a dispositional attribution style are more likely to support punitive measures, as they believe that the offenders engaged in crime by their own choice and deserved to be punished. Individuals who possess a situational attribution style, on the other hand, are more likely to support rehabilitative efforts, as these tend to be aimed at helping the offender

PAGE 15

10 adjust to adverse circumstances that may have contri buted to their offending behavior (Cullen et al., 1985; Young, 1991). Although research has demonstrated empirical support for attribution theory, the question of who adopts a dispositional versus situational attributi on style remains largely unanswered. Contemporary research has assumed that religious affiliations influence peoples attribution styles, which, in turn, shape their correctional attitudes (Carroll and Payne, 1977; Hawkins, 1981; Cullen et al., 1985; Young, 1991; Grasmick and McGill, 1994). Specifically, the doctr ines of fundamentalist faith groups promote a dispositional attribution style, pres umably because they emphasize a str ongly literal Biblical interpretation, including the Old Testament focus on punishment and facing Gods wrath. By definition, individuals who possess a dispositional attribution style believe that crime is a result of free will, which should then make them more likely to support punitive sanctions, such as capital punishme nt (Grasmick and McGill, 1994). Conversely, members of moderate or liberal religious denominations should adopt a situational attribution style as a result of their exposure to more tolerant moral messages. Individuals who possess a situational attribution style believe that crime is a function of external forces, and therefore, offenders should be treated rather than punished. Thus, members of non-fundamentalist faith groups should oppose capital punishment as a result of their situational a ttribution styles (Grasmick and McGill, 1994). The present study investigates the linkages am ong race, religion, and attitudes toward capital punishment by engaging in several le vels of questioning. Initially, the independent direct effects of race and religion on death penalty attitudes are examined to establish whether they, alone, affect individuals attitudes toward capital punishment. Th e processes through which religious teachings direct individuals to choose opinions on secular social issues like the death penalty, are then addressed within the framework of attribution theory. We predict that individuals attribution styles will be determined by the theological emphasis of the faith group to which they subscribe. However, the strength of this relationship will be positively associated with the extent to which individuals identify religion as a salient institution, as well as their levels of involvement in their faith groups. The nature of the relationship between theological emphasis and attribution style may also be affected by race or ethnicity. Among individuals who identify religi on as important and who are highly involved in their faith groups, we expect this re lationship to play out as follows: fundamentalist

PAGE 16

11 denominations typically provide a forum that encourages their members to adopt conservative ideological beliefs (and hence, a punitive theological emphasis). Non-fundamentalist faith groups, on the other hand, will likely encourage their members to identify with more moderate or liberal social and political beliefs (and hence, adopt a more progressive theological emphasis). When race and faith group are considered, we e xpect white fundamentalists to be the most punitive and black non-fundamentalists to be the most progressive in their theological emphases (see Britt, 1998; Applegate, Cullen, Fisher, and Vander Ven, 2000). Individuals who possess a punitive theological emphasis will also adopt a dispositiona l attribution style, a nd individuals who posse ss a more progressive theological emphasis will identify with a situational attribution style. Individuals attribution styles, in accordance with attribution th eory, should be predictive of their view s on crime and justice policies, with people possessing a dispositional attr ibution style supporting more pun itive philosophies and respondents with a situational attribution styl e advocating rehabilitative responses to crime. The final path in the attribution theory model reflects a positive associ ation between support for punitive policies and support for capital punishment. Conversely, support for reha bilitative measures should be positively related to opposition to the death penalty. The purposes of this study, then, are twofold: (1 ) to examine the impact of race and religion on attitudes toward capital punishment, and (2) to test hypotheses derived from a ttribution theory which attempt to account for how religious teachings influence opini ons regarding more secular social issues like capital punishment. The thesis is organized in the following manne r. Chapter Two examines research on public attitudes toward capital punishment, fo cusing on demographic trends and motivations for attitudes toward the death penalty. Chapter Three provides a review of the literature on religion, punitive attitudes, and capital punishment support. Attribution theory provides the framework for this discussion. Chapter Four details the methodology of this study, including the sample and procedure, measurement of the variables, and the analytic strategy. Chapter Five provides the re sults from the preliminary an alysis of the effects of race and religion on capital punishment attitudes, and Ch apter Six reports the fi ndings from our test of attribution theory. Chapter Seven concludes with a discussion of the findings and their implications.

PAGE 17

12 CHAPTER 2 RESEARCH ON ATTITUDES TOWARD CAPITAL PUNISHMENT General Trends Over Time Although over 19,000 convicted criminals have been executed in the U.S. since the first occurred in Virginia in 1608 (Death Penalty Information Center, 2001a), it was not until 1953 that the first major American public opinion poll, a Gallup poll, regarding capital punishment was taken (Carroll, 2000). The 1953 Gallup Poll revealed that 68 percent of the respondents supported the death penalty for people convicted of murder, while 25 percent opposed it. Support steadily declined through the 1950s and 1960s, culminating in an all-time low of 42 percent in 1966 (Bohm, 1991). That year, 47 percent of Americans opposed capital punishment, and 11 percen t had no opinion. Over the next three decades, the public grew increasingly punitive (Durham, 1988; Pillsbury, 1989); pr o-death penalty sentiments reached new heights in 1994 with 80 percent in favor of capital punishment (Carroll, 2000). Bohm (1991) observed that since their beginning, the Gallup Polls have revealed two long-term and five short-term trends. The first long-term trend demonstrated that public support for capital punishment declined by 28 percen t between the years 1953 and 1966. During that time period, opposition increased by 18 percent and the proportion of respondents who did not know or had no opinion grew by 10 percent (Bohm, 1991). The trend culminated with a de facto moratorium on executions, beginning in 1967, and a de jure moratorium as a result of the historic Furman decision, in which the Supreme Court ruled that capital punishment, as it was administered, was unconstitu tional. The second long-term trend began in the mid-1960s. It demonstrated that death penalty support increased by 32 percent between 1966 and 1985, while opposition decreased by 28 percent, and the per cent of respondents with no opinion declined by 3 percent (Bohm, 1991; Carroll, 2000). Indeed, a 1972 Gallup public opinion survey revealed that 50 percent of Americans favored capital punishment for individuals convicted of murder. The percent of individuals in 1981 who favored the death penalty had risen to 66 percent, while only 25 percent of the citizenry opposed it. By 1985, support had reached a level of 74 percent, its highest ev er, and opposition was at 19

PAGE 18

13 percent, its record low. Between 1985 and 1994, a ttitudes toward capital punishment varied, with most Americans favoring it (70 to 80%) and few opposing it ( 16 to 22%). The 1994 poll reflected a new all-time high for death penalty support (80% ); however, since then, support ha s steadily declined, and opposition has steadily increased. The year 2000 reflected the closest gap between support and opposition in decades, with support at 66 percent, its lo west in decades, and opposition at 27 percent (The Gallup Organization, 2000). The shift in public attitudes to ward capital punishment that is refl ected in the two long-term trends is largely a function of the rapid social changes that occurred during the 1960s. They included the Civil Rights movement, student protests, and the due process revolution, all of which may have contributed, to varying degrees, to a decrease in the publics desire to punish offenders. The soci al and political climate of the 1960s also involved de-industrialization and high levels of crime, drug use, and promiscuity, which were reflected in this period of social unrest. One product of the social unrest of the 1960s was an upward shift in punitive attitudes among the public (see McCorkle, 1993). Demographic Variation in Public Attitudes Toward Capital Punishment As established above, when Americans are aske d questions about the death penalty, the results generally reveal high levels of support and low levels of opposition. Yet, there is consid erable evidence of variation in public sentiment, particularly when demographic variables are considered (Vidmar and Ellsworth, 1974; Bedau, 1982; Zeisel and Gallup, 1989; Bohm, 1991; Young, 1991; Barkan and Cohn, 1994; Durham, Elrod, and Kinkade 1996; Borg, 1997). Assessing how attitudes vary both within and across demographic categories may provide insight into the structure and dynamics of capital punishment sentiment and thereby reveal knowledge pert aining to evolving sta ndards of decency ( Trop v. Dulles 1957, 101). Table 1 depicts levels of support for, opposition to, and undecided attitudes toward capital punishment, averaged across 21 Gallup Polls conducted between 1953 and 1986, for a variety of demographic categories: city size, region of the country occupation, income, education level, age, gender, political party affiliation, race, and religion. Of the 21 Gallup Polls concerning death penalty attitudes taken befo re 1986, there are data on city size for 16. Residents of cities with over one million people were slightly more likely to support and less likely to oppose capital punishment than residents of c ities with less than one million people (Bohm, 1991).

PAGE 19

14 Table 1. Levels of Support for and Opposition to Capita l Punishment by Demogra phic Variables, Averaged Across 21 Gallup Polls, 1953-1986 Support (%) Opposition (%) Undecided (%) City Size 1 million and over 61% 30% 11% Under 1 million 55-56% 35-36% 8-10% Region South 55% 36% 10% West 64% 29% 7% East 61% 31% 9% Midwest 58% 34% 9% Occupation Professional/business 58% 35% 7% Manual 58% 33% 9% Clerical/sales 64% 28% 9% Income Level High 64% 30% 6% Low 50% 39% 11% Education Level Grade School Graduate 55% 33% 13% High School Graduate 61% 31% 9% College Graduate 58% 36% 6% Age Under 30 57% 36% 8% 30-49 60% 32% 9% 50 and over 60% 30% 11% Sex Females 53% 37% 10% Males 65% 29% 6% Political Party Democrat 55% 36% 10% Republican 65% 27% 9% Independent 58% 34% 9% Race Black 41% 48% 11% White 61% 31% 8% Religion Catholic 62% 30% 8% Protestant 59% 32% 9% *Adapted from Bohm, R.M. (1991). American death penalty opinion, 1936-1986: A critical examination of the Gallup Polls. In R.M. Bohm (Ed.) The Death Penalty in America: Current Research Cincinatti, OH: Anderson.

PAGE 20

15 On average, 55 to 56 percent of respondents who resided in cities with populations under one million people supported capital punishment, while 35 to 36 per cent were opposed. In cities with over one million people, the mean for death penalty support was 61 percent, while the mean for opposition was 30 percent. An average of eight to ten percent of respondents in cities with under one million residents and eleven percent of respondents in cities with over one million people were und ecided (Bohm, 1991). One explanation for the higher levels of support and lowe r levels of opposition among people in large cities may be that crime tends to be concentrated in urban ar eas. Thus, urban residents should be more likely to support punitive measures that would re move offenders from their streets. Data on region of the country were available for every year of the poll except one (Bohm, 1991). Surprisingly, Bohm (1991) found that people residing in the South generally revealed the lowest levels of support for and the highest levels of opposition to capital punishment, with means of 55 and 33 percent, respectively. Conversely, the West demonstrated the highest level of support and the lowest level of opposition, followed by the East and the Midwest, respectively (Bohm, 1991). This result contradicts expectations, as 60 percent of all executions betw een 1930 and 1980, and 93 percent in the post-Furman era, occurred in the South (Bohm, 1991). One possible explanation for this anomaly is the high number of poor blacks living in the South, where the judiciary is almost all white. Other explanations have suggested that structural racial biases in the execution of blacks in the South may be associated with attitudes toward the death penalty (Keil and Vito, 1992; Borg, 1997). Further, in her analysis of the subculture of punitiveness, Borg (1997) determined that the southern subculture of punitiveness (p. 41) may only apply to certain individuals, particularly the politically conservative, religious fundamentalists, and the racially intolerant. Zeisel and Gallups (1989) examination of the 1985 and 1986 polls revealed that the West demonstrated the highest level of support for capital punishment, followed by the South, the Midwest, and the East, respectively. Among the 20 polls that included informati on on respondents occupations, there was little variation between the categories, as the mean percen tages for favoring the death penalty were 58 percent for both the professional and business and the manual categories and 64 percent for the clerical/sales category (Bohm, 1991). Mean percenta ges for opposition were 28 percent fo r clerical/sales, 33 percent for manual, and 35 percent for professional and business. Mean percentages for respondents who did not

PAGE 21

16 express an opinion were approximately 9 percent for manual and clerical/sales and 7 percent for professional and business (Bohm, 1991). The low mean level of support for capital punishment among workers in manual positions may be explained by the fact that they are comprised of a greater proportion of minorities. Since offenders, particularly those sentenced to death, are members of racial or ethnic minorities of lower or working class backgrounds (F lanagan and McCleod, 1983; Baldus, Pulaski, and Woodworth, 1983; Eckland-Olsen, 1988; Young, 1991; Dieter, 1993; Bedau, 1997), it seems logical that persons in similar positions woul d be opposed to capital punishment. The low support and high opposition by the wealthy, on the other hand, is likely a function of their higher levels of education, which exposed them to more socially and politically liberal view points than less educated individuals would receive. In the 19 polls for which there were data on respondents income, individuals in the highest socioeconomic group demonstrated higher levels of s upport for and lower levels of opposition to capital punishment than those people in the lowest income group. On average, 64 percent of individuals in the top income category support capital punishment compared to 50 percent of those in the bottom income category (Bohm, 1991). Relatedly, an average of 30 percent of the individuals in the highest income level opposed capital punishment, while an average of 39 percent of individuals in the lowest income level opposed it (Bohm, 1991). Smith (1990) also discovered a consistent relationship between low socioeconomic status and opposition to the d eath penalty, and he attributed the differentiation to the desire for order among the wealthy and perceptions of injustices within the criminal justice system among the poor. Information on respondents education was availabl e in 16 of the 21 polls. With a few exceptions, college graduates were the most likely to oppose capital punishment (mean=36%), followed by grade school graduates (mean=33%), and high school graduates (mean=31%). Mean levels of support, on the other hand, were 55 percent for grade school gradua tes, 61 percent for high school graduates, and 58 percent for college graduates (Bohm, 1991). The relationship between education level and capital punishment attitudes, however, may be spurious (S mith, 1990). In other words, because high school graduates tend to make more money than individuals who did not graduate, the apparent effect of their education levels on their attitudes to ward capital punishment may actually be a function of their income, as income and education tend to be highly correlated. Although income created a confounding effect between grade school graduates and high school graduates, Smith (1990) argued that it suppressed the relationship

PAGE 22

17 between high school and college gradua tes. That is, because the direct effect of education on support for capital punishment is negative for high school and college graduates, but the effects of education on income and income on support for capital punishment are bot h positive, the positive path between income and capital punishment suppresses the true negative net effect of education on capital punishment. Two distinct patterns emerged with regard to the effect of age on death penalty sentiment. Until 1960, those under 30 were more likely to favor and le ss likely to oppose the death penalty than people over 50. Individuals between 30 and 49 usually fell betw een the two groups. After 1960, the trend reversed, and respondents over 50 were more likely to support and less likely to oppose the death penalty than individuals under 30. People between the ages of 30 and 49 fell closer to the 50 and over age group (Bohm, 1991). The pattern of death pe nalty attitudes with regard to age is likely the function of a cohort effect. Individuals who grew up prior to the 1960s (i n the under 30 category who exhibited higher levels of support for the death penalty) were the same people who aged into the over 50 category following the 1960s. Thus, they maintained their support for capital punishment, while their younger counterparts who grew up in the 1960s, not surprisingl y, exhibited comparatively less fa vorable attitudes toward the death penalty. In all 21 Gallup Polls, as well as in other public opinion polls, the proportion of males who favored the death penalty exceeded the proportion of females. Conve rsely, a consistently greater percentage of females than males opposed the d eath penalty (Zeisel and Gallup, 1989; Bohm, 1991; Ellsworth and Gross, 1994). There was a 12 percent difference in the mean levels of support for capital punishment, with an average of 65 percent of males and 53 percent of females favoring it. Further, there was an 8 percent difference in opposition levels, with an average of 29 percent of males and 37 percent of females against the death penalty (Bohm, 1991). Consistent with earlier polls, the 2000 Gallup Poll reflected a 10 percent difference between mens and womens support for capital punishment percent to 61 percent, respectively (Jones, 2000). Several schol ars have explained gender differences in variation toward capital punishment as a function of differences in the socialization process (Smith, 1990; Gelles and Strauss, 1976; Vidmar and Miller, 1980) Females are generally socialized to be empathic, nurturing, kind, and understanding, while males are raised to be t ough and unfeeling toward other human beings. These

PAGE 23

18 gender roles then tend to be translated to attitudes and behaviors. Their application to death penalty attitudes suggests that males should be more supportive and females more opposed. An evaluation of the Gallup Polls, 19 of which include information on re spondents political party affiliations, demonstrates that Democrats generally exhibited the highest levels of opposition and the lowest levels of support; Republicans registered the lowest levels of opposition and the highest levels of support; and Independents fell between them (National Opin ion Research Center Poll 1972-1977 in Bedau, 1982; Bohm, 1991; Jones, 2000). These trends are not surpri sing, as Republicans tend to hold both politically and socially conservative beliefs, including support for tough on crime policies like capital punishment. The year 1957 was unusual, though, as there was only a 2 percent difference between Democrats and Republicans for both support for and oppositi on to the death penalty (Bohm, 1991). Among the demographic variables, racial differences reflected the largest differentials in death penalty attitudes (Zeisel and Gallup, 1989; Bohm, 1991). On average, ove r the 50-year period, 61 percent of whites supported capital punishment, as opposed to 41 percent of blacks. Thus, there was a 20 percent difference in the mean levels of support between wh ites and blacks. Regarding opposition, 48 percent of blacks were against the death pena lty, compared to 31 percent of wh ites, revealing a 17 percent mean difference. With the exceptions of the years 1953 and 1965, a greater percentage of blacks than whites have identified themselves as undecided on the issue of capital punishment (Bohm, 1991). Barkan and Cohns (1994) investigation of the sources of racial variation in death penalty support showed that white support for capital punishment was associated with prejudice against blacks. Young (1991), on the other hand, found that whites disbelief that poverty causes crime influenced their support for capital punishment, while black opposition to the death pena lty was associated with their dist rust in the police. This was consistent with expectations, as punitive sanctions, in cluding death sentences, have been disproportionately unfavorable for blacks (Radelet, 1981; Baldus et al., 1983; Radelet and Pierce, 1985; Eckland-Olsen, 1988). In 17 Gallup Polls, there was information on re spondents religions. Only in 1965 did more Catholics than Protestants favor the death penaltythere were no differences between these faith groups in any other year (Bohm, 1991). An average of 59 percent of Protestants and 62 percent of Catholics favored capital punishment, while 33 percent of Protestants and 30 percent of Catholics opposed it. About 8

PAGE 24

19 percent of both groups were undecided (Bohm, 1991). The apparent sim ilarities between faith groups in their attitudes toward the death penalty were probably a function of measurement error, specifically combining all Protestants into one category, as well as ignoring differences in the strength of respondents religious convictions. More valid measures may be needed because of ideol ogical differences between fundamentalist and non-fundamentalist members of the sa me denomination that would not appear in faith groups. The difference between Catholics and Prot estants in 1965 may have been a consequence of Kennedys assassination, as Ke nnedy was Catholic (Bohm, 1991). Although there are some inconsistencies, overa ll trends reflect a pattern of support for and opposition to capital punishment. Specifically, whites, Protestants, Republicans, males, the relatively undereducated, residents of large cities, persons who live in the South, and people who have a lot of money tend to support the death penalty. Non-protestants, liberals, females, individuals who are relatively welleducated, residents of smaller cities, people who reside outside the South and the West, and people who do not have a lot of money are more likely to possess unfavorable attitudes toward capital punishment. Although such patterns are interesting and important the sources of support for and opposition to capital punishment are necessary to provide a more comp rehensive understanding of public opinion toward the death penalty. Motivations for Attitudes Toward the Death Penalty It is necessary to understand th e demographic variations in leve ls of support for and opposition to capital punishment, but this, alone, is not sufficient for a clear, comprehensive picture of public sentiment toward the death penalty in America, as only 8 percent of the variance in death penalty opinions has been explained by demographic variables (Fox, Radelet, a nd Bonsteel, 1991). Another issue that needs to be addressed is the motivation behind the attitudeswh y and under what circumstances do Americans support or oppose capital punishment? The four major punishment philosophies (deterrence, incapacitation, retribution, and rehabilitation) provi de possible rationales for the varia tion in capital punishment attitudes. Deterrence and incapacitation, which combined have been called utilitarian philosophies (Thomas and Foster, 1975; Zeisel and Gallup, 1989) and retributi on are generally viewed as sources of support for capital punishment (Vidmar and Ellsworth, 1974; Tyler and Weber, 1982; Warr and Stafford, 1984; Zeisel

PAGE 25

20 and Gallup, 1989; Cook, 1998), while s upport for rehabilitation has been asso ciated with anti-death penalty sentiments (Warr and Stafford, 1984). The deterrence philosophy of punishment seeks to reduce crime by using the offender as an example to make potential offenders realize that the cost of the punishment outweighs any potential benefits s/he may receive from committing the crime. Such a realization, accord ing to deterrence theory, occurs when the punishment is certain, severe, and sw ift. Although there has not been definitive evidence supporting the deterrent value of capital punishment, several scholars have argued that the opposite, a brutalization effect, occurs (Cochran, Chamlin, a nd Seth, 1994; see also Paternoster, 1991 for a thorough discussion of this issue). In spite of evidence to the contrary, deterrence is a frequent and primary reason people have cited for supporting capital punishment (V idmar and Ellsworth, 1974; Tyler and Weber, 1982; Warr and Stafford, 1984; Zeisel and Gallup, 1989). Several opinion polls have assessed public belief in the deterrent value of capital punishment. A 1972 poll conducted in Texas asked wh ether respondents belie ved fewer murders would be committed if individuals who were sentenced to death were actually executed. Most of the sample (52%) said yes; 36 percent said no; and 12 percent said they did not know (Vidmar and Ellsworth, 1974). Likewise, among individuals in a 1973 Iowa poll who supported the restor ation of the death penalty, 44 percent believed in its deterrent value (Vidmar and Ellsworth, 1974). One comprehensive effort toward determining motivations for death penalty support was undertaken in the 1973 Harris Survey. Fifty-six percent of the entire sample agreed that capital punishment was a more effective deterrent than life imprisonment; 32 percent disagreed. Among supporters of capital punishme nt, the percent of individuals who believed it was a better deterrent than life in prison rose to 76 percent, and among opponents, the level of agreement dropped to 29 percent (Vidmar and Ellsworth, 1974). Thus, there appears to be some evidence of an association between belief in deterrence and support for capital punishment. A study in Volusia County, Flor ida illustrated the complex nature of support for utilitarianism in general as a primary source of favorable attitudes to ward the death penalty (T homas and Foster, 1975). Thomas and Foster (1975) considered the interdepende nce of respondents perceptions of increasing crime rates, fear of victimization, willingness to employ punitive sanctions, a nd belief in deterrence, and they determined that the rising crime rate, coupled with fear of victimization and belief in the effectiveness of

PAGE 26

21 punishment led to an increased willingness to employ m easures of punitiveness, such as the death penalty (see also Vidmar and Ellsworth, 1974). Interestingly, Tyler and Weber (1982) found that although authoritarianism was the strongest predictor of capital punishment support, the interaction between fear of crime and belief in deterrence was an additional explanatory variable. Overall, however, their results indicated that symbolic political and social beliefs were more influential on attitudes toward the death penalty than instrumental beliefs, such as fear of crime and belief in deterrence. Nevertheless, these symbolic attitudes were, themselves, the major source of the belief in deterrence (Tyler and Weber, 1982). Results of these studies suggest that the belief in the efficacy of deterrence explains at least some of the variation in death penalty s upport, although further research is nece ssary to establish more definite conclusions. Indeed, there is some evidence to refute earlier indications that belief in deterrence is related to public attitudes toward the death penalty. For example, Ellsworth and Ross (1983) found that the majority of individuals would maintain their positions on capital punishment, even if they were wrong about its deterrent value. Likewise, Gallup polls in 1985 and 1986 revealed that support for the death penalty was not a function of individuals beliefs in its deterrent value. More recently, a 2000 Gallup Poll asked Americans who supported capital punishment why they favored itof the 12 available responses, deterrence ranked third, but only 8 percent of responde nts cited it as a basis for their support (The Gallup Organization, 2000). Incapacitation generally has not been found to be a primary basis of support for capital punishment; however, its importance may be underestimated in studies of public opinion (Bedau, 1997). Sixteen percent of death penalty supporters in one study indicated it as a reason (Zeisel and Gallup, 1989), and only 20 percent in another named it as the princi pal goal of punishment (Warr and Stafford, 1984). A 2000 Gallup Public Opinion Poll illustrate d that incapacitation is a relativel y low priority for death penalty supportersit ranked seventh out of the twelve bases for support provided in the response optionsand only 4 percent indicated that it was a reason for th eir support (The Gallup Organization, 2000). Gibbs (1978) ascertained that the incapacitation argument, as a basis for the implementation of the death penalty, would be effective only to the extent that it recognized that capital punishment must be imposed frequently and that the crime being punished was of a repetitive nature.

PAGE 27

22 Bowers (1993), on the other hand, argued that the prevalence of incapacitation in forming death penalty attitudes was overlooked due to the failure among researchers to include alternatives to capital punishment. Indeed, several opinion polls across th e U.S. have shown that although between 70 and 80 percent of the population endorsed the death penalty for murderers, less than 33 percent did so when life without parole plus restitution (L WOP+R) was an option (Bowers, 1990; Paternoster, 1991). Zeisel and Gallup (1989) found similar results in their study: 71 percent of the sample favored capital punishment, but 19 percent would oppose it if LWOP was an available alternative. Further, they found that 16 percent of their respondents would oppose capital punishment if they were convinced that it was not a deterrent, and among those who initially expressed opposition to capital punishment (21%), 4 percent would favor it if they were convinced that it was a deterrent (Zeise l and Gallup, 1989). Bowers (1993) also found that including a stipulation that the murderer would work in prison to provide monetary compensation for the victims family further decreased support for capital punishment. Thus, there seems to be support for the argument that death penalty responses change when alternatives are provided. As deterrence and incapacitation contribute to the social goal of reducing crime, retributive sentiments are a reflection of moral guidelines (B edau, 1997). Whereas most research has acknowledged the importance of retribution in shaping arguments for the death penalty (Vidmar and Ellsworth, 1974; Warr and Stafford, 1984; Zeisel and Gallup, 1989; Pate rnoster, 1991), Bedau (1997) refuted it, both as a punishment goal and as a legitimate basis for the support of capital punishment. He argued that the death penalty does not rectify anything and, therefore, the notion of retribution should be banned. He further contended that confident judgments regarding the offe nder and the context of the crime, which should have implications on the imposed sentence, can only be made when the individuals situation is objectively considered and not clouded by anger and other such emotions. Relatedly, Justice Marshall contended in his dissent in Gregg v. Georgia that there are two types of retribution, both of which are constitutionally and morally inadequate. First, he addresses the claim that the death penalty is necessary because it pre-empts th e citizenry from seeking private vengeance, and states that capital punishment clearly does not accomplish this re sult. Second, he consid ers the purely retributive justification of capital punishment, wh ich states that the death penalty is morally acceptable because it ends the lives of convicted murderers. Justice Marshall opined that the notion that society can make and carry

PAGE 28

23 out the judgment that a murderer deserves to die is proscribed by the Eighth Amendment. After all, the Eighth Amendment protects human dignity, for which capital punishment has at its very roots the total denial. Nonetheless and contrary to Bedau (1997) and Jus tice Marshall, most of the research, as well as the decisions of the United States Supreme Court, have established citizens willingness to endorse retribution as a primary basis of support for the death penalty (Vidmar and Ellsworth, 1974; Gregg v. Georgia, 1976; Warr and Stafford, 1984; Zeisel and Ga llup, 1989; Ellsworth and Gross, 1994; Cook, 1998). Whereas deterrence was once a more socially acceptable justification fo r capital punishment than retribution (Vidmar and Ellsworth, 1974; Thomas and Foster, 1975), social norms have changed and retribution is now seen as a legitimate reason to s upport the death penalty (Warr and Stafford, 1984; Fox et al., 1991). Consistent with this logic, lex talionis has been cited more frequently than deterrence among proponents of the death penalty since 1981 (Ellsworth and Gross, 1994). Results from the 1973 Harris Survey revealed a figure as high as 81 percent of the sample who agreed with statements measuring retribution as the basis for their support of capital punishment (Vidmar and Ellsworth, 1974). A Seattle Survey also showed overwhelming favor for retribution as the primary sentencing goal with regard to the death penalty (Warr and Stafford, 1984). There was a 22-percentage point difference between retribution and the next most popular sentiment, incapacitation. Although retribution did not receive support from the majority of the sample (42%), two-thirds of the respondents ranked it as one of their top three choices (Warr and Stafford, 1984). Forty-eight percent of the sample in the 1985 Gallup Poll reported retribution as the primary reason for their approval of capital puni shment, with 30 percent citing their belief in the philosophy an eye for an eye and an additional 18 percent naming just de sertsi.e., murderers deserve to be executed (Paternoster, 1991). An eye for an eye or fits the crime was the number one reason given by death penalty supporters for th eir positions on capital punishment in a 2000 Gallup Poll. An additional 13 percent of supporters cited reasons related to retr ibution, as well: fair punish ment (6%), they deserve it (5%), and serve justice (2%) (The Gallup Organization, 2000). Although punitiveness as a symbolic (versus inst rumental) response to crime does not entirely coincide with retribution, the two concepts parallel each other in many ways, particularly in that both involve the desire to punish the offender based upon the gravity of the offense and without concern for the

PAGE 29

24 crime control potential of the sanctions (Thomas and Foster, 1975; Tyler and Weber, 1982). Analyses of the instrumental and symbolic responses to crime ty pically include fear of victimization (Thomas and Foster, 1975; Stinchcombe, Adams, Heimer, Schepple, Smith, and Taylor, 1980), which is associated with the deterrence goal of reducing the crime rate and, thus considered an instrumental response to crime. These studies also tend to include measures of dogmatic or authoritarian persona lity, which are associated with having a punitive orientation, thereby making it a symbolic response to crime. Tyler and Weber (1982) tested the symbolic approach, specifically hypothesizing that authoritarianism would be associated with harsher punishments, including executions. The re sults were consistent with their hypotheses, as authoritarianism was the strongest predictor of death penalty support (Tyler and Weber, 1982). Cook (1998) expanded this body of research by measuring desire to punish with opinions regarding the death penalty and abortion. She found that individuals who favored capital punishment and opposed abortion displayed more punitive sentimen ts than respondents who were against capital punishment but in favor of a womans choice to have an abortion (Cook, 1998). A more in-depth analysis of variables related to the just deserts philosophy de monstrated that gender (males 2.4 times more likely than females), punitiveness toward sinners (almost twice as likely), political conservativism (1.8 times more likely), and opposition to euthanasia and suicid e (1.78 greater odds) were statistically significant correlates of retributive opinion (Cook, 1998). Whereas attitudes favorable toward punitive po licies, including the d eath penalty, reflected support for the deterrence, incapacitation, and retri butive sentencing philosophi es, non-punitive sentiments were associated with the rehab ilitative doctrine (Warr and Stafford, 1984; Applegate et al., 2000). Warr and Stafford (1984, p. 102) found th at rehabilitation was a secondary, ye t important, goal of punishment, as less than 20 percent of their sample identified it as a pr imary goal, but 59 percent ranked it in the top three. Education level and age were significantly related to goals of punishment, such that as education level increased, support for punitive philosophies decreased, and older res pondents were significantly more likely to employ the retributive doctrine than younger ones (Warr and Stafford, 1984). Many other studies have brought to light dimensions of public opi nion that are not purely punitive, but rather support the commitment of understanding a nd resources toward addressing the sources of offending behavior (Cullen, Cullen, and Wozniak, 1988; Gottfredson and Taylor, 1985; Skovron, Scott, and

PAGE 30

25 Cullen, 1988). Cullen, Skovron, Scott, and Burton (1990), however, were among the first to investigate support for punitive versus rehabilitative measures for di fferent offenses. They asked respondents whether they believed rehabilitation would be helpful for violent, non-violent, juvenile, and adult offenders and found that individuals tende d to believe less in rehabilitation for adult and violent offenders than for juvenile and non-violent offenders. McCorkle (1993) further developed this line of research, examining how support for punitive and rehabilitative measures varied across six offenses (robbe ry, rape, molestation, burglary, drug sale, and drug possession). They f ound that, although respondents expressed a high degree of punitiveness regardless of the cr ime, they were significantly mo re punitive toward offenders who committed crimes against people and less so toward nonviolent and drug offenders (primarily burglary and drug possession). Few studies have been conducted on the relati onship between public support for rehabilitation and support for capital punishment. Applegate et al. (2000) undertook an investigation that indirectly examined this relationship. Specifically, they examined the link between possessing a forgiving theological emphasis and correctional attitudes. They f ound that individuals who held a fo rgiving theological emphasis were more supportive of rehabilitation and less punitive than ot hers; they were also less likely to support capital punishment. These findings seem to suggest that there is an associa tion between support for rehabilitation and opposition to the death penalty. Although deterrence, incapacitation, and retributi on have all been associated with support for capital punishment, their independent explanatory power has varied over time. Deterrence and retribution have been consistently cited with greater frequenc y than incapacitation as a motivation for support for capital punishment. Until the early 1980s, deterrence wa s the most common justification for support for the death penalty; however, as research steadily inva lidated the deterrence argument (Cochran et al., 1994), and as retribution became more socially acceptable (F ox et al., 1991), it has been the most frequently cited reason for supporting capital punishment Understanding motivations for public attitudes toward the death penalty is necessary; however, it may not be sufficient for a clear and comprehensive view of public sentiment toward the death penalty in America. Anot her issue that needs to be addressed is the specific circumstances under which Americans support or oppose capital punishment.

PAGE 31

26 Changes in the Question Wording of the Dependent Variable Recent literature has increasingly focused on que stion wording and speci fic circumstances under which support for capital punishment wavers or st rengthens to acquire a more comprehensive understanding of public attitudes towa rd the death penalty in Ameri ca. Although some research has concluded that question wording was a largely insignificant determinant of capital punishment support (e.g., Bedau, 1997), most scholars agree that there is high variation in levels of support, depending on the context of the question (Vidmar and Ellsworth, 1974; Zeisel and Gallup, 1989; Paternoster, 1991; Durham et al., 1996; Borg, 1998). Question wo rding varied in specificity accord ing to the purpose of each study. The 1953 and 1965 Gallup polls asked about favoring the death penalty for women and juveniles, respectively, and found that 68 percent of the population favored capital punishment in 1953, and 65 percent favored it for women. Forty-five percent of the population favored capital punishment for murder in 1965, but only 23 percent favored it for juveniles (Vidmar and Ellsworth, 1974). A Minnesota Poll conducted in May, 1973 revealed that 49 percent of the sample favored a mandatory death sentence for the murder of a law enforcement officer, 59 percent suppor ted it when a kidnapper or hijacker killed someone, 58 percent for the assassination of a federal official, and 39 percent in the cases of crimes against the federal government (Vidmar and Ellsworth, 1974). The Harris Survey, conducted in June, 1973, concluded that support for a mandatory death sentence under a va riety of circumstances was never higher than 41 percent (Vidmar and Ellsworth, 1974; Bedau, 1982). The difference between the Minnesota Poll and the Harris Survey was response optionsin the Minnesota Poll, respondents had to choose either yes or no; whereas, in the Harris Survey, the response options depends and not sure appeared (Vidmar and Ellsworth, 1974). The June, 2000 Gallup Poll revealed that most Americans who support capital punishment (37%) do so with reservations, while an almost equal percentage support it without reservations (28%) as oppose it (26%) (Jones, 2000). A second major issue regarding response options was the inclusion of life without parole (plus restitution sometimes added: LWOP, LWOP+R). Several opinion polls across the United States have shown that although between 70 and 80 percent of the population endorsed the death penalty for murderers, less than 33 percent did so when LWOP+R was an option (Bowers, 1990; Paternoster, 1991). Zeisel and Gallup (1989) found similar results in their study: 71 percent of the sample favored capital punishment, but

PAGE 32

27 19 percent would oppose it if LWOP was available. A Gallup Poll conducted in August and September, 2000 reported that when life imprisonment with abso lutely no possibility of parole was an option, support for the death penalty fell from 67 percent to 49 percent (Carroll, 2000). Given the recent media attenti on to the arbitrary and discrimi natory application of capital punishment, as well as the potential execution of innocents (particularly the execution of Texas inmate Gary Graham), a Gallup Poll conducted in June, 2000 asked Americans whether they believed the death penalty is applied fairly. Results showed that 51 percent believe it is applied fairly, while 41 percent do not believe it is applied fairly. Further, among individuals who believe it is applied fairly, an overwhelming majority (86%) support capital punishment. On the othe r hand, persons who believe it is applied unfairly are divided percent oppose it, and 44 percent support it, suggesting that there is a strong correlation between belief in the fairness of cap ital punishment and favorable attitudes toward it (Jones, 2000). Similar to the 2000 Gallup Survey findings, a nationwide NBC/Wall Street Journal Poll reported that 42 percent of respondents supported a moratorium on capital punishment until questions about its fairness were resolved; 42 percent did not believe that Americas current system of capital punishment was administered fairly. Relatedly, a 2000 Newsweek Poll indicated that 82 percent of respondents believe that states should make it easier for death row inmates to introduce evidence that could prove their i nnocenceeven if it meant delaying the death penalty process (Death Penalty Information Center, 2001c). (Recall that the Supreme Court prohibited new evidence of actual innocence without the introduction of at least one other constitutional issue in federal post-conviction proceedings in Herrerra v. Collins). Another method of ques tioning respondents a bout support or opposition to the death penalty is the factorial survey. The structure of the factorial survey is such that variables are built into vignettes by randomly varying the specific values of these factors. Durham et al. (1996), for instance, used 2 versions of 17 homicide scenarios to determine the level of support for capital punishment. They found that the version of the homicide, which was a proxy measure for aggravating and mitigating circumstances, was a significant predictor of level of support for the dependent variable in 6 of the vignettes. When they ranked the 34 vignettes, degrees of support ranged from 24.4 per cent to 93.2 percent. They additionally found that only 5.2 percent of the respondents opposed the deat h penalty under all conditions, while 13.1 percent endorsed it in every vignette (Durham et al., 1996). Thus, there appears to be overwhelming support for

PAGE 33

28 the argument that question wording and response options significantly a ffect public attitudes toward capital punishment. Over the past 50 years, capital punishment attitude s have been the subject of major public opinion polls, as well as scholarly endeavors in the field of criminology. Although there has clearly been a progression in the understanding of public attitudes toward the death penalty particularly with regard to demographic correlations and bases of support, the origin s of these beliefs remain unclear. Some research has suggested that the source of deat h penalty attitudes may be rooted in religion. Chapter Three discusses the extant literature on religion, including the role of race in religion, and how they affect public attitudes toward capital punishment.

PAGE 34

29 CHAPTER 3 RELIGION, RACE, AND SUPPORT FOR CAPITAL PUNISHMENT Until recently, research on religion and punitive a ttitudes made only crude distinctions between faith groups, particularly between Protestants, Catho lics, and Jews (e.g., Blumstein and Cohen, 1980; Tyler and Weber, 1982; Bohm, 1991). Results of this research have largely reported that religion did not affect punitive attitudes. For example, 16 of 17 Gallup polls taken prior to 1995 indicated that Protestants and Catholics were not significantly di fferent from one anothe r on their attitudes toward capital punishment. No data were provided for Jews, members of other religions, or nonaffiliates. Although Blumstein and Cohen (1980) found that Catholics, Protestants, and Jews were more punitive than individuals with no affiliation, they also did not find significant differences among these faith groups. Consistent with previous findings, Tyler and Weber (1982) found no effect of religious a ffiliation on attitudes toward the death penalty. Recent studies have broken this pattern in favor of a more in-depth examin ation of the effects of religion on death penalty sentiments. Grasmick, Dave nport, Chamlin, and Bursik (1992) contended that it is because of the strong punitive overtones in JudeoChristian teachings and their belief in a literal interpretation of the Bible that Protestant fundamentalists are attracted to the retributive punishment philosophy. Their hypothesis was supported, as fundame ntalist Protestants were significantly more likely to support the retribution doctrine than liberal/moderate Protestants, Catholics, and individuals with no religious affiliation. These differences were maintained when the effects of religious salience and sociodemographic variables were controlled. The differe nces between fundamentalists and liberal/moderate Protestants and Catholics, however became non-significant when the eff ect of a measure of Biblical literalness was controlled, thereby suggesting that Bi blical literalness rendered a statistical association between faith group and support for retribution that is causally spurious. Although the distinction between fundamentalists and moderates/liberals is an improvement in the measure of Protestantism, Grasmick et al. (1992) maintain some of the measur ement crudeness in earlier studies both by failing to consider Jews in

PAGE 35

30 their analyses and by lumping liber al and moderate Protestants into one category when there may be notable differences in their social and political attitudes. In a follow-up study, Grasmick, Cochran, Bursik, and Kimpel (1993a) examined the effects of religious affiliation on levels of support for five punitiv e justice policies that included juvenile and adult death penalty, harsh courts, stiffer laws, and police use of deadly force. They found that liberal/moderate Protestants were significantly less retributive than evangelical/fundamentalist Protestants on every justice preference item, except police use of deadly force. Catholics fell between these two groups on the same four items; however, the only relationship that attained significance was that they were more likely to support juvenile executions than liberal/moderate Protestants, but less likely than evangelical/fundamentalist Protestants (Grasmick et al., 1993a). Again, Jews were not considered in the analyses, and liberal and moderate Protestants were combined into one group. Unlike previous research, Britt (1998) found no eff ect of membership in a fundamentalist church on death penalty support. He suggests three possible reas ons for this finding. The first is due to different measures of death penalty support. Rather than di chotomize death penalty a ttitudes into favorable or unfavorable responses, Britt used five Likert-s cale response options, rangi ng from strongly oppose to strongly favor capital punishment. Secondly, the non-significance of the effect of fundamentalism on death penalty support may be the result of differences in measures of religious affiliation, as Britt (1998) compared fundamentalist Protestants with one other category, which included all other mainstream faith groups, rather than separating out m oderate/liberal Protestants, Catholic s, and Jews. A third possibility could be that the non-significance of the effect of fundamentalism on support for capital punishment was actually a function of differences between black and white fundamentalists views. In other words, fundamentalist Protestants tend to be more proscr iptive (i.e., intolerant or prohibitive) than nonfundamentalists, which make them more likely to align with a conservative social agenda, including issues such as pornography, drinking, and abortion. Non-funda mentalists, on the other hand, tend to be attracted to a more liberal social agenda, which includes civ il rights, the environment, and other social justice concerns (Hall, 1997). However, blacks tend to hold mo re liberal views than whites. Thus, a distinction must be made between black and white fundamentalists.

PAGE 36

31 There is reason to believe that there are substantive racial differences between faith groups, particularly when Blacks and Hispanics are compared to whites (Br itt, 1998). The uniqueness of African American religion is rooted in its history, particularly slavery. Religion was multi-functional for blacks, as it was the only permissible means through which they could seek relief from oppression, while also trying to realize the goals of freedom and equality thr ough protest and action (Was hington, 1964). The common suffering of blacks during the Nineteenth Century, from which their religion was established, has created a unity that still prevails today. Here and there this folk religion may be identifiable with a given congregation, yet wherever and whenever the sufferi ng is acute, it transcends all religious and socio-economic barriers which sepa rate Negroes from Negroes. There are Negroes who are Protestants, there are Negroes who are Christians, there are Negroes in churches.There are Negro religious institutions which developed out of the folk religion. And it is this historical folk religion which unites all Negroes in a brotherhood which takes precedence over their individual pattern for the worship of god, or the lack thereof. The root of this folk religionis racial unity for freedom and equality. Every ecclesiastical expression of Negro congr egations and institutions is but a variation or frustration of this theme (Washington, 1964: 30-31). Accordingly, blacks should f unction differently than whites in their re ligious institutions. Specifically, they should possess different styl es of worship, receive different moral messages, and emphasize different dimensions of religiosity than whites. Hispanics, like blacks, have a cultural experience that creates a unique context through which they recognize and synthesize cultural, social, and religious goals. The major themes in Hispanic ministry reflect their group struggles: (1) inculturation or evangelization of culture; (2) the promotion of social justice or structural transformation; (3) modernity and secularization; (4) popular Catholicism/religiousness ( Dolan and Deck, 1994). The goal of evangelizing is conceived of as bot h personal conversion and structural change, and although there has been a bifurcation among religious leaders, many agree that personal conversions must occur in order for more just and humane structures to be built. The less emphasized, but important goal of social conversion is also contained in the first theme under inculturation. This is a particularly difficult goal to realize, as it requires a critical cultural consciousness, which is inhibited both by the ontology of American individualism and by the Hispanics negative attitudes and stereotypes of the oppressive American culture. Struct ural change remains an ideal in Hispanic ministry, but it remains largely ignored, particularly when the clergy is Hispanic, which is often the case. Hispanics

PAGE 37

32 are suspicious and uncomfortable with politics in the U.S., and thus, social teachings of the church remain largely uncommunicated. The goals of modernization and secularization may only be achieved through dialogue between Anglican and Hispanic cultures a nd through a critical reading of the Gospel, while simultaneously recognizing that the world is entering a post-modern era. In other words, a resolution must be made between traditional role perceptions and contemporary American values (e.g., progress and egalitarianism) (Dolan and Deck, 1994). Because of their unique group experien ce, Hispanics are more likely to use religious faith groups to achieve diffe rent purposes than either whites or blacks, but particularly whites. Thus, they are likely to r eceive different moral messa ges and emphasize different facets of religiosity than other races and ethnicities. With regard to capital punishment, Hispanics, like blacks, will likely be supportive of social justice concerns, thereby making them less likely to support the death penalty than whites. This should occur equally across their faith group memberships, although Hispanics who are part of moderate and non-proscriptive faith groups should be less likely to support capital punishment than Hispanics who are members of fundamentalist faith groups. In addition to social scientific research c onducted on the effects of r ace and religion on attitudes toward capital punishment, various faith groups have also set forth, in position papers and formal doctrines, their official stances on issues such as capital punishment. In the 1970s, numerous Christian groups supported capital punishment, perhaps the most notable being the National Association of Evangelicals (NAE), which represented more than 10 million conservative Christians and 47 denominations and the Moral Majority. The Christian Coalition, NAEs successor, has maintained its support for capital punishment, as well (Death Penalty Information Center, 2001b). Currently, the Southern Baptist Church, Jehovahs Witnesses, Mormons, Seventh Day Adventists and Fundamentalist and Pentecostal churches are among the only religious groups that continue to take a stand in favor of the death penalty. Ironically, the American Lutheran Church (different from the Lutheran Church in Am erica) has issued a statement in favor of capital punishment, contending that there is scriptural support for the states right to execute its citizens and that death is a necessary sentence to main tain order in society (Capital Punishment: What the Religious Community Says). To the contrary, the Jewish faith, the Roman Catholic Church, and most Protestant denominations including Baptists, Episcopalians, Lutherans, Methodists, Presbyterians, and the United Church of Christ, have spoken out against capital punishment (Capital Punishment: What the

PAGE 38

33 Religious Community Says; Death Penalty Information Center, 2001b; 2001d). These documents comport with the conclusions established in extant research. Given these official positions and supporting literature, the followi ng testable hypothesis has been derived: 1. As religious conservativism increases, support for capital punishment will increase. Relatedly, the current study also examines the interactive effects of religion and race on capital punishment attitudes. With this in mind, the followi ng predictions have been de veloped, consonant with Britts (1998) findings: 2. As religiosity increases, the effect of religious conservativism on capital punishment support will increase. 3. Across faith groups, support for capital punish ment will be greater among whites than among Blacks and Hispanics. 4. Among whites, as religiosity increases, the effect of religious conservativism on capital punishment support will increase. 5. Among Blacks and Hispanics, as religiosity increases, the effect of religious conservativism on capital punishment support will remain unchanged. Attribution Theory Although several studies have addressed why we should anticipate racial/ethnic and faith group differences in attitudes toward capital punishment, th ey fail to fully explor e the underlying processes through which religion may shape attitudes for and agai nst capital punishment. Attr ibution theory provides such an explanation. Specifically, it posits that people occupy one of two different attribution styles with regard to placing blame and/or causation for deviant and criminal behavior. The first is a dispositional attribution style, which is character ized by the belief that crime is chosen. In other words, offenders commit crimes as a result of selfish, but rational motivations. A situational attribution style, on the other hand, is characterized by the belief that crime is caused by environmental forces (Heider, 1958; Young, 1991; Grasmick and McGill, 1994). Although attributi on theory was initially a psychological theory of motivation, it has since been applied to criminology. The concepts of situational and dispositional attribution styles reflect the crim inological debate between the cla ssical and positivist schools of thought (Cullen et al., 1985). Specifically, the dispositional position attributes blame to the individual offender,

PAGE 39

34 thus legitimizing punishment, while the situational attr ibution style looks into th e environment for social causes of crime, thereby emphasizing reformation and re-integrative efforts, rather than punitive measures (Shaver, 1975; Grasmick and McGill, 1994). These attr ibution styles tend to re flect individuals attitudes toward other issues, as well, including support for formal sanctions that are imposed by the criminal justice system on offenders. Specifically, individuals with a dispositional attribution style are more likely to support punitive measures, as they believe that the o ffenders engaged in crime by their own choice and deserve to be punished. Individuals who possess a situational attribution style, on the other hand, are more likely to support rehabilitative efforts, as these tend to be aimed at help ing the offender adjust to adverse circumstances that may have contributed to their offending behavior (Hawkins, 1981; Cullen et al., 1985; Young, 1991). Early research on attribution theory demonstrated that the characteristics of the dispositional and situational attribution styles were consistent with the facets of the theory. Hawkins (1981), for example, categorized criminal offenses into crimes of violence committed by adults, property crimes committed by adults, and crimes committed by juveniles. Across a ll three types, he found that his sample of college students preferred more severe sanctions for crimes attributed to dispositional factors than crimes attributed to situational factors. Further, Cullen et al. (1985) examined the effect of attribution style, which they labeled classical (dispositional) and positivist (situational) on general pun itiveness, the effectivene ss of rehabilitation as a punishment goal, and support for capital punishment. Based on survey data from a random sample of 200 residents of Galesburg, Illinois, Cullen et al. (1985) determined that the positivist (i.e., situational) attribution scale had a significant negative effect on general punitiveness and capital punishment support and a significant positive effect on the belief that reha bilitation would be an effective punishment goal. That is, the more likely individuals were to attribute crime to environmental factors, the less likely they were to possess punitive attitudes, including support for capital punishment, and the more likely they were to support rehabilitation as a punishment goal. These findings held for their sa mple of both the general public as well as criminal justice practitioners. Young (1991) examined the influence of measures of situational and dispositional attribution styles on death penalty support. He found that, among whites, the belief that poverty causes crime

PAGE 40

35 (situational attribution style) was inversely related to support for capital punishment. Conversely, he found that, among whites, the belief that criminals commit crimes for rational self-serving reasons (dispositional attribution style) was positively rela ted to death penalty support. Young s results further suggested that blacks were more affected by perceptions of procedural and distributive justice than by attribution style. Early research on attribution theory demonstrated empirical support for its main tenets regarding the characterization of dispositional and situational attr ibution styles. It failed, however, to explore the relationship between religion and attribution style. Grasmick and his colleagues (1992; 1993a; 1993b; 1994) employed attribution theory as a mechanism to explain how religious di fferences shape attitudes toward crime and justice. Grasmick et al. (1992) examined the effect of Protestant fundamentalism on the retributive sentencing philosophy. Consistent with their expect ations, they found that fundamentalists were more likely to support the retributive doctrine of punishment than both liberal/moderate Protestants and individuals with no religious affiliati on. Although this study did not provide a direct test of attribution theory, it was an important pre-cursor, as it provided a link between religion, particularly fundamentalism, and retributive sentiment. Recall that religion may serve as an important group from which individua ls adopt their attitudes and model their behaviors. Given the inherently mora l nature of religious groups, the socialization process that occurs within them, particularly the faith groups that espouse proscriptive doctrines, may be an important source of the internalization of retributiv e emotions upon which attitudes toward issues such as capital punishment are based (Vidmar and Miller, 1980). Applying this notion of individual socialization to attribution theory, Grasmick, Bursik, and Blackwell (1993b) contended that, b ecause fundamentalist Protestants believe that behaviors which go against Biblical teachings are a reflection of bad character, they should be more likely to possess a dispositional than a situational attribution style. Am ong whites, their hypothesis was supported, as Biblical literalness was significantly and directly related to the tendency to attribute juvenile crime to dispositional causes. In other words, the more literally individuals interpreted the Bible, the more likely they were to attribute juvenile crime to offender character. Dispos itional attribution style, in turn, was significantly and positively related to punitiveness toward juveniles.

PAGE 41

36 Grasmick et al. (1993a) further considered the role of attribution theory when they hypothesized that, because of their tendency to adopt a dispos itional attribution style, evangelical/fundamentalist Protestants would be more likely to support punitive sa nctions for offenders than would members of other religious faith groups. Indeed, they found that Protes tant fundamentalists were significantly more punitive than liberal/moderate Protestants on four out of five punitive justice items, including both adult and juvenile death penalty support, support for ha rsher courts, and support for stiffer laws. Grasmick and McGill (1994) provided the first co mplete and direct app lication of attribution theory to the religion-pun itiveness relationship. They found th at, among whites, Biblical literalness positively affected the likelihood of attributing juvenile crime to di spositional causes. Dispositional attribution, in turn, had a significant positive effect on the desire to puni sh juvenile offenders. Further, when controls were included, attribution style accounted for half of th e effect of Biblical literalness on punitiveness, and the remaining effect fell to nonsignificance. Thus, attr ibution style was, as hypothesized, the intervening variable. In other words, Biblical literalness affected individuals attribution styles, which then affected punitiveness. It is important to note, however, that the effect of Biblical literalness fell to non-significance when the analysis was run for blacks. Thus, fundamentalist Protestant beliefs do not appear to generate the same punitiv eness among blacks as they do whites (Grasmick and McGill, 1994). Although attribution theory provides an explanation for how religion shapes arguments for and against capital punishment, it largely fails to explore how a forgiving and tolerance-oriented theological emphasis may also lead to the possession of a situational attribution style, which, in turn, should shape attitudes unfavorable to capital punishment. Appl egate et al. (2000) focused on the link between progressive theological emphases and correctional attitudes. Their analyses indicated that religiosity had a significant effect on correctional attitudes. Specifi cally, individuals who possessed a more forgiving theological emphasis were more supportive of rehabilita tion for offenders, less likely to believe their local courts were too lenient, and advocated less punitive approaches to criminals in general and to a hypothetical offender. They were also significantly less supportive of capital punishment. These relationships remained when contro ls were included (Applegate et al., 2000). Further, Biblical literalness and perceptions of a punitive God were significant and positively associated with punitive attitudes and

PAGE 42

37 negatively related to rehabilitation support. In ot her words, individuals w ho possessed fundamentalist religious beliefs were more likely to favor punitive measures and oppose rehabilitative ones than nonfundamentalists. These relationships retained thei r significance when demographic and other relevant controls were included. Although Applegate et al. (2000) add the dimension of forgiving theological emphasis, they fail to fully explore attribution style as a possible linkage between theological emphasis and the adoption of various philosophies of punishment. Previous research has suggested that Biblical literalness plays an important role in individuals attribution styles, with higher degrees of Biblical literalness associated with more dispositional attribution styles (Grasmick et al., 1993b). It also indicates that attribution style is a consistent predictor of punishment philosophy, with possessi ng a dispositional attribution st yle predicting support for more punitive policies (including capital puni shment) and possessing a situationa l attribution style predicting a more rehabilitative approach to offenders (Cullen et al., 1985; Young, 1991; Grasmi ck et al., 1993a). It neglects, however, to systematically specify the entire causal process by which religion and perceptions of crime causation affect attitudes toward capital punishment The present study provi des a direct test of attribution theory, specifically improving upon extant research by delineating the process through which attitudes toward the death penalty are shaped by an individuals theological emphasis. Applegate et al. (2000) found that both theological emphasis and Biblical literalness were associated with attitudes toward correctional treatment, including attitudes toward capital punishment. They did not, however, examine the possible link between theological emphasis and Biblical literalness. There is extensive evidence that different religious orientations produce different images of God, with more conservative religions adopting a more punitive image of God and more liberal religions perceiving God as more understanding and forgiving. It follows, then, that individuals pe rceptions of God will affect their interpretation of His word. People who perceive God as rigid and condemning should be more likely to adopt a literal interpretation of the Bible, while peopl e who perceive God as more malleable and forgiving should be more likely to allow for a more interpretative approach to the Bible. The latter group of individuals might, for instance, believe that parts of the Bible are literal, but other parts are symbolic or figurative, have contextualized meanings, etc. More formally stated, 6a. There will be a positive and direct relations hip between a punitive theological emphasis and a

PAGE 43

38 more literal interpretation of the Bible. 6b. Conversely, a more progressive theological emphasis will be negativ ely associated with Biblical literalness. Consistent with Grasmick et al.s (1993b) findings, we also asse rt that Biblical literalness will affect individuals attribution styles. We predict the following with regard to the relationship between Biblical literalness and attribution style: 7a. A more literal interpretation of the Bible will be directly predictive of a dispositional attribution style for adult and juvenile offenders. 7b. Conversely, a less literal interpretation of the Bible will be predictive of a more situational attribution style for adult and juvenile offenders. Keeping with the propositions of attribution theory and in accordance with the findings of Cullen et al. (1985), Young (1991), and Grasmick et al. (1993a), which indicated that there was relationship between attribution style and individuals positions on punishment philosophy our correctional system should adopt, we establish the following hypotheses: 8a. A more dispositional attribution style will have a direct positive effect on possessing a more punitive punishment philosophy for a dult and juvenile offenders; and 8b. Conversely, a more situational attribution st yle will be directly and negatively related to possessing a more punitive sentiment toward adult and juvenile offenders. Previous research has examined various indi cators of punitiveness and capital punishment as two separate dependent variables. Due to the conceptual linkages between Biblical literalness and attitudes toward the death penalty, we model them as distin ct from one another. Once we have established a relationship between attribution styl e and correctional punishment philos ophy, we are then in a position to understand why Biblical literalness affects capital punishment supportit is because individuals who possess a more punitive punishment philosophy tend to favor capital punishment more than individuals who adopt a more reformative approach to adult and juve nile offenders. Accordingly, the final path in the processual model suggests: 9a. A more punitive punishment philosophy will be directly associated with capital punishment support; and 9b. Conversely, a more rehabilitative philosophy to ward the correctional treatment of adult and juvenile offenders will be directly and nega tively related to support for the death penalty. Early research on religion and attitudes toward crime and punishment issues was limited to differences between faith groups. Later research exte nded the concept of religion to include measures of personal religiosity (e.g., religious commitment and i nvolvement, religious beliefs) and included measures

PAGE 44

39 of Biblical literalness. These developments were important, because they provided a clearer and more accurate picture of why religion aff ects death penalty attitudes. Despite this progress in the research literature surrounding religion and punitive attitudes, the re search thus far has been largely atheoretical. Attribution theory appears to offer a substantive explanation that demonstrates the process of how religion affects punitive attitudes, including attitudes toward cap ital punishment; however, it fails to explicate the entire causal process through which relig ion affects death penalty attitudes. Based on the body of research that has evolved, we established a comprehensive causal model for this relationship. Summarily, we believe that theological emphasis will affect the degree to which individuals literally interpret the Bible; that Biblical literalness will determine individuals a ttribution styles (Grasmick et al., 1993a; Grasmick et al., 1993b; Grasmick and McGill, 1994); that individuals attribution st yles will predict the punishment philosophies that they support and oppose (Cullen et al., 1985; Grasmick and McGill, 1994); and that individuals punishment philos ophies will affect their attit udes toward capital punishment.

PAGE 45

40 CHAPTER 4 METHODOLOGY Sample and Procedure The data for this investigation were obtained from a survey administered to individuals called for jury service in Hillsborough County, Florida during the fall of 1999. Th e jury pool consists of a random selection of Hillsborough Count y residents18 and over with a drivers license or Florida identification. Groups that are ineligible to serve include those under 18 (Att.Gen. 1973.Laws 1973, c. 73-21, the Adult Rights Law, 743.07), convicted felons whose civil rights have not been restored [Fla. 1950, F.S.A. .04(2), 40.02(5), 40.07(2), 913.03, F.S.A. Const. art. 16, ], those individuals under prosecution for any crime [Fla. App. 1974, Wests F.S.A. .01, 40.07, 40.07(1)], people no longer residing in Hillsborough County (C.A.Fla. 1978, Wests F.S.A. .01 et seq.; U.S.C.A. Const.Amends. 6, 14), and non-U.S. citizens (C.A.Fla. 1975, 28 U.S.C.A. 1865). Other groups who may decline to serve include expectant mothers and unemployed parents with full custody of a child under six [Fla. 1984; Wests F.S.A. .01(1)], full-time law enforcement officers and investigative personnel [Fla. 1984; Wests F.S.A. .013(2)], and individuals over 70 years of age (Op.Atty.Gen., 070-115, Aug. 25, 1971). Despite the potential biases in under-representing these groups of people, the jury pool is, perhaps, the best readily available random sample of adults. There are also other advantages associated with this population. For example, many are called to report for jury duty, but end up waiting in the courthouse for part or all of the day. These individuals have been called upon to perform a civil duty. They have agreed to sacrifice their time, while tolerating other inconveniences (e.g., waking up at an early hour, traffic, parking, etc.), as well. Yet despite their frustrations, when they arrive at the courthouse, they sit and wait for, minimally, two hours; some wait all day. Understandably, these potential jurors are bored and exasperated. The administration of a lengthy questionnaire related to issues of crime and justice, then, is convenient as a time filler and beneficial in that participation in this study at least serves as a proxy for the fulfillment of

PAGE 46

41 citizenship duties. This situation also facilitates the development of an extensive and comprehensive questionnaire, as there were minimal time constraints on the respondents. Once it had been established that the jury pool would provide an appropriate sample for survey research, several meetings between researchers at the University of South Florida (USF) and the Prosecutors Office, the Public Defenders Office, and the staff at the Courthouse for the Thirteenth Judicial Circuit in the State of Florida ensued. During these meetings, details concerning administration and scheduling were addressed, a nd drafts of the questionnaire were exchanged for suggestions and comments. A final instrument was then submitted to the Institutional Review Board (IRB) at USF. The survey underwent an expedited review process and permission to proceed wa s granted. A final copy of the questionnaire may be found in Appendix A, and an informed consent form is located in Appendix B. A copy of the codebook for the survey may be obtained from the author. The jury pool is called on Monday and Wednesday mornings of every week. Data collection occurred over two different weeks in a one-month period. Following the first week of data collection, a few weeks elapsed prior to the second survey administration. Three researchers distributed questionnaires and informed consent forms on every chair in the central courtroom prio r to 7:30 a.m. when the potential jurors arrived. On the first day, however, a smaller courtroom was used in addition to the central courtroom, which increased the number of potential respondents by approximately 60 people. The researchers circulated among the potential jurors to assist with questions and collect the survey instruments. One researcher also made an announcement on each of the days detailing the purpose of the study and reminding the respondents that their participation was voluntary, their identities would remain anonymous, and that the information they provided would be held with the strictest assurance of confidentiality. The four days of data collection generated a total of 655 surveys. For the analysis in the present study, a sub-sample was selected that included only those subjects who were black, white, or Hispanic. All others (American Indian or Alaskan Natives and Asia ns) were deleted from the analysis. This selection process resulted in a final sample size of 636 respondents. The demographic characteristics in the present sample of poten tial jurors in Hillsborough County were largely reflective of the overall Hillsborough County population. With regard to gender, 51 percent of Hillsborough County residents were ma les, while 49 percent were females, according to the 1990 Census

PAGE 47

42 Report. Of the potential jurors surveyed in this study, 54.1 percent were females and 45.9 percent were males. Race and ethnicity were meas ured by separate items that refl ect similar measures in the U.S. Census. According to the Demographics USA C ounty Edition, 1998, 82.3 percent of residents in Hillsborough County, Florida were Cau casian and 15.2 percent were African American. In our sample, a similar distribution emerged: 79.9 percent of respondents identified themselves as Caucasian and 11.9 percent as African American. In Hillsborough County, Hispanics comprise 2.4 pe rcent of the population; in the present sample, 10.4 percent identify themselves as Hispanic. Our sample was similar to Hillsborough County reside nts in their level of educational attainment: 18 percent of Hillsborough County residents have graduated from high sc hool; 26 percent of the potential jurors in our sample were high school graduates. Thirty-four pe rcent of Hillsborough County residents attended college, but did not graduate.7 percent of our sample attained this level of education. Similar percentages were college graduates (24% of Hillsbor ough County residents, 24.4% of our sample) and had a graduate or professional degree (11% of Hills borough County residents, 9. 4% of our sample). Although it was not possible to compare our sa mple to Hillsborough County residents across all income categories because they were measured with different categories and with different definitions of income (effective buying income or after tax income in the Demographics USA Report as opposed to annual household income in our survey), the income level, as determined by the median incomes appear to be similar. The median after tax income for H illsborough County residents was $40,900, while the median income category for the jury pool in our sample was that category that ranged from $40,000 to $45,000. Unfortunately, although our survey instrument in cluded several other demographic characteristics, they were unavailable for Hillsborough County residents. Based on the da ta that were available, though, our sample appears to be an accurate portrayal of Hillsborough County overall. Keep in mind that although information on ages of both populations are available, it would be inappropriate to draw any comparisons, because the jury pool is limited to individuals ages 18 and over. Measurement of Variables Capital Punishment Attitudes The dependent variable, attitudes toward capital punishment, is measured with items reflecting general and circumstantial support for/opposition to capital punishment fo r both juveniles and adults.

PAGE 48

43 Although most previous studies have used rather crude measures of death penalty support (Grasmick et al., 1993a; Grasmick et al., 1993b; Britt, 1998; Applegate et al., 2000), a more in-depth operationalization is used in the present study for three reasons. First, support for capital punishment may vary by type of offense and characteristics of the offender, such as age (Young, 1991). Second, since public opinion is a legitimate basis for the implementa tion of the death penalty, it is imperative to understand all of the qualitative distinctions between the conditions unde r which support may vary (Young, 1991). Finally, because prior research has established that support for the death penalty substantia lly declines when general support is distinguished from support in concrete situ ations (Ellsworth and Ross, 1983), respondents in the present study were asked to indicate their support for capital punishment across a variety of Likert-type statements. Response options ranged from 1 = strongly disagree to 5 = strongly agree. The two general items state, Generally speaking, I support the death pe nalty for adults, and Generally speaking, I support the death penalty for juveniles. The mean levels of support on these two items were 3.87 and 2.80, respectively. Beyond these and in keeping with the recommendations of Ellsworth and Ross (1983) and Durham (1988), circumstantial support for the death penalty was measured for both adults and juveniles across six specific offense-types. Th ey included: the murder of a criminal justice employee (mean = 3.59, 3.22, for adults and juveniles, respectively), a planned murder (mean = 3.95, 3.50), the murder of a child (mean = 3.95, 3.39), an unplanned murder (mean = 3.02, 2.83), serial rape (mean = 3.37, 3.06), and child molestation (mean = 3.27, 2.93). Based on the mean levels of support for these offender and offense-specific items, it is clear that respondents largely favor or hold ne utral attitudes toward capital punish ment. In fact, slight opposition (mean support less than 3.0) appeared in only 3 of the 14 items: general juvenile support and support for capital punishment for juveniles who commit an unplanned murder or who molest a child. Consistent with previous research, though, respondents degree of capital punishment support varied across these different circumstances. Table 2 presents the distribution of support for capital punishment on these 14 items. As Table 2 reflects, this sample of Hillsbor ough County resident s exhibited a higher level of capital punishment support for adults than juveniles across every offense type, as well as on a general level. Moreover, respondents indicated greatest support for the death penalty for adults on a general level (66%) and when adults committed planned murder (66.8%). The highest level of juve nile death penalty support

PAGE 49

Table 2. Percentage Distribution of Capital Punishment Attitudes by Specific Offenders and Offense Types Variable SD 5 D 4 N 3 A 2 SA 1 Mean Adults-general support 8.3% 3.9% 21.7% 24.2% 41.8% 3.87 Adults-murder of cj employee 8.3% 4.4% 33.0% 28.3% 225.9% 3.59 Adults-planned murder 7.9% 1.7% 23.6% 20.9% 45.9% 3.95 Adults-murder of a child 7.9% 1.7% 25.9% 16.8% 47.6% 3.94 Adults-unplanned murder 7.5% 15.1% 52.5% 17.6% 7.2% 3.02 Adults-multiple rapes 9.1% 12.1% 35.5% 18.9% 24.4% 3.37 Adults-child molestation 8.8% 15.1% 37.6% 17.1% 21.4% 3.27 Juveniles-general support 22.3% 17.3% 30.5% 17.8% 12.1% 2.80 Juveniles-murder of cj employee 8.8% 11.2% 45.1% 18.6% 16.4% 3.23 Juveniles-planned murder 8.3% 8.2% 36.0% 20.4% 27.0% 3.49 Juveniles-murder of a child 8.3% 8.2% 41.7% 19.5% 22.3% 3.39 Juveniles-unplanned murder 10.2% 19.0% 53.1% 12.7% 4.9% 2.83 Juveniles-multiple rapes 10.2% 16.0% 44.8% 15.9% 13.1% 3.06 Juveniles-child molestation 10.7% 18.7% 47.6% 12.4% 10.5% 2.93 44

PAGE 50

45 was for those juveniles who committed planned murder (47.4%); respondents al so reported a relatively favorable attitude for juveniles who killed other kids (41.8%). For both adults and juveniles, the greatest proportion of respondents opposed capital punishment for unpl anned murder, rape, and child molestation. Observe, however, that the highest opposition to juvenile capital punishment was the general item (39.6%).The 14 death penalty opini on items were collapsed into one 14-item additive capital punishment support scale. The range of scores on this scale was 14 (opposed capital punishment under every circumstance) to 70 (supported capital punishment under every circumstance). The mean for the sample was 46.76, reflecting a moderate level of support for cap ital punishment. Interestingly, despite conceptual distinctions in capital punishment a ttitudes by offender and/or offense ty pe, the data reveal no empirical benefit to separating the items. Indeed, when the 14 items were entered into a principal components factor analysis, the results suggested that a single factor solution best fit the data. Specifically, according to the Kaiser Rule, three factors had eigenvalues greater than 1.00 (values = 8.55, 1.39, and 1.27 of factors 1,2, and 3, respectively). When the Sc ree Discontinuity Test was applied, however, the greatest difference in the eigenvalues occurred between the first two factors, suggesting the presence of only one meaningful factor. The 14 items all loaded well, with loadings ranging from .61 (general support for juvenile capital punishment) to .87 (support for the death penalty for juveniles who committed a planned murder). See Appendix C for a breakdown of the item-to-scale co rrelations. Finally, this 14-item additive scale produced a Cronbachs Alpha reliability coefficient of .95, indicating a highly reliable scale. Measures of Religion Recall that the first set of hypotheses to be test ed addresses the independent and interactive effects of race and religion on capital punishment attitudes. Thus, the independent variables in these analyses are religious conservativism, measured by current denominational affiliation; religiosity; and race. Religious Conservativism Faith groups serve as membership groups that provide a context for which individuals may evaluate themselves in relation to the dictates of their faiths. Potential jurors in this sample were asked to identify their faith groups from one of the following re sponse options: Catholic, Jewish, Protestant, none, or other. Given the prevalence of different Protestant s ects, as well as their doctrinal differences, individuals

PAGE 51

46 who identified themselves as Protestants were also as ked to specify to which denomination they belonged: Baptist, Episcopalian, Lutheran, Methodist, Presbyterian, Fundamentalist, Pentecostal, or Other. Almost half (49.8%) of the sample reported that they were Protestants, while nearly a quarter (23.7%) indicated that they were Catholics. Non-affiliates comprised 11.9 percent of the participants, followed by others (7.9%), and Jews (2.5%). The remaining 4.7 percent were missing. Protestants, in turn, indicated the following denominational distribution: Baptist (23.0%), Episcopalian (4.1%), Lutheran (3.1%), Methodist (8.5%), Presbyterian (4.6%), Fundamentalist (1 .1%), Pentecostal (3.8%), and Other (2.7%). Because of the small percentage of blacks and Hispanics in each faith group, we trichotomized faith group into members of liberal, moderate, and conservative denominations to measure religious conservativism. Consistent with previous categorizations (Smith, 1990), liberal faiths included the Protestant denominations Episcopalian and Presbyteri an, as well as individuals who are not members of any faith group. Moderate religions included Catholic Jewish, and the Protestant denominations Lutheran and Methodist. Finally, conservative faiths were comprise d of the Protestant sects Baptist, Fundamentalist, Pentecostal, and Other Protestant faiths. Non-Protes tant members of other faith groups and missing data were coded as moderate, as this category represente d the mean, the median, and the mode (Smith, 1990). Respondents received a score of 1 if they were members of a liberal religion, 2 if they were part of a moderate religion, and 3 if they identified themselves as members of a conservative faith group. Of the 636 respondents, 129 (20.3%) were categorized as religious liberals, 319 (50.2%) as religious moderates, and 188 (29.6%) as religious conservatives. Given the findings of previous studies (Grasmick et al., 1992; Grasmick et al., 1993a; Britt, 1998) regarding the significant differences between conserva tive Protestants and members of other faith groups, we employed a third measure of religious conservativ ism, which makes a dichotomous distinction between conservative Protestants (i.e., Baptists, Fundamentalists, Pentecostals, and Other Protestants, all coded 1) and members of all other faith groups and non-affiliate s (coded 0). A frequency di stribution showed that 191 (30%) of the Hillsborough County respondents we re members of a conservative Protestant denomination, and 445 (70%) were members of a different faith group or non-affiliatesi.e., nonconservative Protestants.

PAGE 52

47 Religiosity A single 9-item religiosity scale was created to assess the effect of religiosity on death penalty attitudes. The 9 indicators tapped different, but related aspects of religious salience, commitment, and involvement. They are: 1. In a typical month (4 weeks long), how many times do you attend worship services? This item contained ordinal level response options. 2. How many church-related groups or organiza tions do you belong to (include Bible studies, youth groups, educative activitie s, choral groups, etc.)? This item was an interval level measure, with responses being the number of groups to which respondents belonged. 3. How often do you take part in the activitie s and organizations of a church or place of worship, other than attending services? This item was an ordinal measure, with response options ranging from 0 (never) to 6 (more than once a day). 4. In a typical month, how often do you pray outside of worship services (do not include table blessings)? This, too, was an ordinal measure. Possible responses ranged from 0 (never) to 4 (once a day). 5. In a typical month, how often do you have quiet time that focuses on religious or spiritual development (e.g., read religious material, meditate, etc.)? The response categories were the same as they were for the item above. 6. Religion is a very important part of my life. (coded 1 = strongly agree to 5 = strongly disagree) 7. I would describe myself as very religious. (coded 1 = strongly agree to 5 = strongly disagree) 8. When I have decisions to make in my everyday life, I usually try to find out what God wants me to do. (coded 1 = strongly agree to 5 = strongly disagree) 9. Religion influences how I live my life. (coded 1 = strongly agree to 5 = strongly disagree) The composition of this scale was based on a compilation of items taken from related literature. The first item, which is indicative of respondents attendan ce at worship services, has been widely used as a measure of religious commitment/involvement (Bock, Cochran, and Beeghley, 1987; Cochran, Beeghley, and Bock, 1988; Beeghley, Bock, and Cochran, 1990; Clarke, Beeghley, and Cochran, 1990; Cochran and Beeghley, 1991; Cochran, Beeghley, and Bock, 1992; Young, 1992; Cochran, Chamlin, Beeghley, Harnden, and Blackwell, 1996). Pa rticipation in religious organizations (items 2 and 3) has also been a prevalent measure of religio sity, although much of the previous research (Bock et al., 1987; Cochran et al., 1988; Beeghley et al., 1990; Clarke et al., 1990; Cochran and Beeghley, 1991) has dichotomized it into whether or not responde nts were members of a religious organization. The present study expanded this dimension of commitment/involvement to include both the number of

PAGE 53

48 organizations, as well as the amount of time partic ipants spent engaged in activities related to the organization(s). Young (1992) developed the concept of devoti onalism (analogous to religious commitment), which he operationalized with concepts related to attendance at services, frequency of prayer, and frequency with which individuals read the Bible at home. In agreement with Youngs conceptualization of devotionalism, two items were developed to ascertain the frequency with which respondents engaged in prayer and quiet time, (w hich included, but was not limited to reading the Bible at home). Finally, the last four statements in the religiosity scale assess the salience of religion in respondents daily lives. These are a replication of a scale created by Grasmick and his colleagues (Grasmick et al., 1993a; Grasmick et al., 1993b; Grasmick and McGill, 1994) and validated by their research, as well as in subsequent studies (e.g., Curry, 1996; Applegate et al., 2000). Items were standardized into z-scores, as they were measured on different metrics. These 9 standardized items were entered into principal components factor analysis, which suggested that a single factor solution best fit the data. The Kaiser Rule demonstrated two eigenvalues greater than 1 (5.56 and 1.17 on factors 1 and 2, respectively); however, an application of the Scree Discontinuity Test revealed the presence of only one substantive factor, as the greatest difference between the eigenvalues was between the first two values. F actor loadings ranged from .63 (number of churchrelated groups to which respondents belonged) to .87 (the degree to which respondents agreed that they turned to God for guidance in their daily decisions) (see Appendix D). The linear composite of these 9 standardized items had an alpha reliability of .92. Race Religious influences do not operate in a vacuum Numerous secular influences may enhance or attenuate the relationship between religion and a ttitudes toward capital punishment. Race is one of these influences, as racial differences in both socio-political views and religious affiliations have appeared in previous studies (Young, 1991; Barkan and Cohn, 1994; Britt, 1998). Race, in the present study, was measured by as king respondents to indicat e their race according to the following racial categories used by the federal government: American Indian or Alaskan Natives, Asians, Blacks or African Americans, Native Hawaiia n or other Pacific Islanders, and Whites. Given

PAGE 54

49 the unique role that religion plays in Hispanic cultures in America, we also asked one question to determine respondents ethnic origins: Are you of Hispanic, Latino, or Spanish origin (includes a person of Cuban, Mexican, Puerto Rican, South or Ce ntral American, or other Spanish culture or race, regardless of race)? As race was, the description of ethnicity was based on th at used by the federal government. The variable used in the analysis was a dummy variable, coded 0 for non-Hispanic whites and 1 for non-whites. The non-white category included only Blacks and Hispanicsthe remaining racial groups were not considered, because there were t oo few members of these groups to analyze. In fact, this was also the case with Blacks and Hispanic sseparated, there were too few members of each group to provide meaningful results Although it may be ideal to examine the unique effects of race and ethnicity, the di chotomizing of race/ethnicity is conceptually sound, as both Blacks and Hispanics are oppressed populations in the United States, and religion plays a prevalent and unique role in both cultures. Given their experiences with oppression, we would expect that both groups would hold more liberal social and political views than whites. Of the 636 respondents in this investigation, 18 percent were non-white, and 82 per cent were white. Table 3 provides the correlation matrix for the death penalty scale, the religion variables, the race variable, and the control variables. Table 4 provides descripti ons of the religion variables and the race variable, including their means and standard deviations. Measures from Attribution Theory Following the initial hypotheses, which are expected to establish a relationship between religion and attitudes toward the death penalty, it is necessa ry to determine the process through which religion affects capital punishment attitudes. As we discusse d previously, attribution theory provides a useful framework from which to analyze th e religion-capital punishment attitude s relationship. Keeping with the tenets of attribution theory, we developed a path model that reflects the indirect relationship between religion and death penalty attitudes. First, we measure religion by theological emphasis, rather than faith group or religiosity, as it is likely that the theological emphasis of a faith gr oup translates into punitive or progressive secular attitudes. We developed a 6-item theological emphasis scale; the first 3 items listed below were taken

PAGE 55

50 Table 3. Correlation Matrix of Death Penalty Scale Variables, Religion Variables, the Race Variable, and Control Variables sex race age education income ideology abortion sexroles Sex 1.00 -0.04 0.04 0.03 0.09* 0.01 0.01 0.17* Race 1.00 0.13* -0.04 -0.07 -0.13* 0.04 0.03 Age 1.00 -0.02 0.06 0.10* 0.05 0.19* Education 1.00 0.36* 0.01 -0.13* -0.16* Income 1.00 0.09* -0.11* -0.06 Ideology 1.00 0.33* 0.24* Abortion 1.00 0.37* Sexroles 1.00 Socwlfre Cjscale No religion Jewish Catholic Episcopalian Presbyterian Methodist Lutheran Baptist Fundamentalist Rel. tradition Cons. Prot Religiosity Dpscale Ad-gen. dp supp Ad-murder cj emp Ad-plan murder Ad-murder child Ad-unplan murder Ad-rape Ad-molest child Juv-gen. dp supp. Juv-murder cj emp Juv-plan murder Juv-murder child Juv-unplan murder Juv-rape Juv-molest child Continued on next page

PAGE 56

51 Table 3 continued socwlfre cjscale no religion Jewish Catholic Episcopalian Presbyterian Sex 0.19* -0.07 0.04 0.03 0.02 -0.08* -0.04 Race -0.21* -0.21* 0.03 -0.05 0.11* -0.07 -0.05 Age 0.01 0.13* -0.18* 0.10* -0.03 0.01 0.08* Education 0.09* 0.12* -0.04 0.10* 0.07 0.12* 0.09* Income 0.15* 0.16* -0.04 0.11* 0.06 0.05 0.08* Ideology 0.27* 0.19* -0.18* 0.06 -0.00 0.01 0.03 Abortion 0.13* 0.07 -0.19* -0.16* 0.03 -0.09* -0.06 Sexroles 0.21* -0.11* -0.10* -0.08* -0.08 -0.04 -0.01 Socwlfre 1.00 0.16* 0.02 -0.06 -0.01 0.08* 0.01 Cjscale 1.00 -0.07 0.03 0.05 0.08* 0.08* No religion 1.00 -0.06 -0.21* -0.07 -0.08* Jewish 1.00 -0.09* -0.03 -0.04 Catholic 1.00 -0.11* -0.12* Episcopalian 1.00 -0.04 Presbyterian 1.00 Methodist Lutheran Baptist Fundamentalist Rel. tradition Cons. Prot Religiosity Dpscale Ad-gen. dp supp Ad-murder cj emp Ad-plan murder Ad-murder child Ad-unplan murder Ad-rape Ad-molest child Juv-gen. dp supp. Juv-murder cj emp Juv-plan murder Juv-murder child Juv-unplan murder Juv-rape Juv-molest child Continued on next page

PAGE 57

52 Table 3 continued Methodist Lutheran Baptist Fund. Rel. tradition cons. Prot religiosity Sex -0.05 0.01 0.02 0.04 0.04 0.04 -0.14* Race -0.06 -0.09* -0.07 -0.03 -0.03 -0.08* 0.07 Age 0.05 0.06 0.07 0.08* 0.14* 0.11* 0.21* Education 0.06 0.05 -0.16* -0.05 -0.16* -0.18* -0.05 Income 0.08* 0.02 -0.08* -0.11* -0.11* -0.14* -0.06 Ideology -0.01 -0.03 0.16* 0.14* 0.22* 0.22* 0.28* Abortion -0.05 -0.11* 0.15* 0.27* 0.32* 0.29* 0.47* Sexroles -0.04 -0.01 0.16* 0.14* 0.21* 0.22* 0.21* Socwlfre 0.04 0.02 0.01 0.01 -0.02 0.02 -0.11* Cjscale 0.03 0.06 -0.06 0.02 -0.05 -0.04 0.08* No religion -0.11* -0.07 -0.20* -0.10* -0.57* -0.24* -0.40* Jewish -0.05 -0.03 -0.09* -0.05 -0.02 -0.11* -0.08* Catholic -0.17* -0.10* -0.30* -0.16* -0.07 -0.37* -0.05 Episcopalian -0.06 -0.04 -0.11* -0.06 -0.31* -0.13* -0.05 Presbyterian -0.07 -0.04 -0.12* -0.06 -0.34* -0.14* -0.03 Methodist 1.00 -0.05 -0.16* -0.09* -0.04 -0.20* 0.08* Lutheran 1.00 -0.10* -0.05 -0.02 -0.12* -0.04 Baptist 1.00 -0.15* 0.70* 0.83* 0.21* Fundamentalist 1.00 0.24* 0.34* 0.43* Rel. tradition 1.00 0.83* 0.42* Cons. Prot 1.00 0.32* Religiosity 1.00 Dpscale Ad-gen. dp supp Ad-murder cj emp Ad-plan murder Ad-murder child Ad-unplan murder Ad-rape Ad-molest child Juv-gen. dp supp. Juv-murder cj emp Juv-plan murder Juv-murder child Juv-unplan murder Juv-rape Juv-molest child Continued on next page

PAGE 58

53 Table 3 continued dpscale ad-gen dp ad-murder cj ad-plan ad-murder ad-unplan ad-rape supp employee murder child murder Sex 0.09* 0.08 0.08 0.05 0.04 0.00 0.03 Race -0.14* -0.16* -0.18* -0.12* -0.13* -0.10* -0.12* Age 0.01 -0.01 -0.04 -0.05 -0.03 0.06 0.02 Education -0.14* -0.01 -0.01 -0.03 -0.04 -0.13* -0.14* Income 0.07 0.08* 0.13* 0.07 0.08 0.04 0.01 Ideology 0.17* 0.14* 0.16* 0.13* 0.14* 0.09* 0.09* Abortion -0.11* -0.16* -0.07 -0.11* -0.10* -0.09* -0.10* Sexroles 0.04 -0.03 -0.00 -0.03 -0.03 0.01 0.02 Socwlfre 0.22* 0.20* 0.23* 0.21* 0.19* 0.13* 0.15* Cjscale 0.06 0.16* 0.13* 0.13* 0.14* 0.09* 0.01 No religion 0.02 0.03 0.02 0.06 0.05 0.02 0.04 Jewish -0.02 -0.03 -0.04 0.01 0.03 -0.02 0.03 Catholic -0.03 -0.02 -0.06 -0.07 -0.04 -0.05 -0.00 Episcopalian -0.00 0.03 0.03 0.03 -0.02 0.02 -0.01 Presbyterian -0.03 -0.04 -0.03 -0.02 -0.02 -0.04 -0.02 Methodist -0.02 0.00 0.01 0.01 0.01 -0.02 -0.04 Lutheran 0.05 0.08 0.05 0.07 0.07 0.07 0.06 Baptist 0.12* 0.09* 0.11* 0.05 0.05 0.12* 0.03 Fundamentalist -0.04 -0.04 -0.02 -0.01 -0.04 -0.01 -0.05 Rel. tradition 0.06 0.02 0.05 -0.00 -0.01 0.07 -0.02 Cons. Prot 0.09* 0.06 0.09* 0.04 0.03 0.11* -0.00 Religiosity -0.09* -0.10* -0.10* -0.11* -0.13* -0.00 -0.06 Dpscale 1.00 0.71* 0.85* 0.81* 0.82* 0.75* 0.78* Ad-gen. dp supp 1.00 0.67* 0.65* 0.65* 0.50* 0.50* Ad-murder cj emp 1.00 0.83* 0.83* 0.65* 0.60* Ad-plan murder 1.00 0.93* 0.55* 0.65* Ad-murder child 1.00 0.57* 0.66* Ad-unplan murder 1.00 0.52* Ad-rape 1.00 Ad-molest child Juv-gen. dp supp. Juv-murder cj emp Juv-plan murder Juv-murder child Juv-unplan murder Juv-rape Juv-molest child Continued on next page

PAGE 59

54 Table 3 continued ad-molest juv-gen dp juv-murder cj juv-pl an juv-murder juv-unplan juv-rape juvchild support employee murder child murder molest Sex 0.00 0.20* 0.14* 0.11* 0.09* 0.05 0.04 0.00 Race -0.07 -0.07 -0.16* -0. 11* -0.11* -0.05 -0.08* -0.05 Age -0.04 0.06 0.01 0.00 0.03 0.11* 0.03 0.05 Education -0.17* 0.09* 0.09* -0.10* -0.15* -0.18* -0.19* -0.23* Income -0.03 0.10* 0.12* 0.09* 0.07 0.04 -0.03 -0.06 Ideology 0.05 0.14* 0.18* 0.19* 0.21* 0.10* 0.12* 0.10* Abortion -0.08* -0.12* -0.07 -0.07 -0.05 -0.07 -0.08* -0.06 Sexroles 0.02 0.09* 0.05 0.05 0.07 0.03 0.09* 0.07 Socwlfre 0.11* 0.21* 0.25* 0.23* 0.19* 0.11* 0.12* 0.06 Cjscale -0.02 -0.04 0.07 0.08 0.04 0.02 -0.06 -0.05 No religion -0.01 -0.01 0.00 -0.00 0.02 -0.02 0.02 -0.04 Jewish -0.02 -0.05 -0.03 -0.03 -0.02 -0.02 0.00 -0.01 Catholic 0.05 0.01 -0.10* -0.02 -0.06 -0.04 0.01 0.05 Episcopalian -0.03 -0.02 0.02 -0.01 -0.00 -0.02 -0.00 -0.03 Presbyterian -0.04 -0.04 -0.03 -0.03 -0.00 -0.03 -0.02 -0.01 Methodist -0.03 -0.02 0.02 0.00 -0.02 0.00 -0.07 -0.06 Lutheran 0.03 0.02 0.01 0.02 0.02 0.00 0.03 -0.01 Baptist 0.07 0.12* 0.14* 0.11* 0.14* 0.14* 0.06 0.08* Fundamentalist -0.05 -0.04 -0.02 -0.01 -0.02 -0.02 -0.04 -0.04 Rel. tradition 0.05 0.08* 0.08 0.08 0.07 0.10* 0.02 0.06 Cons. Prot 0.04 0.09* 0.11* 0.10* 0.11* 0.12* 0.04 0.05 Religiosity -0.07 -0.09* -0.08* -0.06 -0.08* 0.02 -0.05 -0.04 Dpscale 0.74* 0.62* 0.84* 0.87* 0.86* 0.73* 0.79* 0.74* Ad-gen. dp supp 0.42* 0.54* 0.55* 0.56* 0.53* 0.41* 0.40* 0.32* Ad-murder cj emp 0.55* 0.40* 0.75* 0.75* 0.71* 0.52* 0.50* 0.43* Ad-plan murder 0.55* 0.28* 0.62* 0.75* 0.67* 0.40* 0.48* 0.40* Ad-murder child 0.58* 0.28* 0.61* 0.72* 0.70* 0.41* 0.49* 0.44* Ad-unplan murder 0.53* 0.43* 0.59* 0.57* 0.58* 0.81* 0.49* 0.51* Ad-rape 0.78* 0.32* 0.50* 0.56* 0.53* 0.41* 0.81* 0.66* Ad-molest child 1.00 0.29* 0.47* 0.50* 0.49* 0.44* 0.67* 0.81* Juv-gen. dp supp. 1.00 0.62* 0.55* 0.58* 0.55* 0.49* 0.44* Juv-murder cj emp 1.00 0.80* 0.80* 0.68* 0.62* 0.54* Juv-plan murder 1.00 0.86* 0.61* 0.64* 0.55* Juv-murder child 1.00 0.67* 0.65* 0.61* Juv-unplan murder 1.00 0.58* 0.60* Juv-rape 1.00 0.80* Juv-molest child 1.00 p<.05

PAGE 60

55 Table 4. Descriptions of the Deat h Penalty Scale Variables, the Relig ion Variables, the Race Variable, and the Control Variables (N = 636) a Variable Description %/Mean SD Support for capital 14-item additive scale b 46.76 12.58 Punishment Race/ethnicity 0 = whites 0.18 0.39 1 = Blacks and Hispanics Faith group c None 11.9% Jewish 2.5% Catholic 23.7% Presbyterian 4.6% Episcopalian 3.8% Methodist 8.3% Lutheran 3.1% Baptist 22.6% Fundamentalist 7.4% Religious tradition: Liberal None, Presb., Episc. 20.3% Moderate Jew, Cath., Meth., Luth. 50.2% Conservative Bapt., Fund., Pent., other Prots. 29.6% Conservative Protestant 0 = no (includes all other faiths) 70.5% 1 = yes (Bapt., Fund., Pent., Other) 29.6% Religiosity 9-item additive scale d 14.54 7.78 Sex 0 = females 0.46 0.50 1 = males Age Ordinal measure with higher number 43.74 11.93 representing higher age Education Ordinal measure with higher number 5.02 1.57 representing more formal education Income Ordinal measure with higher number 5.60 1.88 representing a higher income bracket Ideology Likert item, ranging from 1=very liberal 3.20 0.95 to 5=very conservative Abortion attitudes 5-item additive scale e 12.99 4.92 Continued on next page

PAGE 61

56 Table 4 continued Attitudes toward sex roles 3-item additive scale e 5.53 2.34 Attitudes toward social 2-item additive scale e 6.11 2.19 welfare Attitudes toward the police 4-item additive scale e 14.75 3.31 a The analyses only include Blacks, whites, and Hi spanics. All others (American Indians or Alaskan Natives and Asians) were deleted from the analyses. b Appendix C describes the construction of this composite scale. c Faith group, religious tradition, and membership in a conservative Protestant faith group are all measures of religious conservativism; each of these measures are nominal level. d Appendix D describes the construction of this composite scale. e Appendix F describes the construction of these composite scales.

PAGE 62

57 Applegate et al.s (2000) forgiveness scale. The 6-item scale was comprised of the following statements: 1. To receive forgiveness, it is important to love the sinner. 2. It is important to hate the sin, but love the sinner. 3. God teaches us that even if someone has lived a life of crime, s/he should be forgiven for their offenses if they are truly repentant (sorry). 4. The God I believe in sympathizes with people. 5. The God I believe in understands when people sin. 6. The God I believe in loves people, re gardless of their behaviors and beliefs. The response options for these items ranged from 1 = strongly disagree to 5 = strongly agree. We recoded the items so that higher scores on the scale were indicative of a more punitive theological emphasis. When these items were entered into a principal components factor analysis, the Kaiser Rule suggested the presence of two possible factors (eigenvalues = 3.09 and 1.09). An application of the Scree Discontinuity Test, however, demonstrated that the greatest di fference was between the first two values, thereby revealing the existence of only one substantial factor. The item-to-scale correlations were moderate to high (see Appendix E), ranging fro m .59 (God understands when people sin) to .81 (it is important to hate the sin, but love the sinner). The overall reliability of this scale was high (Cronbachs Alpha = .81). The possible range of scores on this scale ranged from 6 (the most progressive) to 30 (the most punitive). Potential jurors in Hillsborough County displayed a moderately progressive theological emphasis, with a mean score of 12.29 on the scale. Appendix E displays the full results of this factor analysis. The relationship between religion and capital puni shment attitudes is argued to be indirect, operating through a process of attribution. Just as previous research has revealed that possessing a strong belief in Biblical literalness is associated with individuals adopting a dispositional attribution style (Grasmick et al., 1993b), we, too, believe that this relationship will exist. Moreover, we believe that respondents theological emphasi s will affect the degree to which th ey literally interpret the Bible. Specifically, respondents who indicate that they have a more punitive theological emphasis should interpret the Bible more literally than respondents w ho have a more progressive theological emphasis, and individuals with a more progressive theol ogical emphasis should adopt a more interpretative reading of the Bible than their more punitive counterparts.

PAGE 63

58 We utilized five statements from various (but similar) adaptations of Grasmicks Biblical literalness scale (Grasmick et al., 1992; Grasmick et al., 1993b; Grasmick and McGill, 1994), which have been highly reliable in previous studies (e.g., Applegate et al., 2000). The five items that comprised the Biblical literalness scale in the present study included: 1. The Bible is the actual word of Godto be taken literally word for word. 2. Miracles really happenednot just stories. 3. Jesus truly rose from the dead. 4. Those who do not accept God will go to hell. 5. Hell actually exists. The response options for all 5 items ranged from 1 = strongly disagree to 5 = strongly agree. Thus, the higher the respondents scores, the more literally they interpreted the Bible. When these items were entered into a principal components factor analysis, the Kaiser Rule revealed the presence of a single factor solution (eigenvalue = 3.68). Factor loadings ranged from .81 (the Bible is the actual word of God) to .91 (miracles really happened). The scale produced a Cronbachs Alpha of .91, indicative of a highly reliable scale. Appendix E reports the results of this factor analysis. The range of scores on this scale could vary from 5 (strongly disagreed with every statement) to 25 (strongly agreed with every statement). Respondents in this sample, as a whole, interpreted the Bible with a fairly high degree of literalness (mean = 17.80). The model, in accordance with previous research (Cullen et al., 1985; Young, 1991; Grasmick et al., 1993a), predicts that there should be a direct effect of Biblical lite ralness on attribution style. We operationalized attribution style with one 10-item s cale, which reflected respondents beliefs about why adult and juvenile offenders committed crimes. We did not adopt a scale that had already been created; however, many of the items were taken from previous studies on attribution theory (especially Grasmick and McGill, 1994, but also Cullen et al., 1983). The 10 items included the following: 1. Most adult offenders commit crimes because they have bad characters. 2. Most juvenile offenders commit crim es because they have bad characters. 3. Most adult offenders commit crimes because they are too lazy to find a lawful way out of a bad situation. 4. Most juvenile offenders commit crimes because they are too lazy to fi nd a lawful way out of a bad situation. 5. Most adult offenders commit crimes as a way of coping with poor living conditions. 6. Most juvenile offenders commit crimes as a way of coping with poor living conditions. 7. Most adult offenders commit crimes because th eir home lives as children were lacking in love, discipline, and supervision.

PAGE 64

59 8. Most juvenile offenders commit crimes because their home lives as children were lacking in love, discipline, and supervision. 9. Most adult offenders commit crimes because of outside influences (e.g ., peer pressure, money problems). 10. Most juvenile offenders commit crimes because of outside influe nces (e.g., peer pressure, money problems). These statements were Likert-type items with response options ranging from 1 = strongly disagree to 5 = strongly agree. The situational attribution style items (5-10) were re-coded, so higher scores were demonstrative of a less situational attribution style. When the ten items were entered into a principal components factor analysis, the Kaiser Rule suggested the presence of two factors (eigenvalues = 3.30 and 1.52); however, the Scree Discontinuity Test re vealed that the greates t difference between the eigenvalues occurred between the first two values. Moreover, although the scales reliability is low (Cronbachs Alpha = .40), deleting any of the items from the scale would not substantially improve it. The factor loadings were low, ranging from .52 (adults commit crimes to cope with poor living conditions) to .63 (juveniles commit crimes to cope with poor living conditions). Appendix E reports the results of this factor analysis. The possible values on this scale ranged from 10 (highest degree of situational attribution style) to 50 (highest degree of dispositional attribution style). Respondents mean on the attribution scale was moderate at 27.45. Based on previous studies (e.g., Grasmick and McGill, 1994), possessing a dispositional or a situational attribution style shoul d be associated with having a f undamentally punitive or progressive philosophy toward convicted offende rs. We tested this proposition by hypothesizing a direct path between attribution style and attitudes toward the correctional treatment of adult and juvenile offenders. Again, we utilized one scale to measure attitudes toward system treatment for adult and juvenile offenders. The scale was comprised of twelve items, which operationalize the general sentencing philosophies of retribu tion, incapacitation, specific and ge neral deterrence, restitution, and rehabilitation. The items included the following statements: 1. When determining the sentence for adult offenders who have been legally convicted of crimes, the circumstances of the offense, such as weapon used, planning, injury to the victim, and amount of loss are not important. Instead, the sentence should be solely based on the seriousness of the offense committed. 2. When determining the sentence for juvenile offende rs who have been legally convicted of crimes, the circumstances of the offense, such as weapon used, planning, injury to the victim, and amount of loss are not important. Instead, the sentence should be solely based on the seriousness of the offense committed.

PAGE 65

60 3. The best way to deal with adults who have been legally convicted of crimes is to lock them up so they are not able to harm anyone again. 4. The best way to deal with juveniles who have been legally convicted of crimes is to lock them up so they are not able to harm anyone again. 5. The best way to deal with adults who have been legally convicted of crimes is to punish them harshly so they will learn their lesson and not commit other crimes. 6. The best way to deal with juveniles who have been legally convicted of crimes is to punish them harshly, so they will learn their lesson and not commit other crimes. 7. The best way to deal with adults who have been legally convicted of crimes is to punish them to the extent that they will pay their debt to society for the crimes they have committed. 8. The best way to deal with juveniles who have been legally convicted of crimes is to punish them to the extent that they will pay their debt to society for the crimes they have committed. 9. The best way to deal with adults who have been legally convicted of crimes is to punish them harshly to set an example, so that others will not commit crimes. 10. The best way to deal with juveniles who have been legally convicted of crimes is to punish them harshly to set an example, so that others will not commit crimes. 11. The best way to deal with adults who have been legally convicted of crimes is to treat and help them so that they do not commit other crimes. 12. The best way to deal with juveniles who have b een legally convicted of crimes is to treat and help them so that they do not commit other crimes. Response categories for each of these items ranged from 1 = strongly disagree to 5 = strongly agree. The rehabilitation item was re-coded, so higher scores on the scale were reflective of a less rehabilitative sentencing philosophy. The items were entered into a principal components factor analysis. The Kaiser Rule suggested the presence of three factors (eigenvalues = 4.22, 1.95, and 1.52); however the greatest difference between the eigenvalues was between the firs t two, thereby meeting the single factor criterion of the Scree Discontinuity Test. It is also of interest to note that the overall reliability of the scale was high (Cronbachs Alpha = .79), and deleting any of the twelve items would not substantially improve it. Factor loadings ranged from .01 (juveniles s hould be rehabilitated) to .83 (juven iles should be punished to set an example so others will not commit crimes). Factor load ings for each item, as well as the full results of this factor analysis are reported in Appendix E. Possible scores ranged from 12 (most rehabilitative approach toward offenders) to 60 (most punitive philosophy toward system treatment of o ffenders). Respondents in this sample were almost right in the middle of the continuum, with an average score of 37.32. Finally, we predicted that a general punitive or rehabilitative punishment philosophy would be related to attitudes toward capital punishment, meas ured as described above. Specifically, a punitive punishment philosophy would be pos itively associated w ith capital punishment support, and a more rehabilitative sentiment toward offenders would be associated with opposition to capital punishment. The correlation matrix for the variables in the attribution theory analysis appears in Table 5.

PAGE 66

61 Table 5. Correlation Matrix of Variable s in the Attribution Theory Analysis sex race age education income ideology abortion sexroles sex 1.00 -0.04 0.04 0.03 0.09* 0.01 0.01 0.17* race 1.00 0.13* -0.04 -0.07 -0.13* 0.04 0.03 age 1.00 -0.02 0.06 0.10* 0.05 0.19* education 1.00 0.36* 0.01 -0.13* -0.16* income 1.00 0.09* -0.11* -0.06 ideology 1.00 0.33* 0.24* abortion 1.00 0.37* sexroles 1.00 socwlfre cjscale theol. emph. punitive t.e. Bib. lit attrib.style pun. phil. dpscale Continued on next page

PAGE 67

62 Table 5 continued socwlfre cjscale theol. emph. punitive t. e. Bib. lit attrib. style pun. phil. dpscale sex 0.19* -0.07 0.11* 0.07 -0.04 0.03 0.04 0.09* race -0.21* -0.21* -0.09* 0.09 0.04 -0.02 0.01 -0.14* age 0.01 0.13* -0.05 0.00 0.06 0.11* 0.00 0.01 education 0.09* 0.12* 0.03 -0.11* -0.18* -0.14* -0.17* -0.14* income 0.15* 0.16* 0.09* -0.07 -0.13* -0.02 -0.04 0.07 ideology 0.27* 0.19* -0.12* 0.08* 0.28* 0.13* 0.14* 0.17* abortion 0.13* 0.07 -0.33* 0.11* 0.44* 0.10* -0.02 -0.11* sexroles 0.21* -0.11* -0.05 0.10* 0.22* 0.07 0.12* 0.04 socwlfre 1.00 0.16* 0.13* -0.05 -0.02 0.22* 0.17* 0.22* cjscale 1.00 -0.05 -0.08 0.07 -0.01 0.03 0.06 theol. emph. 1.00 -0.25* -0.53* 0.09* 0.10* 0.15* punitive t.e. 1.00 0.42* 0.08 0.24* 0.13* Bib. lit. 1.00 -0.02 0.13* 0.11* attrib. style 1.00 0.13* 0.22* pun. phil. 1.00 0.45* dpscale 1.00 Significant at p<.05

PAGE 68

63 Control Variables Several opinion polls have suggest ed that socio-demographic variab les may be related to attitudes toward capital punishment. Thus, we controlled for them in the present study. Sex is coded as a dummy variable (0 = female). Age is measured in number of years. Education is measured on an ordinal scale, ranging from 1 = grade school or less to 8 = an advanced degree. Income is measured by combined household annual income, also an ordinal scale, ranging from 1 = under $10,000 to 10 = $150,000+. In addition to socio-demographic variables, we control fo r several socio-political variables, as they, too, are likely to be related to both religious affiliation and death penalty attitudes. First, political ideology is measured by a single item that asked respondents to id entify their social and political views as: (1) very liberal, (2) somewhat liberal, (3) middl e of the road, (4) somewhat conserva tive, or (5) very conservative. Note that we also asked respondents which, if any, political party they were a member of, but due to the risk of multicollinearity, this variable was not controlled for in the analysis. Second, a five-item additive abortion scale was created (alpha = .87). This scale was measured with five indicators of support for abortion under varying circumstances. Specifically, these circumstances included: (1) under any circumstances, (2) when a womans health or life is endangered, (3) when the fetuss health or life is endangered, (4) when the pregnancy was the result of a rape, and (5) when the mother cannot afford to raise the child. These five items were entered into a principal components factor analysis, from which the Kaiser Rule suggested a single factor solution (eigenvalue = 3.29). Factor loadings ranged from .74 (consideration of abortion when mother was poor) to .87 (consideration of abortion when the pregnancy was th e result of rape) (see Appendix F). Third, a three-item additive gender roles scale was created (alpha = .70). Specifically, respondents were asked whether women s hould have an equal role with men in running businesses, the military, and the family. Response options ranged from 1 = strongly agree to 5 = strongly disagree. These three items were also entered into a principal components factor analys is, and the Kaiser Rule, again, indicated the presence of a single factor solution (eigenvalue = 1.96). Fact or loadings ranged from .77 (women should have an equal role with men in the military to .84 (women s hould have an equal role with men in businesses) (see Appendix F).

PAGE 69

64 Fourth, a two-item social welfare scale was created (alpha = .67). Participants were asked whether they believed that the government should improve the social and economic conditions of minorities through: (1) affirmative action and (2) social welfare programs. Again, the possible responses ranged from 1 = strongly agree to 5 = strongly disagree. A principal components factor analysis revealed that the two items successfully created one scale (eigenvalue = 1.50). Finally, an additive four-item criminal justice scale was creat ed based on respondents attitudes toward the police (alpha = .80). Respondents were o ffered Likert-type statements, to which they were asked to indicated the degree to which they agreed or disagreed with the following statements: (1) most police officers are prejudiced; (2) most police officers abuse their di scretion; (3) most police officers are untrustworthy; and (4) most po lice officers act professionally when dealing with citizens (reverse coded). These items were entered into a principal components factor analysis, and the Kaiser Rule supported a single factor solution (eigenvalue = 2.56). Factor loadings ranged from .75 (belief that most police officers are prejudice) to 82 (belief that most police officers are untrustworthy). Appendix F provides the eigenvalues and the item-to-scale correlations for each of the control variables. Table 3 provides the correlation matrix for the religion and the control variables. Notice that the correlations between the control variable items were largely non-significant, thereby showing no explanatory overlap and enabling each to possess unique explanatory powerin other words, there does not appear to be a problem with multicollinearity. Analytic Strategy The statistical analysis will pro ceed in two stages. First, we us e analysis of variance (ANOVA) to assess significant differences in attitudes toward capital punishment by faith. We then employ OLS regression to determine whether the variables in each pa th are, as we predicted, directly related to one another. The strength and significance of the effect of the independent variable on the dependent variable in each path is indicative of the su ccess of attribution theory as a c onceptual framework through which the relationship between religion and a ttitudes toward capital punishment may be understood. We also utilized OLS regression to assess all of the possible indirect effects of the variables on one another (e.g., the possible effect of theological emphasis on attribution style or Biblical literalness on punishment philosophy). These effects will not be reported in great detail, as they are not the focus of the analysis.

PAGE 70

65 Table 6. Descriptions of the Variables in the Test of Attribution Theory (N = 636) a Variable Description Mean STD Theological Emphasis Progressive 6-item additive scale b 12.29 4.78 Punitive 4-item additive scale b 10.60 3.41 Biblical literalness 5-item additive scale b 17.80 5.25 Attribution style 10-item additive scale b 27.45 4.21 Punishment philosophy 12-item additive scale b 37.32 7.10 a The analyses include only Blacks, whites, and Hispanics. All others (American Indians or Alaskan Natives and Asians) were deleted from the analysis. b Appendix E describes the construction of these composite scales.

PAGE 71

66 CHAPTER 5 RESULTS: THE EFFECTS OF RACE AND RELIGION ON DEATH PENALTY ATTITUDES We first conducted an Analysis of Variance ( ANOVA) to assess significant differences in attitudes toward capital punishment by faith (hypothesis 1). Table 7 contains the full results of the ANOVA procedure for all three measures of religious conservativism. Prior to classifying faith groups into the liberal/moderate/conservative categories (which, recall, wa s due to the small number of individuals in some faiths and the small number of minorities in all faiths), we examined the differences in mean levels of overall support for capital punishment between each of the nine faith groups (see line 1 of Table 7). Inconsistent with our expectations, both the Tukey and Scheffe comparison of means tests revealed the presence of only one significant relationship: Baptists exhibited significantly higher levels of support for capital punishment overall (mean=49.54) than members of other faith groups (mean=45.95). The overall model, however, was not significant. When the faith groups were classified into religious traditions (liberal, moderate, and conservative), we expected to find a higher number of significant relationships, given the larger number of individuals in each group. This was not the case. Results of the Tukey and Scheffe comparison of means tests demonstrated no appreciable mean differences in overall death penalty attitudes between any of the classifications, and the overall model was also not significant (F=2.52, p>.05). Because conservative Protestant faiths are reputed to be particularly pros criptive in their moral messages, we conducted an ANOVA to assess mean di fferences between members of these faiths and members of other faith groups. Line 1 of Table 7 reports the results of this model. Consistent with our expectation, conservative Protestants displayed significantly higher levels of support for capital punishment (mean=48.44) than members of other religions (mean=46.04). The overall model was significant (F=4.86, p<.05), but it explained almost none of the variation in capita l punishment attitudes (R=.008). Given the surprisingly weak association between religious affiliation and death penalty attitudes, we assessed faith group differences in capital punish ment support by specifically examining support for capital punishment by offender and offense types (i.e ., the 14 individual items that comprised the overall

PAGE 72

67 Table 7. Analysis of Variance of Death Penalty Attitudes by Religious Conservativism, Measured by Faith Group, Religious Tradition, and Membership in a Conservative Protestant Denomination Dependent Variable F-value F-value F-value faith group religious tradition conservative Protestant dp-support scale 1.53 1 2.52 4.86* (.008) adult-general support 1.20 2.03 adult-murder of cj employee 2.74 5.04* (.008) adult-planned murder 2.26 1.00 adult-murder of child 0.89 0.42 adult-unplanned murder 3.63* (.011) 7.10* (.011) adult-rape 0.09 0.01 adult-child molestation 0.75 1.01 juvenile-general support 2.85 5.19* (.008) juvenile-murder of cj employee 4.51* (.014) 8.28* (.013) juvenile-planned murder 3.54* (.011) 6.02* (.009) juvenile-murder of child 5.46* (.017) 8.05* (.013) juvenile-unplanned murder 4.35* (.014) 8.92* (.014) juvenile-rape 0.46 0.93 juvenile-child molestation 1.09 1.58 1 This overall model was non-significant; however, the ANOVA procedure revealed a significant difference between Baptists and members of other faiths (F=9.16, p<.05). 2 The R values are reported in parenthesis next to those models that are significant. Significant at p<.05

PAGE 73

68 scale). Due to the small number of members in most religions, we conducted ANOVAs for only the latter two measures of religious conservativism, religious tradition and membership in a conservative Protestant faith group. When death penalty support is analyzed by offe nder and offense types and faith group is measured by religious tradition, 5 of the 14 models attain significance (see Table 7). These models include support for capital punishment for juveniles who murder a criminal justice employee (F=4.51, p<.05), commit a planned murder (F=3.54, p<.05), murder a child (F=5046, p<.05), and commit an unplanned murder (F=4.35, p<.05) and adults who commit an unplanned mu rder (F=3.63, p<.05). All of the models were weak, though, as they explained only 1.1 percent (juve niles who commit a planned murder and adults who commit an unplanned murder) to 1.7 percent (juveniles who murder a child) of the variation in death penalty attitudes. Interestingly, four of the five significant models involve the execution of juvenile offenders, suggesting greater polarization among faith traditions with respect to juvenile rather than adult capital punishment. Also interesting, in all five of the significant models, the Tukey and Scheffe comparison of means tests revealed that the only significant differences in mean support for capital punishment were between individuals affiliated with moderate and conservative faith groups, with members of conservative traditions displaying higher levels of support for the death penalty in all of the models. There are no distinguishable differences betw een members of liberal and conservative religions. One possible explanation for this unexpected finding may be that the religious liberal category contains individuals claiming no religious affiliation. There is no guarantee, in retrospect, that individuals without religious affiliation are actually liberal in their political or social ideologies. When the measure of the independent variable is dichotomized (i.e., conservative Protestants/nonconservative Protestants), similar findings to the previous ones occur. Specifically, 7 of the 14 models attain significance: general support for juvenile capita l punishment (F=5.19, p<.05), death penalty support for adults (F=5.04, p<.05) and juveniles (F=8.28, p<.05) who murder a criminal justice employee, juveniles who commit a planned murder (F=6.02, p<.05), juvenile s who murder a child (F=8.05, p<.05), and adults (F=7.10, p<.05) and juveniles (F=8.92, p<.05) w ho commit an unplanned murder. The mean level of support for capital punishment was higher for conservativ e Protestants than members of other faith groups in all of these models. None of the models explaine d a substantial amount of variation in attitudes toward

PAGE 74

69 capital punishment. The weakest model (adults w ho committed an unplanned murder) accounted for 1.1 percent of the variance, and the strongest (juven iles who committed an unplanned murder) accounted for 1.4 percent. Note that all five of the models that were significant when faith groups were trichotomized were also significant when the independent variable was dichotomized. The only change was the addition of two significant models, general support for juvenile capital punishment and adults who murder a criminal justice employee. Summarily, the first hypothesis that there would be a positive relationship between religious conservativism and support for capital punishment received mixed support. Although there were some significant differences between moderate and conserva tive faith groups, being a member of a conservative Protestant denomination appeared to be the best predictor of support for capital punishment. (Recall that the conservative category of the religious tradition measure was comprised of only Protestant denominations.) Indeed, when conservative Protesta nts were compared with members of other faith groups, they displayed significantly higher levels of support for capital punishment in half of the offender and offense-specific models. It is premature to draw any conclusions based on these findings alone, however. The effect of individuals commitment to their religion and the importance of religion in their daily lives must also be considered, as group membership combined with the degree to which individuals identify the group as important (religiosity) affect thei r attitudes and behaviors. We address the effect of religiosity in our second hypothesis: as religiosity increases, the effect of religious conservativism on capital punishment support w ill increase. We employ Ordinary Least Squares (OLS) Regression to address this hypothesis and, again, use religious tradition and membership in a conservative Protestant faith group as measures of relig ious conservativism. Table 8 reports the results for these analyses. See also Appendix G for the effects of the control variables in both sets of models. In the religious tradition models, religious tradition significantly affected death penalty attitudes in six of the models: adults who rape (b=-.37, p<.05) the overall juvenile scale (b=-2.22, p<.05), general support for juvenile capital punishment (b=-.53, p<.05), juveniles who commit a planned murder (b=-.45, p<.05), juveniles who rape (b=-.44, p<.05), and juveniles who molest a child (b=-.35, p<.05). Somewhat surprisingly, the effect of religious conservativism on capital punishment support was negative; however, this effect occurred when religiosity was controlled. When the interactive effects of religious tradition and

PAGE 75

70 Table 8. OLS Regression of Death Penalty Attitudes by Religious Conservativism (Measured by Religious Tradition and Membership in a Conservative Protestant Denomination) and Religiosity Dependent Variable religious traditi on religious tradition*religiosity b se(b) B b se(b) B adult-overall scale -1.11 0.90 -0.14 0.08 0.05 0.29 adult-general support -0. 23 0.16 -0.16 0.02 0.01 0.34 adult-murder of cj employee 0.04 0.16 0.03 0.00 0.01 0.08 adult-planned murder -0.11 0.16 -0.08 0.01 0.01 0.23 adult-murder of child -0.05 0.17 -0.04 0.01 0.01 0.12 adult-unplanned murder -0.11 0.13 -0.23 0.01 0.01 0.29 adult-rape -0.37* 0.17 -0.25 0.02 0.01 0.35 adult-child molestation -0.28 0.17 -0.20 0.01 0.01 0.24 juvenile-overall scale -2.22* 0.85 -0.29 0.16* 0.05 0.57 juvenile-general support -0.53* 0.17 -0.35 0.04* 0.01 0.67 juvenile-murder of cj employee -0.12 0.15 -0.09 0.02* 0.01 0.37 juvenile-planned murder -0.45* 0.16 -0.31 0.03* 0.01 0.61 juvenile-murder of child -0.19 0.15 -0.14 0.02* 0.01 0.38 juvenile-unplanned murder -0.14 0.13 -0.12 0.01 0.01 0.33 juvenile-rape -0.44* 0.15 -0.33 0.02* 0.01 0.51 juvenile-child molestation -0.35* 0.15 -0.28 0.02* 0.01 0.38 conservative Protestant conservative Protestant*religiosity b se(b) B b se(b) B adult-overall scale -0.94 1.45 -0.06 0.10 0.08 0.13 adult-general support -0. 24 0.26 -0.09 0.02 0.01 0.17 adult-murder of cj employee 0.10 0.25 0.04 0.01 0.01 0.06 adult-planned murder -0.26 0.26 -0.10 0.02 0.01 0.18 adult-murder of child -0.10 0.27 -0.04 0.01 0.01 0.09 adult-unplanned murder -0.09 0.21 -0.04 0.01 0.01 0.13 adult-rape -0.44 0.27 -0.17 0.02 0.01 0.16 adult-child molestation 0.08 0.27 0.03 -0.00 0.01 -0.03 juvenile-overall scale -0.90 1.38 -0.06 0.11 0.07 0.15 juvenile-general support -0.23 0.28 -0.08 0.03 0.01 0.19 juvenile-murder of cj employee 0.06 0.24 0.03 0.01 0.01 0.09 juvenile-planned murder -0.37 0.26 -0.14 0.03* 0.01 0.26 juvenile-murder of child -0.11 0.25 -0.05 0.02 0.01 0.15 juvenile-unplanned murder 0.07 0.21 0.03 0.00 0.01 0.04 juvenile-rape -0.34 0.25 -0.14 0.02 0.01 0.15 juvenile-child molestation 0.03 0.24 0.01 -0.00 0.01 -0.02 *Significant at p<.05

PAGE 76

71 religiosity were considered, the cross product term attained significance in 7 of the 8 juvenile models, but none of the adult models, thereby corroborating our ear lier claim that the religi on-death penalty attitudes relationship may be polarized among faith groups with respect to executing juvenile, but not adult, offenders. The effects in these models were weak, but consistent with our hypothesis. So, as religiosity increases, the effect of religious conservativism on capital punishment support also increasesfor juvenile offenders: overall juvenile support scale (b=.16, p<. 05), general support for juvenile capital punishment (b=.04, p<.05), juveniles who murdered a criminal justice employee (b=.02, p<.05), juveniles who committed a planned murder (b=.03, p<.05), juveniles who murdered a child (b=.02, p<.05), juveniles who raped (b=.02, p<.05), and juveniles who molested a child (b=.02, p<.05). All 16 of the full models attained significance and possessed variable explanatory power ranging from 10.8 percent (adults who raped) to 19.5 percent (overall juvenile death penalty scale). The F-values and R values are as follows: overall adult scale (F=7.16, p<.05; R=.170), general support for adult capital punishment (F=7.76, p<.05; R=.182), adult murder of a criminal justice employee (F=5.69, p<.05; R=.140), adult planned murder (F=5.60, p<.05; R=.138), adult murder of child (F=5.75, p<. 05; R=.142), adult unplanned murder (F=4.69, p<.05; R=.118), adult rape (F=4.23, p<.05; R=.108), adult child molestation (F=4.72, p<.05; R=.119), overall juvenile scale (F=8.47, p<.05; R=.195), general support for juvenile capital punishment (F=7.21, p<.05; R=.171), juvenile murder of a criminal justice employee (F=6.97, p<.05; R=.166), juvenile planned murder (F=7.18, p<.05; R=.171), juvenile murder of child (F=6.78, p<.05; R=.163), juvenile unplanned murder (F=4.80, p<.05; R=.121), juvenile rape (F=5 .24, p<.05; R=.131), and juvenile child molestation (F=5.20, p<.05; R=.130). Unlike the religious tradition models, religious conservativism measured by membership in a conservative Protestant faith group did not affect capital punishment support in any of the 16 models when all other variables were controlled. Table 8 reports the results of these models and Appendix G reports the effects of the control variables. Th e interactive effect of membership in a conservative Protestant faith and religiosity affected death penalty support in only one model, juveniles who committed a planned murder (b=.03, p<.05). Note that this single significant effect is a juvenile model, and it is in the expected direction. All of the 16 overall models were significant and explained between 8.6 percent (adults who molested a child) and 18.8 percent (general support for adult capital punishment) of the variation in death

PAGE 77

72 penalty attitudes. The F values and R values were as follows: adult death penalty scale (F=9.47, p<.05; R=.165), general support for capital punishment for adu lts (F=11.04, p<.05; R=.188), adult murder of a criminal justice employee (F=9.08, p<.05; R=.160), adult planned murder (F=7.65, p<.05; R=.138), adult murder of child (F=7.45, p<.05; R=.135), adult unpl anned murder (F=6.85, p<.05; R=.125), adult rape (F=4.49, p<.05; R=.086), adult child molestation (F=4.48, p<.05; R=.086), juvenile death penalty scale (F=10.36, p<.05; R=.178), general s upport for juvenile death penalty (F=9.02, p<.05; R=.159), juvenile murder of a criminal justice employee (F=9.76, p<.05; R=.169), juvenile planned murder (F=9.01, p<.05; R=.156), juvenile murder of child (F=8.61, p<.05; R=.153), juvenile unplanned murder (F=6.93, p<.05; R=.127), juvenile rape (F=5.57, p<.05; R=.104), and juvenile child molestation (F=5.14, p<.05; R=.097). We turn now to our third hypothesis: across faith groups, support for capital punishment will be greater among whites than among Blacks and Hispanics. An ANOVA revealed that, indeed, there were significant differences between whites and non-whites in their death penalty attit udes (F=12.63, p<.05). The Tukey and Scheffe comparison of means tests indicated that whites (mean=47.60) exhibited a significantly higher level of support for capital punishment than non-whites (mean=43.06). The model explained 2 percent of the variance in death penalty attitudes. Table 9 reports these findings. When death penalty support is considered for specific offenders and offense types, the data demonstrated significant mean differences between whites and non-whites in 10 of the 14 models: general support for adult capital punishment (F=16.93, p<.05), support for the death penalty for adults (F=20.94, p<.05) and juveniles (F=16.79, p<.05) who murder a cr iminal justice employee, adults (F=8.56, p<.05) and juveniles (F=7.50, p<.05) who commit a planned murder adults (F=10.55, p<.05) and juveniles (F=8.02, p<.05) who murder a child, adults who commit an unplanned murder (F=6.74, p<.05), and adults (F=8.88, p<.05) and juveniles (F=3.87, p<.05) who rape. All of the models were weak in that they explained little variance in capital punishment attitudes; the R valu es ranged from .006 (support for the death penalty for juveniles who rape) to .032 (support for the death penalty for adults who murder a criminal justice employee). The remaining four models (general support for juvenile capital punishment, capital punishment support for juveniles who commit an unpla nned murder, and for adults and juveniles who molest a child) were not significant. Whites displa yed more favorable attitudes toward capital punishment than non-whites in all ten of the significant models. See Table 9 for a summary of these findings.

PAGE 78

73 Table 9. Analysis of Variance of Death Penalty Attitudes by Race Dependent Variable F-value mean (whites) mean (non-whites) dp support scale 12.63* (.020) 1 47.60 43.06 adult-general support 16.93* (.026) 3.97 3.45 adult-murder of cj employee 20.94* (.032) 3.69 3.15 adult-planned murder 8.56* (.013) 4.02 3.66 adult-murder of child 10.55* (.016) 4.02 3.62 adult-unplanned murder 6.74* (.011) 3.07 2.81 adult-rape 8.88* (.014) 3.44 3.07 adult-child molestation 3.43 juvenile-general support 2.91 juvenile-murder of cj employee 16.79* (.026) 3.31 2.85 juvenile-planned murder 7.50* (.012) 3.56 3.22 juvenile-murder of child 8.02* (.012) 3.45 3.12 juvenile-unplanned murder 1.73 juvenile-rape 3.87* (.006) 3.10 2.87 juvenile-child molestation 1.85 1 The R values are reported in parentheses next to those models that are significant. Significant at p<.05

PAGE 79

74 The ANOVA revealed strong support for our third hypothesis, regardless of whether the dependent variable was the death penalty attitudinal scale or the items comprising the scale. Whites consistently displayed higher levels of support for capital punishment than Blacks and Hispanics. The scope of these findings, though, is largely untelling. Although race does constitute a group through which attitudes and behaviors may be affected, members of each race also belong to other groups, making it necessary to understand how attitudes are affected with the presence of multiple groups. We analyze the differential effects of religious conservativism on capital punishment support while controlling for religiosity among whites and Blacks/Hispanics separately in our fourth and fifth hypotheses: among whites, as religiosity increases, the effect of religious conservativism on capital punishment support will increase; and among Blacks and Hisp anics, as religiosity in creases, the effect of religious conservativism on capital punishment support will remain unchanged. Consistent with our previous analyses, we utilize religious tradition and me mbership in a conservative Protestant faith group as measures of religious conservativism. Hypothesis 4 regarding the effect of religi ous conservativism on capital punishment support for whites received only slight support. In the religious tradition models, re ligious tradition negatively affected death penalty support among whites in four of the juvenile circumstances, indicating that the more conservative their religious tradition, the less likely whites were to support capital punishment. The four models included the overall juvenile death penalty s cale (b=-1.77, p<.05), general support for juvenile capital punishment (b=-.43, p<.05), juveniles who comm it a planned murder (b=-.37, p<.05), and juveniles who rape (b=-.37, p<.05). Table 10 reports the results for both the religious tradition and membership in a conservative Protestant faith group models for whites. Appendix H displays the effects of the control variables in these models. More important than the effect of religious tradition on capital punishment support is the test of our hypothesis, which addre sses the effect of religious conservativism on capital punishment support for whites as their degree of religiosity increases. When the interactive effects of religious conservativism and religiosity are considered, they attain significance in the same four models as religious tradition; however, their relationship to capital punishment suppor t is positive in the direction we expected. Notice that the effects weaken when religiosity is cons idered. The unstandardized Beta coefficients are: overall juvenile death penalty scal e (b=.14, p<.05), general support for juvenile capital

PAGE 80

75 Table 10. OLS Regression of Death Penalty Attitudes by Religious Conservativism (Measured by Religious Tradition and Membership in a Conservative Protestant Denomination) and Religiosity Among Whites Dependent Variable religious traditi on religious tradition*religiosity b se(b) B b se(b) B adult-overall scale -0.72 0.95 -0.09 0.06 0.13 0.22 adult-general support -0. 15 0.17 -0.10 0.01 0.01 0.23 adult-murder of cj employee 0.09 0.16 0.06 0.00 0.01 0.04 adult-planned murder -0.04 0.17 -0.03 0.01 0.01 0.18 adult-murder of child -0.01 0.17 -0.01 0.00 0.01 0.07 adult-unplanned murder -0.07 0.13 -0.07 0.01 0.01 0.29 adult-rape -0.29 0.18 -0.19 0.01 0.01 0.28 adult-child molestation -0.25 0.18 -0.17 0.01 0.01 0.20 juvenile-overall scale -1.77* 0.89 -0.23 0.14* 0.05 0.53 juvenile-general support -0.43* 0.18 -0.28 0.03* 0.01 0.52 juvenile-murder of cj employee -0.04 0.15 -0..03 0.01 0.01 0.31 juvenile-planned murder -0.37* 0.17 -0.26 0.03* 0.01 0.60 juvenile-murder of child -0.15 0.17 -0.11 0.02 0.01 0.36 juvenile-unplanned murder -0.11 0.13 -0.10 0.01 0.01 0.35 juvenile-rape -0.37* 0.16 -0.28 0.02* 0.01 0.51 juvenile-child molestation -0.29 0.16 -0.23 0.02 0.01 0.35 conservative Protestant conservative Protestant*religiosity b se(b) B b se(b) B adult-overall scale -0.76 1.49 -0.05 0.08 0.08 0.12 adult-general support -0. 19 0.26 -0.07 0.02 0.01 0.15 adult-murder of cj employee 0.08 0.25 0.03 0.01 0.01 0.08 adult-planned murder -0.24 0.27 -0.09 0.02 0.01 0.18 adult-murder of child -0.13 0.27 -0.05 0.01 0.01 0.09 adult-unplanned murder -0.12 0.21 -0.06 0.02 0.01 0.16 adult-rape -0.32 0.29 -0.12 0.01 0.02 0.11 adult-child molestation 0.17 0.28 0.06 -0.01 0.02 -0.08 juvenile-overall scale -0.55 1.42 -0.04 0.11 0.08 0.16 juvenile-general support -0.15 0.29 -0.06 0.02 0.02 0.15 juvenile-murder of cj employee 0.08 0.25 0.03 0.01 0.01 0.10 juvenile-planned murder -0.28 0.26 -0.11 0.03* 0.01 0.28 juvenile-murder of child -0.12 0.26 -0.05 0.02 0.01 0.18 juvenile-unplanned murder 0.02 0.21 0.01 0.01 0.01 0.10 juvenile-rape -0.22 0.26 -0.09 0.02 0.01 0.15 juvenile-child molestation 0.12 0.24 0.05 -0.00 0.01 -0.04 *Significant at p<.05

PAGE 81

76 punishment (b=.03, p<.05), juveniles who commit a pla nned murder (b=.03, p<.05), and juveniles who rape (b=.02, p<.05). All of the 16 overall models are si gnificant and have R values ranging from .094 (adults who rape) to .190 (general support for adult death penalty ). The F values and R values are: overall adult death penalty scale (F=6.9, p<.05; R=.171), general support for adult capital punishment (F=7.42, p<.05; R=.190), adult murder of a criminal justice employee (F=4.96, p<.05; R=.136), adult planned murder (F=5.39, p<.05; R=.145), adult murder of a child (F=5.41, p<.05; R=.146), adult unplanned murder (F=4.69, p<.05; R=.129), adult rape (F=3.29, p<.05; R=.094), adult child molestation (F=4.67, p<.05; R=.129), overall juvenile capital punishment scale (F=7.26, p<.05; R=.187), general support for juvenile death penalty (F=6.65, p<.05; R=.174), juvenile murder of a criminal justice employee (F=5.89, p<.05; R=.157), juvenile planned murder (F=6.15, p<.05; R =.163), juvenile murder of a child (F=5.77, p<.05; R=.154), juvenile unplanned murder (F=4.58, p<.05; .126), juvenile rape (F=4.15, p<.05; R=.116), and juvenile child molestation (F=4.49, p<.05; R=.124). When we changed the measure of religious c onservativism to membership in a conservative Protestant faith group, we found different results of the effect of religious conservativism on capital punishment support among whites. Our hypothesis received almo st no support. Indeed, membership in a conservative Protestant faith did not significantly affect whites attitude s toward capital punishment in any of the models. More importantly, the more religious white members of conservative Protestant faith groups were, the more likely they were to support capital punishment in only one modeljuveniles who committed a planned murder (b=.03, p<.05). This isolated finding is untelling with regard to the effect of religious conservativism on capital punishment support. Our failure to find significant results is more shocking, especially in light of th e body of literature that suggests that highly religious white members of conservative Protestant faith groups display consid erably conservative social and political attitudes, including support for the death penalty. All 16 of th e overall models attained significance and possessed explanatory power, ranging from 7.3 percent (adults who rape) to 21.2 percent (general support for capital punishment for adults). The F values and R values in each model were as follows: overall adult capital punishment scale (F=8.50, p<.05; R=.168), general support for adult death penalty (F=11.32, p<.05; R=.212), adult murder of a criminal justice employee (F=7.78, p<.05; R=.156), adult planned murder (F=6.72, p<.05; R=.138), adult murder of a child (F=6.68, p<.05; .137), adult unplanned murder (F=6.75,

PAGE 82

77 p<.05; R=.138), adult rape (F=3.31, p<.05; R=.073) adult child molestation (F=4.43, p<.05; R=.095), overall juvenile death penalty scale (F=9.41, p< .05; R=.183), general capital punishment support for juveniles (F=8.74, p<.05; R=.172), juvenile murder of a criminal justice employee (F=8.68, p<.05; R=.171), juvenile planned murder (8.57, p<.05; R=. 169), juvenile murder of a child (F=7.83, p<.05; R=.157), juvenile unplanned murder (F=6.63, p<.05; R=.136), juvenile rape (F=4.35, p<.05; R=.094), and juvenile child molestation (F=4.57, p<.05; R=.098). Our fifth hypothesis that relig ious conservativism should not a ffect death penalty attitudes among Blacks and Hispanics, regardless of their levels of religiosity, was almost entirely supported. Religious tradition was only significant in the general death pena lty support for juveniles model (b=-1.71, p<.05). Although this finding seems to suggest that as religious conservativism increases, death penalty support decreases, keep in mind that the effect of religiosity, which likely mediates faith group effects, is not considered. Indeed, when religiosity is considered, religious conservativism measured by religious tradition is associated with increased support for capita l punishment. This effect is weak and is also observed in the general juvenile death penalty model (b=.12, p<.05). More important is the larger picture that as religiosity increases among Blacks and Hispanics, religious conservativism has no significant effect on capital punishment attitudes. Only three of thes e overall models attained significance: the overall juvenile death penalty scale (F=2.06, p<.05; R =.285), general support for juvenile capital punishment (F=2.32, p<.05; R=.310), and juvenile unplanned murder (F=2.16, p<.05; R=.295). Notice that the explanatory power of these models is substantially higher than that of the previous models that included both whites and non-whites and whites only. When we measured religious conservativism with membership in a conservative Protestant faith group, the results are very similar to our findings for non-whites when religious conservativism was measured with religious tradition. Membership in a conservative Protestant faith group was not significant in any of the 16 non-white models. The interaction term conservative Protestant faith*religiosity was not significant in 15 of the 16 models, suggesting that, re gardless of their degrees of religiosity, Black and Hispanic members of conservative Protestant denominations were not more likely than Black and Hispanic members of other faith groups to support the death penalty. The model in which the interactive effect of religious conservativism and religiosity was signifi cant (general support for juvenile capital punishment)

PAGE 83

78 revealed that more highly religious Black and Hispanic members of conservative Protestant faith groups displayed higher levels of capital punishment support than their less religious counterparts (b=.11, p<.05). The results of these models, as well as for the religious tradition models, are reported in Table 11, and the effects of the control variables are reported in Appendix I for both the religious tradition and membership in a conservative Protestant denomination models. Fi ve of these sixteen overall models were significant, but explained less variati on in capital punishment attitudes than the religious tradition models: adult unplanned murder (F=1.86, p<.05; R=.177), the overa ll juvenile death penalty scale (F=2.34, p<.05; R=.213), juvenile planned murder (F=1.85, p<.05; R=.176), juvenile unplanned murder (F=3.03, p<.05; R=.259), and juvenile rape (F=2.32, p<.05; R=.211). Summarily, our hypotheses regarding the race-religion/religiosity-death penalty attitudes relationship received slight to mi xed support, at best. Support for hypothesis 3 recei ved mixed support limited to when religious tradition was the measure of religious conservativism and to support for capital punishment for juveniles. Support for hypothesis 4 was also limited to support for the death penalty for juveniles, although even in these models, only half were significant when religious conservativism was operationalized as religious tradition, and only one was significant when it was operationalized as membership in a conservative Protestant faith group. Support for hypothesis 5 received the most support and was not limited to support for juvenile capital punishment. It is important to consider, however, the possibility that the findings (lack of significant religious conservativism*religiosity effects) regarding Black and Hispanic conservative Protestants could be a function of their low membership in these denominations. There were 74 total non-whites, only 12 of whom identified themselves as members of conservative Protestant faiths.

PAGE 84

79 Table 11. OLS Regression of Death Penalty Attitudes by Religious Conservativism (Measured by Religious Tradition and Membership in a Conservative Protestant Denomination) and Religiosity Among Non-whites Dependent Variable religious traditi on religious tradition*religiosity b se(b) B b se(b) B adult-overall scale -3.34 3.26 -0.42 0.20 0.18 -0.42 adult-general support -1. 00 0.60 -0.68 0.07 0.03 1.09 adult-murder of cj employee -0.12 0.58 -0.09 0.01 0.03 0.22 adult-planned murder -0.32 0.60 -0.22 0.02 0.03 0.28 adult-murder of child -0.07 0.60 -0.05 0.01 0.03 0.14 adult-unplanned murder -0.24 0.47 -0.21 0.01 0.03 0.32 adult-rape -1.02 0.57 -0.76 0.05 0.03 0.89 adult-child molestation -0.57 0.61 -0.40 0.03 0.03 0.48 juvenile-overall scale -4.77 3.02 -0.62 0.28 0.17 0.88 juvenile-general support -1.71* 0.59 -1.12 0.12* 0.03 1.91 juvenile-murder of cj employee -0.43 0.54 -0.32 0.03 0.03 0.60 juvenile-planned murder -0.92 0.60 -0.62 0.05 0.03 0.79 juvenile-murder of child -0.18 0.51 -0.14 0.02 0.03 0.30 juvenile-unplanned murder -0.02 0.45 -0.02 0.00 0.03 0.07 juvenile-rape -0.79 0.49 -0.65 0.03 0.03 0.56 juvenile-child molestation -0.72 0.53 -0.56 0.03 0.03 0.59 conservative Protestant conservative Protestant*religiosity b se(b) B b se(b) B adult-overall scale -3.32 5.47 -0.20 0.22 0.29 0.25 adult-general support -1. 15 1.04 -0.36 0.08 0.05 0.49 adult-murder of cj employee 0.32 0.93 0.11 -0.01 0.05 -0.08 adult-planned murder -0.25 0.99 -0.08 0.02 0.05 0.12 adult-murder of child 0.35 1.02 0.11 -0.01 0.05 -0.08 adult-unplanned murder 0.13 0.76 0.05 -0.00 0.04 -0.00 adult-rape -1.79 0.97 -0.60 0.09 0.05 0.60 adult-child molestation -0.92 1.00 -0.31 0.05 0.05 0.35 juvenile-overall scale -4.42 5.08 -0.27 0.21 0.27 0.26 juvenile-general support -1.53 1.01 -0.48 0.11* 0.05 0.68 juvenile-murder of cj employee -0.24 0.85 -0.09 0.02 0.04 0.12 juvenile-planned murder -1.14 1.01 -0.36 0.05 0.05 0.33 juvenile-murder of child 0.15 0.89 0.05 -0.01 0.05 -0.07 juvenile-unplanned murder 0.44 0.70 0.19 -0.03 0.04 -0.22 juvenile-rape -1.28 0.86 -0.46 0.04 0.05 0.31 juvenile-child molestation -0.83 0.90 -0.30 0.03 0.05 0.20 *Significant at p<.05

PAGE 85

80 CHAPTER 6 RESULTS: ATTRIBUTION THEORY Although the analysis so far has established a weak association between religion variables and death penalty attitudes, it must be recalled that the rela tionship may be at best indi rect through the effect of religion on variables taken from attribution theory. The following analyses address the links among religion, operationalized as theologi cal emphasis and Biblical literalne ss, attribution style, punishment philosophy, and attitudes toward capital punishment. Table 12 reports the results of two OLS regressions that model the impact of theological emphasis on the dependent variable, Biblical literalness. The ten control variables are also included in both models. Model 1 includes the six-item composite scale described in the methods section. This regression produced a significant theological emphasis effect, but in the opposite direction that we predicted (hypotheses 6a and 6b). Individuals who possessed a less forgiving theological emphasis adopted a less literal interpretation of the Bible than their more forgiving counterparts (b=-.46, p<.05). There is no evidence of such a relationship in previous research, so it is difficu lt to speculate any concep tual reason why such an anomalous finding occurred. Recall, however, that there was little variati on in the respondents scores on the forgiving theological emphasis scale. In fact, over three-fourths of the respondents had scores of 14 or lower (out of a range of 6 to 30, 14 or less constituted only the lower third of the scale). To determine if the low variation was the reason for the negative relationship between theological emphasis and Biblical literalness, we examined the effect of a four-item punitiveness scale on Biblical literalness. The four items (full descriptions of which are available in Appendix E) are related to individuals fearing Gods punishment after they have sinned, believing th at evildoers will go to hell, that God condemns sinners, and that God gets even with sinners. These items did not scale w ith the six forgiveness items that comprised the first theological emphasis scale. They di d, however, form a single factor solu tion when entered into a principal

PAGE 86

81 Table 12. OLS Regression of Biblical Literalness by Theological Emphasis, Measured by a Forgiving Theological Emphasis Scale and a Punitive Theological Emphasis Scale Model 1: Forgiving Theological Emphasis Scale DF SS MS F Model 11 7060.44 641.86 38.37* Error 624 10437 16.73 Total 635 17497 R=.404 Biblical literalness b se(b) B theological emphasis -0.46* 0.04 -0.42 race -0.19 0.12 -0.05 sex 0.08 0.34 0.01 age -0.00 0.01 -0.01 educate -0.39* 0.11 -0.12 income -0.10 0.09 -0.04 ideology 0.87* 0.19 0.16 sexroles 0.17* 0.08 0.04 abortion 0.22* 0.04 0.20 socwlfre -0.10 0.08 -0.04 police 0.08 0.05 0.05 Model 2: Punitive Theological Emphasis Scale DF SS MS F Model 11 6578.45 598.04 34.18* Error 624 10919 17.50 Total 635 17497 R=.376 Biblical literalness b se(b) B theological emphasis 0.55* 0.05 0.36 race -0.14 0.13 -0.03 sex -0.54 0.35 -0.05 age 0.01 0.01 0.01 educate -0.24* 0.12 -0.07 income -0.13 0.10 -0.05 ideology 0.83* 0.20 0.15 sexroles 0.10 0.08 0.04 abortion 0.34* 0.04 0.32 socwlfre -0.20* 0.08 -0.08 police 0.13* 0.05 0.08 *Significant at p<.05

PAGE 87

82 components factor analysis. The scale was moderately reliable, with a Cronbachs Alpha of .63. Item-toscale correlations are also provided in Appendix F. The regression results, reported in Model 2 of Table 12, supported our hypotheses (6a and 6b), as individuals who adopted a more punitive theological emphasis also interpre ted the Bible more literally than respondents with a less punitive theological emphasis (b =.55, p<.05). Thus, we are likely correct in assuming that our earlier finding was a product of low variation from a skewed sampling distribution, rather than anything substantially meaningful. The summary statistics at the top of Table 12 demonstrate that both of the overall models were statistically significant (F=38.37, p<.05, F=34.18, p<. 05 for the first and second models, respectively). The explanatory power of the independent variable, alone was moderate; when the control variables were deleted from the models, the explanatory power of theological emphasis fell from 40.4 percent to 27.6 percent in the first model and from 37.6 percent to 17.8 percent in the second model. The absence of the control variables did not diminish the models explained variance by as much as it has in previous studiestypically, the explanatory power of religion variables, alone, is only about 10 percent (Grasmick et al., 1993a; Grasmick and McGill, 1994; Applegate et al., 2000). Further, the absence of the control variables did not affect the nature of the relations hip between individuals theological emphasis and the degree to which they interpreted the Bible literally. Turning to our next hypotheses (7a and 7b), Table 13 reports the effect of Biblical literalness on attribution style, with the inclusion of the ten control variables. Much to our surprise, a more literal interpretation of the Bible was significantly related to possessing a situational attribution style (b=-.07, p<.05). The distribution of the sample on both variables was about normal, so it is unlikely that this finding is a product of a statistical anomaly. The overall re gression model of Biblical literalness on attribution style was significant (F=6.63, p<.05), but weak, as the combination of Biblical literalness and the ten control variables explained only 10.5 percent of the variance in individuals attribution styles. Interestingly, though, when we removed the control variables from the analysis, we found that, instead of the explanatory power of Biblical literalness increasing, it fell to non-significant. Thus, it appears that the negative effect of Biblical literalness on possessing a dispositional attribution style reflects a suppressor effect. Indeed, when we examined the correlations between the explanatory and response

PAGE 88

83 Table 13. OLS Regression of Attribution Style by Biblical Literalness DF SS MS F Model 11 1175.42 106.86 6.63* Error 624 10056 16.12 Total 635 11231 R=.105 attribution style b se(b) B Biblical literalness -0.07* 0.04 -0.09 race -0.15 0.12 -0.05 sex -0.11 0.33 -0.01 age 0.04* 0.01 0.12 educate -0.44* 0.11 -0.17 income -0.00 0.09 -0.00 ideology 0.36 0.19 0.08 sexroles -0.12 0.08 -0.07 abortion 0.07 0.04 0.08 socwlfre 0.44* 0.08 0.23 police -0.07 0.05 -0.06 *Significant at p<.05

PAGE 89

84 variables and the three significant control variables in the model (age, education level, and attitudes toward social welfare), we found that education was significantly and negatively associated with Biblical literalness (r=-.18, p<.05) and possessing a dispositional attribution style (r=-.14, p<.05). It appears, then, that by ignoring (rather th an controlling for) respondents education level, we infl ate the number of people with a situational attribution style who interpret the Bible literally. This association actually does not existthere is no significant relationship between the degree to which respondents interpret the Bible literally and their attribution styles Our hypotheses remain unsupported, but there is no evidence to the contrary, either, as there initially appeared to be. Although this finding does not comport with previous research on the relationship between attribution theory and Biblical litera lness (e.g., Grasmick and McGill, 1994) it is somewhat of a theoretical extension of the findings of Applegate et al. (2000). Even though Applegate et al. (2000) found that Biblical literalness was associated w ith punitiveness (which is consisten tly associated with a dispositional attribution style), they did not find the relationships they expected between Biblical literalness and possessing a punitive theological empha sis (both measures of religious fundamentalism) and the two dependent variables, capital punishment support and beliefs that individuals local courts were too lenient. They maintained, in spite of these anomalous findings, that the relationships between Biblical literalness and perceptions of a punitive God and punitive responses to convicted criminals persisted. The lack of a significant relationship between Biblical literalness and possessing a dispositional attribution style in the present study, combined with the inconsistent findings of Applegate et al. (2000) regarding the effect of fundamentalism on punitiveness are indicative of the ch anging nature of the institution of religion. Even highly fundamentalist religious individuals no longer s eem to be the extreme dogmatists that they have been reputed to be. It is also po ssible that the Bible contains both dispositional and situational attributions and that some literalists place greater emphasis on the fo rgiving aspects of the New Testament, rather than the vengeful Yahweh of the Old Testament. The data do not contain such distinctions, however, so this proposition remains speculative. Our third path (hypotheses 8a and 8b) predicted a positive relationship between possessing a dispositional attribution style a nd supporting a more punitive punishment philosophy toward convicted adult and juvenile offenders. Table 14 reports the mode l from this regression procedure. Indeed, there was

PAGE 90

85 Table 14. OLS Regression of Punishment Philosophy by Attribution Style DF SS MS F Model 11 3716.30 337.85 7.45* Error 624 28298 45.35 Total 635 32015 R=.116 punishment philosophy b se(b) B attribution style 0.29* 0.07 0.17 race 0.08 0.20 0.02 sex 0.08 0.56 0.01 age -0.03 0.02 -0.05 educate -0.74* 0.19 -0.16 income -0.07 0.16 -0.02 ideology 0.81* 0.32 0.11 sexroles 0.29* 0.13 0.10 abortion -0.22* 0.06 -0.15 socwlfre 0.36* 0.14 0.11 police 0.10 0.09 0.05 *Significant at p<.05

PAGE 91

86 a moderate effect of dispositiona l attribution style on having a punitiv e punishment philosophy (b=0.29, p<.05). Consistent with our hy potheses, individuals who adopted a more classical position on crime causation (i.e., believed that crime was a product of free will) were more likely to support punitive measures to deal with convicted offenders. Conve rsely, respondents who were more reflective of a positivist position on crime control (i.e., crime is caused by environmental forces) were more likely to support a rehabilitative approach to convicted offenders This overall model was statistically significant, but weak (R=.116). Deleting the control variables from the model reduced the explanatory power of the independent variable to 4.7 percent. Finally, the last direct relationship we hypothesized (9a and 9b) was that possessing a more punitive philosophy toward th e correctional systems treatment of o ffenders would be predictive of higher levels of support for capital punishment among respondents. The OLS regression analysis indicated that this was indeed the case (see Table 15 for the results of this analysis). Punishment philosophy, in fact, was a relatively strong predictor of capital punishment support (b=0.68, p<.05). The overall model was significant (F=21.60, p<.05) and moderately strong, explaining 27.6 percent of the variation in death penalty attitudes. Removing the control variables from the analysis left an explained variance of 20.2 percent for the effect of punishment philosophy on capital punishment. The data have demonstrated some support for attribution theory; however, a comprehensive interpretation of the findings is contingent upon the indirect relationships between the variables. We conducted OLS regression for the six indirect paths between variables in our model: the effects of the two theological emphasis scales on attr ibution style, punishment philosophy, and death penalty attitudes, the effects of Biblical literalness on punishment philosophy and death penalty attitudes, and the effect of attribution style on death penalty attitudes. Although we found significant eff ects of possessing a punitive theol ogical emphasis (the four-item scale comprised of items reflecting a punitive theol ogical emphasis) on punishme nt philosophy (b=0.49, p<.05) and death penalty support (b=0.47, p<.05) in the expected directions, these effects were not as strong as the direct effect of th eological emphasis on Biblical literaln ess. There was not a significant relationship between possessing a punitive theological emphasis and attribution style. Interestingly and contrary to our findings of the effect of possessing a forgiving theological emphasis on Biblical literalness,

PAGE 92

87 Table 15. OLS Regression of Death Pena lty Attitudes by Punishment Philosophy DF SS MS F Model 11 27733 2521.20 21.60* Error 624 72832 116.72 Total 635 100566 R=.276 death penalty attitudes b se(b) B punishment philosophy 0.68* 0.06 0.38 race -0.48 0.32 -0.05 sex 1.33 0.89 0.05 age 0.00 0.04 0.00 educate -1.19* 0.30 -0.15 income 0.48 0.25 0.07 ideology 1.79* 0.51 0.13 sexroles -0.19 0.21 -0..04 abortion -0.43* 0.10 -0.17 socwlfre 0.80* 0.22 0.14 police 0.09 0.14 0.02 *Significant at p<.05

PAGE 93

88 we found that the indirect effects were more ali gned with our expectations. Possessing a forgiving theological emphasis was signi ficantly associated with possessing a situational attribution style, as well as opposition to the death penalty. Neither of these effects, however, was as strong as the direct effect (in the opposite direction we predicted) of a progressive theological emphasis on Biblical literalness. Possessing a progressive theological emphasis was not significantly associated with individuals correctional treatment philosophies. Recall that the relationship between Biblical literalness and attribution style was significant, but inconsistent with our hypotheses. Biblical literalness also significan tly affected punishment philosophy (b=0.19, p<.05) and support for capital punishment (b=0.42, p<.05), but notice that these effects are positive. Finally, we found that attr ibution style did not a ffect attitudes toward the death penalty, thus lending credence to our decision to separate puni shment philosophy and attitudes toward capital punishment. The direct and indirect relationships between theo logical emphasis, Biblical literalness, attribution style, punishment philosophy, and attit udes toward the death penalty larg ely supported our hypotheses. In fact, Biblical literalness was the only variable that displayed effects inconsistent with our expectations. First, it did not produce a significant effect on individuals attribution styles, such that individuals who possessed a more literal interpretation of the Bible were not more likely to adopt a dispositional attribution style, and individuals who interpreted the Bible less literally were not more likely to adopt a situational attribution style. Second, Biblical literalness was significantly related to attitudes toward capital punishment. These seemingly anomalous relationships are likely a reflection of two different types of right to lifers in conservative faith groups: first, t hose individuals who are interested in the protection of all lives (e.g., are opposed to abortion, capital punishment, euthanasia, etc.) and second, those individuals who are interested in the protection of innocent lives (e.g., are opposed to abortion and euthanasia, but support the death penalty). These individuals theological emphases are what segregates them, thereby producing the significant relationships we found in the attribution theory analysis. Individuals who adopt a more conservative theological emphasi s adopted a more literal interpretati on of the Bible; they were more punitive, and more supportive of capital punishment than individuals with a more progressive theological emphasis. However, the negative relationship between Biblical literalness and attribution style seems to

PAGE 94

89 suggest that even members of conservative faith gr oups may adopt a more positivist approach to crime causation, allowing for the possible influence of external factors, such as poor home life, poverty, etc. These individuals still tend to be more punitive than i ndividuals who adopt a more progressive theological emphasis, but nevertheless, they seem to be more liberal than they have been reputed to be.

PAGE 95

90 CHAPTER 7 DISCUSSION AND CONCLUSION The Race-Religion-Death Pena lty Attitudes Relationship Several pieces of research have advanced the st udy of the effects of race and religion on attitudes toward criminal justice policies, including the death penalty. Much of the earlier research on this topic is atheoretical and fails to develop th e different dimensions of religion. More recent investigations, however, have utilized theoretical frameworks to analyze the race-religion-death penalty attitudes relationship, and they have broadened the conceptualization of religi on beyond the typical crude distinctions. This more recent body of research has revealed several trends. First, members of conservative Protestant faith groups are more retributive toward adult and juvenile offenders; they believe that the courts are too lenient; and they are more supportive of capital punishment than members of other faith groups (Young, 1992; Grasmick et al., 1993a; Grasmick et al., 1993b; Grasmi ck and McGill, 1994; Britt, 1998; Applegate et al., 2000). Second, when the effects of both race and relig ion are considered, signifi cant interactive effects emerge, with white fundamentalists exhibiting the hi ghest level of support for capital punishment and black non-fundamentalists displaying the lowest level of support for capital punishment (Young, 1992; Britt, 1998). Third, attribution theory ha s been successfully utili zed as a theoretical framework to explain the relationship between individuals attribution styles and their punitive or rehabilitative orientations toward offenders (Cullen et al., 1983; Gr asmick et al., 1993a; Grasmick et al., 1993b; Grasmick and McGill, 1994). The present study aimed to provide a more comprehensive explanation of public attitudes toward capital punishment, first by determining the degree to which faith group, religiosity, and race affected public attitudes and second, by usi ng attribution theory to provide the process through which religion affects death penalty attitudes. In spite of support we found for the hypotheses regarding membership in different faith groups and races, the conclusion remains that we found little support for the faith groupreligiosity-capital punishment and the faith group-religiosity-race-capita l punishment attitudes

PAGE 96

91 relationships. Our findings, then, that whites and members of more conservative religious traditions, particularly conservative Protestants, were more likely to support capital punishment than non-whites and members of more liberal religious traditions, but that there were limited significant religiosity by race by faith group effects were inconsistent with the previous trends demonstrated in the literature and somewhat puzzling. Clearly, there is something about fundamentalist Protestant religions that either draws people with certain qualities to them (in which case their commitment to and involvement with the group would not matter if they had already adopted the types of be haviors and beliefs that the group required) or changes the behaviors and beliefs of their members even without a high level of commitment and involvement from them. Criminological research has b een lacking a processual explan ation for how religion shapes individuals attitudes and behaviors related to social and political issues. The attitudinal research on religion and punitiveness thus far has tended to focus on the effect of possessing conservative religious beliefs or punitive orientations (Young, 1992; Grasmi ck et al., 1993a; Grasmick et al., 1993b; Grasmick and McGill, 1994; Britt, 1998). In focusing on the funda mentalist and proscriptive elements of religion, criminologists have risked neglecting the possible eff ects of other dimensions of religion, including more forgiving moral messages and even the level at which religion is measured (e.g., faith group, congregation, etc.) without which a comprehensive understanding of its effects on i ndividuals attitude s and behaviors cannot be understood. Theological emphasis is a different and arguably more precise dimension of religion, as it is a concrete reflection of individuals perceptions of G od, whereas faith group is a more structural level measure of religion that may or may not be reflectiv e of individuals personal beliefs, because it imposes on members a theological position. In other words, there is likely more within group than between group variation among members of faith groups. Faith group, then, regardless of how it is measured, is a crude measure of affiliation that presumes sameness among its members; divisions within faith groups are not reflected by this measurement. Our test of attribution theory, while addressing many of the puzzling conclusions left by our initial analyses, also left many unanswered questions. Our findings, as well as Applegate et als (2000), are moving in a new direction theoretica lly, as both pieces of research seem to indicate that the differences

PAGE 97

92 between fundamentalists and non-fundamentalists are not as stark as they have been treated in previous research, particularly the body of research conduc ted by Grasmick and his colleagues (1992, 1993a, 1993b, 1994). Applegate et al. (2000) found no relationship between theological emphasis or Biblical literalness and attitudes toward capital punishme nt; however, they did find signifi cant effects of both of these measures of fundamentalism on general punitiveness to ward offenders, as well as on punitiveness toward a hypothetical offender. Furthermore and consistent with our earlier statement regarding neglected dimensions of religion in social scientific research, Applegate et al. (2000) also found that individuals who adopted a more forgiving theologi cal emphasis were significantly less punitive than more fundamentalist individuals: they were less likely to support capital punishment, less likely to agree that their local courts were too lenient, less punitive toward offenders in general, and less punitive toward a hypothetical offender. Our findings were consistent with Applegate et als (2000) in that we, too, found that individuals who held less strongly to fundamentalist religious te nets (i.e., possessed a more progressive theological emphasis) were significantly less likely to support the death penalty than their fundamentalist counterparts. Our analyses, however, detailed how this relations hip between religion and capital punishment attitudes developed, and it established a more precise theoretical framework that explained the direct and indirect effects between the variables in the processual model derived from attribution theory. The results from the path model in the present study support findings in previous studies that found a relationship between conserva tive religious beliefs, a dispositiona l attribution style, and punitive sentiments toward convicted offenders, including support for capital punishment. More importantly, however, the present study is among the first pieces of research to examine the effect of more progressive elements of religion and shed light on individuals more compassionate sentiments toward convicted offenders. Indeed, we found that possessing a less forgiving theological emphasis was associated with a less literal interpretation of the Bible. Further, we utilized several situational items in our attribution scale, and found that individuals who possessed more of a s ituational attribution style were more supportive of rehabilitative correctional policies than individuals who possessed more of a dispositional attribution style, and finally that these individuals (who were more supportive of rehabilita tive policies) were less likely to support the death penalty than individuals who favored more punitive correctional policies.

PAGE 98

93 Theoretical and Policy Implications The results in the present study demonstrate that the conceptualization of the relationship between attitudes toward criminal justice policies and religi on should extend beyond religi on as a basis for punitive sentiments. Certain religious values may contribut e to feelings of compassion and understanding toward convicted offenders, which would likely be associated with less punitiv e policies to deal with them. Moreover, no existing criminological theory, nor any other theory which has been applied to the study of religion and criminal justice attitudes, confines its propositi ons to religion as an institu tion of social control. This presumption has been adopted by social scientists, and it has created a void in the extant literature on religion as a predictor of criminal justice attitudes. Clearly, the study of public attitude s is very important. Politicians use public appeal as one of the major forces behind the justification of get tough measures, which empirically appear to be largely ineffective. As we develop a more sophisticat ed understanding of why individuals support or oppose criminal justice policies, we will be in a better position to utilize different mediums and different types of presentations of information to inform individuals about the policies that they are supporting. The United Statess treatment of its worst offende rs (murderers) is a reflection of its level of civilization. Furthermore, we are in the midst of an interna tional battle against human rights viol ations, among which one of the focal issues is the institution of capital punishment. Understanding why individua ls support and oppose the taking of a human lifeeven if it is a person who has been convicted of murderand understanding how they arrive at these positions is one of fundamental importance in developing an intellectual dialogue about the death penalty, devoid of false rhetoric and in possibly swaying the political and constitutional pendulums. Religion is one of the most influential institutions in individuals lives and has consistently been viewed as one of the reasons behind public punitiveness. While this may certainly be the case, it is equally important to recognize that religion may also be a major reason why individuals oppose punitive correctional policies, including the death penalty. Many individuals across different faiths perceive God as a loving and understanding God, and in trying to follow His example, believe that there is hope for redemption for all people and, acco rdingly, support rehabilitative approaches to convicted offenders.

PAGE 99

94 Future Research This study contributed to an important dialogue about the ba ses for attitudes toward capital punishment and how these attitudes are formed. It is necessary, however, in order to expand our understanding of this topic, to acknowledge the limitati ons of our research and provide ideas for future investigations. One of the weaknesses in this study was its geographical limitation; our sample was drawn from a single county in Florida. The differences between our findings and the findings of Grasmick and his colleagues could very well be a function of location. Grasmicks respondents were residents of Oklahoma, which is in the middle of the Bible Belt and, not surprisingly, more fundamentalist than our sample from the Tampa area, which is more cosmopolitan. Our study utilized more in-depth measures of re ligion, attribution style, general punitiveness, and capital punishment attitudes than had been adopted in pr evious research. Future research should continue to develop even more sophisticated measures, particularly of religion. The concept of a progressive theological emphasis should be further developed, a nd other dimensions of religion should be created and comprehensively considered in one analytic strategy. Perhaps the most important consideration that our study neglected was the level at which we measured faith group. Instead of examining separate faith groups, future endeavors should explore smaller units of analysis, such as congregation, to better assess religious-based between group differe nces on attitudes toward capital punish ment. Future research should also attempt to obtain larger sample sizes, as it would facilitate a more confident interpretation of racial and ethnic religious-based differences toward social policies like capital punishment. With regard to the study of religion and cap ital punishment attitudes, future endeavors should consider the effect of religion on not only death penalty attitudes, but other right to life issues, as well. Just as capital punishment may be viewed as a criminal justice policy, it may also be viewed as a right-to-life issue. The examination of capital punishment attitude s in such a context will likely shed new light on theoretical applications to the study of religion and crime. Finally, due to the cross-sectional nature of our data, it was impossible to definitively determine whether the causal order of the relationships we pred icted was accurate. Longit udinal data on religion and attitudes toward social and political issues, such as the death penalty, would be very enlightening.

PAGE 100

95 Moreover, they would provide a much clearer picture of the utility of the theories that have been applied to this body of research thus far.

PAGE 101

96 REFERENCES Coker v. Georgia 433 U.S. 584 (1977). Crampton v. Ohio 402 U.S. 183 (1971). Enmund v. Florida 458 U.S. 782 (1982). Ford v. Wainwright 477 U.S. 399 (1986). Furman v. Georgia 408 U.S. 238 (1972). Gregg v. Georgia 428 U.S. 153 (1976). Harris v. Alabama 513 U.S. 504 (1995). Herrerra v. Collins 113 S.Ct. 853 (1993). Jurek v. Texas 428 U.S. 262 (1976). Lockett v. Ohio 438 U.S. 586 (1978). McCleskey v. Kemp 481 U.S. 279 (1987). McGautha v. California 402 U.S. 183 (1971). Morgan v. Illinois 112 S.Ct. 2222 (1992). ODell v. Netherland 117 S.Ct. 1969 (1997). Penry v. Lynaugh 492 U.S. 302 (1989). Proffitt v. Florida 428 U.S. 242 (1976). Roberts v. Louisiana 428 U.S. 325 (1976). Robinson v. California 370 U.S. 660 (1962). Stanford v. Kentucky 492 U.S. 361 (1989). Sumner v. Schuman 483 U.S. 66 (1987). Thompson v. Oklahoma 487 U.S. 815 (1988). Tison v. Arizona 481 U.S. 137 (1987). Trop v. Dulles 356 U.S. 86 (1958). Weems v. U.S. 217 U.S. 349 (1910). Wilkins v. Missouri 492 U.S. 361 (1989). Witherspoon v. Illinois 391 U.S. 510 (1968). Woodson v. North Carolina 428 U.S. 280 (1976). Zant v. Stephens 462 U.S. 862 (1983).

PAGE 102

97 Applegate, B.K., Cullen, F.T., Fisher, B.S., &Vander Ven. (2000). Forgiveness and fundamentalism: Reconsidering the relationshi p between correctional attitudes and religion. Criminology, 38, 719-754. 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, 202-209. Bedau, H.M. (1997). The death penalty in America: Current controversies City: Oxford University Press. Beeghley, L., Bock, E.W., &Cochran, J.K. (1990). Religious change and alcohol use: An application of reference group and socialization theory. Sociological Forum, 5, 261-278. Blumstein, A. & Cohen, J. (1980). Sentencing of convicted offenders: An analysis of the publics view. Law and Society Review, 14, 223-261. Bock, E.W., Cochran, J.K., & Beeghley, L. (1987). Moral messages: The relative influence of denomination on the religiosity-alcohol relationship. Sociological Quarterly, 28, 86-105. Bohm, R.M. (1991). American death penalty opinion, 1936-1986: A critical examination of the Gallup polls. In R.M. Bohm (Ed.), The death penalty in America: Current Research Cincinatti, OH: Anderson. Borg, M.J. (1997). The southern subculture of punitiveness? Regional variation in support for capital punishment. Journal of Research in Crime and Delinquency, 34, 25-45. Bowers, W. (1993). Capital punishment and contemporary values: Peoples misgivings and the courts misperceptions Law and Society Review, 27 Britt, C. (1998). Race, religion, and support for the death penalty. Justice Quarterly, 15, 175-191. Carroll, J. (2000). Poll Summary: Where America stands on the death penalty. www.gallup.com Carroll, J. & Payne, J.W. (1977). Crime seriousness, recidivism risk, and causal attributions in judgements of prison term by students and experts. Journal of Applied Psychology, 62, 595-602. Clarke, L., Beeghley, L., & Cochran, J.K. (1990). Religiosity, social class, and alcohol use: An application of reference group theory. Sociological Perspectives, 33, 201-218. Cochran, J.K., Beeghley, L., & Bock, E.W. (1988). Religiosity and alcohol behavior: An exploration of reference group theory. Sociological Forum, 3, 256-276. Cochran, J.K., Chamlin, M.B., Beeghley, L., Harnden, A., & Blackwell, B.S. (1996). Religious stability, endogamy, and the effects of persona l religiosity on attitudes toward abortion. Sociology of Religion, 57, 291-309. Cochran, J.K., Chamlin, M.B., & Seth, M. (1994). Deterrence or brutalization?: An impact and assessment of Oklahomas return to capital punishment. Criminology, 32 Cook, K.J. (1998). A passion to punish: abortion opponents who favor the death penalty. Justice Quarterly, 15, 329-346. Coyne, R. &Entzeroth, L. (1994). Capital punishment and the judicial process Durham, NC: Carolina Academic Press.

PAGE 103

98 Cullen, F.T., Clark, G.A., Cullen, J.B., & Mathers, R.A. (1985). A ttribution, salience, and attitudes toward criminal sanctioning. Criminal Justice and Behavior, 12, 305-331. Cullen, F.T., Cullen, J.B., & Wozniak, J.F. (1988). Is rehabilitation dead? The myth of the punitive public. Journal of Criminal Justice, 16, 303-317. Cullen, F.T., Skovron, S.E., Scott, J.E., & Burton, V.S., Jr. (1990). Public support for correctional treatment: The tenacity of rehabilitative ideology. Criminal Justice and Behavior, 17, 6-18. The Death Penalty Information Center. (2001a). The death penalty in 2000: Year end report. www.deathpenaltyinfo.org The Death Penalty Information Center. (2001b). History of the death penalty, part I. www.deathpenaltyinfo.org The Death Penalty Information Center. (2001c). Summaries of recent poll findings. www.deathpenaltyinfo.org The Death Penalty Information Center. (2001d). Statements and resolutions from religious organizations and leaders. w ww.deathpenaltyinfo.org. Dieter, R.C. (1993). Sentence for life: American s embrace alternatives to the death penalty. The Death Penalty Information Center, 3-5. Dolan, J.P. & Deck, A.F. (1994). Hispanic catholic culture in the U.S.: issues and concerns Notre Dame: University of Notre Dame Press. Durham, A.M. (1988). Judgements of appropriate punishment: The effects of question type. Journal of Criminal Justice, 17, 75-85. Durham, A.M., Elrod, P.H., & Kinkade, P.T. (1996). Public support for the death penalty: Beyond Gallup. Justice Quarterly, 13, 705-736. Ellsworth, P.C. & Gross, S.R. (1994). Harden ing of attitudes: Ameri cans 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 and Delinquency, 29, 116-169. Flanagan, T.J. & McLeod, M. (1983). Sourcebook of Criminal Justice Statistics 1982 Washington, DC: U.S. Department of Justice. Fox, J., Radalet, M., & Bonsteel, J. (1991). Death penalty opinion in the post-Furman years. New York University Review of Law and Social Change, 18, 499-528. Gallup Organization. (2000). Gallup poll topics: A-Z, death penalty. www.gallup.com Gelles, R. & Strauss, M. (1976). Family experience and public support of the death penalty. In H.A. Bedau & C.M. Pierce (eds.), Capital Punishment in the U.S. New York: AMS. Gibbs, J.P. (1978). The death penalty, retribution, and penal policy. Journal of Criminal Law and Criminology, 69, 291-299.

PAGE 104

99 Gottfredson, S.D. & Taylor, R.B. (1985). Attitude s of correctional policy makers and the public. In S.D. Gottfredson & S. McConville (eds.), Americas correctional crisis: Prison populations and public policy (pp. 57-75). New York: Greenwood. Grasmick, H.G., Bursik, R.J., & Blackwell, B.S. (1993b). Religious beliefs and public support for the death penalty for juveniles and adults. Journal of Crime and Justice, 16, 59-86. Grasmick, H.G., Cochran, J.K., Bursik, R.J., & Kimpel, M. (1993a). Religion, punitive justice, and support for the death penalty. Justice Quarterly, 10, 289-314. Grasmick, H.G., Davenport, E., Chamlin, M.B., & Bursik, R.J. (1992). Protestant fundamentalism and the retributive doctrine of punishment. Criminology, 30, 21-45. Grasmick, H.G. & McGill, A.L. (1994). Religi on, attribution style, and punitiveness toward juvenile offenders. Criminology, 32, 23-46. Hawkins, D.F. (1981). Causal attribution and punishment for crime. Deviant Behavior, 2, 207230. Heider, F. (1958). The psychology of interpersonal relations New York: John Wiley & Sons. Jones, J. (2000). Slim majority of Americans think death penalty applied fairly in this country. www.gallup.com Liebman, J.S. (2000). The overproduction of death. Columbia Law Review, 100, 2030. McCorkle, R.C. (1993). Research note: Punish and rehabilitate? Public attitudes toward six common crimes. Crime and Delinquency, 39, 240-252. Ofshe, R. (1972). Reference conflict and behavior. In J. Berger, M. Zelditch, & B. Anderson (eds.), Sociological theories in progress (pp. 88-116). Boston: Houghton Mifflin. Paternoster, R. (1993). Assessing Capriciousness in capital cases. Law and Society Review, 27, 111-124. Pillsbury, S.H. (1989). Understanding pe nal reform: The dynamic of change. Journal of Criminal Law and Criminology, 80, 726-780. Radelet, M. (1992). In spite of innocence: erroneous convictions in capital cases Boston: Northeastern University Press. Skovron, S.E., Scott, J.W., & Cullen, F.T. (1989). Prison crowding: Public attitudes toward strategies of population control. Journal of Research in Crime and Delinquency, 25, 150-169. Smith, T.W. (1990). Classifying Protestant denominations. Review of Religious Resarch, 31, 225245. Stinchcombe, A.L., Adams, R., Heimer, C.A., Schlepple, K.L., Smith, T.W., & Taylor, D.G. (1980). Crime and punishment: Changing attitudes in America San Fransisco: Jossey-Bass. Thomas, C.W. & Foster, S.C. (1975). A sociological perspective on support for capital punishment. American Journal of Orthopsychiatry, 45, 641-657. Tyler, T.R. & Weber, R. (1982). Support for the death penalty: Instrumental response to crime or symbolic attitude? Law and Society Review, 17, 21-45.

PAGE 105

100 Vidmar, N. & Ellsworth, P.C. (1974). Public opinion and the death penalty. Stanford Law Review, 26, 1245-1270. Warr, M. & Stafford, M. (1984). Public goals of punishment and support for the death penalty. Journal of Research in Crime and Delinquency, 21, 95-111. Washington, J.R. (1964). Black religion: the negro and Christianity in the United States Boston: Beacon Press. Wood, M. & Hughes, M. (1984). The moral basis of moral reform: Status discontent versus culture and socialization as explanations of anti-pornography social movement adherence. American Sociological Review, 49, 86-99. Young, R.L. (1991). Race, conceptions of crime and justice, and support for the death penalty. Social Psychology Quarterly, 54, 67-75. Young, R.L. (1992). Religious orientation, race, and support for the death penalty. Journal for the Scientific Study of Religion, 31, 76-87. Zeisel, H. & Gallup, A.M. (1989). Death penalty sentiment in the U.S. Journal of Quantitative Criminology, 5, 285-296.

PAGE 106

101 APPENDICES

PAGE 107

102 APPENDIX A JURY POOL SURVEY Fall, 1999 University of South Florida Department of Criminology Tampa, FL 33620 Thank you for participating in our survey. Your answers to the following questions will give us information about your attitudes regarding crime and justice. Please complete the questionnaire as follows. 1. Please answer the questi ons in the order they appear. 2. Circle only one answer for each question unless directed otherwise. 3. There are no right or wrong answerswe are interested in your opinions. 4. Please DO NOT write your name on the questionnaire. 5. Your responses are anonymous and will not be tied to you in any way. 6. You may skip any question you do not wish to answer. 7. You may stop filling out the questionnaire at any time. 8. Please ignore the numbers next to the response options; they are for computer coding purposes only. Thank you, again, for your participation! Continued on next page

PAGE 108

103 Appendix A continued A. PLEASE ANSWER THE FOLLOWING QUESTIONS ABOUT YOU AND YOUR BACKGROUND. CIRCLE WHICH APPLIES. 1. Sex 0. Female 1. Male 2. Race 1. American Indian or Alaskan Native (a person having origins in any of the original peoples of North and South America, including Central America, and who maintain tr ibal affiliation or community attachment). 2. Asian (a person having origins in any of the original peoples of the Far East, Southeast, Asia or the Indian subcontinent, including, for example, Cambodia, China, India, Japan, Korea, Malaysia, Pakistan, the Philippine Islands, Thailand, and Vietnam). 3. Black or African American (a person having origins in any of the black racial groups or Africa. This term includes Haitian or Negro). 4. Native Hawaiian or other Pacific Islander (a person having origins in any of the original peoples of Hawaii, Guam, Samoa, or other Pacific Islanders). 5. White (a person having origins in any of the original peoples of Europe, the Middle East, or North America). 3. Are you of Hispanic, Latino, or Spanish origin? (includes a person of Cuban, Mexican, Puerto Rican, South or Central American, or other Spanish culture or race, regardless of race) 1. yes 0. no 4. What is your age? _____ 5. Do you have any children (biological, stepchildren, adopted, etc.)? 1. Yes 0. No If so, what are their ages? ___ ___ ___ ___ ___ ___ ___ ___ 6. How much education have you completed? 1. grade school or less 2. some high school 3. high school graduate 4. 1 or more years of technical, vocational, or trade school 5. some college 6. college graduate 7. 1 or more years of graduate, law, or medical school 8. advanced degree (e.g., Masters, Ph.D., J.D., M.D., etc.) Continued on next page

PAGE 109

104 Appendix A continued 7. What is the combined income of everyone in your household? 1. under $10,000 2. $10,000-$19,000 3. $20,000-$29,000 4. $30,000-$39,000 5. $40,000-$49,000 6. $50,000-$69,000 7. $70,000-$89,000 8. $90,000-$119,000 9. $120,000-$149,000 10. $150,000+ 8. In what region of the country did you grow up? 1. North 2. East 3. South 4. West 5. Midwest 6. Other: _______________ 7. Dont know 9. Which of the following best describes the area in which you currently live? 1. the City of Tampa 2. a suburb of Tampa (e.g. Brandon, Temple Terrace, Carrolwood, North Tampa) 3. another Hillsborough County community (e.g. Sun City, Wimouma, Plant City) 4. rural Hillsborough 5. Other: ____________________ 10. How are you registered as a voter? 1. Republican 2. Democrat 3. Independent 4. Other:___________________ 5. I am not registered. 11. Circle the position which best describes your social and political views. 1. Very-----2. somewhat --3. middle of --4. somewhat -----5. very liberal liberal the road conservative conservative 12. With what religion, if any, do you identify? 1. Catholicism 2. Judaism 3. Protestantism 4. None 5. Other: _____________________ Continued on next page

PAGE 110

105 Appendix A continued 13. If you are Protestant, what denomination are you? 1. Baptist 2. Episcopalian 3. Lutheran 4. Methodist 5. Presbyterian 6. Fundamentalist 7. Other: ___________________ 14. Were your above answers in questions 12 and 13 the faith that you grew up in? 1. Yes _____ 0. No _____ 9. Not applicable to me _____ 15. If you answered no to number 15, what faith did you grow up in? Please be specific (e.g. Southern Baptist, Roman Catholic, Conservative Judaism, none, etc.) _____________________ B. THE NEXT GROUP OF QUESTIONS ASKS YOU ABOUT YOUR EXPERIENCE WITH AND FEAR OF CRIME. 16. In the past twelve months, have you been the victim of a crime? Check all that apply. 1. Violent ________ 2. Property _______ 17. How afraid are you of being a victim of property crime? 1. Not at all afraid 2. Somewhat afraid 3. Afraid 4. Very afraid 18. How afraid are you of being a victim of violent crime? 1. Not at all afraid 2. Somewhat afraid 3. Afraid 4. Very afraid 19. Do you believe that you will be a victim of a property crime in the future? 1. It is very unlikely. 2. It is somewhat unlikely. 3. It is somewhat likely. 4. It is very likely. 20. Do you believe that you will be a victim of a violent crime in the future? 1. It is very unlikely. 2. It is somewhat unlikely. 3. It is somewhat likely. 4. It is very likely. Continued on next page

PAGE 111

106 Appendix A continued C. THE FOLLOWING QUESTIONS ASK ABOUT YOUR VIEWS ON VARIOUS SOCIAL AND POLITICAL ISSUES. 21. Women should have an equal role with men in running businesses. 1. strongly agree 2. agree 3. uncertain 4. disagree 5. strongly disagree 22. Women should have an equal ro le with men in running the military. 1. strongly agree 2. agree 3. uncertain 4. disagree 5. strongly disagree 23. Women should have an equal role with men in running the family. 1. strongly agree 2. agree 3. uncertain 4. disagree 5. strongly disagree 24. Abortion should be allowed under any circumstances. 1. strongly agree 2. agree 3. uncertain 4. disagree 5. strongly disagree 25. Abortion should be allowed when a womans health or life is in danger. 1. strongly agree 2. agree 3. uncertain 4. disagree 5. strongly disagree 26. Abortion should be allowed when the fetus health or life is in danger. 1. strongly agree 2. agree 3. uncertain 4. disagree 5. strongly disagree Continued on next page

PAGE 112

107 Appendix A continued 27. Abortion should be allowed when the pregnancy is the result of incest or rape. 1. strongly agree 2. agree 3. uncertain 4. disagree 5. strongly disagree 28. Abortion should be allowed when the mother cannot afford to raise the child. 1. strongly agree 2. agree 3. uncertain 4. disagree 5. strongly disagree 29. Abortion should not be allowed under any circumstances. 1. strongly agree 2. agree 3. uncertain 4. disagree 5. strongly disagree 30. The government should make every effort to improve the social and economic condition of women a nd other racial and ethnic minorities through affirmative action policies. 1. strongly agree 2. somewhat agree 3. uncertain 4. somewhat disagree 5. strongly disagree 31. The government should make every effort to improve the social and economic condition of women and other raci al and ethnic minorities through welfare programs. 1. strongly agree 2. somewhat agree 3. uncertain 4. somewhat disagree 5. strongly disagree Continued on next page

PAGE 113

108 Appendix A continued D. THE NEXT GROUP OF QUESTIONS ASKS ABOUT YOUR RELIGIOUS PRACTICES AND BELIEFS. We recognize that these questions refer to statements often associated with Judeo-Christia n religious beliefs and, therefore, may not reflect your personal beliefs. If you are uncomfortable responding to these statements, feel free to check the code after each statement that indicates, Ratin g this question is not appropriate given my personal beliefs. 32. In a typical month (4 weeks long), how many times do you attend worship services? 0 1 2 3 4 5 or more 9. Rating this question is not appropriate given my personal beliefs. 33. How many church-related groups or orga nizations do you belong to (include Bible studies, youth groups, educative activ ities, choral groups, etc.)? _____ 9. Rating this question is not a ppropriate given my personal beliefs. 34. How often do you take part in the activitie s and organizations of a church or place of worship other than attending services? 0. never 1. 1-2 times per year 2. 1-2 times per month 3. 1-2 times per week 4. 3-5 times per week 5. daily/almost daily 6. more than once a day 9. Rating this question is not a ppropriate given my personal beliefs. 35. In a typical month, how often do you pray outside of worship services? (do not include table blessings). 4. more than once a day 3. daily/almost daily 2. once or twice a week 1. once or twice a month 0. never 9. Rating this question is not appr opriate given my personal beliefs. 36. Religion is a very important part of my life. 5. strongly agree 4. somewhat agree 3. uncertain 1. somewhat disagree 1. strongly disagree 9. Rating this question is not appr opriate given my personal beliefs. Continued on next page

PAGE 114

109 Appendix A continued 37. I would describe myself as very religious. 5. strongly agree 4. somewhat agree 3. uncertain 2. somewhat disagree 1. strongly disagree 9. Rating this question is not appr opriate given my personal beliefs. 38. When I have decisions to make in my everyday life, I usually try to find out what God wants me to do. 5. strongly agree 4. somewhat agree 3. uncertain 2. somewhat disagree 1. strongly disagree 9. Rating this question is not appr opriate given my personal beliefs. 39. Religion influences how I live my life. 5. strongly agree 4. somewhat agree 3. uncertain 2. somewhat disagree 1. strongly disagree 9. Rating this question is not appr opriate given my personal beliefs. 40. In a typical month, how often do you have quiet time that focuses on religious or spiritual development (e.g., read religious material, meditate, etc.) 4. more than once a day 3. daily/almost daily 2. once or twice a week 1. once or twice a month 0. never 9. Rating this question is not appr opriate given my personal beliefs. 41. I believe I have had a re ligious born-again experience. 1. yes 0. no 9. Rating this question is not appr opriate given my personal beliefs. 42. I believe everyone must have a bor n-again experience in order to be saved. 5. strongly agree 4. somewhat agree 3. uncertain 2. somewhat disagree 1. strongly disagree 9. Rating this question is not appr opriate given my personal beliefs. Continued on next page

PAGE 115

110 Appendix A continued 43. After I do something wrong, I fear Gods punishment. 5. strongly agree 4. somewhat agree 3. uncertain 2. somewhat disagree 1. strongly disagree 9. Rating this question is not appr opriate given my personal beliefs. 44. People who do evil in this world will eventually suffer in Hell. 5. strongly agree 4. somewhat agree 3. uncertain 2. somewhat disagree 1. strongly disagree 9. Rating this question is not appr opriate given my personal beliefs. 45. In order to receive Gods forgiveness, it is important that we forgive those who sin against us. 5. strongly agree 4. somewhat agree 3. uncertain 2. somewhat disagree 1. strongly disagree 9. Rating this question is not appr opriate given my personal beliefs. 46. It is important that we hate the sin, but love the sinner. 5. strongly agree 4. somewhat agree 3. uncertain 2. somewhat disagree 1. strongly disagree 9. Rating this question is not appr opriate given my personal beliefs. 47. God teaches us that even if someone has lived a life of crime, s/he should be forgiven for their offenses if they are truly repentant (sorry). 5. strongly agree 4. somewhat agree 3. uncertain 2. somewhat disagree 1. strongly disagree 9. Rating this question is not appr opriate given my personal beliefs. 48. The God I believe in condemns sinners. 5. strongly agree 4. somewhat agree 3. uncertain 2. somewhat disagree 1. strongly disagree 9. Rating this question is not appr opriate given my personal beliefs. Continued on next page

PAGE 116

111 Appendix A continued 49. The God I believe in sympathizes with people. 5. strongly agree 4. somewhat agree 3. uncertain 2. somewhat disagree 1. strongly disagree 9. Rating this question is not appr opriate given my personal beliefs. 50. The God I believe in understands when people sin. 5. strongly agree 4. somewhat agree 3. uncertain 2. somewhat disagree 1. strongly disagree 9. Rating this question is not appr opriate given my personal beliefs. 51. The God I believe in gets back at sinners. 5. strongly agree 4. somewhat agree 3. uncertain 2. somewhat disagree 1. strongly disagree 9. Rating this question is not appr opriate given my personal beliefs. 52. The God I believe in loves people regardless of their behavior and beliefs. 5. strongly agree 4. somewhat agree 3. uncertain 2. somewhat disagree 1. strongly disagree 9. Rating this question is not appr opriate given my personal beliefs. E. BELOW IS A LIST OF BIBLE VERSES. PLEASE INDICATE THE EXTENT TO WHICH YOU BELIEVE EACH SHOULD BE INTERPRETED LITERALLY. We recognize that the questions in this section refer to statements often associated with Judeo-Christia n religious beliefs and, therefore, may not reflect your personal beliefs. If you are uncomfortable responding to these statements, feel free to check the code after each statement which indicates, Rating this question is not appropriate given my personal beliefs. 53. Anyone who hits a man so hard that he dies shall surely be put to death. 4. the true/literal words of God 3. the inspired or interpreted words of God 2. more or less influenced by human beings 1. myth or simply not true 9. Rating this question is not appr opriate given my personal beliefs. Continued on next page

PAGE 117

112 Appendix A continued 54. Obey the government, for God is the one who has put it there. There is no government anywhere that God has not placed in power. 4. the true/literal words of God 3. the inspired or interpreted words of God 2. more or less influenced by human beings 1. myth or simply not true 9. Rating this question is not appr opriate given my personal beliefs. 55. If her eye is injured, injure his; if her tooth is knocked out, knock out his; and so on-hand for hand, foot for foot, burn for burn, wound for wound, lash for lash. 4. the true/literal words of God 3. the inspired or interpreted words of God 2. more or less influenced by human beings 1. myth or simply not true 9. Rating this question is not appr opriate given my personal beliefs. 56. Your heavenly father will forgive you if you forgive those who sin against you; but if you refuse to forgive them, he will not forgive you. 4. the true/literal words of God 3. the inspired or interpreted words of God 2. more or less influenced by human beings 1. myth or simply not true 9. Rating this question is not appr opriate given my personal beliefs. 57. Then Jesus asked the man, what do y ou want? Lord, he pleaded, I want to see! And Jesus said, All right, begin seeing! Your faith has healed you. 4. the true/literal words of God 3. the inspired or interpreted words of God 2. more or less influenced by human beings 1. myth or simply not true 9. Rating this question is not appr opriate given my personal beliefs. 58. If someone slaps you on one cheek, let him slap the other too! If someone demands your coat, give him your shirt besides. 4. the true/literal words of God 3. the inspired or interpreted words of God 2. more or less influenced by human beings 1. myth or simply not true 9. Rating this question is not appr opriate given my personal beliefs. Continued on next page

PAGE 118

113 Appendix A continued F. THE FOLLOWING ITEMS CONTINUE TO ASK YOU ABOUT ASPECTS OF YOUR RELIGIOUS BELIEFS. We recognize that the questions in this section refer to statements often associated with Judeo-Christia n religious beliefs and, therefore, may not reflect your personal beliefs. If you are uncomfortable responding to these statements, feel free to check the code after each statement that indicates, Rating th is question is not appropriate given my personal beliefs. 59. I believe the Bible is the actual word of God and it is to be taken literally, word for word. 5. strongly agree 4. somewhat agree 3. uncertain 2. somewhat disagree 1. strongly disagree 9. Rating this question is not appr opriate given my personal beliefs. 60. I believe the miracles described in the Bible really happened; they are not just stories. 5. strongly agree 4. somewhat agree 3. uncertain 2. somewhat disagree 1. strongly disagree 9. Rating this question is not appr opriate given my personal beliefs. 61. I believe Jesus truly rose from the dead. 5. strongly agree 4. somewhat agree 3. uncertain 2. somewhat disagree 1. strongly disagree 9. Rating this question is not appr opriate given my personal beliefs. 62. I believe that those who do not accep t God will go to hell after their death. 5. strongly agree 4. somewhat agree 3. uncertain 2. somewhat disagree 1. strongly disagree 9. Rating this question is not appr opriate given my personal beliefs. 63. I believe that Hell actually exists. 5. strongly agree 4. somewhat agree 3. uncertain 2. somewhat disagree 1. strongly disagree 9. Rating this question is not appr opriate given my personal beliefs. Continued on next page

PAGE 119

114 Appendix A continued G. THE NEXT GROUP OF QUESTIONS ASKS ABOUT YOUR THOUGHTS ABOUT PEOPLE. SOME OF THESE QUESTIONS REFER TO PERSONS WHO HAVE BEEN LEGALLY CONVICTED OF CRIMES. 64. I believe most police officers are prejudiced against minorities. 5. strongly agree 4. somewhat agree 3. uncertain 2. somewhat disagree 1. strongly disagree 65. I believe most police officers abuse their discretion. 5. strongly agree 4. somewhat agree 3. uncertain 2. somewhat disagree 1. strongly disagree 66. I believe most police officers try hard, but do not have the resources to do their job effectively. 5. strongly agree 4. somewhat agree 3. uncertain 2. somewhat disagree 1. strongly disagree 67. I believe most police officers are trustworthy and do their job effectively. 5. strongly agree 4. somewhat agree 3. uncertain 2. somewhat disagree 1. strongly disagree 68. I believe most police officers act professionally when dealing with citizens. 5. strongly agree 4. somewhat agree 3. uncertain 2. somewhat disagree 1. strongly disagree 69. Most people who commit crimes were born to be criminals. 5. strongly agree 4. somewhat agree 3. uncertain 2. somewhat disagree 1. strongly disagree Continued on next page

PAGE 120

115 Appendix A continued 70. Most offenders commit crimes because they have little or no self-control. 5. strongly agree 4. somewhat agree 3. uncertain 2. somewhat disagree 1. strongly disagree 71. Most adult offenders commit crimes as a way of coping with poor living conditions (e.g., extreme poverty, violence in the home, marital problems, etc.) 5. strongly agree 4. somewhat agree 3. uncertain 2. somewhat disagree 1. strongly disagree 72. Most adult offenders commit crim es because they have bad characters. 5. strongly agree 4. somewhat agree 3. uncertain 2. somewhat disagree 1. strongly disagree 73. Most adult offenders commit crimes b ecause they are too lazy to find a lawful way out of a bad situation. 5. strongly agree 4. somewhat agree 3. uncertain 2. somewhat disagree 1. strongly disagree 74. Most adult offenders commit crimes because our society offers them little opportunity to get a job and make money. 5. strongly agree 4. somewhat agree 3. uncertain 2. somewhat disagree 1. strongly disagree 75. Most adult offenders commit crimes becau se their home lives as children were lacking in love, discipline, and supervision. 5. strongly agree 4. somewhat agree 3. uncertain 2. somewhat disagree 1. strongly disagree Continued on next page

PAGE 121

116 Appendix A continued 76. Most adult offenders commit crimes because of outside influences (e.g., peer pressure, money problems). 5. strongly agree 4. somewhat agree 3. uncertain 2. somewhat disagree 1. strongly disagree 77. When determining the sentence for adult offende rs who have been legally convicted of crimes, the circumstances of the offense, such as weapon used, planning, injury to victim, amount of loss, are not important. Instead, the sentence should be based so lely on the seriousness of the offense committed. 5. strongly agree 4. somewhat agree 3. uncertain 2. somewhat disagree 1. strongly disagree 78. The best way to deal with adults who have been legally convicted of crimes is to lock them up so they are not able to harm anyone again. 5. strongly agree 4. somewhat agree 3. uncertain 2. somewhat disagree 1. strongly disagree 79. The best way to deal with adults who have been legally convicted of crimes is to punish them harshly so they w ill learn their lesson and not commit other crimes. 5. strongly agree 4. somewhat agree 3. uncertain 2. somewhat disagree 1. strongly disagree 80. The best way to deal with adults who have been legally convicted of crimes is to punish them to the extent that they will pay their debt to society for the crimes they have committed. 5. strongly agree 4. somewhat agree 3. uncertain 2. somewhat disagree 1. strongly disagree 81. The best way to deal with adults who have been legally convicted of crimes is to punish them harshly to set an example so that others will not commit crimes. 5. strongly agree 4. somewhat agree 3. uncertain 2. somewhat disagree 1. strongly disagree Continued on next page

PAGE 122

117 Appendix A continued 82. The best way to deal with adults who have been legally convicted of crimes is to treat and help them so that they do not commit other crimes. 5. strongly agree 4. somewhat agree 3. uncertain 2. somewhat disagree 1. strongly disagree 83. Most juvenile offenders commit crimes as a way of coping with poor living conditions (e.g., poverty, violence in the home, lack of supervision, broken home). 5. strongly agree 4. somewhat agree 3. uncertain 2. somewhat disagree 1. strongly disagree 84. Most juvenile offenders commit crimes because they have bad characters. 5. strongly agree 4. somewhat agree 3. uncertain 2. somewhat disagree 1. strongly disagree 85. Most juvenile offenders commit crimes because they are too lazy to find a lawful way out of a bad situation. 5. strongly agree 4. somewhat agree 3. uncertain 2. somewhat disagree 1. strongly disagree 86. Most juvenile offenders commit crimes because they are bored and have nothing to do (e.g., work, play sports, help with family chores, etc.). 5. strongly agree 4. somewhat agree 3. uncertain 2. somewhat disagree 1. strongly disagree 87. Most juvenile offenders commit crimes because of outside influences (e.g., peer pressure, money problems). 5. strongly agree 4. somewhat agree 3. uncertain 2. somewhat disagree 1. strongly disagree Continued on next page

PAGE 123

118 Appendix A continued 88. Most juvenile offenders commit crimes because their home life is lacking in love, discipline, and supervision. 5. strongly agree 4. somewhat agree 3. uncertain 2. somewhat disagree 1. strongly disagree 89. When determining the sentence for juvenile offenders who have been legally convicted of crimes, the circumstances of the offense, such as the weapon used, planning, injury to victim, amount of loss, are not important. Instead, the sentence should be based solely on the seriousness of the offense committed. 5. strongly agree 4. somewhat agree 3. uncertain 2. somewhat disagree 1. strongly disagree 90. The best way to deal with juveniles who have been legally convicted of crimes is to lock them up so they are not able to harm anyone again. 5. strongly agree 4. somewhat agree 3. uncertain 2. somewhat disagree 1. strongly disagree 91. The best way to deal with juveniles who have been convicted of crimes is to punish them harshly so they will learn their lesson and not commit other crimes. 5. strongly agree 4. somewhat agree 3. uncertain 2. somewhat disagree 1. strongly disagree 92. The best way to deal with juveniles who have been convicted of crimes is to punish them to the extent that they will pay their debt to society for the crimes they have committed. 5. strongly agree 4. somewhat agree 3. uncertain 2. somewhat disagree 1. strongly disagree 93. The best way to deal with juveniles who have been convicted of crimes is to punish them harshly to set an example so that others will not commit crimes. 5. strongly agree 4. somewhat agree 3. uncertain 2. somewhat disagree 1. strongly disagree Continued on next page

PAGE 124

119 Appendix A continued 94. The best way to deal with juveniles who have been convicted of crimes is to treat and help them so that they do not commit other crimes. 5. strongly agree 4. somewhat agree 3. uncertain 2. somewhat disagree 1. strongly disagree H. THE FINAL GROUP OF QUESTIONS ASKS FOR YOUR OPINION ON THE DEATH PENALTY. 95. Generally speaking, I support the death penalty for adults. 5. strongly agree 4. somewhat agree 3. uncertain 2. somewhat disagree 1. strongly disagree 96. Generally speaking, I support the death penalty for juveniles. 5. strongly agree 4. somewhat agree 3. uncertain 2. somewhat disagree 1. strongly disagree For questions 97-108, please check the column that best reflects your thoughts about whether adult or juvenile offenders should be sentenced to death for the following crimes (e.g., for stealing a loaf of bread if a juvenile X ): SENTENCED TO DEATH Strongly Strongly Disagree Disagree Uncertain Agree Agree 1 2 3 4 5 97. for the murder of a employee when an adult ___ ___ ___ ___ ___ is the killer 98. for the murder of a criminal justice ___ ___ ___ ___ ___ employee when a juvenile is the killer 99. for murder that is intentional and well-pl anned ___ ___ ___ ___ ___ when the killer is an adult Continued on next page

PAGE 125

120 Appendix A continued 100. for murder that is intentional and wellplanned ___ ___ ___ ___ ___ when the killer is a juvenile 101. for the murder of a ___ ___ ___ ___ ___ child when the killer is an adult 102. for the murder of a ___ ___ ___ ___ ___ child when the killer is a juvenile 103. for unplanned murder ___ ___ ___ ___ ___ when the killer is an adult 104. for unplanned murder ___ ___ ___ ___ ___ when the killer is a juvenile 105. for multiple rapes when ___ ___ ___ ___ ___ the rapist is an adult 106. for multiple rapes when ___ ___ ___ ___ ___ the rapist is a juvenile 107. for molesting a child ___ ___ ___ ___ ___ when the molester is an adult 108. for molesting a child ___ ___ ___ ___ ___ when the molester is a juvenile Continued on next page

PAGE 126

121 Appendix A continued FOR THOSE WHO SUPPORT CAPITAL PUNISHMENT: If you support capital punishment, please indi cate how important each of the reasons below are to you when thinking about the issue with regard to offenders who have been legally convicted of crimes. Not at all Somewhat Important Very Important Important Important 1. 2. 3. 4. 109. Because the offender _____ _____ _____ _____ deserves to die 110. So the offender _____ _____ _____ _____ cannot re-offend 111. To teach a lesson _____ _____ _____ _____ to others in society so they will not offend 112. To give the victims _____ _____ _____ _____ family a sense that justice has been done Continued on next page

PAGE 127

122 Appendix A continued FOR THOSE WHO OPPOSE CAPITAL PUNISHMENT Please indicate how important each of the reasons below are to you when thinking about capital punishment with regard to offenders who have been legally convicted of crimes. Not at all Somewhat Important Very Important Important Important 1. 2. 3. 4. 113. Because it is _____ _____ _____ _____ morally wrong 114. Because everyone _____ _____ _____ _____ is redeemable 115. Because certain racial _____ _____ _____ _____ and ethnic minority groups are more likely to receive a death sentence 116. Because only the poor _____ _____ _____ _____ get the death penalty 117. Because of the _____ _____ _____ _____ possibility of executing innocent people 118. Because it is cruel _____ _____ _____ _____ and unusual punishment YOU HAVE FINISHED THE SURVEY. THANK YOU FOR YOUR PARTICIPATION!

PAGE 128

123 APPENDIX B YOUR PARTICIPATION IN THE FO LLOWING QUESTIONNAIRE IS VOLUNTARY. YOU ARE UNDER NO OBLIGATION TO TAKE OR COMPLETE THE INSTRUMENT. Tampa Jury Pool Survey Research Project on Attitudes Toward Crime and Justice RESEARCH INFORMATION AND RESPONDENT CONSENT TO PARTICIPATE Professors John K. Cochran and Kathleen M. Heid e in the Department of Criminology, and Sheila Schlaupitz, a masters student in the Department of Cr iminology at the University of South Florida, request your participation in a research project examining public attitudes toward crime and justice. This project has been approved by the Department of Criminology and the Institutional Review Board (IRB) at the University of South Florida. Our staff of graduate criminology students at the University of South Florida will administer questionnaires to jury pool participants in Hillsborough C ounty such as yourself. Your participation in this project will be very valuable to increasing our knowledge regarding how citizens view issues related to crime and punishment. However, your participation is entirely voluntary; you are under no obligation to participate. Moreover, if at any time during the survey you feel you do not wish to continue filling it out, you may simply stop. Likewise, you may decline to answer any question. Should you choose to participate, please also note that your identity will remain anonymous and the information you provide will not be tied to you in any way. No one outside the research staff will have access to your answers; your answers will be recoded into numeric, machine-readable code and the questionnaires will be destroyed. Finally, only st atistics summarizing findings from this study will be published or shared with the public. We hope that you will choose to participate. If you choose not to participate, we ask that you please complete Section A, questions 1 through 9 of th e survey. Your willingness to participate indicates your voluntary consent. Should you have any questions, please feel free to contact any of us. Please retain this document. Thank you, and have a great day! John K. Cochran, PhD. Kathleen M. Heide, PhD. Associate Professor/Associ ate Chair Full Professor (813) 974-9569 (813) 974-9543

PAGE 129

124 Appendix C. Factor Analysis Results for the Overall Death Penalty Scale Item-to-Scale Correlations General support for adult death penalty 0.66 Death penalty support for adults who murder a criminal justice employee 0.82 Death penalty support for adults who commit a planned murder 0.77 Death penalty support for adults who murder a child 0.79 Death penalty support for adults who commit an unplanned murder 0.71 Death penalty support for adults who commit a rape 0.74 Death penalty support for adults who molest a child 0.69 General support for juvenile death penalty 0.55 Death penalty support for juveniles who murder a criminal justice employee 0.81 Death penalty support for juveniles who commit a planned murder 0.84 Death penalty support for juveniles who murder a child 0.83 Death penalty support for juveniles who commit an unplanned murder 0.69 Death penalty support for juveniles who commit a rape 0.75 Death penalty support for juveniles who molest a child 0.69 Cronbachs Alpha=0.95 Eigenvalue=8.55

PAGE 130

125 Appendix D. Factor Analysis Results for the Religiosity Scale Item-to-Scale Correlations Frequency of attendance at worship services 0.73 Number of church-related groups 0.56 Frequency of participation in activities outside of services 0.65 Frequency of prayer outside of services 0.73 Respondents believe religion is an important part of life. 0.72 Degree to which respondents describe themselves as religious 0.77 Respondents seek God when making decisions 0.81 Religion influences how respondents live their lives 0.77 Frequency of quiet time 0.72 Cronbachs Alpha=0.92 Eigenvalue=5.56

PAGE 131

126 Appendix E. Factor Analysis Results for the Attribution Theory Variables Theological Emphasis Item-to-Scale Correlations To receive forgiveness, you must forgive sinners 0.66 Important to hate the sin, but love the sinner 0.68 If sorry, person who lives life of crime should be forgiven 0.63 God sympathizes with people 0.51 God understands when people sin 0.46 God loves people regardless of their behaviors and beliefs 0.51 Cronbachs Alpha=0.81 Eigenvalue=3.09 Gpunitiv (alternative theological emphasis scale used in the first path analysis due to possible problems with the sampling distribution) Item-to-Scale Correlations Respondents feared Gods punishment after wrongdoing 0.42 People who do evil will eventually go to hell 0.48 God condemns sinners 0.43 God gets back at all sinners 0.30 Cronbachs Alpha=0.63 Eigenvalue=1.89 Biblical Literalness Item-to-Scale correlations Bible is actual word of God to be taken word-for-word 0.71 Miracles really happenednot just stories 0.84 Jesus truly rose from the dead 0.76 Those who do not accept God will go to hell 0.76 Hell actually exists 0.79 Cronbachs Alpha=0.91 Eigenvalue=3.68 Attribution Style Item-to-Scale correlations Adults commit crimes because th ey have bad characters -0.02 Adults commit crimes because they are lazy 0.05 Adults commit crimes to cope with poor living conditions 0.20 Adults commit crimes as a result of a lacking home life 0.23 Adults commit crimes as a result of outside influences 0.23 Juveniles commit crimes because they have bad characters 0.10 Juveniles commit crimes becau se they are lazy 0.05 Juveniles commit crimes to cope with poor living conditions 0.26 Juveniles commit crimes as a result of a lacking home life 0.28 Juveniles commit crimes as a result of outside influences 0.20 Continued on next page

PAGE 132

127 Appendix E continued Cronbachs Alpha=0.40 Eigenvalue=3.30 Punishment Philosophy Item-to-Scale Correlations Adult sentence/based on severity of offense only 0.25 Adult sentence/lock them up, so they cannot re-offend 0.46 Adult sentence/lock them up to scare them into not re-offending 0.65 Adult sentence/punish to the extent that they repay debt to society 0.30 Adult sentence/punish them to scare others into not offending 0.63 Adult sentence/treat and help them so they do not re-offend 0.12 Juvenile sentence/based on severity of offense only 0.31 Juvenile sentence/lock them up, so they cannot re-offend 0.60 Juvenile sentence/lock them up to scare them into not re-offending 0.69 Juvenile sentence/punish to the extent that they repay debt 0.39 Juvenile sentence/punish them to scare others into not offending 0.68 Juvenile sentence/treat and help them so they do not re-offend 0.09 Cronbachs Alpha=0.79 Eigenvalue=4.22

PAGE 133

128 Appendix F. Factor Analysis Results for the Control Variable Scales Attitudes toward sex roles Item-to-Scale Correlations Women should have an equal role as men in business 0.60 Women should have an equal role as men in the military 0.51 Women should have an equal role as men in the family 0.53 Cronbachs Alpha=0.70 Eigenvalue=1.96 Attitudes toward abortion Item-to-Scale Correlations Abortion allowable under any circumstance 0.68 Abortion allowable when moms health is in danger 0.65 Abortion allowable when fetus health is in danger 0.74 Abortion allowable when pregnancy is the result of a rape 0.76 Abortion allowable when mom cannot afford to raise the child 0.63 Cronbachs Alpha=0.87 Eigenvalue=3.29 Attitudes toward social welfare Item-to-Scale correlations Support affirmative action to help minorities 0.50 Support welfare to help minorities 0.50 Cronbachs Alpha=0.67 Eigenvalue=1.50 Attitudes toward the police (the criminal justice scale) Item-to-Scale correlations Most police officers are prejudiced against minorities. 0.60 Most police officers abuse their discretion. 0.68 Most police officers are trustworthy/do their jobs effectively. 0.62 Most police officers act professionally when dealing with citizens. 0.61 Cronbachs Alpha=0.80 Eigenvalue=2.56

PAGE 134

129 Appendix G. OLS Regression of Control Variables on Death Penalty Attitudes Model 1: OLS regression of control variables on the overall adult death penalty scale for religious tradition*religiosity DF SS MS F Model 13 3748.86 288.37 7.16* Error 454 18280 40.26 Total 467 22029 R=.170 b se(b) B age -0.00 0.03 -0.12 educate -0.52* 0.22 -0.12 income 0.30 0.17 0.08 ideology 0.86* 0.37 0.11 sexroles 0.03 0.15 0.01 abortion -0.35* 0.07 -0.25 socwlfre 0.46* 0.15 0.15 police 0.15 0.10 0.07 Model 2: OLS regression of control variables on general support for adult death penalty for religious tradition *religiosity DF SS MS F Model 13 130.38 10.03 7.76* Error 454 586.85 1.29 Total 467 717.23 R=.182 b se(b) B age 0.00 0.01 0.00 educate -0.02 0.04 -0.03 income 0.04 0.03 0.00 ideology 0.19* 0.07 0.14 sexroles -0.00 0.03 -0.00 abortion -0.07* 0.01 -0.27 socwlfre 0.07* 0.03 0.13 police 0.04 0.02 0.10 Continued on next page

PAGE 135

130 Appendix G continued Model 3: OLS regression of control variables on adult murder of a criminal justice employee for religious tradition *religiosity DF SS MS F Model 13 88.91 6.84 5.69* Error 454 545.87 1.20 Total 467 634.79 R=.140 b se(b) B age -0.00 0.00 -0.04 educate -0.03 0.04 -0.04 income 0.07* 0.03 0.11 ideology 0.13 0.06 0.10 sexroles 0.01 0.03 0.02 abortion -0.04* 0.01 -0.16 socwlfre 0.06* 0.03 0.12 police 0.02 0.02 0.07 Model 4: OLS regression of control variables on adult planned murder for religious tradition*religiosity DF SS MS F Model 13 96.54 7.43 5.60* Error 454 601.84 1.33 Total 467 698.38 R=.138 b se(b) B age -0.00 0.01 -0.03 educate -0.06 0.04 -0.08 income 0.04 0.03 0.07 ideology 0.12 0.07 0.09 sexroles -0.00 0.03 -0.01 abortion -0.05* 0.01 -0.18 socwlfre 0.10* 0.03 0.17 police 0.03 0.02 0.09 Continued on next page

PAGE 136

131 Appendix G continued Model 5: OLS regression of control variable s on adult murder of a child for religious tradition*religiosity DF SS MS F Model 13 100.76 7.75 5.75* Error 454 611.47 1.35 Total 467 712.23 R=.142 b se(b) B age -0.00 0.01 -0.01 educate -0.07 0.04 -0.09 income 0.04 0.03 0.07 ideology 0.17* 0.07 0.13 sexroles -0.01 0.03 -0.02 abortion -0.04* 0.01 -0.16 socwlfre 0.08* 0.03 0.14 police 0.03 0.02 0.09 Model 6: OLS regression of control variables on adult unplanned murder for religious tradition*religiosity DF SS MS F Model 13 49.59 3.81 4.69* Error 454 369.37 0.81 Total 467 418.97 R=.118 b se(b) B age 0.00 0.00 0.05 educate -0.08* 0.03 0.13 income 0.04 0.02 0.08 ideology 0.04 0.05 0.04 sexroles -0.00 0.02 -0.01 abortion -0.05* 0.01 -0.23 socwlfre 0.03 0.02 0.06 police 0.02 0.01 0.08 Continued on next page

PAGE 137

132 Appendix G continued Model 7: OLS regression of control variables on adult rape for religious tradition*religiosity DF SS MS F Model 13 76.65 5.90 4.23* Error 454 632.76 1.39 Total 467 709.41 R=.108 b se(b) B age 0.01 0.01 0.05 educate -0.12* 0.04 -0.15 income 0.04 0.03 0.05 ideology 0.12 0.07 0.09 sexroles 0.01 0.03 0.02 abortion -0.05* 0.01 -0.19 socwlfre 0.07* 0.03 0.13 police -0.01 0.02 -0.02 Model 8: OLS regression of control variables on adult child molestation for religious tradition*religiosity DF SS MS F Model 13 82.68 6.36 4.72* Error 454 611.40 1.35 Total 467 694.07 R=.119 b se(b) B age -0.00 0.01 -0.03 educate -0.15* 0.04 -0.19 income 0.02 0.03 0.04 ideology 0.09 0.07 0.06 sexroles 0.03 0.03 0.05 abortion -0.06* 0.01 -0.23 socwlfre 0.05 0.03 0.09 police 0.00 0.02 0.01 Continued on next page

PAGE 138

133 Appendix G continued Model 9: OLS regression of control variables on the overall juvenile death penalty scale for religious tradition *religiosity DF SS MS F Model 13 3909.87 300.76 8.47* Error 454 16116 35.50 Total 467 20026 R=.195 b se(b) B age 0.02 0.03 0.04 educate -0.75* 0.20 -0.18 income 0.34* 0.16 0.10 ideology 1.16* 0.35 0.16 sexroles 0.13 0.14 0.04 abortion -0.37* 0.07 -0.27 socwlfre 0.32* 0.14 0.11 police -0.08 0.09 -0.04 Model 10: OLS regression of control variables on general support for juvenile death penalty for religious tradition *religiosity DF SS MS F Model 13 136.48 10.50 7.21* Error 454 660.92 1.46 Total 467 797.41 R=.171 b se(b) B age 0.01 0.01 0.05 educate -0.07 0.04 -0.09 income 0.07* 0.03 0.10 ideology 0.23* 0.07 0.16 sexroles 0.03 0.03 0.05 abortion -0.07* 0.01 -0.26 socwlfre 0.07* 0.03 0.13 police -0.04* 0.02 -0.10 Continued on next page

PAGE 139

134 Appendix G continued Model 11: OLS regression of control variables on juvenile murder of a criminal justice employee for religious tradition *religiosity DF SS MS F Model 13 96.83 7.45 6.97* Error 454 484.92 1.07 Total 467 581.75 R=.166 b se(b) B age 0.00 0.00 0.00 educate -0.08* 0.04 -0.11 income 0.08* 0.03 0.14 ideology 0.15* 0.06 0.12 sexroles 0.01 0.02 0.03 abortion -0.05* 0.01 -0.21 socwlfre 0.07* 0.02 0.14 police -0.00 0.02 -0.00 Model 12: OLS regression of control variable s on juvenile planned murder for religious tradition*religiosity DF SS MS F Model 13 118.04 9.08 7.18* Error 454 573.93 1.26 Total 467 691.97 R=.171 b se(b) B age 0.00 0.00 0.00 educate -0.09* 0.04 -0.11 income 0.06 0.03 0.09 ideology 0.18* 0.07 0.13 sexroles 0.01 0.03 0.03 abortion -0.05* 0.01 -0.22 socwlfre 0.08* 0.03 0.14 police 0.01 0.02 0.02 Continued on next page

PAGE 140

135 Appendix G continued Model 13: OLS regression of control variables on juvenile murder of a child for religious tradition*religiosity DF SS MS F Model 13 103.74 7.98 6.78* Error 454 534.64 1.18 Total 467 638.38 R=.163 b se(b) B age 0.00 0.00 0.01 educate -0.13* 0.04 -0.18 income 0.07* 0.03 0.11 ideology 0.24* 0.06 0.18 sexroles 0.02 0.02 0.03 abortion -0.05* 0.01 -0.20 socwlfre 0.04 0.03 0.08 police -0.00 0.02 -0.00 Model 14: OLS regression of control variable s on juvenile unplanned murder for religious tradition*religiosity DF SS MS F Model 13 49.43 3.80 4.80* Error 454 359.82 0.79 Total 467 409.25 R=.121 b se(b) B age 0.01* 0.00 0.10 educate -0.10* 0.03 -0.17 income 0.04 0.02 0.09 ideology 0.07 0.05 0.07 sexroles -0.01 0.02 -0.04 abortion -0.04* 0.01 -0.21 socwlfre 0.02 0.02 0.06 police -0.01 0.01 -0.02 Continued on next page

PAGE 141

136 Appendix G continued Model 15: OLS regression of control variables on juvenile rape for religious tradition*religiosity DF SS MS F Model 13 77.11 5.93 5.24* Error 454 513.44 1.13 Total 467 590.56 R=.131 b se(b) B age 0.00 0.00 0.03 educate -0.12* 0.04 -0.17 income 0.01 0.03 0.02 ideology 0.15* 0.06 0.12 sexroles 0.04 0.02 0.08 abortion -0.05* 0.01 -0.22 socwlfre 0.03 0.03 0.05 police -0.03 0.02 -0.08 Model 16: OLS regression of control variable s on juvenile child molestation for religious tradition*religiosity DF SS MS F Model 13 71.92 5.53 5.20* Error 454 483.39 1.06 Total 467 555.31 R=.130 b se(b) B age 0.00 0.00 0.03 educate -0.15* 0.04 -0.22 income 0.01 0.03 0.01 ideology 0.13* 0.06 0.11 sexroles 0.03 0.02 0.06 abortion -0.05* 0.01 -0.24 socwlfre 0.00 0.02 0.01 police -0.01 0.02 -0.03 Continued on next page

PAGE 142

137 Appendix G continued Model 17: OLS regression of control variable s on the overall adult death penalty scale for membership in a conservative Protestant faith*religiosity DF SS MS F Model 13 4869.80 374.60 9.47* Error 622 24617 39.58 Total 635 29487 R=.165 b se(b) B age -0.02 0.02 -0.03 educate -0.51* 0.18 -0.12 income 0.20 0.15 0.06 ideology 0.84* 0.30 0.12 sexroles -0.08 0.12 -0.03 abortion -0.31* 0.06 -0.22 socwlfre 0.53* 0.13 0.17 police 0.13 0.08 0.07 Model 18: OLS regression of control variables on general support for adult death penalty for membership in a conservative Protestant faith*religiosity DF SS MS F Model 13 182.03 14.00 11.04* Error 622 788.66 1.27 Total 635 970.68 R=.188 b se(b) B age -0.00 0.02 -0.03 educate -0.02 0.03 -0.03 income 0.02 0.03 0.03 ideology 0.16* 0.05 0.12 sexroles -0.01 0.02 -0.02 abortion -0.07* 0.01 -0.27 socwlfre 0.08* 0.02 0.15 police 0.04* 0.01 0.11 Continued on next page

PAGE 143

138 Appendix G continued Model 19: OLS regression of control variables on adult murder of a criminal justice employee for membership in a conservative Protestant faith*religiosity DF SS MS F Model 13 136.86 10.53 9.08* Error 622 720.85 1.16 Total 635 857.71 R=.160 b se(b) B age -0.01 0.00 -0.06 educate -0.03 0.03 -0.05 income 0.07* 0.02 0.11 ideology 0.13* 0.05 0.11 sexroles -0.01 0.02 -0.02 abortion -0.03* 0.01 -0.15 socwlfre 0.09* 0.02 0.16 police 0.03 0.01 0.07 Model 20: OLS regression of control variables on adult planned murder for membership in a conservative Protestant faith*religiosity DF SS MS F Model 13 128.79 9.91 7.65* Error 622 805.80 1.30 Total 635 934.58 R=.138 b se(b) B age -0.01 0.00 -0.05 educate -0.05 0.03 -0.06 income 0.03 0.03 0.04 ideology 0.14* 0.05 0.11 sexroles -0.02 0.02 -0.04 abortion -0.05* 0.01 -0.19 socwlfre 0.09* 0.02 0.17 police 0.03* 0.01 0.09 Continued on next page

PAGE 144

139 Appendix G continued Model 21: OLS regression of control variables on adult murder of a child for membership in a conservative Protestant faith*religiosity DF SS MS F Model 13 129.42 9.96 7.45* Error 622 830.76 1.34 Total 635 960.18 R=.135 b se(b) B age -0.00 0.00 -0.03 educate -0.06 0.03 0.08 income 0.04 0.03 0.06 ideology 0.17* 0.06 0.13 sexroles -0.02 0.02 -0.04 abortion -0.04* 0.01 -0.15 socwlfre 0.07* 0.02 0.13 police 0.04 0.02 0.10 Model 22: OLS regression of control variables on adult unplanned murder for membership in a conservative Protestant faith*religiosity DF SS MS F Model 13 73.15 5.63 6.85* Error 622 510.62 0.82 Total 635 583.77 R=.125 b se(b) B age 0.00 0.00 0.04 educate -0.09* 0.03 -0.15 income 0.03 0.02 0.06 ideology 0.04 0.04 0.04 sexroles -0.01 0.02 -0.03 abortion -0.04* 0.01 -0.21 socwlfre 0.06* 0.02 0.13 police 0.02 0.01 0.06 Continued on next page

PAGE 145

140 Appendix G continued Model 23: OLS regression of control variables on adult rape for membership in a conservative Protestant faith*religiosity DF SS MS F Model 13 82.35 6.33 4.49* Error 622 878.34 1.41 Total 635 960.68 R=.086 b se(b) B age 0.00 0.00 0.01 educate -0.12* 0.03 -0.16 income 0.02 0.03 0.02 ideology 0.12* 0.06 0.09 sexroles -0.00 0.02 -0.01 abortion -0.04* 0.01 -0.17* socwlfre 0.08* 0.02 0.14 police -0.01 0.02 -0.03 Model 24: OLS regression of control variables on adult child molestation for membership in a conservative Protestant faith*religiosity DF SS MS F Model 13 79.20 6.09 4.48* Error 622 846.74 1.36 Total 635 925.94 R=.086 b se(b) B age -0.00 0.00 0.04 educate -0.13* 0.03 -0.17 income 0.00 0.03 0.01 ideology 0.07 0.06 0.06 sexroles 0.00 0.02 0.00 abortion -0.04* 0.01 -0.15 socwlfre 0.06* 0.02 0.11 police -0.01 0.02 -0.02 Continued on next page

PAGE 146

141 Appendix G continued Model 25: OLS regression of control variables on the overall juvenile death penalty scale for membership in a conservative Protestant faith*religiosity DF SS MS F Model 13 4845.68 372.74 10.36* Error 622 22382 35.98 Total 635 27228 R=.178 b se(b) B age 0.02 0.02 0.04 educate -0.83* 0.17 -0.20 income 0.31* 0.14 0.09 ideology 1.12* 0.29 0.16 sexroles 0.04 0.12 0.01 abortion -0.30* 0.06 -0.22 socwlfre 0.51* 0.12 0.17 police -0.07 0.08 -0.04 Model 26: OLS regression of control variables on general support for juvenile death penalty for membership in a conservative Protestant faith*religiosity DF SS MS F Model 13 170.37 13.11 9.02* Error 622 903.27 1.45 Total 635 1073.64 R=.159 b se(b) B age 0.01 0.00 0.06 educate -0.10* 0.03 -0.12 income 0.08* 0.03 0.11 ideology 0.17* 0.06 0.13 sexroles 0.02 0.02 0.04 abortion -0.06* 0.01 -0.23 socwlfre 0.11* 0.02 0.19 police -0.03* 0.02 -0.09 Continued on next page

PAGE 147

142 Appendix G continued Model 27: OLS regression of control variables on juvenile murder of a criminal justice employee for membership in a conservative Protestant faith*religiosity DF SS MS F Model 13 135.00 10.38 9.76* Error 622 661.85 1.06 Total 635 796.85 R=.169 b se(b) B age -0.00 0.00 -0.01 educate -0.10* 0.03 -0.13 income 0.08* 0.02 0.13 ideology 0.16* 0.05 0.13 sexroles -0.00 0.02 -0.00 abortion -0.04* 0.01 -0.17 socwlfre 0.10* 0.02 0.19 police 0.00 0.01 0.01 Model 28: OLS regression of control variables on juvenile planned murder for membership in a conservative Protestant faith*religiosity DF SS MS F Model 13 146.60 11.28 9.01* Error 622 778.39 1.25 Total 635 924.99 R=.159 b se(b) B age -0.00 0.00 -0.01 educate -0.09* 0.03 -0.12 income 0.06* 0.03 0.09 ideology 0.20* 0.05 0.15 sexroles -0.00 0.02 -0.00 abortion -0.05* 0.01 -0.20 socwlfre 0.10* 0.02 0.18 police 0.01 0.01 0.03 Continued on next page

PAGE 148

143 Appendix G continued Model 29: OLS regression of control variables on juvenile murder of a child for membership in a conservative Protestant faith*religiosity DF SS MS F Model 13 130.84 10.06 8.61* Error 622 726.89 1.17 Total 635 857.73 R=.153 b se(b) B age 0.00 0.00 0.02 educate -0.13* 0.03 -0.18 income 0.06* 0.03 0.09 ideology 0.23* 0.05 0.19 sexroles 0.00 0.02 0.00 abortion -0.04* 0.01 0.15 socwlfre 0.07* 0.02 0.13 police -0.00 0.01 -0.00 Model 30: OLS regression of control variables on juvenile unplanned murder for membership in a conservative Protestant faith*religiosity DF SS MS F Model 13 71.80 5.52 6.93* Error 622 495.86 0.80 Total 635 567.66 R=.127 b se(b) B age 0.01* 0.00 0.10 educate -0.12* 0.03 -0.20 income 0.04* 0.02 0.08 ideology 0.07 0.04 0.07 sexroles -0.02 0.02 -0.04 abortion -0.04* 0.01 -0.19 socwlfre 0.06* 0.02 0.13 police -0.00 0.01 -0.01 Continued on next page

PAGE 149

144 Appendix G continued Model 31: OLS regression of control variables on juvenile rape for membership in a conservative Protestant faith*religiosity DF SS MS F Model 13 82.75 6.37 5.57* Error 622 710.32 1.14 Total 635 793.07 R=.104 b se(b) B age 0.00 0.00 0.02 educate -0.14* 0.03 -0.19 income 0.01 0.02 0.01 ideology 0.16* 0.05 0.14 sexroles 0.02 0.02 0.05 abortion -0.04* 0.01 -0.19 socwlfre 0.05* 0.02 0.10 police -0.03* 0.01 -0.08 Model 32: OLS regression of control variables on juvenile child molestation for membership in a conservative Protestant faith*religiosity DF SS MS F Model 13 71.33 5.49 5.14* Error 622 663.89 1.07 Total 635 735.23 R=.097 b se(b) B age 0.00 0.00 0.04 educate -0.16* 0.03 -0.23 income -0.00 0.02 -0.00 ideology 0.14* 0.05 0.12 sexroles 0.01 0.02 0.03 abortion -0.03* 0.01 -0.15 socwlfre 0.03 0.02 0.06 police -0.02 0.01 -0.06 *Significant at p<.05

PAGE 150

145 Appendix H. OLS Regression of Control Variables on Death Pe nalty Attitudes Among Whites Model 1: OLS regression of control variables on the overall adult death penalty scale for religious tradition*religiosity among whites DF SS MS F Model 12 2997.00 249.75 6.49* Error 380 14616 38.46 Total 392 17613 R=.170 b se(b) B age 0.00 0.03 0.00 educate -0.48* 0.23 -0.11 income 0.22 0.18 0.06 ideology 0.99* 0.40 0.13 sexroles 0.06 0.16 0.02 abortion -0.36* 0.08 0.27 socwlfre 0.55* 0.16 0.18 police 0.19 0.10 0.09 Model 2: OLS regression of control variables on general support for adult death penalty for religious tradition*religiosity among whites DF SS MS F Model 12 107.09 8.92 7.42* Error 380 456.91 1.20 Total 392 564.00 R=.190 b se(b) B age 0.00 0.01 0.01 educate -0.01 0.04 -0.01 income 0.01 0.03 0.02 ideology 0.21* 0.07 0.15 sexroles 0.00 0.03 0.00 abortion -0.07* 0.01 -0.30 socwlfre 0.10* 0.03 0.18 police 0.05* 0.02 0.12 Continued on next page

PAGE 151

146 Appendix H continued Model 3: OLS regression of control variables on adult murder of a criminal justice employee for religious tradition *religiosity among whites DF SS MS F Model 12 67.41 5.62 4.96* Error 380 430.24 1.13 Total 392 497.65 R=.136 b se(b) B age -0.01 0.01 -0.05 educate -0.03 0.04 -0.04 income 0.06* 0.03 0.10 ideology 0.17* 0.07 0.13 sexroles 0.01 0.03 0.02 abortion -0.04* 0.01 -0.18 socwlfre 0.08* 0.03 0.15 police 0.03 0.02 0.08 Model 4: OLS regression of control variables on adult planned murder for religious tradition*religiosity among whites DF SS MS F Model 12 80.96 6.75 5.39* Error 380 475.92 1.25 Total 392 556.88 R=.145 b se(b) B age -0.00 0.01 -0.01 educate -0.05 0.04 -0.06 income 0.03 0.03 0.05 ideology 0.16* 0.07 0.12 sexroles 0.00 0.03 0.00 abortion -0.05* 0.01 -0.20 socwlfre 0.10* 0.03 0.18 police 0.04* 0.02 0.12 Continued on next page

PAGE 152

147 Appendix H continued Model 5: OLS regression of control variable s on adult murder of a child for religious tradition*religiosity among whites DF SS MS F Model 12 82.50 6.87 5.41* Error 380 483.34 1.27 Total 392 565.84 R=.146 b se(b) B age -0.00 0.01 -0.01 educate -0.06 0.04 -0.08 income 0.04 0.03 0.06 ideology 0.19* 0.07 0.14 sexroles -0.00 0.03 -0.00 abortion -0.04* 0.01 -0.19 socwlfre 0.10* 0.03 0.17 police 0.05* 0.02 0.12 Model 6: OLS regression of control variables on adult unplanned murder for religious tradition*religiosity among whites DF SS MS F Model 12 43.14 3.60 4.69* Error 380 291.41 0.77 Total 392 334.55 R=.129 b se(b) B age 0.00 0.00 0.06 educate -0.07* 0.03 -0.12 income 0.03 0.03 0.07 ideology 0.04 0.06 0.04 sexroles -0.01 0.02 -0.02 abortion -0.05* 0.01 -0.26 socwlfre 0.04 0.02 0.09 police 0.03 0.01 0.10 Continued on next page

PAGE 153

148 Appendix H continued Model 7: OLS regression of control variables on adult rape for religious tradition*religiosity among whites DF SS MS F Model 12 54.78 4.56 3.29* Error 380 527.55 1.39 Total 392 582.33 R=.094 b se(b) B age 0.01 0.01 0.06 educate -0.12* 0.04 -0.15 income 0.03 0.03 0.04 ideology 0.11 0.08 0.08 sexroles 0.02 0.03 0.04 abortion -0.05* 0.01 -0.21 socwlfre 0.08* 0.03 0.15 police -0.00 0.02 -0.00 Model 8: OLS regression of control variables on adult child molestation for religious tradition*religiosity among whites DF SS MS F Model 12 73.37 6.11 4.67* Error 380 497.62 1.31 Total 392 571.00 R=.129 b se(b) B age -0.00 0.01 -0.02 educate -0.15* 0.04 -0.19 income 0.02 0.03 0.03 ideology 0.11 0.07 0.08 sexroles 0.03 0.03 0.07 abortion -0.06* 0.01 -0.26 socwlfre 0.06* 0.03 0.11 police Continued on next page

PAGE 154

149 Appendix H continued Model 9: OLS regression of control variables on the overall juvenile death penalty scale for religious tradition *religiosity among whites DF SS MS F Model 12 2982.86 248.57 7.26* Error 380 13004 34.22 Total 392 15987 R=.187 b se(b) B age 0.02 0.03 0.03 educate -0.72 0.22 -0.18 income 0.29 0.17 0.09 ideology 1.30* 0.38* 0.18 sexroles 0.11 0.15 0.04 abortion -0.38* 0.07 -0.30 socwlfre 0.41* 0.15 0.14 police -0.08 0.10 -0.04 Model 10: OLS regression of control variables on general support for juvenile death penalty for religious tradition*religiosity among whites DF SS MS F Model 12 112.44 9.37 6.65* Error 380 535.07 1.41 Total 392 647.51 R=6.65 b se(b) B age 0.00 0.01 0.04 educate -0.09* 0.04 -0.11 income 0.05 0.03 0.08 ideology 0.27* 0.08 0.18 sexroles 0.02 0.03 0.04 abortion -0.07* 0.02 -0.28 socwlfre 0.09* 0.03 0.16 police -0.04* 0.02 -0.10 Continued on next page

PAGE 155

150 Appendix H continued Model 11: OLS regression of control variables on juvenile murder of a criminal justice employee for religious tradition *religiosity among whites DF SS MS F Model 12 71.26 5.94 5.89* Error 380 383.05 1.01 Total 392 454.31 R=.157 b se(b) B age -0.00 0.00 -0.01 educate -0.07 0.04 -0.11 income 0.07* 0.03 0.13 ideology 0.20* 0.07 0.17 sexroles 0.01 0.03 0.02 abortion -0.05* 0.01 -0.25 socwlfre 0.07* 0.03 0.14 police -0.00 0.02 -0.00 Model 12: OLS regression of control variable s on juvenile planned murder for religious tradition*religiosity among whites DF SS MS F Model 12 87.52 7.29 6.15* Error 380 450.99 1.19 Total 392 538.51 R=.163 b se(b) B age -0.00 0.01 -0.01 educate -0.06 0.04 -0.09 income 0.05 0.03 0.08 ideology 0.20* 0.07 0.15 sexroles 0.01 0.03 0.02 abortion -0.06* 0.01 -0.25 socwlfre 0.09* 0.03 0.17 police 0.01 0.02 0.04 Continued on next page

PAGE 156

151 Appendix H continued Model 13: OLS regression of control variables on juvenile murder of a child for religious tradition*religiosity among whites DF SS MS F Model 12 80.77 6.73 5.77* Error 380 443.48 1.17 Total 392 524.25 R=.154 b se(b) B age -0.00 0.01 -0.01 educate -0.14* 0.04 -0.19 income 0.06* 0.03 0.10 ideology 0.28* 0.07 0.21 sexroles 0.02 0.03 0.03 abortion -0.05* 0.01 -0.21 socwlfre 0.06* 0.03 0.11 police -0.00 0.02 -0.01 Model 14: OLS regression of control variable s on juvenile unplanned murder for religious tradition*religiosity among whites DF SS MS F Model 12 41.24 3.44 4.58* Error 380 284.90 0.75 Total 392 326.14 R=.126 b se(b) B age 0.01 0.00 0.10 educate -0.09* 0.03 -0.16 income 0.03 0.03 0.07 ideology 0.10 0.06 0.09 sexroles -0.03 0.02 -0.07 abortion -0.04* 0.01 -0.22 socwlfre 0.03 0.02 0.07 police -0.01 0.01 -0.02 Continued on next page

PAGE 157

152 Appendix H continued Model 15: OLS regression of control variables on juvenile rape for religious tradition*religiosity among whites DF SS MS F Model 12 56.64 4.72 4.15* Error 380 432.43 1.14 Total 392 489.07 R=.116 b se(b) B age 0.00 0.01 0.03 educate -0.11* 0.04 -0.15 income 0.01 0.03 0.01 ideology 0.14* 0.07 0.11 sexroles 0.04 0.03 0.08 abortion -0.05* 0.01 -0.24 socwlfre 0.05 0.03 0.01 police -0.03 0.02 -0.07 Model 16: OLS regression of control variable s on juvenile child molestation for religious tradition*religiosity among whites DF SS MS F Model 12 56.01 4.67 4.49* Error 380 394.92 1.04 Total 392 450.94 R=.124 b se(b) B age 0.00 0.00 0.02 educate -0.05* 0.04 -0.22 income 0.01 0.03 0.01 ideology 0.12 0.07 0.10 sexroles 0.03 0.03 0.07 abortion -0.05* 0.01 -0.25 socwlfre 0.02 0.03 0.04 police -0.01 0.02 -0.04 Continued on next page

PAGE 158

153 Appendix H continued Model 17: OLS regression of control variable s on the overall adult death penalty scale for membership in a conservative Protestant faith*religiosity among whites DF SS MS F Model 12 3848.27 320.69 8.50* Error 506 19089 37.73 Total 518 22938 R=.168 b se(b) B age -0.02 0.02 -0.04 educate -0.47* 0.19 -0.11 income 0.11 0.16 0.03 ideology 0.95* 0.34 0.13 sexroles -0.09 0.14 -0.03 abortion -0.31 0.07 -0.24 socwlfre 0.60* 0.14 0.20 police 0.20* 0.09 0.10 Model 18: OLS regression of control variables on general support for adult death penalty for membership in a conservative Protestant faith*religiosity among whites DF SS MS F Model 12 155.86 12.99 11.32* Error 506 580.59 1.15 Total 518 736.44 R=.212 b se(b) B age -0.00 0.00 -0.04 educate -0.01 0.03 -0.01 income 0.01 0.03 0.01 ideology 0.18* 0.06 0.14 sexroles -0.02 0.02 -0.04 abortion -0.07* 0.01 -0.30 socwlfre 0.11 0.02 0.20 police 0.05 0.02 0.14 Continued on next page

PAGE 159

154 Appendix H continued Model 19: OLS regression of control variables on adult murder of a criminal justice employee for membership in a conservative Protestant faith*religiosity among whites DF SS MS F Model 12 103.64 8.64 7.78* Error 506 561.42 1.11 Total 518 665.06 R=.156 b se(b) B age -0.01 0.00 -0.08 educate -0.04 0.03 -0.05 income 0.06* 0.03 0.10 ideology 0.17* 0.06 0.14 sexroles -0.02 0.02 -0.04 abortion -0.03* 0.01 -0.15 socwlfre 0.09* 0.02 0.18 police 0.03* 0.02 0.09 Model 20: OLS regression of control variables on adult planned murder for membership in a conservative Protestant faith*religiosity among whites DF SS MS F Model 12 100.34 8.36 6.72* Error 506 629.47 1.24 Total 518 729.81 R=.138 b se(b) B age -0.00 0.00 -0.04 educate -0.03 0.03 -0.05 income 0.01 0.03 0.02 ideology 0.17* 0.06 0.13 sexroles -0.02 0.03 -0.04 abortion -0.04* 0.01 -0.19 socwlfre 0.09* 0.02 0.17 police 0.04* 0.02 0.12 Continued on next page

PAGE 160

155 Appendix H continued Model 21: OLS regression of control variables on adult murder of a child for membership in a conservative Protestant faith*religiosity among whites DF SS MS F Model 12 100.97 8.41 6.68* Error 506 637.80 1.26 Total 518 738.77 R=.137 b se(b) B age -0.00 0.00 -0.04 educate -0.05 0.03 -0.07 income 0.03 0.03 0.04 ideology 0.16* 0.06 0.13 sexroles -0.02 0.03 -0.04 abortion -0.03* 0.01 -0.15 socwlfre 0.09* 0.03 0.16 police 0.05* 0.02 0.13 Model 22: OLS regression of control variables on adult unplanned murder for membership in a conservative Protestant faith*religiosity among whites DF SS MS F Model 12 63.72 5.31 6.75* Error 506 398.05 0.79 Total 518 461.77 R=.138 b se(b) B age 0.00 0.00 0.02 educate -0.09* 0.03 -0.15 income 0.02 0.02 0.03 ideology 0.05 0.05 0.05 sexroles -0.02 0.02 -0.05 abortion -0.04* 0.01 -0.22 socwlfre 0.07* 0.02 0.16 police 0.03* 0.01 0.10 Continued on next page

PAGE 161

156 Appendix H continued Model 23: OLS regression of control variables on adult rape for membership in a conservative Protestant faith*religiosity among whites DF SS MS F Model 12 55.82 4.65 3.31* Error 506 712.14 1.41 Total 518 767.96 R=.073 b se(b) B age 0.00 0.00 0.01 educate -0.12* 0.04 -0.15 income 0.00 0.03 0.00 ideology 0.11 0.07 0.09 sexroles 0.00 0.03 0.01 abortion -0.05* 0.01 -0.19 socwlfre 0.08* 0.03 0.15 police -0.00 0.02 -0.01 Model 24: OLS regression of control variables on adult child molestation for membership in a conservative Protestant faith*religiosity among whites DF SS MS F Model 12 70.36 5.86 4.43* Error 506 669.45 1.32 Total 518 739.81 R=.095 b se(b) B age -0.00 0.00 -0.04 educate -0.13* 0.04 -0.18 income -0.01 0.03 -0.02 ideology 0.10 0.06 0.08 sexroles 0.01 0.03 0.02 abortion -0.04* 0.01 -0.18 socwlfre 0.07* 0.03 0.13 police Continued on next page

PAGE 162

157 Appendix H continued Model 25: OLS regression of control variables on the overall juvenile death penalty scale for membership in a conservative Protestant faith*religiosity among whites DF SS MS F Model 12 3912.65 326.05 9.41* Error 506 17530 34.64 Total 518 21443 R=.183 b se(b) B age 0.01 0.02 0.01 educate -0.76* 0.18 -0.19 income 0.25 0.15 0.08 ideology 1.24* 0.33 0.18 sexroles -0.01 0.13 -0.00 abortion -0.31* 0.07 -0.24 socwlfre 0.59* 0.13 0.20 police -0.03 0.08 0.01 Model 26: OLS regression of control variables on general support for juvenile death penalty for membership in a conservative Protestant faith*religiosity among whites DF SS MS F Model 12 148.55 12.38 8.74* Error 506 716.49 1.42 Total 518 865.04 R=.172 b se(b) B age 0.01 0.00 0.05 educate -0.10* 0.04 -0.12 income 0.07* 0.03 0.10 ideology 0.21* 0.07 0.15 sexroles 0.01 0.03 0.01 abortion -0.07* 0.01 -0.26 socwlfre 0.13* 0.03 0.22 police -0.03 0.02 -0.08 Continued on next page

PAGE 163

158 Appendix H continued Model 27: OLS regression of control variables on juvenile murder of a criminal justice employee for membership in a conservative Protestant faith*religiosity among whites DF SS MS F Model 12 107.73 8.98 8.68* Error 506 523.32 1.03 Total 518 631.06 R=.171 b se(b) B age -0.00 0.00 -0.04 educate -0.09* 0.03 -0.13 income 0.07* 0.03 0.12 ideology 0.21* 0.06 0.18 sexroles -0.01 0.02 -0.02 abortion -0.05* 0.01 -0.21 socwlfre 0.10* 0.02 0.20 police 0.01 0.01 0.03 Model 28: OLS regression of control variables on juvenile planned murder for membership in a conservative Protestant faith*religiosity among whites DF SS MS F Model 12 119.97 10.00 8.57* Error 506 589.99 1.17 Total 518 709.96 R=.169 b se(b) B age -0.00 0.00 -0.03 educate -0.07* 0.03 -0.09 income 0.05 0.03 0.08 ideology 0.20* 0.06 0.16 sexroles -0.01 0.02 -0.02 abortion -0.05* 0.01 -0.21 socwlfre 0.11* 0.02 0.20 police 0.02 0.02 0.05 Continued on next page

PAGE 164

159 Appendix H continued Model 29: OLS regression of control variables on juvenile murder of a child for membership in a conservative Protestant faith*religiosity among whites DF SS MS F Model 12 107.83 8.99 7.83* Error 506 580.86 1.15 Total 518 688.69 R=.157 b se(b) B age 0.00 0.00 0.00 educate -0.12* 0.03 -0.17 income 0.05 0.03 0.09 ideology 0.25* 0.06 0.21 sexroles -0.00 0.02 -0.01 abortion -0.03* 0.01 -0.15 socwlfre 0.08* 0.02 0.15 police 0.00 0.02 0.01 Model 30: OLS regression of control variables on juvenile unplanned murder for membership in a conservative Protestant faith*religiosity among whites DF SS MS F Model 12 61.80 5.15 6.63* Error 506 393.07 0.78 Total 518 454.87 R=.136 b se(b) B age 0.01 0.00 0.07 educate -0.12* 0.03 -0.20 income 0.03 0.02 0.05 ideology 0.08 0.05 0.08 sexroles -0.04 0.02 -0.10 abortion -0.04* 0.01 -0.19 socwlfre 0.07* 0.02 0.16 police 0.00 0.01 0.01 Continued on next page

PAGE 165

160 Appendix H continued Model 31: OLS regression of control variables on juvenile rape for membership in a conservative Protestant faith*religiosity among whites DF SS MS F Model 12 59.25 4.94 4.35* Error 506 573.93 1.13 Total 518 633.18 R=.094 b se(b) B age 0.00 0.00 0.00 educate -0.12 0.03 -0.17 income -0.00 0.03 -0.00 ideology 0.15* 0.06 0.13 sexroles 0.03 0.02 0.05 abortion -0.04* 0.01 -0.20 socwlfre 0.06* 0.02 0.12 police -0.02 0.02 -0.06 Model 32: OLS regression of control variables on juvenile child molestation for membership in a conservative Protestant faith*religiosity among whites DF SS MS F Model 12 56.19 4.68 4.57* Error 506 519.04 1.03 Total 518 575.23 R=.098 b se(b) B age 0.00 0.00 0.03 educate -0.15* 0.03 -0.22 income -0.01 0.03 -0.02 ideology 0.14* 0.06 0.12 sexroles 0.02 0.02 0.03 abortion -0.04* 0.01 -0.17 socwlfre 0.04 0.02 0.09 police -0.01 0.01 -0.04 *Significant at p<.05

PAGE 166

161 Appendix I. OLS Regression of Control Variables on Death Pena lty Attitudes Among Non-whites Model 1: OLS regression of control variables on the overall adult death penalty scale for religious tradition*religiosity among non-whites DF SS MS F Model 12 905.66 75.47 1.50 Error 62 3111.86 50.19 Total 74 4017.52 R=.225 b se(b) B age -0.04 0.09 -0.06 educate -0.46 0.71 -0.10 income 0.82 0.58 0.18 ideology -0.14 1.00 -0.02 sexroles -0.02 0.42 -0.01 abortion -0.30 0.25 -0.16 socwlfre -0.26 0.50 -0.07 police -0.37 0.32 -0.15 Model 2: OLS regression of control variables on general support for adult death penalty for religious tradition*religiosity among non-whites DF SS MS F Model 12 33.47 2.79 1.64 Error 62 105.25 1.70 Total 74 138.72 R=.241 b se(b) B age 0.00 0.00 0.02 educate -0.07 0.13 -0.08 income 0.24* 0.11 0.28 ideology 0.11 0.18 0.07 sexroles 0.05 0.08 0.09 abortion -0.08 0.05 -0.25 socwlfre -0.11 0.09 -0.16 police -0.05 0.06 -0.11 Continued on next page

PAGE 167

162 Appendix I continued Model 3: OLS regression of control variables on adult murder of a criminal justice employee for religious tradition*religiosity among non-whites DF SS MS F Model 12 20.49 1.71 1.07 Error 62 98.90 1.60 Total 74 119.39 R=.172 b se(b) B age -0.00 0.02 -0.03 educate 0.01 0.13 0.01 income 0.14 0.10 0.18 ideology -0.11 0.18 -0.08 sexroles 0.03 0.08 0.06 abortion -0.05 0.05 -0.15 socwlfre -0.03 0.09 -0.05 police -0.04 0.06 -0.09 Model 4: OLS regression of control variables on adult planned murder for religious tradition*religiosity among non-whites DF SS MS F Model 12 27.98 2.33 1.36 Error 62 106.34 1.72 Total 74 134.32 R=.208 b se(b) B age -0.02 0.02 -0.15 educate -0.13 0.13 -0.14 income 0.13 0.11 0.15 ideology -0.18 0.18 -0.11 sexroles -0.04 0.08 -0.08 abortion -0.06 0.05 -0.17 socwlfre 0.06 0.09 0.08 police -0.09 0.06 -0.19 Continued on next page

PAGE 168

163 Appendix I continued Model 5: OLS regression of control variable s on adult murder of a child for religious tradition*religiosity among non-whites DF SS MS F Model 12 32.62 2.72 1.61 Error 62 104.66 1.69 Total 74 137.28 R=.238 b se(b) B age -0.00 0.02 -0.02 educate -0.05 0.13 -0.05 income 0.10 0.11 0.11 ideology 0.02 0.18 0.01 sexroles -0.06 0.08 -0.10 abortion -0.02 0.05 -0.07 socwlfre -0.05 0.09 -0.07 police -0.07 0.06 -0.14 Model 6: OLS regression of control variables on adult unplanned murder for religious tradition*religiosity among non-whites DF SS MS F Model 12 17.35 1.45 1.39 Error 62 64.57 1.04 Total 74 81.92 R=.212 b se(b) B age 0.00 0.01 0.02 educate -0.10 0.10 -0.14 income 0.14 0.08 0.21 ideology -0.05 0.14 -0.04 sexroles 0.02 0.06 0.04 abortion -0.05 0.04 -0.20 socwlfre -0.06 0.07 -0.10 police -0.05 0.05 -0.14 Continued on next page

PAGE 169

164 Appendix I continued Model 7: OLS regression of control variables on adult rape for religious tradition*religiosity among non-whites DF SS MS F Model 12 17.47 1.46 0.94 Error 62 96.48 1.56 Total 74 113.95 R=.153 b se(b) B age -0.00 0.02 -0.02 educate -0.04 0.13 -0.05 income 0.08 0.10 0.10 ideology 0.13 0.18 0.09 sexroles -0.02 0.07 -0.04 abortion -0.01 0.04 -0.04 socwlfre -0.02 0.09 -0.03 police -0.07 0.06 -0.16 Model 8: OLS regression of control variables on adult child molestation for religious tradition*religiosity among non-whites DF SS MS F Model 12 14.85 1.24 0.72 Error 62 107.15 1.73 Total 74 122.00 R=.122 b se(b) B age -0.02 0.02 -0.12 educate -0.10 0.13 -0.11 income 0.01 0.11 0.01 ideology -0.06 0.19 -0.04 sexroles 0.01 0.08 0.02 abortion -0.02 0.05 -0.07 socwlfre -0.06 0.09 -0.09 police -0.02 0.06 -0.04 Continued on next page

PAGE 170

165 Appendix I continued Model 9: OLS regression of control variables on the overall juvenile death penalty scale for religious tradition*religiosity among non-whites DF SS MS F Model 12 1062.70 88.56 2.06* Error 62 2665.64 42.99 Total 74 3728.35 R=.285 b se(b) B age 0.04 0.08 0.06 educate -0.52 0.66 -0.11 income 0.56 0.54 0.13 ideology 0.47 0.93 0.06 sexroles 0.37 0.39 0.13 abortion -0.32 0.24 -0.18 socwlfre -0.44 0.47 -0.12 police Model 10: OLS regression of control variables on general support for juvenile death penalty for religious tradition*religiosity among non-whites DF SS MS F Model 12 45.51 3.79 2.32* Error 62 101.16 1.63 Total 74 146.67 R=.310 b se(b) B age 0.02 0.02 0.16 educate 0.08 0.13 0.09 income 0.15 0.10 0.17 ideology 0.19 0.18 0.12 sexroles 0.15 0.08 0.25 abortion -0.06 0.05 -0.17 socwlfre -0.13 0.09 -0.18 police -0.05 0.06 -0.11 Continued on next page

PAGE 171

166 Appendix I continued Model 11: OLS regression of control variables on juvenile murder of a criminal justice employee for religious tradition*religiosity among non-whites DF SS MS F Model 12 27.23 2.27 1.63 Error 62 86.16 1.39 Total 74 113.39 R=.240 b se(b) B age 0.01 0.01 0.06 educate -0.04 0.12 -0.05 income 0.11 0.10 0.14 ideology -0.05 0.17 -0.03 sexroles 0.03 0.07 0.06 abortion -0.03 0.04 -0.09 socwlfre 0.03 0.08 0.05 police -0.03 0.05 -0.07 Model 12: OLS regression of control variable s on juvenile planned murder for religious tradition*religiosity among non-whites DF SS MS F Model 12 35.83 2.99 1.78 Error 62 104.25 1.68 Total 74 140.08 R=.256 b se(b) B age 0.00 0.02 0.03 educate -0.15 0.13 -0.17 income 0.06 0.11 0.07 ideology 0.03 0.18 0.02 sexroles 0.05 0.08 0.09 abortion -0.03 0.05 -0.09 socwlfre -0.04 0.09 -0.05 police -0.06 0.06 -0.13 Continued on next page

PAGE 172

167 Appendix I continued Model 13: OLS regression of control variables on juvenile murder of a child for religious tradition*religiosity among non-whites DF SS MS F Model 12 25.02 2.09 1.70 Error 62 75.86 1.22 Total 74 100.88 R=.248 b se(b) B age 0.00 0.01 0.03 educate -0.05 0.11 -0.07 income 0.10 0.09 0.13 ideology 0.05 0.16 0.04 sexroles 0.03 0.07 0.06 abortion -0.05 0.04 -0.16 socwlfre -0.07 0.08 -0.13 police -0.01 0.05 -0.04 Model 14: OLS regression of control variable s on juvenile unplanned murder for religious tradition*religiosity among non-whites DF SS MS F Model 12 24.36 2.03 2.16* Error 62 58.23 0.94 Total 74 82.59 R=.295 b se(b) B age 0.01 0.01 0.06 educate -0.15 0.10 -0.21 income 0.15 0.08 0.22 ideology -0.06 0.14 -0..05 sexroles 0.05 0.06 0.11 abortion -0.08* 0.03 -0.30 socwlfre -0.02 0.07 -0.04 police -0.03 0.04 -0.09 Continued on next page

PAGE 173

168 Appendix I continued Model 15: OLS regression of control variables on juvenile rape for religious tradition*religiosity among non-whites DF SS MS F Model 12 23.84 1.99 1.76 Error 62 69.84 1.13 Total 74 93.68 R=.255 b se(b) B age -0.00 0.01 -0.02 educate -0.12 0.11 -0.17 income 0.03 0.09 0.05 ideology 0.16 0.15 0.13 sexroles 0.02 0.06 0.05 abortion -0.03 0.04 -0.11 socwlfre -0.09 0.08 -0.16 police -0.06 0.05 -0.15 Model 16: OLS regression of control variable s on juvenile child molestation for religious tradition*religiosity among non-whites DF SS MS F Model 12 20.70 1.73 1.30 Error 62 82.05 1.32 Total 74 102.75 R=.202 b se(b) B age 0.00 0.01 0.01 educate -0.10 0.12 -0.12 income -0.02 0.09 -0.03 ideology 0.15 0.16 0.11 sexroles 0.04 0.07 0.09 abortion -0.05 -0.04 -0.16 socwlfre -0.12 0.08 -0.19 police 0.00 0.05 0.01 Continued on next page

PAGE 174

169 Appendix I continued Model 17: OLS regression of control variable s on the overall adult death penalty scale for membership in a conservative Protestant faith*religiosity among non-whites DF SS MS F Model 12 897.76 74.81 1.57 Error 104 4969.47 47.78 Total 116 5867.23 R=.153 b se(b) B age -0.03 0.06 -0.04 educate -0.56 0.54 -0.12 income 0.77 0.41 0.19 ideology -0.01 0.72 -0.00 sexroles 0.01 0.29 0.00 abortion -0.37* 0.17 -0.22 socwlfre 0.02 0.36 0.00 police -0.29 0.22 0.13 Model 18: OLS regression of control variables on general support for adult death penalty for membership in a conservative Protestant faith*religiosity among non-whites DF SS MS F Model 12 27.48 2.29 1.31 Error 104 181.51 1.75 Total 116 208.99 R=.132 b se(b) B age -0.00 0.01 -0.03 educate -0.09 0.10 -0.10 income 0.12 0.08 0.16 ideology -0.01 0.14 -0.01 sexroles 0.05 0.05 0.10 abortion -0.07* 0.03 -0.25 socwlfre -0.07 0.07 -0.11 police -0.02 0.04 -0.05 Continued on next page

PAGE 175

170 Appendix I continued Model 19: OLS regression of control variables on adult murder of a criminal justice employee for membership in a conservative Protestant faith*religiosity among non-whites DF SS MS F Model 12 21.87 1.82 1.32 Error 104 143.36 1.38 Total 116 165.23 R=.132 b se(b) B age -0.00 0.01 -0.04 educate -0.02 0.09 -0.03 income 0.14* 0.07 0.21 ideology -0.06 0.12 -0.05 sexroles 0.03 0.05 0.07 abortion -0.06* 0.03 -0.22 socwlfre 0.04 0.06 0.06 police -0.03 0.04 -0.07 Model 20: OLS regression of control variables on adult planned murder for membership in a conservative Protestant faith*religiosity among non-whites DF SS MS F Model 12 30.14 2.51 1.61 Error 104 162.19 1.56 Total 116 192.32 R=.157 b se(b) B age -0.02 0.01 -0.15 educate -0.11 0.10 -0.12 income 0.12 0.07 0.17 ideology -0.04 0.13 -0.03 sexroles -0.02 0.05 -0.04 abortion -0.08* 0.03 -0.26 socwlfre 0.07 0.07 0.11 police -0.03 0.04 -0.07 Continued on next page

PAGE 176

171 Appendix I continued Model 21: OLS regression of control variables on adult murder of a child for membership in a conservative Protestant faith*religiosity among non-whites DF SS MS F Model 12 31.34 2.61 1.56 Error 104 174.36 1.68 Total 116 205.69 R=.152 b se(b) B age -0.01 0.01 -0.04 educate -0.07 0.10 -0.08 income 0.11 0.08 0.15 ideology 0.09 0.14 0.06 sexroles -0.00 0.05 -0.06 abortion -0.06 0.03 -0.20 socwlfre -0.02 0.07 -0.03 police -0.03 0.04 -0.07 Model 22: OLS regression of control variables on adult unplanned murder for membership in a conservative Protestant faith*religiosity among non-whites DF SS MS F Model 12 20.50 1.71 1.86* Error 104 95.36 0.92 Total 116 115.86 R=.177 b se(b) B age 0.01 0.01 0.10 educate -0.07 0.07 -0.10 income 0.13* 0.06 0.23 ideology -0.05 0.10 -0.05 sexroles 0.01 0.04 0.04 abortion -0.05 0.02 -0.22 socwlfre -0.02 0.05 -0.03 police -0.05 0.03 -0.17 Continued on next page

PAGE 177

172 Appendix I continued Model 23: OLS regression of control variables on adult rape for membership in a conservative Protestant faith*religiosity among non-whites DF SS MS F Model 12 23.20 1.93 1.29 Error 104 156.25 1.50 Total 116 179.45 R=.129 b se(b) B age 0.00 0.01 0.01 educate -0.10 0.10 -0.11 income 0.07 0.07 0.10 ideology 0.12 0.13 0.09 sexroles -0.02 0.05 -0.04 abortion -0.04 0.03 -0.13 socwlfre 0.04 0.06 0.06 police -0.07 0.04 -0.19 Model 24: OLS regression of control variables on adult child molestation for membership in a conservative Protestant faith*religiosity among non-whites DF SS MS F Model 12 15.10 1.26 0.79 Error 104 166.04 1.60 Total 116 181.15 R=.083 b se(b) B age -0.01 0.01 -0.06 educate -0.10 0.10 -0.12 income 0.07 0.07 0.10 ideology -0.05 0.13 -0.04 sexroles -0.02 0.05 -0.04 abortion -0.02 0.03 -0.06 socwlfre -0.02 0.07 -0.03 police -0.06 0.04 -0.16 Continued on next page

PAGE 178

173 Appendix I continued Model 25: OLS regression of control variables on the overall juvenile death penalty scale for membership in a conservative Protestant faith*religiosity among non-whites DF SS MS F Model 12 1159.54 96.63 2.34* Error 104 4294.12 41.29 Total 116 5453.66 R=.213 b se(b) B age 0.06 0.06 0.09 educate -0.99 0.50 -0.21 income 0.71 0.38 0.18 ideology 0.47 0.67 0.06 sexroles 0.21 0.27 0.08 abortion -0.30 0.16 0.19 socwlfre -0.01 0.34 -0.00 police -0.37 0.21 -0.17 Model 26: OLS regression of control variables on general support for juvenile death penalty for membership in a conservative Protestant faith*religiosity among non-whites DF SS MS F Model 12 33.51 2.79 1.71 Error 104 170.18 1.64 Total 116 203.69 R=.165 b se(b) B age 0.01 0.01 0.11 educate -0.06 0.10 -0.07 income 0.09 0.08 0.12 ideology 0.05 0.13 0.03 sexroles 0.08 0.05 0.16 abortion -0.04 0.03 -0.12 socwlfre -0.01 0.07 -0.01 police -0.06 0.04 -0.14 Continued on next page

PAGE 179

174 Appendix I continued Model 27: OLS regression of control variables on juvenile murder of a criminal justice employee for membership in a conservative Protestant faith*religiosity among non-whites DF SS MS F Model 12 24.13 2.01 1.73 Error 104 121.10 1.16 Total 116 145.23 R=.166 b se(b) B age 0.01 0.01 0.10 educate -0.09 0.08 -0.11 income 0.10 0.06 0.16 ideology -0.04 0.11 -0.03 sexroles 0.03 0.04 0.07 abortion -0.02 0.03 -0.09 socwlfre 0.05 0.06 0.09 police -0.04 0.04 -0.11 Model 28: OLS regression of control variables on juvenile planned murder for membership in a conservative Protestant faith*religiosity among non-whites DF SS MS F Model 12 35.91 2.99 1.85* Error 104 168.31 1.62 Total 116 204.22 R=.176 b se(b) B age 0.00 0.01 0.02 educate -0.18 0.10 -0.20 income 0.12 0.08 0.16 ideology 0.12 0.13 0.08 sexroles 0.02 0.05 0.03 abortion -0.05 0.03 -0.18 socwlfre 0.03 0.07 0.04 police -0.04 0.04 -0.11 Continued on next page

PAGE 180

175 Appendix I continued Model 29: OLS regression of control variables on juvenile murder of a child for membership in a conservative Protestant faith*religiosity among non-whites DF SS MS F Model 12 27.20 2.27 1.80 Error 104 131.12 1.26 Total 116 158.32 R=.172 b se(b) B age 0.00 0.01 0.01 educate -0.17 0.09 -0.21 income 0.11 0.07 0.17 ideology 0.09 0.12 0.08 sexroles 0.02 0.05 0.05 abortion -0.06* 0.03 -0.23 socwlfre -0.01 0.06 -0.01 police -0.03 0.04 -0.09 Model 30: OLS regression of control variables on juvenile unplanned murder for membership in a conservative Protestant faith*religiosity among non-whites DF SS MS F Model 12 28.83 2.40 3.03* Error 104 82.42 0.79 Total 116 111.25 R=.259 b se(b) B age 0.02 0.01 0.18 educate -0.12 0.07 -0.17 income 0.14* 0.05 0.25 ideology -0.02 0.09 -0.02 sexroles 0.04 0.04 0.12 abortion -0.05* 0.02 -0.22 socwlfre -0.01 0.05 -0.02 police -0.05 0.03 -0.15 Continued on next page

PAGE 181

176 Appendix I continued Model 31: OLS regression of control variables on juvenile rape for membership in a conservative Protestant faith*religiosity among non-whites DF SS MS F Model 12 32.79 2.73 2.32* Error 104 122.29 1.18 Total 116 155.08 R=.211 b se(b) B age 0.01 0.01 0.07 educate -0.20* 0.08 -0.25 income 0.08 0.06 0.12 ideology 0.16 0.11 0.13 sexroles 0.01 0.05 0.02 abortion -0.05 0.03 -0.17 socwlfre -0.01 0.06 -0.02 police -0.08* 0.04 -0.21 Model 32: OLS regression of control variables on juvenile child molestation for membership in a conservative Protestant faith*religiosity among non-whites DF SS MS F Model 12 22.40 1.87 1.43 Error 104 135.46 1.30 Total 116 157.86 R=.142 b se(b) B age 0.01 0.01 0.07 educate -0.18 0.09 -0.22 income 0.07 0.07 0.10 ideology 0.10 0.12 0.08 sexroles 0.00 0.05 0.01 abortion -0.03 0.03 -0.10 socwlfre -0.05 0.06 -0.08 police -0.06 0.04 -0.17 *Significant at p<.05


xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001461871
003 fts
006 m||||e|||d||||||||
007 cr mnu|||uuuuu
008 040406s2003 flu sbm s000|0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0000228
035
(OCoLC)54938284
9
AJQ2283
b SE
SFE0000228
040
FHM
c FHM
090
HV6025
1 100
Schlaupitz, Sheila M.
0 245
Race, religion, and attitudes toward capital punishment
h [electronic resource] :
a test of attribution theory /
by Sheila M. Schlaupitz.
260
[Tampa, Fla.] :
University of South Florida,
2003.
502
Thesis (M.A.)--University of South Florida, 2003.
504
Includes bibliographical references.
516
Text (Electronic thesis) in PDF format.
538
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
500
Title from PDF of title page.
Document formatted into pages; contains 181 pages.
520
ABSTRACT: The present study attempts to provide a more thorough understanding of public attitudes toward capital punishment. Two theories are tested toward this end, utilizing a random sample taken from the jury pool in Hillsborough County, Florida. First, An indirect test of reference group theory demonstrates the degree to which faith group, religiosity, and race affect public attitudes toward the death penalty. Surprisingly, there is little, albeit mixed, support for reference group theory as it attempts to explain attitudes toward capital punishment. Using the relationships between religion and capital punishment attitudes that we presumed would emerge from the test of reference group theory, we developed a processual model to more accurately describe how the relationships between religion and capital punishment operates. A test of attribution theory provided the vehicle through which this processual model could be tested. Although it seemed that the model was dependent on significant findings between the relationships in reference group theory, this was not the case. Indeed, we found many relationships between certain dimensions of religion and capital punishment attitudes that were consistent with previous research. Moreover, this piece of research is among the first to examine the effect of possessing a more progressive theological emphasis on death penalty attitudes. It addresses important theoretical and empirical questions regarding the direct and indirect relationships between religion and capital punishment attitudes.
590
Adviser: Sellers, Christine S.
653
death penalty.
fundamentalism.
protestantism.
public opinion.
religiosity.
690
Dissertations, Academic
z USF
x Criminology
Masters.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.228