RELIGION, PUNITIVE JUSTICE, AND SUPPORT FOR THE DEATH PENALTY*

Transcription

1 RELIGION, PUNITIVE JUSTICE, AND SUPPORT FOR THE DEATH PENALTY* HAROLD G. GRASMICK JOHN K. COCHRAN ROBERT J. BURSIK JR. M'LOU KIMPEL University of OHahoma The increased punitiveness in the criminal justice system, stimulated at least to some extent by public opinion, has coincided with the revival of Protestant evangelicalism and fundamentalism and with their followers' active involvement in politics and policy debates. Previous research on the determinants of preferred justice policies in the public either ignored religion or relied on a simple distinction among Protestants, Catholics, and Jews. The present study argues that evangelical/fundamentalist Protestants are more inclined to attribute crime to offenders' dispositional characteristics than to situational factors. Consequently they are expected to be more punitive than members of other groups. Survey data from a sample of adults in a southwestern city reveal greater punitiveness among evangelical/fundamentalist Protestants on four of five policy issues, includh~g support for the death penalty both for adults and for juveniles. Criminal justice researchers have failed to seriously explore the connection in the public between religion and justice policy preferences, including attitude toward the death penalty. This fail- ure is perhaps ironic in view of the central role of religion in Durkheim's theory of punishment (see Garland 1983). In David Garland's (1990) recent book Punishment and Modern Society, for which Garland received a Distinguished Scholar Award from the American Sociological Association, he documents that "throughout the history of penal practice religion has been a major force in shaping the ways in which offenders are dealt with" (203). Yet a recent review of research on public opinion concerning crime and * Funding for this research was provided by the College of Arts and Sciences, University of Oklahoma, as part of the annual Oklahoma City Survey conducted by the Department of Sociology and directed by Harold G. Grasmick and Robert J. Bursik Jr. JUSTICE QUARTERLY, Vol. 10 No. 2, June Academy of Criminal Justice Sciences

2 290 RELIGION AND DEATH PENALTY punishment (Warr 1991) does not even mention religion. I Garland's work surely will be influential in thinking about why society punishes and about the functions of punishment as an institution in the larger social structure and culture. The present research, a modest first step, explores the link between religious differences and variation in justice policy preferences in the current American public. PUBLIC OPINION AND SENTIMENTS REGARDING PUNISHMENT The past two decades have been an era of increased punitiveness in the criminal justice system in the United States (see Finckenauer 1988; Pillsbury 1989). In precisely the same period, evangelical and fundamentalist Protestants, labeled by the media as the New Christian Right, became active in politics and public policy debates (see Corbett 1990:209-30; Fowler 1984; Wald 1992). The 1976 Supreme Court case of Gregg v. Georgia (428 U.S. 153) played two roles in the movement toward a more punitive system of criminal justice. First, it opened the gates for states to inflict the death penalty; second, it ruled that public opinion was a legitimate basis for formulating justice policy (428 U.S. 184). Subsequently, states across the country developed new death penalty legislation. Beginning with California, through either direct legislation or the formation of sentencing commissions, they enacted penal policies based on a "just deserts" philosophy (see von Hirsch, Knapp, and Tonry 1987) and, in varying degrees, on a rejection of the rehabilitative model (see Kramer, Lubitz, and Kempinen 1989). As Pillsbury observes, "Recent penal changes have been driven by a perceived need for harsher punishment" (1989:763). In fact, they have accomplished this goal. The same movement, initiated by the Sentencing Reform Act of 1984, occurred at the federal level. As Nagel (1990:884) notes, the U.S. Sentencing Commission was guided in part by public opinion data (see Flanagan and Jamieson 1988:142-43), a fact suggesting that "sentences served were considerably and consistently more lenient than public estimates of what ought to be the normative societal response." The other trend, the involvement of what the media labeled the New Christian Right in public policy issues, was not anticipated by the social sciences. Hadden (1987) observes that sociology i Among the Moral Majority's ten "amendments," one item states the prolife position but makes capital punishment an exception: "Amendment Ill. We believe that, apart from justified capital punishment, no medical or judicial process should be introduced that would allow the termination of life before its natural or accidental completion."

3 GRASMICK, COCHRAN, BURSIK AND KIMPEL 291 had been dominated by a "secularization paradigm," which assurned that in the long run in a rational, bureaucratic, modern society, religion would be inconsequential in the public sphere. The effects of religion on politics and public policy were not worthy of research. Consequently sociologists were ill-prepared to understand the role of religion in policy controversies in recent years. Guth, et al. (1988) provide a similar assessment of political science. Scant attention has been given to the possible religious roots of punishment sentiments in contemporary research. Although some observers (e.g., CuUen et al. 1985; Finckenauer 1988; Langworthy and Whitehead 1986) have noted the connection between conservative politics and increased punitiveness, they have overlooked the role of religion in this process. Instead, the trend away from rehabilitative efforts toward harsher punishment and just deserts typically is explained as a response to Martinson's (1974) purported evidence that "nothing works" in regard to rehabilitation, coupled with an increased fear of crime and an increase in the victimization rate resulting from a rising crime rate (see discussions in Bayer 1981; Rankin 1979; Sechrest, White, and Brown 1979; Sheley 1985). Researchers, however, generally have found no relationship between victimization or fear of crime and punitiveness in the public (Cullen et al. 1985; Stinchcombe et al. 1980; Taylor, Scheppele, and Stinchcombe 1979). Martinson himself was careful to qualify his conclusions and to note exceptions. Some scholars remain convincecl that rehabilitation can be a viable goal of punishment (e.g., Andrews et al. 1990; Cullen and Gendreau 1988; Cullen and Gilbert 1982; Gendreau and Ross 1987). In the academic world, scholars continue to debate about how to interpret the scientific evidence (see Lab and Whitehead 1990). Yet policy makers live in a different world, one in which at least some segments may care little about the data and instead may base their justice policy preferences on their religious convictions. In a recent commentary, Barry Krisberg (1991:143), president of the National Council on Crime and Delinquency, confronted the issue openly. Kxisberg views the fit between punitive justice policies and the "reemergence of puritanism as a dominant force in American political and social discourse." He comments, "In this formulation, evil is a virtually inevitable character flaw of certain individuals," and "social and economic forces are assumed to have little impact on deviant behavior." Krisberg concludes that from this point of view, "punishment becomes the primary societal response to offenders, and religious conversion is the only acceptable rehabilitation program."

4 292 RELIGION AND DEATH PENALTY In this study we offer a rationale for predicting that individuals affiliated with evangelical and fundamentalist Protestant denominations are more punitive than others in their criminal justice policy preferences. Then we examine the relationships between a more refined measure of religious affiliation than has been used in past research and various kinds of justice policy preferences, using survey data from a community in the southwest. The dependent variables are support for the death penalty for juveniles and for adults, and views about the harshness of courts, the severity of prescribed penalties, and the use of deadly force by police. CLASSIFYING RELIGIOUS AFFILIATIONS The literature has become rife with controversy concerning our independent variable. For research purposes, social scientists occasionally have attempted to group the large number of Protestant denominations into a small number of categories that vary along some dimension of religious conservatism to religious liberalism. For example, Glock and Stark's (1965) influential classification scheme contained fundamentalist, conservative, moderate, and liberal categories. The scheme developed recently by Smith (1990) for use with General Social Survey data consists of fundamentalists, moderates, and liberals. We use Smith's classification in the present research. Any such scheme immediately creates disagreement on two grounds. First, some scholars argue that the term fundamentalist, though used casually in the media, among the public, and by some researchers, has a very precise theological and historical meaning (see Kellstedt and Smidt 1991; Marsden 1987). These scholars stress the link between the contemporary fundamentalist movement, whose most visible spokesperson has been Reverend Jerry Falwell and his organization the Moral Majority, and a series of pamphlets titled The Fundamentals, published from 1910 to From a historical and theological perspective, fundamentalists should not be confused with other categories, especially evangelicals. People in churches which are lumped into the "fundamentalist" category but which do not belong to this theological tradition become resentful, no doubt because of genuine doctrinal differences but probably also, as Lechner (1990:95) notes, because "conservative evangelicals resent the term because of the stigma associated with it and because they clearly want to separate themselves from what they consider to be extreme fundamentalism." Others writers, however, insist that what these groups have in common is more important for research than the differences

5 GRASMICK, COCHRAN, BURSIK AND KIMPEL 293 among them, and they see no practical reason to make any fine distinctions (see Barr 1977; Peek, Lowe, and Williams 1991). The common ground between fundamentalists, narrowly defined, and what others would call evangelical or conservative evangelical denominations is 1) an emphasis on interpreting the Bible literally, 2) an emphasis on personal salvation, and 3) a "passion for evangelism and missions" (Kellstedt and Smidt 1991:260; also see Bruce 1983; Jelen 1989; Johnston 1991; Marsden 1991). Most observers, then would, consider fundamentalists a subgroup within evangelicalism (see KeUstedt and Smidt 1991; Smidt, 1983, 1988), viewing fundamentalists as even more strongly committed than other evangelicals to the notion of Biblical inerrancy, as separatists in their lifestyle (also see Ammerman 1987), and as less willing to cooperate with others who do not share their religious convictions (also see Marsden 1987). For our present purposes we use the term evangelical/fundamentalist Protestant denominations, although Smith's (1990) classification scheme, which we adopt, uses the label fundamentalist. Smith's classification is based on a combination of survey data concerning belief in a literal interpretation of the Bible, surveys of church leaders, previous classification schemes, and memberships of denominations in various national organizations of churches. The category he calls fundamentalist actually is a combination of what others would call evangelicals and fundamentalists because the scheme relies so heavily on Biblical literalness. In other words, we adopt his categories but suggest that the one he calls fundamentalist would more appropriately be called evangelical/ fundamentalist. There simply exists no classification that separates these two categories. We choose not to use the term conservative Protestant denominations because the term conservative implies a somewhat consistent political ideology. We do not wish to give the impression that evangelical and fundamentalist Protestants are more conservative than others on all political issues; evidence suggests that they are not (see Wald 1992). Classification schemes lead to a second controversy. Not every individual affiliated with the denominations we call evangelical/ fundamentalist Protestants necessarily adheres to a literal interpretation of the Bible, and some individuals in the other categories no doubt are Biblical literalists. Thus one might argue that persons' religious beliefs, not their religious affiliations, are the preferred measure of evangelical/fundamentalist Protestantism (see Kellstedt and Smidt 1991). For some purposes, no doubt, this argument is valid, and in our analysis we include a measure of personal

6 294 RELIGION AND DEATH PENALTY belief in a literal interpretation of the Bible. Yet, we focus primarily, though not exclusively, on the affiliation measure for two reasons. First, the scant previous research on religion and punishment sentiments relies exclusively on the simple distinction among Protestants, Catholics, and Jews (e.g., Blumstein and Cohen 1980; Flanagan and Maguire 1990:160-61; Harris 1986; Hindelang 1974; Tyler and Weber 1982). Such studies find no effect of religious affiliation. Despite a large number of published studies on attitudes toward the death penalty (for an annotated bibliography, see Radelet and Vandiver 1988), our earnest search uncovered not a single study that distinguished among Protestant denominations as an independent variable. The primary aim of the present article is to demonstrate that once a distinction is made within the Protestant category, the affiliation effect undetected in previous studies will appear Our second reason for emphasizing denominations is the potential role of churches as mobilizing organizations. Evangelical and fundamentalist Protestant churches have all the ingredients to be a powerful force in public policy debates, not only at the national level but also at state and local levels. As Wald, Owen, and Hill (1988) note, churches are by far the most extensive network of voluntary organizations in the United States, providing an organizational base for communicating political messages and for mobilizing policy sentiments into political action. In their more recent work (Wald et al. 1990), these authors find evidence that evangelical and fundamentalist churches are the most effective in instilling political consensus and cohesion among their members. Compared to others, these churches demand more extensive participation and commitment from members and are more intolerant of the questioning of church doctrine. When members of such churches "come to view a particular issue of public policy in religious terms, as a theological imperative, their support for the preferred option is complete and wholehearted" (Wald et al. 1990:211). To emphasize the importance of affiliation, however, as we do in the present research, is not to suggest that personal religious beliefs are unimportant. In fact, we replicate our analysis of the effect of affiliation, replacing affiliation with a measure of the individual's adherence to a literal interpretation of the Bible. We report similarities and differences between the two analyses. In the conclusion, we suggest other dimensions of religious beliefs which should be considered in future research as possible determinants of punishment sentiments.

7 GRASMICK, COCHRAN, BURSIK AND I~MPEL 295 In other areas of social science research, evidence is accllmulating that affiliation with an evangelical or fundementalist Protestant denomination has a direct effect on individuals' positions on various policy issues, especially issues concerning the family, gender roles, and sexuality (e.g., Cochran and Beeghley 1991; Grasmick, Wilcox, and Bird 1990; Hertel and Hughes 1987; Luker 1984; Peek and Brown 1980; Peek et al. 1991; Tedin 1978; Wood and Hughes 1984). Wald et al. (1990) label this complex of issues "moral conservatism." Although many scholars appear to suggest that this complex is limited to matters related to the family and sexuality (for a summary of this view, see Grasmick et al. 1990), we propose below that it also includes a harsh punitive response to those who break the law. A few existing studies offer some indirect evidence of a possible connection between evangelical/fundamentalist affiliation and punitiveness toward those who violate the rules. In her study of sentencing behavior among judges in Georgia, Myers (1988) reports that with controls for offender and offense characteristics and other characteristics of judges, judges affiliated with evangelical/ fundamentalist Protestant denominations are more likely than others to sentence with imprisonment rather than with parole. Grasmick, Davenport, et al. (1992) find that in a sample of adults, individuals identifying themselves with evangelical/fundamentalist Protestant denominations are more likely to endorse retribution as the goal of punishment than are other Protestants, Catholics, and those claiming no religious affiliation. (The number of Jews in their sample was too small for inclusion in the analysis.) Grasmick, Morgan, and Kennedy (1992) offer evidence that support for corporal punishment in the schools is significantly stronger among evangelical/fundamentalist Protestants than among others. RELIGION AND ATTRIBUTION Hadden (1987) is correct in observing that theories about why religion has its effect are underdeveloped, and the development of such a theory is beyond our intentions. Even so, recent discussions of religion in the context of attribution theory in social psychology can help us to understand why evangelicals and fundamentalists might advocate more punitive criminal justice policies than others. Attribution theory (Heider 1958) attempts to explain the causes that individuals ascribe to events and behavior; recently efforts have been made to apply the theory to religion. Some writers have devised schemes to predict conditions under which individuals will make religious rather than secular attributions (Shaver

8 296 RELIGION AND DEATH PENALTY 1985:137-54; Spilka, Shaver, and Kirkpatrick 1985). More interesting for our purposes, however, is the discussion of possible denominational differences in "attributional styles" (Gorsuch and Smith 1983; Slupfer, Hopkinson, and Kelley 1988; Lupfer and Wald 1985). Vold (1958:282) noted long ago that views on what causes crime will affect views on what should be done about it (also see Cullen et al. 1985). In addition to an emphasis on Biblical literalness, evangelical/ fundamentalist Protestants are preoccupied with individuals' "character" because character is a manifestation of salvation status (Wuthnow 1976; Zeman 1981; also see Krisberg 1991). Behavior in accord with Biblical teachings demonstrates good character; behavior that defies religious prescriptions and proscriptions, which typically are incorporated into secular law, reflects bad character. The emphasis on character as the key determinant of behavior is linked to the distinction between dispositional and situational attributions emphasized in attribution theory (see Shaver 1975). Dispositional attributions explain behavior as a consequence of characteristics of the actor. Situational attributions interpret behavior in terms of influences of the environment in which it occurs. Because of their concern with character, it has been suggested that evangelicals and fundamentalists are more likely to adopt a dispositional than a situational attribution style, and thus to commit what attribution theorists call "the fundamental attribution error"--the tendency to overestimate the importance of dispositional causes while underestimating the relevance of situational causes. Thus Lupfer and Wald (1985:299) suggest that such persons will be inclined to explain behavior in terms of an actor's character and will be insensitive to situational causes (also see Lupfer et al. 1988:390). In this context, Cullen et al. (1985:324) speculate that the public tendency to blame crime on offenders rather than on their social surroundings has been increasing. Although these authors do not make the connection, the period to which they refer corresponds to the period of growth of evangelical and fundamentalist Protestantism. When a dispositional attribution style is adopted, the causal explanation of crime is straightforward: crimes are committed by bad people. On the basis of such a premise, then, criminality can be equated with sinfulness, which must be punished (also see Warr 1989:818-19; for a description of this view of criminals in seventeenth-century Massachusetts, see Zeman 1981). Shaver (1975) argues that a dispositional rather than a situational attribution makes an actor more blameworthy from the perspective of the one making the attribution, and thus deserving of greater punishment.

9 GRASMICK, COCHRAN, BURSIK AND ~MPEL 297 In fact, Carroll (1978) reports that dispositional versus situational attributions influence decisions about parole: offenders whose crimes are attributed to dispositional factors are perceived as poor risks and are more likely to be denied parole. Furthermore, in a vignette study of college students, Hawkins (1981) shows that individuals who attribute a crime to dispositional rather than situational factors assign a higher preferred punishment score to the crime across a wide variety of types of crime for both adult and juvenile offenders. In their surveys of the public and of criminal justice policy makers and practitioners, Cullen et al. (1985) find that people who attribute crime to the offenders' character are significantly more likely to favor the death penalty than those who consider social factors to be the cause of crime, score significantly higher on a scale measuring punitiveness toward offenders, and score significantly lower on a scale measuring support for rehabilitation as a valuable goal of criminal justice policy (also see Harem 1989; for a discussion of other somewhat relevant issues and findings from attribution theory, see Perlman 1980). Therefore we are suggesting that evangelical/fundamentalist Protestants are more inclined than others to blame crime on the offender's disposition. This attribution, in turn, calls for more severe punishment. Although the present study does not directly examine the relationship between religious affiliation and attribution style, this argument serves as the basis for our prediction that those affiliated with evangelical/fundamentalist Protestant denominations will prefer more punitive criminal justice policies, including the death penalty. Samp~ METHOD Data to compare punitive justice preferences across categories of religious affiliation were collected in spring 1991 as part of the thirteenth annual Oklahoma City Survey, conducted by the Department of Sociology at the University of Oklahoma. A simple random sample of 395 adults (age 18 and older) was drawn from the R.L. Polk Directory for the city, and initial contact was made through a letter stating that a member of the research team soon would try to schedule an appointment for an interview. Attempts to schedule appointments were made in person by interviewers. Members of the target sample who refused to participate or who could not be located were replaced by random selection until a total of 395 face-to-face interviews had been conducted.

10 298 RELIGION AND DEATH PENALTY Unfortunately, 1990 Census data are not yet available for comparison with the sample, so we compared the sample to 1980 Census figures for the community. The sample does not differ significantly from the 1980 population in percent female (54.3% in the sample, 53.2% in the population) or percent white (81.5% in the sample, 83.7% in the population). The mean age of the 1991 sample (46.5), however, is significantly higher than the mean age of the 1980 population (42.7). Most likely, this increase does not reflect sampling bias but rather an actual aging of the adult population, consistent with national trends. In fact, the mean age of the annual Oklahoma City Survey samples has been increasing steadily. We use a listwise deletion of missing cases, which results in an N of 370 for the analysis. The greatest loss of cases, as described below, occurs in the classification of religious affiliation. Measures Dependent variable. Although we agree with Bowers (1990) that more refined measures of the public's punishment sentiments are needed, this is not our mission in the present study. Rather, our objective is to determine whether significant religious differences in conventional punitiveness measures will appear when evangelical/fundamentalist Protestants are separated from other Protestants. Five justice preference items were included in the survey. Each was presented to the respondent in a Likert format with response options of strongly agree, agree somewhat, disagree somewhat, and strongly disagree. For purposes of analysis, we collapsed the two "agree" and the two "disagree" categories, creating dummy variables coded I for those who agreed with the statement. The items and the percentages of respondents who agree with them are as follows: Juvenile Death Penalty: "Sixteen-year-olds who are convicted of first degree murder generally deserve the death penalty" (51.4% agree). Adult Death Penalty: "Adults who are convicted of first degree murder generally deserve the death penalty" (74.9% agree). Harsh Courts: "When someone is tried and convicted of a serious crime, the judge always should impose the most severe punishment which the law allows" (66.8% agree). Stiffer Laws: "Lawmakers should make the penalties for crimes even stiffer than they are now" (80.8% agree).

11 GRASMICK, COCHRAN, BURSIK AND KIMPEL 299 Deadly Force: "Police officers should have the right to shoot a suspect who is fleeing the scene of a burglary" (52.2% agree). No respondent refused to answer any of these questions. 2 Independent variable. Smith's (1990) classification of Protestant denominations, described earlier, is used in coding respondents' reports of their religious affiliation, although we use the label evangelical/fundamentalist rather than fundamentalist. Although Smith identifies three categories of Protestant denominations-- fundamentalist, moderate, and liberal-- our sample contains too few liberals (primarily Presbyterians and Episcopalians) to be included as a separate category. Thus the liberal and the moderate categories are combined into a single category to be compared with evangelical/fundamentalist Protestants. In view of the location of the sample in the Bible Belt, it is not surprising that 46.7 percent of the respondents fall into the evangelical/fundamentalist category. The largest representations are Southern Baptist (n -- 74), other Baptist (n ), and Assembly of God (n -~ 20). Methodists (n -- 43) constitute nearly half of the 26.2 percent who are in the liberal/moderate Protestant Category. We recognize, as does Smith, that this scheme probably contains some classification errors. Nevertheless, it is grounded empirically, and, unlike the scheme presented by Glock and Stark in 1965, it is based on recent data. Also included in the analysis as separate categories are 11.9 percent of the 370 respondents who identify themselves as Catholics and 8.6 percent who claim no religious affiliation. Smith makes no distinction among types of Catholics in his scheme, even though Catholics no doubt vary according to their parish's order along the same traits identified as characteristic of evangelical and fundamentalist Protestant denominations. Excluded from the analysis because of their small numbers are 15 Jews and other non-christians. Smith's classification scheme for Protestants is quite inclusive, containing nearly 150 denominations grouped into one of the three categories. Thus only three additional cases were lost from the analysis because they did not appear in his list. The questionnaire also included four items measuring belief in a literal interpretation of the Bible. Each was answered on a fourpoint Likert scale with response options ranging from strongly 2 We considered creating a single scale from tl~ese five items, but the interitem correlations were not high, and the reliability coefficient (alpha) for the scale was less than.60. Furthermore, as revealed in the analysis, religlous affiliation and the control variables do not have consistent effects across the five types of punitiveness.

12 300 RELIGION AND DEATH PENALTY agree to strongly disagree. The items are as follows: 1) I believe the miracles described in the Bible really happened; they are not just stories; 2) I believe in a literal interpretation of the Bible; 3) I believe that Jesus truly rose from the dead; 4) I believe that those who do not accept God will go to hell after their death. A factor analysis showed clearly the presence of a single factor with eigenvalues of 2.59,.66,.53, and.22. The linear composite of z- scores, called Biblical literalness, has a reliability (Cronbach's alpha) of.82. The mean scores on this sale are 1.30 for evangelical fundamentalist Protestants, for liberal/moderate Protestants, for Catholics, and for no affiliation. All pairwise comparisons of evangelical/fundamentalist Protestants with each of the other three categories are highly significant beyond the.001 level, with t-values of for liberal/moderate Protestants, for Catholics, and for no affiliation. Control variables. As controls we include standard socioeconomic and demographic variables that others (see Blumstein and Cohen 1980; Cullen et al. 1985; Langworthy and Whitehead 1986) suggest might be correlated with punitiveness. Gender is a dummy variable coded 1 for males (45.1%); race is a dummy variable coded 1 for whites (83.2%). Age (mean = 46.9, standard deviation ) and education (mean = 13.5, standard deviation = 2.7) are interval variables measured in years. Post-high school training such as trade school is treated as equivalent to a year of college. Family income, measured in thousands of dollars, also is included as a control. In the entire sample, 16 respondents refused to report their income. We used regression procedures to estimate the incomes for these cases based on years of education and the household head's occupational prestige score. For seven of these, however, we were unable to determine an occupational prestige score. These seven, plus the 18 lost in classifying religious affiliation, constitute the 25 cases omitted from the analysis. Among the 370 cases included, the mean family income, in thousands, is 37.1 (standard deviation ). Political party identification is another important control variable. As noted earlier, increased punitiveness has been linked to political conservatism; we want to control for the possibility that the punitiveness of evangelical/fundamentalist Protestants is an artifact of their political allegiance, especially their attraction to the Republican Party in recent years (see Wald 1992), rather than a direct effect of their religion. Democrats make up 45.9 percent of the sample; Republicans, 37.1 percent; independents, 17.0 percent.

13 GRASMICK, COCHRAN, BURSIK AND KIMPEL 301 In other contexts (e.g., Hertel and Hughes 1987), researchers have noted the importance of controlling for religiosity in examining the effects of denominational affiliation. This step allows for the possibility that the dependent variable is affected by the higher level of religiosity among evangelicals and fundamentalists, rather than by their denominational affiliation per se. Our measure of personal religiosity is designed to capture the concept of "religious identity salience"mthe importance that individuals attach to religion as a guide for day-to-day decisions (Wimberley 1989). We included four Likert items to measure this concept, with four response options ranging from strongly agree to strongly disagree. The items are as follows: 1) Religion is a very important part of my life; 2) I would describe myself as very religious; 3) Religion should influence how I live my life; 4) When I have decisions to make in my everyday life, I usually try to find out what God wants me to do. Principal-components analysis shows clearly the presence of a single factor (eigenvalues = 3.10,.41, 27,.21). The religiosity scale is the linear composite of z-score transformations of these four items and has an alpha of.90. Evangelical/fundamentalist Protestants score significantly higher on this scale than any of the other three affiliation category (~ ~.001 for liberal/moderate Protestant and for no affiliation; p for Catholic). Bivariate Relationships. ANALYSIS The simple bivariate relationships between the religious affiliation categories and support for the punitive justice policies are presented in Table 1 in the form of one-way analyses of variance. Caution should be used in interpreting the significance of these relationships because the F-test refers to overall differences among the categories. Our concern is with differences between evangelical/fundamentalist Protestants and each of the other three; these comparisons will be made in the multivariate analysis. The proportion favoring a punitive policy is greatest among evangelical/fundamentalist Protestants for Juvenile Death Penalty, Adult Death Penalty, Harsh Courts, and Stiffer Laws. Deadly Force does not fit this pattern. Furthermore, Catholics tend to fall almost exactly halfway between liberal/moderate Protestants and evangelical/fundamentalist Protestants in proportions favoring Harsh Courts and Stiffer Laws. For the two death penalty measures, however, Catholics more closely resemble liberal moderate Protestants. Finally, across all five policy issues, those claiming no affiliation are the group least likely to be punitive, a

14 302 RELIGION AND DEATH PENALTY Table 1. Bivariate Relationships (ANOYA) between Religious Affiliation and Preferences for Punitive Justice Policies No Religious Affiliation Catholic Liberal/Moderate Protestant Evangelical/Fundamentalist Protestant Eta 2 F3.~ P Proportion Favoring Punitiveness Juvenile Adult Death Death Harsh Stiffer Deadly Penalty Penalty Courts Laws Force finding consistent with other evidence (see Blumstein and Cohen 1980:240; Flanagan and Magnire 1990:160-61). Multivariate Analysis Table 2 reports the multivariate analyses in the form of logistic regressions. One-tailed tests of significance are appropriate because direction is predicted. In the five equations, the four religious affiliation categories are captured by dummy variables for no affiliation, Catholic, and liberal/moderate Protestants. Evangelical/fundamentalist Protestant is the suppressed category, coded 0 on all three of these dummy variables. Thus the coefficient for any one of the dummy variables compares that affiliation category to evangelical/fundamentalist Protestants, with all other variables controlled. We expect these coefficients to be negative. In the dummy coding of political party identification, Republican is the suppressed category; the dummy variables Democrat and Independent compare each of these two categories to Republicans. Before we examine the effects of religious affiliation, the effects of the control variables merit consideration. First, the religiosity scale consistently fails to have a significant effect across all five policy issues; its sign even fluctuates from issue to issue. Our scale measuring "religious identity salience" clearly makes no direct contribution to shaping respondents' justice policy preferences. In general, compared with previous research, Table 2 offers no surprises concerning the effects of the socioeconomic and demographic variables. For two of the issues (Harsh Courts and Stiffer laws) males actually tend to be less punitive than females, although the differences are not significant at the conventional.05 level. Only for Juvenile Death Penalty are men significantly more

15 Table 2. Logistic Regressions of Preferences for Punitive Justice Policies on Religious Affiliation and Control Variables a Juvenile Death Penalty Adult Death Penalty Harsh Courts Stiffer Laws Deadly Force b SE... p b SE p b SE p b SE p b SE p t~ No Religious Affiliation -1, , , , Catholic , ,458 "255, Liberal/Moderate Protestant , ,635,344, Religiosity Scale , Male, ,441, White Age (years) Education (years) Family Income (thousands of $), ,000,004,500,006,004,094, ,138 Democrat , , Independent,113, , , (intercept) - 0, , ,227 Evangelical/Fundamentalist Protestant variables for political party. is the suppressed category in the dummy variables for religious affiliation. Republican is the suppressed category in the dummy

16 304 RELIGION AND DEATH PENALTY likely to be punitive than women. These findings are essentially similar to those of Cullen et ah (1985): gender was unrelated to their measure of general punitiveness, but males scored higher on a measure of support for capital punishment (which did not distinguish between adults and juveniles). Race effects are not uniform across issues in Table 2, Whites and nonwhites do not differ significantly for Harsh Courts and Stiffer Laws, but whites are significantly more likely than nonwhites to favor the use of deadly force by police, capital punishment for adults, and capital punishment for juveniles. Age has a significant effect only for Deadly Force, where support increases with age. (Cullen et al find that age is unrelated to support for the death penalty but is related positively to their general punitiveness scale.) With the exception of Juvenile Death Penalty, where the effect is not significant, education is linked consistently to less punitiveness, but family income is not significant in any of the equations. Because the family income variable is skewed positively by a few outliers, we reestimated the equations in Table 2 using the log of family income, but the log transformation did not alter any of the conclusions. The tendency of education to reduce punitiveness and of income to have little effect is consistent with Blumstein and Cohen's (1980) results. Findings concerning political party tend to reveal greater punitiveness among Republicans. Although independents do not differ significantly from Republicans in any of the five equations, Democrats are significantly less likely than Republicans to believe that the laws should be stiffer and that the police should be allowed to use deadly force, and significantly less likely to support the death penalty for adults. The direct effects of religious affiliation are readily apparent in Table 2. With the exception of Deadly Force, all other equations offer evidence of greater punitiveness among evangelical/ fundamentalist Protestants than among at least some of the other affiliation categories. The pattern is most pronounced for Juvenile Death Penalty: those claiming no affiliation, Catholics, and liberal/moderate Protestants are all significantly less likely than evangelical/fundamentalist Protestants to favor the execution of juveniles. This finding complements the evidence from Grasmick, Morgan, and Kennedy (1992) that evangelical/fundamentalist Protestants tend more than others to support corporal punishment in the schools (also see Maurer 1982). The findings concerning the religious affiliation categories in the Juvenile Death Penalty equation contrast sharply with those concerning political party.

17 GRASMICK, COCHRAN, BURSIK AND KIMPEL 305 Neither Democrats nor independents differ significantly from Republicans. Even with all other variables controlled, liberal/moderate Protestants and those claiming no affiliation are significantly less likely to favor executing adults than are evangelical/fundamentalist Protestants. Catholics too are less supportive of this policy than evangelical/fundamentalist Protestants, but the difference is not significant at the.05 level. Because of the geographical location of our sample, the proportion Catholic is relatively small. As a result, the standard error for the logistic regression coefficient tends to be inflated, and the significance of the coefficient is attenuated. In fact, in the Adult Death Penalty equation, the standard error is.405 for Catholics and.223 for liberal/moderate Protestants. It is possible that in a national sample, or in a local sample with a higher proportion of Catholics, Catholics also would emerge as significantly less supportive of the death penalty for adults than evo angelical/fundamentalist Protestants. The same argument concerning the consequences of a small number of cases would apply to the category of no affiliation. In fact, the standard error of.484 for the coefficient of that group is even higher than the standard error for Catholics. Nevertheless, the effect of no affiliation is so pronounced that it is highly significant. Again, in the equation for Harsh Courts, all three of the other affiliation categories are less punitive than evangelical/fundamentalist Protestants, but the difference is significant only for liberal/ moderate Protestants. As in the equation for Juvenile Death Penalty, we find no significant political party differences in punitiveness, but a significant religious affiliation effect is present. In other words, political party affiliation has no direct effect on support for harsh courts, but religious affiliationmthe distinction between evangelical/fundamentalist and other Protestants~has such an effect. The statistical insignificance of the coefficient for no affiliation in the Harsh Courts equation is somewhat surprising in view of the substantial bivariate difference between this category and evangelical/fundamentalist Protestants in Table 1. A more detailed analysis revealed that education is primarily responsible for the insignificance of no affiliation in the Harsh Courts equation: those claiming no religious affiliation are more highly educated than evangelical/fundamentalist Protestants, and high levels of education are linked significantly to lower levels of support for harsh courts. For Stiffer Laws, the difference between those claiming no affiliation and evangelical/fundamentalist Protestants, observed in

18 306 RELIGION AND DEATH PENALTY the bivariate table, remains as a significant direct effect when the controls are added in Table 2. Both liberal/moderate Protestants and those claiming no affiliation are significantly less supportive of stiffer laws than are evangelical/fundamentalist Protestants. Catholics also are less punitive on this issue, but the difference is not significant. In summary, Table 2 reveals strong evidence for direct effects of religious affiliation on preference for punitive justice. Liberal/ moderate Protestants are significantly less punitive than evangelical/fundamentalist Protestants on all issues except the use of deadly force by the police. On the issues of harsh courts, stiffer laws, and the death penalty for both adults and juveniles, Catholics fall between the two groups of Protestants, but they are significantly less punitive than evangelical/fundamentalist Protestants only on the issue of executing juveniles. Claiming no affiliation is linked strongly to lower punitiveness than among evangelical/fundamentalist Protestants on the issues of stiffer laws and the death penalty for both juveniles and adults. In general, religious affiliation effects are as strong as political party effects, or stronger, for all issues except deadly force. For that issue, Democrats are significantly less punitive than Republicans, while all religious affiliation effects are insignificant. Substituting the Biblical Literalness Scale for Affiliation Categories We repeated the analysis in Table 2, substituting the Biblical literalness scale for the affiliation categories. In view of the large differences between evangelical/fundamentalist Protestants and the other categories on the Biblical literalness scale, noted earlier, it is not surprising that the results are generally the same. For Juvenile Death Penalty, Biblical literalness had a significant positive direct effect (t = 2.37; p =.009). As in the equation containing the affiliation categories, we found no political party effects. For Adult Death Penalty, Biblical literalness also had a significant positive direct effect (t = 2.50; p ). Here, as in the equation containing the affiliation categories, Democrats were significantly less supportive than Republicans (t = -1.78; p =.038); independents did not differ significantly from Republicans. Biblical literalness had a significant positive direct effect on Harsh Courts (t = 2.29; p ). In keeping with the equation containing affiliation categories, no political party effects were significant.

19 GRASMICK, COCHRAN, BURSIK AND KIMPEL 307 The results for Deadly Force and Stiffer Laws were somewhat different when Biblical literalness replaced the affiliation categories. In Table 2, when we used the affiliation categories, there were clearly no affiliation effects for Deadly Force. Yet when Biblical literalness was used instead of the affiliation categories, its effect on support for Deadly Force was positive and approached statistical significance (t = 1.37; p =.085). Democrats also were significantly less supportive of Deadly Force (t ; p <.001), as they were in Table 2. The difference was reversed for Stiffer Laws. In Table 2, both liberal/moderate Protestants and those claiming no affiliation scored significantly lower on support than evangelical/fundamentalist Protestants. When the Biblical literalness scale was substituted for the affiliation categories, however, Biblical literalness revealed no significant effect (t =.63; p =.264), although the sign was positive. Democrats, as in Table 2, were significantly lower than Republicans on support for Stiffer Laws (t = -1.66; p ). In all five equations in which Biblical literalness replaced the affiliation categories, the scale measuring religiosity failed to have a significant effect; in fact, the sign was not consistent across all five equations. This is exactly the pattern observed in Table 2. Tests for Interaction In addition to the additive models reported in Table 2, we considered possible types of interaction effects suggested by other researchers. First, Hertel and Hughes (1987), in their study of family policy issues, suggest that the effects of religiosity might not be the same for all affiliation categories. In the case of punitive justice preference, it is possible that increased religiosity increases punitiveness among evangelical/fundamentalist Protestants but reduces it among those in the other religious affiliation categories. To consider this possibility initially, we added to the equations in Table 2 product terms formed by multiplying each of the three affiliation dummy variables by the religiosity scale. With one exception, none of these product terms were significant in any of the five equations; this finding suggests no significant interaction. The exception was the term for the liberal/moderate Protestant religiosity scale in the equation for Stiffer Laws. A more detailed analysis revealed a significant inverse effect of religiosity on Stiffer Laws among liberal/moderate Protestants, but no effect among the other affiliation categories. In other words, religiosity reduces punitiveness in reference to stiffer laws, but only among the liberal/moderate Protestants. This significant interaction is masked

20 308 RELIGION AND DEATH PENALTY by the insignificant overall direct effect of religiosity on Stiffer Laws in Table 2. Peek et al. (1991) report ttmt affiliation with a fundamentalist denomination increases sexism among men but not among women. Because the nurturing role is assigned to women in the traditional sexual division of labor, it is possible that religious affiliation has less of an effect on punitiveness for women than for men. To address this issue, we added product terms formed by multiplying each of the dummy variables for religious affiliation by the dummy variable Male to each of the equations in Table 2. None of the product terms were significant, suggesting no affiliation gender interaction, for Harsh Courts, Stiffer Laws, Deadly Force, and Juvenile Death Penalty. The product terms were significant, however, in the Adult Death Penalty equation for no affiliation and for Catholics. A more detailed analysis revealed that among people in this combination of categories, being male significantly increases the likelihood of supporting the death penalty for adults. Among evangelical/fundamentalist and liberal/moderate Protestants combined, however, gender has no effect. The overall effect of Male in the Adult Death Penalty equation in Table 2 is positive but not significant. Significance is lacking because our sample is overwhelmingly Protestant. Thus we have some evidence of a religious affiliation X gender interaction for one justice policy issue, but the nature of the interaction is not what we anticipated from the findings of Peek et al. (1991). Finally, we wanted to consider the possibility of an affiliation X race interaction. Specifically, because minorities are overrepresented as targets of punitive justice, it is possible that their views on justice policy issues are influenced less by their affiliation than by the fact that they are common targets. Unfortunately, the number of nonwhites in the sample (70% of whom are black or American Indian) was too small to allow computation of product terms. Hardly any minority-group members were in the categories of no affiliation (n ~- 2) or Catholic (n = 5). As an alternative, we replicated the logistic regressions in Table 2 for whites only. If whites are influenced more strongly than nonwhites by their religious affiliation, the effects of affiliation observed in Table 2 should be even greater when the equations are estimated for whites only. Generally, this was not the case. The conclusions concerning Juvenile Death Penalty, Harsh Courts, Stiffer Laws, and Deadly Force are identical whether the analysis is based on the entire sample or on whites only. A difference emerges for Adult Death Penalty, however. In Table 2 (the entire sample), those with no affiliation and liberal/moderate Protestants were

21 GRASMICK, COCHRAN, BURSIK AND KIMPEL 309 significantly less supportive of the death penalty for adults than were evangelical/fundamentalist Protestants. Catholics also were less supportive, but the difference was not significant. When we reestimated the equation for whites only, however, all three affiliation categories in the equation were significantly less supportive than evangelical/fundamentalist Protestants, including Catholics (p =.022). A more detailed analysis revealed that nonwhite evangelical/fundamentalist Protestants were less supportive of the death penalty for adults than their white counterparts. When the nonwhites were eliminated from the analysis, the difference between evangelical/fundamentalist Protestants and Catholics was large enough to be significant. DISCUSSION We have not produced, and do not claim to have produced, the definitive study of the relationship between religion and justice preferences. Our sample is relatively small and is drawn from a single community. Only additional research will reveal whether our findings are generalizable to the United States as a whole. Our scheme for classifying denominations (Smith 1990) is relatively new and untested, and fails to distinguish between evangelical and fundamentalist affiliations. Those who believe such a distinction would be important in understanding justice preferences should build on Smith's work and should develop separate categories for these two groups. Furthermore, Smith's scheme does not recognize possibly relevant differences among Catholic orders. This shortcoming probably will be more difficult to overcome without data in addition to those provided by Smith. Our own research, because of the geographical location of the sample and its relatively small size, made it impossible to distinguish between the liberal and the moderate Protestants identified in Smith's scheme; the two categories had to be combined. The number of Jews in the sample was so small that they had to be excluded from the analysis. We are aware that affiliation categories such as those developed by Smith do not directly tap individuals' religious beliefs. In fact, we included a measure of adherence to a literal interpretation of the Bible as an alternative to the affiliation categories. The results differed somewhat, but strong similarities also appeared, especially regarding support for the death penalty for both juveniles and adults. Although we would encourage others to explore additional dimensions of religious beliefsmsuch as having a "bornagain" experience (see Smidt 1983) and being a self-identified fundamentalist (see Kellstedt and Smidt 1991), among others (see

22 310 RELIGION AND DEATH PENALTY Smidt 1989)--as possible determinants of justice policy preferences, we would not dismiss religious affiliation as an unimportant variable. As noted earlier, Wald et al. (1990) stress the importance of churches as organizational actors in translating religious beliefs into policy positions. Garland's (1990) observation that religion has shaped punishment sentiments throughout history should encourage researchers to examine a wide variety of religious variables as possible determinants of public policy preferences in criminal justice. Clearly, the old simple distinction among Protestants, Catholics, and Jews provides no insights. We hope our research has opened the door for other researchers to explore this question more thoroughly. Prior research (e.g., Carroll 1978; Cullen et al. 1985; Hawkins 1981) showed the importance of attributional styles in shaping justice policy preferences. We hope our tentative theory concerning the possible link between evangelical and fundamentalist Protestantism and dispositional attributions, via the emphasis on salvation status, will stimulate others to consider in greater detail precisely how religion affects justice policy preferences. Some readers perhaps will conclude that although our arguments and findings might be interesting, they have no relevance to the debate concerning justice policy. We would counter by asserting that such an interpretation is a continuation, by social and behavioral scientists, of pushing religion into the background, even though it has come increasingly to the foreground in the public. Those scholars and policy makers who criticize the rehabilitative and even the deterrent models of justice should recognize that much of the public support for their position does not stem from the impact of such eloquent critics as James Q. Wilson and Ernst Van den Haag (see Krisberg 1991) or from the retributivist themes in the writings of Hegel and Kant. Instead much public support for the "justice model" probably rests on a religious conviction that crimes are committed by people with weak, sinful characters. We say this not as a criticism but as an acknowledgment. On the other hand, those who are unwilling to lay the rehabilitative model to rest, who continue to insist that justice policy should be aimed not merely at reacting to crime but also at reducing it, must confront the religious roots of the just deserts sentiments in the public (see Grasmick, Davenport, et al. 1992). Any research evidence concerning social causes of crime and the possible effectiveness of carefully designed rehabilitation or intervention programs is unlikely to alter these public sentiments. For much of the public, as Garland (1990) argues, the primary function

23 GRASMICK, COCHRAN, BURSIK AND KIMPEL 311 of punishment may be to express emotional moral outrage rather than to control or reduce crime. REFERENCF~ Ammerrnan, N.T. (1987) Bible Believers: Fundamentalists in the Modern World. New Brunswick: Rutgers University Press. Andrews, D.A., I. Zinger, R.D. Hoge, J. Bonta, P. Gendreau, and F.T. Cullen (1990) "Does Correctional Treatment Work? A Clinically Relevant and Psychological Informed Meta-Analysis." Criminology 28: Barr, J. (1977) Fundamentalism London: SCM. Bayer, R. (1981) "Crime, Punishment, and the Decline of Liberal Optimism." Crime and Delinquency 27: Blumstein, A. and J. Cohen (1980) "Sentencing of Convicted Offenders: An Analysis of the Public's Views." Law and Society Review 14: Bowers, W.J. (1990) '~ne Death Penalty's Shaky Support." The New York Times, May 28, p. 32. Bruce, S. (1983) "Identifying Conservative Protestantism." Sociological Analysis 44: Carroll, J.S. (1978) "Causal Theories of Crime and Their Effect upon Expert Parole Decisions." Law and Human Behavior 2: Cochran, J. and L. Beoghley (1991) "The Influence of Religion on Attitudes toward Nonmarital Sexuality: A Preliminary Assessment of Reference Group Theory." Journal for the Scientifw Study of Religion 30: Corbett, J.M. (1990) Religion in America. Englewood Cliffs, NJ: Prentice-Hall. Cullen, F.T., G.A. Clark, J.B. Cullen, and R.A. Mathers (1985) "Attribution, Salience, and Attitudes toward Criminal Sanctioning." Criminal Justice and Behavior 12: Cullen, F.T. and P. Gendreau (1988) '~nne Effectiveness of Correctional Rehabilitation." In L. Goodstein and D.L. MacKenzie (eds.), The American Prison: Issues in Research Policy, pp New York: Plenum. Cullen, F.T. and K.E. Gilbert (1982) Reaffirming Rehabilitation. Cincinnati: Anderson. Finckenauer, J.O. (1988) "Public Support for the Death Penalty: Retribution as Just Deserts or Retribution as Revenge?" Justice Quarterly 5: Flanagan, T.J. and K.M. Jamieson, eds. (1988) Sourcebook of Criminal Justice Statistics Washington, DC: U.S. Government Printing Office. Flanagan, T.J. and K. Maguire, eds. (1990) Sourcebook of Criminal Justice Statis. tics Washington, DC: U.S. Government Printing Office. Fowler, R.B. (1984) Religion and Politics in America. Metuchen, NJ: Scarecrow Press. Garland, D. (1983) "Durkheim's Theory of Punishment: A Critique." In D. Garland and P. Young (eds.), The Power to Punish: Contemporary Penality and Social Analysis, pp London: Heinemann. (1990) Punishment and Modern Society. Chicago: University of Chicago Press. Gendreau, P. and R.E. Ross (1987) "Revivification of Rehabilitation: Evidence from the 1980s." Justice Quarterly 4: Glock, C. and R. Stark (1965) Religion and Society in Tension. Chicago: Random House. Gorsuch, R.J. and C.S. Smith (1983) "Attributions of Responsibility to GOd: An Interaction of Religious Beliefs and Outcomes." Journal for the Scientific Study of Religion 22: Grasrnick, H.G., E. Davenport, M.B. Chamlin, and R.J. Bursik Jr. (1992) "Protestant Fundamentalism and the Retributivist Doctrine of Punishment." Criminology 30: Grasmick, H.G., C.S. Morgan, and M.B. Kennedy (1992) "Support for Corporal Punishment in the Schools: A Comparison of the Effects of Socio-Economic Status and Religion." Social Science Quarterly 73:

24 312 RELIGION AND DEATH PENALTY Grasmick, H.G., L. Wilcox, and S. Bird (1990) "The Effects of Religious Fundamentalism and Religiosity on Preference for Traditional Family Norms." Sociological Inquiry 60: Guth, J.L., T.G. Jelen, L.A. Kellstedt, C.E. Smidt, and K.D. Wald (1988) "The Politics of Religion in America: Issues for Investigation." American Politics Quarterly 16: Hadden, J.K. (1987) "Toward Desacralizing Secularization Theory." Social Forces 65: Harem, M.S. (1989) "Legislator Ideology and Capital Punishment: The Special Case for Indiana Juveniles." Justice Quarterly 6: Harris, P.W~ (1986) "Over-Simplification and Error in Public Opinion Surveys on Capital Punishment." Justice Quarterly 3: Hawkins, D.F. (1981) "Causal Attribution and Punishment for Crime." Deviant Behavior 2: Heider, F. (1958) The Psychology of Interpersonal Relations. New York: Wiley. Hertel, B.R. and M. Hughes (1987) "Religious Affiliation, Attendance, and Support for Pro-Family Issues." Social Forces 65: Hindelang, M.J. (1974) "Public Opinion Regarding Crime, Criminal Justice, and Related Topics." Journal of Research in Crime and Delinquency 21: Jelen, T.G. (1989) "Biblical IAteralism and Inerrancy: Does the Difference Make a Difference?" Sociological Analysis 49: Johnston, R. (1991) "American Evangelicalism: An Extended Family." In D. Dayton and R. Johnston (eds.), The Variety of American Evangelicalism, pp Downers Grove, IL: Inter-Varsity Press. Kellstedt, L. and C. Smidt (1991) "Measuring Fundamentalism: An Analysis of Different Operational Strategies." Journal for the Scientific Study of Religion 30: Kramer, J.H., R.L. Lubitz, and C.A. Kempinen (1989) "Sentencing Guidelines: A Quantitative Comparison of Sentencing Policy in Minnesota, Pennsylvania, and Washington." Justice Quarterly 6: Krisberg, B. (1991) "Are You Now or Have You Ever Been a Sociologist?" Journal of Criminal Law and Criminology 82: Lab, S.P. and J.T. Whitehead (1990) "From 'Nothing Works' to 'The Appropriate Works': The Latest Stop on the Search for the Secular Grail." Criminology 28: Langworthy, R.H. and J.T. Whitehead (1986) "Liberalism and Fear as Explanations of Punitiveness." Criminology 24: Lechner, F.J. (1990) "Fundamentalism Revisited." In T. Robbins and D. Anthony (eds.), In Gods We Trust: New Patterns of Religious Pluralism in America, pp New Brunswick, N J: Transaction. Luker, K. (1984) Abortion and the Politics of Motherhood. Los Angeles: University of California Press. Lupfer, M., P.L. Hopkinson, and P. Kelley (1988) "An Exploration of the Attributional Styles of Christian Fundamentalists and of Authoritarians." Journal for the Scientific Study of Religion 27: Lupfer, M. and K. Wald (1985) "An Exploration of Adults' Religious Orientations and Their Philosophies of Human Nature." Journal for the Scientific Study of Religion 24: Marsden, G. (1987) Reforming Fundamentalism: Fuller Seminary and the New Evangelicalism. Grand Rapids: Eerdmans. (1991) "Fundamentalism and American Evangelicalism." In D. Dayton and R. Johnston (eds.)~ The Variety of American Evangelicalism, pp Downers Grove, IL: Inter-Varsity Press. Martinson, R. (1974) "What Works? Questions and Answers about Prison Reform." The Public Interest 35: Maurer, A. (1982) "Religious Values and Child Abuse." In R. Hanson (ed.), Institutional Abuse of Children and Youth, pp New York: Haworth Press. Myers, M.A. (1988) "Social Background and Sentencing Behavior of Judges." Criminology 26: Nagel, LH. (1990) "Structuring Sentencing Discretion: The New Federal Sentencing Guidelines." Journal of Criminal Law and Criminology 80: Peek, C.W. and S. Brown (1980) "Sex Prejudice among White Protestants: Like or Unlike Ethnic Prejudice?" Social Forces 59:

25 GRASMICK, COCHRAN, BURSIK AND KIMPEL 313 Peek, C.W., G.D. Lowe, and L.S. Williams (1991) "Gender and God's Word: Another Look at Religious Fundamentalism and Sexism." Social Forces 69: Perlman, D. (1980) "Attributions in the Criminal Justice Process: Concepts and Empirical Illustrations." In P.D. Lipsitt and B.D. Sales (eds.), New D/rectwns in Psycholegal Research, pp New York: Van Nostrand Reinhold. Pillsbury, S.H. (1989) "Understanding Penal Reform: The Dynamics of Change." Journal of Criminal Law and Criminology 80: Radelet, M.L and M. Vandiver (1988) Capital Punishment in America: An Annotated Bibliography. New York: Garland. Rankin, J.H. (1979) "Changing Attitudes toward Capital Punishment." Social Forces 58: Sechrest, L.B., S.O. White, and E.D. Brown, eds. (1979) The Rehabilitation of Criminal Offenders: Problems and Prospects. Washington, DC: National Academy Press. Shaver, K.G. (1975) An Introduction to Attribution Processes. Cambridge, MA: Winthrop. - - (1985) The Attribution of Blame: Causality, Responsibility, and Blameworthiness. New York: Springer-Verlag. Sheley, J.F. (1985) America's 'Crime Problem" An Introduction to Criminology. Belmont, CA: Wadsworth. Smidt, C. (1983) "Born-Again Politics: The Political Behavior of Evangelical Christians in the South and Non-South." In T. Baker, R. Steed, and L. Moreland (eds.), Religion and Politics in the South: Mass and Elite Perspectives, pp New York: Praeger. (1988) "Evangelicals within Contemporary American Politics: Differentiating between Fundamentalist and Non-Fundamentalist Evangelicals." Western Political Quarterly 41: (1989) "Identifying Evangelical Respondents: An Analysis of Born-Again and Bible Questions Used across Different Surveys." In T. Jelen (ed"), Religion and Political Behavior in the United States, pp New York: Praegar. Smith, T.W. (1990) "Classifying Protestant Denominations." Review of Religious Research 31: Spilka, B., P. Shaver, and LA. Kirkpatrick (1985) "A General Attribution Theory for the Psychology of Religion." Journal for the Scientific Study of Religion 24:1-20. Stinchcombe, A.L., R. Adams, C.A. Helmet, K.L Scheppele, T.W. Smith, and D.C. Taylor (1980) Crime and Punishment---Changing Attitudes in America. San Francisco: Jossey-Bass. Taylor, D.C., K.L. Scheppele, and Arthur L. Stinchcombe (1979) "Salience of Crime and Support for Harsher Sanctions." Social Problems 26: Tedin, K.L. (1978) "Religious Preference and Pro/Anti Activism on the Equal Rights Amendment Issue." Pacific Sociological Review 21: Tyler, T.R. and R. Weber (1982) "Support for the Death Penalty: Instrumental Response to Crime or Symbolic Attitude?" Law and Society Review 17: Vold, G. (1958) Theoretical Criminology. New York: Oxford University Press. yon Hirsch, A., K.A. Knapp, and M. Tonry (1987) The Sentencing Commission and Its Guidelines. Boston: Northeastern University Press. Wald, K.D. (1992) Religion and Politics in the United States. 2nd ed. Washington, DC: Congressional Quarterly Press. Wald, K.D, D.E. Owen, and S.S. Hill Jr. (1988) "Churches as Political Communities." American Political Science Review 82: (1990) "Political Cohesion in Churches." Journal of Politics 52: Warr, M. (1989) "What Is the Perceived Seriousness of Crimes?" Criminology 27: (1991) "America's Perceptions of Crime and Punishment." In J.F. Sheley (ed.), Criminology: A Contemporary Handbook, pp Belmont, CA: Wadsworth. Wimberley, D. (1989) "Religion and Role-Identity: A Structural Symbolic Interactionist Conceptualization of Religiosity." Sociological Quarterly 30: Wood, M. and M. Hughes (1984) "The Moral Basis of Moral Reform: Status Discontent vs. Culture and Socialization as Explanations of Anti-Pornography Social Movement Adherence." American Sociological Review 49:86-99.