Coping, defending, and the relations between moral judgment and moral behavior in prostitutes and other female juvenile delinquents

Acknowledgement: This research was supported by Grants 410–87–1115 and 410–91–0510 from the Social Sciences and Humanities Research Council of Canada.
The assistance of Raymond Corrado, Laurene Wilson, Barbara Kirkland, Gary Poole, Kathy Denton, Sandra Vermeulen, Gillian Wark, and Patti Moen is gratefully acknowledged.

Past studies have found that people who adopt what are conventionally considered immoral life-styles—in particular, juvenile delinquents and adult criminals—obtain relatively low scores on the Moral Judgment Interview (MJI; see Blasi, 1980, and Hayes & Walker, 1986, for reviews). The conclusions of these studies, however, are limited on both descriptive and explanatory levels. On a descriptive level most studies have administered outdated versions of the MJI to males, and none has assessed moral judgments about the kinds of moral dilemma actually faced by delinquents. On an explanatory level little has been done to account for the processes that induce low-level moral judgment or that link it to criminal behavior (see Blasi, 1980). The overriding purposes of this study were to assess moral judgment in a sample of female delinquents with the current version of the MJI, to examine the relation between level of moral judgment on the MJI and level of moral judgment about a criminal behavior in which half of the delinquents acknowledged engaging, namely prostitution, and to explore the relations between coping, defending, moral judgment, and criminal behavior.

When Kohlberg (1984) and other theorists claim that low-stage moral judgments give rise to immoral behaviors, they refer to moral judgments about the immoral behaviors in question—in the case of delinquents, moral judgments about their delinquent behaviors. This is not, however, the relation that researchers have examined. Rather, they have assessed the relation between moral judgment on Kohlberg's test and criminal behavior (see Blasi, 1980). This practice has not seemed problematic to Kohlberg (1984) because he assumed that moral judgment is organized in structures of the whole and, therefore, that the level of moral judgment people display on the MJI corresponds to their level of moral judgment on other issues.

There is, however, reason to question whether moral judgment is as consistent as Kohlberg assumes. In an early study Kohlberg, Scharf, and Hickey (1972) found that prisoners in traditional prisons scored at lower stages when they made moral judgments about prison dilemmas than they did when they responded to the dilemmas on the MJI. Colby and Kohlberg (1987) dealt with this anomaly by asserting that the MJI assesses moral competence—the highest stage available to an person—and although people usually perform at their level of competence, they may invoke lower forms of moral judgment in extraordinary contexts, such as traditional prisons.

In a recent series of articles, Krebs and his colleagues (Carpendale & Krebs, 1992; Denton & Krebs, 1990; Krebs, Denton, Vermeulen, Carpendale, & Bush, 1991; Krebs, Vermeulen, Carpendale, & Denton, 1991; Krebs, Vermeulen, & Denton, 1991) have argued that Colby and Kohlberg (1987) did not go far enough in acknowledging the structural flexibility of moral judgment. Although Krebs and his colleagues have agreed that the MJI tends to assess moral competence, they have adduced evidence that people frequently perform below their level of competence. Jurkovic (1980) showed that delinquents may perform at different levels when they respond to different problems, and Krebs and Denton (in press) argued that regressive moral judgments are important in the determination of immoral behavior. Krebs and his colleagues have called for a shift in emphasis in research on moral development from the investigation of moral competence (as with the MJI) to the study of performance factors that corrupt moral judgment about real-life moral issues.

The first goal of this study was to compare the level of moral competence of two groups of female delinquents—a group that acknowledged engaging in prostitution and a group that did not—to that of nondelinquents of the same age on the current version of the MJI. The second goal was to assess the extent to which these groups performed at their level of competence (i.e., the extent to which they based their moral judgments on the same stages they displayed on the MJI) on a moral dilemma we believe is particularly salient to one of the groups—a moral dilemma about prostitution. The third goal of this study was to explore the effect of two performance variables we believe affect moral judgment and juvenile delinquency, namely, coping and defending.

According to Haan, Aerts, and Cooper (1985), people differ in the way they respond to threatening events: Some cope, and others react defensively. Haan et al. outlined 10 complementary coping and defending processes, such as intellectuality and intellectualizing (cognitive), empathy and projection (reflexive–intraceptive), concentration and denial (attention), and sublimation and displacement (affective–impulsive). Although no one has yet assessed coping and defending in juvenile delinquents with a validated test, there is good reason to expect delinquents to possess weaker coping skills and stronger defenses than other juveniles. Lund and Salary (1980) found that delinquents had a higher average true–false ratio than did control subjects on the Tennessee Self-Concept Scale, which indicates that the delinquents had weaker egos and poorer self-control. Portnoy (1986) found that high school students disposed toward delinquency tended to score higher on defensiveness than other students on the Defense Mechanism Inventory (Gleser & Ihilevich, 1969), and other studies have found that people at risk for delinquency score low on coping or high on defensiveness on Haan's (1977) Coping and Defending Test and its derivatives (see Frank & Quinlan, 1976; Jessor & Jessor, 1977; Robins, 1978; Rutter & Giller, 1984; West & Farrington, 1977).

Haan et al. (1985) expected “moral levels [to] be higher when [persons] cope and lower when they defend” (p. 128) on dilemmas that activate these ego processes. In this study we expected the dilemma about prostitution to activate the defenses of the low-coping and highly defensive prostitutes, inducing them to make lower level moral judgments on the prostitution dilemma than on the MJI.

Specifying the relation between the structures of moral judgment on the MJI and the structures of moral judgment people invoke on dilemmas they actually experience helps bridge the gap between moral judgment and moral behavior, but it leaves an important question unanswered, namely, in what ways to low-stage moral judgments about an immoral behavior facilitate or give rise to the behavior? Kohlberg and Candee (1984) suggested two routes. First, low-stage moral reasoning may give rise to immoral prescriptive judgments, such as, it is all right to steal if you need the money, which pave the way for immoral behaviors. Second, low-stage moral reasoning may give rise to the same moral choices as higher stage reasoning but supply less conceptual support for carrying them out. In the first case, people at low stages of moral development are viewed as morally ignorant—they do not know right from wrong. In the second case, people are viewed as morally weak—they know what is right but are not strong enough to behave accordingly. According to Kohlberg and Candee, a particularly importnat source of weakness in low-stage structures is their susceptibility to quasiobligations: “Excuses generated by each stage, to justify failure to act in terms of the moral obligations generated by that stage” (Kohlberg, 1984, p. 522).

The final goal of this study was to examine the relations between moral stage, moral choice, and delinquent behavior. Because the moral dilemmas on the MJI are largely irrelevant to the moral decisions delinquents make in their everyday lives, and because other investigators have found that delinquents do not differ from nondelinquents in their disapprobation of criminal acts (Gordon, 1975), we did not expect moral choices on the MJI to relate either to stage or to delinquency. We did, however, expect prostitutes to make less harsh judgments than other participants about the criminal behavior in which they had engaged.

To summarize, this study was designed to test the following hypotheses: (a) Female juvenile delinquents would score lower than nondelinquents on Colby and Kohlberg's (1987) test of moral judgment (the MJI) and on a moral dilemma about prostitution; (b) female juvenile delinquents would score lower on coping and higher on defending than nondelinquents; (c) prostitutes—especially low copers and high defenders—would score lower on moral maturity than the other groups on the prostitution dilemma, in relation to their level of competence on the MJI; and (d) prostitutes—especially low copers and high defenders—would make weaker prescriptive judgments about prostitution than other groups, but groups would not differ in their prescriptive judgments on the MJI. Finally, we examined the extent to which prescriptive judgments help bridge the gap between moral reasoning and moral behavior.

Sixty female adolescents participated in this study: 20 juvenile delinquents who acknowledged engaging in prostitution, 20 juvenile delinquents who said they had never engaged in prostitution, and 20 nondelinquent control subjects. The juvenile delinquents were recruited from two juvenile containment centers for repeated offenders in Burnaby, British Columbia, Canada, and were invited to participate in a study of their opinions on two moral issues—euthanasia and prostitution. Only 1 delinquent declined. After the 6 initial inmates at the first center were tested, each new admission was invited to participate in the study. After 1 month of testing at the first center, volunteers were solicited from the second center to complete the participant sample. (In all, 83% of the sample of delinquents came from the first center, and 17% came from the second center.) Members of the control group were recruited from a high school in a community adjacent to the first detention center. A teacher at the high school announced to all 10th and 11th grade students that a professor from a local university needed volunteers for a study on opinions about euthanasia and prostitution. Neither the teacher nor the students were told anything else about the study. Virtually all students volunteered. The first 20 available for testing constituted the sample.

Eighty percent of the participants in both the control group and the group of nonprostitute delinquents and 85% of the participants in the prostitute group were White. The mean ages of participants in the control, nonprostitute delinquent, and prostitute groups were, respectively, 16.20 (SD = 0.95), 15.70 (SD = 1.38), and 15.60 (SD = 1.23) years. The mean socioeconomic status (SES) of each group on a scale of low (1) to high (4), estimated according to Duncan's scale (Reiss, 1961), were, respectively, 2.85 (SD = 0.99), 2.35 (SD = 0.99), and 2.05 (SD = 1.05). The three groups did not differ significantly in age, but they did differ significantly in SES, F(2, 57) = 3.21, p < .05. Tukey tests revealed that the mean SES of the two delinquent groups did not differ significantly from each other, but both differed significantly from the control subjects. (Past research on juvenile delinquency has also found that delinquents come from lower SES than nondelinquents; see Braithwaite, 1979; Tolan, 1988; Wolfgang, Figlio, & Sellin, 1972.)

An examination of the criminal records of the two groups of delinquents revealed that all 8 delinquents with records of prostitution had been assigned to the group of prostitutes. The mean number of other offenses—the most frequent of which were breach of probation, theft under $1,000 (Canadian), breaking and entering, theft over $1,000 (Canadian), unlawfully at large, mischief, failing to appear, and possession of narcotics—was 4.3 (SD = 2.66) for the prostitutes and 4.0 (SD = 2.68) for the other delinquents. With the exception of prostitution, there were no notable differences between the two groups in the types of offenses with which they had been charged.

Each participant was interviewed individually in a private room at the high school or detention center. After establishing rapport and confirming in written and verbal form that participation in the study was voluntary, a trained female interviewer read each participant the first dilemma from From B of Colby and Kohlberg's (1987) MJI. This dilemma involves a decision about whether or not to violate the law in order to honor a sick woman's request for euthanasia. The interviewer then asked each participant a series of probing questions in accordance with the procedure outlined by Colby and Kohlberg (1987, pp. 192–194). After each participant had given a complete set of judgments to the first dilemma on the MJI, the interviewer gave her the second dilemma, which involves a conflict between reporting a law-breaker or letting him go unpunished, followed by the appropriate probes.

After the participants responded to the second dilemma on the MJI, they were read a structurally analogous dilemma about prostitution. The protagonist in the prostitution dilemma was a hypothetical 14-year-old girl named Jennifer who was thrown out of her house by her father and invited by a friend to engage in prostitution in order to make money. The interviewer probed participants' ideas about prostitution with questions such as:

Should Jennifer take her friend up on the offer? Why or why not? Would Jennifer be doing something wrong if she had sex for money? Why or why not? Under what circumstances is it right to accept money for sex? Is there anything wrong with prostitution? If so, what and why? If not, why not? Some people say that as long as a hooker and her client both agree to the arrangement, there is nothing wrong with prostitution. Other people say that it doesn't matter whether both parties agree to it; prostitution is still wrong. What do you think? Why? Is there any place for prostitution in society?

In a continuation of the dilemma, Jennifer is caught and pleads guilty to the offense of communicating for the purpose of prostitution. The participants were asked such questions as, “What should the judge do with Jennifer?” and “What circumstances should the judge consider in making his decision? Why?” In all, 10 probe questions followed the prostitution dilemma. After the participants responded to all prescriptive (should) questions, they were asked two questions designed to assess their receptiveness to prostitution. The questions were: “What would you do?” and “Would you ever have sex for money?” Finally, participants were asked if they had ever had sex for money.

Delinquents who acknowledged having sex for money were assigned to the prostitution group; those who said they had never had sex for money were assigned to the nonprostitute group. The average time per interview was approximately 1 hr. The interviewers probed participants' answers extensively but politely, asking them to explain, clarify, and justify their decisions and opinions. All interviews were audiorecorded and transcribed for scoring.

After the moral interviews were conducted, participants were given two subtests from Joffe and Naditch's (1977) paper-and-pencil Coping and Defending Processes Test. The Coping and Defending subtests consist of 65 true–false questions selected from the California Personality Inventory and the Minnesota Multiphasic Personality Inventory on the basis of their ability to predict clinical ratings of the 10 complementary coping and defending mechanisms identified by Haan (1963, 1977) and Kroeber (1963) in their models of ego functioning. An example of a coping process measured by this test is logical analysis (“Let's start at the beginning and figure out what happened”); the complementary defending process is rationalization (“I was trying at first, but one thing after another happened”; Hann et al., 1985, p. 129). Joffe and Naditch (1977) reported that test–retest reliability coefficients for women ranged from “.46 for isolation to .81 for repression and intellectualization. The median reliability was .70” (p. 288). The test provides two summary scores, one for coping and the other for defending. Joffe and Naditch reported that the correlation between scale scores and clinical ratings derived from 14- to 16-hr interviews produced validity coefficients of .34 for coping and .38 for defending. In a validity study by Joffe and Bast (1978), a sample of well-functioning and poorly functioning blind men obtained significantly different scores on the two scales, and “results … regarding the construct validity of the ego scales were generally supportive” (p. 549).

Stage and moral maturity on the Moral Judgment Interview

The complex, 17-step procedure for scoring the MJI is outlined in detail in a two-volume, 1,200-page scoring manual (Colby & Kohlberg, 1987). Scoring involves identifying prescriptive judgments made by interviewees, classifying them by issue, norm, and element, and finding matching criterion judgments in the scoring manual that specify subjects' stage structure. For example, the chosen issue for the criterion judgment, “[the doctor should give the woman the drug even though it is against the law] because it's her life and she can do whatever she wants with it” is life, because the subject advocates breaking the law to save a life; the norm also is life, because life is being valued; the element is having a right, and the stage is 2 (Colby & Kohlberg, 1987, p. 282).

Stage and moral maturity on the prostitution dilemma

As in the procedure used by Krebs et al. (1989), Krebs, Denton et al. (1991), Walker (1988), and others, the prostitution dilemma was scored by identifying prescriptive judgments made by participants during the interview, classifying them by element, and searching for criterion judgments in the MJI scoring manual that matched their stage structure. For example, the judgment that becoming a prostitute is understandable if the person is absolutely desperate for money and cannot find any work anywhere invokes the element of blaming–approving, matches Criterion Judgment 8 (Colby & Kohlberg, 1987, p. 6), and was scored as Stage 2/3. Table 1 contains a sample of judgments about prostitution matched with criterion judgments from Colby and Kohlberg's (1987) manual.

After all prescriptive judgments were classified, they were summed according to the formula supplied by Colby and Kohlberg (1987, p. 186) to obtain global stage scores on a 9-point scale (i.e., Stage 1, Stage 1/2, Stage 2, Stage 2/3, etc.). In addition, a continuous measure of moral maturity, a weighted average score, was computed. This measure is derived by weighing and summing all issue stage scores, dividing by the sum of the weights, and multiplying by 100. A weighted average score of 100 corresponds to Stage 1, a score of 200 to Stage 2, a score of 300 to Stage 3, and so forth.

Prescriptive judgments about law, punishment, and prostitution

Deontic choices, or prescriptive judgments about (a) law (from Dilemma IV of the MJI), (b) punishment (from Dilemma IV′ of the MJI), and (c) prostitution (from the prostitution dilemma), were scored on a 4-point scale (from unequivocally pro to unequivocally con) and summed to obtain aggregate scores. An example of an unequivocally pro-law prescriptive judgment is, “He should obey the law—no matter what.” An example of a moderately con-prostitution judgment is, “She shouldn't do it unless she is desperate.” Each of the aggregate scores for law, punishment, and prostitution was based on answers to six questions (coefficient αs = .72, .69, and .82, respectively). In addition, an aggregate score for receptiveness to prostitution was obtained by summing scores for the two receptiveness questions. (The two receptiveness questions differ from the six prescriptive questions in that the former ask, “Would you …?”, whereas the latter ask, “Should she…?”) The coefficient alpha was .96.

One quarter of the MJI results were randomly selected and scored for reliability. Each of the three dilemmas responded to by each participant (two MJI dilemmas and one prostitution dilemma) was scored independently. Scorers were unaware of all information about the participant, her experimental condition, and her score on other dilemmas. The intraclass correlations between the two raters' moral maturity scores were .68 for the MJI dilemma and .65 for the prostitution dilemma, with 80% exact agreement on both on a 9-point scale. There was perfect agreement in scoring for receptiveness to prostitution and only two 1-point discrepancies in ratings of prescriptive judgments.

We present the results in five sections. In the first four sections, we discuss group differences in (a) receptiveness to prostitution, (b) coping and defending, (c) moral maturity, and (d) prescriptive judgments. In the final section we discuss the relation between moral maturity and prescriptive judgments.

Although it is possible that some of the juvenile delinquents lied about whether or not they had engaged in prostitution, there is good reason to believe that the groups differed as expected. First, the interviewer made every effort to establish the type of rapport conducive to self-disclosure. Second, as we indicate, all 8 delinquents with records of prostitution were correctly assigned to the prostitute group. Finally, Tukey tests revealed that the prostitutes obtained significantly higher mean scores on receptiveness to prostitution (M= 1.47, SD = 0.92) than did the other delinquents (M = 0.05, SD = 0.22) and the control subjects (M = 0.05, SD = 0.22), F(2, 57) = 39.51, p < .0001.

It is notable that 7 of the acknowledged prostitutes scored low on the receptiveness questions, which suggests that although they had engaged in prostitution, they would not do it again. Within the prostitution group, the correlation between receptiveness to prostitution and record of convictions for soliciting was not statistically significant, r(20) = .09. There was, however, a significant positive correlation between receptiveness to prostitution and prescriptive judgments favoring prostitution on the hypothetical dilemma, r(20) = .44, p < .05, for the prostitutes. (There was virtually no variance in the other groups.) The scores of subgroups of (a) convicted (n = 8) and nonconvicted (n = 12) and (b) receptive (n = 13) and nonreceptive (n = 7) prostitutes were compared on all dependent variables. Although no significant differences emerged, the limited power of these analyses must be noted.

A comparison between the mean coping scores of the three groups produced a highly significant result, F(2, 53) = 8.00, p < .01. Tukey tests revealed that the scores of the prostitutes (M = 47.4, SD = 5.4) and other delinquents (M = 43.1, SD = 11.6) were not significantly different, but both were significantly different from the coping scores of the control subjects (M = 54.3, SD = 8.1). The pattern of results for defending was similar to that for coping. The defending scores of the prostitutes (M = 51.7, SD = 11.2) and other delinquents (M = 52.6, SD = 9.1) were not significantly different, but both were significantly higher than the scores of the control subjects (M = 35.4, SD = 10.4), F(2, 53) = 16.82, p < .0001. Coping and defending scores were negatively correlated, r(56) = −.36, p < .01.

A Group (prostitutes, delinquents, and control subjects) × Dilemma (MJI and prostitution) analysis of variance (ANOVA), with repeated measures on dilemma and moral maturity as the dependent variable, produced a significant main effect for group, F(2, 57) = 6.34, p < .01. Tukey tests revealed that the mean MJI scores of the prostitute (M = 239, SD = 25) and nonprostitute (M = 230, SD = 39) delinquents were not significantly different, but both were significantly lower than the MJI score of the control subjects (M = 265, SD = 31). Cast in terms of stages, the modal stage of the control subjects was Stage 3, and their minor Stage was 2; however, the modal stage of the delinquents was Stage 2, with a minor Stage 3. (Analyses of covariance with SES as a covariate also were conducted. They did not alter the significance of the other main effects in any case. )

Participants tended to score lower on the prostitution dilemma (M = 239, SD = 44) than on the MJI (M = 250, SD = 39), F(1, 57) = 3.33, p < .07, but the expected Group × Dilemma interaction was not significant, F(2, 57) = 0.06.

To test the hypothesis that low-coping prostitutes would perform at a lower level of moral maturity than other groups on the prostitution dilemma but not on the MJI, we divided participants in each group at the median into high and low copers and added coping to the ANOVA (Group × Coping × Dilemma), F(2, 50) = 2.84, p < .07. Substituting defending for coping produced a similar but weaker result, F(2, 50) = 1.24. Although not statistically significant, the pattern of means was as expected, except the interaction between coping and dilemma appeared to apply to both groups of delinquents, not just to the prostitutes.

To add power to the analysis, we combined the two delinquent groups on a post hoc basis (see Table 2). The three-way interaction in the associated 2 × 2 × 2 (Group × Coping × Dilemma) ANOVA was significant, F(1, 52) = 8.54, p < .01. Tests for simple effects were conducted to decompose the three-way interaction. To control for familywise error, the significance level was set conservatively at .004 (.05 × 12 comparisons). As expected, the effect for dilemma was significant for only one of the four groups, the low-coping delinquents, who scored lower on the prostitution dilemma than on the MJI, F(1, 52) = 14.74, p < .001 (for the three other groups, Fs < 2.22). Also as expected, the prostitution scores of the low-coping delinquents were significantly lower than the prostitution scores of the low-coping control subjects, F(1, 52) = 15.00, p < .001, but the low-coping delinquents did not differ from the low-coping control subjects on the MJI, F(1, 52) = 1.92, and the high-coping delinquents did not differ from the high-coping control subjects on either dilemma, Fs(1, 52) < 3.30. Finally, the low-coping delinquents scored significantly lower than the high-coping delinquents on the prostitution dilemma, F(1, 52) = 9.34, p < .001, but not on the MJI (F < 1), but the differences between the scores of high- and low-coping control subjects were not significant on either dilemma (Fs < 1.44). For the three-way interaction in a comparable analysis of defending scores, F(1, 52) = 2.58, p = .11.

Group × Coping and Group × Defending ANOVAs on prescriptive judgments about law, punishment, and prostitution revealed only one significant effect, a group difference in prescriptive judgments about prostitution, Fs(2, 50) = 4.26 and 3.95, p < .05. Although all three groups disapproved of prostitution, Tukey tests revealed that the prostitutes (M = 1.83, SD = 0.82) showed significantly less disapprobation than the other juvenile delinquents (M = 2.56, SD = 0.57; the pro–con midpoint was 1.5). The difference between the prescriptive judgments of the prostitutes and the controls (M = 2.13, SD = 0.79) was in the expected direction but not statistically significant.

Neither the correlations between moral maturity on the MJI and prescriptive judgments about (a) law and (b) punishment nor the correlations between moral maturity on the prostitution dilemma and prescriptive judgments about prostitution were statistically significant: rs(59) = −.13, .15, and −.03, respectively.

Although a spate of studies has found that male juvenile delinquents score lower than same-age control subjects on earlier versions of the MJI and that female juvenile delinquents score lower than nondelinquents on other tests of moral development, this study is the first to confirm that female delinquents score lower than same-age control subjects on the current version of the MJI. As in past research on male delinquents, we found that the modal stage for female delinquents is Stage 2/3, as compared with Stage 3(2) for same-age control subjects. This difference is psychologically significant, consitituting a quarter of a stage. In most contexts, the difference between the self-focused, instrumental orientation of Stage 2 and the other-oriented, role-upholding orientation of Stage 3 is equivalent to the difference between a moral and an immoral point of view (see Colby & Kohlberg, 1987, pp. 26–28).

Past investigators have interpreted such differences as evidence that juvenile delinquents possess a lower level of moral competence than do nondelinquents; however, such differences also may stem from performance factors. For example, like the delinquents in this study, most delinquents are tested in institutions, which, as suggested by Kohlberg et al. (1972), may constrain moral judgment. Another factor that may affect performance is the type of dilemma. Although participants tended to make lower level moral judgments on the prostitution dilemma than on the MJI dilemmas, which suggests that level of moral judgment is determined by an interaction between person-related (level of moral competence) and situational (type of moral dilemma) variables (see Krebs, Vermeulen et al., 1991), and although this difference tended to be greatest for the prostitutes, the expected Group × Dilemma interaction was not statistically significant.

A novel finding of this study is that juvenile delinquents scored significantly lower on coping and significantly higher on defending than did matched control subjects. Indeed, group differences in coping and defending were much stronger than group differences in moral maturity. This finding is consistent with Haan et al.'s (1985) contention that “whether people solve problems by defensiveness or coping—either habitually or just at a particular moment—affects the morality of their action” (p. 169). People who scored high on defensiveness displayed life-styles of delinquent behavior. Clearly, the role of coping and defending in the etiology and maintenance of juvenile delinquency deserves more attention from researchers than it has received.

Our hypothesis about the relations between coping and defending and moral maturity was not supported in its original form: The low-coping and highly defensive prostitutes did not perform at a significantly lower level of moral maturity on the prostitution dilemma than did the other groups. In retrospect, we believe this hypothesis was misguided. We did not realize that most of the delinquents who disavowed prostitution (virtually all of whom had been living on the streets before they were incarcerated) had experienced the temptation to engage in the act. We came to believe that the prostitution dilemma was as salient and threatening to the delinquents who denied having sex for money as it was to those who acknowledged engaging in the act. When we combined the two groups of delinquents on a post hoc basis, we found the expected three-way interaction with coping. The general idea that low-coping and highly defensive people perform below their level of moral competence when they respond to dilemmas they find threatening seems worthy of further examination.

Our findings add to growing evidence that the structure of moral judgment, as assessed by the MJI, is related to moral behavior (see Blasi, 1980). The level of analysis in this study was general: A broad-based aspect of personality—stage of moral development—was related to a broad-based measure of moral behavior—a life-style of juvenile delinquency. As pointed out by Ajzen and Fishbein (1980), it is at such matching levels of generality that positive relations are expected between such intrapsychic variables as attitudes and overt manifestations of behavior.

How do low levels of moral judgment predispose people to criminal behavior? The first route suggested by Kohlberg and Candee (1984)—that a low level of moral judgment gives rise to immoral prescriptive judgments—received little support in this study. None of the prescriptive judgments participants made on any of the moral dilemmas covaried significantly with stage, and groups did not differ in their prescriptive judgments on the MJI. Although the prostitutes made weaker prescriptive judgments about prostitution than the other delinquents, it is unclear whether these judgments paved the way for their behavior, as Kohlberg and Candee assumed in their model. It also is possible that the prostitutes adjusted their attitudes after they engaged in the behavior, as suggested by cognitive-consistency theorsts (see Aronson, 1988).

Our findings are more consistent with the second route suggested by Kohlberg and Candee (1984)—the higher the stage structure, the more conceptual support it supplies for moral choices. Although most people agree that it is wrong to break the law and to engage in prostitution, people at different stages of moral development differ in the reasons why they believe it is wrong. Kohlberg (1984) suggested that the types of justification that stem from high stages of moral development are stronger than those that stem from lower stages because they are less susceptible to excuses and evasions of responsibility. Some of the judgments made by delinquents seem consistent with this characterization. Consider, for example, the following: “I haven't tried to abide by the law because it doesn't help me. The law here in Canada is supposed to be ‘innocent until proven guilty’? Well here I am—guilty until proven innocent—why should I obey it when I won't be believed when I'm innocent anyways,” and “Deep down inside, I'm not the kind of person to break the law. I got involved in the wrong crowd, and I got involved in the street and drugs, and I was an alcoholic, and I went crazy.”

In addition, the low-stage judgments of the delinquents seemed weaker than the higher stage judgments of the control subjects in at least three other related ways. First, in the delinquents' orientation to physical consequences, they tended to believe it is permissible to violate the law as long as one does not get caught. This sentiment was expressed clearly by a delinquent responding to a question about whether a doctor ought to perform euthanasia: “I guess I would do it, but I wouldn't tell anybody and just make it look like a suicide… by covering my ass, I'd make it look like a suicide.” Second, the constraints against immoral behavior in the delinquents' Stage 1/2 judgments were individualistic and egocentric—“I might get punished”—whereas higher stage judgments of the nondelinquents, such as, “laws [are] made to protect citizens and to keep law and order and peace in the community,” reflected a more broad-based social perspective and concern for others. In this sense, higher stage judgments contain more reasons for upholding the law. Finally, the higher stage judgments of the control subjects contained a more internalized orientation toward the law. From a Stage 2 perspective, laws are imposed on people externally, whereas, as put by one of our subjects, from a Stage 3/4 perspective, “most of the time the laws are the will of the majority, and most people would agree for them to be made in the first place… They went through a lot of people's supervision before they were decided on, so there must be some wisdom behind them.”

To summarize, the findings of this study support the following conclusions. First, female juvenile delinquents score at lower levels of moral maturity and coping and at a higher level of defensiveness than do their same-age nondelinquent counterparts. Second, there is some indication that intrapsychic processes, such as coping and defending, may constrain level of moral judgment on personally relevant and threatening dilemmas—a possibility that seems worthy of further study. Third, although abstract prescriptive judgments about upholding the law do not seem to relate to criminal behavior, prescriptive judgments about the particular acts in which people have engaged may. Finally, the well-documented relation between moral judgment and moral behavior appears to stem less from what people believe is right and wrong (i.e., from their prescriptive judgments) and more from why they believe an act is right or wrong (i.e., from the structure of their moral reasoning).

1  When we first started testing, we intended to give all participants the coping and defending test in a second session. After 4 (2 prostitutes and 2 non-prostitutes) of the first 8 delinquents we tested were released from custody before completing the coping and defending questionnaire, we decided to give the test after the Moral Judgment Interview and complete all testing in one session. The sample size for the coping and defending measures was, therefore, reduced to 56.

2  The covariate results for socioeconomic status were as follows: (a) moral maturity, F(1, 56) = 1.50; (b) coping, F(1, 52) = 1.55; (c) defending, F(1, 52) = 2.41; (d) prescriptive judgments about law, F(1, 56) = 0.95; (e) prescriptive judgments about punishments, F(1, 56) = 1.49; and (f) prescriptive judgments about prostitution, F(1, 56) = 0.06.

Ajzen, I., & Fishbein, M. (1980). Understanding attitudes and predicting social behavior. Englewood Cliffs, NJ: Prentice-Hall.

Aronson, E. (1988). The social animal (5th ed.). New York: Freeman.

Blasi, A. (1980). Bridging moral cognition and moral action: A critical review of the literature. Psychological Bulletin, 88, 1–45.

Braithwaite, J. (1979). Inequality, crime and public policy. London: Routledge & Kegan Paul.

Carpendale, J., & Krebs, D. (1992). Situational variation in moral judgment: In a stage or on a stage?Journal of Youth and Adolescence, 21, 203–224.

Colby, A., & Kohlberg, L. (Eds.). (1987). The measurement of moral judgment (Vols. 1–2). New York: Cambridge University Press.

Denton, K., & Krebs, D. (1990). From the scene of the crime: The effect of alcohol and social context on moral judgment. Journal of Personality and Social Psychology, 59, 242–248.

Frank, S., & Quinlan, D. (1976). Ego development and female delinquency: A cognitive–developmental approach. Journal of Abnormal Psychology, 85, 505–510.

Gleser, G., & Ihilevich, D. (1969). An objective instrument for measuring defense mechanisms. Journal of Consulting and Clinical Psychology, 33, 51–60.

Gordon, R. (1975). Crime and cognition: An evolutionary perspective. São Paulo, Brazil: International Center for Biological and Medico-Forensic Criminology.

Haan, N. (1963). Proposed model for ego functioning: Coping and defense mechanisms in relationship to IQ change. Psychological Monographs, 77, 1–27.

Haan, N. . (Ed.) (1977). Coping and defending: Processes of self-environment organization. New York: Academic Press.

Haan, N., Aerts, E., & Cooper, B. (1985). On moral grounds: The search for practical morality. New York: New York University Press.

Hayes, S., & Walker, W. (1986). Intellectual and moral development in offenders: A review. Australian and New Zealand Journal of Criminology, 19, 53–64.

Jessor, R., & Jessor, S. (1977). Problem behavior and psychosocial development: A longitudinal study of youth. New York: Academic Press.

Joffe, P., & Bast, B. (1978). Coping and defense in relation to accommodation among a sample of blind men. Journal of Nervous and Mental Disease, 166, 537–552.

Joffe, P., & Naditch, M. (1977). Paper and pencil measures of coping and defending processes. In N.Haan (Ed.), Coping and defending: Processes of self-environment organization (pp. 280–297). New York: Academic Press.

Jurkovic, G. (1980). The juvenile delinquent as a moral philosopher: A structural–developmental perspective. Psychological Bulletin, 88, 709–727.

Kohlberg, L. . (Ed.) (1984). Essays on moral development: Vol 1. The philosophy of moral development. New York: Harper & Row.

Kohlberg, L., & Candee, D. (1984). The relationship of moral judgment to moral action. In L.Kohlberg (Ed.), Essays on moral development: Vol. 2. The psychology of moral development (pp. 498–581). New York: Harper & Row.

Kohlberg, L., Scharf, P., & Hickey, J. (1972). The justice structure of the prison: A theory and intervention. Prison Journal, 51, 3–14.

Krebs, D., & Denton, K. (in press). On the relations between the structure of moral judgmental and moral behavior. In D.Garz, F.Oser, & W.Althof (Eds.), The context of moralityFrankfurt, FRG: M. Suhricamp.

Krebs, D., Denton, K., Carpendale, J., Vermeulen, S., Bartek, S., & Bush, A. (1989). The many faces of moral judgment. In M.Luszez & T.Nettelbeck (Eds.), Psychological development: Perspectives across the life-span (pp. 97–105). Amsterdam: Elsevier Science.

Krebs, D., Denton, K., Vermeulen, S., Carpendale, J., & Bush, A. (1991). The structural flexibility of moral judgment. Journal of Personality and Social Psychology, 61, 1012–1023.

Krebs, D., Vermeulen, S., Carpendale, J., & Denton, K. (1991). Structural and situational influences of moral judgment: The interaction between stage and dilemma. In W.Kurtines & J.Gewirtz (Eds.), Handbook of moral behavior and development: Theory, research, and application (pp. 139–169). Hillsdale, NJ: Erlbaum.

Krebs, D., Vermeulen, S., & Denton, K. (1991). Competence and performance in moral judgment: From the ideal to the real. Moral Education Forum, 16, 7–22.

Kroeber, T. (1963). The coping functions of the ego mechanisms. In R.White (Ed.), The study of lives: Essays on personality in honor of Henry A. Murray (pp. 178–198). New York: Atherton Press.

Lund, N., & Salary, H. (1980). Measured self-concept in adjudicated juvenile offenders. Adolescence, 15, 65–74.

Portnoy, R. (1986). Ego defenses as a predictor of deterrence (Doctoral dissertation, University of Nebraska). Dissertation Abstracts International, 47, 2180B–2181B.

Reiss, A. (1961). Occupations and social status. New York: Free Press of Glencoe.

Robins, L. (1978). Sturdy childhood predictors of adult antisocial behavior: Replications from longitudinal studies. Psychological Medicine, 8, 611–622.

Rutter, M., & Giller, H. (1984). Juvenile delinquency: Trends and perspectives. New York: Guilford Press.

Tolan, P. (1988). Socioeconomic, family and social stress correlates of adolescent antisocial and delinquent behavior. Journal of Abnormal Child Psychology, 16, 317–331.

Walker, L. (1988). The development of moral reasoning. Annals of Child Development, 5, 33–78.

West, D., & Farrington, D. (1977). The delinquent way of life. London: Heinemann Educational.

Wolfgang, M., Figlio, R., & Sellin, T. (1972). Delinquency in a birth cohort. Chicago: University of Chicago Press.

Submitted: December 17, 1990 Revised: April 21, 1992 Accepted: May 7, 1992