The Journal of Social Psychology, 1985, 125(3), 329-333



School of Sociology, University of New South Wales, Australia


Little support exists for the claim by Adorno, Frenkel-Brunswik, Levinson, and Sanford (1950) that punitiveness is associated with authoritarianism. A balanced 12-item Likert scale of punitiveness toward criminals was composed and tested on a random postal sample of 128 Australians. It showed a reliability (alpha) of .78 and was found to predict strict rather than permissive views on child rearing. It did not predict Machiavellianism, psychopathy, satisfaction with personal relationships, or social desirability responding. The findings indicate both convergent and discriminant validity and tend to undermine the view of punitiveness as psychopathological. On a second random doorstep sample of 113 Australians, the scale had a reliability of .83 and showed a highly significant tendency to be correlated with length of sentence thought appropriate for particular crimes. Again, it did not correlate with a measure of psychopathology (trait anxiety). Mean scores in both samples indicated more punitive than impunitive responses. Punitiveness, therefore, was normal rather than psychopathological, but punitiveness itself may be multidimensional.

PUNITIVENESS WAS SEEN by Adorno et al. (1950) as a central characteristic of the authoritarian personality. Punitiveness, thus, was thought to spring from a deep-lying hostility in personality that in turn originated as a response to conflicting feelings about the father. This hypothesis, however, has not been subsequently supported. An early study by Roberts and Jessor (1958) rejected the identification of punitiveness and authoritarianism, and a recent summary of the literature on the problem by Garcia and Griffitt (1978) also showed that authoritarians were not in general punitive (particularly toward criminals). Garcia and Griffitt did, nonetheless, construct a mock-jury study of their own in an attempt to show that authoritarians are more punitive in certain circumstances. This attempt has since been shown by Ray (1983, p. 179) to be specious. Recent results from mock-jury studies also have been inconclusive. Although Werner, Kagehiro, and Stube (1982) found that student jurors were more likely to convict criminals when the jurors themselves were (in student terms) highly authoritarian, Lamberth, Krieger, and Shay (1982) found that authoritarian jurors were more flexible and that they were not particularly likely to favor a guilty verdict. Other mock-jury studies of authoritarianism (e.g., Berg & Vidmar, 1975) might also be viewed as concerned with conservatism rather than with authoritarianism per se (see also Ray, 1976). The present paper was based on the working hypothesis that punitiveness toward criminals can be used as an index of a general tendency to be punitive.

Study 1

Brief mention of the scale used for present purposes has already been made elsewhere (Ray & Ray, 1982). The scale was administered as part of a questionnaire answered by a random postal sample of 128 Australians. Sampling details have been given in the earlier publication.

Also included in the questionnaire are other scales enabling the likely correlates of punitiveness to be explored. These were the MMPI Pd (Psychopathy) scale, the Christie and Geis (1970) Machiavellianism Scale, a short form of the Crowne and Marlowe (1964) Social Desirability Scale, and two ad hoc scales designed to measure permissiveness and strictness toward children and satisfaction with one's own personal relationships. The attitude toward children scale included items such as "Spare the rod and spoil the child" and "Children need plenty of discipline." The relationships scale included items such as "I generally feel comfortable with other people" and "People often seem to want me as a friend." The Punitiveness Scale included items such as "Hanging should be brought back as a penalty for serious crime" and "The life sentence for serious crime should mean life." Its items have been listed in full elsewhere (Ray, 1982). It was hypothesized that punitiveness toward criminals would tend to be correlated with strictness toward children and that, if Adorno et al. (1950) were correct, punitive people should be psychopathic, Machiavellian, dishonest, and have poor personal relationships.

In fact, the only significant correlation between the Punitiveness Scale and the other scales was one of .30 (p > .01) with strictness toward children. This, then, was some concurrent validation for the scale, but punitiveness did not generalize strongly. It was also found that punitive people were slightly more likely to come from larger families (r = .23; p < .05) and that they tended to go to church a little less often (r = -.22; p < .05). There was also a just significant tendency (r = -.20; p < .05) for punitive people to be less psychopathic than impunitive people. The reliabilities (alpha) of the scales were found to be the following: Punitiveness, .78; personal relationships, .76; attitude toward children, .60; and Psychopathy, .85. The mean of the Punitiveness Scale was 42.62 (SD = 6.67), which was virtually a full standard deviation above the scale midpoint of 36. It was thus shown that randomly sampled people tended to be punitive rather than the reverse.

Study 2

The same scale was included in a questionnaire administered to a random doorstep sample of 113 residents of Sydney, Australia. Other results from this survey have been reported elsewhere (Ray, 1982), where fuller methodological details may also be found. As further validation for the new scale, it was proposed to relate punitiveness to the length of prison sentence that respondents saw as appropriate for various sorts of crime (cf. Garcia & Griffitt, 1978; Werner, Kagehiro, & Stube, 1982). We described 21 types of crime (see Ray, 1982, for details) and asked subjects what sentence (in years) they would like to see given to the perpetrator of each. Although the psychopathology scales of the previous study had not shown any positive correlations with the new scale, the Spielberger, Gorsuch, and Lushene (1970) trait-anxiety scale was included in the present survey.

The reliability for the new scale (alpha) was found to be .83, quite high for a short balanced scale. Its mean and standard deviation were, respectively, 39.49 and 7.78. Five response options were allowed for each of the 12 items. Again, the mean score indicated a tendency toward pro-punitive rather than anti-punitive responses, though not as strongly as in the previous study. The presence of rural respondents in the sample of Study 1 may account for the slight difference.

The length of sentence questions were also found to form a reliable scale (alpha = .77), correlating .30 with the Punitiveness Scale. Although this was significant (p < .01), it was low in absolute magnitude and again suggests that punitiveness might be only weakly generalizable. As a validity coefficient, however, it is comparable to those reported for many other scales (e.g., Jackson, 1967).

The correlation with the Spielberger scale was .13 and was non-significant, indicating that punitive people were not particularly anxious. There were no significant correlations between the new scale and age, sex, occupation, or education. Having actually known a criminal was also found to be unrelated to punitiveness.


We have once again found (cf. Brown, 1965) multidimensionality in an area that Adorno et al. (1950) believed to be unidimensional. Not only was punitiveness not related to authoritarianism, but even different types of punitiveness were not highly related. For example, those who wanted to flog criminals (i.e., high scorers on the new scale) often did not want to give criminals long prison terms (Study 2) and often did not believe in beating children (Study I).

Most of the items in the Punitiveness Scale would not seem out of place in a conservatism scale. Many issues of crime and punishment are highly political. Similarly, the items of the attitude toward children scale polarize around old-fashioned authoritarianism on one hand and liberal, progressive stances on the other. Again, the items could be taken as measuring conservatism. The fact that the two scales showed only a low correlation might suggest that conservatism, too, is multidimensional. People who are conservative by one definition of the term are often not conservative by another. The results once again point us toward caution about generalizations. It may also be noted that to Adorno et al. (1950) the notions of authoritarianism toward children and punitiveness could hardly be more strongly associated. They were the core concepts of their theory. Yet, far from being intimately associated, the two attributes were found to share only 9% of common variance. The concept of authoritarianism must also be broken into its component parts if seriously fallacious assumptions are to be avoided. The present results imply that findings about one of the supposed aspects of authoritarianism provide little information about any of the others.


Adorno, T. W., Frenkel-Brunswik, E., Levinson, D. J., & Sanford, R. N. (1950). The authoritarian personality. New York: Harper.

Berg, K. S., & Vidmar, N. (1975). Authoritarianism and recall of evidence about criminal behavior. Journal of Research in Personality, 9, 147-157.

Brown, R. (1965). Social psychology. New York: Free Press.

Christie, R., & Geis, F. L. (1970). Studies in Machiavellianism. New York: Academic.

Crowne, D. P., & Marlowe, D. (1964). The approval motive. New York: Wiley.

Garcia, L. T., & Griffitt, W. (1978). Authoritarianism-situation interactions in the determination of punitiveness: Engaging authoritarian ideology. Journal of Research in Personality, 12, 469-478.

Jackson, D. N. (1967). Personality research form manual. New York: Research Psychologists Press.

Lamberth, J., Krieger, E., & Shay, S. (1982). Juror decision making: A case of attitude change mediated by authoritarianism. Journal of Research in Personality, 16, 419-434.

Ray, J.J. (1976) Do authoritarians hold authoritarian attitudes? Human Relations, 29, 307-325.

Ray, J.J. (1982) Prison sentences and public opinion. Australian Quarterly 54, 435-443.

Ray, J.J. (1983) The workers are not authoritarian: Attitude and personality data from six countries. Sociology & Social Research, 67 (2), 166-189.

Ray, J.J. & Ray, J.A.B. (1982) Some apparent advantages of sub-clinical psychopathy. Journal of Social Psychology, 117, 135-142.

Roberts, A. N., & Jessor, R. (1958). Authoritarianism, punitiveness and perceived social status. Journal of Abnormal Social Psychology, 56, 311-314.

Spielberger, C. D., Gorsuch, R. L., & Lushene, R. E. (1970). Manual for the State/trait anxiety inventory. Palo Alto, CA: Consulting Psychologists Press.

Werner, C. M., Kagehiro, D. K., & Strube, M. J. (1982). Conviction proneness and the authoritarian juror: Inability to disregard information or attitudinal bias? Journal of Applied Psychology, 67, 629-636.

Go to Index page for this site

Go to John Ray's "Tongue Tied" blog (Backup here or here)
Go to John Ray's "Dissecting Leftism" blog (Backup here or here)
Go to John Ray's "Australian Politics" blog (Backup here or here)
Go to John Ray's "Gun Watch" blog (Backup here or here)
Go to John Ray's "Education Watch" blog (Backup here or here)
Go to John Ray's "Socialized Medicine" blog (Backup here or here)
Go to John Ray's "Political Correctness Watch" blog (Backup here or here)
Go to John Ray's "Greenie Watch" blog (Backup here or here)
Go to John Ray's "Food & Health Skeptic" blog (Backup here)
Go to John Ray's "Leftists as Elitists" blog (Not now regularly updated -- Backup here)
Go to John Ray's "Marx & Engels in their own words" blog (Not now regularly updated -- Backup here)
Go to John Ray's "A scripture blog" (Not now regularly updated -- Backup here)
Go to John Ray's recipe blog (Not now regularly updated -- Backup here or here)

Go to John Ray's Main academic menu
Go to Menu of recent writings
Go to John Ray's basic home page
Go to John Ray's pictorial Home Page (Backup here)
Go to Selected pictures from John Ray's blogs (Backup here)
Go to Another picture page (Best with broadband)