The Journal of Social Psychology, 1980, 111, 9-17.


University of New South Wales, Australia



Since the original work on authoritarianism was done in California, it was desired to compare California today both with itself 30 years ago and with other English-speaking cultures. In so doing it was also desired to incorporate the methodological improvements that have come from 30 years of research in this field. Random doorstep samples of people in Los Angeles, Sydney, London, Glasgow, and Johannesburg received, therefore, scales to measure authoritarian personality. The Los Angeles, Johannesburg, and Sydney samples also received scales to measure authoritarian attitudes and racial attitudes. Authoritarian personality was measured by a behavior inventory and authoritarian attitudes by a balanced version of the original F scale. Sample sizes were generally around 100 in each city. There were no significant cross-cultural differences in either authoritarian attitudes or authoritarian personality, but the American sample was significantly less prejudiced than the Australian or South African samples. Scores on original F items gathered in Los Angeles in 1979 were significantly higher than those tabulated by Adorno et al in 1950.


Since its publication in 1950, The Authoritarian Personality, by Adorno, Frenkel-Brunswik, Levinson, and Sanford, has generated an amount of derivative research which may well be a record in social psychology: Less happily, it has also been one of the most heavily criticized works produced (2, 4, 5, 20, 21).

In spite of the criticisms, most psychologists probably still believe that there is at least a grain of truth in the book's basic thesis. For this reason, it seems appropriate to mark the thirtieth anniversary of its publication with an investigation into how well that thesis fits the modern world. Attitudes and practices have changed vastly in almost all departments of life since 1950. That we might have become less authoritarian as part of that change is an obvious hypothesis. There is some evidence to suggest that this has taken place in South Africa in recent years (8), but such evidence cannot, of course, be automatically generalized elsewhere. If authoritarianism has lessened worldwide, the effect of this on the relationship between authoritarianism and prejudice would also need to be examined. To maximize the degree of comparability with the 1950 research, it does seem important that the present investigation should be based on results gathered in California. Thus cross-cultural and diachronic differences may potentially be examined separately rather than being confounded.


Clearly, a straight replication of the original California work cannot and should not be done. The great volume of criticism that has been directed at the methodology of the original studies must be taken into account. The criticisms directed at the measuring instrument most used in the original work (the F or "Fascism" scale) seem particularly important.

The three chief criticisms of the F scale are taken to be the following: (a) its openness to acquiescent response set contamination; (b) its ideological bias; and (c) its lack of behavioral validity.

The impact of the first of these criticisms would appear to have been somewhat reduced by the work of Rorer (19), whose work might be seen as at least shifting the burden of proof onto the critics of the F scale. This burden was, however, rather convincingly shouldered in a rapid reply to Rorer by Peabody (7). Rorer does not appear to have published a reply to Peabody and Peabody's work would in fact appear to have been at least partly an outcome of discussions with Rorer. In any event, the advent of revised F scale versions controlling for the influence of acquiescence (3, 9) makes it relatively easy to meet this objection. In the present study, the Ray (9) balanced F (BF) scale was used in a 14-item short form. This short form has been shown in previous research to have correlations between its original and reversed halves of -.50 (before reverse-scoring). This finding (17) is a fairly convincing "No" to the pessimistic question posed by Christie, Havel, and Seidenberg (5) in the title of their influential article. The problem of ideological bias in the F scale (2, 4, 21) is not as easy to overcome. The work of Rokeach (18) is the best-known attempt to remedy this defect but it is, regrettably, far from obvious that dogmatism is the whole of authoritarianism or even that it is a central component. It is probably most safely viewed as an empirical correlate of authoritarianism. The Rokeach D scale, furthermore, does normally show a bias towards the political right (2, 10). It would also appear that D scale items are equally open to the charge of "ambiguity" (7) levelled at the F scale. Like the F scale, the D scale lacks balance against acquiescence.

In the circumstances, the solution adopted to this problem was to supplement the BF scale with the Ray (12) "Directiveness" scale. This is a measure of authoritarianism in traditional personality scale format (a behavior inventory) which has repeatedly shown no correlation whatsoever with either vote or ideology (12, 13).

The "Directiveness" scale also formed the basis for dealing with the third problem --behavioral validity. The point that the F scale correlates mainly with other pencil and paper measures rather than with observed behavior was first made by Titus and Hollander (21) and confirmed by Titus (20). An extensive review of more recent evidence to the same effect was included in Ray (12). A measure of authoritarian personality that will predict behavior is therefore necessary. In validity studies, the Directiveness scale has been shown to have this property (12). It was used in the present work in a short form of 14 items -- nos. 1, 2, 6, 8, 9, 10, 11, 12, 14, 16, 17, 18, 22, and 25 of the original.

Another scale used in the present research is the Attitude to Blacks scale produced by Heaven and Moerdyk (6) as an adaptation of the Ray (11) Attitude to Aborigines scale. It was decided to focus on attitude to blacks, specifically, rather than ethnocentrism, generally, because of evidence that attitudes to ethnic outgroups may not be as general as was once thought (11).

The two 14-item authoritarianism scales and the 10-item Attitude to Blacks scale were administered to a random doorstep cluster sample of the population of Los Angeles by the market research firm, Facts Consolidated. There were a total of 101 interviews (10% verified) gathered throughout the Los Angeles conurbation. This was defined to include the whole of Los Angeles and Orange counties but not San Bernardino or Ventura counties. The sample included 47 females and 54 males. Average age was 37.3 years, and only whites, for obvious reasons, were interviewed. Blacks who were encountered in the sampling process were asked quick "dummy" questions solely for the purpose of extricating the interviewer. The Los Angeles urban area was chosen as the largest in the State. Thus, although the results cannot be taken as totally representative of the whole state, they are based on a large enough part of it to have considerable significance in their own right. With a population of roughly 14 million, the Greater Los Angeles area is, in fact, the second largest English-speaking conurbation in both the U. S. A. and the world. The sample N of 101 was decided on in view of the small gains in statistical significance to be obtained from larger samples. With an N of 101, correlations explaining less than 4% of the variance will be shown as significant. Effects weaker than this were not thought to be of great interest.

Other studies using similar methodology were carried out in London, Glasgow, Sydney, and Johannesburg. As they are more fully described elsewhere (13, 14, 15), they will be mentioned only briefly when used for purposes of comparison.


The reliabilities of the three scales-as assessed by Cronbach's coefficient "alpha"-were as follows: Authoritarian attitudes (BF scale) .79; Authoritarian personality (Directiveness scale) .73; and Attitude to Blacks .83. The two halves of the balanced F (BF) scale correlated -.45 before reverse-scoring. Although this is less than the -.6 generally found with the full 28-item BF scale, it is nonetheless quite satisfactory for a short form.

Mean scores on the three scales were as follows: BF scale 56.79 (SD 13.91), Directiveness scale 29.28 (SD 5.79), and Attitude to Blacks scale 27.63 (SD 6.72). The various items of the three scales were responded to respectively on seven-, three-, and five-point scales. When converted to the average item score used by Adorno et al., the BF scale mean becomes 4.05 (SD .99). This means that Los Angelenos generally scored almost exactly on the theoretical midpoint of the scale. In similar scoring, the means (and SDs) of the positive and negative halves of the BF scale were 4.35 (1.27) and 3.76 (1.05), respectively. This compares with an overall mean of 3.81 for the original F scale given by Adorno et al. (1) in their Table 9 (VII). The difference between the positively worded half of the BF scale and the positively worded original scale gives rise to a t of 4.76, which is significant at the .Ol level. Thus the 1979 California sample were more authoritarian in their attitudes than were the samples interviewed by Adorno et al. in the late '40s. In fact, of all the individual samples tabulated by Adorno et al. in their Table 12 (VII), only the San Quentin prisoners scored higher than the 1979 Los Angeles sample. As this seemed a rather remarkable result, the scores for the present seven positive items were compared only with the scores for the same seven positive items as tabulated in the 1950 Table 9 (VII). The 1950 scores on these seven items only had a mean of 4.12. Comparing this mean with the present mean gave rise to a t of 2.01, which is still significant though at only the .05 level.

Cross-cultural comparisons of BF score were hindered somewhat by the fact that the scale has usually been administered with five response options per item rather than the seven favored by Adorno et al. The only study in which the BF scale was previously administered with seven-point responses was a postal (mail-out) survey of the Australian state of New South Wales. Five hundred names were selected from the electoral rolls (voter registration in Australia is universal and compulsory) and mailed a questionnaire in early 1979. At the time of writing (five months later) 34% (170) had been returned. The demographic structure (means on age, sex, occupation, and education) of the resulting sample was virtually identical with that of contemporaneous doorstep samples obtained in the Sydney metropolitan area. Thus although the low response rate is some cause for concern, the sample is nonetheless representative in a demographic sense. The BF mean obtained on this sample, then, was 59.79 (SD 12.40). This is not significantly different from the Los Angeles mean. In another random doorstep cluster sample (N = 100) carried out in 1978 in the Johannesburg Greater Metropolitan area (Republic of South Africa) with the use of the BF scale with five-point responses, the South African mean was found to fall within the same range as Australian means (14). It may then be concluded that no differences in authoritarian attitudes have been shown between Australia, South Africa, and the U.S.A.

The Johannesburg sample mentioned above did, however, receive exactly the same version of the "Directiveness" scale as that administered in Los Angeles. The South African mean was 28.07 (SD 5.78). The difference between the two means was not statistically significant. As Johannesburg is the largest city in South Africa (one million out of South Africa's four million whites live in the Witwatersrand area), the sampling frame is highly comparable with that used in California. In each case largest cities were used and only whites were interviewed. It may then be concluded with some confidence that South Africans were no more likely than Californians to display authoritarian personalities.

The same conclusion may be drawn with respect to Australia. In another random doorstep cluster sample of the Sydney metropolitan area done in 1976 the mean on the same short version of the Directiveness scale was 29.69 (SD 6.16; N = 95). This mean is in fact even closer to the California mean, and again Sydney is the largest city (pop. approx. three million) in the country, containing, in fact, over 20% of Australia's population. Although the Sydney sample did not specifically exclude blacks, none were encountered by the interviewers. As Aborigines (blacks) constitute only 1% (approx. ) of Australia's population, this is not inherently surprising.

In random doorstep cluster samples (carried out in 1977) of the London and Glasgow conurbations (both with an N of 100) which received a slightly different short form of the Directiveness scale, mean scores on authoritarian personality were virtually identical in Australia, Scotland, and England (15). Overall, then, it may be concluded that Australia, England, Scotland, South Africa, and California do not differ. in prevalence of authoritarian personalities. This conclusion is based in each case on means in the largest cities of the respective communities. The U. K. samples did not exclude blacks, but only in the London sample were, in fact, any included (N = 3).

Cross-cultural comparisons on attitude to blacks are available for South Africa and Australia. The Johannesburg sample mentioned above received the same scale as that administered in California and produced a mean of 29.75 (SD 4.88). This is significantly higher than the Los Angeles mean (t = 2.54). As has been shown (14), Australian means bracket the South African mean. We have, then, the slightly odd conclusion that Australia with very few blacks and South Africa with very many blacks both show more prejudice than does California with its intermediate number of blacks.

Finally, the correlations with attitude to blacks were as follows: Directiveness scale .01, BF scale .44. The comparable correlations on the Johannesburg sample were -.02 and .54. The comparable correlations in the Australian mail-out survey described above were .10 and .32. In all three countries, then, authoritarian personality bore no relation to racial attitudes but authoritarian attitudes did. People who behaved in an authoritarian way were not particularly likely to be prejudiced but those who had authoritarian attitudes were highly likely to be prejudiced. The correlations between authoritarian personality and authoritarian attitudes were -.21 in Los Angeles, -.19 in Johannesburg, and -.24 in New South Wales. In all three countries, there was a significant but very weak tendency for people with authoritarian attitudes not to behave in an authoritarian way.


Americans customarily make strong distinctions between the different states in the Union. Although to outsiders the similarities are more evident, New York, Texas, and California (for instance) would normally be considered very different places indeed. This being so, the possibility of findings obtained in California being highly generalizable seemed questionable. It is therefore encouraging that the present work has shown results obtained in California to be very similar to those obtained in other parts of the English-speaking world. Californians are not so different after all. The one difference was the slightly lower prevalence of negative attitudes towards blacks. The reason for this must remain moot, but one possibility is the very strong public discouragement given towards negative racial attitudes in the United States. In Australia, where there are very few blacks and in South Africa where blacks are a serious threat, such public discouragement is not as common. Another possibility is that the progress made by American blacks in recent years has altered attitudes towards them. There has been little such progress in Australia or South Africa.

The diachronic differences in Californian attitudes were, however, more surprising. California of mid-1979 was surely more liberal than California of the late 1940's. The present work appears to show a decrease in liberalism. Californians themselves would probably be inclined to point to the fact that the early work was done in Northern California, while the present work was done in Southern California. Northern Californians might, in fact, say that the present results confirm what they have always thought about the South. A more likely explanation for the present results is the inadequacy of the early samples. The term "samples" for the groups of respondents surveyed by Adorno et al. is something of a misnomer. They were simply whatever groups of people happened to be accessible to the authors at the time and were not representative of any specifiable population. Thus although the groups interviewed in the '40s were fairly variegated, they did contain a fairly obvious bias towards people undergoing tertiary education. This "middle-class" bias of the Adorno et al. samples has of course been noted before (4). The present results would then appear to give strong confirmation to this stricture; the fact that the original samples were largely made up of students led to mean scores for the F scale being unrepresentatively low. They were so low in fact that 30 years of fairly steady progress towards a more liberal society has still not brought the general population up to the levels of liberalism then reported.

Nonetheless, the major finding by Adorno et al. -- a strong relationship between their F scale and racial prejudice-has been replicated. It was replicated, in fact, even after the methodological change of control for acquiescence had been incorporated. Thus although Rorer (19) was almost certainly wrong in declaring acquiescent response style to be a "myth" (7, 16), it would appear to be nonetheless true that acquiescence does not have the role in explaining the Adorno et al. results that was once feared.

If the present work has eliminated class-biased sampling and acquiescence as serious considerations in evaluating the Adorno et al. results, it has certainly not eliminated other factors. It has instead brought into sharp focus the very considerable error committed by Adorno et al. when they assumed that attitudes could be used as an index of personality. Far from attitudes and behavior tendency being positively correlated, the present work has shown that there is in fact a slight tendency for a behaviorally valid scale of authoritarian personality to correlate negatively with the F scale: high scorers on the F scale were grossly misidentified by Adorno et al. Whatever they are, they are not authoritarians in any behavioral sense. It has been argued elsewhere (10) that they are, in fact, nothing other than conservatives: The so-called right-wing "bias" in the F scale is in fact the whole of what the scale measures. The fact that the F scale did predict racial attitudes while the Directiveness scale did not would imply then that it is conservatism, not authoritarianism, which predisposes one to racial prejudice.

The present work has therefore shown that the results obtained by Adorno et al. (1) are not seriously affected by (a) the passage of time, (b) acquiescent response set; (c) middle-class sample bias, or (d) being based on California only. The same basic result was obtained 30 years later with balanced scales and representative, worldwide samples, but it has called into serious question the meaning of a correlation between F scale score and racial prejudice. If the F scale is not valid as a predictor of explicitly authoritarian behavior, there was probably nothing Adorno et al. (1) did at all which in fact constituted a study of authoritarianism.


{Articles below by J.J. Ray can generally be accessed simply by clicking on the name of the article. I am however also gradually putting online a lot of abstracts, extracts and summaries from older articles by other authors so if an article not highlighted below seems of particular interest, clicking here or here might just save you a trip to the library}

1. ADORNO, T. W., FRENKEL-BRUNSWIK, E., LEVINSON, D. J., $L SANFORD, R. N. The Authoritarian Personality New York: Harper, 1950.

2. BROWN, R. Social Psychology New York: Free Press, 1965. Chap. 10.

3. BYRNE, D., & BOUNDS, C. The reversal of F scale items. Psychol. Rep., 1964, 14, 216.

4. CHRISTIE, R., & JAHODA, M. Studies in the Scope and Method of The Authoritarian Personality. Glencoe, Ill.: Free Press, 1954.

5. CHRISTIE, R., HAVEL, J., & SEIDENBERG, B. IS the F scale irreversible? J. Abn. Soc. Psychol., 1956, 56, 141-158.

6. HEAVEN, P. C. L., & MOERDYK, A. Prejudice revisited: A pilot study using Ray's scale. J. Behav. Sci., 1977, 2, 217-220.

7. PEABODY, D. Authoritarianism scales and response bias. Psychol. Bull., 1966, 65, 11-23.

8. QUESNELL, F. J., VAN DER SPUY, H. I. J., & OXTOBY, R. In H. I. J. van der Spuy & D. A. F. Shamley, The Psychology of Apartheid. Washington, D.C.: Univ. Press of America, 1978. Chap. 10.

9. RAY, J.J. (1972) A new balanced F scale -- And its relation to social class. Australian Psychologist 7, 155-166.

10. RAY, J.J. (1973) Conservatism, authoritarianism and related variables: A review and an empirical study. Ch. 2 in: G.D. Wilson (Ed.) The psychology of conservatism London: Academic Press.

11. RAY, J.J. (1974) Are racists ethnocentric? Ch. 46 in Ray, J.J. Conservatism as heresy Sydney: A.N.Z. Book Co.

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

13. RAY, J.J. (1979) Does authoritarianism of personality go with conservatism? Australian Journal of Psychology 31, 9-14.

14. RAY, J.J. (1980) Racism and authoritarianism among white South Africans. Journal of Social Psychology, 110, 29-37.

15. RAY, J.J. (1979) Authoritarianism in Australia, England and Scotland. Journal of Social Psychology 108, 271-272.

16. RAY, J.J. (1979) Is the acquiescent response style not so mythical after all? Some results from a successful balanced F scale. Journal of Personality Assessment 43, 638-643.

17. RAY, J.J. (1979) A short balanced F scale. Journal of Social Psychology, 109, 309-310.

18. ROKEACH, M. The Open and Closed Mind. New York: Basic Books, 1960.

19. RORER, L. G. The great response style myth. Psychol. Bull., 1965, 63, 129-156.

20. TITUS, H. E. 'F" scale validity considered against peer nomination criteria. Psychol. Rec., 1968, 18, 395-403.

21. TITUS, H. E., & HOLLANDER, E. P. The California F scale in psychological research: 1950-1955. Psychol. Bull. 1957, 54, 47-64.

University of New South Wales, School of Sociology, P. O. Box 1, Kensington, New South Wales, Australia 2033

Go to Index page for this site

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