Make your own free website on

South African J. Psychology, 1994, 24(4), 231.

(With three post-publication addenda following the original article)


J.J. Ray

Sociology, University of New South Wales, Australia


Duckitt reports that his Subtle Racism scale predicts both racist behaviour and authoritarian attitudes. His scale, however, correlates only negligibly with reported racist behaviour and its correlation with the RWA scale probably tells us more about conservatism than authoritarianism. Conservatives do not, however, usually behave in more racist ways so Duckitt should report his findings on that question.

Is subtle racism orthogonal to racist behaviour?

A recent article by Duckitt [South African Journal of Psychology, 1993, 23(3)], purported to present behavioural validation for a 'Subtle Racism' scale and went on to report that subtle racists were highly likely to be authoritarian.

One can perhaps understand that some authorial pride is at work when we see Duckitt reporting attitude/behaviour correlations of around 0.2 as strong support for the validity of his scale but the plain fact is that a correlation of 0.2 demonstrates only 4% common variance and hence does little to upset the long-recognized orthogonality between attitudes and behaviour in this area (Crosby, Bromley & Saxe, 1980; LaPiere, 1934; Melamed, 1970; Rule, Haley & McCormack, 1971). Clearly subtle racists too are slow to translate attitudes into behaviour and the 'improvement' that Duckitt's scale represents seems therefore still to be demonstrated.

Authoritarianism or Conservatism?

Duckitt correctly noted that many authors have found little relevance for the concept of authoritarianism as an explanation of racist attitudes in South Africa but appears to believe that his findings in some way throw this consensus into question. The finding he alludes to was a 0.73 correlation between his Subtle Racism scale and a short form of the Altemeyer (1981) RWA scale.

So is Duckitt the only marcher who is in step? Sadly, no. The name RWA stands for Right-Wing Authoritarianism so there is no doubt that the scale measures at least in part some sort of Conservative ideology. How do we know which aspect of the RWA scale is producing the correlation with Subtle Racism? Is it the conservative part or the authoritarian part? Even Adorno, Frenkel-Brunswik, Levinson & Sanford (1950) held that the two types of attitudes were separable. Is it authoritarianism or conservatism then that makes co-operative female students at a conspicuously liberal and multi-racial South African university prone to avow racism? Who knows a priori? Duckitt's use of such a foolishly confounded measure as the RWA scale leaves him unable to answer such questions. His findings, as he describes them, are then mere assertion.

Since it was long ago noted (Ray, 1985b) that the RWA scale is highly correlated with other conservatism scales and has very faint claims as a predictor of authoritarianism, it is much more parsimonious to see Duckitt's findings with the RWA scale as yet another example of the hoary finding that conservatives are more ready to acknowledge racist attitudes (but are NOT more ready to behave in racist ways). Why this is so has been the subject of a variety of explanations but Ray (1984 & 1985) has presented considerations in favour of the view that conservatives have less need for denial in this area. Duckitt (1993) has shown that the approval motive is not involved but denial can, of course, be for reasons deeper (or shallower) than any immediate need for approval. A Leftist might, for instance, disavow racist attitudes simply in order to appear consistent when he/she makes his/her usual bold claim that 'All men are brothers'. Conservatives would have no such need.

Ideology and racist behaviour

It may be noted that Duckitt's reported findings regarding 'authoritarianism' fall into the now very tired mould of correlating different attitude measures with one-another. It was noted a very long time ago in this field (Titus & Hollander, 1957) that correlating attitude measures with behaviour produces very different results. The latest example of this would appear to be a study by Gough & Bradley (1993), who constructed a scale of rated racist behaviour and correlated it with a form of the California 'F scale (usually described as measuring authoritarianism). They found a correlation between the two of essentially zero (0.08 -- a great contrast with Duckitt's 0.73) and were reduced to trawling through a vast matrix of correlations with individual items in order to find some shred of (statistically dubious) evidence that would support the California theory of authoritarianism (Adorno et al., 1950).

It would therefore have been much more interesting if Duckitt had presented the correlations between his behaviour ratings and the RWA scale. Did they too show that tendency towards orthogonality that so bedevils the prediction of behaviour in this area?


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

Altemeyer, R. (1981). Right-wing authoritarianism. Winnipeg: University of Manitoba Press.

Crosby, F., Bromley, S. & Saxe, L. (1980). Recent unobtrusive studies of black and white discrimination and prejudice: a literature review. Psychological Bulletin, 87, 546-563.

Duckitt, J. (1993). Further validation of a subtle racism scale in South Africa. South African Journal of Psychology, 23, 116-119.

Gough, B. & Bradley, P (1993). Personal attributes of people described by others as intolerant. In P.M. Sniderman, P.E. Tetlook & E.G. Carmines (Eds), Prejudice, polities and the American dilemma, (pp. 6-85). Stanford, CA: Stanford University Press.

La Piere, R. (1934). Attitudes and actions. Social Forces, 13, 230-237.

Melamed, L. (1970). The relationship between actions and attitudes in a South African setting. South African Journal of Psychology, 1, 19-24.

Ray, J.J. (1984). Half of all racists are Left-wing. Political Psychology, 5, 227-236.

Ray, J.J. (1985a). The psychopathology of the political Left. High School Journal, 68, 415-423.

Ray, J.J. (1985b) Defective validity in the Altemeyer authoritarianism scale. Journal of Social Psychology 125, 271-272.

Rule, B.G., Haley, P & McCormack, J. (1971). Anti-Semitism, distraction and physical aggression. Canadian Journal of Behavioral Science, 3, 174-178.

Titus, H.E. & Hollander, E.P (1957) The California F scale in psychological research: 1950-1955. Psychological Bulletin, 54, 47-64.


1). The abstract given above was submitted along with the original publication but was not printed. Two typos have also been corrected in this version.

2). My criticisms above were swingeing but they were, even so, less comprehensive than they might have been. What Duckitt was TRYING to do was to show that people who agreed with subtly racist statements also admitted to behaving in certain types of racist ways. This is a very weak sort of validation for his subtle racism measure. All it is looking for is consistency between what people SAY they believe and what they SAY they do. It is not independent evidence about the behaviour of the people concerned at all. The concept of validation for a scale normally involves OUTSIDE evidence of how people behave -- often provided by peer ratings (i.e. how OTHER people describe the person in question). That descriptions of one's own behaviour might be inaccurate or misleading for any number of reasons is obvious and it really is something of a wonder that Duckitt's subjects did not impose more consistency on their self-descriptions than they did. In the circumstances, one must conclude that whatever else might be true of Duckitt's scale, its subtlety is not in any doubt. It appears that most of his subjects did NOT see it as expressing racism. Had they done so, much greater consistency in self-descriptions would have been a reasonable expectation of the obliging young ladies concerned.

3). Duckitt (1984) did respond to the published criticisms above with some reasonable and some very unreasonable comments. He notes for instance that by combining his ratings into a scale, he obtains a correlation with his SR attitude scale of .38 and rightly contends that this is a more respectable result than what was obtained by using individual ratings as validity criteria. Whether or not this is "high" or not can however only be judged by comparing it with the results of similar procedures elsewhere and it may be noted that similar methodology in Ray (1987) produced a .79 validity statistic. And it may be noted that the .79 result was obtained from PEER-ratings, not the much less demanding self-ratings that Duckitt used.

His next point is to draw some reasonable distinctions in assessing what constitutes an attitude-behaviour discrepancy and he claims in consequence that the references I cite are poor grounds for asserting such a discrepancy. Since he ends up agreeing that the discrepancy exists, however, I see no point in pursuing that line of argument.

His further claim that the attitude-behaviour discrepancy could be situational I also have no particular quarrel with and in fact the main reason why I (e.g. Ray & Lovejoy, 1983) make a point of using peer-ratings as a validity criterion is to overcome any such limitation. Peer-ratings are particularly valuable precisely because they tell us how a person characteristically behaves rather than how he/she behaves in one or two special circumstances. Duckitt used self-ratings rather than peer-ratings so his data is clearly much more subjective than peer-ratings would be but his ratings would nonetheless seem to have similar claims to measuring characteristic rather than situational behaviour. So a "situational" explanation for his poor results would seem simply irrelevant.

Duckitt next goes on to assert his (unreferenced) faith in the old Adorno et al. (1950) story that a whole host of things covary to form a single "authoritarianism" syndrome. One wonders where he has been. As long ago as 1965 Brown pointed out that the "components" of authoritarianism do NOT in fact all covary and Altemeyer's (1981) comment on the Adorno research quoted in Ray (1983) sums up very well the artifactual nature of such intercorrelations as do appear from time to time. Additionally, Duckitt himself points to the dubious validity of the Adorno F scale so why he continues to accept the conceptualization of authoritarianism that it embodies certainly needs more explanation than he gives.

Duckitt gets into more interesting territory, however, when he responds to my request that he report the correlation between his behaviour self-ratings and the RWA scale. He reports a correlation between the RWA scale and overall rated racist behaviour of .36. This certainly indicates that his co-operating white female students at a liberal university saw some commonality between the statements of the RWA scale and statements about racist behaviour. Those students who felt at ease with the conservative statements of the RWA scale also were slightly more likely to feel at ease in admitting to discriminatory behaviour. That is certainly consistent with what we know of South African society at the time but even so does NOT tell us anything about the relationship between what the students admitted and what they do. And even the relationship between racist admissions and conservative sentiments may give a misleading impression of cause and effect. That BOTH racially discriminatory behaviour AND the sort of conservative sentiments expressed in the RWA scale would tend to be decried in the university environment concerned seems more than a lively possibility and could alone account for the relatively modest correlation observed: Those who adapt best to the university environment avoid both racist admissions and expressions of conservatism. Sniderman, Brody & Kuklinski (1984) provided fairly explicit confirmation that such an effect exists with their demonstration that conservatism and racism are correlated only among more highly educated respondents.

Duckitt then goes on to provide some VERY interesting data about the correlation between RWA score and support for South African political parties. He reports a quite stark gap between the supporters of the generally pro-Apartheid parties and supporters of the generally anti-Apartheid parties. This contrasts markedly with Altemeyer's (1988 p. 239) own finding that the RWA scale does NOT provide any useful discrimination between the supporters of Canadian political parties. Why the difference?

I think the difference is to be explained in terms of my own finding (Ray, 1985) that the RWA scale is principally a measure of conservatism. And there is no doubt that Afrikaner society (the main base of support for Apartheid) at the time was a remarkably conservative and old-fashioned society by world standards. This was epitomized for me by the fact that psychology professors at Rand Afrikaans University in Johannesburg during the Apartheid era customarily walked around the university in a business suit and tie! And the "old-fashioned" nature of that Afrikaner conservatism would appear to be crucial. I have argued (Ray, 1990) at length elsewhere that the general irrelevance of the F scale to voting patterns is explainable by the F scale measuring a very old-fashioned (and hence currently irrelevant) form of conservatism and it seems parsimonious to explain the usual irrelevance of the RWA scale to voting in the same way. It seems likely that the RWA scale too measures a particularly old-fashioned form of conservatism. And the obliging Duckitt has provided an excellent test of just that hypothesis! IF the RWA scale measures a particularly old-fashioned form of conservatism, a particularly old-fashioned group of conservatives should get much elevated scores on it. And Duckitt has given a resounding demonstration of exactly that occurring!

After such excitement his next report of findings is very much of an anti-climax. He showed that old-fashioned conservatives are more polite and co-operative but tries to make much more out of it than that.

His final reference is to a group dynamics study that showed more attention and submission to authority among high scorers on the Adorno F scale. Since a couple of paragraphs earlier he described the F scale as "an extremely poor measure of authoritarianism" one wonders how he thinks he can infer anything at all from use of that scale (if it does not measure authoritarianism surely it must measure something else?) but perhaps we can be generous and let that question ride.

What was shown in the study concerned was that the high F scorers took more heed of the groups which those in authority over them had created. Since there has never been any controversy that people who say (via the F scale) that they respect authority do indeed show a somewhat greater submissiveness to it, this is unremarkable and overlooks the crucial finding by Titus (1968) -- replicated in Ray & Lovejoy, 1983 -- that a slight degree of submissiveness to authority is the ONLY behavioural correlate of the F scale. That the F scale does not also predict the domineering and aggressive behaviour that would seem to lie at the heart of any conception of authoritarianism is the central problem for Duckitt and all those like him who still accept the old Adorno et al (1950) account of things.

In conclusion, I must again express my surprise at Duckitt's continuing devotion to the Adorno conception of authoritarianism. His earlier study in which he showed a zero relationship between the Adorno F scale and any index of psychopathology among a large general population sample of white South Africans (Duckitt, 1983) is an excellent and important piece of work in evaluating the Adorno claims. The "pathological" nature of authoritarianism was of course a major claim of the Adorno et al (1950) work and few people have upset that contention more decisively than John Duckitt!

In the light of such findings showing that the supposed "components" of authoritarianism do not in fact covary, the logical thing to do would surely be to examine each of the supposed "components" of authoritarianism separately on all occasions but it is just that which Duckitt resolutely refuses to do -- which his use of the RWA scale of course exemplifies.


Altemeyer, R. (1981). Right-wing authoritarianism Winnipeg: University of Manitoba Press.

Brown, R.(1965) Social psychology N.Y.: Free Press. Brown, R.(1986) Social psychology (2nd. Ed.) N.Y.: Free Press.

Duckitt, J. (1983) Authoritarianism and adjustment in an authoritarian culture. J. Social Psychology, 121, 211-212.

Duckitt, J. (1994) 'Are subtle racists authoritarian?' Response to Ray. South African J. Psychology, 24 (4), 232-233.

Ray, J.J. (1983) Book review of "Right-wing authoritarianism" by R.A. Altemeyer "Australian Journal of Psychology 35, 267-268.

Ray, J.J. (1985) Defective validity in the Altemeyer authoritarianism scale. Journal of Social Psychology 125, 271-272.

Ray, J.J. (1987) The validity of self-reports. Personality Study & Group Behaviour 7(1), 68-70.

Ray, J.J. (1990) The old-fashioned personality. Human Relations, 43, 997-1015.

Ray, J.J. & Lovejoy, F.H. (1983). The behavioral validity of some recent measures of authoritarianism. Journal of Social Psychology, 120, 91-99.

Sniderman, P.M., Brody, R.A. & Kuklinski, J.H. (1984) Policy reasoning and political values: The problem of racial equality. Amer. J. Polit. Science 28, 75-94.

Titus, H.E. (1968). F scale validity considered against peer nomination criteria. Psychological Record, 18, 395-403.

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)