Make your own free website on

Personality & Individual Differences Vol. 4, No. 6, pp. 699-701, 1983

Race and climate as influences on anxiety


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

(Received 1 February 1983).

Summary -- Lynn (1971) argues that Nordic race and cool climates independently predispose people to low levels of chronic anxiety. He also associates high economic growth and high anxiety. Random population samples of the cities of Sydney (Australia), Munich (W. Germany) and Bombay (India) received the short form of Eysenck's MPI N scale and the means obtained were compared with one another and with the scale's English norms. Two studies of the Parsee communities in Bombay and Sydney were also carried out. Except for the Bombay Parsees, all translations of the scale (into Gujurati, Hindi, Marathi and German) appeared to have been successful as evidenced by satisfactory levels of alpha. There were no significant differences between the Indian, Australian, English and German means -- thus upsetting Lynn's theories. Sydney Parsees were however significantly less anxious than Bombay Parsees -- thus suggesting that Parsee immigration to Australia is a viable solution to the threatened position in which Parsees find themselves in India.


Although for a time it appears to have been considered a rather suspect field of research, the study of national personality characteristics seems to be making something of a comeback among psychologists (Milgram, 1961; Jacob, Teune and Watts, 1967; Terhune. 1970; Lynn and Hampson, 1975; Iwawaki, Eysenck and Eysenck, 1977; Blunt, 1979; Cattell, Graham and Woliver, 1979). Political scientists have long quoted (for instance) De Tocqueville on the characteristics of Americans and epidemiology is a very important evidential support for many modern theories of preventive medicine, so why should not psychologists also test for national difference in their field?

One of the most interesting recent contributions to the field is a book by Lynn (1971). As in his later work (Lynn and Hampson, 1975), in this book Lynn factor analyses epidemiological statistics of various countries and finds two factors which he identifies as Neuroticism and Extraversion. His principal 1971 conclusions would appear to be that warm climates lead to high anxiety and that Nordic race leads to low anxiety. He also relates high rates of national economic growth to high mean levels of anxiety in the relevant populations.

Some support for the latter theory could be inferred from findings by Ray and Kiefl (1984) to the effect that low rates of economic growth are associated with high national means on Achievement Motivation. Since Achievement Motivation is generally negatively associated with Neuroticism in the data used by Ray and Kiefl, this appears to imply that low rates of growth are associated with low anxiety. This is, however, a quite long chain of inference which is rendered rather suspect by the fact that the correlations between the variables concerned are far from high. It is clear that what is needed for Lynn's inferences to become really well established is actual national means on standard indices of anxiety or neuroticism.

Lynn himself has recently made some initial moves in this direction (Lynn, 1981) by presenting national means for many countries derived from Eysenck's `N' and 'E' personality inventories (Eysenck and Eysenck, 1969). He arbitrarily defines the British mean on these two variables as 50 and presents means for other countries in terms of this. As his Appendix 1 illustrates, however, roughly half the data used to compute these `national' means were in fact derived from available groups of university or college students. No matter what `corrections' are made for this, 'national' means derived from such data can only be highly speculative.

The present work, therefore, presents means on a form of Eysenck's N scale which are derived from random population samples in several countries. They are not national samples but rather samples drawn from the largest (or nearly the largest) city in each country. Thus, although it is still not ideal data, the samples are similarly drawn for each country and the means should at least represent some improvement over those presented by Lynn (1981).

The short-form N scale from the Eysenck (1959) MPI was the one used. In survey research there is always pressure towards brevity and the MPI has the attraction of allowing a 'Not sure' response -- which increases subject cooperation.

Following Lynn, the English mean on the scale (6.15; SD 3.33) is used as a starting point for comparing means in several other countries that vary along dimensions suggested by Lynn. The countries involved are India, Australia and Germany. The notoriously hot Indian climate should enable a particularly good test of the hypothesized effects of climate when contrasted with the two Northern European countries of England and Germany. Australia is particularly useful because it represents a warm climate which is nonetheless inhabited by a people of overwhelmingly Northern European ancestry.

Another feature of the work to be reported below is the data on the neuroticism of an unusual minority group common to two of the countries -- the Parsees. The Parsees are a tiny but very affluent Indian ethnic minority who, although they have lived in India for over a thousand years, have for religious reasons remained substantially endogamous for all that time. They originate from Fars in Persia (now Iran) and are still generally very fair in skin colour by Indian standards. Those who have in recent years emigrated to Australia are generally there mistaken for Southern Europeans. Both Lynn (19717 and Eysenck (1967) have suggested that there might be some genetic basis for the various levels of neuroticism observed so it should be interesting to see in what ways the Parsees differ from both the North Indian and Australian communities in which they live. As Persia is a generally quite cool climate, the Parsees might perhaps be expected to show characteristics more similar to the Northern Europeans than to the Indians.

The use of Indian data might at first seem questionable as Lynn makes an explicit point of excluding data from third World countries. He does this, however, only because poverty could have a quite obviously distorting effect on the epidemiological statistics which are his raw material. Poverty is, however, a far from obvious distorting influence on direct measures of personality -- particularly ones as basic as those considered by Lynn.

The hypotheses derived from Lynn's theories for the purposes of the present work are then as follows: Lynn considers that Australians are predominantly Nordic so we expect_ that their mean N-scale scores should be similar to the English but lower than the West Germans; the very warm Indian climate should lead to very high N-scale scores; Germany is the outstanding country for economic growth so it should have an outstandingly high N-scale mean; the Parsees should fall in between the Indian and northern European mean scores on the N scale.


Various data from this study have now been reported in a wide range of publications (e.g. Ray, 1979, 1981). The short N scale was one of five scales administered to a random cluster sample of the Sydney metropolitan area of Australia. There were 95 respondents. Fuller details can be found in the publications cited. Cluster sampling is the method used by most commercial public opinion polls-where it generally gives very accurate results. Although a sample size of 95 may seem small, it should be noted that the level of t required for significance at the 0.05 level changes only from 1.986 to 1.972 as one moves from a sample size of 95-200. For the comparisons made in the present paper, therefore, t will in fact for all intents and purposes be at the asymptote for significance.

In the present author's practice, the three possible responses to each item of the N scale ("Yes", and "?" and "No") are usually scored 3, 2 and 1 so any N-scale means reported elsewhere would not be directly comparable with those given by Eysenck --who uses a 2,1,0 scoring system for the same responses. A scale mean obtained with the one scoring system can in fact be converted to the other simply by adding or subtracting 6 but, to avoid confusion, all means in the present paper will be given in terms of Eysenck's system. The mean obtained in Sydney, then, was 6.63 (SD 3.57). The coefficient alpha reliability of the scale was 0.73 -- very satisfactory for such a short scale.


Full details of this study have already been given in Ray (1982). Briefly, a random cluster sample consisting of 305 people was obtained of the Bombay conurbation in India by a local Bombay market research firm. The questionnaire was the same as that used in Study I but was translated into the three main languages used in Bombay -- Marathi, Gujurati and Hindi. For the purposes of the present analysis, all responses were aggregated regardless of the language in which they were obtained. As even the three main languages would have left substantial numbers of the population unreached, each questionnaire was in fact set out bilingually -- with the English form of each item being given alongside the translated form. In one way or another, therefore, virtually every one reached could be communicated with. The N-scale reliability achieved under these difficult circumstances of communication was therefore most satisfactory -- 0.72. It suggests that little was lost in translation. The mean was 6.86 (SD 3.55).


This study has been fully described in Ray and Kiefl (1984). It was a random cluster sample of 136 respondents in the Munich conurbation of Bavaria, W. Germany. Here the translation of the scale would appear to have excelled itself. The alpha of 0.84 suggests that the items were in fact more internally consistent in this sample than they were for Eysenck's norming sample of English respondents (with a reliability of 0.80). The mean was 5.56 (SD 4.09).


This study was carried out a few months after Study II. It consisted of 90 Parsees living in Bombay. As there was no master list of the Parsee population available and as they are too tiny a minority to show up in useful numbers in a random cluster sample, interviewees were selected by a 'friends and neighbours' system that seeems to have given reasonable representativeness in the past. Fuller details are available elsewhere (Ray, 1983a). The reliability of the scale on this population was only 0.49 -- indicating that the results can be used only with considerable caution. The mean was 7.21 (SD 2.98).


This study consisted of 72 respondents from the Parsee community in Sydney, Australia. A master list of all Parsees living in Australia was available from the Australian Zoroastrian Association and with their cooperation was used as the basis for a mail-out questionnaire. One reminder was sent to non-respondents and reply-paid envelopes were provided. The final response rate was about 50% -- which is unusually high for a postal survey. Further details of the study can be found in Ray (1986). The questionnaire was a bilingual one -- in both English and Parsee Gujurati. All Parsees living in New South Wales were sent a copy but of these all but four lived in the Sydney area. The reliability of the scale was a very satisfactory 0.77 and the mean was 5.84 (SD 3.60). The 't' for the Study IV and V means is 2.63 (P < 0.01).


The five studies were carried out respectively in 1976, 1981, 1982, 1981 and 1982 so were reasonably contemporaneous. It will be seen that the first hypothesis was poorly supported. Although the mean given by Eysenck was derived from a quota sample covering (presumably) the whole of England, the present sample of only Australia's largest city gave a mean so close to the English one as to be virtually indistinguishable. Lynn would presumably interpret this as indicating that race has triumphed over climate. The main point of the first hypothesis was however to compare the Anglo-Saxon mean with the German mean. Here we find that the results run very contrary to Lynn's theorizing. South Germany is very clearly a high-anxiety area in Lynn's scheme of things but the present results in fact show a mean that is low rather than high in relation to the Anglo-Saxon means. Lynn's attempted differentiations by race were then not supported.

As there are probably few major cities outside India itself which are more consistently very hot than Bombay, the results there should be a fairly decisive test for Lynn's theories concerning the effect of climate. What was in fact found, however, was a mean that was virtually indistinguishable from that observed among the English and Australian respondents. Lynn's theory concerning climate would also therefore seem to be not confirmed.

Australia, India and the U.K. have all shown economic growth rates since the Second World War averaging somewhere between 1 and 2% per annum. West German rates of growth in G.N.P. have also dropped to something like this in very recent years but, as is well known, for many years before that Germany was hailed as one of the post-war economic miracles with growth rates averaging close to 8% p.a. The comparison of U.K. and German growth rates is particularly clear -- with Germany having reached a standard of living which is, according to some, twice that of the U.K. from a starting point well behind the U.K. in 1945. We once again therefore have available a particularly strong test of another of Lynn's theories. If anxiety has anything to do with growth, then Germans should clearly have much higher N scores than the British. In fact, of all groups studied, the Germans had the lowest score. As it seems implausible that Germans have suddenly changed their personalities in recent years, we must yet again suspect that Lynn's theories are very wide of the mark.

In view of the consistent failures of Lynn's hypotheses, nothing could in the end reasonably be predicted of the Parsee means. The high mean levels of anxiety of the Bombay Parsees are however very consistent with Parsee fears about their exposed position in post-independence India. Before independence, the British exercised power in India very largely through Parsee intermediaries -- apparently because Parsees were the most economically-competent group in India. As they still to this day manage to maintain an almost Western standard of living amid the incredible poverty that is India generally, it will be seen why Parsees have been very anxious for their security ever since the loss of their British patrons. The Parsee emigration to Sydney and other large cities of the English-speaking world has so far been the most tangible sign of that anxiety. We now however have psychological data which not only confirm directly the hypothesized Parsee chronic anxiety but which also suggest that emigration at least to Australia is a fully adequate solution to the Parsee problem. The Sydney Parsees were second only to the Germans in their low levels of anxiety. As the total world population of Parsees is considerably less than Australia's total annual immigration intake, even a 'flood' of Parsees could not be expected to alter this situation. As Australia appears now to be virtually the only country with comparatively open doors towards immigrants of all races, colours and creeds, it is a very great pleasure for an Australian author to record that what Australia offers can be so helpful to members of another population which finds itself under stress.

As the above programme of research would appear to have disconfirmed Lynn's (1971) theories across the board, we are left with the problem of working out how Lynn's data and the present data can be reconciled. Could the scales in the present work have been poorly translated? This was almost certainly not so. All translations were made by committees of native speakers and the internal consistencies (alpha) observed in the English-speaking and non-English speaking countries were reasonably similar. Although back-translation is widely recommended it was not used on the present occasion as it seemed that the same thing can usually be expressed in a number of different ways in any modern language and a 'failure' of back-translation could well in fact be unimportant. Consensus between translators and internal consistency were felt to be more important safeguards.

One can only say, therefore, that although some generalizations from epidemiological data to personality seem possible, they can only be made with extreme caution. There are differences as well as similarities. Conclusions drawn from the one may be a poor guide to the other.

Perhaps the most interesting aspect of the present work is its relationship to the contention by the Eysencks (Eysenck and Eysenck, 1969) that Neuroticism is an extremely basic variable of personality with strong biological/genetic foundations. This being so, one might well imagine that N scores would be little influenced by culture and might well be fairly invariant from country to country. The present data certainly support that inference. To a remarkable extent, the British mean on Neuroticism is also the international mean.


Blunt P. (1979) Personality patterns in British and South African managers. J. soc. Psychol. 107, 127-128.

Cattell R. B., Graham R. K. and Woliver R. E. (1979) A reassessment of the factorial, cultural dimensions of modern nations. J. soc. Ps ychol. 108, 241-258.

Eysenck H. J. (1959) Manual of the Maudsley Personality Inventory. Univ. London Press.

Eysenck H. J. (1967) The Biological Basis of Personality. Thomas, Springfield, Ill.

Eysenck H.J. and Eysenck S. B. G. (1969) Personality Structure and Measurement. Routledge & Kegan Paul, London.

Iwawaki S., Eysenck S. B. G. and Eysenck H. J. (1977) Differences in personality between Japanese and English. J. soc. Psychol. 102, 27-33.

Jacob P. E., Teune H. and Watts T. (1967) Values, leadership and development: a four-nation study. Soc. Sci. Inf. 7, 49-92.

Lynn R. (1971) Personality and National Character. Pergamon Press, Oxford.

Lynn R. (Ed.) (1981) Cross-cultural differences in Neuroticism, Extraversion and Psychoticism. In Dimensions of Per.sonalit,v: Papers in Honour of H. J. Evsenck, Chap. 12. Pergamon Press, Oxford.

Lynn R. and Hampson S. L. (1975) National differences in extraversion and neuroticism. Br. J. soc. clin. Psychol. 14, 223-240.

Milgram S. (1961) Nationality and conformity. Scient. Am. 205, 45-51.

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

Ray, J.J. (1981) Do authoritarian attitudes or authoritarian personality reflect mental illness? S. African J. Psychology 11, 153-157.

Ray, J.J. (1982) Authoritarianism and achievement motivation in India. J. Social Psychology 117, 171-182.

Ray, J.J. (1983) Ambition and dominance among the Parsees of India. Journal of Social Psychology 119, 173-179.

Ray, J.J. (1986) The traits of immigrants: A case study of the Sydney Parsees. J. Comparative Family Studies 17, 127-130.

Ray, J.J. & Kiefl, W. (1984) Authoritarianism and achievement motivation in contemporary West Germany. Journal of Social Psychology, 122, 3-19.

Terhune K. W. (1970) From national character to national behavior: a reformulation. J. conflict Resol. 14, 203-263.

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)