The Journal of Social Psychology, 1982, 116, 127-138.
THE PROTESTANT ETHIC IN AUSTRALIA*
University of New South Wales, Australia
JOHN J. RAY
A series of five studies is reported to investigate the relationship between religious belief or adherence and psychological characteristics. Samples from 75 to 240 people were used which had at least some claim to general population representativeness. Attributes studied included achievement motivation, task-orientation, success-orientation, Machiavellianism, alienation, dogmatism, authoritarianism, locus of control, neuroticism, social desirability orientation, metaphysical faith, belief in predestination, frugality, belief in the depravity of human nature, and Puritan morality. Twelve different scales measured various aspects of achievement and work motivation. Scales purportedly measuring the Protestant ethic by Blood and by Mirels and Garrett were also included. Not one of the scales showed any Catholic/Protestant differences which were statistically significant. Several, however, differentiated unbelievers from the religious. Unbelievers were those with especially high achievement motivation rather than Protestants; they were also found to be more machiavellian and less authoritarian.
The initial purpose of the program of research reported in this paper was to investigate the applicability of McClelland's (10) reformulation of Weber's (29) Protestant Ethic thesis to modern-day Australia. As will be seen, however, the nature of the findings led to the broadening of the project into a more general survey of the connection between religion and values. Weber himself saw a connection between the emergence of a capitalistic outlook and the emergence of Protestant Christianity. Calvinism in particular, he believed provided a social climate particularly conducive to the emergence of an enterprising, materialistic spirit.
The restriction of the thesis to Calvinism has been difficult to maintain and such a restriction plays no real part in McClelland's thesis, which applies quite as well to Protestantism as a whole as it does to Calvinism. The major impact of his refinement is to bring the thesis into the twentieth century. Where Weber saw a belief in predestination and the search for signs of "election" as the major psychological variables connecting religion and enterprise, McClelland postulated independence training and achievement motivation. Given the desuetude of predestinarian dogma among even nominally Calvinist churches (e.g., Presbyterians) in the twentieth century, the relevance of Weber's thesis to the modern world would be much more questionable without McClelland's reformulation. McClelland claimed simply that the importance of a personal relationship with the Deity among Protestants led to early independence training among children, thence to achievement motivation and thence to actual economic achievement. Obviously, belief in the importance of a personal and individual relationship with the Deity has survived into the modern world among Protestants much better than has concern over predestination.
The guiding questions behind the present research, therefore, were as follows: Does religion today in, fact have any impact on achievement motivation and are there any peculiarly Protestant values left in a very secularized and pluralistic modern world? Being the more specific, it is perhaps the first of these two questions that has most adequately been examined.
B. STUDY I
This study was planned as a preliminary study to sort out which of the various sorts of achievement motivation and related constructs showed most promise in differentiating Protestants and Catholics. Emphasis in making up the battery of tests to be used was on self-report tests because of the questionable reliability of projective tests (5). One projective test was, nonetheless, included: the French (8) "Test of Insight." The Mukherjee (14) forced choice "S.C.T." was included because of its generally good range of validity evidence. The remaining tests were in Likert format: The Costello (3) scales I & II (intrinsic versus extrinsic motivation), the McReynolds and Guevara (11) Failure-avoidance and Success-seeking scales, and the Mehrabian (12) scale for males. Also included was the Ray (24) "AO" scale in its various forms. This range of tests was designed to leave no stone unturned among the possible ways of measuring achievement motivation.
Most of the details of this study have been given elsewhere (18, 24); the sample was a community sample with an N of 75. On the basis of a question about their denominational background, the Ss were divided up into two groups -- 20 Roman Catholic and 51 Protestant. On none of the scales were the differences between Catholics and Protestants significant as determined by t tests.
C. STUDY II
The apparent "death" of the Protestant ethic among Australian Protestants in the above study certainly seemed very complete. It did therefore seem to stand in some need of rechecking with a random doorstep sample and with a somewhat more careful assessment of religious orientation. In this study, therefore, nine denominational categories were used (including "belief in God only") and Ss were asked both the religion of their upbringing and their religion at the time of interviewing.
The scales were the short Ray "AO" scale (22) and the Ray (25) "catchphrase" AO scale. The latter is a semiprojective instrument, whereas the former showed especially good validity characteristics in comparative validity studies (24).
Again, most details of this study are available elsewhere (25); the sample was a random doorstep sample of 145 Sydney people. As a correlation explaining only 2 % of the variance would be shown as significant at the .05 level with this N, a larger sample was thought unnecessary. It was found that no scale showed a significant relationship with religion, however categorized, as determined by chi squares.
D. STUDY III
At this point it seemed that whatever its applicability elsewhere, the McClelland version of the Weber thesis was certainly not applicable in present-day Australia. The findings did, however, reflect what Veroff et al. (28) had found using projective tests in the United States. They were certainly not the first negative results obtained in a test of McClelland's thesis.
It seemed, then, that one might have to look beyond achievement motivation to find something distinctive about the Protestant outlook. Many possibilities might have been investigated, but it was thought most promising to examine attitudes directly concerned with religion. The aspiration was to survey all the major elements in the traditional Protestant outlook. As some limitation had to be placed on the study, however, only six such elements were in fact selected: asceticism, belief in the "fallenness" or depravity of human nature, belief in predestination, frugality, metaphysical faith, and moralism or "Puritan" ethics. Although these concepts were not, of course, an exhaustive catalogue of what constitutes Protestantism, they were at least more comprehensive than those used to guide much previous research and were, additionally, ones that should have been familiar to most people with a Protestant background.
New scales were written to reflect sentiments characteristic of the first five of these orientations. For the sixth, the Ray (15) Protestant Ethic (PE) scale (15) was deemed sufficient, though in a slightly shortened form of only 14 items. A new asceticism scale was constructed and used even though there is a well-established existing scale centering on this concept, the Mirels and Garrett (13) scale. This was done to give some alternative to the rather pugnacious tone of the Mirels and Garrett items and because that scale has virtually no control against acquiescent response set. To aid comparability with previous research, however, the Mirels and Garrett scale was also used.
The battery containing these scales and demographic questions was administered to a sample of 141 people approached from a market research trailer located in various Sydney metropolitan shopping centers from time to time. Respondents filled out the questionnaire before they sampled and evaluated various products. Most were under 35 years of age and there was only a slight preponderance of women. A wide range of occupations was also encompassed.
The reliabilities ("alpha") of the scales were as follows: Asceticism (10 items) .57, Depravity (seven items) .64, Predestination (11 items) .69, Frugality (10 items) .77, Metaphysical belief (eight items) .84, Mirels and Garrett (19 items) .80, and Protestant moralism (14 items) .76.
Only two scales predicted all the other elements of the Protestant ethic: the Metaphysical Belief scale and the Ray (15) Protestant Moralism scale. The Mirels and Garrett scale predicted only Asceticism and Frugality. The Predestination scale predicted only Asceticism, Metaphysical Belief, and Moralism. The new Asceticism scale seemed superior to the Mirels and Garrett measure, since it predicted Predestination and Moralism in addition to those variables predicted by the Mirels and Garrett scale. The Belief in Depravity scale was disappointing. Whether or not its correlation with the other measures in the battery was significant, it was on all occasions negative in sign. At least in modern times, therefore, negativism or cynicism about human nature does not form part of the Protestant outlook. To the contrary, the Protestant-oriented person trusts and thinks well of man. He is generously disposed towards others rather than misanthropic.
Given the methodological and empirical deficiencies of the Mirels and Garrett scale and the limited thematic range of the Ray (15) scale, it seemed desirable at this point to use the large pool of items available to construct a new balanced scale with comprehensive coverage of traditional Protestant beliefs by combining all the items of all the scales previously used into one superscale and by calculating a total score on this superscale for each person in the sample. The correlation of each item in the individual scales with this total was then examined and the nine highest-correlating pro-Protestant and the nine highest correlating anti-Protestant items were selected. The 18 items of this new scale are given in the Appendix. Its internal reliability was .82 (alpha).
The correlations with demographic variables revealed two important further problems with the Mirels and Garrett scale. It was not significantly related to religious attendance, religious background, or present religious belief. In a scale purporting to measure the Protestant ethic, this seemed a fundamental defect in validity. The second problem was that it did correlate significantly (r = -.219) with education. This means that it predicts poor education better than anything else.
By contrast, the Ray (15) Protestant Moralism scale predicted church attendance (Chi squared = 21.08, df = 12, p < .05), though not religious background or present religious belief. It lacked the education contamination of the Mirels and Garrett scale; also, the correlation (r = .103) was in the opposite direction. This latter finding, taken in conjunction with another that the Ray scale correlated significantly (r = .181) with self-assigned social class, suggests that what was called Protestant Morality could nowadays perhaps be called middle-class morality.
The new eclectic (PE) scale was, moreover, markedly superior to either of the earlier scales. It predicted both church attendance and present religious belief at better than the <.01 level of significance. It was unrelated to education but nearly (r = .142) related to self-assigned social class. It was also significantly related to both the Mirels and Garrett scale (r = .360) and the Ray scale (r = .594). It is then clearly the most valid of the three.
An interesting feature in the prediction of religious belief was that the difference in Protestant ethic scores lay not between Protestants and Catholics but between "believers" and "unbelievers," (with Unbelievers being agnostics or atheists). This again tends to suggest that in modern-day Australian society, Catholics are as Protestant as Protestants themselves in most socially relevant respects.
E. STUDY IV
Since no scale so far had succeeded in discriminating Protestants from Catholics, it seemed that the net would have to be cast even wider in choosing constructs for examination. It was decided, therefore, to put together a set of short scales to measure the most popular constructs in current social psychological research. The chosen constructs were as follows: Internal/External control, Authoritarianism, Dogmatism, Alienation, Achievement motivation, Neuroticism, Machiavellianism, and Social Desirability (Approval motivation). The scales measuring the various constructs were, respectively; The Rotter (27) "I-E" scale, the Ray (21) "Directiveness" scale, the Ray (17) "BD" scale Mark II, the Ray (19) "GA" scale, the Ray (22) "AO" scale, the Eysenck (6) MPI, the Christie and Geis (2) "Mach IV," and the Crowne and Marlowe (4) Social Desirability scale.
Because of the large number of constructs involved, all scales had to be used in short forms only. Where short forms did not already exist, they were devised on the basis of available item analyses; i.e., an attempt was made to select only those items which correlated most highly with the total score on the full scale. All scales were administered in Likert format. The sample has been more fully described elsewhere [see Ray (23)]; 87 Sydney people were selected to provide satisfactory quotas relating to the basic demographic variables of age, sex, occupation, and education; they were indistinguishable from the distribution observed in contemporaneous random doorstep samples (e.g., Study II).
The coefficients "alpha" of the scales used were as follows: .57 for I-E, .60 for authoritarianism, .52 for Dogmatism, .72 for alienation, .72 for achievement motivation, .71 for Neuroticism, .54 for Machiavellianism, and .77 for Social Desirability. These are generally rather low, an inevitable consequence of the decision to use short scales of only eight to 12 items each.
The significant relationships again arose from differences between unbelievers and the rest: Unbelievers both by background and present conviction were high on Machiavellianism (Chi squared = 51.13 and 43.93, dfs = 24, p < .01). By present belief only, unbelievers were also low on authoritarianism (Chi squared = 47.88, df = 24). Evidently, unbelievers preferred to exercise social control subtly rather than directly.
F. STUDY V
Whatever else they had done, none of the studies so far had detected any attribute discriminating between Protestants and Catholics. It is, however, difficult to believe that the sort of differentiation described by Weber has entirely vanished from the modern world. Perhaps the differences are greatly attenuated, but surely there should still be some trace of them. Perhaps the tests used so far were simply not sensitive enough in the detection of weak effects.
The two most obvious ways of increasing the sensitivity of a test are to increase the reliability of the measuring instrument and to increase the sample size. At first glance, the first requirement did not seem difficult. There are many high-reliability tests of achievement motivation in existence [e.g., Jackson (9) and Mehrabian (12)]. On second glance, however, all these tests proved to have been constructed for student use only. Their reliability on representative population samples and in different societies is simply unknown. The Mehrabian test used in Study II had a reliability of .64; its original reliability on students was .91. Even very good student-normed tests may fail badly with general population samples.
Thus, in spite of the large number of achievement motivation tests available (7), it seemed necessary to construct a new test. The best approach under the circumstances seemed to be simply to combine items from existing tests that had been normed and validated on general population samples. Other things being equal, increasing the length of a test increases its reliability. There seemed, however, to be only four published tests designed and validated for general population samples: the Ray (15, 18, 20, 22, 24) "AO" test, the Ray (24) Success-Orientation and Task-Orientation tests, and the Ray (16) Task-Orientation test. As the first three tests have many items in common and are very similar in style, it was decided to combine their items. When the "AO" test was used in its 1979 short form, a new scale of 30 different items emerged: 15 proachievement and 15 antiachievement. This scale, then, was inserted into a questionnaire together with the Ray (16) Task-orientation scale and several other scales. The other scales were primarily designed to replicate somewhat the findings of Study III and comprised accordingly the new eclectic PE scale from that study plus scales to measure belief in Predestination, Frugality, and Metaphysical faith. Also included were two PE scales widely used elsewhere, the Blood (1) scale and the Mirels and Garrett (13) scale. With assistance from the item analyses from study IV, the Mirels and Garrett scale was shortened to the most reliable 12 items only.
The questionnaire was mailed on this occasion to 1000 people selected at random from the New South Wales electoral rolls. A total of 240 were returned. This n is sufficiently large to show statistically significant effects explaining as little as 1.5% of the variance. Again the sample was found to be representative in terms of age, sex, occupation, and income when compared with contemporaneous random doorstep studies carried out in the Sydney metropolitan area.
The reliability of the scales is as follows: New "AO" scale .85, 14-item "AO" scale .78, 1980 TO scale .76, 1980 SO scale .72, 1973 TO scale .80, Blood scale .40, Mirels and Garrett scale .75, eclectic PE scale .78, Frugality scale .69, Metaphysical faith scale .87, and Predestination .79.
For religious background, the main significant Chi squared observed (37.59, df = 24) was for the new "AO" scale. Unbelievers obtained higher scores than those with a denominational background and were, in other words, more achievement motivated. They also showed less metaphysical belief.
For present religious belief, there were many more significant relationships. None, however, reflected Catholic-Protestant differences. Unbelievers were found to score lower on the Blood, Mirels and Garrett, eclectic PE, Frugality, Predestination, and Metaphysical faith scales. They were, in a word, less influenced by the traditional social teachings of the churches.
A methodological finding of some interest concerned the Blood scale. It was the only scale to show no significant correlation between its positively and negatively keyed halves. Instead of the expected high negative correlation, the supposedly oppositeiy-worded halves correlated in fact only .079. This is very strong evidence of zero construct validity. If the items of supposedly opposite meaning are shown in fact not to correlate, what in fact do they mean? No one knows. Previous research using this scale must therefore be viewed as uninterpretable. The low reliability of this scale would also cast grave doubts on its usefulness.
When it employs human Ss at all, over 90% of psychological research appears to be done with available groups of American college students as Ss. This makes its generalizability to the population at large unknown. The present program of research has eschewed this approach in favor of the much greater labor of gathering Australian samples with at least some claim to representativeness. Needless to say, in no case was an ideal sample in fact achieved. Unlike the marbles in the statistician's barrel, people frequently refuse to be sampled. This applies no matter what sampling method is used. The present results can therefore be interpreted only with the most extreme caution.
What they show, however, is a noteworthy consistency. No attitudinal, personality, or motivational differences between Protestants and Catholics have been uncovered. If one is prepared to admit data based on student samples -- e.g., Ray and Doratis (26) -- one can even add that there are no Protestant/Catholic differences on ethnocentrism (racial exclusiveness) or religiocentrism (religious exclusiveness).
The generalization of these results to other countries must of course remain speculative. The prospects for generalization to the United States, however, should be remarkably good. The two countries are very similar in language, history, culture, race, ethnicity, urbanization, and standard of living. At roughly 25%, the proportion of Catholics in Australia is even very similar to that in the United States; they also have similar ethnic composition, predominantly Irish but with large numbers of Southern and Eastern European origin also.
The present results cannot of course invalidate the Weber thesis which was concerned with historical events now long past. What was true then need not be true now. The results do, however, serve the useful purpose of putting a somewhat more specific time-frame around the processes Weber described. We now have some reason to believe that the Protestant ethic as such has perhaps run its course.
This is not, of course, in any way to say that the spirit of enterprise or the worship of work is now dead. It is only to say that all religions have come to share these attributes to an equal degree. The Protestant ethic in substantive terms is certainly not yet dead; it is just no longer Protestant.
No doubt research devoted to examining specific doctrinal stances would still show some differences between the adherents of the various denominations. It hardly needs to be proven that Catholics respect the Pope and that Methodists tolerate divorce. What the present research has shown is that these doctrinal differences may not be accompanied by other psychological differences. Some such differences may yet emerge, but the range of possibilities considered in the present research program has surely eliminated most of the obvious candidates (and some non-obvious ones).
If there is a religiously-polarized "ethic" left in contemporary society, then, it would appear to be the "atheist ethic." Atheists are in a sense the ultimate Protestants; they reject all that the Catholic Church stands for. Even this ethic, however, does not in quantitative terms differ from the societal norms very much. The differences are more notable for their extensiveness than their intensity.
The nature of the ethic is still a little surprising. As one might expect, the rigors and strictures of the churches' social teachings are rejected but this does not at the same time lead to a general slide into apathy, indolence, and self-indulgence. The attitudes of unbelievers (including both atheists and agnostics) appear in fact to be more highly work- and achievement-motivated than denominational adherents. If these differences persist and if they were in fact what gave historical Protestants their economic advantages, we may perhaps look forward to a future of atheist elites and believing masses.
Eclectic Protestant ethic scale
1. Eat drink and be merry, for tomorrow we may be dead. R
2. Too much attention today is given to pleasures of the flesh.
3. There is some great plan for the affairs of men, the end of which no mortal eye can foresee.
4. If you've got it, why not spend it? R
5. You can't take it with you, so you might as well enjoy yourself. R
6. Saving always pays off in the end.
7. The only way to get anything worthwhile is to save for it.
8. I believe in God.
9. I believe in life after death.
10. Once you die, that's all there is. R
11. The spirit of God lives within every man.
12. Predestination is a myth. R
13. For girls to keep themselves virgins before they are married is old-fashioned and unnecessary. R
14. You should never speak lies about other people.
15. Stealing is all right as long as you don't get caught. R
16. There's nothing wrong about having sex with another man's wife. R
17. There is no such thing as absolute right or wrong. R
18. If one works hard enough, he is likely to make a good life for himself.
Note: On the norming sample the mean score was 62.38 with an SD of 10.24. Five response categories are allowed from "Strongly Agree" to "Strongly Disagree" with a midpoint "Not sure" always scored "3." If an item is marked "R" above, it is scored "1" for "Strongly Agree." If it is not so marked, it is scored "5" for the same response.
1. BLOOD, M. R. Work values and job satisfaction. J. Appl. Psychol., 1969, 53, 456-459.
2. CHRISTIE, R., & GEIS, F. L. Studies in Machiavellianism. New York: Academic, 1970.
3. COSTELLO, C. A. Two scales to measure achievement motivation. J. of Psychol., 1967, 66, 231-235.
4. CROWNE, D. P., & MARLOWE, D. The Approval Motive. New York: Wiley, 1964.
5. ENTWISLE, D. R. To dispel fantasies about fantasy-based measures of achievement motivation. Psychol. Bull., 1972, 77, 377-391.
6. EYSENCK, H. J. Manual of the Maudsley Personality Inventory London: Univ. London Press, 1959.
7. FINEMAN, S. The achievement motivation construct and its measurement: Where are we now? Brit. J. Psychol. 1977, 68, 1-22.
8. FRENCH, E. Development of a measure of complex motivation. In J. W. Atkinson (Ed.), Motives in Fantasy, Action and Society Princeton, N.J.: Van Nostrand, 1958.
9. JACKSON, D. N. Personality Research Form Manual New York: Research Psychologists Press, 1967.
10. McCLELLAND, D. C. The Achieving Society Princeton, N.J.: Van Nostrand, 1961.
11. MCREYNOLDS, P., & GUEVARA, C. Attitudes of schizophrenics and normals towards success and failure. J. Abn. Soc. Psychol., 1967, 72, 303-310.
12. MEHRABIAN, A. Male and female scales of the tendency to achieve. Educ. & Psychol. Meas., 1968, 28, 493-502.
13. MIRELS, H. I,., & GARRETT, J. B. The Protestant Ethic as a personality variable. J. Consult. & Clin. Psychol, 1971, 36, 40-44.
14. MUKHERJEE, B. N. A forced-choice test of achievement motivation. J. Indian Acad. Appl. Psychol., 1965, 2, 85-92.
15. RAY, J. J. Christianism . . . . The Protestant ethic among unbelievers. J. Christian Educ., 1970, 13, 169-176.
16. RAY, J.J. (1973) Task orientation and interaction orientation scales. Personnel Psychology 26, 61-73.
17. RAY, J.J. (1974) Balanced Dogmatism scales. Australian Journal of Psychology 26, 9-14.
18. RAY, J.J. (1974) Are trait self-ratings as valid as multi-item scales? A study of achievement motivation. Australian Psychologist
19. RAY, J.J. (1974) Who are the alienated? Ch. 52 in Ray, J.J. (Ed.) Conservatism as heresy Sydney: A.N.Z. Book Co.
20. RAY, J.J. (1975) A behavior inventory to measure achievement motivation. J. Social Psychology 95, 135-136.
21. RAY, J.J. (1976) Do authoritarians hold authoritarian attitudes? Human Relations, 29, 307-325.
22. RAY, J.J. (1979) A quick measure of achievement motivation -- validated in Australia and reliable in Britain and South Africa. Australian Psychologist 14, 337-344.
23. RAY, J.J. (1979) The authoritarian as measured by a personality scale Solid citizen or misfit? J. Clinical Psychology 35, 744-746.
24. RAY, J.J. (1980) The comparative validity of Likert, projective and forced-choice indices of achievement motivation. J. Social Psychology, 111, 63-72.
25. RAY, J. J. Measuring achievement motivation by immediate emotional reactions. J. Soc. Psychol., 1981, 113, 85-93.
26. RAY, J.J. & DORATIS, D. (1972) Religiocentrism and ethnocentrism: Catholic and Protestant in Australian schools. Sociological Analysis 32, 170-179.
27. ROTTER, J. B. Generalized expectancies for internal versus external control of reinforcement. Psychol. Monog., 1966, 80, No. 609.
28. VEROFF, J., FELD, S., & GURIN, G. Achievement motivation and religious background. Amer. Sociolog. Rev., 1962, 27, 205-217.
29. WEBER, M. The Protestant Ethic and the Spirit of Capitalism (trans. T. Parsons) London: Unwin, 1930.
Acknowledgement: I would like to thank Mr. Rick Wilson and Probe Pty. Ltd. for making Study III possible.
Replication is one of the cornerstones of science. A new research result will normally require replication by later researchers before the truth and accuracy of the observation concerned is generally accepted. If a result is to be replicated, however, careful specification of the original research procedure is important.
In questionnaire research it has been my observation that the results are fairly robust as to questionnaire format. It is the content of the question that matters rather than how the question is presented (But see here and here). It is nonetheless obviously desirable for an attempted replication to follow the original procedure as closely as possible so I have given here samples of how I presented my questionnaires in most of the research I did. On all occasions, respondents were asked to circle a number to indicate their response.
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)