Tuesday, February 13, 2007



POLITICS: ONE DIMENSION OR TWO?




By John J. Ray (M.A.;Ph.D.)


Describing the entire domain of political attitudes in terms of a single Right/Left dimension does have its problems. Libertarians in particular are always promoting a two dimensional description of political attitudes. The political compass and the "World's Smallest Political Quiz" are popular applications of such theories. Below are two examples of such a description:





Unfortunately, though I am a libertarian, I am also an academic who has specialized in surveys of political attitudes and I regret to report that we seem to be pretty stuck with the old Left/Right schema. So let me give you a quick tour of the academic literature:

Academics have been talking about this for a long time. Various authors (e.g. Eysenck, 1954; Rokeach, 1960; Kerlinger, 1967) have proposed that an adequate description of world politics really needs two dimensions. They propose, for example, that the Left-Right dimension be supplemented by an Authoritarian/Permissive dimension. So that democratic Leftists and Rightists are Permissive Leftists and Rightists whereas Communists and Fascists are Authoritarian Leftists and Rightists.

Although such proposals have considerable intuitive appeal, they do not, unfortunately, seem to coincide with how people's attitudes are in fact organized when we do surveys of public opinion. It is very easy to find people's attitudes polarizing on a Left/Right dimension but nobody has yet managed to show in a satisfactory way any polarization of attitudes on the postulated second dimension (Ray, 1980 & 1982 -- online here and here).

The summary from my 1982 paper published in The Journal of Social Psychology reads as follows:

The Eysenck/Rokeach/Kerlinger theory that social attitudes are two dimensional suffers from disagreement about what the second dimension should be called and how it should be measured. The present work tests the proposal that there is a dimension of libertarian/authoritarian attitudes orthogonal to radicalism/conservatism. A set of items designed to maximize the likelihood of such dimensions appearing was administered to a random postal sample of Californians. No real evidence of the proposed second dimension appeared. It was concluded that authoritarianism is a personality variable only.


"Orthogonal" is statistician-speak for "unrelated" or "at right-angles to". The correlations observed in the data showed that people who agreed with one conservatism statement, no matter how it was conceived, tended to agree with all other conservatism statements. Similarly for expresions of Leftist views. But "authoritarian" or "libertarian" statements did not cluster together at all. Your being a libertarian on one issue did not mean that you would tend to be libertarian on other issues. So it is solely the Left/Right dimension on which people in fact polarize.

The account of Left/Right attitudes given in my monograph suggests why this is so. For a start, the assumption that Fascists or Nazis are Right-wing is false. Hitler himself energetically claimed to be a socialist and Mussolini (the founder of Fascism) was a lifelong Marxist. The evidence for this has been summarized at great length elsewhere so will not be elaborated here.

Historically, the core of conservatism has always been a suspicion of government power and intervention and conservatives therefore accept only the minimum amount of government that seems needed for a civil society to function. So it is no wonder that there is no authoritarian version of conservative ideology. If it were authoritarian it could not be conservative.

Leftism, on the other hand, IS intrinsically authoritarian and power-loving and will always therefore tend in the direction of government domination. It is only non-authoritarian to the extent that is thwarted by external influences (such as democracy) from achieving its aims. Leftists in democratic societies do of course commonly deny authoritarian motivations but that is just part of their "cover". Deeds speak louder than words.

It may be worth noting that one researcher (Goertzel (1987) did manage to derive two personality measures that seem to correspond fairly closely with the usual proposals for what underlies political orientation. He called his two measures "tendermindedness-toughness" and "intuitiveness-consistency" and the latter was certainly very reminiscent of conservatism. So how did these two personality measures correlate with a range of political beliefs? Negligibly in all cases -- see his Table 4. Once again we see that a two-dimensional account of what underlies political orientation just does not work out empirically.


References:

Eysenck, H.J. (1954) The psychology of politics. London: Routledge
Goertzel, T. G. (1987) Authoritarianism of personality and political attitudes. J. Social Psychology, 127 (1), 7-18.
Kerlinger, F. N. (1967). Social attitudes and their criterial referents: A structural theory. Psychological Review, 74, 110-122.
Rokeach, M. (1960) The open and closed mind. N.Y.: Basic Books.






A TECHNICAL CAVEAT

As conservatives characteristically know and as Leftists characteristically deny, reality tends to be peskily complex -- and the present subject is no exception. I wish therefore to point out now some of the complexities not mentioned above.

The dimensionality of political attitudes was a topic of great interest to me throughout my research career and many of my published papers bear on it. I attempted an answer to the dimensionality question in the second paper I ever wrote, in fact -- written in 1968 and published in 1973. So I hope I can be forgiven if I lapse somewhat into technicalities in what follows:

The summary of what I want to say is this: While two robust orthogonal dimensions cannot be extracted from normal political discourse, two strong oblique dimensions can routinely be extracted.

To elaborate: There is a way to divide up normal political discourse into two dimensions but those two dimensions tend to be correlated. They are not independent. Knowing a person's position on one dimension will give you a weak prediction of his/her position on the second dimension. Ferguson in fact published such a solution in the 1940s, though he forced orthogonality (independence) on his solution.

And the dimensions identified by Ferguson can always be found and they are in fact rather familiar. They are the religious/moral dimension and the economic dimension. In other words, it is perfectly possible for people to be conservative on religious and moral matters but at the same time to be socialistically inclined on economic matters. For an extended discussion of just that combination, see here. There IS a religious Left who favour all sorts of government intervention in economic matters. As Waldman says: "Actually, in 2000, at least 10 million white "evangelical Christians" voted for Gore".

Some attitude combinations are more common than others, however, and it will be no news to anybody to hear that the PREPONDERANCE of religious conservatives also support conservative (non-interventionist) economic policies.

What is surprising, in fact, is how weakly the two attitude clusters are associated. In one of my 1973 papers, I found, for instance, a correlation of only .24 between economic conservatism and social conservatism. It was this weakness of association that enabled Ferguson to force an orthogonal solution (i.e. a picture of independence) on his data.

It would appear, however, that this weakness of association stems from areas of discourse that are less central to overall political orientation -- as, in a 1984 paper, I found that moral issues lay at the centre of all conservatism/Leftism issues. See the second Appendix in the paper. It is no wonder, then, that religious conservatives figure so prominently in America's Grand Old Party at the beginning of the 21st century. It shows, perhaps unsurprisingly, that the most personal issues are also the most important in politics.

In summary, then, religious/moral conservatism and economic conservatism are distinct but related -- and the more important the issue is the more highly they become related. It is perhaps encouraging that psychological research from many years previously in Australia predicted so well what was happening in American politics in the early 21st century.

Reference:

Ferguson, L. (1941) The stability of the primary social attitudes. J. Psychology, 12, 283-288.


FINIS