Worldwide Attitudes

ISSN 1323-9589

Volume 19960114, pages 1-8

Date: 14 January 1996

© Copyright Jonathan Kelley 1995. All rights reserved.

Attitudes towards Political Leaders in Australia, 1995

Jonathan Kelley

Political leaders both symbolise their respective parties and shape their parties’ current policies, so the citizenry’s views about them encapsulate the current state of politics, reflecting the combined effect of both long-term and short-term political factors (Kelley 1995; Bean and Mughan 1989; Marks 1993b). The long-term forces are enduring party loyalties -- those who like the Labor Party tend very strongly to like its leader just because he is the leader, while those who dislike the Labor Party tend to dislike its leader for the same reason, and similarly for the Liberal-National Coalition. The short-term forces are the particular policies and campaign issues the leader has emphasised in recent years, and his own personal image and views. This paper focuses on Australians’ ratings of the leaders of the two main parties: the Labor Prime Minister, Paul Keating, and the Leader of the Opposition, John Howard.

DATA

This analysis is based on a simple random sample of 2,223 Australian citizens. It is a nationwide, representative sample (Kelley, Bean, Evans, and Zagorski 1996; details in Appendix).

RATINGS OF THE LEADERS

We measured attitudes toward leaders using a standard Michigan ‘feeling thermometer’ appropriately labelled (Kelley, Bean, Evans & Zagorski 1995: 8, q1a&b).

"You probably feel favourable about some people or groups and unfavourable about others.  Please show your feelings with this 'feeling thermometer'-- use any number from 0 to 100 for a rating.  100 is the highest, for people you feel very warm or favourable about, and 0 is the lowest, for people you feel very strongly against. If you are neutral, neither for nor against, give a rating of 50. 

1. First, some Australian politicians. How do you feel about..."

 

a. Paul Keating?. . . . . . . . . . . . . . . . . ________

 

b. John Howard? . . . . . . . . . . . . . . . . . ________

 

The answers (grouped for clarity) were:

                      Keating     Howard

Very cool: 0-29          37         16  

Cool: 30-45              19         20  

Neutral: 46-54            9         14  

Warm: 55-70              19         31  

Very warm: 71-100        17         18  

                        ----       ----

   Total                100%       100% 

   Mean                  40         52  

   Cases                2301      2304

 

There is great diversity of opinion about both political leaders, with many Australians very cool toward them and many others very warm. This sharp polarisation about leaders is normal (Kelley 1995: Table 2), reflecting the adversarial ‘us versus them’ nature of democratic politics. 37% were ‘very cool’ to Keating and a further 19% ‘cool’. Only 9% were neutral. On the more sympathetic side, 19% were ‘warm’ and 17% ‘very warm’. The average was 40 points out of 100, well on the cool side of neutral -- indeed, Keating has long been one of the most unpopular Australian politicians (Bean and Kelley 1989).

Far fewer were ‘very cool’ to Howard: 16%, with a further 20% ‘cool’. 14% were neutral. On the sympathetic side, 31% were ‘warm’ and 18% ‘very warm’. The average was 52 points out of 100, just at the neutral point. This is hardly a ringing endorsement -- the most popular Australian politician of recent times was the previous Labor Prime Minister, Bob Hawke, who rated in the 60s in his heyday (Kelley 1995: Figure 1). But at the time these data were collected, toward the middle of 1995, Howard was clearly more popular than Keating, a pattern that has continued, with some ups and downs, to the present.

Influences on Leader Ratings: The ‘Usual Suspects’

Before looking specifically at the effect of ethnic feelings, it is useful to set out the broader context. Table 1 gives a measure of the combined impact of various groups of variables measured by their sheaf coefficient -- a standardised measure that ranges from 0 (no effect) through 1.0 (for a factor that explains everything). Details on method and models are given in the technical appendix; appendix table A gives disaggregated results for the individual variables. We begin with the most familiar element in Australian politics: class.

• In the past, class differences in politics were large, with the working class supporting the left and the upper classes supporting parties of the right, apparently for reasons of straightforward self- interest (eg Lipset 1981). This is especially true in the Anglo-American democracies (Kelley and Evans 1995). But class differences in Australian politics are now small. By 1995, class variables -- education, occupation, and income -- have a combined effect of only .06 on ratings of Keating and .09 on ratings of Howard -- far smaller than many other factors (table 1, lines 9 and 19). This is consistent with many other studies showing that class is no longer very important in most Western democracies, although it was in the past (for Australia: Marks 1993: 139-140; Bean and Kelley 1988a; for other nations: Kelley and Evans 1995; Kelley, McAllister and Mughan 1985). Nor is intergenerational class mobility important in Australia or other nations (Kelley 1992; Turner 1992).

• Background and demographic differences -- specifically age, sex, urban versus rural residence, labor force participation, and union membership -- matter more (table 1, lines 8 and 18). They have a combined effects of .12 on views about Prime Minister Keating. This is due to his greater popularity among union members -- usual for a politician on the Labor side of politics -- and to his greater popularity among men rather than women (appendix table A, panel 2). Background and demographic differences in evaluations of the Leader of the Opposition, John Howard, are even greater, with a combined effect of fully .30. This mainly reflects Howard’s far greater popularity among older rather than younger voters (appendix table A, panel 3). He is also somewhat less popular among union members, as usual for politicians from the conservative side of politics. Some of these differences may reflect self-interest (for example, union members expecting to benefit from pro-union policies of a Labor government) but the age differences probably do not (there are no major policy differences on age pensions or the like).

• Religion matters a little, about as much as class (table 1, lines 10 and 20). Those who believe in God, heaven, the devil, hell and an afterlife are a little less keen on the Labor Prime Minister (.06) and a little more sympathetic to his conservative opponent (.09). Modest religious differences are familiar in Australia and other contemporary Anglo-Celtic democracies and even stronger effects are found in many contemporary Central European democracies (Bean 1995; Kelley and Evans 1995: 169, 173). And in past generations, religious differences were important in many nations (Franklin, Mackie and Valen 1992).

• Ideology is the dominant cleavage in Australian politics, with a massive effect of .44 on evaluation of Keating and .42 on evaluation of Howard (table 1, lines 7 and 17). The most important single element is attitudes toward trade unions (appendix table A, panels 2 and 3), and this has been true for many years (Kelley 1988; Kelley and McAllister 1988; Bean and Kelley 1988a). Those who give unions high ratings on a feeling thermometer, who think they "do a good job for the nation as a whole", whose sympathies are usually for strikers, and who think that unions have too little power in Australia are much warmer towards Keating, and much less sympathetic to Howard.

• Other ideological issues are also important. Those who favour government regulation of medical care are also noticeably more favourable to Keating and less to Howard. Those sympathetic to big business are noticeably more favourable to Howard, although not especially cool to Keating. The same is true of those who oppose government regulation of the economy: They tend to like Howard but neither to like nor dislike Keating. The role of ethnic feelings in assessments of political leaders is discussed in detail in the next issue of WwA (Kelley 1996, in press).

The dominant role of ideology in Australian politics is consistent with a Hegelian emphasis on the importance of ideas. By contrast, it is difficult to reconcile with the rational choice model of human behaviour with its emphasis on individual self-interest. Self-interest varies with people’s objective class and demographic position, not with their ideas about unions, big business, or government regulation. Nor can this be explained away on the argument that objective class and demographic position shape people’s ideology, and so indirectly shape their politics. In my estimates those indirect effects are already attributed to class, not to ideology (see the technical appendix).

REFERENCES

Bean, Clive S. 1991. "Comparison of National Social Science Survey Data with the 1986 Census." National Social Science Survey Report 2(6):12-19. [ISSN 1031-4067]

Bean, Clive S. 1995. "Religion: The Forgotton Cleavage". Paper presented to the annual conference of the Australian Political Studies Association, 1995.

Bean, Clive S. and Jonathan Kelley. 1988a. "Partisan stability and short-term change in the 1987 federal election: Evidence from the NSSS panel survey" Politics 23 (1988): 80-94.

Bean, Clive S. and Jonathan Kelley. 1988b. "The Comings and Goings of Political Leaders" National Social Science Survey Report 1(2):12-19. [ISSN 1031-4067]

Bean, Clive S. and Jonathan Kelley. 1989. "Keating would Cost ALP Votes" National Social Science Survey Report 1(4):2-3. [ISSN 1031-4067]

Bean, Clive S. and Anthony Mughan. 1989. "Leadership Effects in Parliamentary Elections in Australia and Britain" American Political Science Review 83: 1166-1179.

Franklin, Mark, Tom Mackie, and Henry Valen (eds.) 1992. Electoral Change: Responses to Evolving Social and Attitudinal Structures in Western Countries. Cambridge: Cambridge University Press.

Kelley, Jonathan. 1988. "Political Ideology in Australia." Pp. 58-80 in Australian Attitudes, edited by J. Kelley and C. S. Bean. Sydney: Allen & Unwin.

Kelley, Jonathan. 1992. "Social Mobility and Politics in the Anglo-American Democracies." Pp 21-49 in Social Mobility and Political Attitudes: Comparative Perspectives (Frederick C. Turner, ed.) New Brunswick: Transaction Press.

Kelley, Jonathan. 1995. "Public Images of Foreign and Domestic Political Leaders: Australian Evidence". International Social Science Journal 146 (Dec): 601-614.

Kelley, Jonathan and Ian McAllister. 1985. "Class and party in Australia: Comparison with Britain and the USA." British Journal of Sociology 36:383-419.

Kelley, Jonathan and M.D.R. Evans. 1995. "Class and Class Conflict in Six Western Nations." ASR 60 (Apr): 157-178.

Kelley, Jonathan, Clive S. Bean, M.D.R. Evans and Krzysztof Zagorski. 1996. Australia, 1995: International Social Science Survey. Codebook and Machine Readable Data File (Preliminary). Canberra: International Social Science Survey, Institute of Advanced Studies, Australian National University.

Kelley, Jonathan, Clive S. Bean, M.D.R. Evans and Krzysztof Zagorski. 1995. Questionnaire: International Social Science Survey/ Australia 1995. Canberra: International Social Science Survey, Institute of Advanced Studies, Australian National University.

Kelley, Jonathan, Ian McAllister and Anthony Mughan. 1985 "The decline of class revisited: Class and party in England, 1964-1979". American Political Science Review 79 (Sept): 719-737.

Marks, Gary N. 1993a. "Intra- and Extra-familial Political Socialization: The Australian Case and Changes over Time, 1967-1990. Electoral Studies 12: 128-157.

Marks, Gary N. 1993b. "Partisanship and the Vote in Australia: 1967-1990. Political Behaviour 15: 137-166.

Turner, Frederick C. (ed.) 1992. Social Mobility and Political Attitudes: Comparative Perspectives. New Brunswick (USA): Transaction Publishers.

Tables

Table 1. Determinants of attitudes toward Keating 

and Howard. Sheaf coefficients and OLS standardized

partial regression coefficients. Total effects. 

N= 2301. Australia, 1995.[a]

--------------------------------------------------------

                                          Sheaf     t

                                          or OLS   (OLS)

--------------------------------------------------------

Panel A. KEATING

-----------------

1. Ethnicity (sheaf)                       .23

  2. Britain: Sympathy                       -.15  -7.33

  3. Asian migrants: Sympathy                 .15   5.77

  4. Southern European migrants: Sympathy      ns

  Respondent, father or mother born: 

    5. Britain, NZ, USA, etc.                -.06  -3.41

    6. Other non-Australian                   .06   3.18

7. Economic attitudes (sheaf)              .44

8. Demography (sheaf)                      .12

9. Class (sheaf)                           .06

10. Religion (sheaf)                       .06

Panel B. HOWARD

---------------

11. Ethnicity (sheaf)                      .23     

  12. Britain: Sympathy                       .19   9.35

  13. Asian migrants: Sympathy               -.10  -3.99

  14. Southern European migrants: Sympathy    .10   3.99

  15. Respondent, father or mother born:

    16. Britain, NZ, USA, etc.                 ns

    17. Other non-Australian                   ns

18. Economic attitudes (sheaf)             .42

19. Demography (sheaf)                     .30

20. Class (sheaf)                          .09

21. Religion (sheaf)                       .09

--------------------------------------------------------

ns -- Not significantly different from zero, p<.05, 

      two-tailed.

[a] See appendix 1 for the model and appendix table 1 

    for details.

Appendix Table A. Determinants of attitudes toward Keating

and Howard. OLS regression coefficients giving total effects for

individual variables and for sheaf coefficients for groups of

variables. N= 2301. Australia, 1995.

-----------------------------------------------------------------

Panel 1. Means and standard deviations

--------------------------------------

              Mean     St.Dev   Label

Dependent variables:

  tKEATING    40.33    31.39  Thermometer: Keating

  tHOWARD     51.83    26.00  Thermometer: Howard

Ethnicity:

  ANGLO2        .21      .41  R/Mo/Fa Born: GB, NZ, USA, etc

  OTHER2        .21      .41  R/Mo/Fa Born: NW Eur, E Eur, Asia 

  SEURMIG     59.57    19.23  Symp migrants: Greek, Ital (therm)

  ASIAMIG     47.14    21.34  Symp migrants: Asian (thermom)

  tGB         55.12    24.48  Symp: Britain (thermom)

Demography:

  AGE         49.03    15.82  Age (years)

  MALE          .49      .50  Male (=1)

  URBANNOW    70.22    31.04  Urban resident now

  INLF          .64      .48  In labor force

  TUMEMB        .23      .42  Union member

Class:

  EDUC        11.89     3.01  Education (years)

  BLUEC         .28      .45  Occupation now: Blue collar

  FARM          .04      .20  Occupation now: Farm

  FAMINC      47.86    42.19  Family income ($A 000's)

Relgion:

  CATHNOW       .20      .40  Relig now: Catholic

  LNCHATT       .79     1.81  Church attendance: ln(#days)

  CBELIEF     54.89    29.13  Christian belief 

Economic attitudes:

  PROUNION    41.03    18.23  Pro-trade union

  PROBZ       49.89    16.43  Pro Oz business (thermom)

  GOVECON     36.69    13.41  Govt own/regl the economy

  GOVMEDIC    54.63    19.01  Govt own/regl medical care

Controls:

 PNTPARTY      .48      .40   Parents political party (0=Labor)

 TPARTIES    40.84    19.83   Political parties generally 

Panel 2. KEATING

-----------------

Variable           B        SE B       Beta         T  Sig T

Ethnicity (sheaf)                       .23     

  TGB           -.19         .03       -.15     -7.33    .00

  ASIAMIG        .22         .04        .15      5.77    .00

  ANGLO2       -4.95        1.45       -.06     -3.41    .00

  OTHER2        4.53        1.42        .06      3.18    .00

  SEURMIG         ns

Economic attitudes (sheaf)              .44     

  PROUNION       .65         .04        .38     17.66    .00

  GOVMEDIC       .17         .04        .10      4.74    .00

  PROBZ           ns

  GOVECON         ns

Demography (sheaf)                      .12      

  TUMEMB        6.47        1.79        .09      3.60    .00

  MALE          2.90        1.44        .05      2.01    .04

  AGE             ns

  URBANNOW        ns

  INLF            ns

Class (sheaf)                           .06      

  EDUC            ns

  BLUEC           ns

  FARM            ns

  FAMINC          ns

Religion (sheaf)                        .06      

  CBELIEF       -.07         .03       -.07     -2.88    .00

  CATHNOW         ns

  LNCHATT         ns

Controls

  PNTPARTY    -20.24        1.80       -.26    -11.23    .00

  TPARTIES       .34         .03        .21     10.16    .00

Panel 3. HOWARD

---------------

Variable           B        SE B       Beta         T  Sig T

Ethnicity (sheaf)                       .23     

  TGB            .20         .02        .19      9.35    .00

  SEURMIG        .14         .04        .10      3.99    .00

  ASIAMIG       -.13         .03       -.10     -3.99    .00

  ANGLO2          ns

  OTHER2          ns

Economic attitudes (sheaf)              .42     

  PROUNION      -.46         .03       -.32    -14.68    .00

  PROBZ          .19         .03        .12      5.72    .00

  GOVECON       -.15         .04       -.07     -3.35    .00

  GOVMEDIC      -.12         .03       -.09     -4.18    .00

Demography (sheaf)                      .30     

  AGE            .47         .04        .28     11.11    .00

  TUMEMB       -7.57        1.44       -.12     -5.27    .00

  INLF          3.63        1.44        .07      2.53    .01

  MALE            ns

  URBANNOW        ns

Class (sheaf)                           .09      

  FAMINC         .04         .01        .07      2.85    .00

  EDUC            ns

  BLUEC           ns

  FARM            ns

(Appendix Table 1, continued)

Religion (sheaf)                        .09      

  CBELIEF        .06         .02        .07      3.08    .00

  CATHNOW         ns

  LNCHATT         ns

Controls

  PNTPARTY     14.75        1.44        .23     10.23    .00

  TPARTIES       .21         .03        .16      7.58    .00

---------------------------------------------------------------

ns -- Not significantly different from zero, p<.05, two-tailed.

APPENDIX: Sheaf Coefficients

The sheaf coefficients presented in text summarize the combined effect of a group of variables (Heise xx19xx; Kelley and McAllister 1985).The variables are combined additively using weights derived from an ordinary least squares regression. For example the sheaf for class variables’ influence on ratings of Paul Keating stems from a regression (variables are defined in the appendix table):

 KEATING1 =  EDUC + BLUEC + FARM + FAMINC                (eq. 1)

             + AGE + MALE + URBANNOW + INLF + TUMEMB 

             + PNTPARTY + constant

The class variables are those on the first line. Their combined influence is expressed in a single sheaf variable, combining the four class variables into a weighted sum:

   Sheaf1 = .44*EDUC + 1.40*BLUEC -3.96*FARM -.04*FAMINC   (eq. 2)

where the weights are the metric regression coefficients from Equation 1. This sheaf coefficient (ClassSheaf) is then used in a second OLS regression:

   KEATING2 = ClassSheaf                                   (eq. 3)

             + AGE + MALE + URBANNOW + INLF + TUMEMB 

             + PNTPARTY + constant

Equation 3 is mathematically equivalent to Equation 1: It yields idential predicted values, differing only in combining the class variables into a single composite. As in all regression equations, the standardized regression coefficients express the relative influence or importance of each variable. The standardized regression coefficient for the sheaf variables thus gives the relative influence of all aspects of class compared to other variables; it is the 'sheaf coefficient' reported the text. A sheaf coefficient for demographic variables (viz AGE, MALE,URBANNOW, INLF and TUMEMB) is computed analogously.

Equations 1 and 3 give the total influence of class, including a part that comes about indirectly because class position influences other variables (for example, attitudes toward trade unions) which in turn influence attitudes to Keating (Alwin and Hauser classic reference).

The sheaf coefficient giving the combined effect of ethnic feelings is derived similarly, but from the regression equation giving their effect controlling class, demography, and parents’ party (as in Equation 1) and also controlling religion, economic attitudes, and other intervening variables:

  KEATING3 = ANGLO2 + OTHER2 + SEURMIG + ASIAMIG + TGB    (Eq. 4)

             + EDUC + BLUEC + FARM + FAMINC                

             + AGE + MALE + URBANNOW + INLF + TUMEMB 

             + PNTPARTY 

             + CATHNOW + LNCHATT + CBELIEF 

             + PROUNION + PROBZ + GOVECON + GOVMEDIC 

             + TPARTIES + constant

The ethnic variables are those on the first line of Equation 1. Equation 4 is then re-estimated replacing the ethnic variables with the appropriate sheaf variable.

SURVEY INFORMATION ON THE ISSS/A 1995.

The International Social Science Survey / Australia is a nation-wide survey conducted by researchers at the Australian National University and the University of Melbourne. Begun in 1984, it is Australia’s leading academic survey and is devoted entirely to academic research in the social sciences, is non-profit, and is not connected with any business or political party. The survey’s core sponsor is the Research School of Social Sciences at ANU. The ISSS is based on large, representative national samples of all states and territories, drawn from the electoral roll. The detailed and comprehensive survey takes about two hours to complete. It is conducted by mail. The first mailing includes a cover letter from the Australian National University and a postage-paid reply envelope, followed by a further letter about two weeks later. Anyone who did not respond within a month or so is then pursued by up to three more mailings over a six month period. Comparison with the census shows the samples collected in this way to be representative of the Australian population in age, sex, education, occupation, and other characteristics. Dr. Jonathan Kelley is Director and principal investigator of the ISSS; Dr. Clive Bean (Associate Director), Dr. M.D.R. Evans and Dr. Krzysztof Zagorski are co-principal investigators.

The results presented here are from a preliminary sample from the 1995 survey, which was conducted in April through December. The preliminary sample has the first 2,223 cases (the final sample will have a few hundred cases more). Past ISSS/A surveys which have had completion rates around 60% to 65%; closely match the population in all characteristics that can be compared to the census (Bean 1991); and show no systematic difference between the last few hundred cases and the rest of the sample. There is no reason to expect the present survey to differ from previous ones in these respects.

Visitors since 14 September 1998: