Is There an Ethnic Electorate Effect on Representation? Evidence from the 1993 Australian Candidate Study


Research Paper 2 1997-98

Dr Gianni Zappalà
Politics and Public Administration Group
20 October 1997

Contents

Major Issues Summary

Introduction

Data and method

Activity in an ethnic organisation

Ethnicity of the electorate and representatives' attitudes and behaviour towards constituency work

Candidate beliefs with respect to constituency work
Member behaviour with respect to constituency work

Time spent on constituents' problems
Attendance at local community functions

Candidate attitudes to ethnic related issues

Conclusion

Endnotes

References

Major Issues Summary

Recent studies have suggested that the ethnic composition of the electorate which Members of Parliament (MPs) represent may an be an important factor in determining representatives' attitudes and behaviour. For instance, 'ethnic electorates' in Australia also tend to be lower than non-ethnic electorates in socio-economic terms (e.g. income and occupation), and constituents from such electorates often place more demands on their MPs because they face greater problems with issues such as unemployment or social security. This may mean that MPs have to adopt a constituency focus whether they wish to or not. Studies have also shown that in assessing their MPs, ethnic minorities tend to place greater importance on their MPs' approach to constituency work than their approach to particular policy issues or how they vote in Parliament. In the Australian context, for instance, a great deal of representatives' time in ethnic electorates is spent dealing with the immigration problems of their ethnic constituents. In brief, it is reasonable to assume that candidates or Members from 'ethnic electorates' may firstly place greater importance on the constituency face of representation (as opposed to the parliamentary face), and secondly, may spend more time on constituency related activities.

This paper examines whether there was an 'ethnic electorate effect' on candidate attitudes and behaviour in Australia by analysing responses to the 1993 Australian Candidate Study (ACS). It does this by examining whether candidates' responses on several relevant issues varied according to whether they stood for, or were an incumbent Member in, one of three broad 'electorate types': ethnic electorates, non-ethnic urban electorates, or (non-ethnic) rural electorates.

Four main aspects of representational behaviour are explored. First, the degree to which candidates were active in ethnic organisations. Second, candidates' attitudes towards aspects of constituency work according to the type of electorate they stood for in 1993. Third, the actual behaviour of incumbent Members with respect to constituency work (e.g. time spent on constituents' problems and attending functions) according to the type of electorate they represented. Finally, the paper also examines whether the ethnicity of the electorate influenced candidates' opinions on a range of ethnic related issues (e.g. immigration, equal opportunities for immigrants).

The findings suggest that, in general, there did not appear to be differences in attitudes and behaviour which could relate to the type of electorate (i.e. the extent of its ethnic composition). However, when party affiliation by type of electorate was examined, there did appear to be differences. The findings suggest that ALP candidates within ethnic electorates behaved and thought in ways which would appear to be more consistent with the ethnic composition of their electorates (e.g. they spent proportionately more time on dealing with constituents problems and attending functions than did their Coalition counterparts). This is not surprising. It is consistent, for instance, with previous work which has suggested that the ALP has been better at working ethnic electorates and consequently attracting the ethnic vote (at least up until 1993). What is perhaps surprising is that ALP Members and candidates in ethnic electorates continued to work hard in spite of the fact that the bulk of ethnic electorates are also safe ALP seats.

Introduction

Studies of the relationship between ethnicity and representation in Australia have concentrated on whether the composition of representative bodies such as the Parliament, reflect the composition of the wider population (e.g. in terms of gender, ethnicity, age).(1) The debate has therefore focused on the influence of the ethnicity (or gender) of the elected representative. Two recent studies, however, have argued that the ethnic composition of the electorate which Members represent, may at times influence representative attitudes and behaviour more than the ethnic background of a particular Member.(2) This paper seeks to further examine this proposition by examining a range of responses of candidates and Members who stood for the 1993 Federal election. In brief, the assumption is that candidates and Members from 'ethnic electorates' may have different attitudes and behaviour than those from non-ethnic electorates on some aspects of representational activity.(3)

Data and method

The 1993 Australian Candidate Study (ACS) was the third in a series of studies on Federal election candidates initiated in 1987.(4) This series coincides with Australian Federal elections and examines the views on important political and social issues of all major party candidates standing for election.(5) In 1993, a total of 415 responses were received from candidates for the Australian Labor Party (ALP), Liberal, National, and Democrat Parties, as well as the Greens, The Greens (WA) Inc, The Green Party of South Australia and the Green Alliance Senate-NSW (a response rate of 70%). In addition to asking questions on attitudes to particular social and political issues, the survey asked questions relating to the role of the representative. Several background variables such as age, country of birth, parents' country of birth, party affiliation were available. The questionnaires were sent to 'viable' parliamentary candidates on 12 March 1993.(6)

A 'restricted' version of the data file also exists which contains a number of further variables derived from sources other than the questionnaire.(7) This file was used for the analysis presented in this paper. The use of the restricted file allowed linking the respondents to the electorate in which they stood as a candidate. This was necessary in order to examine whether candidates from 'ethnic' electorates behave differently or have different attitudes on particular issues to candidates from non-ethnic electorates. A variable was constructed called type of electorate, consisting of three electorate types and was based on figures derived from 1991 census data and the 1993 electoral boundaries.(8) The candidates' responses to various questions in the 1993 ACS were then tabulated according to the type of electorate in which they stood as candidates or incumbent Members. The three electorate types were:

  • Ethnic electorates: This includes all House of Representative electorates where at least 15% of the total population in that electorate were born in non-English speaking countries (NESC). In 1993 there were 49 such electorates and the mean number of people born in NESCs was 23.98 %. All of these electorates were in urban areas. Such a definition of an ethnic electorate is arbitrary. Some may argue that a higher cut off would more accurately describe an 'ethnic' electorate. The 15% figure was chosen, however, as it approximates the national figure for people born in NESCs and has been used by other analysts.(9)
  • Non-ethnic urban electorates: This includes all non-rural House of Representative electorates where fewer than 15% of the population were born in NESCs. In 1993 there were 51 such electorates and the mean number of people born in NESCs was 8.9 %.
  • Rural electorates: This includes House of Representative electorates defined as rural by the Australian Electoral Commission (AEC). In 1993 there were 47 such electorates and the mean number of people born in NESCs was 5.04 %. Thus all rural electorates were classified as 'non-ethnic' for the purposes of this analysis.

Table 1 shows the frequency distribution of the candidates in the 1993 sample according to type of electorate. The total number of candidates (total N) in the file used for this analysis was 335.

Table 1: Frequency distribution of candidates by electorate type

Electorate type          Frequency       Per cent 
Ethnic                          79           23.6 
Non-ethnic urban               111           33.1 
Rural                           85           25.4 
Senate                          60           17.9 
Total                          335          100.0 

A second way in which some of the data were tabulated was according to the country of origin of the candidate, or more precisely, whether the candidate or the candidate's parents were born in NESCs. This variable also consists of three types:

  • NESBI Candidates: includes those candidates who were born in a NESC;
  • NESBII Candidates: includes all those born in Australia who indicated that at least one parent was born in a NESC;
  • Non-NESB Candidates: those who were born in English speaking countries (ESC) (including Australia) or candidates born in any NESC but whose parents were born in an ESC (including Australia).

Table 2 shows the frequency distribution of the candidates according to country of origin.

Table 2: Frequency distribution of candidates by country of origin

Origin                              Frequency        Per cent 
NESBI                                      23             6.9 
NESBII                                     21             6.3 
Born in ESC                                                   
(including Australia)                     288            86.0 
Unknown                                     3             0.9 
Total                                     335           100.0 

Table 3 shows the party affiliations of all the candidates who responded to the survey. It suggests that there was an even distribution between the two major parties, the ALP and the Liberal Party, with close to 10% from the National Party and 27% from the Democrats.

Table 3: Frequency distribution of the party affiliation of candidates

Party                    Frequency       Per cent 
Liberal                         97           29.0 
ALP                             97           29.0 
National                        32            9.6 
Democrat                        90           26.9 
Green (all)                     19            5.7 
Total                          335          100.0 

For the purposes of the analysis, party affiliations were allotted to three main groups:

  • ALP (97);
  • Coalition (Liberals + Nationals) (129);
  • Other (Democrats + Greens) (109) .(10)

Given the constituency focus of the study, the population of the sample used for most of the analysis was for House of Representatives candidates only. Of the 335 in the sample, 275 (or 82%) of the candidates were standing for the House of Representatives, and 60 (or 18%) were standing for the Senate. Furthermore, of the 335 candidates, only 92 (or 27%) were successfully elected in the 1993 election.

Activity in an ethnic organisation

A useful starting point is the degree to which candidates were active in ethnic organisations. Studies have shown that such organisations play a key role in a Member of Parliament's (MP) constituency activity.(12) Table 4 shows the responses for those who indicated that they had been or were active in an ethnic organisation by type of electorate.

Table 4: Activity of candidates in an ethnic group organisation by type of electorate

(%)

Degree            Ethnic       Non-     Rural     Total 
of activity       (n=47)     ethnic    (n=37)   (n=145) 
                             (n=61)                     
Active (a)          31.9       24.6      21.6      26.3 
Not active          68.1       75.4      78.4      73.7 
Total              100.0      100.0     100.0     100.0 

(a) Combines the responses to those that said they were 'very active' and 'somewhat active'.

Table 4 suggests that just over one quarter (26%) of all House of Representatives candidates who stood for the 1993 federal election, and who responded to this question, had been active in an ethnic group organisation. Candidates who stood in ethnic electorates were more likely to have been active in such an organisation (just under one third) compared with just under one quarter for non-ethnic urban electorates and just over one fifth in rural seats.

Table 5: Activity of candidates in an ethnic organisation by type of electorate and party affiliation

(%)


Degree             Ethnic electorates            Non-ethnic electorates                 Rural              Total    
of              ALP      Lib/N     Other        ALP     Lib/N      Other       ALP      Lib/N     Other           
activity     (n=13)     (n=17)    (n=17)     (n=21)    (n=22)     (n=18)    (n=14)     (n=13)    (n=10)   (n=145) 
Active         38.5       41.2      17.7       28.6      22.7       22.3      28.5          -      40.0      26.1 
Not active     61.5       58.8      82.3       71.4      77.3       77.7      71.5      100.0      60.0      73.9 
Total         100.0      100.0     100.0      100.0     100.0      100.0     100.0      100.0     100.0     100.0 

Table 5 also shows the party affiliation of responding candidates within the different types of electorate. For example, 39% of all responding ALP candidates and 41% of all responding Coalition candidates who stood in ethnic electorates indicated they had been or were active in an ethnic organisation. This compares with only 29% of responding ALP and 23% of responding Coalition candidates in the non-ethnic urban electorates. Within the rural electorates, it is the ALP (29%) and 'Other' (40%) candidates who stand out as being active in an ethnic organisation, with none of the responding Coalition candidates in such electorates ever being active.

Several features stand out from the above tables. First, is the relatively high proportion of all candidates who indicated being active in an ethnic organisation (26%).(13) This confirms the importance that such organisations have been found to play in the political process, particularly at a constituency level.(14) Second, the ethnicity of the electorate seems to affect the degree to which candidates are active in ethnic organisations, with candidates from ethnic electorates appearing to be more likely to be active than those from non-ethnic electorates.

It is also reasonable to assume that activity in an ethnic organisation captures a degree of candidate empathy or identification with ethnic groups. If this is the case, then candidates (from both political parties) from ethnic electorates are more likely to identify or empathise with the ethnic communities in their electorates. This may suggest that having a low number of NESB MPs relative to the wider population does not necessarily mean that ethnic issues are ignored. In particular, non-NESB MPs who represent ethnic electorates must establish and communicate with their ethnic constituencies to be successful.(15) Third, while Coalition candidates within ethnic electorates were more active than their counterparts in non-ethnic electorates, ALP candidates seemed more likely to be active in an ethnic organisation across all types of electorates.

Was there any difference in the degree of activity in an ethnic organisation by a candidate's country of origin? These breakdowns are shown in Tables 6 and 7. Given the small number of candidates from NESB in the sample (and the large number of missing observations), these results should be viewed with caution. To compensate for this, the responses from Senate candidates were also included.

Table 6: Activity in an ethnic organisation by candidates' country of origin (a)

(%)

Degree of                 NESBI       NESBII     All NESB    Non-NESBs         All 
activity                 (n=15)       (n=13)       (n=28)      (n=147)     (n=175) 
Active                     60.0         23.1         42.9         23.1        26.3 
Not active                 40.0         76.9         57.1         76.9        73.7 
Total                     100.0        100.0        100.0        100.0       100.0 

(a) Includes House of Representatives and Senate candidates.

Table 6 suggests that NESBI candidates were more likely to be active in an ethnic organisation. The equal proportions for the NESBII and non-NESB respondents suggests that ethnicity may play less of a role amongst the second generation. The higher number of NESB candidates active in ethnic organisations relative to non-NESB candidates also suggests that the former may be more sensitive to NESB issues/needs than the latter. That is, there may be some validity to the mirror representation argument that ethnic constituents are better represented by elected representatives from ethnic backgrounds. This is also consistent with previous work which suggests that MPs from ethnic backgrounds have higher degrees of parliamentary responsiveness to their ethnic sub-constituencies than their non-ethnic counterparts.(16)

Table 7: Activity in an ethnic organisation by candidates' country of origin and

party affiliation (a)

(%)


Degree                         NESBI                      NESBII                        Non-NESB  
of activity           ALP      Lib/N     Other     ALP     Lib/N      Other       ALP      Lib/N     Other 
                    (n=4)      (n=7)     (n=4)   (n=5)     (n=1)      (n=7)    (n=51)     (n=52)    (n=44) 
Active               75.0       42.9      75.0    40.0         -       14.3      23.5       21.2      25.0 
Not active                                                                                                 
                     25.0       57.1      25.0    60.0     100.0       85.7      76.5       78.8      75.0 
Total               100.0      100.0     100.0   100.0     100.0      100.0     100.0      100.0     100.0 

(a) Includes House of Representatives and Senate candidates.

Table 7 suggests that NESB candidates who were members of the ALP were more likely to be active in an ethnic organisation than their NESB Coalition counterparts. NESBII ALP candidates were also more likely to be active in an ethnic organisation compared to their NESBII counterparts. There was little difference in ethnic organisation activity across political party affiliation amongst non-NESB candidates. Although the number of responses was small, the findings are consistent with previous work which has shown that the ALP, at least since 1972, has made a greater effort at both recruiting ethnic members(17) and at least until 1996, been more successful at wooing ethnic voters.(18)

Ethnicity of the electorate and representatives' attitudes and behaviour towards constituency work

Previous studies have suggested that the ethnic composition of the electorate can be an important factor in determining how a representative behaves in the constituency.(19) This manifests itself in several ways. First, ethnic electorates in Australia also tend to be lower in socio-economic terms (e.g. income and occupation).(20) Constituents from such electorates often place more demands on their MPs because they face greater problems with issues such as unemployment or social security than constituents from higher socio-economic electorates.(21) This may mean that MPs have less choice in terms of their 'presentational style'.(22) In brief, they may have to adopt a constituency focus whether they wish to or not.(23) Second, studies have shown that in assessing their MPs, ethnic minorities tend to place greater importance on their MPs' approach to constituency work than their approach to particular policy issues or how they vote in Parliament.(24) In the Australian context for instance, a great deal of representatives' time in ethnic electorates is spent dealing with the immigration problems of their ethnic constituents.(25) In brief, it is reasonable to assume that candidates or members from ethnic electorates may firstly place greater importance on the constituency face of representation(26), and secondly, may spend more time on constituency related activities.(27) This section examines this proposition using some of the relevant responses from the survey.

Candidate beliefs with respect to constituency work

Table 8 shows by type of electorate, how candidates saw the importance of helping constituents with their problems. It suggests that the overwhelming majority of all candidates believed that helping constituents with their problems was a very important aspect of an MP's job. Contrary to the above proposition, however, relatively fewer candidates from ethnic electorates felt this way compared to those from non-ethnic electorates.

Table 8: The importance of helping constituents with their problems by type of electorate

(%)

Degree                        Ethnic     Non-ethnic          Rural                
of                       electorates    electorates    electorates          Total 
importance                    (n=74)        (n=102)         (n=81)        (n=257) 
Very important                  75.7           89.2           82.7           83.3 
Fairly important                21.6           10.8           16.0           15.6 
Not very important               2.7              -            1.2            1.2 
Not at all important               -              -              -              - 
Total                          100.0          100.0          100.0          100.0 

What happens once party affiliation is taken into account? Table 9 suggests that the greatest within electorate type difference occurs among ethnic electorates. ALP candidates from such electorates, for instance, were more likely to feel that helping constituents with their problems was very important (94%) than either Coalition (79%) or 'Other' (59%) candidates. Although ALP candidates seemed to have a strong constituency focus in all types of electorates, the relative strength of this feeling was most marked in ethnic electorates. One possible explanation is the fact that the bulk of these electorates were, and are, held by the ALP and candidates and members may consider constituency work, especially ethnic related, to be an important source of electoral support. This is consistent with in-depth studies of the political process in ethnic electorates.(28) Interestingly, Coalition candidates were most likely to view helping constituents with problems as being very important (95%) in non-ethnic urban electorates.

Table 9: The importance of helping constituents with their problems by type of electorate and party affiliation

(%)

Degree                      Ethnic                Non-ethnic          Rural electorates   
of                        electorates            electorates                           
importance           ALP   Lib/N   Other     ALP   Lib/N   Other    ALP   Lib/N   Other 
                    (n=18)  (n=29)  (n=27)  (n=30)  (n=40)  (n=32) (n=26)  (n=33)  (n=22) 
Very important        94.4    79.3    59.3    93.3    95.0    78.1   84.6    78.8    86.4 
Fairly important                                                                          
                       5.6    20.7    33.3     6.7     5.0    21.9   15.4    18.2    13.6 
Not very important     7.4       -       -       -      -     3.0       - 
Not at all important     -       -       -       -       -       -      -       -       - 
Total                100.0   100.0   100.0   100.0   100.0   100.0  100.0   100.0   100.0 

Another important aspect of constituency work for MPs, especially in ethnic electorates, is attending local community functions. Research has shown how attendance at functions serves an important communications role for MPs, providing a forum to both present themselves to their ethnic sub-constituency and learn about their constituents' needs and problems.(29) We would therefore once again expect candidates from ethnic electorates to place more importance on attending local functions than those from non-ethnic electorates. This result is not borne out by the responses in Table 10.

Table 10: The importance of attending local community functions by type of electorate

(%)

Degree of                   Ethnic      Non-ethnic         Rural             
importance             electorates     electorates   electorates       Total 
                            (n=74)         (n=102)        (n=80)     (n=256) 
Very important               48.6             52.0          45.0        48.8 
Fairly important              43.2            43.1          48.8        44.9 
Not very important             8.1             4.9           6.3         6.3 
Not at all                       -               -             -           - 
  important                                                                    
Total                        100.0           100.0         100.0       100.0 

When the responses are also viewed according to party affiliation (Table 11), however, a similar pattern to that seen above is evident. Although candidates from both major parties viewed attending local community functions as important (i.e. either 'very' or 'fairly'), ALP candidates from ethnic electorates were more likely to view attending local community functions as being very important (78% compared to 55% for Coalition candidates). While ALP candidates in all electorate types were more likely to also agree with this, the difference with other candidates was not as marked in non-ethnic and rural electorates. This finding is consistent with case study evidence which shows that attending local ethnic community functions in ethnic electorates is an important means for Labor politicians to build up support and trust amongst particular ethnic sub-constituencies, needed for internal factional reasons as well as wider electoral support.(30)

Table 11: The importance of attending local community functions by type of electorate and party affiliation

(%)

Degree                  Ethnic electorates          Non-ethnic            Rural electorates    
of importance                                      electorates                             
                        ALP   Lib/N   Other     ALP   Lib/N   Other      ALP   Lib/N   Other 
                     (n=18)  (n=29)  (n=27)  (n=30)  (n=40)  (n=32)   (n=25)  (n=34)  (n=21) 
Very                                                                                         
  important            77.8    55.2    22.2    60.0    57.5    37.5     52.0    41.2    42.9 
Fairly important                                                                             
                       22.2    44.8    55.6    36.7    40.0    53.1     48.0    52.9    42.9 
Not very                                                                                     
  important               -       -    22.2     3.3     2.5     9.4        -     5.9    14.3 
Not at all                                                                                   
  important               -       -       -       -       -       -        -       -       - 
Total                 100.0   100.0   100.0   100.0   100.0   100.0    100.0   100.0   100.0 

Member behaviour with respect to constituency work

The above section dealt only with candidate attitudes towards constituency work. The survey also asked only those who were current or past Members of the House of Representatives to estimate the actual time (i.e. real behaviour) they spent on constituency related work. This section examines the responses of then current and past MPs to time spent dealing with constituents' problems and attending local community functions. The proposition is once again that Members from ethnic electorates should spend more time on these activities than those from non-ethnic electorates.

Time spent on constituents' problems

Table 12 suggests that the overwhelming majority (71%) of all the then current and past MPs spent over 15 hours per month dealing with constituents' problems. Contrary to the proposition, the responses suggest that MPs from rural electorates spent the most amount of time dealing with constituents' problems, followed by MPs from non-ethnic electorates. In contrast, just over half of MPs from ethnic electorates spent more than 15 hours per month on constituents' problems.

Table 12: Amount of time spent by MPs dealing with constituents' problems by type of electorate (a)

(%)

Hours                   Ethnic         Non-ethnic          Rural               
per                electorates        electorates    electorates           All 
month (b)               (n=19)             (n=34)         (n=16)        (n=69) 

0-15                      47.4               26.5           12.5          29.0 
>15                       52.6               73.5           87.5          71.0 
Total                    100.0              100.0          100.0         100.0 

(a) Refers to then current and past House of Representatives Members only.

(b) Collapses into two categories responses to: none, up to 5, 5-10, 11-15, 16-20, more than 20.

Introducing party affiliation, however, once again seems to change the story (Table 13). It suggests that ALP MPs within ethnic electorates spent more time dealing with constituents' problems than Coalition MPs from the same type of electorates. Similarly, Coalition MPs from non-ethnic urban electorates spent more time on constituent problems than their ALP colleagues.

Table 13: Amount of time spent on constituents' problems by type of electorate and party affiliation

(%)

Hours          Ethnic electorates          Non-ethnic            Rural electorates    
per                                       electorates                              
month        ALP    Lib/N   Other     ALP   Lib/N    Other     ALP   Lib/N    Other 
(a)        (n=10)    (n=8)   (n=1)  (n=14)  (n=17)    (n=3)   (n=8)   (n=8)    (n=0) 
0-15         40.0     50.0   100.0    42.9    11.8     33.3    12.5    12.5        - 
>15          60.0     50.0       -    57.1    88.2     66.7    87.5    87.5        - 
Total       100.0    100.0   100.0   100.0   100.0    100.0   100.0   100.0        - 

(a) Collapses into two categories responses to: none, up to 5, 5-10, 11-15, 16-20, more than 20.

Attendance at local community functions

Table 14 suggests that the majority of then current or past Members (54%) spent more than 15 hours per month attending local community functions. Similar to the findings for time spent on constituents' problems, proportionately more Members from rural electorates spent 15 hours or more per month attending local community functions, followed by those from non-ethnic urban electorates and then those from ethnic electorates.

Table 14: Amount of time spent by MPs attending local community functions by type of electorate (a)

(%)

Hours                   Ethnic      Non-ethnic           Rural                 
per                electorates     electorates     electorates             All 
month (b)               (n=19)          (n=34)          (n=16)          (n=89) 
0-15                      57.9            41.2            18.8            46.0 
>15                       42.1            58.8            81.3            54.0 
Total                    100.0           100.0           100.0           100.0 

(a) Refers to then current and past House of Representatives Members only.

(b) Collapses into two categories responses to: none, up to 5, 5-10, 11-15, 16-20, more than 20.

Turning to time spent by party affiliation (Table 15), a similar pattern to that seen above emerges once again. A Labor MP from an ethnic electorate spent more time attending local community functions than did a non-Labor MP from that same type of electorate. The reverse was the case for the non-ethnic urban electorates.

Table 15: Amount of time spent by MPs attending local community functions by type of electorate and party affiliation.

(%)

Hours          Ethnic electorates     Non-ethnic electorates     Rural electorates   
per           ALP    Lib/N   Other      ALP   Lib/N    Other     ALP   Lib/N   Other 
month      (n=10)    (n=8)   (n=1)   (n=14)  (n=17)    (n=3)   (n=8)   (n=8)   (n=0) 
0-15         50.0     62.5   100.0     57.1    17.6    100.0    25.0    12.5       - 
>15          50.0     37.5       -     42.9    82.4        -    75.0    87.5       - 
Total       100.0    100.0   100.0    100.0   100.0    100.0   100.0   100.0       - 

A consistent picture seems to be emerging. In brief, the moral of the story seems to be that ALP MPs (and candidates) who represented ethnic electorates, worked their constituencies more than did their Coalition counterparts from those same electorates. Such a picture is consistent with the long held view that the ALP has been better at winning (at least until the 1996 election) the supposed 'ethnic vote' than the Coalition. Any ethnic electorate effect on candidate and Member attitudes, therefore, seems to manifest itself through the political party that those candidates or Members belonged to.

Candidate attitudes to ethnic related issues

Most studies which have examined the opinions of political elites (such as candidates) to ethnic related issues have been concerned with the degree to which it divergs from that of their respective constituents.(31) Yet does the ethnicity of the electorate influence elite opinion on such issues? This section reports the responses of candidates to several ethnic related issues according to the type of electorate they stood for.

Table 16: Candidates who included 'immigration and multiculturalism' in the four most important problems facing the country by type of electorate

(%)

      Ethnic             Non-ethnic         Rural electorates   
   electorates          electorates                             
       35.7                 35.7                  28.6          

Note

Based on responses to a question which asked all candidates to select from a list of 14 the four most important problems facing the country. One item in this list was 'immigration and multiculturalism'.

Table 16 shows the responses of candidates who included 'immigration and multiculturalism' in the four most important problems facing the country by type of electorate. This response is difficult to interpret for two reasons. First, because it combines two separate, albeit related, issues in the one category. Whether a candidate indicated immigration and multiculturalism, for instance, does not tell us whether he or she included it because they saw the present levels of immigration as being too high, or because they were critical of multiculturalism. Second, candidates may have included 'immigration and multiculturalism' in their list of top four problems because they saw multiculturalism, for instance, as an important policy priority rather than a problem. In any case, there does not appear to be any support for an ethnic electorate effect in this case.

Table 17: Candidates who included 'immigration and multiculturalism' in the four most important problems, by type of electorate and party affiliation

(%)

                     Ethnic    Non-ethnic        Rural              
Party           electorates   electorates  electorates          All 
affiliation          (n=20)        (n=20)       (n=16)       (n=56) 
ALP                    40.0          15.0         25.0         26.8 
Lib/Nat                25.0          45.0         56.3         41.1 
Other                  35.0          40.0         18.8         32.1 
Total                 100.0         100.0        100.0        100.0 

A more interesting pattern emerges, however, when the party affiliation of the candidates is also introduced (Table 17). Several observations can be made. First, only 56 (or 20%) of the 275 House of Representatives candidates who responded included immigration and multiculturalism in their list of top four problems facing the country. Second, proportionately more Coalition candidates (41%) indicated 'immigration and multiculturalism' as one of the four problems, followed by 'Other' candidates (32%) and then the ALP (27%). Third, and perhaps most interesting is the within electorate breakdown. Although table 16 suggested a relatively even distribution across electorate types, there seems to be a bigger variation according to party affiliation within electorate types. Within ethnic electorates, for instance, 40% of candidates who indicated 'immigration and multiculturalism' were from the ALP compared to 25% for the Coalition. This pattern was reversed in the non-ethnic urban and rural electorates.

One interpretation of why proportionately more ALP candidates from ethnic electorates included 'immigration and multiculturalism' in their list, is that ALP candidates from ethnic electorates were more sensitive to immigration and multicultural issues (if their inclusion of 'immigration and multiculturalism' in the top four was indicative of policy priority). Another interpretation may be that ALP candidates and Members were more likely to rank immigration as a 'problem' because they have more direct experience of immigration related issues through their constituency case work.(32) In contrast, the fact that Coalition candidates in non-ethnic electorates were more likely to see immigration and multiculturalism as a problem is consistent with the approach towards these issues that was contained in the Liberal Party Fightback! policy document.(33) Given the ambivalent nature of the question, however, any firmer conclusion is precluded.

Other issue-based responses concerned 'equal opportunities for migrants' and the number of migrants allowed into Australia at the time of the survey (1993). Table 18 shows the responses regarding 'equal opportunities for migrants' by type of electorate.

Table 18: Candidate attitudes to 'equal opportunities for migrants' by type of electorate

(%)

Equal                      Ethnic     Non-ethnic          Rural              
opportunities         electorates    electorates    electorates          All 
for migrants (a)           (n=76)        (n=109)         (n=82)      (n=267) 
Gone too far                 11.8           17.4           11.0         13.9 
About right                  57.9           46.8           63.4         55.1 
Not gone far enough          30.2           35.7           25.6         31.1 
Total                       100.0          100.0          100.0        100.0 

(a) Collapses into three categories responses to: 'gone much too far' 'gone too far' 'about right' 'not gone far enough' 'not gone nearly far enough'.

Just over half of all candidates (55%) thought that 'equal opportunities for migrants' were about right. However, just under a third of all candidates believed that changes had not gone far enough. Was there much difference by type of electorate? One surprising result was that 37% of candidates in non-ethnic electorates thought changes had not gone far enough compared to 30% of candidates in ethnic electorates.

Table 19: Candidates' attitudes on 'equal opportunities for migrants' by type of electorate and party affiliation

(%)

Equal                  Ethnic electorates         Non-ethnic         Rural electorates   
opportunities                                     electorates                           
for migrants (a)       ALP   Lib/N   Other     ALP   Lib/N   Other    ALP   Lib/N   Other 
                    (n=20)  (n=29)  (n=27)  (n=32)  (n=41)  (n=36) (n=27)  (n=35)  (n=20) 
Gone too far           5.0    17.2    11.1     9.4    29.3    11.1      -      20    10.0 
About right           50.0    65.5    55.6    34.4    58.5    44.4   59.3    77.1    45.0 
Not gone                                                                                  
  far enough          45.0    17.2    33.3    56.3    12.2    44.5   40.7     2.9    45.0 
Total                100.0   100.0   100.0   100.0   100.0   100.0  100.0   100.0   100.0 

(a) As for Table 18.

What emerges when we also look at party affiliation? The results (Table 19) seem to be as expected with a greater proportion of ALP candidates (45%) from ethnic electorates indicating that changes had not gone far enough, compared to only 17% of Coalition candidates. This pattern is even stronger in the non-ethnic electorates, with over half of ALP candidates indicating that changes had not gone far enough compared to only 12% of Coalition candidates in non-ethnic urban electorates, and 41% of ALP candidates compared to 3% of Coalition candidates in rural electorates. These responses suggest that party affiliation is more important than electorate type in determining candidates' attitudes.

Table 20: Candidates' attitudes towards the number of migrants allowed into Australia by type of electorate

(%)

No. of migrants                    Ethnic    Non-ethnic         Rural              
allowed into                  electorates   electorates   electorates          All 
Australia (a)                      (n=68)        (n=97)        (n=68)      (n=233) 
Gone too far                         48.5          54.6          52.9         52.4 
About right                          51.5          45.4          47.1         47.6 
Not gone far enough                     -             -             -            - 
Total                               100.0         100.0         100.0        100.0 

(a) Responses collapsed as for Table 18.

A contentious issue in the politics of immigration in Australia has been the size and composition of the immigration program.(34) Table 20 shows the candidates' attitude by type of electorate to the number of migrants being allowed into Australia. Several features stand out. First, the majority of all candidates (52%) believed that the number of migrants allowed into Australia had gone too far or much too far. Although candidates from ethnic electorates were more likely to say that the number of migrants was about right (the only group where a majority of candidates said this), the high number of candidates (49%) from ethnic electorates who said that numbers had gone too far is surprising, although is consistent with the responses in Table 17.

Table 21: Candidates' attitudes towards the number of migrants allowed into Australia by type of electorate and party affiliation

(%)

No. of migrants      Ethnic electorates        Non-ethnic         Rural electorates   
allowed into                                  electorates                            
Australia           ALP    Lib/N   Other     ALP   Lib/N   Other     ALP   Lib/N   Other 
                 (n=15)   (n=28)  (n=25)  (n=29)  (n=38)  (n=30)  (n=21)  (n=30)  (n=17) 
Gone too far           6.7     71.4      48    10.3    81.6    63.3     9.5    73.3    70.6 
About right           93.3     28.6      52    89.7    18.4    36.7    90.5    26.7    29.4 
Not gone                                                                                    
  far enough             -        -       -       -       -       -       -       -       - 
Total                100.0    100.0   100.0   100.0   100.0   100.0   100.0   100.0   100.0 

How does party affiliation influence these views? Once again, ALP candidates from ethnic electorates were much more likely to view the number of migrants allowed in as about right (93%), compared to their Coalition counterparts (29%). The overwhelming majority of Coalition and 'Other candidates' in non-ethnic electorates (especially rural) were of the view that the number of migrants allowed into Australia had gone too far. The similarity of ALP candidate responses across electorate types, however, would suggest that party affiliation, rather than the ethnicity of the electorate, was the most important determining factor in this case.

Conclusion

This paper examined several responses to the 1993 Australian Candidate Study to explore the proposition that candidates and Members from electorates with high ethnic concentrations have different representational attitudes and behaviour to their counterparts from non-ethnic electorates. Was this borne out? In general there did not appear to be differences which could relate to the type of electorate (i.e. the extent of its ethnic composition).(35) However, when party affiliation by type of electorate was examined, there did appear to be differences. The lack of an explicit 'ethnic electorate effect' on representation in this case does not mean that the ethnic composition of an electorate is unimportant. Such a finding may have more to do with the nature and appropriateness of the data source and methods used to examine the issue. Studies which have used more qualitative based approaches, for instance, have reached different conclusions.(36)

Looking at candidate attitudes and behaviour towards constituency work, there appeared to be no 'pure' ethnic electorate effect. There did appear to be some evidence, however, that attitudes and behaviour towards constituency work varied according to the party affiliation of the candidates within the three electorate types examined. In short, the findings suggested that ALP candidates within ethnic electorates behaved and thought in ways which were more consistent with the ethnic composition of their electorates. This should not come as a surprise. It is consistent, for instance, with previous work which has suggested that the ALP has been better at 'working' ethnic electorates and consequently attracting the ethnic vote.

What is perhaps surprising is that ALP Members and candidates in ethnic electorates continued to work hard in spite of the fact that the bulk of ethnic electorates are also safe ALP seats.(37) The received wisdom is that a representative who comes from a safe seat might engage in less constituency work than a representative from a marginal seat. This is referred to as the 'marginality hypothesis', although empirical tests of this hypothesis have reached mixed results.(38)

A similar story to that of the type of electorate effect on constituency behaviour emerged with respect to candidate attitudes to ethnic related issues. While there was some evidence of an 'ethnic electorate effect', most differences in candidate attitudes which emerged were more likely to be related to the candidate's party affiliation than the ethnic composition of the electorate. In general, ALP candidates from ethnic electorates had positions on ethnic related issues which could be assumed to have been shared by the majority of their ethnic constituents.

Endnotes

  1. Jupp, B. York, & A. McRobbie, The Political Participation of Ethnic Minorities in Australia, AGPS, Canberra, 1989.
  2. Zappalà, The Parliamentary Responsiveness of Australian Federal MPs to their Ethnic Constituents, (Information and Research Service Research Paper No.8), Department of the Parliamentary Library, Canberra, 1997; G. Zappalà, Four Weddings, a Funeral and a Family Reunion: Ethnicity and Representation in Australian Federal Politics, AGPS, Canberra, 1997. This latter work was completed as part of the 1996 Australian Parliamentary Fellow's research project, and was based on an ethnographic case study of an MP in a highly 'ethnic' electorate as well as a content analysis of the parliamentary interventions of a group of MPs from both ethnic and non-ethnic electorates between 1983 and 1996.
  3. As well as the effects on representation which may be associated with the composition of the electorate, studies have also suggested that the 'marginality' of the electorate is an important influence on representational behaviour (see references in Note 38). The analysis in this paper, however, cannot separate out these two influences. Suffice it to note that the bulk of 'ethnic electorates' in Australia (with a few key exceptions) are also safe ALP seats. We return to this issue in the final section.
  4. The fourth ACS was conducted in 1996. Given the changed political environment since the 1993 survey, any findings presented here should not be extrapolated to the present situation.
  5. The survey was designed and conducted by I. McAllister, R. Jones, D. Denemark and D. Gow. A copy of the data file and questionnaire are held at the Social Science Data Archives (SSDA), at the Australian National University. They bear no responsibility for the analysis and interpretation of the data in this paper. See also I. McAllister, R. Jones, D. Denemark, and D. Gow, Attitudinal Responses and Data User's Guide for the Australian Candidate Study, 1993, Social Science Data Archives, ANU, Canberra, 1994.
  6. Viable candidates were those from all the major parties, identifiable Green and other environmental candidates, and other candidates whom it was anticipated would obtain more than 10% of the first preference vote.
  7. The restricted version of the file was created by Professor Ian McAllister and is only released on approval from the principal investigators. I thank Sophie Holloway from the SSDA for providing the access and use of this file.
  8. Kopras, Comparisons of 1991 Census Characteristics: Commonwealth Electoral Divisions, (Parliamentary Research Service Background Paper No.20), Department of the Parliamentary Library, Canberra, 1993.
  9. Jupp, 'The Ethnic Dimension', in C. Bean, S. Bennett, M. Simms, and J. Warhurst (eds.) The Politics of Retribution-The 1996 Federal Election, Allen & Unwin, Sydney, 1997; Zappalà, Four Weddings, op. cit.
  10. It can be seen from Table 3 that the bulk of the 'Other' category were candidates from the Democrats.
  11. Where the total n does not equal 335 means that not all respondents answered that particular question. Percentages may not always total to 100 due to rounding.
  12. Zappalà, Four Weddings, op. cit., Ch.2, 3.
  13. This represented 30% of those who were successfully elected.
  14. Unikoski, Communal Endeavours: Migrant Organisations in Melbourne, ANU Press, Canberra, 1978; Zappalà, Four Weddings, op. cit., Ch.2, 3.
  15. Zappalà, Four Weddings, op. cit., Ch.2.
  16. Zappalà, The Parliamentary Responsiveness, op. cit.
  17. Allan, 'Ethnic Politics in the ALP', in P.R. Hay, J. Halligan, J. Warhurst and B. Costar (eds.) Essays on Victorian Politics, Warrnambool Institute Press, Warrnambool, 1985.
  18. McAllister, 'Ethnic Issues and Voting in the 1987 Federal Election', Politics 23(2), 1988, pp. 219-47.
  19. Jewell, 'Legislator-Constituency Relations and the Representative Process', Legislative Studies Quarterly, 8(3), 1983, pp.303-37; Zappalà, Four Weddings, op. cit.
  20. Kopras, Comparisons of 1991, op. cit.
  21. Studlar & I. McAllister, 'Constituency Activity and Representational Roles among Australian Legislators', Journal of Politics, 58(1), 1996, p.77.
  22. Fenno, Home Style-House Members in their Districts, Little Brown, Boston, 1978.
  23. Searing, 'The Role of the Good Constituency Member and the Practice of Representation in Great Britain', Journal of Politics, 47, 1985, pp. 348-81.
  24. Cain, J. Ferejohn, & M. Fiorina, The Personal Vote: Constituency Service and Electoral Independence, Harvard University Press, Cambridge, Mass., 1987, pp.42-3.
  25. Zappalà, Four Weddings, op. cit., Ch.3.
  26. Norton, & D.M. Wood, Back From Westminster: British Members of Parliament and their Constituencies, University Press of Kentucky, Lexington, Ky., 1993.
  27. Zappalà, Four Weddings, op. cit.
  28. ibid.
  29. ibid.
  30. ibid.
  31. McAllister, 'Immigration, Bipartisanship and Public Opinion' in J. Jupp & M. Kabala (eds.), The Politics of Australian Immigration, AGPS, Canberra, 1993; I. McAllister, 'Public Opinion, Multiculturalism, and Political Behaviour in Australia', in C. Kukathas, Multicultural Citizens: the Philosophy and Politics of Identity, Centre for Independent Studies, St. Leonards, NSW, 1993; S. Jackman, 'Race and Immigration in Australian Political Ideology: Political Elites versus ''the Mainstream''', Paper presented at the Research School of Social Science, ANU, 7 April, 1997.
  32. Zappalà, Four Weddings, op. cit., Ch.3.
  33. See the various contributions in Migration Action, 15(1) February, 1993.
  34. Jupp & M. Kabala, The Politics of Australian Immigration, op. cit.
  35. The data does suggest, however, that there may be a 'rural electorate effect' operating on representation, which requires further research.
  36. Zappalà, Four Weddings, op. cit.
  37. Economou, 'An Overstated Electoral Importance? A Note on ''Ethnic'' Voting and Federal Electoral Outcomes', People and Place, 2(4), 1994, pp. 45-51; Jupp, The Ethnic Dimension, op. cit.
  38. See for example, P.E. Converse & R. Pierce, Political Representation in France, Harvard University Press, Cambridge, Mass., 1986; I. Crewe, 'MPs and their Constituents in Britain: How Strong are the Links?' in V. Bogdanor (ed.), Representatives of the People?, Gower, Aldershot, 1985; Studlar & McAllister, op. cit.

References

Allan, L., 'Ethnic Politics in the ALP', in P.R. Hay, J. Halligan, J. Warhurst and B. Costar (eds.) Essays on Victorian Politics, Warrnambool Institute Press, Warrnambool, 1985.

Cain, B., Ferejohn J. & Fiorina, M., The Personal Vote: Constituency Service and Electoral Independence, Harvard University Press, Cambridge, Mass., 1987.

Converse, P.E. & Pierce, R. Political Representation in France, Harvard University Press, Cambridge, Mass., 1986.

Crewe, I., 'MPs and their Constituents in Britain: How Strong are the Links?' in V. Bogdanor (ed.), Representatives of the People?, Gower, Aldershot, 1985.

Economou, N., 'An Overstated Electoral Importance? A Note on ''Ethnic'' Voting and Federal Electoral Outcomes', People and Place, 2(4), 1994, pp. 45-51.

Fenno, R.F., Home Style-House Members in their Districts, Little Brown, Boston, 1978.

Jackman, S., 'Race and Immigration in Australian Political Ideology: Political Elites versus''the Mainstream''', Paper presented at the Research School of Social Science, ANU, 7 April, 1997.

Jewell, M.E., 'Legislator-Constituency Relations and the Representative Process', Legislative Studies Quarterly, 8(3), 1983, pp.303-37.

Jupp, J., York, B. & McRobbie, A., The Political Participation of Ethnic Minorities in Australia, AGPS, Canberra, 1989.

Jupp, J & Kabala, M., The Politics of Australian Immigration, AGPS, Canberra, 1993.

Jupp, J., 'The Ethnic Dimension', in C. Bean, S. Bennett, M. Simms, and J. Warhurst (eds.) The Politics of Retribution-The 1996 Federal Election, Allen & Unwin, Sydney, 1997.

Kopras, A., Comparisons of 1991 Census Characteristics: Commonwealth Electoral Divisions, (Parliamentary Research Service Background Paper No.20), Department of the Parliamentary Library, Canberra, 1993.

McAllister, I., 'Ethnic Issues and Voting in the 1987 Federal Election', Politics 23(2), 1988, pp. 219-47.

McAllister, I., 'Immigration, Bipartisanship and Public Opinion' in J. Jupp & M. Kabala (eds.), The Politics of Australian Immigration, AGPS, Canberra, 1993.

McAllister, I., 'Public Opinion, Multiculturalism, and Political Behaviour in Australia', in C. Kukathas, Multicultural Citizens: the Philosophy and Politics of Identity, Centre for Independent Studies, St. Leonards, NSW, 1993.

McAllister, I., Jones, R., Denemark, D. and Gow, D., Attitudinal Responses and Data User's Guide for the Australian Candidate Study, 1993, Social Science Data Archives, ANU, Canberra, 1994.

Norton, P. & Wood, D.M., Back From Westminster: British Members of Parliament and their Constituencies, University Press of Kentucky, Lexington, Ky., 1993.

Searing, D.D., 'The Role of the Good Constituency Member and the Practice of Representation in Great Britain', Journal of Politics, 47, 1985, pp. 348-81.

Studlar, D.T. & McAllister, I.,'Constituency Activity and Representational Roles among Australian Legislators', The Journal of Politics, 58(1), 1996, pp.69-90.

Unikoski, R., Communal Endeavours: Migrant Organisations in Melbourne, ANU Press, Canberra, 1978.

Zappalà, G., The Parliamentary Responsiveness of Australian Federal MPs to their Ethnic Constituents, (Information and Research Service Research Paper No.8), Department of the Parliamentary Library, Canberra, 1997.

Zappalà, G., Four Weddings, a Funeral and a Family Reunion: Ethnicity and Representation in Australian Federal Politics, AGPS, Canberra, 1997.

 

Facebook LinkedIn Twitter Add | Email Print
Back to top