The strange case of migrant Australians, political parties, and participation

It happens regularly (to me, at least) that exploratory investigations of data pose more questions than they answer.  Of course, as grad students we are all taught the importance of having a clear hypothesis before commencing our research, not fishing for research questions, not finding results by iterative searching, and so on.   So, in the interests of revealing some skeletons in my research closet, I present one current, unexpected puzzle.  Hopefully, as a result, at least one grad student, somewhere, will take heart that not all research projects are well scripted.

migrants_voting_NAA

Naturalized migrants voting in the 1969 election. National Archives of Australia

(Full disclosure: this post is also at least partially self-serving.  I need help with this.)

The 2013 Australian Election Study over-sampled first generation migrants, enabling – for the first time – meaningful analysis of subsets of migrant respondents (total n = 3,955; immigrant N = 1,090). In other words, instead of being constrained to studying a nebulous subgroup of respondents whose only common trait is ‘not being born in Australia’, there are enough respondents born in Asia, enough born in Europe, and so on, to study them as respective subgroups. (Of course, it is arguable that ‘being born in Asia’ is only marginally less nebulous than ‘being born overseas’; that is certainly the case, but to really drill down into individual countries of origin would require a survey of many tens of thousands of respondents.)

With data in hand, I began a fairly inductive process, looking at how immigrants – grouped by region of birth – participate in politics in Australia (i.e. my regular topic of interest). Clive Bean had previously found that immigrants – vis a vis native-born citizens – ‘have no participatory disadvantage’ in Australia, and I had made a similar, incidental finding in my doctoral thesis. This contrasts with evidence from similar liberal democracies, wherein immigrants tend to participate at lower rates than native-born citizens.

With the 2013 data, I immediately found that immigrants don’t just ‘not have a disadvantage’. Rather, some first-generation migrants participate at much higher rates than native-born Australians.  The table below shows the mean rates of participation in a range of activities, scored ordinally. Standard deviations are in parentheses. Asterisks represent statistically significant differences from native-born Australians (the statistically significant differences each hold in OLS regressions with demographic, political interest, and party identification covariates). Anglo-American-born respondents include those born in Britain, Ireland, Canada, United States, and New Zealand.

  Australia
(n=2865)
Anglo-America
(n=286)
Europe
(n=195)
Asia
(n=269)
During campaign (1=not at all, 4=frequently)
Discussed politics 2.94 (.88) 2.98 (.89) 2.76** (.98) 2.76** (.85)
Talked to people about vote 1.47 (.78) 1.39 (.74) 1.49 (.82) 1.62** (.89)
Worked for party or candidate 1.28 (.69) 1.27 (.65) 1.20 (.56) 1.41** (.75)
Went to meetings or rallies 1.06 (.40) 1.14 (.50) 1.08 (.35) 1.23** (.56)
Contributed money 1.08 (.40) 1.05 (.50) 1.08 (.38) 1.14* (.42)
Discussed politics  online 1.39 (.77) 1.35 (.76) 1.34 (.74) 1.51* (.76)
In previous five years (0=no; 1=yes)
Contacted official (offline) .16 (.36) .20 (.40) .13 (.34) .12 (.32)
Contacted official (online) .18 (.38) .20 (.46) .12* (.33) .13* (.33)
Signed petition (offline) .45 (.50) .44 (.50) .31** (.46) .29** (.46)
Signed petition (online) .30 (.46) .30 (.46) .23* (.42) .20** (.40)
Protest/march .11 (.31) .09 (.29) .09 (.28) .09 (.29)
Worked with others .20 (.40) .18 (.38) .25 (.43) .14* (.35)

Two things stand out. First, significant differences in rates of participation are most common, and also largest, among Asian-born citizens. Second, those differences go in two distinct directions.  Reading down from the top of the table, Asian and European-born migrants are significantly less likely to discuss politics (generally), than are Australian or Anglo-American-born respondents. However, Asian-born migrants are disproportionately likely to discuss how they vote with others. Likewise, they are significantly more likely to discuss politics online (which contrasts with the ‘discuss politics’ generally measure). Perhaps most interestingly, they are far more likely to have worked for a party or candidate, attended political meetings or rallies, and contributed money to a party or candidate during the 2013 federal election campaign.

However, political activities outside the campaign period – and less supportive of the state, generally – tell an entirely different story, Rates of petition-signing are ten percentage points lower among these migrant groups than among non-migrants.  Communal work to achieve political aims is also lower among Asian-born migrants, although European-born migrants are slightly (though not significantly) more likely to report doing this.

There are distinct ethnic bases to political participation in Australia, with migrants from Asian and European backgrounds the most likely to engage in state-affirming, party-based activities. Asian-born migrants are the least likely to engage in activities that contest the state, such as signing petitions. Preliminary multivariate analyses suggest that these phenomena are not explained by respondents’ trust in the political system; it is not the case that, for instance, that migrants from authoritarian backgrounds report higher levels of trust in the political system, which leads to these state-affirming activities.

On the contrary, Asian-born migrants report the highest levels of democratic satisfaction, the lowest mean strength of party identification, are the most likely to think government is run by big interests, and the least likely to agree that politicians know what people think. And yet the are the most likely to donate time and money to those politicians. (Incidentally, they are also among the most likely to vote if not compelled.)

There is something systematic occurring here, but what might explain it?  Rather than rushing in with statistical tests and modelling, I want to get a better handle on the cultural and social contexts in which this political activity occurs.  And eventually, I want to discover why this particular subgroup of Australians appears to be the last holdout against dwindling public engagement with our major political parties.

Jill Sheppard is a lecturer at the Australian National University

Our ‘Research Spotlight’ posts enable POP Politics members to present their latest research, projects and programs to a wider audience. Unless expressly acknowledged, the Research Team are not involved or affiliated with, and make no representations about the research projects listed in this series: these posts are for information and discussion purposes only.