Maori Electoral Participation: Why is there low Maori electoral participation - 2006 Analysis by UMR Research

Maori Electoral Engagement - A Review

UMR Research 

2006

This report draws together data from a number of sources, the most important of which being a combined sample of UMR omnibus surveys from August 2002 to December 2004.

The UMR omnibus is a nationwide telephone survey of a representative sample of n=750 New Zealanders aged 18 years or over. During the specified period the UMR omnibus was conducted every four weeks. It is New Zealand’s longest running and most consistently taken omnibus poll, having been conducted since September 1991.

Some questions are asked in virtually every omnibus, including:

• Current vote
• Previous vote
• National mood (right track/ wrong track)
• Most demographics
• Government performance
• Government performance in managing the economy
• Economic expectations
• Interest rate expectations
• Standard of living expectations
• Unemployment expectations.

A number of questions relevant to this study are asked more sporadically, or began being asked only towards the end of the period designated above. As a result, the combined sample for these questions is considerably smaller, and in some cases should be treated as indicative only.

A number of questions relevant to this study are asked more sporadically, or began being asked only towards the end of the period designated above. As a result, the combined sample for these questions is considerably smaller, and in some cases should be treated as indicative only. It should be noted that the Electoral Commission has asked us to mine data collected previously for the information presented in this report. Because the questions were asked for clients other than the Electoral Commission, the questions were not designed to fit directly with similar sources of information such as the New Zealand Election Survey. There are a number of areas where we have seldom asked questions (if at all), such as interest in politics and efficacy, because they were not felt to be of particular relevance to the clients for whom the questions were originally asked. We can, however, provide some insight into these areas by way of implication (e.g. being able to express an opinion on a party leader suggests a certain base level of knowledge and therefore interest in politics).

Similarly, the omnibus does not address some demographic indicators, such as educational
achievement, that may well have an impact on likelihood of voting.
The primary focus of this analysis is on Māori non-voters. The following definitions have been used for
the purposes of this analysis:
• ‘Non-voters’ are those who admit not voting at the 2002 election, even though they were old
enough to do so
• A respondent qualifies as ‘Māori’ if they identify themselves as Māori in a multiple response
question (i.e. someone who says they are both Māori and Samoan qualifies as Māori for the
purpose of this research). This is in line with the question used in the census.
We compare Māori non-voters with three other groups:
• Māori voters
• Non-Māori non-voters
• Non-Māori voters.

This type of comparison helps us to identify which characteristics are associated with being a Māori
voter, which are associated with being a non-voter, and which are unique to Māori non-voters.

The number of Māori non-voters recruited in each n=750 nationwide sample is of course small, and we
have used the combined sample of August 2002 to December 2004 to ensure that the number of Māori
non-voters is large enough for robust analysis.

We have chosen to exclude all those aged 18 or 19 from the sample, in order to remove distortions
caused by those who did not vote at the July 2002 election because they were not old enough. There
will be a small number of respondents from surveys conducted from August to December 2004 who
had just turned 20 when they were surveyed and were therefore aged less than 18 years in July 2002.

As age is collected in bands rather than as birthdates we have no way of excluding those who fall into
this category, although the number of people in this category is likely to be negligible given the overall
sample sizes.

The table below outlines the sample sizes for the four groups.

Māori non-voters 330  Margin of Error ±5.4

Māori voters 1811 Margin of Error  ±2.3

Non-Māori non-voters  Margin of Error 2309 ±2.0

Non-Māori voters 16418  Margin of Error  ±0.8

*Margins of error expressed for a 50% figure at the 95% confidence level

We have also provided some reworked data from the New Zealand Election Study 2002. This
information is useful because it is more directly targeted at issues of political participation (as opposed
to party preference), although the fact that the combined sample size is considerably smaller (a total of
n=5579 spread over 3 surveys compared with n=20,868) means that it is not able to provide as
substantial a sub-sample size for Māori non-voters).

We are due to provide updated data on equivalent results for the period after the 2005 election in
November 2006. In 2005 the UMR omnibus began being conducted every two weeks instead of every
four weeks. While not all of the questions above are included in every omnibus now the survey has
increased to once every two weeks, we should still have a substantial sample of Māori non-voters to
analyse by November.

*************
Key Research Question: Why is there low Maori electoral participation?
The table below lists the five hypotheses included in the research brief to explain lower participation, along with the evidence that we have found for and against them.

Possible reason 1: Low levels of participation are not about being Māori, they are about poverty and education levels and the different age profile.
Possible reason 1: What the analysis shows? Multivariate analysis clearly supports this hypothesis. Using a very large database of records, we have found that the strongest demographic predictor (out of ethnicity, gender, age, personal income and household income) to be age, followed by gender and household income. For example, 28% of Māori 20-30 year olds in our sample said that they did not vote at the 2002
election, compared with 29% of all 20-30 year olds (12% of all respondents aged 20 years or over in our sample admitted not voting). This hypothesis does not, however, completely explain low turnout, and it is apparent that attitudes towards politics also have a part to play.

Possible reason 2: Māori are strongly engaged in iwi and Māori politics, just not in national elections. 
Possible reason 2: What the analysis shows? 
Māori non-voters are less likely to be enrolled on the Māori roll, and our previous research suggests that this means that they are less likely to be involved in ‘Māori society’. We therefore have no evidence to support this hypothesis.

Possible reason 3: Turnout in Māori electorates is lower because Māori on the Māori roll have strong Māori identity and so are more  involved in iwi politics and not involved in national politic
Possible reason 3: What the analysis shows? 
Māori non-voters are less likely to be enrolled on the Māori roll, and our previous research suggests that this means that they are less likely to be involved in ‘Māori society’. We therefore have no evidence to support this hypothesis.

Possible reason 4:Māori are interested in politics but do not take part because 
they think their voice is not heard or of value. 
Possible reason 4: What the analysis shows? 
Māori non-voters are generally less likely to have strong opinions about political figures. This suggests that the main issue is interest rather than efficacy.

Possible reason 5:Candidates are more important 
than party to Māori voters 
Possible reason 5: What the analysis shows? 
 have no evidence from our polling to prove or disprove this hypothesis. At the same time, the number of party votes cast was higher than the number of electorate votes cast at the 2005 election in each of the seven Māori seats. As our analysis suggests that Māori enrolled on the Māori roll are likely to be more conscious of their Māori heritage, the 2005 election results therefore do not support this hypothesis. On the other hand, our data suggests that Māori enrolled on the Māori roll are more likely to vote than those enrolled on the
general roll, which may be related to the opportunity to vote for a Māori candidate.

Click here for full copy of report http://www.elections.org.nz/sites/default/files/plain-page/attachments/Maori%20Electoral%20Engagement.pdf