A "main" ethnic group? Ethnic self-prioritisation among New Zealand youth.
Kukutai, Tahu ; Callister, Paul
Abstract
Since 1991 a growing share of the New Zealand population has
reported more than one ethnic group in the census, with rates especially
high among children. A key challenge arising from the collection of
ethnicity data is deciding where to count people who record more than
one group. In this paper we explore how a self-prioritised measure of
main ethnicity may facilitate and improve the usage of multiple-ethnic
data. We do so using 2006 data from wave one of the Youth Connectedness
survey of early adolescents. We find that three-quarters of youth who
recorded more than one ethnic group were able to choose a main group
when asked to do so. Though we have reservations about using a main
ethnicity measure to output ethnic data, we see promise for research
that seeks to better understand identification processes and their
relations with ethnic identity and inequality.
INTRODUCTION
The view that race and ethnicity are socially and politically
constructed markers of difference rather than objective traits of human
beings is unremarkable in the social sciences (Omi and Winant 1994). In
other forums, however, the belief in the idea of distinct races
endures--testament to its powerful rendering through legal, bureaucratic
and "scientific" designations, racial ideologies, and everyday
interactions (Callister and Didham 2009). Although the globalisation of
migration flows and the removal of prescriptive identity rules and
classifications have begun to challenge long-held notions that
individuals belong to a single race or ethnic group, change has been
slow to filter through to official statistics.
Among census-taking nations, New Zealand is one of a small number
that explicitly allows for identification with multiple ethnic groups
(Kukutai and Thompson 2007, Morning 2008). Since the introduction of the
ethnic group question in the 1991 census, a growing share of the New
Zealand population has reported belonging to more than one group. As
Table 1 shows, in 1991 just 5% of New Zealanders identified with more
than one ethnic group; by 2006 this had doubled, though the increase has
not been monotonic. In all years multi-ethnic identification has been
especially pronounced among younger people and among Maori and Pacific
peoples. The latter groups are of interest to policy makers, in part
because of their comparative socio-economic disadvantage.
In New Zealand, as in other Anglo settler states (United States,
Canada, Australia), ethnicity and related terms such as "race"
and "indigeneity" are important variables in social research
and policy. Among those who work with ethnicity data in New Zealand
there is a broad consensus that allowing people to choose more than one
group is desirable to best reflect the nation's ethnic milieu
(Didham 2005). However, giving effect to complex ethnic identification
presents a number of challenges in terms of measurement, analysis and
dissemination. How should people who choose to identify with multiple
groups be statistically represented? What weight should be given to
statistical requirements versus individual identification decisions?
What does identification with more than one group even mean? As Bhopal
(2004) notes, there is no easy answer to such questions:
The increasing acceptance of sexual unions that cross ethnic and
racial boundaries is adding both richness and complexity to most
societies. The way to categorise people born of such unions is unclear
and the current approaches are inadequate, partly because the number of
potential categories is huge. (Bhopal 2004:444)
In this paper we explore how a self-prioritised ethnicity measure
may help advance the understanding of complex ethnicity data. Allowing
people to choose a main ethnic group was one of several approaches for
managing multiple-ethnic data identified in the 2004 Report of the
Review of the Measurement of Ethnicity (Statistics New Zealand 2004).
However, with the exception of Kukutai (2004, 2008), little research has
been conducted on ethnic self-prioritisation. We attempt to address this
dearth by exploring whether a main ethnicity prompt delivers useful
information that cannot be captured by the officially sanctioned
methods.
We begin with a discussion about multiple-ethnic identification in
surveys and some ways of reporting and analysing such data. We then
provide an empirical analysis of self-prioritisation using data from the
first wave of Victoria University's longitudinal Youth
Connectedness survey of early adolescents. Only summary data are
presented as a full technical paper by Kukutai (2008) is available on
the Statistics New Zealand website. Given that young people will
significantly influence the nation's ethnic terrain in coming
years, it is valuable to have insights into their identification
decisions. Three questions inform the following analysis:
* Can young people who identify with multiple ethnic groups choose
a main ethnic group when asked to do so?
* If so, what group is prioritised?
* How does a young person's readiness to choose between his or
her ethnicities vary across specific ethnic group combinations?
We conclude with some thoughts about the role that main ethnicity
could play in the future in research and policy making.
MULTIPLE-ETHNIC IDENTIFICATION
Patterns of ethnic identification, including how people are
designated in the census, are important for various reasons. In terms of
policy and planning, ethnicity data are routinely used to identify
population parameters and characteristics, often in ways that influence
the distribution of valued resources. Patterns of ethnic identification
are also of sociological import as they "reflect and affect the
surrounding social world" (Liebler 2004:702). In New Zealand
(Callister 2003, Keddeli 2007, Kukutai 2007) and elsewhere (Brunsma
2005, Roth 2005, Tafoya et al. 2004, Xie and Goyette 1997) there is
ample evidence that the ethnic labels people choose or are designated
are not simply reflections of their parental ethnicities, but are
mediated by a range of factors. These may be categorised as:
* structural--for example, the ethnic composition of the
neighbourhood, ethnic group status differences, and ethnic politics
* personal- for example, life-cycle stage and the ties linking
individuals and their families
* contextual--for example, how, where and why ethnic identification
was elicited (Burton et al. 2008, see also Carter et al. 2009).
The matter of defining who is multi-ethnic is not straightforward.
Goldstein and Morning's (2000) research on the multiple-race
population in the United States suggests at least three ways of
conceptualising a multiple-ethnic population in New Zealand: by
ancestry, by ethnic identification in the census, and by parental
ethnicities. The disconnect between boundaries based on ancestry,
parental ethnicities and self-identification varies, depending on the
groups involved, the context and the time period. To illustrate this
complexity, 643,977 people reported Maori ancestry in the 2006 census,
but just over 80% of them (522,577) identified as Maori by ethnicity. By
comparison, an estimated 7,876,568 people reported American Indian
ancestry in the 2000 US census, but only 4,315,865 people (representing
55% of the American Indian descent population) racially identified as
American Indian (Brittingham and de la Cruz 2004). (2)
The statistical construction of a multiple-ethnic population is
only possible if people are permitted to identify as such (inputs), and
their identification decisions are tabulated in a way that their number
can be determined (outputs). Morning (2008) has noted three possible
ways in which census forms allow for multi-ethnic identification:
permitting the respondent to check off more than one category; offering
a generic mixed-response option that, in effect, creates a single ethnic
category (e.g. "Mestizo"); and, specifying exact combinations
of interest.
In New Zealand, all three approaches have been used to document
complex ethnic and racial identities. From 1874 through to 1921
inter-racial mixing was captured by the use of the
"half-caste" category. (3) The vast majority of half-castes
were half-caste Maori-Europeans, who were further distinguished on the
basis of those who lived as Maori (i.e. in a kin-group village setting)
and those who lived as Europeans. According to the 1921 census report
the total number of half-caste Maori-Europeans was only about one-sixth
of the number of Maori (49,635). Though the tables show half-caste
Europeans (4,236) outnumbered half-caste Maori (3,116), the latter
category was almost certainly underestimated. This is because many Maori
with a European parent or grandparent chose not to acknowledge their
mixed heritage, or were simply identified as Maori (Buck 1924).
The 1926 census introduced a new complexity by requiring
respondents to quantify their heritage more precisely in terms of
fractions. (4) The examples accompanying the question varied over time,
but among the specific combinations named were "European-Indian
quarter-caste" (1926); "1/2 Maori--1/2 Indian" (1945);
and "7/8 European + 1/8 Maori" (1981). From the mid-1980s two
significant changes occurred that affected the reportage of complex
identities. First, the collection of fractional data was abandoned in
the 1986 census and tick-boxes for ethnic origin groups were introduced
with the instruction to "tick the box or boxesthat apply to
you". Second, all references to origins were removed from the
census questionnaire in 1991 and replaced by the term "ethnic
group". The concept of ethnic group is intended to capture a
person's current cultural affiliation rather than the ethnic
origins of their ancestors.
"COUNTING" COMPLEX ETHNICITY
The acknowledgement of complex ethnicity in the New Zealand census
has, in some ways, been less problematic than deciding how such people
ought to be statistically represented. For many decades post-enumeration
rules were used to allocate people who acknowledged their mixed descent
to one race group. In the early period of census data collection it was
standard practice to tabulate racial mixtures in census publications
because officials had a keen interest in "miscegenation" and
what it supposedly represented: the rate of Maori absorption into the
European population (Kukutai forthcoming). However, when comparing Maori
with the general--predominantly European--population, some rule of
designation was needed. Until 1921 half-castes were allocated to the
Maori or European population depending on their mode of living. After
1926 a "half or more" rule was used that allocated those with
half or more Maori blood into the Maori population. This was a
unilateral form of prioritisation as many half-castes could demonstrably
have been counted with Europeans. (5) Paradoxically, though half-caste
denoted an "in-between" statistical category, the allocation
of half-castes to either the Maori or European population served to
solidify the notion of Maori and European as separate races (Kukutai
forthcoming). The system of eliciting data on fractional identities, but
allocating people who recorded "mixed" race or descent to a
single group, continued in some form through to 1981.
Once fractions were dispensed with, new rules of allocation were
needed. Between 1986 and 1991 Statistics New Zealand, as well as most
government agencies and researchers, relied primarily on the
prioritisation of ethnic groups in order to simplify the presentation of
the data. Under this system, Maori had priority coding, followed by
Pacific peoples, then Asian, other ethnic groups besides European,
followed by "Other European" and, finally, New Zealand
European (Allan 2001). (6) Under prioritisation, a person reporting, for
example, as both Maori and Samoan was classified only as Maori. When
prioritisation of ethnic (origin) responses was first introduced in
1986, multiple-ethnic group reporting was under 5%. However, as the
share of the population recording diverse ethnicities grew, so too did
the distorting effect of prioritisation on statistics, particularly for
Pacific peoples (Didham 2005). Despite its drawbacks,
prioritisation-by-proxy is still used in some areas of education and
health research (for a summary of education-related research see Leather
2009; for examples of health research see Chan et al. 2008, Sundborn et
al. 2008).
When undertaking its review of ethnicity in 2004, a number of
outputting options were put to Statistics New Zealand. One was to report
main single and multiple combinations. Other suggested methods, most of
which reduce responses to a single ethnicity, were:
* publish total counts
* randomly allocate multi-ethnic people to a single ethnic category
* use a fractional ethnicity model
* develop a system that can "predict" likely main ethnic
group
* let people choose their own main ethnic group.
The strengths and weaknesses of some of these systems have been
explored in this journal (Callister 2004). (7) Although New Zealand is
unique in having a long history of collecting multiple-race/ethnicity
data, the contemporary challenges involved in dealing with such data are
not uncommon. In the U.S., bureaucrats and researchers face similar
issues with the introduction of multiple-race reporting in the 2000
census. One of the proposed solutions has been to predict
probabilistically the main ethnicity of people who record more than one
race, thereby yielding data that conforms to traditional mutually
exclusive race categories (Liebler and Halpern-Manners 2008).
There are a number of possible reasons underlying the demand for
single ethnic group data. Perhaps the weakest one is that some data
users find ethnic complexity conceptually difficult to deal with and
feel more comfortable when people are placed in seemingly clear-cut
groups. Methodological concerns also have a role. People who affiliate
with more than one group are not readily accommodated within standard
statistical techniques that often require mutually exclusive categories.
This problem is averted when data are collected using some form of
"mixed" single category (e.g. "White and Black
Caribbean" in the UK census). In the absence of such a category,
data users often create mutually exclusive categories by using single
and combination ethnic categories (e.g. Maori, European, and
Maori-European), some form of prioritisation, or simply omitting people
who give complex responses (see Liebler and Halpern-Manners 2008).
The demand for single output ethnicity data may also arise from
political concerns, such as access to resources or political
representation. In the U.S. the decision to allow more than one racial
group to be collected was opposed by groups concerned that it might
decrease the counts of some important minority groups (Korgen 1998). The
desire to maximise the size of particular groups can be particularly
important where resources are at stake. In New Zealand, when
schools' decile funding was still determined, in part, by
enrolments of Maori and Pacific students, schools had an explicit
incentive to maximise the counts of those students. In the health
sector, some funding has an ethnicity weighting based on census data.
For example, Services to Improve Access funding is made available to
primary health organisations to reduce health inequalities by improving
access to primary care services by high-need groups, particularly Maori,
Pacific people and people on low incomes (Ministry of Health, no date).
Maximising counts may also influence wider EEO policies because they set
a baseline target for representation in public and private sector
organisations.
Knowing a "main" ethnicity could be important for a range
of reasons. In New Zealand we have little understanding of why people
record more than one ethnic group, or what such responses are
signalling. Where a historically and socially meaningful
"blended" group has not arisen, it is problematic to assume
people who report two or more groups have an equal sense of affiliation
with all groups (implied by total response), or see it as a unique
blended identity (combination response). Some people may be trying to
accurately report ancestries that are complex, but their lived
experiences may be primarily shaped by their affiliation, or appearance
of belonging to, one group. A self-prioritised main ethnicity prompt
would provide additional information with which to make more nuanced
distinctions.
A main ethnicity designation could also be helpful in clarifying
the relationship between ethnicity and other outcomes of interest.
Because ethnic relations are often hierarchical with regard to the
distribution of power, prestige and resources, it may be useful to be
able to identify people who affiliate more closely with groups that have
a history of disadvantage. Previous research (Chapple 2000, Kukutai
2004, forthcoming) has found that those who have a strong Maori
identification seem to be more disadvantaged than those with
"thinner" ties to Maori identity. Finally, there are some
programmes, such as stop smoking campaigns and mental health programmes,
that are tailored towards particular ethnic groups. Where appropriate,
self-prioritisation data could be used to ensure that such programmes
are more effectively targeted.
ETHNIC SELF-PRIORITISATION AMONG EARLY ADOLESCENTS
Connectedness in Young New Zealanders: Social Connectedness,
Transitions, and Well- being" (the YC survey) is a three-year
survey of early adolescents undertaken by the Roy McKenzie Centre at
Victoria University, aided by the New Zealand Council for Educational
Research. The first wave of data collection was taken in 2006 (n =
2,174); the second wave in 2007 (n = 1,914); and the final wave was
recently completed. The rationale for the YC survey was to collect data
that enabled researchers to explore how connectedness to family, peers,
school and community affect wellbeing in early adolescents. It includes
a suite of ethnicity items relating to ethnic identification, language
use, cultural knowledge and community ties. Of the 78 schools included
in the survey, two-thirds were located in Wellington and Wairarapa; the
remainder were from Kapiti, Taranaki, Hawke's Bay and Auckland.
Initially the YC team aimed for a 50:50 split between New Zealand
European and Maori youth, but this was not achievable. In wave one, the
majority of participants were in Years 6, 8 and 10. A fuller description
of the sample can be found in Kukutai 2008. The selective coverage of
the YC survey means the findings are not nationally representative and
therefore cannot be generalised to all early adolescents in New Zealand.
Nevertheless, it offers a rich source with which to explore the patterns
of ethnic identification and self-prioritisation in ways that are not
possible in nationally representative surveys such as the census.
Table 2 shows the distribution of single, dual and multiple (three
or more) ethnic group responses in the YC survey, as well as for
children aged 10-14 years at the time of the 2006 Census. It shows that
the percentage of youth recording more than one ethnic group in the YC
survey (30%) was double that of 10-14-year-olds in the census (16%). The
recording of three or more ethnic groups was especially pronounced in
the YC survey, but the percentage that failed to record a response to
the ethnic group question was lower.
There are several reasons that may account for the higher
proportion of dual and multi-ethnic reporting in the YC survey,
including:
* the selectivity of participating schools and participants
* the regional concentration of the survey
* the prompt to the ethnic group question, which may have
encouraged the recording of
"symbolic" ethnicities (Gans 1979) alongside those with
which youth held a more meaningful attachment (8)
* the way in which the YC survey was promoted and framed in the
lead-up to being carried out
* self-reporting in the YC survey versus a high likelihood of proxy
reporting in the census (9)
* different modes of data collection--computer-assisted in the YC
survey versus a write-in questionnaire for the census (for a discussion
of questionnaire mode effects, see Dillman and Christian 2005).
Table 3 provides a more detailed breakdown of the ethnic group
responses recorded in both surveys using combination categories. If
standard classification procedures are followed and data are aggregated
at the highest level into broad ethnic groupings or categories, then
combined pan-ethnic responses (e.g. New Zealand European and British;
Samoan and Tongan) are not treated as dual or multi-ethnic. (10) Given
that our key interest is in people who report more than one group, we
have preserved the recording of complex ethnic group responses.
Table 3 shows that, among youth who recorded one ethnic group in
the YC survey, the percentage of European, Maori and Pacific youth
closely resembled the census distributions, whereas Asian and
"other" ethnic groups were under-represented (see Kukutai 2008
for disaggregated data on specific groups and how "other"
responses were categorised). A Maori-European combination was recorded
by 12.5% of YC participants, compared with 8.6% of early adolescents in
the census. However, dual identification was lower overall in the
census. If we restrict the comparison to dual responses, the share of
Maori-European combinations was somewhat higher in the census
(25,311/41,685* 100) than in the YC survey (266/480* 100).
The next-most-common dual combination was for two European
ethnicities--one of which was typically New Zealand European. When some
other European group was recorded it was usually British (e.g. English,
Scottish). Among the children who recorded three ethnic groups, the most
common combination was Maori in conjunction with Pacific and European
ethnic groups. Only slightly fewer youth recorded Maori in combination
with two European ethnic groups. Analysis not shown here found that the
reporting of more than one group also varied significantly, depending on
the group. Of all youth who identified as Maori, two-thirds did so as
part of a dual or multiple ethnic response. For European responses,
however, less than one-third were reported as part of a combined
response. There was no significant age difference in the number of
ethnic groups reported, and a small gender difference, with boys more
likely than girls to record just one ethnic group (73% versus 68%, p
< .05 level of significance).
We now turn to our primary interest: whether dual and multi-ethnic
youth in the YC survey were willing and/or able to self-prioritise a
main ethnic group when asked to do so. The focal question asked:
"If you belong to more than one ethnic group, do you have a main
ethnic group? Which is the main ethnic group you belong to?" In
addition to the nine categories given in the ethnic group question, the
following responses were also provided: "I belong to just one
ethnic group", "I have no main ethnic group", "It
depends on who I am with", and "Don't know". The
question did not immediately follow the ethnic group question, but
followed a series of items related to how participants felt about their
nominated ethnic group(s). For the 641 youth who recorded at least two
ethnic groups, their prioritisation responses are shown in Table 4.
The key finding is that three-quarters of dual and multi-ethnic
youth in the survey were able to self-prioritise a main ethnic group
when prompted. (11) The share of participants who indicated they did not
know, or did not have, a main ethnic group was fairly similar at 8.0%
and 9.8%, respectively. In the absence of cognitive testing, the
distinction between not knowing and not having is unclear. Whereas the
former implies lack of knowledge or indecision, the latter suggests a
more conscious rejection of having to choose. However, it may be that
not knowing and not having a main ethnic group are simply different ways
of articulating discomfort with having to choose between elements of
one's ethnic identity. Non-prioritisation may also denote feeling
an equal sense of belonging to two or more groups, or occupying a kind
of "third space" (Bhaba 1990) where new identities or
"hybridities" are forged.
Just under 3% explicitly indicated that their main ethnic group was
situational, depending on who they were with. This was considerably
lower than shifts in ethnic identification found in U.S. surveys (Harris
and Sim 2002), but is not surprising. People who change their
identification in different contexts may not necessarily be aware of
this, nor respond in such a way when asked about it directly. As
expected, very few dual and multi-ethnic youth stated that they belonged
to just one group. Finally, almost 5% of dual and multi-ethnic youth did
not record a response to the main ethnic group question, which is
notably higher than the proportion who skipped the ethnic group question
(1.6%). There is no way of knowing why participants did not respond to
the prioritisation prompt.
Ideally it would be useful to know how prioritisation responses
varied across specific ethnic group combinations. Unfortunately the
modest sample size, and the predominance of Maori--European responses,
precluded detailed analysis. Given these limitations, Table 5a simply
shows the percentage of youth who were able to prioritise for the five
largest combinations. Table 5b extends the analysis to specify which
group was prioritised in European-Maori combinations, in other dual
responses, and in three or more responses. The small numbers preclude a
detailed analysis of the non-prioritised responses, such as "no
main" or "don't know" for each combination.
The ability to self-prioritise was highest among youth who reported
dual affiliations that included Maori, and lowest among European-Pacific
youth. However, the small number in four of the combined categories, and
the lack of statistical significance, means the results are indicative
only. The willingness and/or ability to self-prioritise was somewhat
higher for participants who recorded three or more ethnic groups than
for those who recorded a dual combination other than European and Maori.
Although one may expect that choosing a main group would be more
difficult the greater the number of ethnic groups an individual
identified with, it may be that third or fourth ethnicities are largely
"symbolic" (Gans 1979) in terms of holding meaning in everyday
life.
Table 5b shows that, of the youth who identified as both Maori and
European and selfprioritised, the majority chose European over Maori,
with a difference of about 10 percentage points. Interestingly, Kukutai
(2004) found a similar pattern of self-prioritisation responses among
women identified as Maori and European in the 1995 New Zealand Women:
Family, Education and Employment (NZW: FEE) survey of women aged 20-59
years. Among the 183 women who recorded Maori and European ethnic
groups, 42% identified mainly as European, 37% identified mainly as
Maori, and the remainder could not choose. (12) Among those who did
self-prioritise, the European/Maori split was 54/46 - remarkably close
to the distribution of responses in the YC survey.
Some may interpret these findings as evidence of a bias towards
European ethnicities among people with both Maori and European heritage.
Such an interpretation would be misleading. In the first instance, our
analysis is based on self-identification rather than ancestry or
parental ethnicities. As we noted earlier, these concepts are not
synonymous. Previous parent-child studies suggest that some youth who
identified as Maori only or European only in the YC survey would be
considered multi-ethnic on the basis of their parental ethnicities
(Callister 2003, Howard and Didham 2005, Kukutai 2007). All of those
studies found that children with one Maori and one European parent were
more likely to be designated as Maori--only than as European only. (13)
Historically, the pattern has been for people of MaoriEuropean heritage
to simplify their ethnic self-identification to Maori (Buck 1924, Metge
1964, Pool 1991). These complex patterns highlight the need to be clear
about the conceptual basis underlying the use of concepts such as
"mixed" and "multi-ethnic".
SUMMARY AND CONCLUSION
This research was undertaken with a view to exploring how ethnic
self-prioritisation may help us better understand the complexity of
multiple-ethnic data and its potential usage as an output method. The
key finding was that almost three-quarters of youth who identified with
more than one ethnic group in the YC survey were willing and able to
choose a main group when asked to do so. This was in spite of the
inclusion of viable alternatives which gave participants every
opportunity to opt out of choosing.
Because Maori-European reports comprised the vast majority of dual
ethnicity responses, we were particularly interested in which group was
most often prioritised. Of the dual identified Maori-European children
who could choose a main group, more chose European over Maori. This
would have several implications if used as a method of outputting
ethnicity data. To illustrate, if the YC pattern of prioritisation
responses were applied to the 2006 census data for all ages, the Maori
Ethnic Group count would be reduced from 565,329 to 395,051. (14) In the
context of population-based funding, this would decrease the portion
allocated to Maori, and decrease the size of the Maori population
relative to Europeans. Substantively, this is similar to the effect the
old system of prioritisation-by-proxy had on non-Maori groups, notably
Pacific peoples, but which continues to be used in some areas of health
and education research. As an output method, self-prioritised data would
result in the loss of information that may be important in research and
policy making. However, arguably this is what currently occurs with
total count data in which multiple-ethnic identification is
"hidden". There is also the matter of how to allocate those
who cannot or will not choose a main ethnic group. In the YC survey
about one-fifth of the youth who reported more than one group either did
not know, or did not have, a main ethnic group. Not being able to choose
is a valid response that would need to be accounted for conceptually and
statistically.
The potentially significant effects that self-prioritisation would
have on ethnic group counts may have political implications. For
example, some people may feel that a main ethnicity measure is a tool
designed to reduce the count of Maori. As Kukutai and Didham (2009) note
elsewhere in this issue, ethnicity and ethnic groups are created and
sustained through intergroup processes that, at times, involve
competitive struggles for political power and material resources.
However, it is for methodological and substantive (rather than
political) reasons that we do not think it appropriate to use
self-prioritisation as a data reduction tool in isolation from other
outputting options.
Our analysis has focused on the implications of a main ethnicity
prompt for outputting purposes, but it also raises the question of how
to deal with ethnic complexity at the input stage. When filling in paper
or computer surveys we do not know what people intend their response to
mean. For example, some youth who checked the Maori and New Zealand
European boxes in the YC survey may see both groups as reasonably
distinct. In such cases, self-prioritisation could signal the group with
which youth felt a stronger connection or affiliation. Others recording
the same two groups may see their response as denoting a blended group
distinct from either Mfiori or European. Whether this group is
Maori--European or European-Maori might be showing up in the
self-prioritisation data. In-depth qualitative research would be needed
to understand how complex ethnicity is being constructed for
adolescents.
Although self-prioritisation is unlikely to receive wide support as
a method for reducing the complexity of multiple-ethnic data, it could
have some important uses for expanding complexity. For example, previous
analysis of main ethnicity data has shown important socio-economic and
demographic differences between people who identified as both Maori and
European but self-prioritised as Maori versus European (Kukutai 2004).
Given the empirical relationship that has consistently been demonstrated
between ethnicity and socioeconomic disadvantage in New Zealand, the
information delivered by a main ethnic group question may be valuable
for better specifying the association between ethnicity and
socioeconomic outcomes that a good deal of policy research is concerned
with.
The limited scope of this paper means that its chief contribution
has been to describe ethnic identification patterns rather than to
identify the factors underlying response patterns, or to explain what
responses to a main ethnic group question may mean. Fortunately, the
inclusion of the main ethnicity prompt in all waves of the YC study,
along with a raft of questions relating to different kinds of ethnic
attachment, means these questions can be pursued within the period of
adolescence. Understanding how ethnicity may change over an
individual's life cycle would not only require longitudinal data
from birth to death but also regular questioning about ethnic
affiliation. Future research that attempts to better specify the
relationship between ethnic identification, ethnic identity and
attachment, and stratification will be valuable for advancing the
understanding of ethnic identification dynamics and the purposes for
which ethnic data can best be used.
REFERENCES
Allan, J. (2001) Classification and Issues: Review of the
Measurement of Ethnicity, Statistics New Zealand, Wellington.
Bhaba, H. (1990) "The third space: Interview with Homi
Bhaba" in J. Rutherford (ed.) Identity, Community, Culture,
Difference (pp. 207-221), Lawrence and Wishart, London.
Bhopal, R. (2004) "Glossary of terms relating to ethnicity and
race: For reflection and debate" Journal of Epidemiology and
Community Health, 58:441-445.
Brittingham, A. and G. de la Cruz (2004) Ancestry 2000: Census 2000
Brief U.S. Census Bureau, Washington, D.C.
Brunsma, D. (2005) "Interracial families and the racial
identification of mixed-race children: Evidence from the early childhood
longitudinal study" Social Forces, 84:1131-1157.
Buck, P. (1924) The Passing of the Maori: Extract from Transactions
and Proceedings of the New Zealand Institute, Government Printer,
Wellington.
Burton, J., A. Nandi and L. Platt (2008) Who Are the UK's
Minority Ethnic Groups? Issues of Identification and Measurement in a
Longitudinal Study, Working Paper 2008-26, Institute for Social &
Economic Research, University of Essex.
Callister, P. (2003) "The allocation of ethnicity to children
in New Zealand: Some descriptive data from the 2001 Census" paper
presented at the Population Association of New Zealand Conference, 3-4
July, Christchurch, www.callister.co.nz
Callister, P. (2004) "Maori/non-Maori ethnic
intermarriage" New Zealand Population Review, 29(2):89-105.
Callister, P. and R. Didham (2009) "Who are we?: The Human
Genome Project, race and ethnicity" Social Policy Journal of New
Zealand, 36.
Carter, K.N., M. Hayward, T. Blakely and C. Shaw (2009) "How
much and for whom does self-identified ethnicity change over time in New
Zealand?: Results from a longitudinal study" Social Policy Journal
of New Zealand, 36.
Census and Statistical Office (1927) Results of a Census of the
Dominion of New Zealand, Government Printer, Wellington.
Chan, W.C., C. Wright, T. Riddell, S. Wells, A.J. Kerr, G. Gala and
R. Jackson (2008) "Ethnic and socioeconomic disparities in the
prevalence of cardiovascular disease in New Zealand" New Zealand
Medical Journal, 121(1285):11-20.
Chapple, S. (2000) "Maori socio-economic disparity"
Political Science, 52(2):101-15.
Didham, R. (2005) Understanding and Working with Ethnicity
Data." A Technical Paper, Statistics New Zealand, Wellington.
Dilhnann, D. and L. Christian (2005) "Survey mode as a source
of instability in responses across surveys" Field Methods,
17:30-52.
Gans, H. (1979) "Symbolic ethnicity" Ethnic and Racial
Studies, 2:1-20.
Goldstein, J. and A. Morning (2000) "The multiple-race
population of the United States: Issues and estimates" Proceedings
of the National Academy of Sciences, 97:62306235.
Harris, D. and J. Sim (2002) "Who is multiracial? Assessing
the complexity of lived race" American Sociological Review,
67:614-627.
Howard, S. and R. Didham (2005) Ethnic Intermarriage and Ethnic
Transference amongst the Maori Population: Implications for the
Measurement and Definition of Ethnicity, Statistics New Zealand,
Wellington.
Keddell, E. (2007) "Cultural identity and the Children, Young
Person's and Their Families Act 1989: Ideology, policy and
practice" Social Policy Journal of New Zealand, 32:49-71.
Korgen, K.O. (1998) From Black to Biracial: Transforming Racial
Identity among Americans, Praeger, Westport, CT.
Kukutai, T. (forthcoming) "Building brown boundaries in New
Zealand: Representations of Maori identity in the Census" in P.
Skol and P. Axelsson (eds.) Indigenous Peoples and Demography: The
Complex Relation between Identity and Statistic (chapter 2), Berghahn
Books, New York and Oxford.
Kukutai, T. (2004) "The problem of defining an ethnic group
for public policy: Who is Maori and why does it matter?" Social
Policy Journal of New Zealand, 23:86-108.
Kukutai, T. (2007)"White mothers, brown children: Ethnic
identification of Maori-European children in New Zealand" Journal
of Marriage and the Family, 69:1150-1162.
Kukutai, T. (2008) Ethnic Self-Prioritisation of Dual and
Multi-Ethnic Youth in New Zealand: A Discussion Paper, Statistics New
Zealand, Wellington.
Kukutai, T. and R. Didham (2009) "In search of ethnic New
Zealanders: National naming in the 2006 Census" Social Policy
Journal of New Zealand, 36.
Kukutai, T. and V. Thompson (2007) "Doing diversity 'down
under': New Zealand's ethnic enumeration practices in global
perspective" paper presented at the Population Association of New
Zealand Conference, 3-4 July, Te Papa Museum, Wellington, New Zealand.
Leather, F. (2009) Prioritising Data in an Increasingly
Multi-ethnic Society, IPS Working Paper WP09/05, Institute of Policy
Studies, Wellington.
Liebler, C. (2004) "Ties on the fringes of identity"
Social Science Research, 33:702-723.
Liebler, C.A. and A. Halpern-Manners (2008) "A practical
approach to using multiple-race response data: A bridging method for
public-use microdata" Demography, 45( 1 ): 143155.
Mays, V.M., N.A. Ponce, D.L. Washington and S.D. Cochran (2003)
"Classification of race and ethnicity: Implications for public
health" Annual Review of Public Health, 24:83110.
Metge, J. (1964) A New M6ori Migration: Rural and Urban Relations
in Northern New Zealand, Berg Publishers, Oxford, UK.
Ministry of Health (no date) Primary Health Care: Services to
Improve Access, Ministry of Health, Wellington.
Morning, A. (2008) "Ethnic classification in global
perspective: A cross-national survey of the 2000 Census round"
Population Research and Policy Review, 27(2):239-272.
Omi, M. and FI. Winant (1994) Racial Formations in the United
States: From the 1960s to the 1990s (2hd ed.), Routledge & Keegan
Paul, New York.
Pool, I. (1991) Te Iwi Maori, Auckland University, Auckland.
Roth, W. (2005) "The end of the one-drop rule? Labeling of
multiracial children in black intermarriages" Sociological Forum,
20:35-67.
Statistics New Zealand (2004) Report of the Review of the
Measurement of Ethnicity, Statistics New Zealand, Wellington.
Statistics New Zealand (2005) Statistical Standard for Ethnicity
2005, Statistics New Zealand, Wellington.
Sundborn, G., P.A. Metcalf, D. Gentles, R.K.R. Scragg, D. Schaaf,
L. Dyall, P. Black and R. Jackson (2008) "Ethnic differences in
cardiovascular disease risk factors and diabetes status for Pacific
ethnic groups and Europeans in the Diabetes Heart and Health survey
(DHAH) 2002-2003" New Zealand Medical Journal, 121 (1281):28-39.
Tafoya, S., H. Johnson and L. Hill (2004) Who Chooses to Choose
Two: Multiracial Identity of Children with Parents of Different Races,
Russell Sage Foundation and Population Reference Bureau, Washington,
D.C.
Xie, Y. and K. Goyette (1997) "The racial identification of
biracial children with one Asian parent: Evidence from the 1990
Census" Social Forces, 76:547-70.
Tahu Kukutai (1)
Stanford University
Paul Callister
Associate Professor
Institute of Policy Studies
Victoria University of Wellington
(1) Acknowledgements We would like to acknowledge the support of
the Youth Connectedness project leaders, Paul Jose and Jan Pryor, who
kindly granted us access to the YC data. as well as the project's
Maori research advisor, Wally Penetito. The YC project is carried out by
the Roy McKenzie Centre, Victoria University, and funded by the
Foundation for Research. Science and Technology. We also wish to thank
Deborah Potter of Statistics New Zealand for her support in instigating
this research, as well as Robert Didham for providing us with the data
shown in Table I. We also appreciate Melinda Webber's readiness to
share with us the preliminary results from her Ethnicity, and
Achievement Survey. Finally, the comments provided by two anonymous
reviewers on an earlier draft of this paper were also very helpful. Any
omissions or errors of fact are ours alone.
Correspondence All correspondence can be directed to Tahu Kukutai
at tkukutai@tahatu.co.nz
(2) In tact the proportion is almost certainly lower because some
people who reported American Indian race would not have reported
American Indian ancestry.
(3) Quotation marks are used on first mention to indicate that the
term is not a neutral descriptor, but are omitted in subsequent
mentions, in keeping with the historical usage of the term.
(4) Though the term "race" was replaced by the language
of "descent" on the 1926 census questionnaire, both terms
continued to be used interchangeably up until the 1976 census.
(5) Maori with non-European heritage (e.g. Indian-Maori"
Polynesian-Maori) were subject to a variety of allocation rules. From
1916 to 1951 they were allocated to the "race alien"
population, regardless of the reported degree of Maori blood: from 1956
such people were subjected to the usual half or more rule, except when
the non-European race was Polynesian. In the latter case, individuals
with Maori and any degree of Polynesian descent were counted only in the
Maori population from 1951 until the 1966 census.
(6) Prioritising data has not been unique to New Zealand. For
example, Mays et al. (2003) set out a variety of ways that US agencies
have prioritised multi-race/ethnic data when it has been available.
(7) In addition to these proposed methods, there has also been some
discussion on whether a measure of "cultural strength" could
be developed using a range of variables (e.g. language use. tribal
affiliation) collected in surveys.
(8) The wording of the ethnicity question in the YC survey was the
same as for the census, but was preceded by the prompt: "Every
person is part of an ethnic group, sometimes two or more ethnic groups.
Some names of ethnic groups are: Samoan, Chinese. Maori. Tongan, New
Zealand European."
(9) Although there is no way of knowing who completes an
individual's census form, researchers have often worked on the
assumption that children under the age of 15 years (or sometimes under
18 years) have their form completed for them by a third person,
typically a parent or caregiver (Brunsma 2005, Roth 2005).
(10) The Standard Classification of Ethnicity is a four-tier
representation of the nation's ethnic composition, with level 1
representing the simplest form and level 4 the most complex (Statistics
New Zealand 2005). Level 1 comprises six categories: European, Maori,
Pacific Peoples, Asian, Middle Eastern, Latin American, African (MELAA),
and Other Ethnicity: for economy and clarity, MELAA is subsumed in the
Other ethnicity category in this paper.
(11) We note that this result aligns with the results from another
recent survey of New Zealand youth. The Ethnicity and Achievement Survey
of Year 9 students was undertaken in 2008 by Melinda Webber as part of
her doctoral dissertation research. Of the 756 youth surveyed, 227 (30%)
reported more than one group, of which 183 (81%) reported a main ethnic
group (M. Webber, preliminary findings from the Ethnicity and
Achievement Survey. 2008, personal communication, 2009).
(12) The main ethnic group question immediately followed the ethnic
group question. Women who reported more than one group were asked:
"Please tell me which one of these is the main ethnic group you
identify with?" Possible responses included "'more than
one" and a combined "don't know/no".
(13) Kukutai's study (2007) used data from the NZW:FEE in
which mothers reported their child's ethnicity. Callister (2003)
and Howard and Didham (2005) used census data. which makes it impossible
to know who reported the child's ethnicity, though it is highly
unlikely that most children (0-14 years) would be given the opportunity
to self-identify in this particular collection. In addition, in the
census data we do not know whether the parents chose only one ethnic
group for themselves when their own ancestry may have been more complex.
Therefore census-based studies on the transmission of ethnicity to
children give us only partial information, and more detailed
ethnographic studies are needed to understand the transmission of
ethnicity between generations.
(14) The number of people who reported more than one ethnic group
including Maori was 266,934, of whom 119.803 would be counted as
European, 96,656 as Maori, 2,165 as some other non-Maori, non-European
grouping, and the remainder would be unable to be allocated to a single
group (i.e. did not prioritise). The number who identified as mainly
Maori would then be added to the number of single-ethnic Maori (n =
298,395).
Table 1 More Than One Ethnic Group Reported, Census of Population
and Dwellings, 1991-2006
More than one ethnicity reported
Census N
year
Total NZ, Total NZ, Maori,
all ages 0-14 all ages
1991 166,158 77,172 111,357
1996 536,757 181,338 249,933
2001 324,090 145,194 231,555
2006 400,428 164,262 266,934
More than one ethnicity reported
Census % (1)
year
Total NZ, Maori, all
Total NZ 0-14 ages
1991 5.0 19.3 25.6
1996 15.5 (2) 45.2 47.8
2001 9.0 34.2 44.0
2006 10.4 38.1 47.2
(1) Percentage of people with a valid ethnic group response.
(2) In 1996 an "Other European" tick-box was included, along
with a sub-list that specified English, Irish, Australian,
Scottish and Dutch ethnic groups. This led to an increase in
the reporting of those groups. The tick-box was dropped from
subsequent census questionnaires.
Table 2 Single, Dual and Multiple Ethnic Group Reporting in
Wave 1, Youth Connectedness survey, 2006, and 2006 Census of
Population and Dwellings, Youth aged 10-14 years
YC survey 2006 census
Number of ethnic N % N %
groups reported
One 1,492 68.7 244,854 80.1
Two 480 22.1 41,688 13.6
Three or more 161 7.4 8,214 2.7
Don't know 5 0.2 351 -
Not stated 35 1.6 10,899 3.6
Total 2,174 100.0 306,006 100.0
Note: "-" indicates figure is too small to be expressed.
Table 3 Ethnic Groups Reported in Wave 1, Youth Connectedness survey,
2006, and 2006 Census of Population and Dwellings
YC survey 2006 census
Number % Number %
One ethnic group
European 1,049 49.1 147,501 50.0
Maori 217 10.2 30,969 10.5
Pacific 146 6.8 17,826 6.0
Asian 40 2.0 23,406 7.9
Other ethnic group 40 1.9 25,152 8.5
Two ethnic groups
European & Maori 266 12.5 25,311 8.6
European & European 60 2.8 1,587 0.5
European & Pacific 36 1.7 3,765 1.3
Maori & Pacific 25 1.2 2,970 1.0
Other dual combinations 93 4.4 8,052 2.7
Three ethnic groups
European, Maori & 25 1.2 2,637 0.9
Pacific
Two European & Maori 22 1.0 2,163 0.7
Other combinations 64 3.0 1,734 0.6
At least four ethnic
groups
Four 38 1.6 1,206 0.4
Five or more 12 0.7 477 0.2
Total 2,134 100.0 294,756 100.0
Table 4 Ability to Self-prioritise Ethnic Group, Youth that Reported
More than One Ethnic Group, Wave 1, Youth Connectedness survey, 2006
Self-prioritisation response N %
Able to choose main ethnic group 474 74.0
No main ethnic group 63 9.8
Depends on who with 17 2.7
Don't know main ethnic group 51 8.0
Belongs to just one 7 1.1
No response stated 29 4.5
Total people 641 100.0
Table 5a Self-prioritisation of Youth who Reported More Than One
Ethnic Group, Five Largest Combinations, Wave 1, Youth Connectedness
survey, 2006
Prioritised a main Total
ethnic group
Combinations N % N *
European & Maori 208 81.9 254
European & European 43 74.1 58
European & Pacific 21 67.7 31
Maori & Pacific 20 83.3 24
European, Maori & Pacific 20 80.0 25
Total 312 ... 393
* Excludes no response stated (n = 20).
Note: "..." indicates not applicable.
Table 5b Self-prioritisation of Youth who Reported More Than One
Ethnic Group, Select Combinations, Wave 1, Youth Connectedness
survey, 2006
Self-prioritisation response N %
European & Maori
European 114 54.8
Maori 92 44.2
Non-Maori, non-European 2 1.0
208 100.0
Other dual combinations
European 80 54.1
Maori 19 12.8
Non-Maori, non-European 49 33.1
148 100.0
Three or more ethnic groups
European 49 41.5
Maori 38 32.2
Non-Maori, non-European 31 26.3
Total 118 100.0
Note: n = 474.
