Kōtuitui
New Zealand Journal of Social Sciences Online
Pacific Islands Families Study: factors associated with living in
extended families one year on from the birth of a child*
Michelle Poland
Families Commission
Level 5, AMI House
63 Albert Street
Auckland, New Zealand
michelle.poland@nzfamilies.org.nz
Janis Paterson
janis.paterson@aut.ac.nz
Sarnia Carter
Wanzhen Gao
Lana Perese
Auckland University of Technology
Private Bag 92006
Auckland 1142, New Zealand
Steven Stillman
Motu Economic & Public Policy Research
PO Box 24390
Wellington, New Zealand
stillman@motu.org.nz
Abstract Using data from the “Pacific Islands
Families: the first two years of life (PIF) study”, this paper explores
the factors associated with the living arrangements of mothers with a
one-year-old Pacific child. Three living arrangements are considered:
(1) those children who live with immediate family members only; (2)
children who live with extended family members including at least one
of their grandparents; and (3) children who live with extended family
members but do not live with any of their grandparents. Findings of
interest include: mothers who were born in New Zealand are much more
likely to live in an extended family with the child’s grandparent(s)
compared to those who came to New Zealand as an adult; and mothers with
high New Zealand identity and low Pacific identity being more likely to
live with extended family members other than the child’s grandparent(s)
relative to those with low New Zealand and high Pacific identity.
Keywords living arrangements; Pacific Islanders;
New Zealand
Introduction
As one of the fastest growing population subgroups in New Zealand,
Pacific peoples (those residents with a Pacific Islands heritage) form
an integral part of New Zealand society (Cook et al. 2001). In the 2000
census, over 230,000 people were of Pacific ethnicity, making up 6.5%
of the New Zealand population (Statistics New Zealand 2002). The
Pacific population in New Zealand is diverse. Samoan people make up the
largest group (49.6%), followed by Cook Island Maori (22.7%), Tongan
(17.6%), Niuean (8.7%), Fijian (3.0%), Tokelauan (2.7%) and Tuvalu
Islanders (0.8%)1. This ethnic diversity
is manifest in differing cultures, languages and access to (and
utilisation of) health and social services.
New Zealand has a long history of Pacific migration with many people
migrating from the Pacific Islands during the 1960s (Meleisea &
Schoeffel 1998; Bedford & Didham 2001). Migration between Pacific
Island groups and countries around the Pacific was fuelled by the
search for employment and a higher standard of living (Macpherson
1991), resulting in 400,000 people of Pacific ethnicity living in rim
countries of the Pacific by the mid 1990s (Ward 1996). However, Pacific
people became significantly socio-economically disadvantaged in New
Zealand society after the reforms in the early 1990s. For example, in
1987–88, 23.4% of the economic family units2
that contained a Pacific adult were in poverty3
compared to 13.8% of all economic family units, and, by 1992–93, this
gap had widened drastically to 50% of all economic family units that
contained a Pacific adult compared to 27.9% of all economic family
units (Ministry of Social Development 2005).
The employment rate and labour force participation of Pacific people
is lower than in the total population, as well as the annual median
income. Pacific people are more likely to have restricted access to
higher education, home ownership, and functional amenities such as
automobiles and telephones. They are also over represented in multiple
adverse health and social statistics (Bathgate et al. 1994; Statistics
New Zealand 2002) and much more likely to be living in a crowded
household (Ministry of Social Development 2005).
While most New Zealand families live in a nuclear family
household—that is, parent(s) and their children—a significant minority
live with extended family members. In 1996, there were 67,068 extended
family households which made up 7.1% of all families (Statistics New
Zealand 1998). These extended family households can be separated into
three main types: (1) those that contain only one generation (e.g.
siblings, cousins); (2) those that contain two generations (e.g.
siblings, cousins and their children); and (3) those that contain three
or more generations of family members (e.g. grandparent(s), parent(s)
and children). The most common extended family household type in New
Zealand in 1996 contained three or more generations (49.2%) followed
closely by those containing two generations (43.6%) (Statistics New
Zealand 1998). In New Zealand, Pacific people are much more likely to
live in a household with extended family members than the general
population. In the 1996 census, 41% of Pacific people lived in extended
families compared to 23% of Maori, 23% of Asian and 6% of European
people (Statistics New Zealand 1998).
Living with extended family members can prove to be beneficial in a
number of areas. It can reduce living costs, particularly housing
costs; allow for greater specialisation within the household in such
areas as housework, meal preparation and labour market participation;
and provide a supportive environment. Families with children can
particularly benefit from living with extended family members. Research
shows that grandparents may have a direct effect on child development
as sources of attachment, affection and knowledge, as well as an
indirect effect on children through their support for parents which
reduces parental stress and allows parents to spend more time with
their children (Tinsley & Parke 1984). Research on extended
families in Korea found that due to these factors, grandparents may
increase the resilience of the children in extended families (Jin Hwang
& St James-Roberts 1998).
However, extended family living can also have negative impacts on
the family. Household crowding, in particular, is often associated with
negative health outcomes for members of the household (Gray 2001). In
three-generation extended family households, role conflicts between the
children’s grandmothers and mothers may also result in adverse
consequences for both parent and child development (Unger & Cooley
1992; Spieker & Bensley 1994).
Gordon et al. (2004) investigated the causal relationship between
extended family living and lower parenting confidence as measured in a
sample of youth mothers with premature low-weight babies. They found
that the mothers with low parenting competence were more likely to
choose to live with family members rather than those living with family
members developing a low level of parenting competence.
New migrants are a group of people who often live with extended
family members. Blank & Torrecilha (1994) argue that for Latino
immigrants to the United States (US), this is primarily related to
lifecycle rather than financial or cultural reasons. They found that
those who had young children or older adults who needed care were much
more likely to live with extended family members.
Pérez (1994) investigated the family living arrangements of a
sample of eighth and ninth graders who had at least one parent who was
not born in the US and who had themselves either been born in the US or
had lived in the US for at least five years. The author used a logistic
multivariate regression using the presence of the child’s grandparents
in the household as the dependent variable. It was found that social
marital status, mother’s education, father’s socio-economic status
index and Cuban origin were all significant in explaining this extended
family living arrangement. The author concluded that the conditions of
migration and the socio-economic status of the family may be a more
important factor than culture and the extent of assimilation.
The aim of the present study is: (1) to examine the prevalence of
living within an extended family one year after the birth of a Pacific
child; and (2) to describe the factors associated with living in
extended families.
Methods
The data used in this study are drawn from the Pacific Islands
Families: first two years of life (PIF) study. The PIF study is a
longitudinal investigation of a cohort of 1398 infants born at
Middlemore Hospital, South Auckland, during the year 2000. Middlemore
Hospital was chosen as the site of recruitment of the cohort as it has
the largest number of Pacific births in New Zealand and is
representative of the major Pacific ethnicities. All potential child
participants were selected from live births at Middlemore Hospital
where the child had a least one parent who identified themselves as
being of a Pacific ethnicity and who was also a New Zealand permanent
resident. All procedures and interview protocols had ethical approval
from the National Ethics Committee.
Information concerning the current and recent circumstances of the
families was obtained through interviews conducted at three points of
time: shortly after birth; at 12 months of age; and at 24 months of age
(Paterson et al. 2006). Only the data from the 12-month interviews with
the mothers4 have been used in this study
because in that wave, mothers who lived with extended family members
were asked why they chose to live in an extended family.
Information collected from the mother and used in our analysis
included age, ethnicity, parity5,
religion, English speaking competency, social marital status, education
and cultural orientation. Cultural orientation was measured with a
modified version of the General Ethnicity Questionnaire developed by
Tsai et al. (2000). This measure is based on the widely used concept of
“acculturation”, the process used to refer to the changes that groups
and individuals undergo when they come into contact with another
culture (Berry & Kim 1988). Berry (1980) identified four different
varieties of acculturation: “assimilation”, “integration”, “separation”
and “marginalisation”. In assimilation, cultural identity is
relinquished and the individual moves into the larger society. We
categorised mothers based on their acculturation scores for affinity
toward New Zealand mainstream culture or identification with Pacific
Islands. Hence “assimilated” mothers had high New Zealand
identification and low Pacific identification. Integration involves
maintenance of cultural integrity but also the movement to become an
integral part of the larger society (high New Zealand, high Pacific).
Separation refers to self-imposed withdrawal from the larger society
(high Pacific, low New Zealand) and marginalisation refers to losing
the essential features of one’s culture, but not replacing them by
entering the larger society (low Pacific, low New Zealand).
Results
Ninety-six percent (N = 1590) of potentially eligible mothers of
Pacific infants who had been born between 15 March and 17 December 2000
gave consent to be visited in their homes when the infant was six weeks
old. Of the 1477 mothers contacted and who met the eligibility
criteria, 1376 (93.2%) agreed to participate in the study. A more
conservative recruitment rate of 87.1% includes mothers who consented
to contact and were either confirmed eligible or of indeterminable
eligibility due to the inability to trace. This recruitment and
beginning of the cohort should remain in the sample but the next part
should be the 12-month cohort.
A second wave of interviews took place when the children were 12
months old. Eighty-eight percent (N = 1207) of the primary respondents
interviewed in the first wave were interviewed again in the second wave
and 17 new respondents participating resulted in a total of 1224
respondents for the 12-month interviews. Of those interviewed in both
waves, 34% were New Zealand born, 48% self-identified as Samoan, 20% as
Tongan, 17% as Cook Island Maori, 5% as Niuean, 3% as Other Pacific and
7% as Non Pacific. The Other Pacific group includes mothers identifying
equally with Pacific and Non-Pacific groups and those identifying with
Pacific groups other than Samoan, Tongan, Cook Island Maori or Niuean.
The Non-Pacific group refers to mothers of infants fathered by Pacific
men. The mean age of mothers was 30.1 years (SD = 6.1), 81% were
married or in de-facto partnerships and 32% had post-school
qualifications.
The analysis in this paper are presented in two parts. The first
part uses cross tabulations to describe the proportion of mothers
living in extended families by various characteristics of the mothers.
This allows an initial description of the sample and the living
arrangements of the members. The second part uses a multinomial
logistic regression to assess the association of various factors with
living arrangement while controlling for possible confounding factors.
The variables examined are the mothers’ age, ethnicity, parity, social
marital status, stage of arrival in New Zealand (whether New Zealand
born, came to New Zealand as a child or came to New Zealand as an
adult), religion, whether the mother speaks English fluently or not,
acculturation and education6.
Two notable variables which are excluded from our analysis are
income and employment status. While a family’s income is likely to
influence their choice of living arrangement, their choice of living
arrangement is also likely to influence their income. For example, a
couple living with extended family members may be doing so in order to
reduce costs given a low income. Or it may be that living with extended
family members allows a family to choose to earn little. In addition,
the PIF Study only collects information on weekly household income
which is directly related to whether an individual lives in an extended
family household as these households typically contain a greater number
of working-age people. Employment status is also endogenously
determined with living arrangements. Extended family members living in
the same household may provide low-cost childcare which allows mothers
to be employed, while preferences towards employment may influence the
choice of family living arrangements. We control for educational
attainment in our regression analysis which captures some of the effect
that income and employment status have on living arrangements, as
educational attainment is positively correlated with both income and
employment (Maani & Maloney 2004).
Half of the members in the sample live in a nuclear family and the
other half live in an extended family household. Of those who reported
living with extended family members, 61% live in a household in which
at least one of the child’s grandparents are present compared to 39%
living in an extended family household in which none of the child’s
grandparents are present.
Table 1 displays a cross tabulation of mothers’ characteristics by
family type. Two family household types are distinguished: a nuclear
family household and an extended family household. Extended family
living is shown to be more prevalent for mothers who are aged less than
30 compared to those aged 30 and over. A greater proportion of Cook
Island Maori and Other Pacific mothers were found living in an extended
family compared to Pacific mothers in general. A smaller proportion of
Samoan and Tongan mothers were found to be living in an extended family
relative to a nuclear family.
As parity increases the proportion of mothers living in an extended
family decreases. A similar pattern can also be seen with the stage of
arrival to New Zealand, with a greater proportion of those born in New
Zealand found to be living in an extended family and a smaller
proportion of those who came to New Zealand as an adult doing so. More
mothers who were not living with a partner were found to be living with
extended family members than those with a partner. In terms of
acculturation, segregationalists were much less likely to be found
living in an extended family. Mothers do not seem to differ in
educational attainment in their choice of living arrangement.
Table 2 displays the reasons people gave for living with extended
family members. Mothers were allowed to give up to three reasons.
Seventeen respondents’ replies are missing or not applicable, 457
respondents gave one reason, 123 respondents gave two reasons, and 13
respondents gave three reasons. The three most common reasons given
were financial reasons for the nuclear family (38%), taking care of the
extended family (21%) and financial reasons for the extended family
(16%). The size of the ‘other’ category (31%) demonstrates the wide
range of responses to this question. Only 3% of mothers who live with
extended family members stated that they did so for cultural reasons.
While the cross-tabulations allow us to examine the characteristics
of mothers that live in different types of households, we are not able
to tell which characteristics are directly associated with having
different living arrangements. For example, mothers with low
education or parity may be more likely to live in extended families
because they are also younger. A multinomial logistic regression was
used with the family type as the dependent variable in order to assess
the association of each individual factor controlling for all other
variables.
Three different living arrangements were defined for the purposes of
this analysis. These were the nuclear family (which included both
“couple with children” and “single parent with children” household
types), an extended family living situation in which at least one of
the child’s grandparents were living in the household, and other
“extended” family types which did not contain any of the child’s
grandparents.
These three categories were considered desirable because the sample
size in each group was fairly large. In addition, it was thought that
the characteristics of the mothers who were living with at least one of
their child’s grandparents may differ significantly in some respects to
the characteristics of the mothers who lived in extended families in
which none of the child’s grandparents were present. Nuclear families
that include people that are unrelated to the family have been put in
the last category. The number of families in this situation is small,
and it is considered that many of the factors associated with living
with unrelated people will be similar to those associated with living
with relatives.
Table 3 outlines the results of the regression.7
The results are displayed in terms of odds ratios. The odds ratio
compares the likelihood of a mother with a given characteristic (such
as aged over 35) living in an extended family rather than a nuclear
family relative to the likelihood that a mother with the reference
characteristic (such as aged under 25) lives in an extended family
rather than a nuclear family. A 5% level of significance was used.
The results show that mothers aged 30 to 34 were less likely to be
living in an extended family living arrangements than those aged under
25. There was no significant difference between those aged less than 25
and those in the 25 to 29 years and 35 years and over age groups.
Ethnicity appears to be important in predicting who will live in an
extended family situation without the child’s grandparents present but
does not seem as important for the extended family situation in which
the grandparents are present. Those mothers who were not of a Pacific
ethnicity were less than half as likely to live in an extended family
situation with the grandparents present compared to mothers who were of
a Pacific ethnicity. Mothers of Cook Island Maori background were one
and a half times more likely to be found in an extended family
situation without the grandparents present than mothers of Samoan
ethnicity. The mothers of the other ethnic groups, including
Non-Pacific, were no more or less likely to be in this living
arrangement as the mothers of Samoan ethnicity (at 5% significance).
Parity appears to be a particularly important factor for those
living in an extended family household which includes the child’s
grandparent(s). The greater the parity, the less likely the family is
to be in this type of living arrangement. Parity does not seem to be as
important for those living in an extended family situation without the
presence of the child’s grandparent(s). However, it does become
significant when the mother has four or more children. These families
are half as likely to be in this living arrangement compared to those
with one child.
Mother’s who came to New Zealand as an adult appear to be half as
likely to live in an extended family situation with the child’s
grandparent(s) compared to mothers who were born in New Zealand. This
characteristic has no significant impact on those in an extended family
living situation. Social marital status has a large impact on whether
the child is living in an extended family situation. Those mothers who
did not have a partner living with them are much more likely to live in
both extended family situations relative to those who did have a
partner living with them.
Acculturation does not appear to have any impact on those living in
an extended family situation with the child’s grandparent(s). However,
mothers who were considered to be segregationalists were found to be
half as likely as assimilationists to live in an extended family
without the child’s grandparent(s) present. The mothers’ level of
educational attainment appears to have no impact on living
arrangements.
Discussion
The significant variables associated with living with extended family
members include the mother’s age, ethnicity, parity, stage of arrival
to New Zealand, social marital status and acculturation. The age effect
together with the parity effect signals a lifecycle influence. Mothers
who are young (under 30 years) or who have fewer children are more
likely to be living in an extended family. This is consistent with the
extended family acting as a support for members when their children
(and they themselves) are young. Although income is not included in
this analysis, it too usually follows a lifecycle pattern of being low
when a person is young, growing as they get older and decreasing again
as they reach retirement and thus might be related to these findings.
Non-Pacific mothers were less than half as likely to live in an
extended family with at least one of the child’s grandparents present
than Pacific mothers. It is interesting to note, however, that they are
found to be just as likely to live in an extended family in which none
of the child’s grandparents are present. It is possible that this
reflects differing cultural expectations about the role of parents and
grandparents (particularly the mother and grandmother) in caring for
and parenting the children.
Characteristics of the mother associated with living in an extended
family in which the child’s grandparent(s) are present are different
than those of mothers living in an extended family situation in which
none of the child’s grandparent(s) live in the same household. For
example, those who came to New Zealand as an adult are much less likely
to live in an extended family with the grandparent’s present than those
who are New Zealand born. However, there is no significant difference
between stages of arrival in New Zealand in terms of extended family
living in which none of the child’s grandparents are present. One
possible explanation for this may be that the child’s grandparents may
not reside in New Zealand. The residence of the child’s grandparents is
not collected in the dataset and so it is not possible to check for
this.
Another area in which the characteristics of the mother differ
depending on the extended family type is that of parity of four or
more. While these families are less likely to live with extended family
members, they appear to be more likely to do so in a living arrangement
such that the child’s grandparent(s) are not living in the same
household, compared to one in which they are. It may be that the role
conflicts between grandparents and parents living in the same household
found in some research (Unger & Cooley 1992: Spieker & Bensley
1994) are driving this difference.
Another possible explanation is that of reverse causality. Perhaps
those who live with at least one of the child’s grandparents tend to
have fewer children. If it is the responsibility of that particular
family to provide care for the child’s grandparent, they may choose to
have fewer children in order to enable them to provide adequate care
for the grandparent. However, this seems unlikely as Pacific people
tend to have children at a fairly young age, the grandparents are
likely to be fairly young also, and they are probably more in a
position to provide support to the family rather than needing large
amounts of support themselves.
Finally, the differences in the parity between the two types of
extended family could be a result of selection by either the
grandparent or the family. The grandparent may choose to live with
whichever of their children have the fewest children of their own. Or
else, those mothers who wish to work may have fewer children and live
with the child’s grandparent(s) so that they may provide assistance
with childcare, whereas those mothers who choose not to work may be
inclined to have more children and to live as a nuclear family due to a
reduced need for childcare support.
Conclusion
This paper describes the living arrangements and associated
characteristics of mothers with a one-year-old Pacific child and
identifies a number of characteristics associated with extended family
living. These mothers are more likely to live with extended family
members when they are young, single or a new parent. In addition,
living arrangements differ by ethnicity, with non-Pacific mothers much
less likely to be living with their child’s grandparent(s) and Cook
Island Maori mothers much more likely to be living in an extended
family without the child’s grandparent(s) present. We also found that
different characteristics are associated with living in an extended
family household in which the child’s grandparents are present versus
living in an extended family household in which the child’s
grandparent(s) are not present.
One major limitation of this study is that it is not possible to
tell whether these results reflect different preferences of the parents
of the sample children, different constraints faced by these parents or
the preferences/availability of potential extended family members. For
example, mothers who came to New Zealand as an adult were less than
half as likely as those who were born in New Zealand to live in an
extended family with at least one of the child’s grandparents present,
but it is not possible to tell if this result reflects an increased
presence of grandparents in New Zealand for these families rather than
an increased preference to living with grandparents. In general,
without having detailed information on both the nuclear family and all
potential extended family members, it is not possible to separate out
these pathways.
Another limitation is the absence of a good measure of the
socio-economic status of the family. Economic resources have been found
to be significant in explaining family living arrangements in other
pieces of research (Angel & Tienda 1982; Pérez 1994; Glick
2000). However, we did find that there is an insignificant relationship
between mother’s education and living in an extended family, which
suggests that income effects may not be important in the context of
this study. It is worth noting though that father’s education would
have formed a better proxy (but unfortunately this is not available in
the current data), as it is likely that the father is the primary
earner for the majority of families. In addition, qualifications may be
measured too broadly to identify any specific effects.
It is also important to note that these results pertain to a
particular sample of Pacific children living in Auckland and it is not
possible to generalise for all Pacific people or for all people in New
Zealand. There also may be other characteristics, such as grandparent’s
health or the mother’s parenting skills, that are significantly
associated with the family living arrangement but are not included in
this study. However, overall this study does highlight that living in
an extended family is most common among mothers that are young, single
or a new parent and is not strongly related with most other
characteristics of these mothers, including their education, religion,
English speaking skills and acculturation.
References
Angel R, Tienda M 1982. Determinants of extended household structure:
cultural pattern or economic need? The American Journal of Sociology
87(6): 1360–1389.
Bathgate M, Donnell A, Mitikulena A, Borman B, Roberts A, Grigg M
1994. The health of Pacific Islands people in New Zealand. Wellington,
Public Health Commission.
Bedford RD & Didham RA 2001. Who are the ‘Pacific Peoples’?
Ethnic identification and the New Zealand Census. In: Macpherson C,
Spoonley P, Anae M ed. Tangata O Te Moana Nui: the evolving identities
of Pacific peoples in Aotearoa/New Zealand. Palmerston North, Dunmore
Press.
Berry JW 1980. Acculturation as varieties of adaptation. In:
Palldilla A ed. Acculturation: theory, models and some new findings.
Boulder, Westview. Pp. 9–25.
Berry JW, Kim U 1988. Acculturation and mental health. In: Dasen P,
Berry JW, Satorius N ed. Handbook of cross-cultural psychiatry: towards
application. London, Sage. Pp.207–236.
Blank S, Torrecilha RS 1994. Understanding the living arrangements
of Latino immigrants: a life course approach. International Migration
Review 32(1): 3–19.
Cook L, Didham R, Khawaja M 2001. The shape of the future: on the
demography of Pacific peoples. In: Macpherson C, Spoonley P, Anae M ed.
Tangata o te moana nui: the evolving identities of Pacific peoples in
Aotearoa/New Zealand. Palmerston North, Dunmore Press. Pp. 44–65.
Glick J 2000. Nativity, duration of residence and the life course
pattern of extended family living in the USA. Population Research and
Policy Review 19(2): 179–198.
Gordon R, Chase-Lansdale P, Brooks-Gunn J 2004. Extended households
and the life course of young mothers: understanding the associations
using a sample of mothers with premature, low birth weight babies.
Child Development 75(4): 1013–1038.
Gray A 2001. Definitions of household crowding and effects of
crowding on health: a literature review. Research Series Report 1.
Wellington, Ministry of Social Policy.
Jin Hwang H, St James-Roberts I 1998. Emotional and behavioural
problems in primary school children from nuclear and extended families
in Korea. Journal of Child Psychology and Psychiatry 39: 973–979.
Maani S, Maloney T 2004. Returns to post-school qualifications: new
evidence based on the HLFS Income Supplement (1997–2002). Report to the
Labour Market Policy Group. Wellington, Department of Labour.
Macpherson C 1991. The changing contours of Samoan identity. In:
Spoonley P, Pearson D, Macpherson C ed. Nga take: ethnicity and racism
in New Zealand/Aotearoa. Palmerston North, Dunmore Press.
Meleisea M, Schoeffel P 1998. Samoan families in New Zealand: the
cultural context of change. In: Adair V, Dixon R ed. The family in
Aotearoa New Zealand. Auckland, Longman. Pp. 158–178.
Ministry of Social Development 2005. The social report 2005: te
pūrongo oranga tangata. Wellington, Ministry of Social Development.
Paterson J, Tukuitonga C, Abbott M, Feehan M, Silva P, Perecival T,
Carter S, Cowley-Malcolm E, Borrows J, Williams M, Schluter P 2006.
Pacific Islands families: first two years of life study—design and
methodology. The New Zealand Medical Journal. Retrieved 15 January
2006, from http://www.nzma.org.nz/ journal/119–1228/1814/
Pérez L 1994. The household structure of second-generation
children: an exploratory study of extended family arrangements. The
International Migration Review 28(4): 736–747.
Spieker S, Bensley L 1994. Roles of living arrangements and
grandmother social support in adolescent mothering and infant
attachment. Developmental Psychology 30: 102–111.
Statistics New Zealand 1998. Families and households: 1996 census of
population and dwellings. Wellington, Statistics New Zealand.
Statistics New Zealand 2002. Pacific progress: a report on the
economic status of Pacific Islands peoples in New Zealand. Wellington,
Statistics New Zealand.
Tinsley BJ, Parke RD 1984. The contemporary impact of the extended
family on the nuclear family: grandparents as support and socialisation
agents. In: Lewis M ed. Beyond the dyad. New York, Plenum Press. Pp.
161–194.
Tsai JL, Ying Y, Lee PA 2000. The meaning of being Chinese and being
American: variation among Chinese and American young adults. Journal of
Cross-Cultural Psychology 31: 302–32.
Unger DG, Cooley M 1992. Partner and grandmother contact in black
and white teen families. Journal of Adolescent Health 13: 546–552.
Ward RG 1996. Expanding worlds of Oceania: implications for
migration. Japan Centre for Area Studies Symposium Series 3: 61–78
*This paper has been co-authored by the Families
Commission and the Pacific Islands Families Study team at Auckland
University of Technology with assistance from Motu Economic and Public
Policy Research. As part of its work, the Families Commission is
committed to working with other agencies to make use of existing data
to inform its programme of work. The Families Commission has an ongoing
interest in promoting Pacific families’ social and economic
development. Any views expressed are the sole responsibility of the
authors and do not purport to represent those of the Families
Commission or Motu Economic and Public Policy Research.
1 Individuals are allowed to report up
to three ethnicities in the New Zealand Census. The percentages
reported here include individuals who report multiple Pacific
ethnicities and thus add up to more than 100%.
2 An economic family unit is “a group
of co-resident people whose financial affairs are common or have been
merged to the extent the people are substantially interdependent”
(Ministry of Social Development 2005: 163).
3 Those in poverty are defined as
those who have net-of-housing-cost incomes below the 60% line
(benchmarked to the 1998 median).
4 The word “mother” is used to refer to
the primary respondent. While 1201 (98.1%) of the primary respondents
were in fact the biological mothers of the children in the cohort, 23
were not. Of these, twelve were adoptive mothers, four were
grandmothers, two were foster mothers, two were biological fathers, one
was a grandaunt, one a cousin and one a non-related caregiver (Paterson
et al. 2006).
5 Note that parity in this study
refers to the number of children the mother has at this point in time.
A parity of two therefore means that the mother has just had her second
child.
6 Fathers’ education has not been
included as this information is missing for a third of the fathers.
7 The sample is choice based (made up
of individuals who chose to have a child) and, as such, many of the
factors associated with family living arrangements are unlikely to be
truly exogenous. For example, those mothers who live with their parents
may have been more likely to have a child given the support available.
We tried a series of regressions in order to be cautious about the
impact of endogeneity. The first contained age, ethnicity, parity and
stage of arrival in New Zealand. The second added social marital
status, which was found to be significant and did not alter any of the
other results. The third added religion, spoken English competency and
acculturation. These were not jointly significant and did not alter any
of the previous results. The fourth used a restricted sample of those
aged over 25 and included education. The reason for this was the
concern that pregnancy may have interrupted educational attainment for
those aged less than 25. The smaller sample size increased the standard
errors and as such reduced the significance of some variables.
Education itself was not significant. When education was added to the
whole sample, it did not alter any of the previous results and its
addition was not very significant.
Table 1 Mothers’ characteristics by family type
|
Mothers’ characteristics
|
Child lives in a nuclear family (%)
|
Child lives in an extended family (%)
|
|
Overall
|
50.2
|
49.8
|
|
Age
|
<25
|
32.8
|
67.1
|
|
25–29
|
45.6
|
54.3
|
|
30–34
|
63.7
|
36.3
|
|
35+
|
63.3
|
36.8
|
|
Ethnicity
|
Samoan
|
52.8
|
47.2
|
|
Cook Island Maori
|
41.1
|
58.9
|
|
Niuean
|
49.1
|
50.8
|
|
Tongan
|
53.3
|
46.7
|
|
Other Pacific
|
40.5
|
59.4
|
|
Non Pacific
|
50.6
|
49.3
|
|
Parity
|
1
|
31.7
|
68.2
|
|
2
|
47.0
|
53.0
|
|
3
|
56.4
|
43.5
|
|
4+
|
68.4
|
31.6
|
|
Stage of arrival to New Zealand
|
NZ born
|
39.9
|
60.1
|
|
As a child (<15)
|
45.3
|
54.8
|
|
As an adult (15+)
|
60.1
|
40.0
|
|
Social marital status
|
Not partnered
|
25.7
|
74.3
|
|
Partnered
|
56.0
|
43.9
|
|
Religion
|
Some
|
50.4
|
49.7
|
|
None
|
47.3
|
52.7
|
|
Speaks English fluently?
|
Yes
|
45.4
|
54.5
|
|
No
|
58.0
|
42.0
|
|
Acculturation
|
Assimilationist
|
44.4
|
55.4
|
|
Segregationalist
|
79.8
|
40.4
|
|
Integrator
|
45.5
|
54.4
|
|
Marginal
|
46.1
|
53.9
|
|
Education
|
No formal
|
50.5
|
49.5
|
|
Secondary school
|
49.6
|
50.3
|
|
Post-school
|
50.4
|
49.5
|
Table 2 Reasons given by mothers for living in an
extended family household
|
Reason live with extended family
|
Frequency
|
%
|
|
Financial/cheaper for mother and nuclear family
|
233
|
38
|
|
Other
|
192
|
31
|
|
Taking care of extended family
|
130
|
21
|
|
Financial/cheaper for extended family
|
98
|
16
|
|
To help others
|
81
|
13
|
|
Cultural reasons
|
18
|
3
|
|
Missing/not applicable
|
17
|
3
|
|
Total number of mothers living in an extended
family
|
610
|
100
|
Table 3 Multinomial logistic regression odds ratios1
|
Mothers’ characteristics
|
Child lives with extended family including
grandparent(s)2
|
Child lives with extended family excluding
grandparent(s)1
|
|
Age
|
<25†
|
1.00
|
1.00
|
|
25-29
|
0.93
|
0.97
|
|
30-34
|
0.54***
|
0.57**
|
|
35+
|
0.58*
|
0.81
|
|
Ethnicity
|
Samoan†
|
1.00
|
1.00
|
|
Cook Island Maori
|
1.06
|
1.94***
|
|
Niuean
|
0.58
|
1.66
|
|
Tongan
|
1.00
|
1.50*
|
|
Other Pacific
|
1.49
|
2.61*
|
|
Non Pacific
|
0.46**
|
1.55
|
|
Parity
|
1†
|
1.00
|
1.00
|
|
2
|
0.51***
|
0.64*
|
|
3
|
0.37***
|
0.71
|
|
4+
|
0.20***
|
0.46***
|
|
Stage of arrival in
New Zealand
|
New Zealand born†
|
1.00
|
1.00
|
|
As a child
|
0.72
|
1.48
|
|
As an adult
|
0.44***
|
1.42
|
|
Social marital status
|
Not partnered†
|
1.00
|
1.00
|
|
Partnered
|
0.28***
|
0.38***
|
|
Religion
|
Some†
|
1.00
|
1.00
|
|
None
|
1.12
|
1.16
|
|
Spoken English
|
Competent†
|
1.00
|
1.00
|
|
Not competent
|
0.93
|
1.58*
|
|
Acculturation
|
Assimilationist†
|
1.00
|
1.00
|
|
Segregationalist
|
1.08
|
0.54**
|
|
Integrator
|
1.24
|
0.91
|
|
Marginal
|
1.04
|
1.00
|
|
Education
|
No formal†
|
1.00
|
1.00
|
|
Secondary school
|
0.90
|
1.11
|
|
Post-school
|
1.22
|
1.40*
|
1The dependent variable is family type.
2The reference family type is a nuclear family.
*p < 0.10; **p < 0.05; ***p < 0.01; †default state.
This year's abstracts | Journal home page | All abstracts | Publishing home page
