12
TAHU H. KUKUTAI Stanford University White Mothers, Brown Children: Ethnic Identification of Maori-European Children in New Zealand Studies of multiethnic families often assume the ethnic identification of children with the minor- ity group results from the minority parent. This study examines an alternate view that main- stream parents also play an important role in transmitting minority ethnicity. It explores this argument using data from New Zealand on the ethnic labels mothers assign to their Maori- European children. It finds that European mothers are just as disposed as Maori mothers to designate their child as Maori, either ex- clusively or in combination. Two explanations, grounded in ethnic awareness and gendered inheritance, are proposed. Although neither satisfactorily predicts maternal designation de- cisions, the readiness of European mothers to identify their child as Maori underscores their role in diffusing Maori ethnicity. For parents who have partnered with someone from a different ethnic group, one of the funda- mental decisions they face is how to ethnically identify their child. In theory, this decision should be straightforward: Simply identify the child in a way that reflects both parental backgrounds. In practice, designation decisions are far more complex. Many of the ethnic labels assigned to multiethnic children do not faithfully reflect their ethnic heritage but are shaped by processes and constraints within and beyond the bounds of family life. Recognizing these complexities, social scien- tists have generally eschewed making sweeping statements about ethnic labeling processes, focus- ing instead on the dynamics that occur within spe- cific kinds of interethnic partnerships (Qian, 2004; Tafoya, Johnson, & Hill, 2004). Even so, two as- sumptions about the intergenerational transmis- sion of ethnicity have consistently transcended the peculiarities of time and place. One is the notion that minority parents are the sole bearers and transmitters of minority ethnicity. The other is the notion that mainstream parents are less in- vested than their minority partners in decisions about how to ethnically identify their child. The pervasiveness of both assumptions are reflected in how parent-child studies try to account for the influence of the minority parent on designation decisions but pay relatively little attention to part- ners from the dominant group (Liebler & Kana’- ianupuni, 2003/2004; Roth, 2005; Saenz, Hwang, & Aguirre, 1995; Xie & Goyette, 1997). As a counterpoint to the literature, this article explores whether mainstream parents also act as agents of cultural continuity vis-a `-vis the deci- sions they make about their child’s ethnic iden- tification. It examines this question within the context of Maori-European intermarriage in New Zealand using data on how mothers ethni- cally identify their children. New Zealand is an ideal empirical context in which to consider this broader theoretical question. To treat parental Department of Sociology, Stanford University, 450 Serra Mall, Building 120, Room 160, Stanford, CA 94305-2047 ([email protected]). Key Words: ethnic identification, interethnic marriage, intergenerational transmission, Maori ethnicity. 1150 Journal of Marriage and Family 69 (December 2007): 1150–1161

White Mothers, Brown Children: Ethnic Identification of Maori-European Children in New Zealand

Embed Size (px)

Citation preview

Page 1: White Mothers, Brown Children: Ethnic Identification of Maori-European Children in New Zealand

TAHU H. KUKUTAI Stanford University

White Mothers, Brown Children: Ethnic

Identification of Maori-European Children

in New Zealand

Studies of multiethnic families often assume theethnic identification of children with the minor-ity group results from the minority parent. Thisstudy examines an alternate view that main-stream parents also play an important role intransmitting minority ethnicity. It explores thisargument using data from New Zealand on theethnic labels mothers assign to their Maori-European children. It finds that Europeanmothers are just as disposed as Maori mothersto designate their child as Maori, either ex-clusively or in combination. Two explanations,grounded in ethnic awareness and genderedinheritance, are proposed. Although neithersatisfactorily predicts maternal designation de-cisions, the readiness of European mothers toidentify their child as Maori underscores theirrole in diffusing Maori ethnicity.

For parents who have partnered with someonefrom a different ethnic group, one of the funda-mental decisions they face is how to ethnicallyidentify their child. In theory, this decision shouldbe straightforward: Simply identify the child ina way that reflects both parental backgrounds.In practice, designation decisions are far morecomplex. Many of the ethnic labels assigned to

multiethnic children do not faithfully reflect theirethnic heritage but are shaped by processes andconstraints within and beyond the bounds offamily life.

Recognizing these complexities, social scien-tists have generally eschewed making sweepingstatements about ethnic labeling processes, focus-ing instead on the dynamics that occur within spe-cific kinds of interethnic partnerships (Qian, 2004;Tafoya, Johnson, & Hill, 2004). Even so, two as-sumptions about the intergenerational transmis-sion of ethnicity have consistently transcendedthe peculiarities of time and place. One is thenotion that minority parents are the sole bearersand transmitters of minority ethnicity. The otheris the notion that mainstream parents are less in-vested than their minority partners in decisionsabout how to ethnically identify their child. Thepervasiveness of both assumptions are reflectedin how parent-child studies try to account for theinfluence of the minority parent on designationdecisions but pay relatively little attention to part-ners from the dominant group (Liebler & Kana’-ianupuni, 2003/2004; Roth, 2005; Saenz,Hwang, & Aguirre, 1995; Xie & Goyette, 1997).

As a counterpoint to the literature, this articleexplores whether mainstream parents also act asagents of cultural continuity vis-a-vis the deci-sions they make about their child’s ethnic iden-tification. It examines this question within thecontext of Maori-European intermarriage inNew Zealand using data on how mothers ethni-cally identify their children. New Zealand is anideal empirical context in which to consider thisbroader theoretical question. To treat parental

Department of Sociology, Stanford University, 450 SerraMall, Building 120, Room 160, Stanford, CA 94305-2047([email protected]).

Key Words: ethnic identification, interethnic marriage,intergenerational transmission, Maori ethnicity.

1150 Journal of Marriage and Family 69 (December 2007): 1150–1161

Page 2: White Mothers, Brown Children: Ethnic Identification of Maori-European Children in New Zealand

labeling decisions as reflective of how parentsactually view their child requires the necessary,but not necessarily sufficient, condition that gen-uine choices exist. Exercising ‘‘ethnic options’’means more than simply having the opportunityto select from one or more ethnicities on a form;it also entails the legitimization of identificationchoices (Davis, 1991). Historically high rates ofMaori-European intermarriage and a legacy ofmultiethnic data collectionmeanNewZealandershave long had the choice to acknowledge theirchild’s mixed parentage for purposes mandatedby the state (Brown, 1983; Pool, 1991). Designa-tion decisions in Maori-European families arethus more likely to reflect parental orientationsat a given point in time than bureaucratic con-straints or a history of prescriptive racial identifi-cation rules.

By examining the ethnic labels assigned toMaori-European children, this article exploreswhether Europeanmothers play a role in the inter-generational diffusion of Maori ethnic identity. Itposes three questions: (a) How willing are Euro-pean mothers to assign Maori ethnicity to theirMaori-European child;? (b) How do their label-ing choices differ from those of mothers whoidentify as Maori;? and (c) What factors are asso-ciated with their designation decisions? The fol-lowing section provides the historical contextfor Maori-European intermarriage and reviewskey theoretical arguments about the intergenera-tional transmission of ethnic identity.Hypothesesregarding the ethnic identification of Maori-European children are proposed and tested usingquantitative methods. The article concludes witha discussion of how the findings provide insightsinto the complexities ofMaori ethnicity transmis-sion and the role that mainstream parents mightplay in the diffusion of minority ethnic identity.

BACKGROUND

Maori are regarded as the tangata whenua orfounding peoples of New Zealand, havingmigrated from eastern Polynesia between 900and 1300 AD (Walker, 1990). New Zealand’s‘‘European’’ population comprises persons ofmostly British origins, many of whom are thedescendants of 19th-century migrants. The Maoriterm Pakeha (King, 1991) remains a popular col-loquialism to describe White New Zealanders,but opposition to its usage has seen it graduallydisappear from official statistics (Statistics NewZealand, 2004). Until the 1970s, New Zealand

was predominantly a nation of two peoples buthas undergone significant diversification withimmigration waves from the Pacific and Asia. In2001, Europeans constituted 80% of the popula-tion, Maori were 14.7%, and Pacific and Asianpeoples were 6.5% and 6.6%, respectively (Sta-tistics New Zealand, 2001). Persons who identi-fied with multiple ethnic groups were counted ineach, so that the sum of the four broad ethniccategories exceeds the total population.

As official definitions regarding group mem-bership can influence ethnic dynamics and iden-tification long after their demise (Davis, 1991),it is useful to consider how persons of Maoriand European parentage were historically classi-fied. In the early days of census taking, half-casteMaori Europeans were allocated to theEuropean or Maori population, depending onwhether enumerators determined their ‘‘style oflife’’ to be Maori or European oriented. The1891 census recorded 2,760 half-castes livingas Maori (i.e., probably residing in customarykin-group settings) and 2,184 living as Euro-peans (Office of the Registrar-General, 1892;Table 1). After 1926, all persons who wererecorded as ‘‘half or more Maori blood’’ werecounted as Maori, and persons below the thresh-old were counted as European. Increasingly re-fined distinctions were made so that by 1951,the census report shows children classified as1/8 Maori and 7/8 European. Scholars have notedthat the institution of fractional identities andtheir associated nomenclature (i.e., half-caste) re-flected Victorian ideas of race as well as politicalefforts to circumscribe the size of indigenouspopulations (Kukutai, 2004; Smith et al., 2006).This was certainly true for Maori although, dis-tinct from American Indians (Snipp, 1997) andNative Hawaiians (Kauanui, 2005), they wererarely called upon to authenticate their bloodclaims. Instead, there was wide latitude to inter-pret the western construct of blood from a Maoriparadigm, in which the most important marker ofidentity was whakapapa or genealogy connectingthe individual to kinship networks of whanau(immediate and extended family), hapu (sub-tribe) and iwi (tribe; see Walker, 1990). For manyMaori, the presence of European ancestry, evenparentage, did not undermine an individual’sclaim to Maori identity. Indeed, many more‘‘full’’ Maori were reported in the census thanwas biologically possible (Buck, 1924; Pool,1991). After the fractional approach was replacedin 1986 by the current question on self-identified

Ethnic Identification of Maori-European Children 1151

Page 3: White Mothers, Brown Children: Ethnic Identification of Maori-European Children in New Zealand

ethnic group (Statistics New Zealand, 2004), pa-rents wanting to acknowledge their child’s mixedheritage could do so by recording multiple ethnicgroups (e.g., Maori and European). This approachcontrasts with that in the United States where pre-scriptive legal definitions such as the ‘‘one-drop’’rule and the practice of disallowing multiracialreporting in the census suppressed the recogni-tion of racial diversity until the 2000 census(Perlmann & Waters, 2002).

Today the potential for a multiethnic Maori-European population is substantial, with about halfof partnered Maori between the ages of 20 and 64in a relationshipwith a non-Maori (Callister, 2004;Callister, Didham, & Potter, 2005). Yet, with theexception of Harre’s (1966) study of Maori-Euro-pean intermarriage in Auckland and recentexploratory analysis of census and births registra-tion data (Callister, 2003; Howard & Didham,2005), there has been relatively little empiricalresearch conducted on the intergenerational trans-mission of Maori ethnicity, let alone the role thatEuropean parents might play. For this reason, thefollowing theoretical arguments draw primarilyfrom the expansive North American literature.

There, we find two dominant models of ethnicchange within which studies of parental designa-tion decisions have often been framed. The first,the assimilation perspective, predicts the fadingimportance of ethnicity as minorities are incor-porated into the ‘‘host’’ society (Gordon, 1964).Incorporation occurs over generations, involvingone or more processes of acculturation, eco-nomic assimilation, and intermarriage. From theassimilation standpoint, persons who have inter-married are less reliable transmitters of minorityethnicity; thus, mainstream-minority intermar-riage eventually leads to a reduction in the lattergroup. This point is especially salient for indige-nous populations that are theoretically ‘‘closed’’to migration. A wide variety of indicators havebeen employed in parent-child studies to mea-sure the minority parent’s level of assimilationand the effects on their child’s ethnic identifica-tion (Saenz et al. 1995; Xie & Goyette, 1997).

The other dominant model of ethnic change isethnic pluralism or what Gans (1997) calls theethnic retentionist model. This perspective pre-dicts that conditions of modernity encourage thesurvival and revival, rather than the decline, of

Table 1. Assigned Child Ethnicity for Maori, European, and Maori-European Parents in the 1996 Census and the 1995

New Zealand Women: Family, Employment, and Education Survey

Father Mother

1996 Censusa

Maori European Maori and European n

Maori Maori 94.4 0.6 5.0 13,503

European 29.2 16.7 54.3 7,389

Maori and European 32.3 1.7 66.0 3,078

European Maori 33.8 12.5 53.7 6,144

European 0.3 98.6 1.1 255,132

Maori and European 2.8 19.7 77.5 10,698

Maori and European Maori 52.4 1.7 45.9 2,157

European 3.2 30.8 66.0 10,389

Maori and European 6.9 3.2 89.9 4,251

Total 20,694 259,260 32,787 312,741

Father Mother

1995 New Zealand Women: Family, Employment, and Education Survey

Maori European Maori and European n

Maori Maori 98.7 — 1.3 157

European 37.0 12.4 50.6 89

Maori and European 67.6 — 32.4 37

European Maori 25.6 18.3 56.1 82

European — 99.4 0.6 1,641

Maori and European 3.8 41.8 54.4 90

Total 238 1,697 165 2,096

aNZ Census, 1996, custom table. Mothers aged 20 – 59 years.

1152 Journal of Marriage and Family

Page 4: White Mothers, Brown Children: Ethnic Identification of Maori-European Children in New Zealand

ethnicity. One aspect of ethnic retention is ethnicawareness, manifest in a sense of ethnic pride,interest in personal family histories, and aware-ness of ethnic status distinctions and discrimina-tion (Portes, 1984; Xie & Goyette, 1997). Moststudies of ethnic labeling decisions use data forwhich there is no direct measure of parental eth-nic awareness, so proxy measures such as educa-tional attainment are substituted. Education isthought to heighten an appreciation of one’s fam-ily history, as well as exposure to mainstreamprejudice that intensifies the salience of ethnicity.The expectation is that minority parents who arebetter educated are more likely to identify theirchild as a member of the minority group.

Both the assimilation and the retentionist ap-proaches revolve around the motivations andpreferences of the minority parent, reflecting themore general assumption that it is the minorityparent’s job to ‘‘pass on’’ the minority ethnicity.Yet most studies of ethnic labeling use censusdata for which there is no way of knowing whichparent, if any, identified the child. Without thisbasic information, we cannot assume that thetransmission of minority ethnic identity dependson the ethnic orientation and preferences of theminority parent. Mainstream parents, and moth-ers in particular, may also play an important role.

The view that mainstream parents act as agentsof minority cultural continuity is not one that iswidely shared in the literature, yet is entirely con-sistent with the sociological understanding ofethnicity as socially constructed rather than biolog-ically contingent (Nagel, 1994). Even if manymainstreamparents lack thewherewithal to conveythe substantive content of minority ethnicity, theethnic identification of their child is one of thefew arenas in which they can fully participate(Archie, 2005; O’Donoghue, 2004; Twine, 1999).Reporting a child’s ethnicity in a survey mightappear inconsequential but is one way by whicha child’s identity is explicitly declared to othersand thus may be seen as a basic component of eth-nicity transmission. If Europeanmothers play a rolein the transmissionofMaori ethnicity, theyought tobe just as willing asMaori mothers to identify theirchild as Maori. This yields the following:

H1: European mothers should be just as likely asMaori mothers to identify their child as Maori, oras both Maori and European.

One might argue that identifying one’s child asMaori does not necessarily imply an active com-

mitment to transmitting Maori ethnicity, and wealso need to consider the factors that underlieparental decisions. One possible explanation forwhy European mothers identify their child asMaori is that they have ethnic awareness. Theopportunities to cultivateMaori ethnic awarenesshave increased significantly since the 1970s, withthe revival of Maori language and customs, re-configured relations between tribes and the state,and the institutionalization of Maori culture inpublic life (Greenland, 1984; Webster, 1998).The renewed emphasis contrasts with the viewthat European New Zealanders lack a cohesiveethnicity or culture and thus do not have a distinc-tive identity worth transmitting (Wetherell &Potter, 1992). European mothers may alsodevelop an awareness of Maori identity and cul-ture from sources endogamous to families (e.g.,interaction with their child’s and partner’s Maorirelatives), aswell as structural aspects of the envi-ronment (e.g., living in a strongly Maori area).Beyond families, ethnic awareness may beheightened by exposure to prejudice and discrim-ination. In spite of ‘‘ethnic renewal’’ (Nagel,1995), Maori remain an economically disadvan-taged and lower status ethnic group (Kukutai,2004; Te Puni Kokiri/Ministry of Maori Devel-opment, 2000). In a recent book on intermarriage,one European mother expressed frustration atconstantly bearing witness to anti-Maori senti-ment: ‘‘When I hear snide remarks, it sometimesgets a bit too much and I will say, ‘Actually I ammarried to a Maori,’ and they will say, ‘I don’tmean everyMaori. There are always good ones’’’(Archie, 2005, p. 234). If a child looks like astereotypical Maori (Holmes, Murachver, &Bayard, 2001), the juxtaposition between Whitemother and Brown child might also heightenethnic awareness and encourage the attributionof Maori ethnicity. Although persons of Maoriparentage vary greatly with respect to skin colorand facial features, subtle distinctions may stillbe invoked to underscore difference (McDonald,1976; Thomas & Nikora, 1998).

The data used here do not provide direct meas-ures of ethnic awareness, but several indirectmeasures may be used: living in a populousMaori area and having a tertiary education. Inaddition, mothers who are ethnically awareshould tend to emphasize their child’s Maori eth-nicity. This yields the following hypothesis:

H2: European mothers who assign Maori ethnic-ity to their child should (a) have higher levels of

Ethnic Identification of Maori-European Children 1153

Page 5: White Mothers, Brown Children: Ethnic Identification of Maori-European Children in New Zealand

education and (b) live in strongly Maori areas.This should be especially evident for those whoidentify their child solely as Maori.

An alternative explanation for why Europeanmothers identify their child as Maori is that theyare enacting gender dynamics, rather than ethnicawareness per se. Several parent-child studieshave found that childrenwith oneminority ethnicparent are more likely to be reported as a memberof the minority group if the minority parent is thefather (see Roth, 2005; Saenz et al., 1995; Xie &Goyette, 1997). One interpretation is that ethnicsurnames typically reveal heritage and thusencourage the attribution of minority ethnicity.Maori surnames disclose Maori ethnicity, butgenerations of intermarriage mean many personswith Maori fathers do not have Maori surnames.Identification decisions may also be influencedby a more general orientation toward a patrilinealconception of ethnic inheritance (Roth). This ori-entation may mean that European mothers iden-tify their child as Maori only because they areacting in a manner consistent with broader patri-archal rules. If this is the case, Maori paternityshould be a significant predictor of the child beingassigned Maori ethnicity, especially if the paren-tal relationship is still intact.

H3: Children should be more likely to be as-signed Maori ethnicity when the Maori parent isthe father. This bias should be pronounced whenthe parental union is intact.

METHOD

I test these hypotheses using data from the 1995New Zealand Women: Family, Employmentand Education survey. The nationally representa-tive survey was not specifically designed withethnic labeling designations in mind but is wellsuited to the task because it tells us which parentidentified the child and that a biological parent-child relationship exists. These signify major ad-vantages over census data for which none of theseconditions can be definitively known (Xie &Goyette, 1997). It also includes persons in fami-lies where the parental union had dissolved.The inclusion of the latter is particularly impor-tant forMaori children because a disproportionatenumber are raised in sole-parent families andtheir exclusion might result in a partial view ofethnic identification dynamics (Callister, 2003).Given the modest sample size I include all per-

sons born to mothers in an interethnic unionrather than focus exclusively on children aged15 years or younger. By including the full rangeof mother-child pairings, it is possible to testwhether maternal designation decisions differ,depending on the age of their child.

The sampling strategy employed was randommultilevel stratified clustering with an oversam-ple of Maori to ensure sufficient numbers forregional analysis (Marsault et al., 1997). The sur-vey was designed so that all questions—eventhose relating to husbands and partners—wereanswered by women respondents in their ownhomes. With respect to their own ethnic identifi-cation, women were asked: ‘‘Which ethnic groupor groups do you belong to’’? Theywere read andshown a list of 10 ethnic groups that includedMaori, New Zealand European or Pakeha andother European. The New Zealand European orPakeha and other European categories are com-bined because of the latter’s small number andconsistent with the standard practice of aggregat-ing all European ethnic groups into a genericEuropean category (Statistics New Zealand,2004). Women who reported more than one eth-nic group were asked which group they identifiedwith the most. For each biological child, the fol-lowing question was asked: ‘‘What ethnic groupdoes the child belong to’’? The ethnicity of thefather was not self-reported but reported by themother. The most likely source of error from thiswas simplification, women reporting fathers aseither Maori or European rather than as bothMaori and European. Comparisons with the dis-tribution of Maori-European fathers in the 1996census suggest simplification did occur. Becauseof the potential unreliability of those responses,fathers reported as both Maori and Europeanwere excluded from the analysis.

This procedure leaves four kinds of Maori-European parental combinations: Europeanmothers and Maori fathers, Maori mothers andEuropean fathers, Maori-European mothers andEuropean fathers, Maori-European mothersand Maori fathers. The last two kinds of unionsmay be considered both intra- and intermarriagebecause parents not only share one ethnic affilia-tion but also differ. It should be emphasized thatthe data are unlikely to give an accurate picture ofthe number of parents who are themselves theproducts of intermarriage because persons whoaremultiethnic often simplify their reported iden-tity (Harris&Sim, 2002;Waters, 1990).What thedata provide ismore sociologically interesting—an

1154 Journal of Marriage and Family

Page 6: White Mothers, Brown Children: Ethnic Identification of Maori-European Children in New Zealand

insight into the identification decisions of parentswilling to acknowledge their own mixed heritage.Where there is more than one child per woman,the youngest biological child is selected, givingafinal sample size of298Maori-European families.

The model used for the multivariate analysis ismultinomial logistic regression that simulta-neously controls for child, parental, and commu-nity characteristics. Given that unstandardizedcoefficients are often difficult to interpret in termsof substantive meaning, I also report the expo-nentiated betas in the forms of odds ratios (Ag-resti, 1996). An odds ratio above or below 1.0represents the percent increase or decrease inthe odds of a child being assigned a particular eth-nicity (vs. the reference category), given a one-unit increase in the independent variable.

Measures

The dependent variable is the assigned ethnicidentity of the child coded 1 ¼ Maori, 2 ¼Maori and European, and 3 ¼ European (base).The key explanatory variables capture differentaspects of parental ethnicity. Maternal ethnicityis a polytomous variable coded 1 ¼ Maori, 2 ¼European, and 3 ¼Maori and European (base).A second maternal variable captures the ethnicorientation of mothers, coded 0 ¼ Europeanand 1 ¼ Maori. Mothers are coded as havinga Maori ethnic orientation if they identifiedexclusively as Maori or if they reported bothMaori and European ethnicities but identifiedmainly as Maori. Paternal ethnicity is a binaryvariable coded 0 ¼ European and 1 ¼ Maori.The effect of gendered inheritance is capturedin a dummy variable that indicates which parentis Maori where 0 ¼ mother and 1 ¼ father. Forunions where the mother is Maori-Europeanand the father is Maori, the father is coded asthe Maori parent.

There are two indirect measures of ethnicawareness. The first is a dummy variable forwhether the mother has a tertiary qualificationcoded 0 ¼ no and 1 ¼ yes. Degree (e.g., bache-lor’s) and nondegree (e.g., diploma) qualifica-tions are combined to overcome the limitationsof the sample size. The second indirect measureof maternal ethnic awareness is the percentage ofMaori in the territorial authority where themother resides. A territorial authority is analo-gous to a county in the U.S. census and is theconventional measure used in New Zealand tomeasure regional effects.

To control for the effects of child age onmater-nal designation decisions, I include a continuousvariable for age in years. Separate dummy varia-bles are included to capture whether the parentalrelationship was still intact and whether the childresided with the mother (0 ¼ no, 1 ¼ yes). Con-trol variables for maternal attributes are alsoincluded. Maternal age is captured by a continu-ous variable for age in years. It is includedbecause older mothers may be less likely toidentify their children as multiethnic, either asa function of ageing or because they were raisedwhen narrow ideas about race and racial identityprevailed (Harris & Sim, 2002). Separatedummy variables control for whether themother was in the paid labor force, was in a legalmarriage, and had at least two dependent chil-dren (0 ¼ no, 1 ¼ yes). Finally, to capture theeffect of social context on maternal labeling de-cisions, a measure of the relative group positionof Maori and non-Maori is included. It is thedifference in the median personal income of theMaori and total (Maori and non-Maori) popula-tion in the territorial authority where the motherresides, expressed in 1,000s. Living in an areawhere Maori are particularly disadvantaged islikely to heighten the salience of Maori ethnic-ity, but whether this would promote or discour-age mothers from labeling their child as Maoriremains an open question.

RESULTS

Table 1 presents the distribution of ethnic desig-nations of children for all combinations ofMaori,European, and Maori-European parents in theNew Zealand Women: Family, Employmentand Education survey as well as in the 1996 cen-sus undertaken the following year. The surveysare not strictly comparable because we do notknow who reported the child’s ethnicity in thecensus, and it is limited to children aged 15 yearsor younger living in two-parent households.Nevertheless, as New Zealand’s only nationwideenumeration, the census provides a valuable con-text in which to interpret maternal designationdecisions.

Several basic patterns can be determined. First,unions involving two European parents repre-sented the majority of partnerships in both sur-veys, reflecting the European dominance in thetotal population. Of the various kinds of Maori-European intermarriages, the most commonwere

Ethnic Identification of Maori-European Children 1155

Page 7: White Mothers, Brown Children: Ethnic Identification of Maori-European Children in New Zealand

those where one partner identified as Europeanand the other as both Maori and European. Infamilies where the parental ethnicities were iden-tical, the ethnicity assigned to the child wasalmost always the same, but if parental ethnicitiesdiverged, therewas considerable variation in howthe child was identified. For example, many chil-dren with a Maori-European parent and a Maoriparent were identified exclusively as Maori,but this was rarely the case for children witha Maori-European parent and a European parent.Overall, a bias toward Maori ethnicity is evident,with more children of Maori and European par-entage being designated Maori ethnicity thannot. These patterns are consistent with the find-ings of other descriptive studies that used 2000– 2004 birth registration data and the 2001 census(Callister, 2003; Howard & Didham, 2005).

Table 2 focuses solely on children of Maori-European parentage in the NewZealandWomen:Family, Employment and Education survey, pro-viding a more detailed overview of the character-istics of children and their mothers. Of the 298Maori-European children included here, onlyhalf were identified as both Maori and European,much less than if the identification process faith-fully reflected both parental backgrounds. Twofifths of the remaining children were designatedsolely European and three fifths as solely Maori.Of the latter, 40% were so designated by theirEuropean mothers. Indeed, children designatedsolely Maori were the most likely to be labeledas such by their European mothers, consistentwith the hypotheses that they play an importantrole in the intergenerational transfer of Maoriethnic identity. Maori mothers were the mostlikely to identify their child in terms of bothparental backgrounds, perhaps reflecting theirgreater willingness to view Maori ethnicity ininclusive terms (McDonald, 1976). By compari-son, mothers who identified as both Maori andEuropean were the least likely to assign Maoriethnicity to their child, although this varied de-pending on whether they identified primarily asMaori or as European (figures in parentheses inTable 2). The overwhelming majority of Maori-European mothers who self-prioritized as Maoridesignated their child as solely Maori comparedto a handful of mothers who self-prioritizedas European. Because mainly Maori womentended to partner with Maori men (and theopposite was true for mainly European women),it is unclear whether the variation in their desig-nation decisions reflected differences in their

ethnic attachment, partnering choices, or both.With regard to fathers, a strong associationbetween paternal and child ethnicity is clearlyevident, also providing support for the genderedinheritance hypothesis (H3).

Interestingly, there were no differences in theage of children assigned different ethnic labelsnor in their coresidence status, but the status ofthe parental relationship differed. Just less thanhalf of the children designated solely Maori hadparents who were still together compared to twothirds of those designated solely European orMaori and European. These differences may bebetter understood by considering the characteris-tics of their mothers. Mothers who identified asboth Maori and European were the most likelyto be in a relationship with the child’s fatherand to be married. Much lower proportions ofMaori women and European women were ina union with their child’s father, and the differ-ence was even more pronounced among thosewho identified their child solely asMaori (figuresin parentheses in Table 2). Of the Europeanmoth-ers who designated their child as solely Maori,only 42% were in a relationship with the child’sfather, and for Maori mothers, it was 29%. Itmay be that couples who are intermarried butwho share at least one ethnic affiliation may bemore likely to stay together. It also suggests thatthe willingness of European women to identifytheir children as Maori does not depend on theirbeing in a relationship with the child’s father. Ifpatriarchal norms about the transmission ofMaori ethnicity exist, they must be diffusedthrough channels other than European mothersand Maori fathers living in the same household.

The lack of ethnic differences in maternal eco-nomic status and residential location is curious,given the well-documented disadvantage ofMaori relative to Europeans and their greaterpropensity to live in areas in which they are over-represented (Te Puni Kokiri/Ministry of MaoriDevelopment, 2000). Additional analysis of allmothers in the survey (not shown here) suggeststhat the lack of differences results from theselectivity of intermarried mothers. When theanalysis was extended to all mothers, ethnic dif-ferences in educational attainment, labor forcestatus, and residence were found, and in the ex-pected direction.

The bivariate analysis in Table 2 providessome support for H1 (European women readilyidentify their children as Maori) and H3 (the des-ignation of Maori ethnicity is influenced by

1156 Journal of Marriage and Family

Page 8: White Mothers, Brown Children: Ethnic Identification of Maori-European Children in New Zealand

Maori paternity) but not for H2 (European moth-ers who assign Maori ethnicity to their childrendo so out of a sense of ethnic awareness). It islikely that themeasures used are poormanifest in-dicators of the latent and infinitely more complexconstruct of ethnic awareness and that morenuanced indicators would better capture its ef-fects. With regard to the explanatory power ofH1 and H3, further clarity can be revealedthrough multivariate analysis in the form of mul-tinomial logistic regression. Table 3 considersseparately the direct effects ofmaternal and pater-nal ethnicity on maternal designation decisions.

Model 1 shows that, when other factors such asthe status of the parental relationship are ac-counted for, an association betweenmaternal eth-nicity and labeling decisions still exists. The oddsof a European mother identifying her child assolely Maori and as both Maori and European(vs. European) were 3.7 times and 2.8 timeshigher than those of Maori-European mothers.For Maori mothers (vs. Maori-European moth-ers), the odds of assigning Maori ethnicity werealso higher but not significantly so. Separatemodels (not shown) were also run that includedinteraction terms for tertiary-educated European

Table 2. Characteristics of Children With Maori and European Parentage and Their Mothers in the New Zealand Women:

Family, Employment, and Education Survey

Child

Maori European Maori and European n

Total (N) 82 64 152 298

Median of age (years) 10.7 9.4 9.4 298

Resides with mother 80.5 81.3 86.8 250

Parental union intact* 48.8 68.8 65.1 183

Mother’s ethnicity*

Maori 25.6 23.4 30.3 82

European 40.2 17.2 29.6 89

Maori and European 34.2 59.4 40.1 127

Father’s ethnicity***

Maori 70.7 17.2 37.5 126

European 29.3 82.8 62.5 172

Mother’s ethnic orientationa,*

Maori (***) 54.0 (90.9) 30.0 (8.8) 47.4 (40.9) 123 (41)

European 46.0 (9.1) 70.0 (91.2) 52.6 (59.1) 148 (59)

Mother

Maori European Maori and European n

Total (N) 82 89 127 298

Median age (years) 37.9 37.5 36.6 298

Married* 43.9 48.3 60.6 156

In a relationship with child’s fatherb,*** (**) 52.4 (28.6) 51.7 (42.4) 73.8 (71.4) 182 (40)

Has two or more dependent children** 35.4 47.2 57.5 144

Has a tertiary qualificationb

34.2 (38.1) 43.8 (54.6) 41.7 (39.3) 120 (37)

Is in paid work 50.0 59.6 52.0 160

% of Maori living in territorial authority 20.1 (16.3) 16.2 (16.9) 18.2 (20.2) 298 (82)

Difference in Maori and total personal

median income in territorial authority (000)

4.4 4.5 4.4 298

Note: Pearson chi-square and one-way analysis of variance tests were used to test bivariate associations and differences in

means, respectively.aExcludes mothers with no main ethnicity (n ¼ 27). Figures in parentheses are for Maori-European mothers.

bFigures in

parentheses are for mothers who identified their child exclusively as Maori.

*p, .05. **p, .01. ***p , .001.

Ethnic Identification of Maori-European Children 1157

Page 9: White Mothers, Brown Children: Ethnic Identification of Maori-European Children in New Zealand

mothers and European mothers who were still ina relationshipwith their child’sMaori father. Nei-ther term was significant nor did their inclusionyield better fitting models. The child’s coresi-dence status and mother’s marital status wereexcluded as control variables because of theirhigh correlations with other included variables.

Model 2 reconfigures maternal ethnicity todistinguish between Maori-oriented andEuropean-orientedmothers and shows a clear dif-ference in their designation decisions. Motherswho identified solely or primarily as Maori werethree times more likely than European-orientedmothers to identify their child solely as Maori.The influence ofMaori paternity onmaternal des-ignation decisions is tested in Model 3; its effectappears to be very strong. Although the odds ofa child being identified exclusively asMaori werealmost 12 times and 3 times higher, respectively,if the father (vs. themother)was theMaori parent,care is needed in interpreting this result. WhenModel 1 was re-run, and the base category waschanged to European mothers, no significant dif-

ferences were found between their designationdecisions and those of Maori mothers. The effectof Maori paternity must therefore result from thedesignation decisions of Maori-European moth-ers. The labels that these mothers chose for theirchild varied greatly, depending on whether thefather was Maori or European. Of the 37 Maori-European mothers who had a child with a Maoriman (in which cases the father was coded as theMaori parent), almost two thirds identified theirchild solely as Maori. By comparison, very fewof the 90Maori-Europeanwomenwhohad a childwith a European man designated their child assolely Maori.

Given this asymmetry and the small samplesize, it is useful to look at the designation patternsin the 1996 census for further clues. There, wefind that in families where one parent was Maoriand the other was European, children were morelikely to be designated exclusively asMaori if theMaori parent was the mother (34%) versus thefather (30%). In unions involving a Maori anda Maori-European, the effect of Maori maternal

Table 3. Summary of Logistic Regression Analysis for Variables Predicting Maternal Designation of Child as Maori (n ¼ 82)

or Maori and European (n ¼ 152) Versus European (reference category), Controlling for Background Variables (n ¼ 298)

Predictor

Child’s Assigned Ethnicity

Maori Maori and European

B SE B eB

B SE B eB

Maternal ethnicity

Maori .58 .44 1.8 .68 .38 2.0

European 1.31** .45 3.7 1.02* .40 2.8

Constant �.80 �.66

v2

28.13**

df 16

Maternal ethnic orientation

Maori 1.14** .38 3.13 .73* .34 2.08

Constant �.91 .21

v2

22.25*

df 14

Father is Maori 2.46*** .42 11.70 1.09** .38 2.97

Constant �1.97 �.40

v2

60.77***

df 14

% Assigned ethnicity 27.5 51.1

Note: Controls are child’s age, parental union intact, maternal age, maternal education, at least two dependent children, and

% of Maori in territorial authority (omitted from the table). eB ¼ exponentiated B. Model 1: Maori and European mother is the

reference category. Model 2: European-oriented mother is the reference category. Excludes mothers who did not report a main

ethnic group, n ¼ 27. Model 3: Maori mother is the reference category.

*p , .05. **p, .01. ***p, .001.

1158 Journal of Marriage and Family

Page 10: White Mothers, Brown Children: Ethnic Identification of Maori-European Children in New Zealand

ethnicity was especially pronounced. The appar-ently stronger effect of Maori maternity in thecensusmeans it is unlikely that Europeanmotherswho identified their child solely as Maori werefollowing a generalized norm about the patrilin-eal transmission of Maori ethnicity. Rather, theywere doing it for reasons that these data have notbeen able to adequately capture. It is worth notingthat, in the census, only 3% of children witha European mother and a Maori-European fatherwere designated solely Maori, but a third wereidentified exclusively as European. This patternsuggests that the readiness of European mothersto identify their child as Maori is likely to dependon whether the father identifies as solely Maori.There are a number of possible reasons for thisresult. Maori ethnicity in the broadest sense(e.g., identity, culture, networks) may be lesssalient in families where the Maori parent alsoidentifies as European. In addition, Europeanmothers who partner with Maori men may bemore ethnically aware than those who partnerwith Maori-European men.

CONCLUSIONS

This article has tried to expand the literature onthe intergenerational transmission of ethnicityby considering the role that mainstream parentsplay in decisions about how to label their multi-ethnic child. Obviously, there are limitations tousing survey data to examine a process that ismuch deeper than the single act of designatingan ethnicity to one’s child when asked to bya stranger.Wedo not knowhowclosely these eth-nic labels reflect the child’s self-identification orhow mothers would identify their child in someother context or when asked in some other way.Nor do we know how European fathers may havedesignated their child, given the opportunity, andthus if this dynamic is peculiar to mothers or per-tains to European parents of part-Maori childrenmore generally. But insofar as the ethnic designa-tions thatmothersmadewere likely to reflect theircurrent orientation toward their child, they pro-vide a useful if partial window into the innerworkings of ethnic identificationwithin the rubricof Maori-European families.

A key finding to emerge from this study is theapparent willingness of European mothers toassign Maori ethnicity to their child. That Euro-pean mothers are just as likely as Maori mothersto identify their child as Maori, even if it meansdenying their own ethnicity, challenges the

assumption that minority parents are the onlytransmitters of minority ethnic identity. Rather,the diffusion of ethnicity across generationswould appear to involve a more complex set ofdynamics inwhich bothmainstream andminorityparents participate.

Using quantitative data not specifically de-signed to probe ethnic designation dynamics, thisstudy has incurred the same limitations as manyother parent-child studies that use ethnic labelingdecisions in the census. The most important lim-itation has been operationalizing the factors thatunderlie parental designation decisions. I haveadvanced several hypotheses, but the partialexplanatory power of the models suggests otherfactors ought to be considered. Ideally one wouldwant directmeasures of both parent’s level of eth-nic awareness with respect to affective feelingstoward the minority group (e.g., pride) andawareness of status distinctions. Subjective andobjective measures of the child’s physicalappearance (e.g., skin color) would also be usefulbecause physical markers such as brown skindenote difference and may encourage attributionof Maori ethnicity.

A clear finding of this study is that racial andethnic identity is not ‘‘passed’’ across generationsin a predictable, linear fashion. Many childrenwho have the option of being labeled as Maoriand European are identified as either Maorior European. For the first-generation Maori-European families, where neither parent ac-knowledges mixed ethnic heritage (Daniel,1996), the bias is toward being identified asMaori. The transmission of Maori ethnicity be-comes less predictable, however, as the profileof the parental union becomes more European.Given that couples in which one partner identi-fied as both Maori and European and the otheras European constitute the majority of Maori-European intermarriages, this pattern has long-term implications for the transmission of Maoriethnicity. A critical issue will be the ways inwhich political and popular understandings ofthis expanding Maori-European group evolve.For example, it is unclear whether parents whoidentify their child as both Maori and Europeansee them as having two separate ethnicities ofequal importance or a single hybrid ethnicitythat lies between both groups (Howard & Did-ham, 2005). An older study conducted byMcDonald (1976) suggested that Maori tendedto view Maori-European identification as abicultural claim of belonging to both groups,

Ethnic Identification of Maori-European Children 1159

Page 11: White Mothers, Brown Children: Ethnic Identification of Maori-European Children in New Zealand

whereas Europeans tended to interpret it as anidentity that transcended both groups. Unfortu-nately, there have not been any recent studiesthat might provide us with clues about whetherparents interpret the meaning of Maori-Euro-pean identity differently, depending on theirown ethnic identification.

Finally, understanding the factors associatedwith ethnic identification in New Zealand mayprovide insights into the future trajectories ofother indigenous peoples in placeswheremultira-cial reporting has not yet come of age. Althoughgroups have their own distinctive histories, thereare some striking similarities between the experi-ences of Maori and those of Native Americans(Snipp, 1997), Canadian Aboriginals (Guimond,1999), and Native Hawaiians (Kana’iaupuni &Liebler, 2005). They include high rates of inter-marriage with the majority group, differentiationin legal and policy contexts that include particularrights, and overrepresentation in the lower socio-economic strata. In recent years, there has beenmuch interest in how children of mixed indige-nous-mainstream unions are identified, in partbecause their designation decisions affects thesize and structure of indigenous populationsand, with it, the allocation of resources. Thatindigenous peoples have not faded into the main-stream makes them an interesting focal group forstudying processes of ethnic change and reten-tion, both within the nexus of families and atthe macrolevel.

NOTE

This paper has its genesis in my MA thesis, which exploredMaori–non-Maori intermarriage within the broader contextof Maori ethnic identification patterns. Earlier versions werepresented at the Annual Meeting of the Population Associa-tion of America in 2005 and the ‘‘How do identities matter’’?workshop at Stanford University.Many thanks are owed to C.Matthew Snipp, Monica McDermott, Michael Rosenfeld,Carolyn Liebler, and several anonymous reviewers for theirinsightful comments. Any errors or omissions are my own.

REFERENCES

Agresti, A. (1996). An introduction to categoricaldata analysis. New York: Wiley.

Archie, C. (2005). Skin to skin: Intimate, true storiesof Maori-Pakeha relationships. Auckland, New

Zealand: Penguin.

Brown, P. G. (1983). An investigation of official eth-nic statistics. Wellington, New Zealand: Depart-

ment of Statistics.

Buck, P. H. (1924). The passing of the Maori. Wel-

lington, New Zealand: Government Printer.

Callister, P. (2003). The allocation of ethnicity tochildren in New Zealand: Some descriptive datafrom the 2001 census. Paper presented at the Popu-

lation Association of New Zealand Conference,

Christchurch, New Zealand.

Callister, P. (2004). Maori/non-Maori intermar-

riage. New Zealand Population Review, 29(2), 89 –105.

Callister, P., Didham, R. A., & Potter, D. (2005).Eth-nic intermarriage in New Zealand. Statistics NewZealand working paper. Wellington, New Zealand:

Statistics New Zealand.

Daniel, G. R. (1996). Black and white identity in the

new millennium: Unsevering the ties that bind. In

M. P. Root (Ed.), The multiracial experience:Racial borders as the new frontier (pp. 121 – 139).

Thousand Oaks, CA: Sage.

Davis, F. J. (1991). Who is Black? One nation’sdefinition. University Park: Pennsylvania State

University.

Gans, H. J. (1997). Toward a reconciliation of

‘‘assimilation’’ and ‘‘pluralism’’: The interplay of

acculturation and ethnic retention. InternationalMigration Review, 31, 875 – 892.

Gordon, M. M. (1964). Assimilation in American life.New York: Oxford University Press.

Greenland, H. G. (1984). The politics of Maori cul-tural revival. Unpublished MA thesis, Univeristy

of Auckland, New Zealand.

Guimond, E. (1999). Ethnic mobility and thedemographic growth of Canada’s Aboriginal pop-ulations from 1986 to 1996. In A. Belanger (Ed.),

Report on the demographic situation in Canada1998 – 1999 (Statistics of Canada Catalogue no.

91-209-XPE, pp. 187 – 200). Ottawa: Statistics

Canada.

Harre, J. (1966).Maori and Pakeha: A study of mixedmarriages in New Zealand. Wellington, New Zea-

land: Reed.

Harris, D., & Sim, J. (2002). Who is multiracial? As-

sessing the complexity of lived race. AmericanSociological Review, 67, 614 – 627.

Holmes, K., Murachver, T., & Bayard, B. (2001).

Accent, appearance and ethnic stereotypes in New

Zealand. New Zealand Journal of Psychology, 30,79 – 86.

Howard, S., & Didham, R. A. (2005). Ethnic inter-marriage and ethnic transference amongst theMaori population. Wellington, New Zealand: Sta-

tistics New Zealand.

Kana’iaupuni, S. M., & Liebler, C. A. (2005). Pon-

dering Poi Dog: Place and racial identification of

1160 Journal of Marriage and Family

Page 12: White Mothers, Brown Children: Ethnic Identification of Maori-European Children in New Zealand

multiracial Native Hawaiians. Ethnic and RacialStudies, 28, 687 – 721.

Kauanui, J. K. (2005). The multiplicity of Hawaiian sov-

ereignty claims and the struggle for meaningful auton-

omy. Comparative American Studies, 3, 283 – 299.King, M. (1991). Pakeha: The quest for identity in

New Zealand. Auckland, New Zealand: Penguin.

Kukutai, T. (2004). The problem of defining an eth-

nic group for public policy: Who is Maori and why

does it matter? Social Policy Journal of New Zea-land, 23, 86 – 108.

Liebler, C. A., & Kana’ianupuni, S. M. (2003/2004).

Pacific identities: Patterns in the racial identifica-

tions of mixed-race Pacific Islanders. Journal ofIntergroup Relations, 30(4), 23 – 48.

Marsault, A., Pool, I., Dharmalingham, A., Hillcoat-

Nalletamby, S., Johnstone, K., Smith, C., &

George, M. (1997). Technical and methodologicalreport: New Zealand Women: Family, Employmentand Education Survey. Hamilton, New Zealand:

University of Waikato.

McDonald, G. (1976). The categories Maori and

Pakeha as defined by research workers and by self-

report. New Zealand Journal of Educational Stud-ies, 11, 37 – 49.

Nagel, J. (1994). Constructing ethnicity: Creating

and recreating ethnic identity and culture. SocialProblems, 41, 1001 – 1026.

Nagel, J. (1995). American Indian ethnic renewal:

Politics and the resurgence of identity. AmericanSociological Review, 60, 947 – 965.

O’Donoghue, M. (2004). Racial and ethnic identity

development in White mothers of biracial, Black-

White children. AFFILIA, 19, 68 – 84.

Office of the Registrar-General. (1892). Results ofa census of the colony of New Zealand. Welling-

ton, New Zealand: Author.

Perlmann, J., & Waters, M. (Eds.) (2002). The newrace question: How the census counts multiracialindividuals. New York: Russell Sage Foundation.

Pool, I. (1991). Te Iwi Maori: A New Zealand popu-lation past, present & projected. Auckland, NewZealand: Auckland University Press.

Portes, A. (1984). The rise of ethnicity: Determinants

of ethnic perceptions among Cuban exiles in Miami.

American Sociological Review, 49, 383 – 397.Qian, Z. (2004). Racial/ethnic identification of chil-

dren of intermarried couples. Social Science Quar-terly, 85, 746 – 766.

Roth, W. D. (2005). The end of the one-drop rule?

Labeling of multiracial children in black intermar-

riages. Sociological Forum, 20, 35 – 67.

Saenz, R., Hwang, S., & Aguirre, B. (1995). Persis-

tence and change in Asian identity among children

of intermarried couples. Sociological Perspectives,38, 175 – 194.

Smith, L., McCalman, J., Anderson, I., Smith, S.,

Evans, J., & Beer, J. (2006). Fractional identities:The political arithmetic of Aboriginal Victorians.Paper presented at the International Workshop on

Indigenous Identity in Demographic Sources, Umea

University, Umea, Sweden.

Snipp, C. M. (1997). Some observations about racial

boundaries and the experiences of American Indi-

ans. Ethnic and Racial Studies, 20, 669 – 689.

Statistics New Zealand. (2001). Census 2001, ethnicgroups (Table 2a). Wellington, New Zealand: Author.

Statistics New Zealand. (2004). Report of the reviewof the measurement of ethnicity. Wellington, New

Zealand: Author.

Tafoya, S. M., Johnson, H., & Hill, L. E. (2004).

Who chooses to choose two: Multiracial identity ofchildren with parents of different races. Washing-

ton, DC: Russell Sage Foundation and Population

Reference Bureau.

Te Puni Kokiri/Ministry of Maori Development.

(2000). Progress towards closing the social andeconomic gaps between Maori and non-Maori: Areport to the Minister of Maori Affairs. Welling-

ton, New Zealand: Author.

Thomas, D., & Nikora, L. (1998). Maori, Pakeha and

New Zealander: Ethnic and national identity among

New Zealand students. Journal of InterculturalStudies, 17, 29 – 40.

Twine, F. W. (1999). Bearing blackness in Britain:

the meaning of racial difference for White birth

mothers of African-descent children. Social Identi-ties, 5, 185 – 209.

Walker, R. (1990). Ka whawhai tonu matou: Strug-gle without end. Auckland, New Zealand: Penguin.

Waters, M. (1990). Ethnic Options: Choosing identi-ties in America. Berkeley: University of California

Press.

Webster, S. (1998). Patrons of Maori culture: Power,theory and ideology in the Maori renaissance.Dunedin, New Zealand: University of Otago Press.

Wetherell, M., & Potter, J. (1992). Mapping the lan-guage of racism: Discourse and the legitimation ofexploitation. New York: Columbia University

Press.

Xie, Y., & Goyette, K. (1997). The racial identifica-

tion of biracial children with one Asian parent:

Evidence from the 1990 census. Social Forces, 76,547 – 570.

Ethnic Identification of Maori-European Children 1161