Not Indian, Not African: Classifying the East African Asian Population in Aotearoa New Zealand

Abstract

This paper explores the challenges of measuring and classifying the East African Asian population in Aotearoa New Zealand. As a particularly diverse country, New Zealand has a significant and varied population of immigrants from South Asia, including India, Pakistan and Bangladesh, along with immigrants of South Asian origin, from Fiji, Southeast Asia, the Caribbean and East Africa. New Zealand’s system of ethnic classification relies on self-identification, with a broad definition of ethnicity encompassing heritage, ancestry, culture, language and feelings of belonging. However, the collection of this information at a granularity that enables detailed analysis is constrained for the South Asian population, regardless of origin or identification. People are typically presented with the choice of selecting “Indian” ethnicity as a tick box, or providing ethnicities under “Other” as write-in descriptors, which in turn are coded to a limited set of categories within the classification being used. This practice potentially conceals a diversity of ethnicities, which can only partially be hinted at by responses to questions relating to religions, languages and birthplaces, especially for second or third-generation descendants of migrants. Ethnic classification at the highest level, moreover, includes East African Indians as Asian, rather than African, reflecting diasporic heritage as a shorthand for ancestry and overlooking the relevance of layers of identity associated with the double diaspora. Drawing on Peter J. Aspinall’s work on collective terminology in ethnic data collection and categorizing the “Asian” ethnic group in the UK, this paper looks at the overlaps and disconnects between heritage, ethnicity and national belonging in classifying less clearcut identities. We explore the strengths and limitations of New Zealand’s self-identification approach to ethnic identity, and query what exactly is being asked of groups on the margins: when self-identification does not match external perception, are we looking for geographic, cultural, or genetic origins? A focus on the East African Asian population in Aotearoa highlights the complexity of identity for diasporic groups with distant ancestral links with India, as lived experience of cultural connection extends far beyond the bounds of ethnicity, language and even ancestry.

1. Introduction

As a small country of just over 5 million people in the South Pacific, Aotearoa New Zealand has a particularly diverse and transnational population. With a long history of settler colonialism, contemporary Aotearoa is a country of immigration, and nearly 29% of the population was born overseas, with Britain, China and India representing the three largest migrant sources (Sawyer 2013; Spoonley 2015; Wood 2022). Today, as an officially bicultural nation with a multicultural population, there are over 200 ethnic groups in the country, and 17.3% of the population falls into the official category of Asian (Statistics New Zealand 2024). As a way of addressing and better measuring this diversity, New Zealand’s system of ethnic classification has developed and changed over the past decades, and now relies on self-identification, with a broad definition of ethnicity encompassing heritage, ancestry, culture, language and feelings of belonging.

This paper explores the challenges of measuring and classifying such diversity by focusing on the East African Asian population: a small group with complex origins, which does not fit easily within the current classification frameworks. New Zealand has a significant and varied population of immigrants from South Asia more broadly, including from India, Pakistan, Sri Lanka and Bangladesh, along with immigrants of South Asian origin or descent, who immigrated from elsewhere, such as from Fiji, Southeast Asia, the Caribbean and East Africa. This distinction proves important in terms of migration paths and cultural connections: with key differences between direct migrants, who have shifted directly from the Indian sub-continent, and “twice migrants” like the East African Asians (Bhachu 1985, 1990).

The South Asian populations in Aotearoa have a long history, with small numbers of Indian migrants initially arriving as temporary sojourners in the 1800s and early 1900s, many en route to or returning from work as indentured labourers in Fiji (Nachowitz 2015; Beaglehole 2013; Leckie 2007). Numbers of migrants were low, remaining in the hundreds before 1920, and much like the early Chinese population, Indian migrants experienced significant racism from the white British settlers, being described as amoral “Hindoos”. However, unlike the Chinese, the Indians were British subjects, often arriving as domestic servants of British military and civil service migrants from India, making it initially more difficult to pass legislation that would explicitly exclude them from the new British colony (Beaglehole 2013; Hussain 2019; Leckie 2021). However, by 1920, the government passed legislation to limit Indian immigration, except for those who had family already in New Zealand, effectively preventing new migrants from arriving (Leckie 2007, 2021). In the 1970s, immigration policy liberalization ended the de facto white immigration policy, but it was not until the 1987 Immigration Act that migration from South Asia was able to meaningfully increase (Leckie 2021; DeSouza 2011). By 2013, there was a significant population of South Asian heritage, with more than 143,000 people describing their ethnicities as Indian, 10,000 as Fijian Indian, 9000 as Sri Lankan and 3000 as Pakistani. A decade later, in 2023, these numbers were even higher, with 255,264 people describing themselves as Indian, 23,808 as Fijian Indian, 23,661 as Sri Lankan and 8091 as Pakistani. India is now the largest source of migrants (by citizenship) for the country, with an estimated 46,400 (±600) arrivals and an estimated 4300 (±200) departures for the year ended May 2024 (Statistics New Zealand 2024).

The South Asian communities in Aotearoa are thus diverse in their origins, and not at all homogenous in terms of culture, religion, language and transnational ties: there have been multiple waves of immigration and diaspora formation in the country (Leckie 2007; Hussain 2019). However, as in other Western contexts, including the UK, South Asian immigrants are often perceived as sharing a generalized “South Asian culture”, regardless of their origins, migration histories or cultural backgrounds (Bhachu 1990; Gilbertson 2007). This diverse population, regardless of origin or identification, is then typically presented with the choice of selecting “Indian” ethnicity as a tick box or, at best, providing ethnicities under “Other” as write-in descriptors, concealing a diversity of ethnicities, religions, languages and birthplaces. Ethnic classification at the highest level in New Zealand includes all Indian ethnicities as Asian. East African Indians and South African Indians are thus coded as Asian, rather than African, reflecting diasporic heritage as a shorthand for ancestry and overlooking both the relevance of layers of identity associated with the double diaspora and the high levels of diversity of communities subsumed under that label.

We seek to draw out some of these complexities of classification, heritage and identification, drawing on Peter J. Aspinall’s work on collective terminology in ethnic data collection. Using the example of the East African Asian population in Aotearoa, this paper looks at the overlaps and disconnects between heritage, ethnicity and national belonging in classifying less clearcut minority identities. Self-identification does not always match the available categories, as experienced by the first author’s father, who is East African Asian and deliberately chooses to avoid the “Indian” tick box, while also not describing himself as Indian in everyday life1. As Aspinall notes, a wide range of terminology is used (often inconsistently) to describe minority ethnic groups, and broad, pan-ethnic terms such as Indian and Asian are often cumbersome and ambiguous (Aspinall 2002). Across the Western world, a mix of ethnic, racial, national and cultural terms are used to describe and classify population groups, often reflecting the prevailing political ideology of the state in question: as illustrated by the race/ethnicity duality highlighting Hispanic origin in the US, and the particular use of Asian to refer to South Asia in the UK (Aspinall 2002, 2003), as well as the shift towards self-identified ethnicity in New Zealand.

2. East African Asians

This paper focuses on the identities and classifications of East African Asians in Aotearoa, a group with a particular, multi-generational history of complex migration and settlement. As descendants of an immigrant community from South Asia that settled in East Africa, primarily Kenya, Uganda and Tanganyika (Tanzania) (Kahyana 2003), East African Asians are also sometimes described as Ugandan or Kenyan Asians (Anders et al. 2018), Kenyans/Ugandans of South Asian origin (Herzig 2010) or more broadly as Asian Africans (Ozawa 2016). East African Asians were born or have roots in Africa, although their family migration paths began in South Asia (Mattausch 1998; Ozawa 2016). As Murji describes, this identity proves multifaceted in terms of personal identification, creating an “apparent paradox of looking Asian but feeling some greater sense of connection to Africa” (Murji 2008, p. 180). There are important ethnographic and sociological dimensions at play, as Brooks and Singh succinctly noted: “it is apparent to us that the East African experience is so important that it cuts across ethnic divisions and … it is appropriate to group East African Asians together, whilst recognizing their ethnic ‘identity’” (Brooks and Singh 1978).

Trade and migration links between South Asia and East Africa pre-date European colonialism, with cross-regional contact and settlement occurring for millennia (Cosemans 2022; Anders et al. 2018, 2021; Loimeier 2018; Ozawa 2016; McLaughlin [2014] 2021). However, it was British colonial rule that triggered large-scale migration and settlement from South Asia in the late nineteenth century, with 38,000 indentured male labourers moving from India to Uganda and Kenya to build the railways. While most of these forced migrants eventually returned to British India (Dyson 2018, p. 155), around 7000 remained in East Africa. They were eventually joined by growing numbers of migrants, including increasing numbers of women (Hand 2011), often from famine-hit Gujarat at the turn of the century: encouraged by the British and seeking better opportunities by shifting from one colonial territory to another (Ozawa 2016; Anders et al. 2018; Mattausch 1998). Many of these migrants worked as traders, teachers, shopkeepers or small business owners, as well as lower-level colonial administrators and soldiers (Cosemans 2022; Herzig 2010). This migrant population was diverse, divided along ethnic and religious lines, as well as caste, migratory path and region of origin. Some of the larger communities included Sikh, Visha Oswal, Ismaili Khoja, Bohra, Lohana, Patel, Gujarati Brahmin, Parsee, Punjabi Hindu and Muslim, and Goan (Herzig 2010; Held 2022, p. 41).

As in many other colonized contexts, colonial East Africa was firmly racially stratified: the white population made up the upper class, the Asians formed the middle class, and the majority African population were at the bottom (Kahyana 2003; DeSouza 2011). As Cosemans describes: “The general perception was that the British had wealth and power, the Asians wealth without power and the Africans neither wealth nor power” (Cosemans 2022, p. 211). The Asian population enjoyed greater economic success than the Africans, frequently working in the public sector and viewed as a buffer between the colonizer and the colonized (Ozawa 2016; Mattausch 1998). Of course, this stereotype of the financially savvy, Asian “middleman” foregrounds men and their socioeconomic roles, often overlooking the positioning of Asian women in colonial East Africa. Hand’s 2011 analysis of women’s memoirs highlights the ambiguities and diversities in the roles of the Asian population, in terms of both socioeconomic opportunities and the ways in which Asian wives and daughters sought to create their own space within rigid colonial hierarchies and family structures, finding new ways to belong in-between ethnic groups. Overall, the Asian migrant population grew rapidly from 1900, as seen in Kenya, where the Asian population went from 11,800 in 1911 to 176,000 in 1962 (Herzig 2010). In 1948, there were 33,767 people identified as Indian and 1448 as Goan in the Ugandan population, which compared with only 3448 Europeans, in a total population of just under 5 million (Morris 1968, p. 182). Across the whole of East Africa, prior to independence in the 1960s, there were roughly 364,500 East African Asians, most of them holding UK passports and classified as Citizens of the UK and Colonies (Rathore-Nigsch and Schreier 2016; Cosemans 2022).

The decolonization of East Africa in the 1960s radically altered the position of the Asian population. The colonial administration had ensured a relatively privileged social and economic status for the Asian communities while also maintaining a significant degree of segregation from the African population (Anders et al. 2018). The British had at the same time promoted a narrative of Asians as exploitative, deflecting attention and hostility from the white colonizers towards the in-between Asian population (Kahyana 2003; Cosemans 2022). Upon independence, this privilege disappeared, but the segregation and negative portrayals remained, creating significant tension between the African nations and the Asians in the newly independent states. As the new states worked towards post-colonial governance through policies of “Africanisation”, the Asian populations were highlighted as not belonging: lacking true blood ties to the land, being unethically successful, and uncomfortably linked to the British colonizers (Anders et al. 2018; Held 2022).

The independent governments in Nairobi, Kampala and Dar Es Salaam each passed legislation directed at the Asian population, limiting property rights, curtailing business and trading, and restricting the right to citizenship and residence. As many then lost the right to live and work in their home countries, Asians began to leave the region from the late 1960s. These targeted policies in the region had a significant impact. For example, in 1967–68, more than 40% of the Asian population left Kenya as a result. This culminated in August 1972, when Ugandan President Idi Amin ordered the expulsion of all Asians in 90 days, expelling an entire population of 60,000 from the country (Mattausch 1998; Cosemans 2022).

The East African Asians leaving Africa were then faced with the question of where to go. As holders of British passports, and in some cases citizens of Britain2, for many, Britain seemed like the obvious choice in leaving Kenya, Uganda and Tanzania. The 1962 Commonwealth Immigration Act allowed British passport holders living in independent Commonwealth countries to enter the UK, a law passed to facilitate the return of white Britons and their families (Anders et al. 2018; Mattausch 1998). However, as the dramatically deteriorating position of Asians in East Africa became clearer, it was understood that there were a significant number of non-white British passport holders who now needed a place to live. In the face of growing hostility to non-white immigration, the 1968 Commonwealth Immigration Act was passed, preventing the entry of UK passport holders from Independent Commonwealth countries, unless they met the requirements of descent: having one parent or grandparent who had been born or naturalized in Britain. In reaction to this legislation, India (which already did not allow dual citizenship) also reversed its policy of free entry, cutting off a second option for migration (Cosemans 2022; Mattausch 1998; Murji 2008). This deliberately exclusionary legislation3 left much of the East African Asian population stateless: pushed out of their homes and disavowed by their former colonial power, as Britain pressed forward with the position that former subjects, regardless of their passports, were not entitled to full citizenship (Anders et al. 2018; Mattausch 1998; Sawyer 2013).

As the British empire further disintegrated, the Asian migrant population of Africa became the collateral damage of decolonization, undergoing a second stage of migration and dispersal around the world (Aspinall and Chinouya 2016; Cosemans 2022). Although Britain largely denied responsibility for those fleeing or forced out of East Africa, Amin’s actions compelled the British government into accepting a larger number of migrants than they would have liked, while they also sought assistance from the international community, “… to shoulder part of the burden of winding up empire” (Cosemans 2022, p. 212). Assisted by UNHCR, East African Asians were relocated all over the globe: the first group of refugees to be resettled by the international community on this scale (Cosemans 2022). The Ugandan Asians were the most extreme example, with 27,000 moving to Britain, 6000 to Canada, 1000 to the USA, 1000 to India and Pakistan, and 4000 to twenty-five other countries, including Malawi, Australia and New Zealand (Anders et al. 2018; Rathore-Nigsch and Schreier 2016; Cosemans 2022).

Described by Bhachu (1985) as “twice migrants”, the East African Asian diaspora now exists across the world, with complex and multiple migration routes and feelings of belonging: a transnational, diverse group with unique historical, geographic and cultural connections (Murji 2008). Categorization of this group is not easy. While this could be seen as an identity defined by lived migration “routes” rather than generational “roots” in India, it is important to note that these routes and roots can intersect and overlap (Murji 2008). Such migratory paths were not always simple, and not always only “twice”: some East African Asians moved multiple times, seeking and being denied access to Britain and shifting between third and fourth destinations as a result (Cosemans 2022; Murji 2008). This lived complexity makes classification more difficult, complicating what it means to belong, and how home is understood (Ozawa 2016).

There is a small population of East African Asians in Aotearoa, although the actual count is not available from the census. The absence of a specific category, and the presence of an “Indian” tick-box, means that the majority would have been coded as the more general “Indian”, with a very few who may have responded with a specific identity such as “Ugandan Indian” being coded to “Indian not elsewhere classified”. Small numbers migrated to New Zealand in the 1960s as the East African countries became independent. In 1972, as Amin expelled the East African Asian population from Uganda, New Zealand agreed to resettle 244 of the refugees at the request of UNHCR and Britain. Initially, the government requested only those who were already in Britain, who held professional and technical qualifications in demand in the country and were competent in English. However, these requirements proved almost impossible to fulfil, and official policy shifted to accept refugees from transit camps in Europe, admitting those who had the greatest need for resettlement (Beaglehole 2013; Leckie 2021). Despite the small numbers, the arrival of this group was controversial, as Asian migrants (South or East Asian alike) were frequently portrayed as unsuitable and not beneficial for society. Regardless of the media debate, upon arrival, the refugees were offered work and accommodation: “The offers, particularly of jobs, suggested that, alongside the suspicion and hostility discussed earlier, there was abundant goodwill towards the refugees” (Beaglehole 2013, p. 74).

Complex Identities: Not Indian, Not African

East African Asians then have multiple ties to distinct and diverse places around the world. The ancestral link to India is often highlighted, perhaps as a result of relying on phenotype for categorization, although such links to South Asia are particularly complex (Herzig 2010; Matthews 2002). For this diasporic community, origins may be linked to India, but this is not their country of origin, their home. Several studies highlight how India holds an ambivalent place for many East African Asians, as for other migrants across contexts and generations: when visiting, it is seen as simultaneously as a return and a reminder of not belonging (see Matthews 2002; Rathore-Nigsch and Schreier 2016 for example; also, Murji 2008). Instead, such an identity seems to lean towards a deterritorialized type of transnationalism, complicating how we understand the relationships between space, place and belonging (Matthews 2002).

The links with Africa seem very different. For many, Africa represents a homeland, a place where the community was rooted before migration (Ozawa 2016). While India is tied to heritage, Africa signifies belonging, with the culture and linguistic mixtures developed by East African Asians remaining distinct from direct migrants from India (Rathore-Nigsch and Schreier 2016). The shared experiences of Asian migrants in Africa have created a feeling of cultural connection; although placed in historical context, it is not easy for East African Asians to claim an African identity for themselves (Held 2022; Murji 2008).

At the intersection of these two complex country connections, East African Asians have created cultural markers of belonging based around language, food, religion and shared histories of (twice) migration. Language highlights this complexity, with many East African Asians speaking multiple languages across generations, including English and often two or more Indian languages such as Gujarati, Punjabi, Sindhi, Marathi, Konkani and Kutchi, as well as Swahili from East Africa4 (Rathore-Nigsch 2015; Rathore-Nigsch and Schreier 2016). Religion has also proven a particularly important form of identity, highlighting the diversity of the Asian migrants to Africa, and the continued strength of these religious and cultural links: religion provides an alternative, less nationally tied identity, as seen in Matthews’ study of the Ismaili community in Canada (Matthews 2002). Murji (2008) suggests that for this community, now largely settled in Western countries such as Canada, the UK, the US and New Zealand, as for many second-generation immigrant groups, this new national affiliation also becomes important. However, it is tempered by the social/ethnic relations in the new country of settlement, with significant weight still being attached to the colour of one’s skin. He goes on to say, “… although we might call ourselves Asian, or be seen as such, it is an imagined and diasporic identity that conceals a good deal about the histories and geographies of movement, settlement and identity” (Murji 2008, p. 179).

3. Questions of Classification: Asian, South Asian and East African Asian

Peter J. Aspinall authored a significant body of work addressing questions of terminology in ethnic classifications and how these terms of categorization are derived. In particular, he looked at the terms Asian, South Asian and Indian, and their different meanings, national/geographical implications and undertones. Race and ethnicity have been measured in different ways around the world, from the national census and birth/death records to identity cards and household surveys (Aspinall 2012; Morning 2008). Yet, such classifications have been heavily based on the idea of singular, mutually exclusive categories (Aspinall 2017). The national census is often the main form of classification as the most visible and well-known way in which the state measures and classifies identities and communities (Kertzer and Arel 2002). The construction and logic behind each national census are inextricable from the context of each country, developing from colonial pasts, contemporary power dynamics, shifting political interests, international pressure, and each particular historical trajectory of measurement (Rocha and Aspinall 2020; Anderson 1991; Nobles 2000; Kertzer and Arel 2002).

The terms used in ethnic classifications at the state level can, therefore, be seen as “fictive unities … which, although only partially and situationally meaningful to people thus labelled, are nonetheless employed by them in their dealings with officialdom” (Aspinall 2003, p. 94). These can be established either through eliciting respondents’ self-categorizations (as self-ascribed and possibly contradictory), or through the external judgement of meaningful distinctions by scientific observers (as validated and tested for social/external meaning) (Aspinall 2003).

Ethnic classifications in Aotearoa are no exception. The classification of race and ethnicity in Aotearoa New Zealand, has a long history, shifting with colonial and post-colonial governance and management of the population. Aotearoa was one of the first countries to define and measure ethnicity as socially constructed, shifting from a colonial and regressive classification system to one of the most flexible: defining ethnicity as voluntary, individual and multiple, and also recognizing multiple ethnic identifications for more than three decades (Didham and Rocha 2020). These shifts can be traced over the twentieth century, with “race” and fractions of origin/blood quantum (such as half-caste) being used in the national census until 1951, followed by “race” and “descent” until 1970. “Race” disappeared from the census in 1971 entirely, being replaced by “descent” and “origin”, before the 1972 Statistics Act outlined that the census must query “ethnic origin”, conceptualized as self-perceived. The 1976 census then queried “ethnic origin” as subjective, a key shift in data collection and conceptualization (Callister et al. 2006; Callister 2011; Callister and Kukutai 2009; Rocha and Wanhalla 2018; Didham and Rocha 2020). Following the 1983 Department of Statistics report, the census then began to measure “ethnic groups”, recognizing that individuals could belong to more than one group at a time (Statistics New Zealand 2009; Didham and Rocha 2020). The question wording shifted from 1991 to 1996 to 2001, but each question specified the subjective nature of ethnicity and allowed the selection of multiple groups—up to six ethnic groups per respondent by 2001 (Cormack 2010).

This deliberate move away from fixed conceptualizations of race toward subjective understandings of ethnicity has set Aotearoa apart from many other countries (Callister 2004; Howard and Didham 2007). Only 14% of census-taking states conceptualize ethnicity in this way (Morning 2008), with many countries instead using different terms and concepts such as ancestry, birthplace and race, or even avoiding classification around origin altogether (Rocha and Aspinall 2020; Kukutai and Broman 2016). Ethnicity is currently defined as “… the ethnic group or groups that people identify with or feel they belong to. Ethnicity is a measure of cultural affiliation, as opposed to race, ancestry, nationality or citizenship. Ethnicity is self-perceived and people can belong to more than one ethnic group” (Statistics New Zealand 2005, p. 2). Moreover, ethnic mobility is significant, so the direct relationship between ancestry and ethnicity is fluid (Didham 2016).

Statistics on ethnicity in Aotearoa are currently classified at four different levels of detail. The first and highest level provides a range of broad groupings and is often used in research and policy-making. These top-level groups are European, Māori, Pacific Peoples, Asian, Middle Eastern/Latin American/African and Other, and the sum of these groupings can exceed the total population due to multiple inter-group ethnicity selections (Statistics New Zealand 2009). Callister (2011) notes that apart from Māori, these high-level groupings are not individual ethnic groups, but rather broad, highly diverse, collections of groupings of ethnicities: “… they are not strictly ‘who we are’, but whom statistical agencies or researchers group us with” (Callister 2011, p. 121).

The second level contains more specific but still general groupings, often with a national tie, including specific Pacific Island countries (Samoan, Cook Islands Māori, Tongan, Niuean, Tokelauan and Fijian) and broader groups such as Indian, Chinese, Middle Eastern, and Latin American. These are then specified in more detail at the third level, and the fourth most granular level, provides 233 detailed categories, although this does not include a category for East African Asian/Indian5: see

Table 1. Census counts of Level 1 categories as compared to Indian and other South Asian detailed responses 2013–2023.

		2013	2018	2023
Level 1	European	2,969,391	3,297,864	3,383,742
	Maori	598,602	775,836	887,493
	Pacific	295,941	381,642	442,632
	Asian	471,708	707,598	861,576
	MELAA	46,953	70,332	92,760
	Other	67,752	58,053	56,133
	Total Stated	4,011,399	4,699,755	4,993,923
	Not Stated	230,649	0	0
	Total Population	4,242,048	4,699,755	4,993,923
Indian	Indian nfd	143,520	221,916	255,261
	Bengali	165	201	261
	Fijian Indian	10,926	15,132	23,808
	Indian Tamil	300	315	1074
	Punjabi	324	459	4152
	Sikh	213	192	9111
	Anglo Indian	324	381	357
	Malaysian Indian	n/a	474	606
	South African Indian	n/a	1632	2340
	Indian nec	786	348	411
	total	155,178	239,193	292,092
Other South Asian	Afghan	3417	5250	8721
	Bangladeshi	1623	2337	3552
	Eurasian	1368	1389	1296
	Pakistani	3261	6135	8094
	Sinhalese	1020	9171	14,436
	Sri Lankan nec	0	3	0
	Sri Lankan nfd	9561	4245	5100
	Sri Lankan Tamil	732	3501	4221
	Total Sri Lankan	11,274	16,830	23,661

for more detail (Statistics New Zealand 2009; Didham and Rocha 2020).

“Asian” is thus utilized in a unique way in Aotearoa (Rasanathan et al. 2006). It is used as a top-level category in ethnic classification in Aotearoa, but not listed as a tick box on the census, which instead lists:

• New Zealand European
• Māori
• Samoan
• Cook Islands Māori
• Tongan
• Niuean
• Chinese
• Indian
• Other (e.g., Dutch, Japanese, Tokelauan. Please state)

Census respondents are grouped as “Asian” if they select “Chinese”, “Indian” or enter an “Other” response which is classified under the level 1 “Asian” grouping—encompassing East Asian, South Asian, Southeast Asian and Central Asian peoples, but not other ethnicities such as Persian or Turkish which are classified under “Middle Eastern” at level 2 and Middle Eastern/Latin American/African at level 1 (Statistics New Zealand 2009; Rasanathan et al. 2006; Didham and Rocha 2020). Prior to 1996, “Asian” was not widely used as a top-level category except in census, and those respondents were instead often coded as “Other” (Rasanathan et al. 2006). However, Asians were recognized as a complex and growing subpopulation approaching 100,000 people from the mid-1980s, especially following the refugee flows from Southeast and East Asia. They were reported in census statistics as a specific grouping of ethnicities (Statistics New Zealand 1995), albeit excluding Central Asian ethnicities, though also including Iranian ethnicities as Asian. Following the dissolution of the Soviet Union, ethnicities associated with Central Asia could be included under Asian, but from 1996, the category excluded Iranian ethnicities.

Interesting differences can thus be seen between the New Zealand operationalization of “Asian” and the use of the category elsewhere (Aspinall 2003, 2013). “Asian” was highlighted by Aspinall as a particularly problematic term in classification: a very broad term that can describe those originating from continental Asia, or as a shorthand for more specific sub-groups, such as South Asian or East Asian. In the UK, “Asian” is frequently used as a collective term, officially and socially, for those with ethnic/family origins in the Indian subcontinent, as the largest migrant group from Asia in Britain (Aspinall 2003, 2013). “Asian” is used as a broad term in the national censuses of England and Wales and Scotland, as one of the “broad-based ethnicity groupings recognized by the wider society and which provide a point of access to the longer-term processes of ethnogenesis, including Britain’s colonial project in the Indian sub-continent and historical processes of racialization, racism, discrimination and disadvantage” (Aspinall 2013, p. 180).

Used for the first time in the 2001 UK Census, the label sought to capture a range of different populations, but initially excluded East Asians from the classification. In England and Wales, five ethnic options were listed: “White”, “Mixed”, “Asian or Asian British”, “Black or Black British” and “Chinese or other ethnic group”. Under “Asian or Asian British”, respondents could select “Indian”, “Pakistani”, “Bangladeshi” or any other Asian background (write in). The Office for National Statistics (ONS) further specified that certain “write in” responses were then re-classified as Indian (“British Indian” and “Punjabi”), while others were shifted to Pakistani (‘‘British Pakistani” and “Kashmiri”) and others (including “British Asian”, “East African Asian”, “Sri Lankan”, “Tamil”, “Sinhalese”, “Caribbean Asian”, “British Asian”, “Nepalese” and “Mixed Asian”, and “Other Asian, Asian unspecified”) were collected into an “Other Asian” category, which was created for reporting purposes after the census (Aspinall 2000, 2003, 2013). This terminology shifted in 2011 and 2021: “Asian” was still retained as a high-level categorization, but “Chinese” was moved from “Other ethnic group” to sit under the broader “Asian” heading, diversifying the category even further (Aspinall 2013).

By way of comparison, the terms “Asian” and “Indian” are used in different ways in the US and in Canada, reflecting different histories of racialization and meanings around the words themselves (Aspinall 2003). In the US, having formerly grouped large populations together under the heading “Asian and Pacific Islander”, the censuses of 2000, 2010 and 2020 reflected a broad definition of “Asian”. The “race” question included seven Asian groups: “Chinese”, “Vietnamese”, “Filipino”, “Korean”, “Asian Indian” (as distinct from Native American), “Japanese” and “Other Asian” (Aspinall 2013; Rasanathan et al. 2006). In Canada, the question addressed “population group”, and included specific mentions of “South Asian (e.g., East Indian, Pakistani, Sri Lankan, etc.)”, “Southeast Asian” and “West Asian”, a grouping which is not often used elsewhere and included Iranian and Afghan (Aspinall 2013). In Australia, “Asian” frequently refers to groups from Southeast Asia and East Asia (Rasanathan et al. 2006). Aspinall further highlights that the national census of India does not collect data on ethnicity or race, but rather on religion, caste, language and tribe. Thus, “these origin-based group categories are entirely different from what the Indian diaspora populations would have encountered in the censuses of their destination countries” (Aspinall 2013, p. 181).

In New Zealand, the official classification “Asian” includes an extremely diverse grouping of populations, encompassing multiple ethnicities, ancestries and migration pathways (Rasanathan et al. 2006; Callister 2011). Rasanathan et al. highlight that this official meaning does not always align with social and popular meanings: with the social (often media-reinforced), historically-rooted understanding of Asians having a significant racial basis, referring to those who can be identified as from East and Southeast Asia through phenotypical difference (Rasanathan et al. 2006; see Ip and Pang 2006, for further discussion of racialization of Asians). The broader, official categorization is then often used in different public sectors to different ends. In the health sector, for example, ethnicity is often an important indicator (see Aspinall 2000), as it can impact assessment and diagnosis, even targeting interventions to groups with particular susceptibilities to disease. Thus, broader, inherently diverse, categories can prove problematic: both homogenizing an extremely diverse population by assuming shared characteristics, and concealing the reality that many of those classified as Asian, or even Indian, may not identify themselves in this way (Rasanathan et al. 2006). Unlike in the UK, where “Asian” predominantly refers to “South Asian”, “Indian” is not commonly included as “Asian” in social discourse in New Zealand, highlighting this dissonance between popular and official understandings. Within the health sector, “Asian” can thus obscure potentially important differences, while at the same time, excluding those who do not see the category as applying to them (Rasanathan et al. 2006).

Nonetheless, the category “South Asian” is itself too diverse to be usable in some contexts. For those with roots in the Indian subcontinent but without corresponding personal identifications, such as the East African Asians, this can be a particular issue in terms of health. Some research highlights this population as being particularly susceptible to certain health issues, including type-2 diabetes and ischaemic heart disease (Liew et al. 2003; Kuppuswamy and Gupta 2005), a finding not replicated for other groups in the “Asian” category, but also found among other descendants of Indian indentured labour communities, especially among populations descended from survivors of extreme famine events such as the Bengal famine (Seah et al. 2023).

4. Classifying East African Asians

Locating East African Asians within these classifications is, therefore, not simple. Aspinall’s work on the UK census provides several important insights. The East African Asian population itself remains diverse in terms of religious and cultural affiliations, and does not fit easily into the three main UK categories of “Indian”, “Pakistani”, “Bangladeshi” (Murji 2008). In the 2021 census, just over 8000 people chose to write in ‘African Asian’ (plus another 4000 as “mixed” African Asian) (Office for National Statistics 2022): this write-in answer is then reclassified by the ONS into “Other Asian”. Given these small numbers, many of the population instead continue to use ‘Indian’ as a shorthand: as in 2001, where 167,610, or 16% of all those who chose ‘Indian’ in the UK census, were born in Africa, although the numbers of second and later generations (born in the UK) are not easy to ascertain (Aspinall 2013). Choosing ‘African’ as a classification is more difficult, as ‘blackness’ and ‘Africanness’ are often treated as synonymous, while the relationship between them is complex when exploring geographical, historical and ethnic origins (Aspinall 2011). For example, in 2001 UK census only 10.5% of those born in Kenya described themselves as “Black African”, while a significant number were of Indian origin (Aspinall 2011; Aspinall and Chinouya 2016).

In Aotearoa, the numbers of East African Asians are significantly smaller, but the challenges of accurately classifying this population remain. While the country is home to many groups with ancestral origins in the Indian subcontinent, they may also be twice migrants—as “Indians” from Fiji, East Africa or the Caribbean. The category of “Indian” thus itself contains significant diversity in New Zealand and elsewhere. While “Indian” is often used as shorthand for anyone with ancestry in South Asia (see Hussain 2019), there are important differences along generational, linguistic, religious and migration path lines (Callister 2011; Rasanathan et al. 2006; Spoonley 2015). The history of India as a nation-state further complicates diasporic identifications, particularly along religious lines, drawing on the history of partition and the positionings of Hindus, Muslims, Jains and other religious groups. When considering the Indian diasporas, the diversity of the Indian populations in South Asia (Truschke 2019), the interaction between Mughal India and the British Empire (Al-Azami 2024), and consequential crises such as ongoing famines (Dyson 2018) cannot be overlooked.

The use of “Indian” as a tick-box in ethnicity collection instruments thus conceals important population diversity in respect of both the diverse populations in India and the equally diverse Indian diaspora, conflating multiple ethnicities within an overarching label: one which has national/state connotations which are not always applicable. East African Asians fall under either the “Indian” ethnicity category (confounded by the presence of an “Indian” tick-box) but may see themselves as African rather than Indian or even Asian, similar to Fijian Indians with connections to the Pacific (Didham and Rocha 2020). As for other twice migrants, cultural connection and belonging go far beyond ethnicity categories, reaching across language, religion and transnational migration histories (Aspinall 2000; Didham and Rocha 2020).

For East African Asians, we could then estimate numbers by looking at those who write in this response directly, but also by looking at country of birth as related to ethnicity—such as those who select “Indian”, and indicate a country of birth in East Africa. For example, the 2018 census recorded 4965 people who identified as “Indian” and were born in Sub-Saharan Africa, as well as 1740 who were born in Europe and the UK. For those born in NZ, this becomes more complex, as they are not easily visible in the data unless they are children still living at home with at least one Sub-Saharan-born Indian parent. Moreover, ethnicity may not necessarily be the most important form of identity for this group, particularly when measured in such a crude manner. Along with complex geographical linkages, both civic identity and religion provide important markers of identity for East African Asians, as with other South Asian New Zealanders. As Hussain describes, religion, in particular, can be hybridized with national identity in Aotearoa, combining religion as a broad, transcending form of identity with “… Kiwi identity [as] a particularistic national identity associated with citizenship based on birth or achieved citizenship” (Hussain 2019, p. 13).

Despite these complications, in Aotearoa, the census remains the gold standard for measuring and classifying the population. The measurement of ethnicity is particularly complex, and “fictive unities” constructed over time include terms considered races, ethnicities, nationalities, cultures and religions (Aspinall 2003; Didham and Rocha 2020). While New Zealand’s system of self-identification allows for flexibility and personal choice, there are limits, as top-level categories mask more detailed identifications when data are consolidated and utilized in public policy and social sectors.

5. Conclusions

This paper has traced the history and identifications of the East African Asian population in Aotearoa, highlighting the complexity of identity for diasporic groups with distant ancestral links with India. For East African Asians, their lived experience of cultural connection extends far beyond the bounds of ethnicity, language and even ancestry: heritage and belonging are neither simple nor easy to classify. Entwined with experiences of South Asian appearance and ancestry, simultaneous ties to and exclusion from Africa, and contemporary legal belonging in New Zealand, this group exemplifies the challenges of categorizing complex identities and the legacies of colonial classifications (see Murji 2008). Using this group as an example and drawing on the work of Peter J. Aspinall, we have explored the strengths and limitations of New Zealand’s self-identification approach to ethnic identity, querying what exactly is being asked of groups on the margins: when self-identification does not match external perception, are we looking for geographic, cultural or genetic origins?

East African Asian identity provides a valuable case study when exploring changing ethnic classifications, particularly in New Zealand’s flexible and relatively inclusive classificatory context. Increasing ethnic complexity has had inevitable implications for classificatory systems around the world: as populations have become increasingly diverse, transnational and mobile, singular classifications have struggled to accurately reflect the complexity of identities and heritages in multicultural societies such as Aotearoa. This has particularly affected populations with complex histories and realities of belonging, those whose heritages and identities do not easily sit within traditional tools of measurement (Rocha and Aspinall 2020).

As this paper has shown, even the most expansive classification system can fall short when trying to encompass multiple identities and in-between forms of belonging, with identities such as East African Asian highlighting how arbitrary labels and boundaries can be (Murji 2008). For transnational migrants, twice migrants and those with multiple identifications, home and belonging can be complex, and the measurement of these identifications can be even more complicated. As Oonk (2019) observed in a discussion of Mira Nair’s film Mississippi Masala, the interplay of the concepts of ‘home’, ‘origins’, ‘territory’ and ‘disconnection’ runs to the core of diaspora identity.

Looking forward, the measurement and classification of those of South Asian descent may need to become more nuanced if we are indeed seeking to understand both roots and routes. Terms such as “Indian”, “Pakistani”, “Bangladeshi” and “Asian” vary widely in their usage and in their meanings across contexts, and in the face of increasingly complex mixed and transnational identities, further discussion around terminology is needed (Aspinall 2003; Aspinall and Chinouya 2016). As Didham notes, classification cannot be approached lightly, as both terminology and the context in which the politics of identity sit are in continual flux (Didham 2016; Didham and Rocha 2020).

As seen in the use of “Asian” in various Western contexts, such a broad label presents significant problems in terms of concealing diversity, with consequences in terms of counting, population management and even health (Aspinall 1995, 2013). Even the more specific terms, such as “Indian” or “Pakistani”, can be problematic, overlooking important differences in culture, language, religion or migratory history, and potentially resulting in the reification of ‘ethnic’ categories of dubious validity (Aspinall 2000). Fine-grained data are then key, despite the difficulties in collection, allowing us a better understanding of the diversity within and the specific needs of communities. Peter J. Aspinall suggested in 2000 that perhaps other lines of difference may prove fruitful for measurement, exploring the public policy case for including intersecting ethnic, religious and cultural identities (Aspinall 2000). However, perhaps our academic and policy drive for simple, easily bounded identities will not be easily shifted: as Hand queries, do we need to rely so heavily on “clear-cut categories” that keep minority groups in “an ambiguous half-way position” (Hand 2011, p. 102)?