Intermarriage by birthplace and ancestry in Australia.

Khoo, Siew-Ean ; Birrell, Bob ; Heard, Genevieve 等

Sociologists have long regarded intermarriage as a key indicator of ethnic integration. The authors analyse marriage data from the 2006 Australian census. They find that many ethnic groups show low levels of intermarriage in the first generation but that, by the third generation, rates of intermarriage are high. However most groups of migrants from South and East Asia and from the Middle East and Africa have not been in Australia long enough for us to know whether the relationship between length of time in Australia and integration will hold for them as it has for the earlier cohorts of European migrants.

INTRODUCTION

Much interest in the subject of inter-ethnic marriage in Australia stems from the scale of Australia's migration program and concurrent concerns about the extent to which migrants are integrated into Australian society. Overseas studies of intermarriage between immigrants and native-born residents have considered it an important indicator of immigrant integration into the host society. (1) Australian scholars have taken a similar stance. (2) Price (3) has written that 'intermarriage is still the best measure of ethnic intermixture because it breaks down ethnic exclusiveness and mixes the various ethnic populations more effectively than any other social process'. Intermarriage between persons of different ethnic background also affects the social and cultural identities of the next generation, who will be of mixed or multi-ethnic heritage.

The sociological literature suggests that intermarriage by migrants and their descendants will be relatively high where the members of a community achieve upward social mobility. As migrants and their descendants progress through the education system and enter the labour force, the possibility increases of meeting prospective partners outside the community. Participation in schools, universities and the workplace all potentially serve to open up new social relationships and different ways of living which serve to liberate young people from the influence of parents and the ethnic community. The more this occurs, the more those making partnering choices are likely to be influenced by emotional ties developed with prospective partners rather than the preferences of their parents, who may be prescriptive about the ethnic background and economic prospects of the partners of their children.

On the other hand, some migrants belong to communities that place a high value on the maintenance of their values and cultural practices, contributing to strong social cohesion within the group. This may be accomplished by the creation of educational and cultural institutions that limit social encounters outside the community, or even by the proscription of out-marriage. Ethnic endogamy can be seen as an indicator of the strength of group cohesion and ethnic intermarriage as an indicator of its weakening. An additional factor that may contribute to this process in contemporary societies is the extent of electronic communication linkages to the homeland and the relative cheapness of international travel. These developments contribute to the maintenance of the ethnic community's cultural traditions as well as to the ease with which members of the community can return to their homeland to find a spouse.

Intermarriage across ethnic groups may also be related to the social distance between ethnic groups. (4) Persons from ethnic groups that are more similar with regard to social and demographic characteristics, such as educational attainment, residential location and language, for example, are more likely to intermarry because they encounter fewer barriers to social interaction. This hypothesis was supported by an early study of intermarriage among immigrants and the second generation of European ancestries in Australia which shows that persons from ancestry groups that are similar to one another on these social and demographic characteristics are more likely to intermarry. (5)

Using data from the 2006 Australian population census, this paper examines the extent of intermarriage by birthplace and ancestry in Australian society. It compares the intermarriage rate by gender and across successive generations by their ancestry to provide an indication of the extent of intermixing across ethnic groups among second and third generation Australians. It also compares the intermarriage rate by level of education to examine the question of whether education leads to a greater propensity to partner outside the ethnic group, as suggested by the sociological literature discussed above. The paper is part of a larger study of intermarriage in Australia based on the 2006 census data that also examines intermarriage by indigenous status and religion. (6)

DATA AND METHOD

The census asked each person to state his or her birthplace and ancestry. A census guide handed out with the census form suggests that people consider the origins of their parents and grandparents in answering the ancestry question. Individuals can provide a maximum of two main ancestries 'with which they most closely identify, if possible'. (7) Seventy-two per cent of the population stated one ancestry and 28 per cent indicated two ancestries in the census.

In this paper, intermarriage is examined by comparing the country of birth or ancestry of the male and female partners in couple families. Couple families include those who are married as well as those who are in de facto relationships. The analysis is based on only those men and women who state a single ancestry, since the aim is to examine the extent of exogamy among people from each ethnic group. The analysis is also based on couples where both spouses are present in the household on census night, since birthplace and ancestry data are not collected for persons temporarily absent from the household.

For men and women of each specific country of birth or ancestry, the intermarriage rate is calculated as the percentage that is intermarried, which is equal to the number of partnered men (or women) born in country x (or of ancestry y) whose partner is not born in country x (or is not of ancestry y) divided by the total number of partnered men (or women) born in country birth x (or of ancestry y) multiplied by 100.

In this paper, the first generation refers to people who are born overseas and have migrated to Australia. The second generation refers to people who are born in Australia who have one or both parents who are born overseas. The third or more generation refers to people who are born in Australia and whose parents are also born in Australia. It is not possible to separate the third generation from higher order generations.

INTERMARRIAGE BY BIRTHPLACE

The intermarriage rates between Australia-born and overseas-born people are shown by the country of birth of the overseas-born partner in Table 1. The census did not collect information on the timing or place of marriage or the start of a de facto relationship. Therefore, it is not possible to determine if people who are part of a couple where one or both spouses are born overseas have married overseas or in Australia after the arrival of the overseas-born partner(s). Birthplace groups with a low rate of intermarriage with Australia-born persons may be a reflection of the migration of family units (where both spouses would have been born overseas) or a low propensity for exogamy or both.

Table 1: Per cent of overseas-born men and women in couple families
with an Australia-born partner, by country of birth, 2006

 Percent
 Intermarried

Country of birth of overseas-born men/women Male Female

Canada 60.6 60.1
United States of America 57.3 56.6
Thailand 15.6 47.4
Netherlands 50.3 42.3
United Kingdom 43.4 40.8
Japan 14.9 40.6
France 43.8 39.5
Switzerland 40.7 39.2
Germany 45.2 38.6
New Zealand 42.9 38.3
Austria 43.9 36.6
Philippines 8.1 35.6
Ireland 42.3 34.8
Singapore 23.5 28.2
Spain 30.6 26.0
Indonesia 17.4 24.0
Argentina 25.5 24.0
Malaysia 17.0 23.7
Malta 32.7 23.4
Hungary 28.5 22.4
Mauritius 23.6 22.4
Zimbabwe 24.5 21.7
Russian Federation 8.9 21.4
South Africa 22.5 20.7
Chile 19.7 18.3
Poland 18.7 18.0
Fiji 14.2 16.8
Burma (Myanmar) 16.2 16.4
Cyprus 24.6 14.5
Egypt 23.3 14.4
Portugal 19.0 14.1
Italy 29.6 13.5
Ukraine 9.8 12.6
Lebanon 24.3 12.2
Hong Kong 8.5 12.0
Taiwan 2.2 11.8
Romania 12.4 11.3
Croatia 18.3 10.9
Korea, Republic of 1.6 9.9
Turkey 16.8 9.9
Greece 19.4 9.1
Serbia 13.8 9.0
Sri Lanka 10.4 8.9
India 10.7 8.8
Fr Yugo Rep of Macedonia 15.0 8.4
Samoa 10.8 8.3
China 2.4 7.3
Pakistan 10.8 7.0
Iran 9.3 6.8
Viet Nam 2.0 5.2
Cambodia 2.2 5.2
Bosnia and Herzegovina 6.4 4.6
Bangladesh 3.4 2.2
Iraq 4.7 2.1
Sudan 3.7 1.9
Afghanistan 2.7 0.9

Source: 2006 census customised tables

Men and women from North America have the highest rate of intermarriage with the Australia-born, followed by men and women born in the United Kingdom and other Western European countries such as the Netherlands, France and Germany. There is generally no difference between men and women from these countries in their intermarriage rate with Australians. The high rate of intermarriage indicates that there is little social and cultural distance between Australian-born and people from Western European and North American countries.

Women from three Asian countries--Thailand, Japan and Philippines--have much higher intermarriage rates than men from these countries with people born in Australia. This pattern of higher rates of intermarriage for women than men is seen for all the East and Southeast Asian birthplace groups (although the gender difference is not as large as for the three countries mentioned above) and also for migrants from Russia, but not for the South Asian groups, whose intermarriage rates with Australia-born are higher for men than for women. Intermarriage with the Australia-born is also more likely for men than for women from Lebanon, Turkey and other Middle Eastern countries. These gender differences in intermarriage rates are likely to be related to differences in gender roles in Asian and Middle Eastern families. (8)

Birthplace groups with the lowest rates of intermarriage with the Australia-born population are mostly from countries that have been the source of recent refugee and other humanitarian migration, such as Afghanistan, Sudan and Iraq. Their low intermarriage rate reflects the migration of families from these countries, most of whom arrived during the past ten years for resettlement under Australia's Humanitarian migration program.

Table 1 also shows the relatively low intermarriage rates of men and women born in Southern European countries such as Greece and Italy. These rates are related to the migration of family units from these countries in the 1950s and 1960s. With the decrease in migration from these countries after 1970, many of the intra-married couples are now in their older ages. As shown in Table 2, men and women from these countries who are aged 40 and over are much more likely to have a spouse from the same country of origin than are those under the age of 40. Many of the younger men and women are likely to have migrated as children with their parents and to have grown up and partnered in Australia.

Table 2: Per cent of overseas-born partnered men and women with spouses
from the same country of origin by age, 2006

 Men Women

Birthplace Aged 15-39 Aged 40 + Aged 15-39 Aged 40 +

Bosnia 65.1 76.9 73.0 80.5
Canada 10.8 13.8 9.9 12.7
China 87.4 81.0 73.7 74.8
Croatia 33.8 64.6 43.0 76.0
Egypt 43.0 53.0 73.1 62.1
Fiji 70.3 76.1 69.2 64.4
Fr Yugo Rep of Macedonia 38.6 84.2 59.2 90.2
Germany 16.5 27.5 15.3 31.1
Greece 13.0 73.1 22.2 84.5
Hong Kong 42.8 62.7 42.6 56.6
India 81.2 71.5 86.6 69.6
Indonesia 70.7 59.0 45.4 44.5
Iraq 80.3 81.7 88.4 86.1
Ireland 19.3 30.5 22.1 35.6
Italy 10.6 60.1 19.1 80.5
Japan 67.3 70.5 23.1 33.8
Lebanon 36.3 78.4 66.1 84.2
Malaysia 39.6 60.1 32.9 50.5
Malta 9.2 53.5 14.3 62.9
Netherlands 15.3 28.0 16.8 42.2
New Zealand 32.6 41.5 35.4 43.0
Philippines 75.2 92.0 41.1 38.5
Poland 45.5 62.2 41.0 66.8
Singapore 29.5 45.1 25.5 34.0
South Africa 43.9 65.1 47.3 61.8
Sri Lanka 71.6 81.5 81.2 80.8
Thailand 54.9 63.0 16.4 14.4
Turkey 51.9 79.0 71.6 84.9
United Kingdom 26.0 42.3 31.4 47.7
United States of America 12.6 18.0 11.8 19.5
Vietnam 82.7 91.1 79.3 86.6

Source: Calculated from 2006 census customised matrix tables

A very high percentage of migrants from more recent source countries of migration, such as China, Vietnam, India and Sri Lanka, also have spouses who are born in the same country (Table 2). While most of the older migrants would have been married before their migration to Australia, it is possible that some of the younger migrants may have sponsored marriage partners from their country of origin under the family migration program for spouse and fiance(e) migration. Data on spouse and fiance(e) visa grants show that China has been the second largest country of origin (after the United Kingdom), and that Vietnam and India are among the top ten source countries of recipients of the partner visas since the late 1990s. (9) A study of spouse migration shows that more than 85 per cent of migrants arriving on partner visas in 1993 to 1995 from China, Vietnam, India, Sri Lanka, Philippines, Lebanon and Turkey are sponsored for migration by Australian residents who are born in the same country. (10)

INTERMARRIAGE BY ANCESTRY AND GENERATION

Intermarriage rates vary by ancestry and generation (Table 3). The majority of persons of the first generation of most Western European ancestries (except the English) have partners of a different ancestry. The relatively low rate for the English is a consequence of the high proportion of Australia-born persons who are of English ancestry. In contrast, only a minority of the first generation of Southern or Eastern European ancestries have partners of a different ancestry. The proportion is even lower for the first generation of Middle Eastern and Asian ancestries. The low proportions intermarried among men and women of these ancestries partly reflect the migration of family units from these respective regions.

Table 3: Percentage of partnered men and women with a spouse of a
different ancestry, (a) by ancestry and generation, 2006

Ancestry 1st generation 2nd generation 3rd+ generation

 Male Female Male Female Male Female

English 41 36 49 48 20 21
Irish 62 59 86 83 71 67
Scottish 65 60 90 88 80 75
Welsh 71 66 96 96 96 94
Austrian 74 65 98 96 * *
Danish 68 61 98 97 98 97
Dutch 62 55 89 88 95 95
Finnish 44 54 93 93 * *
French 61 60 91 93 98 98
German 59 56 91 90 72 69
Swiss 67 57 94 98 * *
Greek 12 9 37 31 67 61
Italian 22 12 51 42 77 74
Maltese 33 28 67 64 79 77
Portuguese 28 25 67 64 * *
Spanish 36 37 87 85 96 98
Bosnian 15 14 44 42 * *
Croatian 26 21 60 59 88 88
Macedonian 10 8 39 35 * *
Serbian 26 17 67 62 96 91
Czech 52 47 96 96 * *
Hungarian 47 36 89 88 * *
Polish 34 34 84 80 95 94
Russian 28 43 74 76 97 94
Ukrainian 44 46 79 75 * *
Arab 19 10 40 39 * *
Armenian 21 15 48 47 * *
Assyrian 9 6 * * * *
Egyptian 24 14 66 58 * *
Afghan 8 4 * * * *
Iranian 19 12 * * * *
Iraqi 14 8 * * * *
Lebanese 11 8 31 21 68 58
Turkish 11 7 25 16 * *
Filipino 8 52 47 76 * *
Indonesian 24 53 58 64 * *
Khmer 10 16 * * * *
Thai 23 81 * * * *
Vietnamese 7 13 48 48 * *
Chinese 6 13 35 48 69 73
Japanese 18 63 * * * *
Korean 6 15 * * * *
Bengali 8 3 * * * *
Indian 11 11 56 58 * *
Sinhalese 14 13 95 86 * *
Pakistani 19 8 * * * *
Sudanese 8 6 * * * *
South African 30 34 92 97 * *
Maori 53 50 89 88 * *
New Zealander 70 69 97 96 * *
Samoan 26 22 * * * *
Tongan 29 25 * * * *
American 82 82 99 99 * *
Chilean 30 34 79 73 * *

Source: 2006 census customised table

Notes: (a) based on sole ancestry response
* less than 100 persons

As expected, the intermarriage rate increases for both men and women from the first to the second generation and from the second to the third or more generations. These patterns point to increasing social interaction between the second and third or more generations of these ethnicities and people outside their ethnic group. Similar patterns were observed in the analysis of 2001 census data. (11) The increase is quite large for some ancestry groups, for example, people of Greek, Lebanese and Chinese ancestries. By the third generation, two-thirds of men and women of these ancestries have partnered outside their ethnic group. Significant increases in intermarriage are also observed from the first to the second generation for other Asian, Middle Eastern and Southern European ancestry groups that do not yet have many people in the third generation who are of marriageable age at this time. The third or more generation of Western European ancestries have very high intermarriage rates of over 90 per cent. The low intermarriage rate of the third or more generation of English ancestry is a notable exception, reflecting the large number of third or more generation Australians who are of English ancestry.

These findings confirm that there appear to be few barriers to social integration in Australia, not just for immigrants from Western Europe but also for those from Eastern and Southern Europe. In the case of those with Eastern European ancestries--including those of Polish, Russian and Serbian backgrounds--there is almost complete out-marriage by the third generation. The Greek case is worth highlighting. Almost all post-World War Two migrants from Greece arrived as couples or families with young children. As has been noted in many studies, the second generation of Greek ancestry have exhibited a relatively low propensity to marry out. (12) The 2006 census results show a similar pattern with only 37 percent of second generation males of Greek ancestry and 31 per cent of females married out. This outcome reflects a strong tendency for first generation Greek families to concentrate residentially and to develop ethnic specific social institutions, including the Greek Orthodox church. Yet despite this ethnic solidarity, by the third generation 67 per cent of men of Greek ancestry and 61 per cent of the women have married out.

The intermarriage rates by ancestry also show patterns by gender that are similar to those indicated in Table 1 by birthplace. Men and women of Western European ancestries have similar rates of intermarriage. There is also not much difference by gender among people of Pacific island ancestries and for men and women of Indian ancestry. However, men of Middle Eastern ancestries are more likely to intermarry than women of these ancestries, while the opposite pattern is observed for men and women of East and Southeast Asian ancestries.

A final indicator of the significance of intermarriage is the extent to which it involves partnering with persons of similarvor different ethnic and racial background. It is possible that marriage across ethnic lines may be confined to persons of similar ethnic origins, such as from other Southern or Eastern European countries in the case of persons of Greek or Italian backgrounds. Alternatively, if intermarriage is predominantly with persons who claim Australian or Anglo-Celtic ancestries it implies a higher degree of social integration into Australian society, which is composed predominantly of persons of English-speaking background. Table 4 provides the information necessary to explore this issue.

Table 4: Second-generation partnered men and women who have a spouse of
a different ancestry, (a) percentage distribution by spouse's ancestry,
2006

 Ancestry of spouse per cent

Ancestry of Australian/NZ/ Other Asian Middle Other *
Individual Anglo-celtic European Eastern

2nd Men 45 37 5 4 8
generation Women 40 42 3 5 10
Greek

Italian Men 64 25 4 2 5
 Women 59 29 2 3 7

Maltese Men 58 31 4 2 5
 Women 56 34 2 3 5

Croatian Men 57 32 4 1 7
 Women 50 39 2 2 8

Serbian Men 58 31 3 2 6
 Women 55 34 2 2 7

Hungarian Men 67 21 5 1 7
 Women 66 27 1 1 6

Polish Men 71 20 3 1 2
 Women 67 26 1 1 6

Russian Men 66 24 3 0 7
 Women 61 28 0 2 8

Lebanese Men 44 36 5 5 11
 Women 35 35 5 8 18

Turkish Men 32 44 6 6 12
 Women 27 35 7 16 15

Vietnamese Men 25 10 44 0 21
 Women 37 13 42 0 9

Chinese Men 61 11 22 1 5
 Women 67 17 8 1 8

Indian Men 66 19 7 3 6
 Women 61 18 11 1 9

Ancestry of Individual Total per cent intermarried

2nd generation Greek Men 100 37
 Women 100 31

Italian Men 100 51
 Women 100 42

Maltese Men 100 67
 Women 100 64

Croatian Men 100 60
 Women 100 59

Serbian Men 100 67
 Women 100 62

Hungarian Men 100 89
 Women 100 88

Polish Men 100 84
 Women 100 80

Russian Men 100 74
 Women 100 76

Lebanese Men 100 31
 Women 100 21

Turkish Men 100 25
 Women 100 16

Vietnamese Men 100 48
 Women 100 48

Chinese Men 100 35
 Women 100 48

Indian Men 100 56
 Women 100 58

Source: 2006 Census customised table

Notes: (a) based on sole ancestry response

* Other ancestries include American, African and Pacific islander
ancestries, other small European, Asian and Middle Eastern groups not
separately identified and 'not stated'.

Among the ancestry groups shown in Table 4, the second generation of Eastern European ancestries such as Polish and Hungarian shows the highest proportion (among those intermarried) with spouses who are of Australian or Anglo-Celtic ancestries. More than 60 per cent of the ancestries who have intermarried also have Australian or Anglo-Celtic partners. The proportion with Australian or Anglo-Celtic partners was lower among the second generation of Southern European ancestries who have intermarried, 50 to 60 per cent for those of Italian, Maltese, Serbian or Croatian ancestry, and less than 50 per cent for the second generation of Greek ancestry. The proportion was 30 to 40 per cent for the Lebanese second generation who are intermarried, and lowest for the Turkish and Vietnamese second generation at 25 to 35 per cent.

Over 40 per cent of the Vietnamese second generation who had partnered a person of a different ancestry had partnered a person of another Asian ancestry, showing a high preference for pan-Asian partnering. Second generation Chinese men also show a relatively high propensity to partner with other Asians, much more so than second generation Chinese women. The reverse pattern is observed for the second generation of Turkish and Lebanese ancestry, with the women more likely than the men to partner with persons of other Middle Eastern ancestry. As noted earlier, these gender differences are related to the attitudes to women and men and their roles in the family in these ethnic communities. (13)

The second generation of the three Asian ancestries shown is less likely to partner with people of Other European ancestries compared with the second generation of the two Middle Eastern ancestries shown. There appears to be less intermixing between the second generation of these Asian ancestries with people of non-English-speaking European ethnicities than between the second generation of Lebanese or Turkish backgrounds with people of the European ethnicities. This may reflect the differences and similarities in the period of migration to Australia of the first generation of these migrant communities.

When the ancestry of the spouses of the intermarried third or more generation is examined for the non-Western European ancestry groups that have an adult third generation of sufficient numbers for analysis, there is remarkable similarity among the ancestry groups in the distribution of their spouses by ethnic origin (table 5). Three-quarters of intermarried third generation men and women of Chinese origin and about 70 per cent of men and women of Lebanese ancestry had partners of Australian for Anglo-Celtic ancestry. There was much less pan-ethnic partnering in the third generation of people of Chinese or of Lebanese ancestry than was observed in the second generation in Table 4. Very few third generation men and women claiming Chinese ancestry had partners of other Asian ancestries; the overwhelming majority of those who had intermarried had spouses of Australian or European ancestry. Similarly, very few of the third generation of Lebanese ancestry who had intermarried had spouses of other Middle Eastern ancestry; the overwhelming majority had partnered with Australians of Anglo-Celtic or other European ancestries.

Table 5: Third generation partnered men and women who have a a spouse
of a different ancestry, (a) percentage distribution by spouse's
ancestry, 2006

 Ancestry of
 spouse

Ancestry of Australian/ Other * Asian Middle Other *
Individual NZ/Anglo-celtic European Eastern

3rd+ generation

Greek Men 69 22 3 2 5
 Women 62 29 1 3 5

Italian Men 78 15 5 1 4
 Women 75 17 1 1 5

Maltese Men 71 22 4 0 3
 Women 68 21 1 2 8

Croatian Men 61 23 6 2 8
 Women 71 23 2 0 5

Polish Men 67 25 2 1 6
 Women 69 21 2 1 7

Russian Men 68 24 4 0 4
 Women 69 22 2 0 8

Lebanese Men 71 20 1 0 8
 Women 69 26 3 0 3

Chinese Men 77 15 2 1 6
 Women 76 17 1 0 6

Ancestry of Individual Total per cent Number of
 intermarried intermarried people

3rd+ generation

Greek Men 100 67 844
 Women 100 61 825

Italian Men 100 77 3497
 Women 100 74 3517

Maltese Men 100 19 403
 Women 100 77 411

Croatian Men 100 88 189
 Women 100 88 183

Polish Men 100 95 593
 Women 100 94 588

Russian Men 100 97 169
 Women 100 94 186

Lebanese Men 100 68 270
 Women 100 58 227

Chinese Men 100 69 520
 Women 100 73 501

Source: 2006 Census customised table

Notes: (a) based on sole ancestry response

* Other ancestries include American, African and Pacific islander
ancestries, other small European, Asian and Middle Eastern groups not
separately identified and 'not stated'.

It appears that by the third generation of the ethnic groups shown in Table 5, the partnering patterns of those who intermarry are more a reflection of the ethnic composition of Australian society than of any preferences based on cultural heritage. According to this measure, a high level of social integration is achieved by the third generation of these ethnic groups.

EDUCATION AND INTERMARRIAGE

Does education lead to a greater propensity to partner outside the ethnic group? Table 6 examines the intermarriage rate by ancestry and level of education. While more educated men and women of some ancestries have higher intermarriage rates, there is no difference by education among men and women of other ancestry groups and in a few ancestry groups, men and women of lower education have higher intermarriage rates than do those who are better educated.

Table 6: Per cent intermarried by gender, education and ancestry,
all partnered men and women, 2006

 Males

Ancestry Degree Other quals Year 11-12 Year 10

Australian 26 20 22 16
Aboriginal 62 21 9 4
Maori 71 64 51 53
New Zealander 69 74 71 74
English 41 31 32 26
Irish 69 69 71 70
Scottish 78 73 78 77
Welsh 85 74 82 82
Dutch 15 71 75 75
German 73 66 80 70
Greek 35 32 30 11
Italian 54 43 45 22
Maltese 79 58 57 35
Spanish 55 43 43 37
Croatian 55 36 44 24
Macedonian 34 21 20 9
Serbian 37 32 32 32
Hungarian 68 51 63 55
Polish 46 45 54 58
Russian 30 41 47 42
Lebanese 27 20 19 11
Turkish 19 18 15 7
South African 30 28 43 48
Chinese 8 9 7 6
Filipino 5 9 13 12
Vietnamese 11 10 7 5
Indian 10 12 18 16
Sinhalese 14 15 24 22

 Females

Ancestry Degree Other quals Year 11-12 Year 10

Australian 27 25 24 20
Aboriginal 55 25 14 6
Maori 60 54 49 55
New Zealander 69 73 70 74
English 37 32 29 26
Irish 68 67 67 66
Scottish 74 72 69 71
Welsh 84 78 68 74
Dutch 77 70 69 68
German 75 65 76 66
Greek 33 28 26 9
Italian 54 39 38 16
Maltese 75 65 56 35
Spanish 59 51 47 34
Croatian 54 39 43 18
Macedonian 34 23 19 7
Serbian 36 24 28 13
Hungarian 67 49 53 42
Polish 49 40 50 58
Russian 44 48 53 56
Lebanese 27 19 12 8
Turkish 21 16 10 4
South African 35 35 38 34
Chinese 19 15 12 9
Filipino 38 52 60 79
Vietnamese 24 13 12 8
Indian 9 16 17 19
Sinhalese 18 12 16 17

Source: 2006 census customised table

People stating Australian, English, Southern European and Middle Eastern ancestries show an increase in intermarriage rates with educational attainment. In contrast, no difference is observed in the intermarriage rate by education for men and women reporting German, Polish, Russian, South African, New Zealander or Sinhalese ancestries. Among men and women of Filipino or Indian ancestry, those who have no post-school qualifications are more likely to intermarry than are those who are better educated. In some groups, such as the Chinese, the effect of education seems to vary by gender, with men showing no difference in intermarriage by education--low rates of intermarriage at all levels of education--but women showing a positive correlation between level of education and inter-ethnic marriage. The effect of education on inter-ethnic marriage appears to be mixed.

Some of the effect of education is also related to age and generation. The younger age cohorts in Australia are better educated than the older age cohorts, due to higher school retention rates and increasing proportions of the younger cohorts going to universities and undertaking other tertiary education and training in recent years than in the past. Since the second generation is also of a younger age group than the first generation, they are also generally better educated than their parents' generation. A disaggregation by generation may be necessary when examining the relation between education and intermarriage for some ancestry groups.

Table 7 shows the intermarriage rate by education and generation for ancestry groups that have younger and better-educated second and third generations. This takes account of the possible effect of the interaction between education and generation on the patterns shown in Table 6. These figures show that, while level of education may be correlated with intermarriage in the first generation, that correlation is no longer observed in the second or third generation for some groups such as Greeks, Italians and Croatians. No relation between education and intermarriage is observed for men of the second generation of Lebanese, Turkish or Macedonian ancestry; but the more educated among the women of these ancestries do have higher rates of intermarriage than do the less educated. It would appear that education has a modest effect in broadening the choice of marriage partners across ethnic boundaries for women of these ancestries.

Table 7: Percent intermarried by gender, education, generation and
ancestry, all partnered men and women, 2006

 Males Female

Ancestry Degree Other Year Year Degree Other Year Year
 quals 11-12 10 quals 11-12 10

English
1st generation 49 39 40 41 42 38 32 38
2nd generation 55 50 51 46 54 51 51 44
3rd+ generation 29 21 21 16 27 23 22 18

Greek
1st generation 33 22 19 7 32 19 16 5
2nd generation 34 37 39 43 32 31 30 31
3rd+ generation 57 75 67 69 59 61 57 76

Italian
1st generation 51 31 35 13 51 25 25 7
2nd generation 53 50 49 53 52 42 40 36
3rd+ generation 80 79 74 79 81 72 73 76

Maltese
1st generation 65 46 41 27 68 52 39 24
2nd generation 85 66 73 58 76 69 66 55

Croatian
1st generation 45 28 34 18 45 28 29 13
2nd generation 62 56 62 74 59 55 61 62

Macedonian
1st generation 24 14 12 6 23 14 10 5
2nd generation 47 38 37 41 43 36 33 31

Lebanese
1st generation 21 14 13 8 20 13 8 6
2nd generation 33 30 34 31 31 23 18 18

Turkish
1st generation 19 16 13 6 22 14 8 4
2nd generation 20 28 24 24 17 18 16 14

Chinese
1st generation 7 7 6 5 17 14 11 8
2nd generation 34 37 42 28 50 52 50 38
3rd+ generation 52 76 60 83 71 75 72 77

Source: 2006 census customised table

The effect of education on intermarriage is also mixed for men and women of different generations of Chinese ancestry. Education appears to increase intermarriage for women but not for men in the first two generations. It has no effect on intermarriage for women in the third generation; however better educated men in the third generation are more likely to marry within the ethnic group than are less educated men. In contrast to the above patterns, for two of the ancestry groups shown in Table 7, the English and Maltese, there is a clear pattern of increasing intermarriage with education within each of the three generations of men and women.

The relationship between level of education and intermarriage is different for different ethnic groups and for men and women of some ethnicities. The variations in patterns as described above suggest the complexity of the relationships between education, ethnicity and marriage/partnering that may be grounded in cultural and generational differences in male and female roles and status in the family.

CONCLUSION

If inter-ethnic partnering is a key indicator of social integration, as suggested by sociologists, then the increase in intermarriage from the first to the second generation, and from the second to the third generation, of Australians of various ethnic backgrounds indicates that social integration is proceeding with each successive generation. It is particularly noteworthy that, by the third generation, the majority of Australians of non-English-speaking background have partnered with persons of different ethnic origins, and that of these, the majority had partnered with persons of Australian or Anglo-Celtic ancestry. These partnering patterns suggest that, while Australian multiculturalism may have encouraged the intergenerational maintenance of ethnic identity, it has not inhibited increased social interaction outside the ethnic group with each successive generation.

For ethnic communities of more recent migrant origin from South and East Asia, the Middle East and Africa, the second generation is still young and not yet of marriageable age and there is no third generation yet. These communities bring with them cultural traditions that are quite different from those people of Southern and Eastern European backgrounds who dominated migration during the 1950s and 1960s. Whether the social integration of the second and third generations of these groups as measured by intermarriage will be similar to those of the non-English-speaking European migrant communities will not be known for several years.

Acknowledgments

Census data discussed in this paper were provided by the Australian Bureau of Statistics through its 2006 Australian Census Analytic Program. We thank the ABS staff for their assistance with the customised datatables.

References

(1) R. Alba, 'Assimilation's quiet tide', The Public Interest, vol. 119, no. 1, 1995, pp 3-18; F. Bean and G. Stevens, America's Newcomers and the Dynamics of Diversity, Russell Sage Foundation, New York, 2003

(2) C.A. Price and J. Zubrzycki, 'The use of intermarriage statistics as an index of assimilation', Population Studies, vol. 16, no. 1, 1962, pp. 58-69; F.L. Jones, 'Ethnic marriage preferences: the classic melting-pot data revisited', Australian Journal of Social Research, vol. 1, no. 1, 1995, pp. 71-85

(3) C.A. Price, The Fertility and Marriage Patterns of Australia's Ethnic Groups, Department of Demography, ANU, Canberra, 1982, p. 100

(4) F.L. Jones and R. Luijkx, 'Post-war patterns of intermarriage in Australia; the Mediterranean experience', European Sociological Review, vol. 12, no. 1, 1996, pp. 67-86

(5) D. Giorgas and F.L. Jones, 'Intermarriage patterns and social cohesion among first, second and later generation Australians', Journal of Population Research, vol. 19, no. 1, 2002, pp. 47-64

(6) G. Heard, S.E. Khoo and B. Birrell, Intermarriage in Australia: Country of Birth, Ancestry, Religion and Indigenous Status, Centre for Population and Urban Research, Monash University, Melbourne, 2009, forthcoming

(7) Australian Bureau of Statistics, Household Guide: How to Complete Your Census Form, Canberra, 2006, p. 7

(8) J. Penny and S.E. Khoo, Intermarriage: A Study of Migration and Integration, Australian Government Publishing Service, Canberra, 1996

(9) Department of Immigration and Citizenship, Population Flows: Immigration Aspects, 2006-07 edition, Canberra, 2008; Department of Immigration and Multicultural Affairs, Population Flows: Immigration Aspects, Canberra, various years

(10) S.E. Khoo, 'The context of spouse migration to Australia', International Migration, vol. 39, no. 1, 2001, pp. 111-132

(11) S.E. Khoo, 'Intermarriage in Australia: patterns by ancestry, gender and generation', People and Place, vol. 12, no. 2, 2004, pp. 35-44

(12) C.A. Price, 'Ethnic intermixture in Australia', People and Place, vol. 1, no. 1, 1993, pp. 6-8

(13) Penny and Khoo, 1996, op. cit.