THE FERTILITY OF IMMIGRANT WOMEN IN AUSTRALIA

Mohammad Jalal Abbasi-Shavazi

The 1991 census did not ask women how many children they had ever had. Because of this, there has been a gap in our understanding of fertility trends. The author uses the `own-children' method which allows him to fill this gap. He analyses the fertility of immigrant women in Australia and discovers that, by 1991, most had lower fertility than Australian-born women. Most also had lower fertility than women of a comparable age in their countries of origin. Second-generation `migrants' had converged even more closely to the Australian norm.

INTRODUCTION

A commonly held view is that immigrants usually have higher fertility than the native-born. This view emanates from the United States where Mexican American immigrants have higher fertility than the native-born.¹ This paper examines the fertility patterns of immigrant groups in Australia during the period, 1977-1991. To measure fertility levels and patterns, the paper uses the own-children method applied to the 1991 Australian Census. First, total fertility rates and age-specific fertility rates for first-generation immigrants are compared to those of Australia- born women and also to those of the populations in the countries of origin. Then, the fertility of women from six selected immigrant groups, British-Irish, Dutch, Greeks, Italians, Poles, and Lebanese, are analysed according to their age at arrival in Australia, and generation.

STUDIES OF IMMIGRANT FERTILITY IN AUSTRALIA 

Several studies have identified marked differences in fertility between immigrants and native-born women in Australia, as well as among immigrant groups themselves.² For example, Day, using the 1961 Census, found very high fertility for Dutch Catholic immigrants in Australia. Higher fertility for Southern European immigrants during the 1970s has also been observed by previous studies. The status of having high fertility then moved to Middle Eastern immigrant women. Immigrants of English-speaking background have always displayed similar fertility patterns to those of the native- born women in Australia.

The question on the number of children ever born was dropped from the 1991 Census. Owing to this, little attention has been paid to immigrant fertility patterns and differentials in recent years. No recent study has been undertaken in Australia to analyse the fertility of the second generation of immigrant groups, who constitute 3.1 million of Australia's population, or to examine the effects of the ages of immigrants at time of arrival on their fertility in Australia. However, analysis of this kind can be conducted using the own-children method.

DATA AND METHOD

In the present study the 1991 Census was the main source of data on the recent fertility of immigrant groups by generation and age at arrival in Australia, whereas the Household Sample Files of the 1986 and 1991 Censuses made it possible to analyse differential fertility for the selected birthplace groups. The fertility rates for immigrant groups in Australia were compared to the rates in country of origin.³

By application of the own-children method, the migrants' current fertility measures for different immigrant groups during the period 1977-91 were calculated. The own-children method of fertility estimation is a reverse-survival technique for estimating age-specific birth rates for years previous to a census or household survey. From the basic household records, enumerated children are first matched to mothers within households. The matched children are then reverse-survived to estimate numbers of births by age of mother in previous years. Reverse-survival is similarly used to estimate numbers of women by age in previous years. After adjustments are made for mis-enumeration and unmatched children, age-specific birth rates are calculated by dividing the number of reverse-survived births by the number of reverse-survived women.⁴ A detailed assessment of the results of the own- children method showed that the application of the method in this study has produced a reliable result.⁵

FERTILITY LEVELS AND PATTERNS OF MAJOR BIRTHPLACE GROUPS

Total fertility rates

Table 1 shows that there were marked variations in total fertility rates among migrant women in Australia over the periods 1977-81, 1982-86 and 1987-91. Additionally, a downward trend was visible in the levels of fertility of most birthplace groups across the periods. Lebanon-born women stood out from other migrant women with 4.2 children per woman in the first period and 3.7 and 3.4 in the latter two periods. Turkey-born women were the second most fertile birthplace group. The fertility of those from the southern European countries of Greece, Italy, Cyprus and Yugoslavia, which was consistently reported as high in the 1970s, had declined to below the Australia-born level by 1987-91.

A comparison between all immigrant groups and the Australia-born revealed six fertility categories. The first category consisted of the Lebanese who displayed the highest total fertility rates during 1977-81, 1982-86 and 1987-91. The second category included women born in Cyprus, Malta, Turkey and Vietnam. These women had higher fertility than native-born women in the period 1977-81 but, despite a declining trend, still had higher fertility than native-born women during 1987-91. Category three consisted of women from the southern European countries of Greece, Italy and Yugoslavia. They had higher fertility than Australia-born women in 1977-81, but their fertility had declined to levels below that of Australia-born women by 1987-91. Birthplace groups included in Category four were those which had lower total fertility rates than Australia- born women during all three periods, but which showed relatively similar trends to the Australia-born across each period. English-speaking-background women, including those born in the United Kingdom and Ireland, New Zealand, North America, and India and Sri Lanka, and also women from Germany and Poland were classified in this group. The fifth category comprised women from China and Taiwan Province, South Africa, Malaysia, the Netherlands and Hong Kong. They displayed similar or lower total fertility rates compared to Australia- born women during 1977-81, and their fertility declined appreciably during the second and third five-year periods. The sixth and final category includes Egypt- born and Philippines-born women who showed irregular fertility patterns over the three periods.

Downward trends were clearly visible in the fertility of all birthplace groups except the Philippines-born, across the three periods. The fertility of most immigrant groups has converged to, or fallen below, that of the host society. Therefore, with time, only a small fraction of migrant women are likely to have higher fertility than Australia-born women.

Fertility rates compared with those in the countries of origin

Table 1 also provides a comparison between the fertility of immigrants in Australia and the fertility of women in the countries of origin. This reveals several patterns. Some birthplace groups displays distinctively lower fertility in Australia relative to the origin populations. Most migrants from non-English-speaking Asian countries, such as those from Turkey, India and Sri Lanka, China, Malaysia, the Philippines and Vietnam, and also, the Egypt-born, displayed considerably lower fertility in Australia during the period 1977-91 than did the populations in their home countries. Women from Cyprus, New Zealand, Poland, China, South Africa and Yugoslavia recorded slightly lower fertility in Australia than prevailed in their home countries. On the other hand, a few birthplace groups such as Italy-born women exhibited higher fertility in Australia than was characteristic of the populations in their home countries. Maltese and Netherlands-born women displayed higher total fertility rates in Australia than did women in their countries of origin during the first two periods, but migrant and origin-country fertility levels were the same during 1987-91. Migrant women from Lebanon, Greece, Germany, the United Kingdom and Ireland, and Hong Kong were among those who recorded very similar total fertility rates to the populations in their home countries.

As Young found,⁶ differences from countries of origin and convergence to the fertility of the host population appeared to be greatest among women from high-fertility countries. Most migrants from Asian countries, such as Malaysia, Turkey, Vietnam, and the Philippines, have in recent times exhibited considerably lower fertility in Australia than has prevailed in their home countries; this is likely to be, at least in part, an indication of the operation of selectivity factors. Disruption effects might also have contributed. Lower fertility in Australia could additionally be related to the fact that effective contraception is widely available, use of which has been a part of immigrant groups adapting to Australian conditions. Then again for some migrants, having come to Australia as children and grown up in Australia will have contributed to lower fertility in the adopted home. And a concentration of migrants, particularly Asian migrants, in metropolitan cities,⁷ could have been another factor reducing their fertility in Australia.

Proportion ever married among the major birthplace groups

One approach to explaining fertility differentials is to look at census proportions of women ever married in age groups 15-19 to 30-34. Groups with early marriage patterns will have larger proportions ever married at younger ages, and their fertility, correspondingly, is likely to be higher at these ages. On the other hand, low proportions ever married in younger age groups indicate a pattern of delayed marriage and childbearing. Age-specific proportions ever married varied among different birthplace groups. Except for the Poland-born, percentages ever married declined between the 1981 and the 1991 Censuses for migrant women in most age groups. Early marriage patterns were found for migrant groups such as the Lebanese and Turks whose high fertility was described earlier. On the other hand, women from Hong Kong, Malaysia, China and New Zealand who exhibited delayed childbearing also had delayed marriage patterns. The marriage patterns for India- and South Africa-born women were midway between those for groups which married early and those which displayed similar marriage patterns to Australia-born women. Women from the United Kingdom and Ireland, the Netherlands and Germany had more-or-less similar marriage patterns to Australia-born women at both censuses. These groups were among those that had displayed similar fertility patterns to Australia-born women.

SELECTED MIGRANT GROUPS FOR FURTHER INVESTIGATION

A more intensive analysis was conducted for six selected groups: the British-Irish, Dutch, Greeks, Italians, Lebanese and Poles. Table 2 presents data on these groups. It allows comparisons to be made for those who arrived in Australia below the age of 15 years and at ages 15 years and over, and for second generation of immigrant groups. The six birthplace groups included in this analysis were selected on the basis of four criteria. First, selected birthplace groups needed to represent the different patterns of fertility and fertility change observed above in comparing the fertility of immigrant women with that of the Australia-born. Second, they needed to be relatively large birthplace groups in Australia, so that analysis of fertility in terms of different ages at arrival would be possible. Third, only birthplace groups which had been in Australia for relatively long periods of time were selected, so that the fertility of the Australia-born second generation could be examined. Fourth, since immigrants from some birthplace groups are of mixed origin, only groups representing relatively `pure' communities should be selected. To maintain this `purity', both first and second generation immigrants were restricted to those who had both their parents born in the particular country.

Selected birthplace groups: fertility rates by age at arrival and generation

Total fertility rates are presented for the six selected birthplace groups in three five-year periods in Table 2. The fertility of the second-generation Lebanese was considerably lower than that of the first generation. Furthermore, Lebanese migrants who came to Australia as children displayed a similar pattern to the second generation. The remarkable fertility decline among the Lebanese, who have had very high fertility in Australia, may indicate the adaptation of this group to the Australian society. Second-generation `migrants' from the United Kingdom and Ireland showed a convergence to the pattern of native-born women by displaying slightly higher fertility than the first generation. The fertility of the first generation may have been disrupted by migration. However, the second generation of the United Kingdom-born, too, appears to have converged to the fertility of their adopted home. For the Poles, fertility is low for both first and second-generation immigrants; it is lower than the levels applying for the Poles in Poland and for the Australia-born. This suggests that emigration to Australia has been selective of Polish women who have a low propensity to have children.

Immigrants from the Netherlands show a very unusual pattern, with fertility being low in the Netherlands itself, higher among immigrants arriving as adults but then being even higher for those arriving as children and for the second generation. What would explain the high fertility of the second generation Dutch? One factor may be that the second generation have lower levels of education than their Dutch parents.⁸ In addition, in contrast to most overseas-born communities, both the Netherlands-born and their children show a high proportion living outside the major cities.⁹

Both first and second-generation Greeks in Australia, surprisingly, have displayed similar fertility patterns to Greeks in Greece, perhaps indicating a continuity of cultural influence of the country of origin in Greek communities in Australia. Younger arrivals display very low fertility compared with older arrivals. Little difference was apparent in the fertility levels of those born in Australia and compared with those who arrived as children and grew up in Australia. It could be argued that, when migrants are a large group in the place of destination, second-generation `migrants' are likely to be communicating with their cousins and other relatives who arrived in Australia at their own age, as well as their counterparts in the host society. Thus they are, on the one hand, able to maintain aspects of the culture and behaviour of their parents' generation; and, on the other hand, adapt to the host society. McDonald argued that gender equity is low in Southern European countries.¹⁰ These values may be maintained among southern European families in Australia.

Overall, the fertility patterns of the first-generation immigrants and the second-generation `immigrants' shifted towards that of native-born women across generations. Furthermore, being born in Australia of migrant parents or arriving at an early age and growing up in Australia had almost the same effect on the fertility of immigrants, particularly those from Italy and Greece.

MULTIVARIATE ANALYSIS

Variations in fertility were found between the six birthplace groups. But the following questions remain. Are these differences in fertility simply the consequence of inter-group variations in the timing of marriage, with groups that married earlier bearing more children than those that married later? Do differences between immigrants and natives, and among the immigrant groups, persist even when comparison is focused on women in the same circumstances? To what extent can the differences in fertility be explained by the different cultural values and norms of these birthplace groups? Are there indications of cultural maintenance among birthplace groups in Australia?

The purpose of this section is to address these questions by examining the recent fertility of the selected birthplace groups. Fertility differentials are examined for two categories of variables: demographic, such as age and marital status; and social, such as age left school, citizenship, religion and English proficiency.

The main data sources used are the one per cent Samples Files of the 1986 and 1991 Australian Censuses. In order to achieve the sample sizes necessary for reasonably precise estimation, the two one per cent Sample Files were combined. The combined data yield essentially an average of fertility in 1983-86 and 1988-91. The limitation of this data set is that it does not allow one to assess the fertility trends observed between the six birthplace groups. The indicator of recent fertility used is whether a woman has an own child/children under three years of age. A logistic regression model predicted individual probabilities of having a child under three for selected immigrant and native-born women aged 15 to 49, taking selected socio-economic variables into account.

Fertility differentials among the selected immigrant groups

Not unexpectedly, age and marital status were by far the most important variables explaining the fertility of the selected birthplace groups. Variation in marriage patterns was an important factor explaining the fertility of immigrants and native-born in Australia. Lebanese Muslim women who were married early were much more likely than native women and other birthplace groups to have a child under three, net of other influences. A similar age pattern of fertility was found for Muslims and Christians, although Muslims had higher fertility in all age groups. The most important finding of this analysis was that variations in fertility among the selected birthplaces were not significant, when controlled for other socio-demographic characteristics, the exception being the Lebanese.

What are the factors behind the high fertility levels of Lebanese-origin women in Australia? Islamic beliefs about family formation and the value of children may be the important factors. High in-marriage, the universality of marriage, and low levels of education and low labour-force participation could be the other main reasons for this pattern.

Some evidence of disruption was found among women recently arrived (0-4 years) in Australia. As expected, lower fertility was found for those who had been in Australia for a short period of time, 0-4 years, than for those who had been in Australia for five years or more. This is consistent with the disruption hypothesis. Some migrants may have postponed their family formation and childbearing after their arrival in Australia, resulting in lower fertility for the first five years of residence in Australia. The catch-up effect was evidenced for those who had been in Australia for 5-9 years.

Overall, there was no significant difference between the fertility of the second generations of immigrant groups, and that of native-born women, taking selected socio-demographic characteristics into account. It can be argued that the fertility of the second generation of immigrants has converged to that of mainstream Australians. Second generation `migrants' have mostly experienced similar social institutions and economic conditions to the Australian population, and therefore are expected to have similar fertility to Australian women. In sum, differences in the fertility patterns of immigrants in Australia are likely to diminish as their socio-economic and demographic characteristics converge to those of the Australian population.¹¹

CONCLUSIONS

The paper examines fertility levels and trends and among immigrants during 1977-91. These trends are compared with trends in the same period in the countries of origin and for the Australian population as a whole. The main conclusion is that, despite recent emphasise on multiculturalism, there is strong evidence of adaptation of immigrants to the patterns of fertility of the Australian population as a whole. The Lebanese in Australia, for example, provide very strong evidence of adaptation. This suggests that the political economy facing immigrants in Australia may have been a more influential determinant of fertility than the cultures that the immigrants brought with them.¹²

The findings of the present study suggest that the situation in Australia is different from that in the United States. Immigrants in Australia have usually had lower fertility than Australian women, and therefore, the effect of their fertility on the population growth rate is very small. As Price¹³ concluded,

The high hopes Arthur Calwell had when launching the post-war immigration program in 1945 -- hopes of importing new people whose high fertility values would help raise the Australian population to twenty million by 1970 - have not been realised.

Given the low fertility of birthplace groups in Australia, it can be concluded that immigration does not compensate for low Australian fertility.

Acknowledgment
This paper is based on the author's PhD thesis, Fertility patterns of selected Australian immigrant groups, 1977-91. Invaluable comments made by Peter McDonald, Gordon Carmichael, Chris Wilson and two referees from People and Place are gratefully acknowledged.

References
¹ J. C. Abma and L. J. Krivo, `The ethnic context of Mexican American fertility', Sociological Perspectives, vol. 34, no. 2, 1991, pp. 145-164; N. A. Fischer and J. P. Marcum, `Ethnic integration, socioeconomic status and fertility among Mexican Americans', Social Science Quarterly, vol. 65, no. 2, 1984, pp. 583-593; A. M. Sorenson, `The fertility and language characteristics of Mexican-American and non-Hispanic husbands and wives', Sociological Quarterly, vol. 29, no. 1, 1988, pp. 111-130.
<sup>2</sup> See W. D. Borrie, Population Trends and Policies: A Study of Australian and Worlds Demography. Australian Publishing Company, Sydney, 1948; L. H. Day, `Family size and fertility', in A. F. Davies and S. Encel (eds), Australian Society: Sociological Introduction, Cheshire, Melbourne, 1965; L. H. Day, Differential fertility in Australia, International Population Conference, London, Liege, International Union for the Scientific Study of Population, 1969; H. Ware, `Immigrant fertility: behaviour and attitudes', International Migration Review, vol. 9, no. 3, 1975, pp. 361-378; F. Yusuf and G. Eckstein, `Fertility of migrant women in Australia', Journal of Biosocial Science, vol. 12, no. 1980, pp. 179-190; F. Yusuf and I. Rockett, `Immigrant fertility: patterns and differentials in Australia, 1971-76', Population Studies, vol. 35 no. 3, 1981, pp. 413-424; C. A. Price, The Fertility and Marriage Patterns of Australia's Ethnic Groups, Department of Demography, Australian National University, Canberra, 1982; L. H. Day, Analysing Population Trends: Differential Fertility in a Pluralistic Society, St Martins Press, New York, 1983; C. M. Young, `Changes in the demographic behaviour of migrants in Australia and the transition between generations', Population Studies, vol. 45, no. 1, 1991, pp. 67-89; S. E. Khoo and J. Shu, Immigrant Family Formation Patterns in Australia, Department of Immigration and Multicultural Affairs, Canberra, 1996; F. Yusuf and S. Siedlecky `Family formation patterns among migrant women in Sydney', Journal of Biosocial Science, vol. 28, no. 1996, pp. 89-99.
<sup>3</sup> Country of origin data were taken from the following sources: for Egypt, B. Bucht and M. A. El-Badry Reflections on Recent Levels and Trends of Fertility and Mortality in Egypt, Cairo Demographic Centre, Cairo, 1984; H. Rashad , Analysis of Recent Fertility Trends in Egypt: The Life Table Approach, Cairo Demographic Centre, Cairo, 1987; for China, X. Wen, Current and desired fertility: reflections on fertility decline in China, PhD Thesis, Australian National University, Canberra, 1993; X. Peng, Demographic Transition in China, Fertility Trends since the 1950s, Clarendon Press, Oxford, 1991; China State Statistical Bureau, China Population Statistics Yearbook 1990, Science and Technology Literature Press, 1991; for the following countries, United Nations, Demographic Yearbooks. New York, 1978-93: Canada, Poland, USA, Netherlands, Malta, Greece (for 1977-1984), Hong Kong, Malaysia (except for 1985), Italy (except for 1983, 1985-1987), Germany (except for 1982), Yugoslavia (except for 1986 and 1991), Cyprus (except for 1986-91), New Zealand. Figures for the following countries (specified years) were taken from the Council of Europe, Recent Demographic Development in Europe and North America 1992, Council of Europe Press, Strasbourg, 1993: Greece, 1985; Italy, 1983, 1985, 1986 and 1987; Germany, 1982; Turkey, 1975 and 1980-91; Cyprus; 1986; UK and Ireland, 1975 and 1980-91. Figures for Yugoslavia in 1991 are from M. Rasevic and M. Petrovic, `Fertility of the population of the FR of Yugoslavia', Yugoslav Survey, vol. 34, no. 3, 1993, pp. 3-38. Figures for South Africa were from Bureau for Information, Official Yearbook of Republic of South Africa: 1989-90, Cape Town, 1990, and W. P. Mostert and J. M. Lotter (eds), South Africa's Demographic Future, Human Science Research Council, Pretoria, 1990. Data for Vietnam are from General Census Steering Committee, Vietnam Population Census 1989 Sample Result, Hanoi, 1990 and J. Banister, Vietnam Population Dynamics and Prospects, University of California, Berkeley, 1993. Total fertility rates for Lebanon are from United Nations, World Demographic Estimates and Projections, 1950-2025, New York, 1988.
⁴ L. J. Cho et al., The Own-Children Method of Fertility Estimation, University of Hawaii Press, Honolulu, 1986, pp. 1-2.
⁵ M. J. Abbasi-Shavazi, `An assessment of the own-children method of estimating fertility by birthplace in Australia', Journal of the Australian Population Association, vol. 14, no. 2, 1997, pp. 167-185.
<sup>6</sup> Young, op. cit., p. 74
<sup>7</sup> B. Birrell, `Overseas migration and metropolitan population', People and Place, vol. 1, no. 2, 1993, pp. 50-52.
⁸ The Bureau of Immigration and Population Research revealed that among Australia-born persons with one or both parents born in the Netherlands, 36 per cent had attained some form of educational or occupational qualification. This was slightly below the figure for the total Australian population (39 per cent) and considerably lower than that for the Netherlands-born (48 per cent). See Bureau of Immigration and Population Research, Community Profiles 1991 Census: Netherlands Born, Australian Government Publishing Service, Canberra, 1994.
⁹ In 1991, 28 per cent lived in a location other than one of the major cities, compared with only 9.5 per cent of all overseas-born from non-English speaking countries. Similar geographical distribution for the second-generation Dutch has been reported by the Bureau of Immigration and Population Research, op. cit.
¹⁰ P. McDonald, Gender equity, social institutions and the future of fertility, Working Paper in Demography, Australian National University, Canberra, 1997.
¹¹ For further information on the regression results see M. J. Abbasi-Shavazi, Fertility Patterns of Selected Australian Immigrant Groups, 1977-91. PhD Thesis, Demography Program, Australian National University, Canberra, 1998.
¹² M. J. Abbasi-Shavazi and P. McDonald, Fertility and multiculturalism: immigrant fertility in Australia, 1977-1991, Paper presented at the IUSSP Conference in Beijing, 11-17 October, 1997.
¹³ Price, op.cit. p. 108

Back to People and Place Home Page