Journal of the Australian Population Association
Vol.8, No.2, 1989
COMMENT
EX-NUPTIAL BIRTHS Siew-Ean Khoo and Peter McDonald (JAPA 5(2), November 1988) raise some interesting questions on this social phenomenon but their interpretation of the slender evidence is not wholly convincing. i. In their zeal to demonstrate the recent growth in ex-nuptial births they provide a "Darrell Huff" type diagram with broken scaling that at first sight suggests that ex-nuptials are about to exceed nuptial births. But when it comes to the age distribution of notmarried women, the picture presented in Table i (and afortiori in figure 2 of the quoted article by Choi and Ruzicka) is quite different: a general fall in ex-nuptial confinement rates since 1971, with a recent small partial recovery only in the 20-24 age group. 2. In seeking an explanation for these uncertain trends, the authors quote two factors. The first, and perhaps less important one, is the alleged greater availability of abortion since the early 1970s which is said to have caused the drop in the rates after 1969, but no evidence of a rise in abortions or a change in inter-state pattern is given. There is also a quaint reference to "the conventional view that ex-nuptial births result from brief encounters or arrangements by women to have children in the single state, .. which was reasonable accurate ten years ago". Have the views and intentions changed since? 3. The second, and more significant claim of the article is the (positive) relation between socio-economic status and ex-nuptiality. This is backed, in Table 4, by AIFS data which show that de facto couples with children have lower educational and occupational and income and housing levels than married couples with children. The statistical validity of this analysis is dubious. The table refers to 130 de facto couples in the survey of whom 28 had children, as against the comparison series of 933 married couples with children, so that any error or ambiguity in a single case shifts the percentage by four points. If one looks at the detail, the contrast between educational standards is as striking as the similarities in female labour force participation, and in the mean and median values of family income but the sample, even if it were adjusted for age, number of children and so on, looks too small for analysis. 4. The authors give brief references to the "culture of poverty" which make unmarried cohabitation and illegitimate births common in America. But this does not explain the up-down-up trend in Australia, nor do suggestions that uncertain economic conditions produce changing marriage patterns among the poor or the rich. Chsnging social convenience, habits, acceptance (to which the authors also allude) might be more important than perceived (or guessed) economic trends. 5. One of the features of the mother's country of birth (Table 6) is that out of the 22 countries listed, the percentage of ex-nuptial
164
births for Vietnam, at 15 per cent, is surpassed only by Australia (16 per cent) and New Zealand (24 per cent) and that the percentage of ex-nuptial births with paternity acknowledged for Vietnam, at 36 per cent, is lower than for any other country shown. The authors suggest that this may be due to differing age and marital status compositions among birthplace groups, and may at least partially reflect high concentrations of unmarried women younger than 25 in Vietnam (and New Zealand) groups. But information from the ABS shows that for mothers giving birth, the age groups 15-19 and 20-24 are somewhat lower for Vietnam than for all countries although they, of course, have the highest ex-nuptial rates, and standardizing for age composition would slightly raise rather than lower the ex-nuptiality rate for Vietnam. Incidentally, at least for women aged 15-39 in 1986, the rates for never married to all women and of ex-nuptial births to unmarried women of those ages were about the same for Vietnam as for all birthplaces. The information from Table 6 can be extended as follows: BIRTHS REGISTERED IN AUSTRALIA IN 1984 Ex-Nuptial Births with Paternity not acknowacknowledged ledged Total
Nuptial
Total
Mother born in Vietnam
No. %
113 5
203 10
316 15
1,792 85
2,108 100
All Births
No. %
23,200 i0
11,429 5
34,629 15
199,406 85
234,034 i00
Taking ex-nuptial births for which paternity is acknowledged, the rate of 5 per cent for Vietnam is well below the Australian and Northern European countries and close to that for Southern European and Asian countries. It is only the figure for paternity not acknowledged that remains relatively high compared with the few countries for which it is quoted in Table 6, and that demographic gremlin might be due more to a communications problem with filling in forms than to different cultural mores. As W. Quesera used to say "It is nonsense to look for sense in statistics that make no sense." 6. The article refers to ex-nuptial births without reference to first and subsequent ones. The effect of de facto relationships on family size was perhaps beyond its scope, but it remains a subject that seems well worth the attention of demographers.
Robert V. Horn Sydney
165
REPLY TO R.V. HORN Horn's comments reflect mainly his misunderstanding and misinterpretation of some of the data, analyses and arguments presented in our paper. We shall therefore attempt to clarify the points he has raised. With regard to his first point, it was not our "zeal to demonstrate the recent growth in ex-nuptial births" that led Figure 1 to be drawn with a break in the scale. We do not believe that zeal has a role in the way research is presented. If the figure had been drawn with an unbroken and a much reduced scale, there would be a large empty space in the middle, and the declining and rising trends in the rates would not be contrasted as clearly. We had assumed that readers usually read the scale when looking at any graph and would see immediately that there was a break in it. Horn appears to have misinterpreted Table l in our article when he says that "when it comes to the age distribution of not-married women, the picture presented in Table 1 ...is quite different..." The table did not show the age distribution of not-married women. It showed the ex-nuptial confinement rates of not-married women. The rates in Table 1 in no way refute the trend in the ex-nuptial component of the TFR in Figure 1 because the denominators are different. Taken together, the two sets of rates indicate that the rise in the ex-nuptial component of the TFR is related mainly to a large increase in the number of not-married women and not so much to an increase in ex-nuptial fertility rates of not-married women although there was an increase in ex-nuptial fertility among women in the 20-24 age group. As we pointed out in our article, 17 per cent of couples with a female partner in this age group were de facto couples. Horn appears to doubt the greater availability of abortion since the early 1970s. Abortion became legal in Australia in the early 1970s. We would be surprised if legalization did not lead to greater availability. As regards evidence of a rise in abortions, there are no data on the number of abortions at the national level. South Australia is the only state where data are available and they show that the number of abortions in 1970 was 1,330; in 1971 it nearly doubled to 2,519 and in 1986, the number was 4,323 (South Australian Year Book, 1974; 1988). Abortion became legal in SA as of January 1970. In his second point, Horn questions whether the "conventional view that ex-nuptial births result from brief encounters or arrangements by women to have children in the single state..", which we said was reasonably accurate ten years ago, has changed since. The data that we have presented in our article showing that a significant proportion of ex-nuptial births now occur to couples who have lived together for some time suggest that that view needs to be changed because it is no longer accurate. In point (3) Horn is mistaken in saying that there is a positive relation between socio-economic status and ex-nuptial fertility. The relationship is negative. Horn is concerned about the small number
166
of cases. We have statistically tested the differences in the proportions in Table 4. Chi-square tests show that the differences in 13 of the 14 pairs of percentage distributions (de facto couples with children compared with de facto couples without children and then with married couples with children) are too large to be due to sampling errors (p < .01 for i0 pairs and p < .025 for 3 pairs). The statistical validity of the data in Table 4 is therefore not dubious as Horn seems to think. The only pair of distributions that did not differ significantly were those for female partners in de facto relationships with children and married couples by occupation or employment status. In his point (4), Horn says that other factors might be more important than economic trends in influencing marriage patterns, but he does not give any data or evidence to support his hypothesis. Studies of marriage patterns in Australia (McDonald, 1974) and couples in de facto relationships (Khoo, 1987) have indicated that people tended to postpone marriage in uncertain economic times and two-thirds of de facto couples said they expected to marry when their financial situation improved. Horn's fifth point reflects his complete misunderstanding of the data presented in Table 6 in our article. He seems to imply there is something wrong in the data for Vietnamese women yet what he says does not contradict the figures in our Table. In fact the data he presents show the same situation. He says that "taking ex-nuptial births for which paternity is acknowledged, the rate of 5 per cent for Vietnam is well below the Australian and Northern European countries...". The 5 per cent is not a rate and the correct proportion of ex-nuptial births for which paternity is acknowledged is 113/316 or 36 per cent, which is what is shown in our Table 6 for Vietnamese women. Horn is right in saying that the proportion of ex-nuptial births to Vietnamese women that were acknowledged is much less than that for women born in Australia or Europe. We also said so on page 171 of our article. He is also right in saying that the percentage for paternity not acknowledged is high. It has to be so because both figures should add to i00 per cent. We are therefore puzzled as to what he was trying to say in his point (5). He says in point (6) that our article refers to ex-nuptial births without reference to first or subsequent ones. On the top of page 169, we pointed out that 45 per cent of births to women aged 20-29 who were living in de facto relationships were first births. Therefore, 55 per cent were second and higher order births, even though we did not spell this out. Also in the middle of page 173, we pointed out that 18 per cent of de facto couples who already had one ex-nuptial birth wanted to have more children but did not plan to marry. Siew-Ean Khoo Adelaide Peter McDonald Melbourne
167
REFERENCES
KHO0, S.E. (1987), "Living together as married: A profile of de facto couples in Australia", Journal of Marriage and the Family, 49 (February): 185-191. McDONALD, P.F. (1974), Marriage in Australia: Age at First Marriage and Proportions Marrying in Australia, 1860-1971, Australian Family Formation Project Monograph No. 2, Australian National University, Canberra. AUSTRALIAN BUREAU OF STATISTICS, South Australian South Australian Office, Adelaide.
Year Book,
197~,
AUSTRALIAN BUREAU OF STATISTICS, South Australian South Australian Office, Adelaide.
Year Book,
1988,
168