ESTIMATING PATERNITY IN AUSTRALIA,

GORDON A. CARMICHAEL, PH.D.* ESTIMATING PATERNITY IN AUSTRALIA, 1976-2010 Demographers have historically studied fertility with reference to female p...
Author: Zoe Richard
0 downloads 2 Views 416KB Size
GORDON A. CARMICHAEL, PH.D.*

ESTIMATING PATERNITY IN AUSTRALIA, 1976-2010 Demographers have historically studied fertility with reference to female populations. As bearers and prime nurturers of children, women have been more reliable sources of data, and their lives are more strongly impacted by parenthood. Greater concern with men’s role in fertility decision-making and child socialization has, however, lately focused more attention on measuring male fertility. This paper estimates age-specific and total paternity rates for Australia during 1976-2010. Australia has long published ages of fathers for marital births, but from 1976, with cohabitation proliferating, also began publishing them for non-marital births where paternity was acknowledged. The methodology developed here estimates annual paternal age distributions for non-marital births where paternity was not acknowledged, enabling age-specific and total paternity rates to be estimated and compared with female fertility rates. Keywords: male fertility, paternity rates and trends, acknowledgement of paternity, non-marital births, Australia

Fertility is by far the largest subdiscipline in contemporary demography. Yet until a very few years ago, women comprised virtually its sole objects of study. Men, if they appeared at all, usually did so as shadows; as partners-by-implication of those engaged in childbearing. Thus did Bledsoe, Lerner and Guyer (2000, p. 1) commence the introduction to their collection of readings on men’s role in fertility, part of an awakening to the topic that received substantial stimulus from the Programme of Action agreed upon at the 1994 Cairo International Conference on Population and Development, and that from the mid-1990s spawned a raft of papers with titles featuring phrases like “It takes two to tango, doesn’t it?” (Corijn, Liefbroer, & de Jong Gierveld, 1996), “Bringing men back in” (Goldscheider & Kaufman, * Australian Demographic and Social Research Institute, Australian National University. Correspondence concerning this article should be addressed to the author, ADSRI, Australian National University, Canberra ACT 0200, Australia. Email: [email protected] FATHERING, VOL. 11, NO. 3, FALL 2013, PP. 256-279. © 2013 by the Men’s Studies Press, LLC. All rights reserved. http://www.mensstudies.com fth.1103.256/$15.00 • DOI: 10.3149/fth.1103.256 • ISSN/1537-6680 • eISSN/1933-026X

256

ESTIMATING PATERNITY IN AUSTRALIA, 1976-2010

1996), “Absent and problematic men” (Greene & Biddlecom, 2000) and “Where are all the men?” (Forste, 2002). The awakening also generated a two-day Conference on Fathering and Male Fertility held in Bethesda in March 1997 to climax a year-long series of workshops at which participants “thought deeply and creatively about how to improve the information available to society on fathers” (Federal Interagency Forum on Child and Family Statistics, 1998, n.p.). Meanwhile, at the Australian National University, Francisco (1996, p. 8) wrote an impassioned plea for a “two-sex demography,” stressing the “complementarity between the sexes”—the proposition that “each sex is mediated by the other and thus both sexes matter,” in tandem holding the key to real understanding of the causal mechanisms underpinning demographic patterns and trends. Why had men’s role in reproduction been overlooked for so long? Coleman (2000, pp. 29-30), echoing Karmel (1947), Hajnal (1948) and Brouard (1977), put it this way: Even in modern societies, demographic data are usually insufficient to describe and analyse formally the contribution of males to fertility and reproduction in anything like the same detail or accuracy as the contribution from females. … it is scarcely possible to test theories with the scanty data available. Fertility has conventionally been measured with reference to females, who as the bearers of children are more reliable sources of data than men are and whose lives, because they are also the prime nurturers of children, are more substantially impacted. The latter can be unaware or uncertain of their paternity of particular children, or else reluctant to acknowledge it because of legal liabilities for support that might follow and/or lack of regular involvement in a child’s life. In Australia, birth registration has only ever recorded details of a child’s father if he was legally married to the mother or, if he was not, if he formally consented to being registered as the father. This has long meant that births of what are known as “ex-nuptial” children—those born outside formal marriages—were often registered without details of the father being recorded. With censuses never broaching the topic of male fertility, or paternity, the other major potential source of data for developed populations has been surveys, but a number of studies have reported seriously incomplete survey coverage of paternity compared to female fertility (Cherlin, Griffith, & McCarthy, 1983; Fikree, Gray, & Shah, 1993; Furstenberg, 1988; Juby & Le Bourdais, 1999; Poulain, Riandey, & Firdion, 1991; Rendall, Clarke, Peters, Ranjit, & Verropoulou, 1999). This appears to relate especially to situations in which the father no longer co-resides with the child, or never did. Children with whom there was never or is no longer regular engagement are ignored and/or, because repartnering creates competing time and resource demands across households, focus and commitment are transferred to progeny and even stepchildren from new relationships (Manlove, Logan, Ikramullah, & Holcombe, 2008; Manning & Smock, 2000; Toulemon & Lapierre-Adamcyk, 2000). Moreover, aside from any tendency to underreport their paternity, non-resident fathers are frequently underrepresented in survey samples (Rendall et al.). Some more recent survey-based studies have argued that their data largely overcome these problems (Guzzo & Furstenberg, 2007; Manlove et al.), but Hynes, Joyner, Peters and Yang DeLeone (2008) caution that different datasets can produce inconsistent findings, with differences in sample composition, data quality or cohort differences all potential explanations. One recent study to have substantially circumvented data problems is Lappegárd, Rønsen and Skrede’s (2011) analysis of

257

CARMICHAEL Norwegian population register data, in which only 1-1.5% of children had no registered father, but few countries have such data. Limited previous research exists on “male fertility” in Australia (Gray, 2002). In this paper the focus is on measuring male fertility, or paternity, trends since the mid-1970s using birth registration data. While Australian vital statistics have long tabulated marital births by age of father as well as by age of mother, until the mid-1970s ex-nuptial births were tabulated only by age of mother, because data for fathers were so often missing. This precluded calculating age-specific paternity rates for males that could be set alongside fertility rates routinely published for females. Pressure only arose through the early 1970s for data to be produced that (a) distinguished by age of mother between ex-nuptial births for which paternity had, and had not, been acknowledged and (b) tabulated ex-nuptial births where paternity had been acknowledged by age of father. It arose as a consequence of informal cohabitation becoming more widespread as feminism and the contraceptive revolution undermined the Judeo-Christian precept that regular intercourse should be confined to marriage, and it was realized that ex-nuptial births would in future no longer be as commonly the product of unintended teenage pregnancies (for which induced abortion would henceforth be a more available resolution option anyway). Instead they would increasingly take place within consensual unions, where they would to a greater extent be planned and fathers would more often be only too happy, and even anxious, to acknowledge paternity. Thus it was that from 1976 the Australian Bureau of Statistics (ABS) began producing data on ex-nuptial births by whether or not paternity had been acknowledged and, if it had, by age of father as well as age of mother. The primary objective here is to use these data and the long available data on nuptial births by age of father to produce annual estimates of agespecific and total paternity rates for the ensuing 35 years that can be set alongside equivalent female fertility rates. The methodology followed obviously also requires estimation of annual distributions of ex-nuptial births for which paternity was not acknowledged by age of father. These births, however, have been such small proportions of total births over the period in question that errors in male paternity rates attributable to this process are almost certainly minor. Following a discussion of the background to the ABS’s decision to begin producing, from 1976, the data relied upon here, two alternative approaches to estimating annual age distributions of fathers of ex-nuptial children who did not acknowledge paternity are outlined. Averaging results from those approaches then effectively yields a third approach, which is the one for which results are reported. Age-specific paternity rates are generated. They are then compared with female age-specific fertility rates for the 35-year period 1976-2010, and the deferment of paternity and childbearing within that period is discussed. The data are subsequently used to generate measures of the dispersion by age of paternity and childbearing. These reveal, unsurprisingly given that male paternity is less biologically age-constrained than is female fertility, greater dispersion for males, but also yield evidence that the gender gap in the degree of dispersion has been narrowing. In combination with the “Background” discussion below, this final interpretative analysis provides something of a potted social history of parenting in Australia since the 1960s. BACKGROUND The history of non-marital pregnancy in Australia (i.e., fertility following non-marital conception) was comprehensively reviewed in the mid-1990s (Carmichael, 1996). This

258

ESTIMATING PATERNITY IN AUSTRALIA, 1976-2010

analysis used long series of data on (a) ex-nuptial births by age of mother and (b) nuptial first births within 0-7 months of marriage by age of mother, each lagged back to year of and age at conception using lexis diagram principles and an assumed gestation period of 38 weeks (Carmichael, 1985). This lagging allowed age-specific non-marital conception rates, which could be partitioned additively according to whether confinement subsequently occurred inside or outside marriage, to be calculated. The focus on conceptions rather than confinements recognized that the former all occurred to unmarried women as a common population at risk, whereas the latter did not. Naturally, only conceptions that resulted in live births were captured, not those ending in miscarriage, stillbirth or, importantly once more openly available, induced abortion. It was shown that dramatic change in the early 1970s had greatly reduced childbearing resulting from non-marital sexual activity. During 1970-1976 the rate of non-marital conception leading to live births plummeted from 51.6 to 29.7 conceptions per 1,000 unmarried women aged 15-44, the component associated with marital confinement falling from 25.4 to 9.8 and its additive counterpart associated with non-marital confinement (ex-nuptial birth) from 26.2 to 19.9. Decline in the former component had especially been focused on age groups 15-19 and 20-24. The rate of non-marital conception at ages 15-19 leading to marital confinement dropped from 29.4 per 1,000 in 1970 to 9.9 per 1,000 in 1976, and would fall further to just 1.3 per 1,000 by 1995 as pregnant teenaged brides became “almost a thing of the past” (Carmichael, 1996, p. 302). The rate at ages 20-24 also fell sharply, from 32.5 in 1970 to 12.5 in 1976 and 4.1 in 1995. These declines reflected the combined impact of (i) wholesale rejection, aided from mid-1973 by introduction of a Supporting Mother’s Benefit, of the norm that marriage was the honorable response to unintended premarital pregnancy, (ii) improved contraception in non-marital sexual encounters (facilitated not least by their occurring increasingly in cohabiting unions), and (iii) greatly improved access to induced abortion following a landmark legal ruling in New South Wales in late 1971. This explicitly admitted “effects of economic and social stresses that may be pertaining at the time” as relevant to establishing danger to a woman’s mental health sufficient to justify abortion (Mr Justice Levine in R. v. Wall et al.). The first of these forces, of course, effectively turned many non-marital conceptions that formerly would have resulted in marital confinements into ones that produced ex-nuptial births instead, as women chose to have their babies unmarried rather than in hasty marriages and fathers, perhaps, no longer felt as obligated to “do the right thing.” Consequently declines in overall and age-specific rates of non-marital conception leading to non-marital (ex-nuptial) confinement were less spectacular. At ages 15-19 and 20-24 the respective 1970-1976 reductions were from 21.2 to 16.7 and from 34.2 to 26.0 conceptions per 1,000 unmarried women. They also did not continue through the 1980s and into the 1990s as did those in rates of non-marital conception leading to marital confinement, because as cohabitation proliferated, ex-nuptial childbearing within consensual unions became an increasingly important feature of the Australian fertility landscape. The transformation of ex-nuptial fertility in Australia from substantially the product of unintended pregnancies conceived in non-co-residential relationships before 1970 to overwhelmingly the (probably mostly planned) product of consensual unions beyond that date is reflected in both its greatly increased proportion of total fertility and its changing maternal age composition. Figure 1 shows trends in overall and age-specific percentages of births that were ex-nuptial between 1908 and 2010. Until the mid-1950s the overall figure re-

259

CARMICHAEL

All ages

% Ex-nuptial

40 30 20 10 0 1908 1918 1928 1938 1948 1958 1968 1978 1988 1998 2008 Year

Ages 15-29 100

% Ex-nuptial

80 60 40 20 0 1908 1918 1928 1938 1948 1958 1968 1978 1988 1998 2008 Year 15-19

20-24

25-29

Ages 25-44 35

% Ex-nuptial

30 25 20 15 10 5 0 1908 1918 1928 1938 1948 1958 1968 1978 1988 1998 2008 Year 25-29

30-34

35-39

40-44

Figure 1. Overall and age-specific percentages of births ex-nuptial, 1908-2010.

260

ESTIMATING PATERNITY IN AUSTRALIA, 1976-2010

mained fairly constant at around five percent. There then began an upswing that only recently has shown any sign of ending, at just under 35%. Through the late 1950s and 1960s it was driven by teenage pregnancies that, among other things, supplied a growing adoption market as parental oversight of courtship weakened in the face of readier access to motor vehicles greatly increasing young people’s mobility, and hence their capacity to find privacy for intimate encounters (Carmichael, 1988). But thereafter it has been driven chiefly by the proliferation of cohabitation and mounting acceptance of childbearing within it. Youthful childbearing these days occurs overwhelmingly outside formal marriages but to cohabiting couples. At the end of the post-war marriage boom in the early 1970s around onethird of births to women aged 15-19 were ex-nuptial (Carmichael, 1988). This figure reached 90% in the mid-1990s, and now hovers just below 95%. At ages 20-24 the early 1970s proportion was below 10%; it passed 60% in 2002 and is now only marginally below twothirds. At ages 25-29 fewer than five percent of births were ex-nuptial in the early 1970s, this figure passing 30% in 2006. There have been sharp upswings at older ages as well (Figure 1). Standardizing to the age structure of the female population of reproductive age in 1990, Table 1 shows trends since 1908 in the age composition of women conceiving children who would be born ex-nuptially. Little change is apparent until the 1960s, when the importance of age group 15-19 rose sharply from 30 to over 40% (the peak was almost 43% in 1972) as teenage pregnancy reached unprecedented levels at a time when oral contraceptives were still rarely available to the unmarried and abortion law reform had yet to occur. Compensating declines occurred at and above ages 25-29. Then beyond 1975 age group 15-19 rapidly decreased in importance again, to little more than 15% by 2010. Counterbalancing increments were largely concentrated at what have latterly become the peak reproductive ages, 25-29 and 30-34, although more recently the shares of older age groups have also risen sharply. This shift in age composition reflects the transition from rampant unplanned teenage childbearing to widespread ex-nuptial childbearing within cohabiting unions. Also reflective of that transition are marked increases, across all reproductive ages and within five-year age groups, in the frequency with which paternity of ex-nuptial children has been acknowledged. As already intimated, data on paternity only began to be tabulated once it was realized that the propensity to acknowledge it was on the rise, so the base from which the trend commenced is unclear. Across all reproductive ages acknowledgement of paternity of ex-nuptial children rose from 47% in 1976 to pass 60% in 1981, 70% in 1986, 80% in 1992 and 90% in 2005 (Figure 2). New Zealand data from searches of the 1966 and 1976 birth registers indicated that 33% and 54% of ex-nuptial births were paternity-acknowledged in those years (Carmichael, 1982, Table 5.6), so Australia’s base was almost certainly well below 47%, and possibly of the order of 30%. Increments by age have also been spectacular, age group 16-19 stabilizing more recently at around 85% of ex-nuptial births paternity-acknowledged and older age groups mostly clustered just above 90%. In individual cases acknowledging paternity does not guarantee parents were cohabiting, but the two circumstances are so likely to coexist that Figure 2 nonetheless clearly implies that, overwhelmingly, ex-nuptial children nowadays are born to cohabiting parents. The marked shift in the maternal age distribution of ex-nuptial births away from teenagers since the early 1970s (Table 1) also leaves little doubt that this is the case.

261

CARMICHAEL Table 1 Age-Standardized* Percentage Distributions by Age of Woman of Conceptions Resulting in Ex-Nuptial Live Births: Australia 1908-2010 Year

15-19

20-24

25-29

30-34

35-39

40-44

Total

1908 1910 1915 1920 1925 1930 1935 1940 1945 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010

28.6 30.3 31.3 29.7 32.5 34.0 32.1 29.6 28.8 30.4 28.3 30.1 37.3 40.9 40.0 37.3 31.2 26.2 23.3 19.4 15.8 15.3

34.3 32.8 33.2 33.3 30.4 29.4 29.9 30.9 32.0 29.3 29.8 28.5 27.0 27.6 28.3 31.4 32.9 33.1 31.4 30.5 27.3 26.6

18.1 18.4 16.7 17.3 16.5 15.7 17.5 17.8 18.3 18.9 20.0 19.0 16.7 15.7 17.2 18.5 20.9 22.6 23.7 24.7 25.7 24.6

10.3 9.9 10.0 10.4 10.5 10.6 10.6 11.8 11.3 12.5 12.4 13.2 11.0 9.4 9.4 8.7 10.7 12.6 14.4 16.5 19.3 19.9

6.5 6.6 6.7 7.0 7.5 7.6 7.6 7.4 7.3 6.8 7.5 7.2 6.3 5.0 4.2 3.2 3.6 4.7 6.1 7.5 10.0 11.1

2.2 2.0 2.0 2.3 2.6 2.8 2.3 2.5 2.2 2.2 2.0 2.1 1.8 1.4 0.9 0.9 0.7 0.8 1.1 1.4 1.8 2.5

100.0 100.0 99.9 100.0 100.0 100.1 100.0 100.0 99.9 100.1 100.0 100.1 100.1 100.0 100.0 100.0 100.0 100.0 100.0 100.0 99.9 100.0

* Standardized to age composition of female population of reproductive age in 1990.

ESTIMATING PATERNITY The core task in estimating paternity is to estimate the age distribution of fathers of exnuptial children who did not acknowledge paternity. The ages of fathers of nuptial children and of ex-nuptial children whose paternity was acknowledged are known since 1976, and fortunately the group for which they are not known is small throughout that period. In the mid-1970s this was so because, while paternity was not acknowledged for more than half of ex-nuptial births, ex-nuptial births were little more than 10% of all births. Births where paternity was not acknowledged were thus generally a little over five percent of total births. Then, as the proportion of total births that were ex-nuptial rose sharply, so, too, in interactive compensation, did the level of acknowledgement of paternity of those births. The proportion of all births for which paternity was not acknowledged consequently declined to below five percent after 1983, below four percent after 1997, and reached a nadir of 3.04% in 2007 (Figure 3).

262

ESTIMATING PATERNITY IN AUSTRALIA, 1976-2010 All ex-nuptial births % Paternity Acknowledged

100 80 60 40 20 0 1976

1981

1986

1991

1996

2001

2006

1996

2001

2006

Year

Ages 16-44

% Paternity Acknowledged

100

80

60

40

20

0 1976

1981

1986

1991 Year

16-19

20-24

25-29

30-34

35-39

40-44

Cumulative %

Figure 2. Overall and age-specific percentages of ex-nuptial births paternity-acknowledged, 1976-2010. 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 1976

1981

1986

1991

1996

2001

2006

Year Nuptial

Ex-nuptial P A

Ex-nuptial P not A

Figure 3. Distribution of births by nuptiality and whether paternity acknowledged, 1976-2010.

263

CARMICHAEL One is therefore seeking to estimate annual paternal age distributions for a relatively small minority of total births, limiting the degree of error likely to occur in estimated paternal age distributions across all births. The problem at hand is essentially one requiring imputation of missing data, and the broad approach adopted is one described by Gelman and Hill (2007, p. 533) as “using information from related observations.” One estimation option of this type might be to assume that the age distribution of fathers of ex-nuptial children who did not acknowledge paternity was identical to that of fathers of such children who did acknowledge paternity. However, as mothers of ex-nuptial children have been appreciably younger where paternity was not acknowledged than where it was acknowledged, that is not a reasonable approach. During 1976-1980, for example, just 28.5% of ex-nuptial births for which paternity was acknowledged, compared to 45.1% of those for which it was not acknowledged, occurred to mothers aged less than 20. During 2006-2010 equivalent figures were 10.2% and 20.0% respectively. Over both periods proportions of mothers of ex-nuptial children whose paternity was and was not acknowledged who were aged 20-24 were by contrast very similar—34.2% and 32.1% during 1976-1980, and 27.4% and 27.9% during 2006-2010. So over the period of focus, mothers of ex-nuptial children whose paternity was not acknowledged have been particularly likely to be aged less than 20 and comparatively less likely than those of paternity-acknowledged ex-nuptial children to be 25 or older. It stands to reason that the fathers of their children would also have been comparatively young. Two approaches to estimating paternal age distributions for ex-nuptial births whose paternity was not acknowledged were adopted. Both were based on comparing paternal and maternal cumulative percentage distributions by age of ex-nuptial births for which paternity was acknowledged, assuming that the gender differential between these distributions was in some way transferrable to paternal and maternal cumulative percentage distributions by age of ex-nuptial births for which paternity was not acknowledged. Under the first approach, for each year across the period 1976-2010 ratios of paternal to maternal cumulative percentages of paternity-acknowledged ex-nuptial births at each single year of age were computed, then applied to cumulative percentages of ex-nuptial births for which paternity was not acknowledged by age of mother to estimate a cumulative percentage distribution of such births by age of father. The assumption underpinning this approach was that the proportionate relationships between paternal and maternal cumulative distributions at each parental age would be the same for ex-nuptial births whose paternity was not acknowledged as for those whose paternity was acknowledged. Having thus estimated cumulative percentage distributions of ex-nuptial births whose paternity was not acknowledged by age of father it was a simple matter, using the known total numbers of such births during each year, to produce estimated distributions of those births by single-year ages of father. Figure 4 shows selected sample cumulative percentage distributions of ex-nuptial births by sex of parent (M or F) and whether paternity had been acknowledged (PA or PnotA) from this approach. In each case the suggestion from these graphs is that mothers of children for whom paternity was not acknowledged were younger than those of children for whom it was acknowledged (compare the dotted and dashed lines) to a greater degree than were fathers of children whose paternity was not acknowledged compared to those of children for whom it was acknowledged (compare the two solid lines). In other words, given that there is every reason to expect that relative youth of both parents is a material factor in relationships that produce ex-nuptial children being sufficiently uncommitted for paternity not to be acknowledged, the first estimation approach yielded results that seemed a bit conservative (not young enough).

264

ESTIMATING PATERNITY IN AUSTRALIA, 1976-2010

1976 plots Cumulative %

100 80 60 40 20 0 15

20

25

30

35

40

45

50

55-59

Age MPA

FPA

FPnotA

MPnotA

1990 plots Cumulative %

100 80 60 40 20 0 15

20

25

30

35

40

45

50

55-59

Age MPA

FPA

FPnotA

MPnotA

2010 plots Cumulative %

100 80 60 40 20 0 15

20

25

30

35

40

45

50

55-59

Age MPA

FPA

FPnotA

MPnotA

Figure 4. Cumulative age distributions of ex-nuptial births by sex of parent and whether paternity acknowledged, selected years using first estimation method.

265

CARMICHAEL

1976 plots Cumulative %

100 80 60 40 20 0 15

20

25

30

35

40

45

50

55-59

Age MPA

FPA

FPnotA

MPnotA

1990 plots Cumulative %

100 80 60 40 20 0 15

20

25

30

35

40

45

50

55-59

Age MPA

FPA

FPnotA

MPnotA

2010 plots Cumulative %

100 80 60 40 20 0 15

20

25

30

35

40

45

50

55-59

Age MPA

FPA

FPnotA

MPnotA

Figure 5. Cumulative age distributions of ex-nuptial births by sex of parent and whether paternity acknowledged, selected years using second estimation method.

266

ESTIMATING PATERNITY IN AUSTRALIA, 1976-2010

A second approach to estimation of paternal age distributions for ex-nuptial births whose paternity was not acknowledged was to base it on maintaining the same absolute (as distinct from proportionate) gender differences between paternity-acknowledged and -not-acknowledged cumulative percentage distributions of births by age of parent. Thus, for each year across the period 1976-2010, the differences at each single year of age between cumulative maternal and paternal percentages of ex-nuptial births for which paternity had been acknowledged were calculated, and these absolute differences were subtracted from maternal cumulative percentages of ex-nuptial births for which paternity was not acknowledged to yield a second set of equivalent paternal cumulative percentages. These were then again used in conjunction with known annual numbers of such births to produce a second set of estimated distributions of births by single-year ages of father. Equivalent plots from this approach to those shown in Figure 4 are presented in Figure 5. Compared to Figure 4, only one line in each graph changes—that designated “MPnotA” (males, paternity not acknowledged). The graphs in Figure 5 portray much younger paternal age distributions for ex-nuptial births where paternity was not acknowledged than do those in Figure 4 (i.e., the solid lines are appreciably further apart). But are they too young? Figure 6, which compares actual (not cumulative percentage) estimated distributions of ex-nuptial births for which paternity was not acknowledged by age of father yielded by the two estimation methods outlined suggests that they may be. These graphs have several interesting features. The pronounced peak across ages 19-21 in 1976 had moderated and shifted to the right by 1990, then flattened considerably by 2010 when the total number of ex-nuptial births for which paternity was not acknowledged was around 25% lower than in 1976. These trends reflect the changing character of ex-nuptial fertility and associated increased preparedness of fathers to acknowledge paternity of ex-nuptial children. Secondly, paternal age distributions for ex-nuptial births obtained using the two estimation methods are well nigh identical beyond age 35, and arguably beyond age 30. The chief difference between them is the transfer under Method 2 of significant numbers of births allocated by Method 1 to age group 20-29 to age group 15-19 (or more accurately, to age group under 20, since births shown in source data as occurring at age 15 actually occurred at ages 15 or younger). Just as one sensed that numbers of fathers of paternity-not-acknowledged ex-nuptial children aged 16 or younger estimated by Method 1 were too low, so one senses that numbers estimated by Method 2 may be too high; that very young mothers neither married to nor cohabiting with their children’s fathers are likely often to have been left in that predicament by somewhat older, not same-aged, males. Consequently, in estimating overall paternity it was decided to use means of the age distributions of fathers of paternity-not-acknowledged ex-nuptial children estimated using Methods 1 and 2 (effectively Method 3). This decision is arbitrary, but hardly problematic, as whatever distributions were chosen—those derived by Method 1, Method 2, or their average—paternity rates for males in their mid-teens were going to be very low. No apology is made for the judgement exercised in concluding (a) that Methods 1 and 2 established lower and upper bounds, respectively, to the youth of age distributions of fathers of paternity-not-acknowledged ex-nuptial births, and (b) that reality probably lies somewhere in between. Both methods adopted straightforward, yet plausible, assumptions as to how gender differences in parental age distributions for paternity-acknowledged ex-nuptial births might inform those for paternity-not-acknowledged ex-nuptial births, but some as-

267

CARMICHAEL

1976 plots 1000

Births

800 600 400 200 0 15

20

25

30

35

40

45

50

55-59

50

55-59

50

55-59

Age Method 1

Method 2

1990 plots 1000

Births

800 600 400 200 0 15

20

25

30

35

40

45

Age Method 1

Method 2

2010 plots 1000

Births

800 600 400 200 0 15

20

25

30

35

40

45

Age Method 1

Method 2

Figure 6. Age distributions of fathers of ex-nuptial children whose paternity was not acknowledged estimated by methods 1 and 2, selected years.

268

ESTIMATING PATERNITY IN AUSTRALIA, 1976-2010

Table 2 Age-specific Paternity Rates per 1,000 Males Derived Using Methods 1 and 2, 1976, 1990 and 2010 1976

1990

Age

Method 1

Method 2

Method 1

Method 2

15 16 17 18 19 20 21 22 23 24 25 26 27 28 29

0.1 0.9 5.7 14.0 27.0 38.5 56.4 74.8 95.8 117.1 137.0 149.2 154.7 155.3 143.6

3.8 4.7 8.0 15.5 26.9 37.5 55.0 73.4 94.2 116.1 135.8 148.1 153.7 154.5 143.0

0.5 1.4 3.7 9.8 17.6 26.5 35.2 45.2 57.3 70.7 85.9 99.8 115.9 126.8 130.7

1.8 2.9 5.7 11.1 18.1 26.1 34.6 44.4 56.6 69.9 85.1 99.2 115.4 126.4 130.4

2010 Method 1

Method 2

0.9 2.1 4.9 9.1 14.4 19.6 24.4 29.8 36.4 42.2 51.0 62.4 74.2 87.1 99.3

1.6 2.8 5.6 9.9 14.9 19.6 24.3 29.6 36.0 41.8 50.6 62.1 74.0 86.9 99.0

sessment of how reasonable the results produced were was always going to need to be made. In the final analysis the decision to adopt Method 3 has very limited implications for the overall age-specific paternity rates estimated.This is so because (a) as previously noted the proportions of annual total births for which paternal ages needed to be imputed are low (35%), and (b) variation in rates across imputation methods is greatest at the youngest ages, where paternity rates are very low, and minimal at older ages where they are far higher. This limited sensitivity to imputation method is illustrated in Table 2, which shows paternity rates obtained using Methods 1 and 2 by single years of age across ages 15-29 for 1976, 1990 and 2010. Data for older ages are not shown simply because differences by imputation method quickly disappear altogether. Method 3 paternity rates graphed and further analysed in the following section lie midway between those yielded by Methods 1 and 2. MALE PATERNITY RATES COMPARED TO FEMALE FERTILITY RATES Trends since 1976 in age-specific paternity rates based on published distributions of nuptial births and paternity-acknowledged ex-nuptial births by age of father, and the means of distributions of paternity-not-acknowledged ex-nuptial births by age of father estimated using Methods 1 and 2, are shown in Figure 7. For comparison, trends in female age-specific fertility rates are shown in Figure 8. It is immediately apparent that paternity rates at ages 15-19 have been markedly lower than fertility rates since 1976, and while falling

269

CARMICHAEL through the late 1970s remained relatively stable beyond 1980. Fertility rates, by contrast, follow fairly continuous downward trajectories until beyond the millennium, although they, too, fell faster through the late 1970s than thereafter. Earlier female data (Carmichael, 2013, Figure 1) show this trend actually dating from 1971, before which teenage fertility had trended up through the 1960s as the sexual revolution unfolded and unmarried teenagers were largely denied access to the oral contraception that enabled married couples to better control their fertility through that decade. The sharp turnaround after 1971 may have been partly a product of emergent consensual partnering facilitating access to the pill for unmarried teenagers, but more crucial was the aforementioned liberalization of abortion law in October of that year. Young men will also have benefitted from the capacity this conferred to avoid unwanted births, and so the downturn in teenaged paternity evident through the late 1970s in Figure 7 undoubtedly represents the tailend of a more pronounced trend that occupied the preceding quinquennium as well. Male paternity and female fertility trends at ages 20-24 in Figures 7 and 8 are not dissimilar, although consistent with expectation male rates are lower. Trends are relentlessly downward until 2006, with rates converging to narrower ranges but paternity/fertility at all times positively associated with age. Clearly the 35-year period under review saw having children in early adulthood become distinctly less popular among both sexes. Slight resurgences in rates after 2006 appear at this time to have been temporary, possibly associated with the Howard Government’s “baby bonus,” introduced in 2004, although its impact on fertility remains a topic of debate (Anderson, 2007; Drago, Sawyer, Sheffler, Warren, & Wooden, 2009; Heard, 2006, 2010; Jackson, 2006; Lain et al., 2009; Lattimore & Pobke, 2008; Parr & Guest, 2011; Read, Crockett, & Watson, 2007; Risse, 2010). At ages 25-29 female fertility rates exhibit a neat crossover (Figure 8). While in 1976 fertility was highest at age 25 and lowest at age 29, by the early 1990s this pattern had reversed and became entrenched over the ensuing two decades as fertility fell sharply at ages 25 and 26 but followed flatter trajectories at ages 28 and 29. There is a hint of something similar having happened perhaps a decade earlier for males in this age group (Figure 7), with paternity rates clustered in a relatively narrow range in 1976 but fanning out thereafter as paternity declined more rapidly at age 25 than it did at age 29. Post-2006 upward kinks are again evident for both sexes, as they were at ages 20-24. The standout feature of the male graph for age group 30-34 in Figure 7 is the marked convergence of single-year age group paternity rates across the period under review. In 1976 male paternity at age 30 was 60 per 1,000 higher than it was at age 34. Thirty years later all five paternity rates were within a range of less than 10 per 1,000. Paternity had declined at ages 30 and 31, especially through the late 1980s and the 1990s, but had risen persistently at older ages, leading to the convergence described. It had also risen again at the younger ages beyond the millennium, by which time the early to mid-30s had clearly become consensus preferred ages for Australian males to have children. Female fertility rates at ages 30-34 also exhibit convergence across the three decades after 1976, but most noticeably through the 1990s. Prior to that they rose more or less in parallel, but then fertility at age 30 turned downward again, that at age 31 plateaued, and at age 32 upward momentum slackened. It was reasserted at these younger ages from about 2001. Clearly Australians have been delaying parenthood—having children less frequently at younger ages and more often at older ones. Maternal ages in the late 20s and early 30s have been the pivots around which this trend has occurred, with fertility initially falling as childbearing was delayed but

270

ESTIMATING PATERNITY IN AUSTRALIA, 1976-2010

20-24

160

160

120

120

ASPR

ASPR

15-19

80

40

80

40

0

0 1976

1981

1986

1991

1996

2001

2006

1976

1981

1986

Year 16 19

17

20 23

25-29

2001

2006

21 24

22

30-34

160

160

120

120

ASPR

ASPR

1996

Year

15 18

80

40

80

40

0

0 1976

1981

1986

1991

1996

2001

2006

1976

1981

1986

Title

1991

1996

2001

2006

Year

25 28

26 29

27

30 33

35-39

31 34

32

40-44

160

160

120

120

ASPR

ASPR

1991

80

40

80

40

0

0 1976

1981

1986

1991

1996

2001

2006

1976

1981

1986

Title 35 38

1991

1996

2001

2006

Year 36 39

37

40 43

Figure 7. Age-specific male paternity rates, 1976-2010.

271

41 44

42

CARMICHAEL 20-24

160

160

120

120 ASFR

ASFR

15-19

80

80

40

40

0

0 1976

1981

1986

1991

1996

2001

1976

2006

1981

1986

15 18

16 19

20 23

17

1996

2001

2006

21 24

22

30-34

160

160

120

120 ASFR

ASFR

25-29

80

40

80

40

0

0 1976

1981

1986

1991

1996

2001

2006

1976

1981

1986

Year 25 28

1991

1996

2001

2006

Year 26 29

27

30 33

31 34

32

40-44

35-39

160

160

120

120 ASFR

ASFR

1991 Year

Year

80

80

40

40

0

0 1976

1981

1986

1991

1996

2001

1

2006

6

11

35 38

16

21

26

31

Year

Year 36 39

40 43

37

Figure 8. Age-specific female fertility rates, 1976-2010.

272

41 44

42

ESTIMATING PATERNITY IN AUSTRALIA, 1976-2010

then rising again as those who had postponed family formation at younger ages chose to initiate it at these ages instead. Paternity at ages 35-39 rose relentlessly during the period under review, and it also trended up persistently at ages 40-44 (Figure 7). As much as having children in one’s 20s became decisively less and less the thing to do, having them in ones mid- to late 30s, and even into one’s 40s became more commonplace. Women, too, became much more likely to have children aged 35-39, and fertility rates in their early 40s also increased (Figure 8). There is very little female fertility beyond age 44, but males continue to have children in reasonable numbers after that age, with paternity rates at age 52 about on a par with female fertility rates at age 44 over the period studied here. Paternity rates at these advanced ages (not shown) have also generally been rising, particularly since the mid-1990s. It is, of course, possible to generate, by summation of male age-specific paternity rates, total paternity rates (TPRs) in identical manner to conventional total fertility rates (TFRs) based on age-specific fertility rates for females. Such rates, summed over ages 15-59 rather than 15-49 in recognition that men’s biological capacity to procreate is less age-constrained than is women’s, are shown in Figure 9. Comfortingly, although hardly surprisingly, the two rates follow similar paths (alarm bells would ring were it not so), declining during the late 1970s and again during the 1990s before rising once more between 2001 and 2008, then dropping off in 2009 and 2010. TPRs tend to be marginally higher than TFRs early in the 35-year period covered, and slightly lower recently. The deferment of paternity and childbearing to which Figures 7 and 8 clearly speak can be rendered more explicit through quartiles of the age distributions of paternity and childbearing, defined as the exact ages at which annual summations of single-year age-specific paternity/fertility rates from youngest to oldest reproductive ages reach quarter, half and three-quarters of their ultimate values as total paternity and fertility rates (Figure 10). Having determined the single-year age group within which a quartile lies, linear interpolation is used to locate the exact age corresponding to it. Female plots have been extended back to 1960 to give a sense of what happened through the years leading up to the widespread 2.5

TFR/TPR

2.0 1.5 1.0 0.5 0.0 1976 1979 1982 1985 1988 1991 1994 1997 2000 2003 2006 2009 Year Female TFR

Male TPR

Figure 9. Total fertility/paternity rates by sex, 1976-2010.

273

CARMICHAEL postponement of parenthood. From the mid-1960s to the early 1970s quartiles of the age distribution of childbearing declined as the post-war marriage boom went through its final throes and oral contraception enabled older couples to better avoid unwanted higher parity births (hence the especially steep decline in the female Q3). The trend to later childbearing began in the early 1970s, the female lower quartile being the first to turn upward as teenage childbearing suddenly fell sharply as induced abortion became more accessible and cohabitation more acceptable, the latter enabling some of the unmarried young to contracept more effectively. But cohabitation would also have a much broader impact, leading in conjunction with the contraceptive revolution represented by the pill to widespread rejection of the notions that one should be married to enjoy regular sexual relations and that such a sex life should lead sooner rather than later to parenthood. The deferment of childbearing and paternity as other things—education, career, travel, having a good time—took center stage in early adulthood, for both sexes, was under way, and Figure 10 shows the trend to have been relentless. Slight interruptions in 2007-2008, mainly evident in lower quartile trend lines and perhaps signalling a particular attraction to the Government’s “baby bonus” among younger couples, seem to have been aberrations rather than forerunners of another major change of direction. Between 1976 and 2010 the female ages by which quarter, half and three-quarters of childbearing took place rose pretty much continuously from 23.0 to 26.6, 26.3 to 30.8 and 29.9 to 34.6 years. Ages by which men experienced quarter, half and three-quarters of their paternity rose likewise from 25.7 to 29.1, 29.2 to 33.2 and 33.4 to 37.3 years. Over that 35-year period median ages at childbearing and paternity essentially became lower quartiles and upper quartiles became medians as quartiles increased by between 3.4 and 4.7 years. The increases were greater for females, and this is reflected in a closing of the gender gap in the degree of dispersion of fertility/paternity by age, and a narrowing of gaps for specific age quartiles (Figure 11). As indicated by respective interquartile ranges, male paternity is more dispersed by age than is female childbearing. This is to be expected when having children is biologically less ageconstrained for men, some of whom undoubtedly use that good fortune to delay committing 38 36 34

Exact Age

32 30 28 26 24 22 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 Year Female Q1

Female Q2

Female Q3

Male Q1

Male Q2

Male Q3

Figure 10. Quartile ages of childbearing 1960-2010, and paternity 1976-2010.

274

ESTIMATING PATERNITY IN AUSTRALIA, 1976-2010

to parenthood. But the evidence of the upper graph in Figure 11 is that while childbearing in Australia became less age-dispersed through the 1960s and much of the 1970s as couples avoided higher parity births through better contraception and sterilization, then dispersion increased again as there developed a differentiation between career mothers and women who viewed children as something to enjoy after several years focused on education, career, travel and their social lives, during the second of these phases the gender gap in dispersion of paternity/fertility by age narrowed considerably. Women became more like men in their propensity to defer family formation, and probably also in their reasons for doing so, but their biological constraint may prevent them ever closing the gap completely. DISCUSSION Australian sources of fertility data and their limitations for studying men’s fertility were reviewed by Gray (2002) prior to presenting an analysis of data from the Negotiating the Lifecourse Survey. Australia’s Perinatal National Minimum Data Set is compiled and reInterquartile ranges

Q3-Q1

9

8

7

6 1960

1965

1970

1975

1980

1985

1990

1995

2000

2005

2010

Year Females

Males

Years of Age

Gender differences in quartiles 4 3.5 3 2.5 2 1.5 1 0.5 0 1976

1981

1986

1991

1996

2001

2006

Year MQ1-FQ1

MQ2-FQ2

MQ3-FQ3

Figure 11. Trends in the age dispersion of fertility/paternity by sex, 1976-2010.

275

CARMICHAEL ported on annually by the Australian Institute of Health and Welfare using data on births routinely supplied to State/Territory health authorities by attending midwives and hospital administrative staff. However, while it covers a range of issues not captured by birth registration data, the title of the publication it gives rise to each year is instructive—Australia’s Mothers and Babies—fathers are completely beyond its purview. Likewise the Australian census has only ever asked women, never men, about the number of children they have had. So birth registration data are the sole official data source that can be used for routine population-based measurement of paternity. This paper has sought to render that source more useful by providing a methodology for estimating the age distribution of fathers whose details are not registered—those associated with so-called “paternity-not-acknowledged” ex-nuptial births. This allows complete age distributions of fathers af all children born in a year to be estimated and hence age-specific and total paternity rates to be calculated and placed alongside equivalent fertility rates for females.The method has the potential to become the basis of generation of paternity rates for publication on an annual basis, particularly whilst ever the proportion of total births for which paternity is not acknowledged is as low as it has been in recent decades. Error in estimating a paternal age distribution for less than five percent of total births is unlikely, provided the method used is essentially plausible, to translate into serious error in the paternal age distribution across all births. Unsurprisingly, not a great deal that is startling emerges from the present exercise. Indeed it could hardly be otherwise. Data for females had long told us that family formation was being delayed, and nobody doubted that, in the broad, men were also part of that trend. Nevertheless, some of the subtler detail from the present analysis is of interest. The trend away from teenage parenthood appears to have occurred more rapidly for males, but lingered somewhat for females after initially being sharp, perhaps because, from 1973, Australia provided welfare assistance that made sole parenthood attractive to some young women. The convergence of age-specific paternity rates at ages 30-34 is especially fascinating. Anecdotally and from qualitative data (Carmichael & Whittaker, 2007; Carmichael, 2013), age 30 seems to be significant to a lot of young Australians as a marker to be passed before they will even contemplate serious relationships and parenthood, and certainly for young men in the early twenty-first century anywhere in their early to mid-30s has become the optimal life cycle phase for implementing any aspiration toward fatherhood. For women the corresponding phase seems to straddle their late 20s and early 30s. In the mid-1970s age group 25-29 was the prime age for men to have children, but not any more. Both sexes have also experienced large increases in childbearing/paternity beyond age 35. Older ages at fatherhood in Australia are clearly not, given the trend in total paternity (Figure 9), a product of men having larger numbers of children. They are indicative of delayed transitions to paternity (i.e., deferment by many men of first births) over the last two to three decades, a trend that is likely to be constraining completed family sizes for at least two reasons. First, the women bearing children by older men are themselves also likely to be older, so that IVF options notwithstanding, biological impediments to female partners conceiving are more likely to be encountered. Second, social impediments to additional paternity may also come into play. Later transitions to fatherhood are apt at times to render the associated lifestyle adjustment more difficult, and this can result in decisions to have fewer children than might once have been contemplated. There can also be concerns, particularly in relationships featuring substantial age differences between the parties, that an older man

276

ESTIMATING PATERNITY IN AUSTRALIA, 1976-2010

may be “too old” to have another child; that capacity to interact physically with an additional child is waning or there is a danger of one’s relationship with that child becoming more grandparental than paternal (or being perceived that way by the child and/or her/his peers and/or other adults). Evidence of these sorts of social constraints related to paternal age on ultimate family size did emerge from a recent Australian study of family formation decisionmaking (Carmichael, 2013). That study also highlighted the scenario of couples who swear off parenthood for years, then are converted to it well into their thirties as attractions of a childless lifestyle wear thin, biological clocks are activated, and/or peer experiences expose them to the joys of parenthood. Couples fitting that scenario without exception commented that their late conversions to the idea of becoming parents had limited them to having no more than two children. As indexed by quartile ages of childbearing and paternity, the trend to later active involvement in reproduction has been less steep for men than for women. Male paternity seems, however, to be more age-dispersed than does female childbearing, although the evidence of Figure 11 is that this gender difference has narrowed of late. The biological realities of male and female participation in reproduction lead one to anticpate greater dispersion by age for men, but clearly in recent decades Australian women, as they have sought to extract maximum satisfaction from educational, career and travel opportunities and freedom of social engagement before embracing parenthood, have been testing the biological limits of their capacity to reproduce to a degree they probably only previously did prior to the fertility transition of the late nineteenth and early twentieth centuries (Ruzicka & Caldwell, 1977). Men, of course, have long been free to prioritize such things in their lives if so inclined, and they, too, have clearly chosen to do so for longer and in larger proportions in recent decades. REFERENCES Anderson, M.J. (2007). Fertility futures: Implications of national pronatalist policies for adolescent women in Australia. In M.J. Albion & P. Collins (Eds.), Education, employment, and everything: Refereed proceedings of the International Women’s Conference (pp. 40-45). Toowoomba, Australia: University of Southern Queensland Women’s Network. Bledsoe, C., Lerner, S., & Guyer, J.I. (2000). Fertility and the male life-cycle in the era of fertility decline. New York, NY: Oxford University Press for the IUSSP. Brouard, N. (1977). Évolution de la fécondité masculine depuis le début du siècle [Evolution of male fertility since the beginning of the century]. Population, 32, 1123-1158. Carmichael, G.A. (1982). Aspects of ex-nuptiality in New Zealand: Toward a social demography of marriage and the family since the Second World War. Unpublished doctoral thesis, Australian National University, Canberra. Carmichael, G.A. (1985). Non-marital pregnancies in New Zealand since the Second World War. Journal of Biosocial Science, 17, 167-183. Carmichael, G.A. (1988). With this ring: First marriage patterns, trends and prospects in Australia. Australian Family Formation Monograph No. 11. Canberra, Australia: Department of Demography, Australian National University and Australian Institute of Family Studies. Carmichael, G.A. (1996). From floating brothels to suburban semirespectability: Two centuries of nonmarital pregnancy in Australia. Journal of Family History, 21, 281-315. Carmichael, G.A. (2013). Decisions to have children in late 20th and early 21st century Australia: A qualitative analysis. SpringerBriefs in Population Studies. Dordrecht, Netherlands: Springer.

277

CARMICHAEL Carmichael, G.A., & Whittaker, A. (2007). Forming relationships in Australia: Qualitative insights into a process important to human wellbeing. Journal of Population Research, 24, 2349. Cherlin. A., Griffith, J., & McCarthy, J. (1983). A note on maritally disrupted men’s reports of child support in the June 1980 Current Population Survey. Demography, 20, 385-389. Coleman, D. (2000). Male fertility trends in industrial countries: Theories in search of some evidence. In C. Bledsoe, S. Lerner & J.I. Guyer (Eds.), Fertility and the male life-cycle in the era of fertility decline (pp. 29-60). New York, NY: Oxford University Press for the IUSSP. Corijn, M., Liefbroer, A.C., & de Jong Gierveld, J. (1996). It takes two to tango, doesn’t it? The influence of couple characteristics on the timing of the birth of the first child. Journal of Marriage and the Family, 58, 117-126. Drago, R., Sawyer, K., Sheffler, K., Warren, D., & Wooden, M. (2009). Did Australia’s baby bonus increase the fertility rate? Working paper 1/09, Melbourne: Melbourne Institute of Applied Economic and Social Research. Federal Interagency Forum on Child and Family Statistics. (1998). Nurturing fatherhood: Improving data and research on male fertility, family formation, and fatherhood. Retrieved October 26, 2011 from http://fatherhood.hhs.gov/CFSForum Fikree, F.F., Gray, R.H., & Shah, F. (1993). Can men be trusted? A comparison of pregnancy histories reported by husbands and wives. American Journal of Epidemiology, 138, 237-242. Forste, R. (2002). Where are all the men? A conceptual analysis of the role of men in family formation. Journal of Family Issues, 23, 579-600. Francisco, A. (1996). Considerations for a two-sex demography: When, why and how should both sexes matter to demography? Unpublished doctoral thesis, Australian National University, Canberra. Furstenberg, F.F. (1988). Good dads—bad dads: Two faces of fatherhood. In A. Cherlin (Ed.), The changing American family and public policy (pp. 193-218). Washington, DC: Urban Institute Press. Gelman, A., & Hill, J. (2007). Data analysis using regression and multilevel/hierarchical models. New York, NY: Cambridge University Press. Goldscheider, F.K., & Kaufman, G. (1996). Fertility and commitment: Bringing men back in. Population and Development Review, 22(Suppl.), 87-99. Gray, E. (2002). What do we know about men’s fertility levels in Australia? People and Place, 10(4), 1-10. Greene, M.E., & Biddlecom, A.E. (2000). Absent and problematic men: Demographic accounts of male reproductive roles. Population and Development Review, 26, 81-115. Guzzo, K.B. & Furstenberg, F.F. (2007). Multipartnered fertility among American men. Demography, 44, 583-601. Hajnal, J. (1948). Some comments on Mr Karmel’s paper. Population Studies, 2, 325-360. Heard, G. (2006). Pronatalism under Howard. People and Place, 14(3), 12-25. Heard, G. (2010). Interpreting Australia’s fertility increase. People and Place, 18(2), 10-18. Hynes, K., Joyner, K., Peters, H.E., & Yang DeLeone, F. (2008). The transition to early fatherhood: National estimates based on multiple surveys. Demographic Research, 18(art. 12), 337375. Jackson, N. (2006). When is a baby boom not a baby boom? Nine points of caution when interpreting fertility trends. People and Place, 14(4), 1-13. Juby, H., & Le Bourdais, C. (1999). Where have all the children gone? Comparing mothers’ and fathers’ declarations in retrospective surveys. Canadian Studies in Population, 26, 1-20. Karmel, P.H. (1947). The relation between male and female reproduction rates. Population Studies, 1, 249-273.

278

ESTIMATING PATERNITY IN AUSTRALIA, 1976-2010 Lain, S.J., Ford, J.B., Raynes-Greenow, C.H., Hadfield, R.M., Simpson, J.M., Morris, J.M., et al. (2009). The impact of the baby bonus payment in New South Wales: Who is having “one for the country”? Medical Journal of Australia, 190, 238-241. Lappegárd, T., Rønsen, M., & Skrede, K. (2011). Fatherhood and fertility. Fathering, 9, 103120. Lattimore, R., & Pobke, C. (2008). Recent trends in Australian fertility. Productivity Commission Staff Working Paper. Canberra, Australia: The Productivity Commission. Manlove, J., Logan, C., Ikramullah, E., & Holcombe, E. (2008). Factors associated with multiple-partner fertility among fathers. Journal of Marriage and Family, 70, 536-548. Manning, W.D., & Smock, P.J. (2000). Swapping families: Serial parenting and economic support for children. Journal of Marriage and Family, 62, 111-122. Parr, N., & Guest, R. (2011). The contribution of increases in family benefits to Australia’s early 21st century fertility increase: An empirical analysis. Demographic Research, 25(art. 6), 214244. Poulain, M., Riandey, B., & Firdion, J.M. (1991). Enquête biographique et register Belge de population: Une confrontation des données [Biographical survey and register of Belgian population: A confrontation of data]. Population, 46, 65-87. Read, D., Crockett, J., & Watson, G. (2007). What’s behind recent fertility trends—Government policy, alarms on biological clocks or lessons learned from childhood? People and Place, 15(2), 22-29. Rendall, M.S., Clarke, L., Peters, H.E., Ranjit, N., & Verropoulou, G. (1999). Incomplete reporting of men’s fertility in the United States and Britain: A research note. Demography, 36, 135-144. Risse, L. (2010). ‘…And one for the country’: The effect of the baby bonus on Australian women’s childbearing intentions. Journal of Population Research, 27, 213-240. Ruzicka, L.T., & Caldwell, J.C. (1977). The end of the demographic transition in Australia. Australian Family Formation Project Monograph No. 5. Canberra, Australia: Department of Demography, Australian National University. Toulemon, L., & Lapierre-Adamcyk, E. (2000). Demographic patterns of motherhood and fatherhood in France. In C. Bledsoe, S. Lerner & J.I. Guyer (Eds.), Fertility and the male lifecycle in the era of fertility decline (pp. 293-330). New York, NY: Oxford University Press for the IUSSP.

279

Suggest Documents