• Share this page to Facebook
  • Share this page to Twitter
  • Share this page to Google+
Is balancing the sex of their children important to New Zealand parents?

Is balancing the sex of their children important to New Zealand parents? Evidence from recent birth data looks at whether New Zealand parents have preferences for sons or daughters, and whether the sex of previous children affects family size.

Are parents of two or three children of the same sex more likely than other couples to have an additional child, and do they do so in order to have at least one girl and one boy?

Our study, based on births between 2007 and 2015, found that parents with children all of the same sex are slightly more likely to have a third or fourth child than parents with at least one boy and one girl. This may have a minor effect on family size. We found, however, that parents did not have a preference for boys over girls or vice versa.


We first report what we found in the New Zealand data and then discuss some of the limitations in this analysis. Finally we point to some of the international research for people who want to pursue the topic further and suggest some aspects of the topic that require further work to add value to our understanding.

We used birth registration data to look at third births to mothers with two previous children and fourth births to mothers with three previous children. We look at these two groups separately.

Based on the sex of their existing children, the birth of a third or fourth child may indicate a desire of a mother and/or her partner for a larger family, or for a mixed-sex family, if we assume that births are planned. From the sex of previous children we may be able to infer any underlying preference for a male or female child.

If a preference for sons and/or daughters is valid then it may be reflected in the number of higher-order (third or fourth) births in New Zealand of families with all children of one sex when we compare these with other families. We found that the sex of existing children does not have a major impact on family size, implying that parents may not have preferences regarding the sex of their subsequent children.

On the other hand, we did find, for three- and four-child families, that New Zealand parents may have a small preference for having a child of each sex, but they may not prefer boys over girls, or vice versa. One possible interpretation of this is that the family size dominates unless for the family there is a desire for a balancing of the sexes, that is to say, to have at least one son and at least one daughter. The cumulative effect can have a minor effect on the total fertility levels.

Children and fertility rates

The number of births, in part, contributes to the age structure of a population now and into the future. Fertility rates are based on the number of births and population size, providing policy-makers with basic information for social and economic planning, such as demand for and location of schools, housing, and aged-care facilities. The Total Fertility Rate for New Zealand tells us that currently women on average have about two children. This does not mean that almost all women have two children because some women have more than two children and some women do not have any children.

The proportion of mothers having more than two children is an increasingly important component in a time when more women are remaining childless. Since the early 1980s, census data shows that the percentage of women who remained childless increased from 9 percent in 1981 to 16 percent in 2013 (Statistics NZ, 2016). If there were no net gain from migration, women who do have children would have to have more than two children each to stop the population from shrinking in the long term.

The sex of children as an incentive for subsequent childbearing is a topic that continues to fascinate people in similar low-fertility societies and societies with shared histories, from Australia (Gray & Evans, 2004; Kippen et al, 2007) to Europe (Schröder et al; 2016, Sævarsson, 2016).

We consider this question in the New Zealand context.

Effect of the sex of existing children on the progression to a third child

Figure 1 shows the percentage of third-child births according to the sex combination of the previous two births between 2007 and 2015.

Figure 1 

Between 2007 and 2015, mothers with two sons accounted for around 29 percent of all third births, while mothers with two daughter accounted for around 26 percent of all third births. This means that of all third births, around 55 percent were to mothers with two previous births of the same sex, compared with 45 percent of those with previously one boy and one girl child.

This suggests that mothers with same-sex children are more likely to have a third child than are mothers of a son and a daughter. The considerable volatility throughout this period is primarily due to the numbers being small. It is possible the gap is widening slightly in recent years, though the period covered is too short to definitely draw that conclusion.

What does this difference tell us about parents’ preferences? There was approximately the same number of same-sex two-child families and different-sex two-child families between 2007 and 2015, and hence we might therefore expect approximately the same number of children born from each of the two groups. Therefore, the difference (55 percent versus 45 percent) could possibly imply a consequence of preference or planning, though other factors such as the spacing and timing of births and the age of mother may play a part.

The small difference between the number of third children born to same-sex two-child families and different-sex two-child families suggests that New Zealand parents with two children of the same sex are slightly more likely to have a third child. This may imply a preference for at least one child of each sex in the resulting family of three children, but does not prove that this is so: we know neither the intention for having a third birth nor whether the sex of the eventual third child was what may have been preferred.

Are parents of two children are more likely to progress to a third child if they already have two boys rather than two girls? It is important not to interpret this necessarily as a preference because there are many factors affecting the sex of births.
Biological reasons dominate because the propensity to carry a child to term varies by sex of the foetus from mother to mother (Lazarus, J. 2002; Austad 2015; Orzack, 2016). Climate, life histories and environment also affect outcomes (West & Sheldon, 2002; Leimar, 1996; Helle et al 2008).

These other factors may be more important for the sex of higher parity births than the preferences of parents, though the overall effect remains small. Figure 1 shows an average difference of about 3 percent between third-child births from mothers with two boys and mothers with two girls.

There was a similar percentage difference in the number of two-boy versus two-girl families in this period, which in part reflects the tendency for more boys to be born than girls regardless of parity. The implication is that the birth of a third child in both these family types is not related to whether the previous children are boys or girls.

In other words there is no direct suggestion in the data of anything resembling a son-preference or daughter-preference. New Zealand parents with two female children are just as likely to progress to a third child as parents of two male children.


Effect of the sex of existing children on the progression to a fourth child

Families with four children are fewer than those with three children. It is also expected that the progression from three children to four children may be influenced by different parental considerations than those which applied to the third birth. In this section we report on parents with three previous children in relation to fourth births to those parents. If there are economic constraints on family size, the intention to stop at three children may be stronger than the intention to stop at two children. However there is the added complication that we do not know if births are in fact generally planned.

In contrast to third births where 45 percent had different-sex previous births and 55 percent same-sex previous births, around 31 percent of fourth births were to same-sex-child families and 69 percent were to families with previous boys and girls.

Figure 2 shows the percentage distribution of fourth births by the sex-combination of previous siblings for the 2007–15 period. We see that around 16 percent of all fourth births were to mothers who had 3 previous sons and around 14 percent were to mothers who had three previous daughters. Among mothers with existing families of at least one son and one daughter, each of the six possible combinations of sons and daughters accounted for around 11–12 percent of the total. This means that around 30 percent of the fourths births were to same-sex-child families and 70 percent to families that already had children of each sex.

Figure 2 

If there were no bias, the proportion of fourth births to same-sex-children families should be similar to the proportion of these families among three-child-families.

When we analysed the odds of a three-child family having a fourth child we found that a family with three children of the same sex was around 4 percent more likely to have a fourth child than one that already included at least one son and one daughter. We did not adjust for whether mothers of three previous children of the same sex were more or less likely to carry a fourth child of that same sex to term, as we assumed this effect is small relative to other factors.

Thus, we concluded that the 4 percent more births could have come from preference or planning, though this cannot exclude other environmental and biological factors.

Is this conclusion valid, and can this then be extrapolated to suggest a gender preference of the fourth child? We found that the difference between the number of fourth children born to same-sex three-child families and different-sex three-child families is significant. Therefore it is possible that New Zealand parents with three children may have a preference for at least one child of each sex in the resulting family of four children, but without knowledge of the relationship between the intentions and the outcomes we cannot assume this to be the case.

For parents of same-sex-children families, there may be a different probability of having a fourth children based on the sex of their children. We looked at the birth outcomes of mothers of boys compared with mothers of girls, to see whether mothers with three boys were more likely to have a fourth child than mothers with three girls. Figure 2 shows that across the study period there was an average difference of 1.6 percent between fourth-child births from mothers with three boys and mothers with three girls. In 2015, there were 2.6 percent more fourth-child births from mothers with three boys than mothers with three girls.

Does this difference mean that parents prefer girls over boys? More boys were born than girls: there are about 2 percent more three-boy families than three-girl families in New Zealand between 2007 and 2015. This difference alone can explain the 1.6 percent difference in the birth numbers.

The difference between the number of fourth children born to three-boy families and to three-girl families is therefore not statistically significant, providing no evidence that New Zealand parents prefer one sex over the other for a family of four children. A fourth child will be (generally) either a boy or a girl and outcomes do not indicate preference. This difference, which is not statistically significant, merely tells us that in a three-child family the probability of three boys is higher than the probability of three girls.


Some issues and conflicts in empirical fertility studies remain difficult to solve, but we have attempted to use the most appropriate data and method to answer the question of the relationship between the sex of children and parity progression in New Zealand.

Birth registration and census data can together shed light on child-bearing behaviour.

Census data provides a broader contextual frame than birth registrations and shows that, in 2013, 95 percent of women all women aged 15 years and over had borne fewer than five children. On average all women had had 1.8 children. This figure includes the 31 percent of women who had not, yet, had children. Women who had children, on average, had given birth to 2.6 children. Children grow up and leave home so that 98 percent of families with children had fewer than five children at home in the 2013 census.

Census questions have the advantage of measuring all childbearing as well as childlessness and delayed childbearing. However, mothers may have given birth to children who have since left home or who have since died, so information is not available for complete families in all cases. Moreover, the number of children in the family may equal the number of children a woman has had, but these children may not be the biological children of the current partnership. For example, they may be children of her current partner from previous relationships, or her own off-spring from previous relationships, or the family may be a single-parent family.

The census data therefore tells us a different part of the story about higher parity childbearing because it is essentially centred on the mother rather than on a specific couple. Her current family situation adds context to this. Census does however, provide a much richer picture of the environment of the mother and her children than birth registrations are able to.

New Zealand censuses asks about the number of children that women aged 15 years and over have given birth to. The Australian census asks a similar question. This has been studied by Rebecca Kippen and colleagues (Kippen et al. 2005, 2007). Their study explored the implied impact of the sex of existing children on parents' fertility decisions, using census data that asked about the number of children born alive to a woman.

For our purpose, birth registration data has some advantages over census data in the New Zealand context. Birth registration data provides the sex of the newborn and the number and sex of the siblings from the same parent relationship.
Moreover, birth registration data provides almost full coverage of births in New Zealand over the study period. Additionally, birth registration data, unlike surveys of attitude or intention, show parents' actual childbearing.

The birth data we used is based on live births registered in New Zealand to mothers resident in New Zealand and to mothers resident overseas between 2007 and 2015. Late registrations (those registered more than two years after the birth occurred) are included. These data are different from most standard statistical tables published by Stats NZ, which exclude late registrations and births to mothers resident overseas.

The data records both birth order and sex of previous children. We chose families with either three or four children. We did not look at larger families. This was because there were only a small number of larger families and our focus was only on the third and fourth births in the birth registration data between 2007 and 2015. To include larger families, we would have had to extract from the subsequent births the available information about previous issue – but many of these births would have occurred outside our window of interest and would have complicated the analysis for very little benefit given that we know nothing beyond the sex and date of birth of that child (eg, we did not know about the characteristics of the mother at the time of the earlier births).

In the years 2007 and 2015, there were 562,755 live births registered in New Zealand. More boys are born than girls, with 289,269 baby boys and 273,486 baby girls during this period. Among these there were 65,433 third births, of which 35,901 were to mothers with either two previous boys or two previous girls. In contrast to this, 6,657 of the 21,816 fourth births were to mothers whose first three children were of the same sex. Table 1 shows that there were in both cases more new babies to families with all male siblings than all female siblings.

Table 1

 Third and fourth births by sex of siblings, birth registrations 2007–15

 Sex of siblings

 Third births

Fourth births 

 Same sex siblings

Different sex siblings 

 Same sex siblings

Different sex siblings 
















Note: Data has been randomly rounded to base 3, to protect confidentiality.
Source: Stats NZ

We looked into whether or not New Zealand parents prefer boys rather than girls, or vice versa, and if they prefer having at least one child of each sex. In order to answer this question, we asked:

  • Are parents with two or three children of the same sex more likely to have an additional child than parents with two or three children of different sexes? 
  • Are parents with two or three sons more likely to have an additional child than parents who have two or three daughters?

This research was placed within some of the prevailing theories of childbearing. We used Hoffman and Hoffman's (1973) 'value-of-children' theory for developed societies as our testing foundation. According to the value of children theory, parents in more advanced societies prefer a mix of sons and daughters for their respective benefits, such as companionship and stimulation. If this theory is correct, we could assume that a third or fourth birth from a same-sex two- or three-child family might come from the desire of having one of each sex in the family (Blake, 1981, 1994). This would of course not guarantee the outcome of a child of the same or different sex as the existing children.

According to Kippen et al. (2007), three methods are commonly used internationally to examine if parents have a particular gender preference:

  • records of sex-selective abortion 
  • focused surveys, and 
  • analysis of mothers progressing to another birth based on the sex of existing children.

The first not only lacks data in New Zealand but sex-selective abortion is not legal in this country. The second option, of using specialised survey data, would be the ideal, but no recent surveys are available in New Zealand. As with the Australian studies, we used the third, most common, method of analysing families' progression to another birth based on the sex of existing children, using birth registration data. This then gives us an indication, by inference, of the parents' preference for the sex of their new baby.

Specifically, we examined the sex of existing children in a family and the impact of this on third and fourth births. We identified if there was a significant percentage difference between the number of third and fourth children born to same-sex-children families and different-sex-children families to examine if parents have a preference for one of each sex; and we identified if there was a significant percentage difference between the number of third and fourth children born to all-boy families and all-girl families, to examine if parents prefer boys over girls or vice versa.

Before carrying out the comparison, we had to readjust the percentages of birth for all family groups (same-sex-children families, different-sex-children families, all-boy families, and all-girl families) to accommodate the average sex ratio at birth. The world's average sex ratio is 107 boys to 100 girls (World Bank, 2011), though the underlying causes and local variations remain poorly understood (Austad, 2015; Orzack, 2016). In New Zealand, between 2007 and 2015, the average ratio was 106 boys to 100 girls, though this varied by age, parity, and geographic region. For this article, we used a sex ratio of 106 males to 100 females. Because more boys were born than girls, we expected more all-boy than all-girl families, and more different-sex than same-sex families. As a consequence, there would be more births from all-boy and different-sex families.

To avoid overstating the differences between groups and therefore over-implying knowledge of causality as to parental preferences, we adjusted for the known sex ratio of births so that we could compare boy-families and girl-families fairly.


How New Zealand compares with other countries

Theories and empirical findings

The literature on sex ratios, sex preferences of children, and parity progression is very large indeed. Some previous theories and empirical studies have focused on parents' preferences for sex of off-spring in developing nations (Hank & Kohler, 2000, 2002), but have relevance too to developed countries.

Economic net utility versus care and emotional support

The main theories suggested that the preferences (for boys or girls) are the result or reflection of the economic, cultural, social, and religious needs in some less developed societies (eg, Caldwell, 1982; Arnold, 1997). In traditional societies, children still provide positive economic net utility. For example, boys are preferred because of their physical strength in agricultural activities and social security, such as in Nepal (Niraula & Morgan, 1996). Daughters may represent an economic investment (Tertilt, 2005). In other societies, daughters are preferred because they are considered to be better at providing care and emotional support (Arnold, 1997; Kağitçibaşi, 1982).

However, these two seemingly plausible theories may not apply in the same way in low fertility regimes and societies with greater gender equity, where children are more likely to lead to time and financial cost (Hoffman &Hoffman, 1973; Hank & Kohler, 2000, 2002; Weston, et al, 2004). This change in children's 'value' as a society develops might partly explain the low fertility rates in OECD countries, which in 2014 was an average of 1.68 children per woman (OECD, 2016).

Value of children

So what drives fertility rates in modern societies and does gender preference change as a result? According to Namboodiri (1972), fertility decisions were sequential in developed countries, instead of being driven by total utility maximisation in traditional nations. In other words, parents decide first on whether to have children or not; then whether to have one child or two children in the family; and lastly whether or not to attempt to achieve a certain gender mix (Rindfuss, et al, 1988; Bulatao, 1981; Blake 1981, 1994).

The 'value of children' paradigm continues to underpin much fertility literature. Hoffman and Hoffman (1973) considered the 'value-of-children' and argued that children also bring intangible benefits, such as expansion of the self, affiliation, accomplishment, social comparison, and lastly economic utility. Hoffman and Manis (1979) found empirical support for the theory. Their survey found that in America the top six reasons for having children were psychological (eg, group ties and affection, stimulation and fun, and expansion of the self); while economic utility and security in old age ranked the last of all reasons to have children.

The value-of-children theory suggests that parents in modern societies may desire a gender mix because of the different benefits that are attached to each sex. That is, each parent might prefer to have at least one child of his or her own (opposite) sex for the purpose of companionship. For example, fathers could, particularly in societies where gender-role equity is less advanced and sexism remains a feature, prefer a son to play certain sports with or a daughter to cook with; while mothers might prefer a daughter to shop with or a son to play certain sports with.

Empirical studies of modern societies have found evidence for such a preference of having at least one child of each sex. Hank and Kohler (2000) found that 11 of the 17 European countries had a preference of having one of each sex. Finland, France, West Germany, Norway, Poland, and Portugal, all countries noted for gender equality, had no preference.

Kippen et al. (2005) showed that the sex of existing children was a significant factor for the number of children in Australian families. Specifically, a two-son or two-daughter family was more likely to have a third child than a family with one son and one daughter. Similarly, a family with three boys or three girls was more likely to have a fourth child than a family with at least one child of each sex. However, Kippen et al. (2005) did not find evidence that parents prefer boys over girls or vice versa.

International research has shown that many factors modify behaviours and outcomes in this arena. Murphy (1992) found that couples who had two breadwinners in the family were less likely to have a third child in Sweden; while Hoem (1993) found Swedish parents with higher education, unlike experiences in some other countries, were more likely to proceed to a third child.

Another example of ambiguous evidence can be found in Brockmann (2001) versus Hank and Kohler (2000). Brockmann (2001) found that West German women in their sample survey did not have a gender preference for children; while Hank and Kohler (2000) found they had a preference for at least one boy and at least one girl.

Yamaguchi and Ferguson (1995) reported an increasing effect of sex-of-previous-children with parents' education level, but looking at more recent emerging trends. Pollard and Morgan (2002) suggested a declining impact of previous children's sex post 1985 when education levels were increasing. This may have some relevance to New Zealand where there is a continuing shift in New Zealand in the gender balance of enrolments in tertiary education, especially where one parent dominates in the decision-making process and there is counterpointing between this trend and changing cultural and social considerations.

International comparison: data and methods

There are two types of data and method combinations used in many of the international empirical studies of fertility.

Determinants of respondents' attitude towards fertility

First, data from surveys of respondents' attitude towards fertility collected together with information about determinants (eg, age, income, and education level) may be combined with modelling the relationship between fertility and its determinants. Survey data provides information about attitudes and behaviours. Brockmann (2001) and Hank and Kohler (2002), for example, surveyed German parents' attitude towards fertility decisions to understand the determinants of fertility. They then modelled and studied the relationship between the determinants and fertility rates. Similar data is available for New Zealand from the 1995 New Zealand Women: Family Employment and Education Survey (Pool et al, 2007), though this has not yet been repeated, so comparable contemporary information is not available for this country.

The limitations common to studies of this type of data and method combination are three-fold:

  • There is inherit sampling error. 
  • There is no direct linkage between attitude and behaviour, that is, no guarantee that the respondents would actually carry out their intention. In fact, Hank and Kohler (2002) found, in their German study, that despite the preference for at least one son and one daughter in the family, it did not affect their fertility behaviour. 
  • Furthermore, studies that apply models to explain fertility rates and its determinants are heavily influenced by the underlying theory and to some extent the researchers' focus.

The apparently conflicting results by Brockmann (2001) and Hank and Kohler (2002) are mainly derived from the focus on different and contrary causal factors. While Brockmann (2001) examined the relationship between post-war welfare policies and fertility behaviour, Hank and Kohler (2002) considered social-cultural factors to be the main determinants of Germany's fertility rates. As in New Zealand, both sets of determinants play a role. There is no simple right or wrong complete answer.

Indirect statistical measure and direct comparison of data sources

The second data and method combination is indirect statistical measure and direct comparison of the data sources. Indirect statistical measures come from sources that are not designed to explore fertility rates and its determinants but nevertheless provide information for fertility research. Data derived from fertility-related questions in a population census are examples of indirect statistical measures.

This combination of data selection and method has some advantages. Census data is almost free of sampling-error and direct comparison of the statistics is not bound by theory or researchers' personal beliefs. However, it is not without problems. The potential scope of a study depends largely on the fertility question asked as well as other information collected in the census. Data selection still can have unexpected outcomes on the analysis. For example, Griffith et al. (1985) argued that the sex-of-previous-children effect is only relevant for partnered parents in their original families and where the coded census family has the same number of children at home as have been born to the current mother. This approach excludes much of the valuable information available in census about the wider picture of childbearing and parity progression.


Further research

Is balancing the sex of their children important to New Zealand parents? is a brief initial analysis. We inferred parents' preferences for the sex of their children by examining the relationship between the sex of a child and the sex combination of the existing children in the family.

Parents, in general, are unable to predetermine the sex of a child. This does not enable us to gain a complete picture of how families make such decisions, because there are a number of other determinants that can affect both the outcome of a subsequent child and the sex of that child.

In order to deliver meaningful insights, further research could include parents' age, education level, financial situation, biological determinants of the sex of a baby, and access to health care. This would require access to more comprehensive data than birth registration provides.

Further substantive work will add to our knowledge about parity progression and the drivers behind childbearing. The data sources used here provide a fertile field for more extensive research. Census data adds several further elements to the picture, and linked census data could provide a longitudinal component that would enable an analysis of changing childbearing patterns over time.

Many questions, however, will need to await new information sources. The Integrated Data Infrastructure could provide more data about fertility decisions and behaviour. Other information would require a specialised survey to capture additional relevant data.


Arnold, F. (1997). Gender preference for children. Demographic and health surveys comparative studies No.23.

Austad, S. (2015). The human pre-natal sex ratio: A major surprise. Proceedings of the National Academy of Sciences of the United States of America, 112, 4839–4840.

Blake, J. (1981). The only child in America: Prejudice versus performance. Population and development review, 7, 43–54.

Blake, J. (1994). Judith Blake on fertility control and the problem of voluntarism. Population and development review, 20, 167–177.

Brockmann, H. (2001). Girls preferred? Changing patterns of gender preference in the two German states. European sociological review, 17, 189–202.

Bulatao, R. A. (1981). Value and disvalues of children in successive childbearing decisions. Demography, 18, 11–25.

Caldwell, J. C. (1982). Theory of fertility decline, only child in America: Prejudice versus performance. London: Academic Press.

Gray, E. & Evans, A (2004, September). Parity progression in Australia: What role does sex of existing children play? Paper presented to Australian Population Association Conference, Canberra.

Griffith, J. D., Koo, H. P, & Suchindran, C. M. (1985). Childbearing and family in remarriage. Demography, 22, 73–88.

Hank, K., & Kohler, H.-P. (2000). Gender preferences for children in Europe: Empirical results from 17 FFS countries. Demographic Research, 12(1), 11–25.

Hank, K., & Kohler, H.-P. (2002). Gender preferences for children revisited: New evidence from Germany. Max Planck Institute for Demographic Research, Working paper 2002–017. Max Planck Institute, Munich.

Helle, S. Helama S., & Jokela J. (2008) Temperature related birth sex ratio bias in historical Sami: Warm years bring more sons. Biology letters, 4, 60–62.

Hoem, B. (1993). The compatibility of employment and childbearing in contemporary Sweden. Acta Sociologica, 36, 101–120.

Hoffman, L. W., & Hoffman, M. L. (1973). The value of children to the parents. In Fawcett, J. T. (Eds.). Psychological perspectives on population. New York: Basic Books, pp 19–76.

Hoffman, L. W., & Manis, J. D. (1979). The value of children in the United States: A new approach to the study of fertility. Journal of marriage and the family, 41, 583&ndash596.

Kağitçibaşi, Ç. (1982). Old-age security value of children: Cross-national socio-economic evidence. Journal of cross-cultural psychology, 13(1), 29–&ndash42.

Kippen, R, Evans, A., & Gray, E. (2007). Parental preference for sons and daughters in a western industrial setting: Evidence and implications. Journal of Biosocial Science, 39(4), 583&ndash97.

Kippen, R., Gray, E, & Evans, A. (2005). The impact on Australian fertility of wanting one of each. People and place, 13(2), 12&ndash20.

Lazarus, J. (2002). Human sex ratios: Adaptations and mechanisms, problems and prospects. In I. Hardy (Ed) Sex ratios: Concepts and research methods. Cambridge: Cambridge University Press. pp 287&ndash311.

Leimar, O. (1996) Life-history analysis of the Trivers and Willard sex-ratio problem. Journal of Behavioural Ecology, 7, 316&ndash325.

Murphy, M. (1992). Economic models of fertility in post-war Britain &ndash a conceptual and statistical re-interpretation. Population studies, 46, 235&ndash258.

Namboodiri, N. K. (1972). Some observations on the economic framework for fertility analysis. Population studies, 26, 185&ndash206.

Niraula, B., & Morgan, P. (1996). Marriage formation, post-marital contact with natal kin and autonomy of women: Evidence from two Nepali settings. Population studies, 50(1), 35&ndash50.

OECD (2016). Fertility rates (indicator). doi: 10.1787/8272fb01-en . Accessed on 17 November 2016.

Orzack, S. (2016). Old and new ideas about the human sex ratio. Significance, February 2016, 25–27.

Pool, I., Dharmalingam, A., & Skeats J. (2007). The New Zealand family from 1840. Auckland: Auckland University Press.

Pollard, M. S., & Morgan, S. P. (2002). Emerging parental gender indifference? Sex composition of children and the third birth. American Sociological review, 67(4), 600–613.

Rindfuss, R. R., Morgan, S. P., & Swicegood, G. (1988). First birth in America: Changes in the timing of parenthood. Berkley, CA: University of California Press.

Schröder, J., Schmiederberg, C., & Brüderl, J. (2016). Beyond the two-child family: Factors affecting second and third birth rates in West Germany. Zeitschrift für Familienforschung, 28(1), 3–18.

Statistics New Zealand. (2016). More women are remaining childless. Retrieved 10 October 2016 from www.stats.govt.nz.

Sævarsson, S. (2016). Children's gender and parents' subsequent fertility and partnerships in Iceland. Master of Science thesis in Economics, University of Iceland, Reykjavik.

Tertilt, M. (2005). Polygyny, fertility and savings. Journal of political economy, 113(6), 1341–1371.

West, S., & Sheldon, B. (2002). Constraints in the evolution of sex ratio adjustment. Science, 295, 1685–1688.

Weston, R., Qu, L., Parker, R., & Alexander, M. (2004). It' not for lack of wanting kids. A report on the fertility decision making project. Melbourne: Australian Institute of Family Studies.

World Bank. (2011). Gender statistics. Retrieved 10 October 2016 from http://databank.worldbank.org.

Yamaguchi, K., & Ferguson, L. R. (1995). The stopping and spacing of childbirths and their birth-history predictors: Rational-choice theory and event-history analysis. American Sociological Review, 60, 272–298.

Citation for this article
Stats NZ (2017). Is balancing the sex of their children important to New Zealand parents? Evidence from recent birth data. Retrieved from www.stats.govt.nz.

ISBN 978-1-98-852814-4 (online)
Published 21 July 2017

  • Share this page to Facebook
  • Share this page to Twitter
  • Share this page to Google+
  • Share this page to Facebook
  • Share this page to Twitter
  • Share this page to Google+