Impact of Chinese one-child policy on sibling structure: experience from rural areas in three provinces.

OBJECTIVES: To study the influence of population policy and boy preference on the sibling structure, that is, in which order and combinations boys and girls are born into families.
DESIGN: A population-based survey with a representative sample of new mothers in 2008-2009 in rural China.
SETTING: Two provinces (Anhui and Shaanxi) where authorisation for a second child was usually given if the first birth was a daughter and one province (Chongqing) where only one child was authorised. PARTICIPANTS: The mothers giving birth in 2008-2009 were identified from family planning and hospital birth registers (including births outside hospitals) (n=5049). Of them, 3673 were interviewed by trained medical university staff members and students using structured questionnaire (response rate 73%). MAIN OUTCOME MEASURES: Children's distribution by sex and families' distribution by children's birth order and sex composition were calculated and compared with theoretical values based on the assumption that family planning policy is fully followed.
RESULTS: The recommended family policy was varyingly followed in the three provinces. In all provinces, there were more second children than allowed. If the policy allowing a second child only after a first-born girl were fully followed, it would result in a sibling structure in which the one-child family is always with a boy and in the two-child family the first one is always a girl. This sibling structure was partly seen in Anhui but weakly in Shaanxi. The policy allowing only one child would result in an equal number of boys and girls, but in Chongqing, there were more boys. In Anhui, unlike the other provinces, there were many more first-born girls than boys, which the authors could not fully explain.
CONCLUSION: Population policy and boy preference influence the actual and relative number of girls and boys and also sibling structure.

Introduction

Previous studies have shown that many more boys than girls are born in China,1–3 and also elsewhere in Asia,4 5 being far more than suggested by the normal ratio (about 1.02 to 1.06). This is due to a strong preference for boys in a situation when either government regulations or a family's own preferences restrict the number of children to one or two. The easy availability of ultrasound and sex-selective abortions is apparently the current main method to guarantee that at least one son is born.3 5 It has been considered that the bias in the sex ratio, besides being a difficult ethical problem, also entails notable social problems when sons grow up and cannot find wives.6

The previous literature has focused on the number of boys and girls. But boy preference and population policy rules may also impact the sibling structure, that is, in which order and combinations boys and girls are born in families. We found no previous studies on sibling structure in modern China. One study found that in several countries in Asia (not including China) and North Africa, girls are more likely to be born into larger families and at earlier parities; these patterns were absent in sub-Saharan countries.7 Sibling structure has been of interest in educational and psychological research but not so much in health or sociological research.2 8–13 However, girls being born into larger families and as older siblings may mean less resources and more family responsibilities and less schooling for girls, both of which influence health.7

The purpose of this study is to investigate the influence of population policy on the sibling structure, that is, in which order and combinations boys and girls are born into families. We report findings from rural areas in three Chinese provinces, based on surveys of new mothers in 2008–2009. In China, the governmental regulations for the number of permissible children vary locally. In two of the study provinces (Anhui and Shaanxi), the usual interpretation of family planning policy was that the second child was authorised (with a 4-year interval) if the first birth was a daughter. In the third province (Chongqing), only one child was authorised. For minorities, the regulations were more relaxed, although they were few in number in the study areas. Anhui province was the wealthiest (per capita annual income US$ 815 in 2008), followed by Chongqing (US$ 758), while Shaanxi was the poorest (US$ 443).14 In Anhui and Chongqing one-third and in Shaanxi one-quarter of the women who had given birth in 2008 were rural-to-urban migrants.

Methods

Theoretical models

If no sex-selective abortions were made, and all families followed the family planning rules and wanted and could have two children (when allowed), and the natural male dominance at birth is ignored, the theoretical sibling composition would be as given in tables 1–3. In Anhui and Shaanxi, once childbearing is completed, for 100 families, there would be150 children, with equal numbers of girls and boys (75 each). Half of the families (50) would have only one child, all of them boys. The other half would have two children, the first-born always a girl and the second born either a boy (25 families) or a girl (25 families). This would result in a society in which single children would always be boys, and for those with siblings, the oldest one would always be a girl. In Chongqing (table 2), all families would have only one child, with equal numbers of girls and boys.

Table 2

Theoretical number (%) of children in 100 families*, if the family planning policy were fully followed, Chongqing study area

	First child		Total
	Girl	Boy
One child	50 (50)	50 (50)
Second child	–	–	100 (100)

This is also the family structure.

Data and analysis

We carried out a survey on a representative sample of new mothers in 2008–2009 in rural China as part of an evaluation of interventions to increase the use and to improve the quality of maternity care.14 15 The survey was made in five counties and utilised cluster random sampling, yielding 5049 women who had given birth between March 2008 and March 2009. We asked about the gender of all children they had given birth to; the experience thus comes mainly from the 2000s. The new mothers were identified from family planning registers in Anhui and hospital birth registers (including births both in and outside hospitals) in Chongqing and Shaanxi.14 Family planning workers or village doctors contacted the mothers by telephone or by home visit before the survey. Due to the data collection method, the likelihood of including mothers with illegal births into the survey sample is unknown in Anhui.

Ethical approval for the study was obtained from the International Centre for Reproductive Health, Ghent University (2008/143). The mothers voluntarily participated in the study. Using a structured questionnaire, trained medical university staff or students interviewed participants at home or in public places. Overall, 73% of mothers (3673 of 5049) completed the interview. The main reasons for non-response were no-one at home at the time of survey (10%) and no or incomplete contact information (7%); 2% refused the interview.

Cross-tabulation was used to describe the sex of the two first-born children (including families having three or more children). Children's distribution by sex and families' distribution by children's birth order and sex composition were calculated and compared with the theoretical numbers in tables 1–3. SAS V.9.1 was used to analyse the data.

Results

Number and sex of children of the first-born child

The actual number of children by the sex is given in table 4. In Anhui, the first-born child was more often a girl (sex ratio 0.69) and the second-born child a boy (sex ratio 1.14). Also in Shaanxi, there were more first-born girls and more second-born boys, but the difference for first-born was smaller (sex ratio 0.93) and for second born larger (sex ratio 1.27).

Table 4

Observed number (%) of children, by child's sex and province

	Anhui		Shaanxi		Chongqing
	Girl	Boy	Girl	Boy	Girl	Boy
First child	896 (59.1)	619 (40.9)	837 (51.9)	776 (48.1)	247 (45.6)	295 (54.4)
Second child	227 (46.7)	259 (53.3)	297 (44.0)	378 (56.0)	94 (44.3)	118 (55.7)
Total	1123 (56.1)	878 (43.9)	1134 (49.6)	1154 (50.4)	341 (45.2)	413 (54.8)

Because childbearing was not yet completed, we could not test the theoretical model of number of children in Anhui and Shaanxi (table 1). However, the experience this far suggests that the theoretical model of the number of children (table 1) will be only partly followed.

Table 1

Theoretical number (%) of children in 100 families, if the family planning policy were fully followed*, Anhui and Shaanxi study areas

	First child		Total
	Girl	Boy
One child	50 (33.3)	50 (33.3)
Second child
Girl	25 (16.7)	–
Boy	25 (16.7)	–
Total	100 (66.6)	50 (33.3)	150 (100)

All families with first-born girl want and can have two children.

In Chongqing, there were relatively more first-born boys (sex ratio 1.19) and many second children. More of the second children were boys (sex ratio 1.26).

Number of families by the sibling structure

In Anhui, of those families who had at least two children, most of the first-borns were girls, as suggested by the theoretical model (table 5). But 14% of the families, rather than 0%, had a boy as the first-born. Of those families who had only one child, somewhat more had a boy. However, as many of these families may not have completed childbearing, we cannot judge what the final composition of one-child families will be.

Table 5

Observed number (%) of families by the number and order of children and by sex, Anhui (1515 families)

	First child			Second child*
	Girl	Boy	Total	Girl	Boy	Total
1 child	476 (46.2)	554 (53.8)	1030 (100)	–	–
2+ children	420 (86.6)	65 (13.4)	485 (100)	223 (46.7)	255 (53.4)	478 (100)
Total	896 (59.1)	619 (41.1)	1515 (100)

Data were missing for seven children.

In Shaanxi, of those families who had at least two children, the first-born was a boy in many (40%) (table 6). However, the number of first boy families was still less than the number of first girl families.

Table 6

Observed number (%) of families by the number and order of children and by sex, Shaanxi (1613 families)

	First child*			Second child
	Girl	Boy	Total	Girl	Boy	Total
1 child	429 (45.7)	509 (54.3)	938 (100)	–	–
2+ children	408 (60.4)	267 (39.6)	675 (100)	297 (44.0)	378 (56.0)	675 (100)
Total	837 (51.9)	776 (48.1)	1613 (100)

Data were missing for one child.

In Chongqing, there was little indication of the theoretical model of only one child: 38% of the families had more than one child (table 7). There were also many two-child families in which the first-born was a boy.

Table 7

Observed number (%) of families by the number and order of children and by sex, Chongqing (542 families)

	First child			Second child*
	Girl	Boy	Total	Girl	Boy	Total
1 child	147 (44.7)	182 (55.3)	329 (100)	–	–
2+ children	100 (46.9)	113 (53.1)	213 (100)	94 (44.3)	118 (55.7)	212 (100)
Total	247 (45.6)	295 (54.4)	542 (100)

Data were missing for one child.

Combining the three provinces, of those families having two or more children, the first-born was a girl in 68% and a boy in 32% (data not shown). The latter figure would be 0% in the theoretical model.

Discussion

Our results from these three rural areas show that the recommended family policy was varyingly followed in different areas. If the policy allowing a second child only after a first-born girl were fully followed, it would result in a sibling structure in which the one-child family is always with a boy and in the two-child family the first one is always a girl. This sibling structure was seen, even though not in pure form, in Anhui province. It was weakly reflected in Shaanxi. In the policy allowing only one child, no such pattern would be seen, but an equal number of boys and girls in one-child families. However, our results from Chongqing showed somewhat more boys and many second-born children. Furthermore, our study found the well-known phenomenon of more boys than girls among second children in Anhui and Shaanxi.

A finding that could not be explained by the recommended family planning policy or sex-selective abortions was the extra number of first-born girls in Anhui and Shaanxi provinces, particularly in Anhui. We do not know the reason for the finding, but in theory, it could be explained by the following factors. First, women do selective abortions for boys in the first pregnancy to have a girl first, as this allows them to have another child later. Second, the mothers may have hidden their first-born boys to allow them to have another child. Third, they may have given wrong information during the interview if they suspected that the researchers communicate with family planning officials. Fourth, mothers with first-born boys were more often away from home at the time of interview than mothers with girls. Fifth, notably more girls are born as a result of environmental contaminants. All factors may have contributed, but we think that the fourth explanation most likely. Our Chinese collaborators had not heard of the first and second explanations. The third explanation is unlikely in the interview situation. But when women were approached to agree about the interview, they had time to think and disappear from home if they did not want to be interviewed on their childbearing which they knew was against the rules. However, we wonder why this did not occur in Chongqing too. Environmental contamination has been proposed, but we know no studies from China supporting an impact of that size.

Anhui has traditionally had a high sex ratio favouring males.16 It is also the richest and most developed of our study areas. Shaanxi is the poorest and has many remote areas. Thus, Anhui results may be indicative of future trends elsewhere in rural China.

Can we trust our results?

All data were based on mothers' self-reports. There is no memory bias on children's sex, but two other biases may have existed: first, those having illegal unauthorised births may have not wanted to tell the number and sex of their previous children correctly. Second, a mother not being at home during the interview or the interviewers not having the right address may have been dependent on the number and sex of children.

We did not analyse for minority status or other reasons that allowed for exceptions of the recommended policy. Considering the small number of minorities in our study areas, ignoring minorities should not introduce bias. Our theoretical model assumed that no sex-selective abortions occur, but the actual numbers and gender of children showed that in all provinces, they must have been common. Many mothers had not completed their families, and a more accurate picture would be obtained from older mothers, though in the latter case, the current or future situation would not be as apparent: Chinese childbearing habits are changing.

More than one child

Because of the commonness of sex-selective abortions and the rule allowing another child if the first one is a girl, the Chinese population policy is sometimes called ‘one-boy policy’. However, the number of families having a second child after the first-born boy suggests that not only the sex but also the number is important. Not simply boys but children in general are valued.

If the finding of more girls in Anhui and Shaanxi is true and women use selective abortions to have a girl first, this is a very interesting finding, indicating how people can find ways of circumventing regulations to fulfil their own wishes. It is possible that to some families, the number of children is more important than the gender of the child.

Sibling structure

There was some indication that the family planning policy in China has created a new sibling structure, and possibly more so in the future, if people more fully accept the policy. If so, is that important? The literature has focused on being one child2 10–13 17 and much less is known of the impact of the order of siblings by sex. Generally, relationships between siblings offer children unique opportunities for learning about self and other and have an effect on children's development and well-being.8 It is known that parental favouritism or differential treatment can contribute to a negative relationship between siblings.9

Basu and de Jong7 found that in some Asian (China was not included) and North African countries, girls are born into larger families and that they have to share resources. This may put them at a disadvantage even when parents do not discriminate against girls. Furthermore, they suggested that the fact that girls are born at relatively earlier parities might also work to their disadvantage. In poorer and larger families, parental responsibility for younger children is passed on to older girls, who may not be able to devote their full time and energy to their own education or recreational activities. However, this is not very likely in the current Chinese culture or two-child families. The new patterns of sibling structure may not have any health implications, but they are likely to influence family dynamics.

Table 3

Family structure in 100 families, if the family planning policy were fully followed*, Anhui and Shaanxi study areas, n (%) of families

	First child			Second child
	Girl	Boy	Total	Girl	Boy	Total
1 Child	–	50 (100)	50 (100)	–	–	–
2+ Children	50 (100)	–	50 (100)	25 (50)	25 (50)	50 (100)
Total	50 (50)	50 (50)	100 (100)

And all families with first-born girl want and can have two children.