China's 1990 national population census revealed that the sex ratio at birth in 1989 was 113.8 (SSB, 1991). Because this was much higher than what is considered a normal sex ratio at birth at 107, the Government and others became concerned about the reasons for this phenomenon. Subsequently a number of studies were undertaken both within and outside China to assess the situation (Hull, 1990; Johansson and Nygren, 1991; Xu and Guo, 1991; Gu and Peng, 1991; Banister, 1992a; Wen, 1992). In October 1992, the sex ratio at birth became the de facto focus of discussions at the "International Seminar on China's 1990 Population Census" held at Beijing, with presentations by scholars from a number of countries (Banister, 1992b; Coale, 1992; Hull, 1992; Johansson and Arvidsson, 1992; Li, 1992; Tu and Liang, 1992; Zeng and others, 1992). In November 1994, the United Nations Population Fund (UNFPA) sponsored the "International Symposium on Issues Related to Sex Preference for Children in the Rapidly Changing Demographic Dynamics in Asia", which was held at Seoul. This Symposium brought together scholars from a number of Asian countries to address the various aspects of the issue and "to make a contribution to the understanding of the complex processes of rapid fertility decline, sex preference for children and adaptations to social and cultural change" (Singh, 1994; Nizamuddin, 1994; Roy, 1994). At the Symposium, the situation in eight Asian countries, namely Bangladesh, China, India, Indonesia, Pakistan, Republic of Korea, Sri Lanka and Thailand, was reviewed and discussed.1

All these activities have greatly improved knowledge and understanding of this issue. Using available statistical information and empirical research findings, this article intends to provide a wrap-up of the current status of knowledge concerning the abnormal sex ratio at birth in China in terms of (a) when it occurs, (b) where it occurs, (c) among whom it occurs, (d) how it occurs, (e) why it occurs, and (f) what can be done about it. While the focus of the discussion is on China itself, for comparative purposes the article also refers to Taiwan Province of China and the Republic of Korea, since the phenomenon of "missing girls" is observed not only in China but also in other parts of East Asia, especially Taiwan Province of China and the Republic of Korea.

____________

* The authors of this article are Baochang Gu, Associate Director and Senior Research Associate, China Population Information and Research Centre (CPIRC), Beijing, and Krishna Roy, Programme Manager, Latin American and Caribbean Operations Research and Technical Assistance Programme, Population Council, Washington, D.C., the United States.

In scrutizing the data from previous population censuses and fertility surveys in China, Coale and Banister (1994:476-477) have demonstrated that the sex ratio at birth was higher than normal in the 1930s and 1940s, which suggests that larger than normal female mortality "resulted from the persistence of the traditional practice of female infanticide".2 During the 1960s and 1970s however, the sex ratio at birth in China "was very close to 106" (Zeng and others, 1993:283) and "by and large within the normal range" (Gu and Li, 1994). And "the proportion of missing girls in each five-year grouped cohort reached a low point at about 2 per cent of the girls born ..." (Coale and Banister, 1994:477). "The decline of excess female mortality after the establishment of the People's Republic was assisted by the action of a strong Government, which tried to modify this custom as well as other traditional practices that it viewed as harmful" (Coale and Banister, 1994:472).

The decade of the 1980s, however, witnessed a rise in the sex ratio at birth, particularly in the late 1980s (Gu and Li, 1994). More specifically, it exceed 108 in the year 1984, and since then has not dropped lower than that level. In their recent study, Park and Cho (1995:60) examined the recorded sex ratios of children by single year of age through age four for China, Taiwan Province of China and the Republic of Korea, and concluded that in the three populations it appears that the sex ratio at birth began to rise "after about 1985", and "the initial year of increase in the sex ratio at birth appears to be around 1985 in the Republic of Korea, 1986 in China and 1987 in Taiwan" (Province of China).3

Why did the sex ratio at birth happen to rise at almost the same time in these three populations, each having quite different social, economic and political contexts? One factor which can be noticed is that they share the traditional cultural trait of son preference. Another factor which deserves more attention is the rapidity of fertility decline in the these populations. Table 1 presents the sex ratio at birth and total fertility rate (TFR) of China, Taiwan Province of China and the Republic of Korea since 1980. Women's fertility in China as measured by the TFR declined dramatically from 5.8 in 1970 to 2.2 in 1980, an astonishing decrease of more than 50 per cent in only one decade; during the decade of the 1980s, the TFR fluctuated around the level of 2.4 (Gu and Yang, 1991; SFPC, 1994; Gu, forthcoming). In the case of Taiwan Province of China, between 1956 and 1983, the TFR fell by 67 per cent from 6.51 to 2.16, and then decreased 22 per cent more to 1.68 in 1986; since then it has oscillated narrowly around the level of 1.8 (Freedman, Chang and Sun, 1994). In the case of the Republic of Korea, the TFR also dropped precipitously from about 6.0 in the early 1960s to the replacement level (2.1) at the beginning of the 1980s, dropping to below the replacement level to as low as 1.6, as shown in table 1 (Hong, 1994).

	China		Taiwan Province of China		Republic of Korea
Year	SRB	TFR	SRB	TFR	SRB	TFR
1980	107.4	2.24	106.4		103.9
1981	107.1	2.63	107.0		107
1982	107.2	2.86	106.9		106.9	2.7
1983	107.9	2.42	106.7	2.16	107.7
1984	108.5	2.35	107.3		108.7	2.1
1985	111.4	2.20	106.6		110.0
1986	112.3	2.42	107.2	1.68	111.9
1987	111.0	2.59	108.3	1.70	109.0	1.6
1988	108.1	2.52	108.2	1.85	113.5	1.6
1989	113.9	2.35	108.6	1.68	112.1
1990	114.7	2.31	110.2	1.81	116.9	1.6
1991	116.1	2.20	110	1.72	112.9
1992	114.2	2.00			114.0
1993	114.1

Sources: Sex ratio at birth (SRB): China: (1980-1989) Gu and Li, 1994: table 1; (1990) SSB, 1991b; (1991) SSB, 1992; (1992) SSB, 1993b; (1993) SSB, 1994. Taiwan Province of China: Chang, 1994: table 4; (1991) Freedman, Chang and Sun, 1994: table 14. Republic of Korea: Park and Cho, 1995: table 6; (1981) Cho and Kim, 1994.

Total fertility rate (TFR): China: SFPC, 1994:68. Taiwan Province of China: Freedman, Chang and Sun, 1994: table 2. Republic of Korea: (1982, 1984, 1987) KIHASA, 1991:18; (1988, 1991) Hong, 1994.

While the fertility rate in China seems to be somewhat higher than in the other two populations, one should keep in mind that the decline is occurring in a country accounting for one-fifth of the world's total population, with considerable regional variations in social, economic and demographic conditions not to mention a stronger population control programme.4 Overall, these three populations all experienced rapid fertility decline to the replacement level within a short period, which continued to drop to below the replacement level during the 1980s. The former period is called the "demographic transition period"; the latter period, the "post-transition period" (Freedman, Chang and Sun, 1994). "Historically, the number of children preferred in Taiwan (Province of China) has been tilted upward by a strong preference for sons. The decrease in the number of children wanted and born during the fertility transition has partly resulted from a decreasing, if persistent, preference for sons". However, along with the rapid fertility decline, the "continuing conflict for many couples between the desire to have only two or three children and the desire to have a son has become evident in Taiwan" (Province of China) (Freedman, Chang and Sun, 1994). This observation should be viewed as applicable to the other two populations as well.

When fertility is high, people may satisfy their sex preference for son(s) through the number of children they are going to have, some of whom almost surely will be boys. However, when fertility declines dramatically people will not be able to have the number of children they would like to have, because of such factors as a strict population policy or social and economic constraints. Therefore, they then switch to a "quality for quantity" strategy of fertility (Gu and Peng, 1991; Gu, 1992). In a socio-cultural setting with a strong son preference but under conditions of rapid fertility decline, couples will be more conscious of their sex preference for children. When they know the limited number of children they are going to have, it is very likely that they would seek various means to ensure that they get the sex of child they most desire (Gu and Li, 1994). In the other words, the distorted sex ratio at birth is a new demographic phenomenon that accommodates both the parents' sex preference for children and the small-family norm (Hong, 1994; Park and Cho, 1995).

It seems that the occurrence of abnormal sex ratios at birth during the mid-1980s in the three aforementioned populations has a lot to do with the rapidity of fertility decline as well as the downward change in the norm concerning desired family size along with the persistence of son preference.

While a higher than normal sex ratio at birth is observed in China as a whole, this does not mean that the phenomenon occurs everywhere and to the same degree nationwide. The sex ratio at birth in 1989, based on data from the 1990 population census, is presented in table 2 for China as a whole and 30 individual provinces, municipalities and autonomous regions (all referred to as provinces hereafter).

With regard to geographical variations in the sex ratio at birth by residence, the city provinces showed the lowest sex ratio at birth, i.e. 108.9, which is close to the normal range. However, it is interesting to note that in town populations, the sex ratio at birth was highest at 111.9, whereas for the county or rural population it was 111.7. With regard to geographic variations in the sex ratio at birth by province, in 1989 it varied among the 30 provinces from as high as 117.4 in Guangxi and 116.7 in Zhejiang to as low as 103.4 in Guizhou and 103.6 in Tibet. While China has an overall sex ratio at birth of 111.3,5 21 of the 30 provinces had a sex ratio at birth higher than 108.0, which means that they are above the normal or acceptable range.

It is interesting to observe the relationship between the sex ratio at birth and the fertility level among the provinces, which table 2 also shows. Among the nine provinces with sex ratios at birth less than 108, well within the normal range, there are some which are socio-economically more advanced, having the lowest fertility in China, much below the replacement level. These include Shanghai and Beijing with TFRs of 1.344 and 1.332, respectively; both of them are virtually entirely metropolitan. In addition, there are other provinces which are among the least developed socio-economically, having concentrated minority populations and the highest fertility in China; the TFRs of these are Tibet (4.222), Xinjiang (3.157), Guizhou (2.963), Yunnan (2.588) and Qinghai (2.468). In the case of China as a whole, it appears that the sex ratio at birth displays an inverted U-shaped relationship with the fertility level and socio-economic development. The phenomenon of abnormal sex ratios at birth is associated more with those areas which are in the process of developing socio-economically and are characterized by an accelerated fertility transition.

Observations on Taiwan Province of China and the Republic of Korea are somewhat different. In his discussion of the former, Chang (1994: table 6) argued that the sex ratio at birth for parity 3 and above is "positively associated with degree of urbanization in Taiwan" (Province of China), with the ratios being 138 for Taipei City, 123 for Kaohsiung City, 121 for other cities and 119 for all counties. However, it may be noted in another account of the fertility situation in that province that "by 1991, all urbanization strata had TFRs below replacement levels" (Freedman, Chang and Sun, 1994: table 6).

Park and Cho (1995:63) examined the sex ratio of children under five years of age by geographical area in the Republic of Korea, and demonstrated that "the sex ratio rose earlier in large cities than in towns or rural areas .... the year the sex ratio at birth reached a consistent level of 110 or more is 1985 for cities, 1986 for towns, and 1988 for rural areas," while in 1980 the sex ratio of young children was "within a normal range throughout (the Republic of) Korea". Taegu had the highest ratio (124.0 for age zero in 1990) among the six largest cities in the Republic of Korea. This is an area, according to Park and Cho (1995:64), "considered to be very conservative, having a strong parochial tie popularly known as the TK (Taegu-Kyungsangpuk province) sentiment", which implies a greater association of high sex ratio with the cultural setting of strong son preference rather than the size of the city.

			Sex ratio at birth
Rank	Area/province	TFR	Total	City	Town	County
	China	2.253	111.3	108.9	111.9	111.7
1	Guizhou	2.963	103.4	99.4	109.0	103.7
2	Tibet	4.222	103.6	112.4	106.0	102.8
3	Xinjiang	3.157	104.1	106.6	104.6	103.6
4	Shanghai	1.344	104.1	103.9	104.0	104.7
5	Qinghai	2.468	104.6	115.3	92.5	103.9
6	Beijing	1.332	107.1	106.1	105.8	108.9
7	Yunnan	2.588	107.3	103.9	105.3	107.6
8	Heilongjiang	1.713	107.3	105.5	106.4	108.6
9	Jilin	1.806	107.8	106.0	107.3	108.5
10	Gansu	2.340	108.4	106.6	112.6	108.5
11	Inner Mongolia	1.967	108.5	105.2	105.3	110.1
12	Hubei	2.496	109.5	108.8	115.0	109.4
13	Ningxia	2.614	109.7	111.8	110.0	109.4
14	Fujian	2.362	109.9	109.4	124.0	108.9
15	Shanxi	2.461	110.1	111.5	109.3	109.9
16	Hunan	2.397	110.1	105.6	111.1	110.5
17	Shaanxi	2.705	110.3	113.6	116.7	109.6
18	Tianjin	1.661	110.4	106.4	107.6	115.4
19	Jiangxi	2.460	110.4	112.8	112.1	109.9
20	Liaoning	1.505	110.5	107.5	107.0	113.2
21	Hebei	2.331	110.9	104.0	108.4	111.9
22	Anhui	2.511	111.3	108.9	107.8	111.0
23	Guangdong	2.512	111.3	114.0	120.5	109.1
24	Sichuan	1.758	112.1	108.9	106.0	112.8
25	Jiangsu	1.939	113.8	112.0	107.3	114.5
26	Shandong	2.124	115.0	113.3	117.2	115.2
27	Hainan	2.932	116.1	111.1	136.2	114.7
28	Henan	2.897	116.2	113.0	113.9	116.6
29	Zhejiang	1.404	116.7	107.5	119.2	118.2
30	Guangxi	2.727	117.4	113.2	110.4	118.1

Source: Sex ratio at birth: SSB, 1991a:45, 427-429.

Total fertility rate: SSB, 1991a:57.

To identify the particular population groups who are more likely responsible for the rise in the sex ratio at birth, table 3 presents the available data by parity of women in China as well as Taiwan Province of China and the Republic of Korea. While the overall sex ratio at birth in China has been rising since the mid-1980s and has remained at a high level, i.e. as high as 114, since 1989, the ratio for parity 1 has been almost constantly at the normal level of around 107, and until 1984, so was the ratio for parity 2. The remaining data all have been above the normal range since the very beginning of the 1980s. The higher is the order of birth, the higher has been the sex ratio at birth, and the more recent is the year, the higher has been the sex ratio at birth. By the end of that decade, they were all at high levels, even above 120, "which is implausible to be explained by either biological or ethnical factors" (Gu and Li, 1994). It is clear that the imbalance between male and female births has been occurring mainly at births of high orders.

This observation also holds true for Taiwan Province of China and the Republic of Korea. The overall ratio for Taiwan Province of China went up in the mid-1980s and remains at the level of around 110, which is lower than that of the two other populations. The sex ratio for parity 1 is normal and for parity 2 more or less acceptable, but for parity 3 and 4, all exceeded 110, and for parity 4 it reached as high as 130 in 1990.

	Parity of women
Year	Total	1	2	3	4	5+
China
1981	107.1	105.1	106.7	111.3	106.5	114.1
1982	107.2	106.6	105.2	109.4	112.9	109.9
1983	107.9	107.8	107.2	109.5	104.7	112.1
1984	108.5	102.5	113.3	113.0	115.3	127.3
1985	111.4	106.6	115.9	114.1	126.9	117.3
1986	112.3	105.4	116.9	123.1	125.3	123.5
1987	111.0	106.8	112.8	118.9	118.6	124.6
1988	108.1	101.5	114.5	117.1	123.1	108.7
1989	113.9	105.2	121.0	124.3	131.7	129.8
1989 (Cities)	110.5	105.6	121.3	128.9	137.3	37.4
1989 (Towns)	114.0	108.0	125.5	127.0	136.3	134.1
1989 (Rural)	114.5	104.8	120.7	123.9	131.2	129.3
1990	114.7
1990*(Cities)	105.5
1990*(Towns)	115.5
1990*(Rural)	116.4
1991	116.1	110.8	122.6	124.4(3+)
1992	114.2	106.7	125.7	126.7(3+)
1993	114.1	105.6	130.2	126.1(3+)
Taiwan Province of China
1987	108	107	108	110	114
1988	108	107	107	112	111
1989	109	107	107	113	121
1990	110	107	109	119	128
1990 (Taipei)	112	108	110	134	156
1990 (Kaohsiung)	109	106	107	122	130
1990 (Cities: all)	109	105	108	119	131
1990 (Rural: all)	110	107	109	117	121
1991	110	107	109	118	130
Republic of Korea
1980	104	106	104	103	102	96
1981	107	106	107	107	113	115
1982	107	106	106	110	113	118
1983	108	106	106	113	121	128
1984	109	107	108	119	132	134
1985	110	106	108	133	157	154
1986	113	108	112	143	161	161
1987	109	105	109	137	150	163
1988	114	108	114	170	199	187
1989	113	105	114	190	217	214
1990	117	109	117	196	234	215
1991	112.9	106.1	112.8	184.7	212.3
1992	114.0	106.4	112.8	195.6	229.0(4+)

Sources: China: (1981-1989) Gu and Li, 1994; (1989) SSB, 1993a; (1990) SSB, 1991; (1991) SSB, 1992; (1992) SSB, 1993b; (1993) SSB, 1994. Taiwan Province of China: (1987-1989, 1991) Freedman, Chang and Sun, 1994; (1990) Chang, 1994. Republic of Korea: (1980-1990) Cho and Kim, 1994; (1991) Park and Cho, 1995; (1992) Hong, 1994.

*Note: The SSB data for 1990 do not provide sex composition by birth order, but only by residence.

With regard to the Republic of Korea, the overall ratio rose from normal levels in the mid-1980s and since 1988 remained at a level above 113, but the ratio for parity 1 has always been within the normal range, which was also the case for parity 2 during most of the 1980s until 1986. The sex ratios at birth for the other parities have been far out of the normal range, even exceeding 200 for parity 4 and above in recent years. This suggests a positive relationship between sex ratio at birth and birth order in recent years in the Republic of Korea (Park and Cho, 1995:65).

Table 4 provides data obtained by computerized analysis of the 1 per cent sample of China's 1990 census. Disaggregated by three variables, it shows the sex ratios of children born between 1 January 1989 and 30 June 1990 who survived to the time of the 1990 census: (a) by the number and sex of previous siblings, (b) by residence and (c) by the educational level of the mother. The table indicates on one hand that the sex ratio of the surviving children aged 0-1.5 born during 1989 and the first half of 1990 was high nationally, i.e. 115.3 surviving male children per 100 surviving female children. On the other hand, it also reveals that, regardless of residence and education of the mother, the sex ratios of surviving children for women with no surviving children or only one son tend to be in the normal or acceptable range. This may suggest that the majority of these women, i.e. those having their first child as well as those with a son and having a second child, do not belong to the group likely to distort the sex ratio at birth. By contrast, the sex ratios of surviving children for women with no son but only daughter(s) were all extremely high, even exceeding 200; also, the sex ratios of surviving children for women with son(s) but no daughter were too low to be normal.

Gender of siblings
	0	1		2			3+			Total
Sex	0male(m)	1m	0m	2m	1m	0m	3+m	1+m	0m
	0female(f)	0f	1f	0f	1f	2f	0f	1+f	3+f
China
Sex ration at birth(SRB)	105.6	101.4	149.4	74.1	116.4	224.9	64.4	121.9	219.4	115.3
Residence
Rural areas
SRB	105.1	101.1	152.9	73.1	114.6	226.6	63.6	119.7	215.9	116.0
Towns
SRB	106.0	100.1	143.6	79.4	120.4	215.2	71.7	125.3	215.6	115.5
Cities
SRB	106.0	103.8	147.7	69.7	116.4	233.5	52.1	125.4	237.0	113.8
Education of mother
< 1 year of schooling
SRB	99.2	99.2	129.5	74.8	115.0	209.2	66.8	119.0	186.0	111.9
1-5 years of schooling
SRB	104.3	99.5	148.0	74.3	116.9	223.7	62.8	117.0	237.3	115.0
6-8 years of schooling
SRB	107.5	105.3	159.9	73.5	118.2	239.2	68.8	146.7	245.8	117.4
9+ years of schooling
SRB	108.1	100.1	157.2	71.6	111.3	228.9	41.2	131.3	223.3	114.1

Notes: 1. Calculated from a 1 per cent sampling of China's 1990 census data; some figures may disagree with those published in 10 per cent tabulations.

2. Twins are not considered owing to poor sex identification.

3. The table excludes children who did not survive to the date of the 1990 census.

4. In some cases, the children enumerated in the household many not be the biological siblings of the child born during the period 1 January 1989 to 30 June 1990.

This pattern provides evidence that a sex-selective process showing son preference is involved in childbearing in China. The imbalance between male and female births has been occurring mainly among women who already have one or more children, but especially among women having daughter(s) but no son, which suggests intentional interference with fertility behaviour owing to strong son preference. However, it may be noticed that, although extremely high sex ratios of surviving children may be observed among women with no son, women with 6-8 years of education (to the junior middle school level) appear to have relatively higher sex ratios at birth, which once again suggests an inverted U-shaped relationship at the individual level, as observed previously at the aggregate level in table 2. Obviously, more research needs to be undertaken on these issues, which include the possibility that some parents in households with more than one child may be aborting male fetuses when their other children are males and they have no female child or children.

In a recent study with the same data set, Poston and others (1995) examined the sex ratio by parity and sex composition of previous births in 29 provinces of China (Tibet was excluded owing to the unavailability of data). That study showed a similar pattern, i.e. the sex ratios at birth for women with previous births of all male children tended to be very low in the provinces. For example, the mean value for parity 3 women with two previous male births was as low as 81; the sex ratios at birth for those with previous births of all female children tended to be very high. Also, the mean value for parity 3 women with two previous female births was as high as 208, and the sex ratios at birth for those with previous births of mixed sex tended to fall in between that range, e.g. the mean value for parity 3 women with two previous births of mixed sex was 112 (Poston and others, 1995).

Freedman, Chang and Sun (1994: table 13), in their discussion of the trends in family-size preferences and family planning in Taiwan Province of China during the period of 1965-1991, examined data on the percentage of married women aged 22-39 who wanted no additional children and the percentage currently using contraception, by number of living children and number of living sons in 1965, 1980, 1985 and 1991. The data revealed that the percentage of women who wanted no additional children and the percentage currently practising contraception increased with the number of living sons they had. Thus, those with no son in 1991 were less likely to want no additional children regardless of the number of children they had. Also, the latter group of women were less likely to practise contraception than women who had at least one son; this trend held true for women with up to four children.6 That study indicates that, in the case of Taiwan Province of China, preference for sons had persisted to 1991, even if on a diminished scale compared with earlier periods. In 1991, "among couples with similar numbers of children, the proportions wanting no more children and the proportions currently practicing contraception increased with the number of living sons they had" (Freedman, Chang and Sun, 1994:324).

Park and Cho (1995: table 7) have shown the sex ratio at birth by birth order following a specified sex sequence of previous births for the Republic of Korea in 1974 and 1991. Whereas in 1974 none of the sex ratios classified by the sex sequence of earlier births was significantly different from the normal level, in the 1991 sample the sex ratio of third children born to women with two daughters, i.e. 136.3, significantly deviated from the normal level. Though the sample size under discussion for the Republic of Korea may be too small to be statistically significant, Park and Cho (1995:68) concluded that "sex-selective abortion appears especially prevalent among families having only daughters".

Several studies (Hull, 1990; Johansson and Nygren, 1991; Banister, 1992a) have hypothesized female infanticide, sex selective abortion, the underreporting of female births, or female adoption as plausible immediate causes for the rise in the sex ratio at birth in China.

Zeng and others (1993: table 3) have demonstrated that sex-differential underreporting of births and sex-selective induced abortion after pre-natal determination were mainly responsible for the increase in the reported sex ratio at birth in China during the 1980s. To identify the contribution of sex-selective abortion to the rise in the sex ratio at birth, a field survey was conducted in 1993 in southern Zhejiang Province, which had a sex ratio at birth higher than 120, according to the 1990 census (Gu and Li, 1994). Table 5 presents the sex ratio of the aborted fetuses by the number and sex of surviving children for that province. About 30 per cent of the abortion cases occurred among women with one daughter but no son; the sex ratio of aborted fetuses was as low as 51.0. For those women with at least one son, the sex ratio of aborted fetuses tended to be normal which suggests less interference. However, for those women with no son, the sex ratio of aborted fetuses was much lower than normal, a phenomenon most likely affected by the sex identification of the fetuses prior to the induced abortion. It would be interesting to compare this table with table 4 on the sex ratio at birth by the number and the sex of siblings: the trend of the sex ratio of aborted fetuses is surprisingly consistent with the trend of the sex ratio of surviving children. Discounting the underreporting of female births, it would seem that sex-selective abortion following sex identification of a fetus by ultrasound or other diagnostic methods is additionally, and likely to be increasingly, responsible for the increase in the sex ratio at birth in China.

Source: Gu and Li (1994: table 7).

While infanticide and abandonment are relatively less responsible for the rise in the sex ratio at birth in China, research also indicates that female babies are more frequently subject to abandonment than male babies. During the decade of the 1980s, the infant mortality rate for girls was increasing while that of boys was decreasing. This was particularly so in rural areas, where as discussed above the sex ratio at birth has been exceptionally high (Gu and Li, 1994).

In Taiwan Province of China, it has been observed that "an increasing number of couples with no sons use amniocentesis when the wife is pregnant to determine the sex of the fetus and to have an abortion performed if the fetus is female" (Freedman, Chang and Sun, 1994:324). According to the 1992 KAP (knowledge, attitude and practice) survey, most of the pregnancies with reported pre-natal sex screening "turned out to be male live births" with a sex ratio of 295 (72.0 per cent for male live births and 24.4 per cent for female live births: see Chang, 1994).

Several studies on the Republic of Korea argue that sex-selective abortion is "the sole cause of rising sex ratios" (Park and Cho, 1995:64). According to the 1988 National Fertility and Family Health Survey, sex-identification tests were conducted in 1.2 per cent of pregnancies surveyed that year. Out of 275 cases tested, 169 were male fetuses and 87 female fetuses, with the sex not being distinguishable in nine cases. While more than 90 per cent (154 cases) of the tested pregnancies indicating male fetuses resulted in normal births, more than 30 per cent (27 cases) of the tested pregnancies indicating female fetuses were terminated by induced abortion, subject to underreporting, according to the report (Hong, 1994: table 9).7 While Park and Cho (1995:68) have argued that "sex-selective abortion appears especially prevalent among families having only daughters", Cho and Kim (1994: table 2c) have noted that the abortion rate tends to increase with parity and time, and is much higher among families with at least one son, which "indicates the changing preferences toward smaller families while the son preferences are maintained". They state that "ill treatment of baby girls, as compared with baby boys, is possible; but overt sex-selective infanticide is beyond imagination. The lag in the rise of the sex ratio in rural areas may be largely explained by a lesser availability there than in urban areas of medical facilities to determine fetal sex. Son preference in rural areas is expected to be stronger than in urban areas" (Park and Cho, 1995:64).

Since the time when the issue of rising sex ratios at birth (SRB) became a major concern, it has often been asserted that the phenomenon in China is a result of the Government's adoption in 1980 of a population policy advocating one child per couple. Although it is true that the sex ratio at birth for high parity women exceeded the normal level from the very beginning of the 1980s as seen in table 3, it is not true for other women since the overall sex ratio at birth did not become abnormal until the mid-1980s, as seen in table 1. Moreover, it would be difficult to explain why a similar phenomenon has also been observed since the mid-1980s in the populations of Taiwan Province of China and the Republic of Korea, both presenting different contexts. As stated by Coale and Banister (1994:476): "Can we assume that only special factors as communism, compulsory family planning, or the one-child policy can cause a sharp rise in the masculinity of younger cohorts, of the type that occurred in China in the 1980s and early 1990s? The answer appears to be `no' because a very similar increasing dearth of young girls is observed in data from the Republic of Korea from 1983 to 1988".

The implication of this situation for policy purposes would be misleading if one were to assume that China should be able to rid itself of the abnormal sex ratios at birth once a two-children-per-couple policy is adopted (Gu, 1994). Some researchers have followed this line of reasoning: "Witness the high sex ratios in Taiwan (Province of China) and (the Republic of) Korea where such drastic fertility policies (as found in China) do not exist. Were China's policy relaxed, the main difference might be that the SRB deviation from normality might not occur until the second parity, instead of, as is now the case in China, after the first parity" (Poston and others, 1995). As shown in table 2, in areas where the population policy of "advocating one child per couple" is most effective, such as in Shanghai and Beijing, the sex ratio at birth has been normal, and "does not demonstrate a 'missing girl' problem" (Banister, 1995). Areas with the most serious abnormality in the sex ratio at birth, such as Guangxi, Hainan and Henan, are also characterized by high fertility and a large proportion of births occurring among the higher-than-one parity couples (see Gu and Yang, 1991 for a detailed discussion of the fertility trends in rural China during the 1980s).

It has been widely noticed, at least in the three populations under discussion, that the rise in the sex ratio at birth seems to have a lot to do with the rapidity of the fertility decline which has occurred in recent decades. "In both populations (China and the Republic of Korea), the total fertility rate fell from about 6.0 in the late 1950s to about 2.6 around 1980, and marital fertility changed from an age pattern indicating little voluntary control to a pattern implying widespread practice of birth limitation" (Coale and Banister, 1994:476). The fertility transition in China accelerated at the beginning of the 1970s with the vigorous implementation of the family planning programme, driven by the policy of "later (marriage), longer (birth spacing) and fewer (children)"; the TFR declined more than 50 per cent in only one decade (Gu and Peng, 1991:50). "Taiwan (Province of China) completed the fertility component of the demographic transition in the 27 years between 1956 and 1983 with the fall of the net reproduction rate (NRR) from 2.84 to 1.01" (Freedman, Chang and Sun, 1994:317). However, this might not have been the case if the cultural setting of son preference were absent. For example, as we have seen in Indonesia, Sri Lanka and Thailand, where son preference appears to be nearly non-existent, a rapid decline in fertility may not necessarily cause the sex ratio at birth to rise to an abnormal level (Gu, 1994). Such is the case also for the most developed areas in China, such as Shanghai and Beijing (Gu and Li, 1994).

The cultural setting of son preference seems to be crucial: when fertility declines dramatically people show that, of the fewer children they have, they still prefer to have son(s) for family labour and for support during their old age (more so in poor and remote areas than other parts of the country). In some cases, the desire to have sons is related to enhancement of the mother's status or the wish to perpetuate the family line (more so in areas with a strongly growing economy). In this context, a recent field survey indicates that the most seriously perceived gender inequality for many women is that they feel that they will be most strongly discriminated against if they simply fail to have a son (Gu and Li, 1994). According to the Korean National Fertility and Family Health Surveys, the ideal number of children preferred by Korean women decreased over the period from the 1960s to the 1990s, but 62.6 per cent of rural women and 34.6 per cent of urban women still insisted that it is necessary to have a son (Hong, 1994), indicating a changing preference towards smaller families, but one that still maintains the traditional preference for sons (Cho and Kim, 1994). According to a KAP survey in Taiwan Province of China in 1991, "the mean preferred number of sons was still greater than the mean preferred number of daughters among both younger and older wives" (Freedman, Chang and Sun, 1994:324; Chang, 1994). "In areas where there is little or no son preference, fertility decline does not bring it on. But where son preference existed alongside high to moderately high fertility, even a modest decline in fertility may exacerbate the existing discrimination against female babies, children, and now fetuses" (Banister, 1995).

Another factor which is often claimed to be responsible for the rising sex ratio at birth in some Asian populations is the introduction during recent years of technologies for pre-natal sex identification. "In the past decade or so, medical advances have opened up the potential for a new form of gender discrimination, that is through prenatal sex identification followed by sex-selective induced abortions" (Leete, 1994). The rise in the sex ratio at birth during the 1980s coincided with the availability and accessibility of medical technology in China, in particular the ultrasound B machine, to detect the sex of fetuses during pregnancy (Zeng and others, 1993:291). Two techniques for pre-natal sex screening, namely chorionic villus sampling (CVS) and amniocentesis, have become available recently in Taiwan Province of China. In view of the fact that CVS can determine the sex of a fetus at the early stage of pregnancy, "women who intend to perform a combination of prenatal sex determination and abortion usually choose CVS for sex screening simply because of the lower risk of induced abortion" (Chang, 1994). In the case of the Republic of Korea, because CVS "is very expensive, only a small number of people can afford it. Amniocentesis is frequently used" as an alternative. "The most often used method in (the Republic of) Korea appears to be ultrasound. It is the least expensive and simplest method, but it is effectively applicable only in later stages of pregnancy" (Park and Cho, 1994:77).8

Nevertheless, the availability and accessibility of technologies for the pre-natal detection of the sex of the fetus and sex-selective abortion may not be regarded as the "root causes" of the phenomenon of abnormal sex ratios at birth. Otherwise, it would be very difficult to explain why it does not occur in more developed areas, such as Shanghai and Beijing in China as well as in many developed countries of the world, where sex-detection technology is widely available (Gu, 1994).

Based on the discussion above, it seems that the phenomenon of abnormal sex ratios at birth observed recently in a number of East Asian populations may be viewed as the outcome of an interplay of four factors, namely (a) the effect of the cultural setting on sex preference, (b) level of socio-economic development, (c) rapidity of fertility decline, and (d) focus of the population programme. Such a situation may be conceptualized as in the flowchart above. At the aggregate level, the problem is more likely to occur in a population in the process of development with a cultural setting conducive to a strong son preference when the fertility level declines rapidly and the population programme is almost solely concentrated on reducing the number of children per woman in a context where the technology for pre-natal sex detection is easily available (Gu, 1994). At the individual level, it may be more likely to occur among couples living in a less urbanized area, where there has been a downward change in the family size norm but persistence of a preference for sons, who having a lower level education, have not even one son but are exposed to the knowledge of or technical means for pre-natal sex detection and sex-selective abortion. This conceptualization may be helpful to our understanding of the inverted U-shaped relationship observed among the various provinces of China (table 2), and among the various population groups covered by table 4.

Family size norm	Population policy \|/	Fertility level
| ---------------> |	Childbearing	/|\ ---------------- \|/
Sex Preference	/|\ Technology	Sex ratio at birth

It has been argued that the phenomenon of abnormal sex ratios at birth is a transitory one, which will eventually disappear when societies move sufficiently closer to being fully modernized. However, this does not necessary mean that development will automatically take care of this problem as can be seen in the cases of Taiwan Province of China and the Republic of Korea, the evidence from which so far is mixed or even discouraging despite the advanced stages of economic development there (Gu, 1994; Banister, 1995). Taiwan Province of China and the Republic of Korea are "some of the most developed places in Asia, yet sex-selective abortion is very widespread .... greater economic development, affluence, education, and knowledge do not necessarily ameliorate son preference or reduce the use of sex-selective abortion" (Banister, 1995). It seems that when fertility declines rapidly in terms of the number of children a couple will have, the tradition of a strong preference for sons over daughters will probably become more salient, and socio-cultural factors will likely be more influential than economic factors in dominating people's fertility behaviour (Gu, 1994). It is argued, therefore, that "economic development is not the solution. Rather, changes in cultural norms are required" (Banister, 1995; Hong, 1994).

Much of the concern about the abnormal sex ratio at birth often focuses on the possible occurrence of the so-called "marriage squeeze", a phenomenon that has been foreseen by social scientists in societies where cohorts of males exceed those of females in number. Such a situation will inevitably result in an aberrant age-sex structure of the population of such a society, since many men, when they reach marriageable age, will be hard pressed to find spouses of compatible age, which in turn may affect the stability of that society (Gu and Peng, 1991:57-59; Zeng and others, 1993:296; Freedman, Chang and Sun, 1994:324; Gu and Li, 1994; Hong, 1994; Park and Cho, 1995:74-75). The discussion on the issue suggests, however, that our knowledge of the nature and magnitude of potential societal adjustments to a demographic imbalance in the sex ratios of cohorts of marriageable age is rudimentary. While the negative as well as positive consequences of the imbalance between the numbers of males and females in a society are still under discussion (Park and Cho, 1994), it has been argued that the focus of the concern itself is male oriented and reflects a high valuation of males and a disregard of the needs of females (Gu, 1994; Banister, 1995).

Such a situation may also shift attention away from the present to the remote future, thus taking the urgency out of the need to find a solution to the problem. A more serious implication of the issue is very much related to the status and well-being of females at home and in society at large, not to mention the fate of the females who are maltreated, abandoned, or whose life was prematurely cut short by abortion or infanticide (Gu, 1994; Greenhalgh and Li, 1995; Banister, 1995). Women without sons may encounter a variety of day-to-day problems and feel inferior to women with sons. The evidence from the aforementioned field survey suggests that for women, especially those in rural areas, their status and position at home is to a great degree dependent upon their ability to bear children, particularly a son. Many women feel that they are discriminated against by society simply because they fail to have a son, which is by far the most seriously perceived result of gender inequality. In trying to counteract discrimination against female fetuses and children, not only should the future dearth of available women, i.e. potential wives, be emphasized, but also more so the negative impacts of sex-selective abortion, female infanticide, and the negative impact of these on today's women and girls (Banister, 1995).

One question that may arise in the discussion of this issue is: why have the sex ratio at birth abnormalities that have occurred in a number of Asian populations not been given the attention that this issue deserves in existing population and family planning programmes? The answer to this question seems to be related to the existing conceptualization of the fertility transition and evaluation of family planning programmes. Conceptually, the "fertility transition" has long been viewed as a single-dimensional process going from high fertility to low fertility, which focuses almost solely on the change in level of fertility, often measured by the number of children born per woman during her lifetime.

Rather than regarding the fertility transition as an integral part of the overall transition from a traditional society to a modern society, it should be considered as a multi-dimensional process, which includes not only how many children to have (level of fertility), but also when to have them (timing of childbearing), and also what to have (sex of child) (Gu, 1994; Ng and Gu, 1995). In reality people never think separately about "how many children to have", "when to have them" and "what sex of child to have". In developing countries where fertility has declined over a short period and in a dramatic manner, mainly as a result of family planning programmes rather than the gradual process of socio-economic development, the culture and ideology concerning the timing of fertility and sex composition of children may not change so quickly as the size of the family. This is probably the reason why the patterns and trends of abnormal sex ratios at birth may be observed not only in China but also in some other East Asian populations with similar cultures and rapid fertility declines.9

Should the "fertility transition" be regarded as a multi- rather than uni-dimensional process, the completion of the "fertility transition" would have to be redefined. Accordingly, population and family planning programmes should be implemented by attempting not only to lower fertility but also to bring about a balance in the sex ratio of births. Although the performance of a family planning programme is evaluated in terms of fertility level, population growth rate and contraceptive prevalence etc., performance should also be evaluated in terms of the degree of son preference in fertility behaviour as measured by the sex ratio at birth (Gu and Li, 1994). By the same token, information, education and communication (IEC) programmes should explicitly urge couples to value equally female and male children in addition to encouraging them to have fewer children (Gu, 1994).

More fundamentally, to curb population growth and achieve a balanced sex ratio at birth, the strategy of "beyond family planning" must be recalled. The family planning programme should be implemented by incorporating other socio-economic improvements such as the improvement of women's status and better maternal and child health care. Family planning, women's status, and maternal and child health care (MCH) should be regarded as three major components of the concept of "reproductive health". In other words, the programme should be more "people oriented" rather than "number oriented". Such programmes should emphasize reproductive health and the improvement of women's status, and should be coordinated with the MCH programme. The human orientation of such programmes should be stressed, particularly in the light of the ICPD Programme of Action (Gu, 1994).

For countries with high fertility and strong son preference but not yet showing an abnormal sex ratio at birth, the experience of the three populations discussed in this article may suggest that it is likely that their fertility decline may also be accompanied by worsening discrimination against female fetuses and children. However, if such countries prepare themselves well and closely monitor aspects such as the sex ratio at birth while implementating their family planning programme, they may be able to reduce greatly the impact of such abnormalities if they begin to occur. When people face conflicts over the number and sex of children to have in the face of rising costs associated with rearing them, they will try every way available to achieve their desired combination of children's sex within the limited number of children they are to have. Some, when they have no choice over these matters prior to the birth of the child, will achieve it after the birth of the child if it turns out to be not of the desired sex, by neglect, mistreatment, abandonment, or infanticide. When pre-natal or even pre-pregnancy determination of sex is available, accessible and affordable, a couple will seek to make their choice of the child's sex before the child is born. A demographic implication of such action is that, while female infant mortality may decline, the sex ratio at birth may increase.

It has been argued that an important strategy is for the Government to ban misuse of pre-natal sex detection. However, such prohibitions may serve more as a moral statement against the selective abortion of female fetuses than as a fundamental solution to the problem. It is not so much that the technology for pre-natal sex selection is to be blamed, but that the culture and attitudes of the people have to be changed (Gu, 1994). Moreover, laws banning the misuse of sex-detection technologies may be counter-productive, since they are likely to be enforced in a discriminatory, uneven manner against women and medical personnel (Banister, 1995). By contrast, the diffusion of knowledge and information about the negative effects on women and children concerning the use of abortion and the application of sex-detection technology may help to discourage people from practising pre-natal sex screening (Chang, 1994).

The above discussion of three East Asian populations may provide some understanding of abnormal sex ratios at birth, but it is still only "tentative until better information confirms or refines" that understanding (Banister, 1995). Nevertheless, based on this discussion, we know fairly well when, where, among whom, how, and why abnormal sex ratios at birth and sex-selective abortion occur and are likely to occur. Therefore, such information should help Governments and NGOs to target population groups for programmes aimed at changing this situation for the better (Gu, 1994).

Consciousness-raising is still a priority strategy in any effort to rectify the situation of abnormal sex ratios at birth. This requires society-wide efforts aimed at emphasizing the value of girls and women, the need for education and health care to be made equally available to girls and boys, and the need for legal changes to promote true equality of the sexes (Banister, 1995). As is brought out in the Bali Declaration, sustainable development should not be regarded as related to population size only (ESCAP, 1992). A comprehensive strategy to raise the status of women should be recognized as the cornerstone of any sustainable development policy, which may subsequently work towards reducing people's desire for sons over daughters no matter what sex-detection technology is or becomes available (Gu, 1994). The study of the sex ratio at birth should not involve merely counting the number of male births relative to female births, but it should be of profound significance to our understanding of the population dynamics in developing countries as we approach the year 2000. It is an issue related on the one hand to the final completion of the fertility transition in a population and on the other hand to the empowerment of women in a society, which is an essential measure in the achievement of truly "sustainable development".

1. In fact, ESCAP and KIHASA had sponsored another meeting entitled "Seminar on Impact of Fertility Decline on Population Policies and Programme Strategies: Emerging Trends for the 21st Century", held at Seoul, Republic of Korea in December 1991, during which the issue had been discussed in the presentations (see Gu and Peng, 1991; Lee and Cho, 1991).

2. Park and Cho (1995:80-81) noted that the reported sex ratios at birth in the Republic of Korea "prior to 1945 were consistently and unusually low. A possible explanation is high fetal mortality, which claims far more male than female lives".

3. Park and Cho (1995) discussed the situations in four populations, namely those of China, Hong Kong, Taiwan Province of China and the Republic of Korea. Since the population structure of Hong Kong is quite different from that of the three other populations, it is thus excluded from the present discussion.

4. Given the higher infant and child mortality rates in some parts of the population than the rates for the more developed societies, the replacement level may have to be set at a level a little higher than 2.1 (Keyfitz, 1984).

5. There are two figures for the sex ratio at birth in China in 1989 (derived from the 1990 population census): one is 113.8 derived from SSB (1991a:461), the other is 111.3 derived from SSB (1991a:45) using different tabulation procedures. See Zeng and others (1993) for an explanation.

6. An exception is the figure for women with 5+ children but no son and currently using contraceptives, i.e. 93 per cent, which is somewhat higher than the figure for women with one son (87 per cent). This may be due to fluctuation because of the small size of the population under observation. See Freedman, Chang and Sun (1994: table 13).

7. Another study by Kim and Joo (1994) reports an estimate of the proportion of births at 3.6 per cent rather than 1.2 per cent following sex identification testing in the Republic of Korea in 1989-1991. See Hong (1994).

8. According to Park and Cho (1995:78), the cost of a chorionic biopsy test is more than half a million won (US$ 625), for amniocentesis 200,000-300,000 won (US$ 250-375), and for ultrasound 60,000 won (US$ 75).

9. A discussion of the dynamics of the three dimensions connected with the process of fertility transition is beyond the scope of this article; however, a brief speculation on the topic follows (Gu, 1992; Ng and Gu, 1995). In a pre-transition population, women tend to have as many as four children or even more; they tend to get married early in life and start to have children immediately after marriage, with the birth interval being as short as biologically and sociologically allowable (Davis and Blake, 1956). The transition may start with a change in tempo rather than quantum, which is a phenomenon that Coale (1973) defined as the "first transition". But the second transition, focusing on a reduction of the number of children born to a woman during her lifetime, will follow, and this results in the historical decline in fertility from high to low levels, mostly as a reduction in marital fertility. For a population reaching a relatively low level of fertility, given the fact that there is not much of an opportunity left for a further decline in terms of the number of children per couple, the timing of marriage and childbearing will once again play a much more important role with regard to further completion of the transition. Meanwhile, mainly owing to the rapidity of the fertility transition, people's desire for children may lag behind. When a couple cannot have the number of children they would like to have for various reasons, they may adopt a "quality for quantity" strategy of fertility and switch to a strategy favouring the quality of the children they desire to have. The tradition of a strong preference for sons over daughters may become a salient aspect along with fertility decline, and the sex ratio at birth may rise accordingly. Finally, by the end of the transition, a situation may occur when women with few children are having them at a relative later age in life, with proper spacing and a balanced sex ratio of children at the aggregate level, which is in essence compatible with that of a modern society. In this sense the three dimensions may play roles at various weights over the different stages of the fertility transition. If it is valid, the focus of the population and family planning programme should vary accordingly in view of the stage of the transition for a given population.

Banister, Judith (1992a). "China: recent mortality levels and trends". Paper presented at the Annual Meeting of the Population Association of America, Denver, Colorado.

__________ (1992b). "Implications and quality of China's 1990 census data". Paper presented at the International Seminar on China's 1990 Population Census, 19-23 October, Beijing.

__________ (1995). "Son preference in Asia - report of a symposium". Paper presented at the Annual Meeting of the Population Association of America, 6-8 April, San Francisco, California.

Chang, Ming-Cheng (1994). "Sex preference and sex ratio at birth: the case of Taiwan"(Province of China). Paper presented at the UNFPA/KIHASA International Symposium on Issues Related to Sex Preference for Children in the Rapidly Changing Demographic Dynamics in Asia, 21-24 November, Seoul, Republic of Korea.

Cho, Nam-Hoon and Il-Hyuun Kim (1994). "Impact of induced abortion on sex ratio at birth in Korea". Paper presented at the UNFPA/KIHASA International Symposium on Issues Related to Sex Preference for Children in the Rapidly Changing Demographic Dynamics in Asia, 21-24 November, Seoul, Republic of Korea.

Coale, Ansley J. (1973). "The demographic transition reconsidered". Proceedings of the International Population Conference (IUSSP) (Liege) 1:53-72.

__________ (1992). "High ratios of males to females in the population of China". Paper presented at the International Seminar on China's 1990 Population Census, 19-23 October, Beijing.

_________ and Judith Banister (1994). "Five decades of missing females in China", Demography 31(3):459-479.

Davis, Kingsley and Judith Blake (1956). "Social structure and fertility: an analytic framework", Economic Development and Cultural Change, 4:211-235.

ESCAP (1992). "Report of the Fourth Asian and Pacific Population Conference: 19-27 August 1992. Bali, Indonesia", Asian Population Studies Series No. 115, ST/ESCAP/1198.

Freedman, Ronald, Ming-Cheng Chang and Te-Hsiung Sun (1994). "Taiwan's transition from high fertility to below-replacement levels", Studies in Family Planning 25(6):317-331.

Greenhalgh, Susan and Jiali Li (1995). "Engendering reproductive policy and practice in peasant China: for a feminist demography of reproductions", Signs: Journal of Women in Culture and Society 20(31) (Spring).

Gu, Baochang (1992). "Fertility and fertility transition: quantity, timing and sex", Population Research (6):1-7 (in Chinese).

__________ (1994). "Synthesis of findings of eight country reports: implications for policy and programs". Discussion note presented at the UNFPA/KIHASA International Symposium on Issues Related to Sex Preference for Children in the Rapidly Changing Demographic Dynamics in Asia, 21-24 November, Seoul, Republic of Korea.

__________ (forthcoming). "Fertility in China: from the 1970s to the 1990s", In: The Interaction of Social, Economic, and Demographic Change in China, Alice Goldstein and Wang Feng (eds.) (Denver, Colorado: Westview Press).

__________ and Shuzhang Yang (1991). "Fertility trends in rural China in the 1980s: cohort effect versus period effect", Asia-Pacific Population Journal 6(4):3-34.

Gu, Baochang and Xizhe Peng (1991). "Consequences of fertility decline: cultural, social and economic implications in China". Paper presented at ESCAP/KIHASA Seminar on Impact of Fertility Decline on Population Policies and Programme Strategies: Emerging Trends for the 21st Century, 16-19 December, Seoul, Republic of Korea. And in Impact of Fertility Decline on Population Policies and Programme Strategies (1992) (Seoul: Korea Institute for Health and Social Affairs) 49-66.

Gu, Baochang and Yongping Li (1994). "Sex ratio at birth and son preference in China". Paper presented at the UNFPA/KIHASA International Symposium on Issues Related to Sex Preference for Children in the Rapidly Changing Demographic Dynamics in Asia, 21-24 November, Seoul, Republic of Korea.

Hong, Moon Sik (1994). "Boy preference and imbalance in sex ratio in Korea". Paper presented at the UNFPA/KIHASA International Symposium on Issues Related to Sex Preference for Children in the Rapidly Changing Demographic Dynamics in Asia, 21-24 November, Seoul, Republic of Korea.

Hull, Terence H. (1990). "Recent trends in sex ratios at birth in China", Population and Development Review 16(1):63-83.

_________ (1992). "Rising sex ratios at birth in China: evidence from the 1990 population census". Paper presented at the International Seminar on China's 1990 Population Census, 19-23 October, Beijing.

Johansson, Sten and Ola Nygren (1991). "The missing girls of China: a new demographic account", Population and Development Review 17(1):35-51.

Johansson, Sten and Arne Arvidsson (1992). "Problems in counting the youngest cohorts in Chinese censuses and surveys". Paper presented at the International Seminar on China's 1990 Population Census, 19-23 October, Beijing.

Keyfitz, Nathan (1984). "The population of China", Scientific American 250(2):38-47.

Lee, Hung-Tak and Nam-Hoon Cho (1991). "Consequences of fertility decline: social, economic and cultural implications in Korea". Paper presented at ESCAP/KIHASA Seminar on Impact of Fertility Decline on Population Policies and Programme Strategies: Emerging Trends for the 21st Century, 16-19 December, Seoul, Republic of Korea. And in Impact of Fertility Decline on Population Policies and Programme Strategies (1992) (Seoul: Korea Institute for Health and Social Affairs) 27-48.

Leete, Richard (1994). "Synthesis". Discussion note presented at the UNFPA/KIHASA International Symposium on Issues Related to Sex Preference for Children in the Rapidly Changing Demographic Dynamics in Asia, 21-24 November, Seoul, Republic of Korea.

Li, Yongping (1992). "Sex ratios of infants and relations with some socioeconomic variables: the results of China's 1990 census and implications". Paper presented at the International Seminar on China's 1990 Population Census, 19-23 October, Beijing.

Ng, Timothy Siu-Man and Baochang Gu (1995). "Dimensions of the fertility transition in the third world: level, timing, and quality". Paper presented at the Annual Meeting of the Population Association of America, 6-8 April, San Francisco, California.

Nizamuddin, M. (1994). "Background and objectives of the symposium". Paper presented at the UNFPA/KIHASA International Symposium on Issues Related to Sex Preference for Children in the Rapidly Changing Demographic Dynamics in Asia, 21-24 November, Seoul, Republic of Korea.

Park, Chai Bin and Nam-Hoon Cho (1995). "Consequences of son preference in a low-fertility society: imbalance of the sex ratio at birth in Korea". Population and Development Review 21(1):59-84.

Poston, Dudley L., Peihang Peggy Liu, Baochang Gu and Terra McDaniel (1995). "Son preference and the sex ratio at birth in China: a provincial level analysis". Paper presented at the Annual Meeting of the Population Association of America, 6-8 April, San Francisco, California.

Roy, Krishna (1994). "Socio-cultural factors and sex preference for children in South and East Asian countries - a review of selected studies". A review paper prepared for the UNFPA/KIHASA International Symposium on Issues Related to Sex Preference for Children in the Rapidly Changing Demographic Dynamics in Asia, 21-24 November, Seoul, Republic of Korea.

Singh, Jyoti (1994). "Keynote address". Presented at the UNFPA/KIHASA International Symposium on Issues Related to Sex Preference for Children in the Rapidly Changing Demographic Dynamics in Asia, 21-24 November, Seoul, Republic of Korea.

State Family Planning Committee of China (SFPC) (1990). Data of China Fertility and Contraception Survey (Whole Country) (Beijing: China Population Press).

__________ (1994). National Report of the People's Republic of China on Population and Development, for the International Conference on Population and Development, Cairo, 5-13 September.

State Statistical Bureau of China (SSB) (1991a). 10 Percent Sampling Tabulation on the 1990 Population Census of the People's Republic of China (Beijing: China Statistics Press).

__________ (1991b). China Population Statistical Yearbook 1991 (Beijing: China Statistics Press).

__________ (1992). China Population Statistical Yearbook 1992 (Beijing: China Statistics Press).

__________ (1993a). Tabulation of the 1990 Population Census of the People's Republic of China (Beijing: China Statistics Press).

__________ (1993b). China Population Statistical Yearbook 1993 (Beijing: China Statistics Press).

__________ (1994). China Population Statistical Yearbook 1994 (Beijing: China Statistics Press).

Tu, Ping and Liang Zhiwu (1992). "An evaluation of the quality of the enumeration of infant deaths and births in China's 1990 census". Paper presented at the International Seminar on China's 1990 Population Census, 19-23 October, Beijing.

Wen, Xingyan (1993). "Effect of son preference and population policy on sex ratio at birth in two provinces in China", Asia-Pacific Population Journal 7(4):25-40.

Xu, Yi and Weiming Guo (1991). "A discussion on the current sex ratio at birth and related issues in China", Population and Economy (5).

Zeng, Yi, Ping Tu, Baochang Gu, Yi Xu, Bohua Li and Yongping Li (1992). "An analysis of the causes and implications of the recent increase in the sex ratio at birth in China". Paper presented at the International Seminar on China's 1990 Population Census, 19-23 October, Beijing.

__________ (1993). "Causes and implications of the recent increase in the reported sex ratio at birth in China". Population and Development Review 19:283-302.

For further information on this material please contact: loftus.unescap@un.org