Child spacing and two child policy in practice in rural Vietnam: cross sectional survey

BMJ 1996; 313 doi: (Published 02 November 1996) Cite this as: BMJ 1996;313:1113
  1. H T Hoa, lecturer in epidemiologya,
  2. N V Toan, lecturer in epidemiologya,
  3. A Johansson, sociologistb,
  4. V T Hoa, lecturer in biologyc,
  5. B Hojer, associate professorb,
  6. L A Persson, associate professord
  1. a Department of Environment, Hygiene and Epidemiology, Hanoi School of Medicine, Hanoi, Vietnam
  2. b Department of Public Health Sciences, Division of International Health Care Research (IHCAR), Karolinska Institute, S-171 77 Stockholm, Sweden
  3. c Department of Biology, Thai Binh Medical College, Thai Binh, Vietnam
  4. d Department of Epidemiology and Public Health, Umea University, S-901 85 Umea, Sweden
  1. Correspondence to: Dr H T Hoa c/o Dr Hojer.
  • Accepted 6 September 1996


Objective: To explore the reproductive pattern of women in rural Vietnam in relation to the existing family planning policies and laws.

Design: Cross sectional survey with questionnaires on reproductive history.

Setting: Tien Hai, a district in Red River Delta area, where the population density is one of the highest in Vietnam.

Subjects: 1132 women who had at least one child under 5 years of age in April 1992.

Main outcome measures: Birth spacing and probability of having a third child.

Results: The mean age at first birth was 22.2 years. The average spacing between the first and the second child was 2.6 years. Mothers with a lower educational level, farmers, and women belonging to the Catholic religion had shorter spacing between the first and second child and also a higher probability of having a third child. In addition, women who had no sons or who had lost a previous child were more likely to have a third child.

Conclusion: Most families do not adhere to the official family planning policy, which was introduced in 1988, stipulating that each couple should have a maximum of two children with 3–5 years' spacing in between. More consideration should be given to family planning needs and perceptions of the population, supporting the woman to be in control of her fertility. This may imply improved contraceptive services and better consideration of sex issues and cultural differences as well as improved social support for elderly people.

Key messages

  • The adherence to the family planning policy was lower among mothers with less education, among farmers, and in Catholic households

  • The perceived necessity of having at least one son was a major factor for not adhering to the two child policy

  • Death of an earlier child increased the tendency to have a third child

  • More consideration should be given to the women's own family planning needs and percep- tions


Population and family planning programmes have been implemented in Vietnam since the early 1960s.1 2 In 1988 a decree on population and family planning policies from the Council of Ministers stipulated a maximum of two children per couple with 3–5 years of spacing between them.3 The objectives of the 1993 policy clearly defined and reconfirmed the previous two child policy and the targets expressed already in 1980.4

The total fertility has declined from over six children per woman in the early 1970s to fewer than four in the late 1980s. The growth rate is still high, however, at 2.3% in 1992.5 Thai Binh province, in the Red River Delta area, has one of the highest population densities in the country and an actively implemented family planning programme.5 6 7

During the current transition period in Vietnam, introduced in 1986, many changes have taken place under the common label doi moi (renovation). Greater responsibility has been given to the individual and the family,8 and a positive impact of economic reforms has been demonstrated—for example, in rice production.9

We studied reproductive patterns and how the family planning policies have been followed during a period when the implementation of the policy was strong and the transition to a market economy was at an early stage.

Subjects and methods

This study was performed in Tien Hai district, one of eight districts of the Thai Binh province in the Red River Delta area in Northern Vietnam. Tien Hai (population 194 000) has 35 communes involved in rice growing, fishing, and salt production. The sample was selected by use of a multistage cluster sampling technique to represent households with children under 5 years of age. Five communes were randomly selected, representing the different modes of production in a proportional way. Several villages in each commune were randomly selected, proportional to the size of the population in the communes. After choosing a random starting point, we took a sample within each village house by house, proportional to the size of the village. The sample comprised 1120 households and 1132 mothers.

The women were interviewed in their homes in April 1992 with a pretested questionnaire by a group of trained field workers. The mothers were asked about their year of birth and total number of past pregnancies and deliveries. Each outcome and date of the pregnancy was recorded. Spacing between births was calculated and related to background factors of the women and sex of the previous child or children. The death of a previous child was defined as the death of any child before the second or third pregnancy.

The educational level of women was classified as no formal education (illiterate or able to read and write without any formal schooling) or primary and secondary education. The religion of the women was classified into two main groups: Catholic or Buddhist and “other,” the second one including those who described themselves as having no religion but often practised Buddhist traditions.

Differences between means were analysed with analysis of variance, and P values of the F statistic were indicated. When the variances differed between the groups the non-parametric Kruskal-Wallis H was analysed and the P value was given.10 The latency period between age at marriage and age at first child was analysed in a multiple linear regression model where age at marriage, education, occupation, religion, and calendar time for birth of first child (before or after 1 April 1987) were independent variables. The time between the first and the second child, and the second and third child, respectively, was computed in bivariate and multivariate analyses by using the Cox proportional hazards model in relation to different background variables.11


Outcome of pregnancy—The 1132 mothers had experienced 1662 pregnancies during the period April 1987 to April 1992, of which 91% ended in the birth of a child and 9% were interrupted by an induced abortion, miscarriage, or stillbirth (table 1).

Table 1

Numbers (percentages) of pregnancies according to outcome in relation to pregnancy number. Data from reproductive health survey, Thai Binh, Vietnam, 1992

View this table:

Marriage and time between marriage and first child—For women who had had their first child in the past five years the median age at marriage was 20.9 years compared with 19.0 for those who had had their first child earlier (P<0.001). The time between marriage and birth of the first child, however, had decreased in the past five years, and consequently there was no significant difference in mother's age at first child between those delivering in the past five years (median 21.2 years) and earlier (21.1 years, P = 0.187). The time was shorter for those marrying later (1.4 v 2.4 years; P<0.001) and also shorter for those having formal education compared with those with no formal education (table 2). Occupation and religion were not significantly associated.

Table 2

Latency period between marriage and first child in relation to independent variables∗ in 1104 mothers with children under 5 years of age at time of study.+ Regression analysis with latency period (decimal years) as dependent variable. Y intercept = 6.302; x2 = 0.12

View this table:

Birth spacing between first and second child—The median duration of spacing (analysed by life table technique) between the first and the second child was 31.4 months. Women with secondary or higher education had significantly longer spacing between the first and the second child than did women with no formal education (37.4 months and 25.1 months, respectively; P = 0.004). The sex of the first child was not associated with the length of the latency period between the first and the second child (table 3), while the death of the first child was related to a shorter interval. Catholic families and younger women had shorter spacing. These associations were found in bivariate analyses as well as in a multivariate model (table 3).

Table 3

Bivariate and multivariate Cox regression analyses of time between first and second child in relation to sex of first child, death of first child, and characteristics of mother. Data from Thai Binh reproductive health survey, April 1992, on 426 mothers delivering for first time from April 1987 to April 1992 inclusive

View this table:

Probability of having a third child—Forty six per cent of mothers in this sample had had three children or more. Families who had previously lost a child, who had no son, in whom the mother had no formal education, who were farmers, and who belonged to the Catholic religion were more likely to have the third child (table 4). In a multivariate model three of these factors were significantly associated with the time to the third child: the sex of the previous children, the death of a previous child, and the religion of the family.

Table 4

Bivariate and multivariate Cox regression analyses of time between second and third in relation to sex of previous child, death of previous child, and characteristics of mother. Data from Thai Binh reproductive health survey, April 1992, on 347 mothers delivering for first time from April 1987 to April 1992 inclusive

View this table:


This study describes the reproductive pattern in relation to existing family planning policies during a period of rapid socioeconomic transition in Vietnam. Although the women in the study area were well informed about the family planning policy,12 most did not adhere to it.


Age at marriage and first birth corresponded to findings from the demographic health survey for rural northern Vietnam.13 14

In 1984 the provincial authorities of Thai Binh, facing an extremely high population density, introduced a rule of five years' spacing between the first two children. In Thai Binh the two child policy was forcefully implemented by introduction of financial penalties and by reduced opportunities to get promotion at work and allocation of land, while those who did not respect the five years' spacing faced much weaker or no sanctions.7 In an analysis of alternatives to the one child policy in China, Bongaarts in 1985 calculated that a two child policy, if accompanied by delayed childbearing through later age at marriage and child spacing more than four years, could be as effective as the one child policy in achieving China's current demographic goals.15 These findings underline the need not to focus exclusively on the two child policy in Vietnam.

Studies from other countries have also shown that education is a crucial factor associated with women's childbearing patterns.16 The tendency to have a third child if a previous child had died illustrates the strength of infant and child mortality as a determinant of population size. It is also well known that there is a strong association between birth spacing and child survival17 as well as physical and cognitive development.18


Women who had no son were more prone to have a third child. In the traditional Vietnamese society only sons can carry on the family line of descent and are responsible for the care of the elderly and the ancestral altar, and families without a son often feel inferior and ashamed.19 After marriage girls become members of the husband's family and are expected to work for their parents-in-law rather than for their parents. There is a well known folk song about daughters saying:

“Con gai la con nguoi ta

Con dau moi thuc me cha mua ve”

(Your daughter is a child of the other

Only a daughter-in-law is your daughter since you have paid for her).


To support the family planning policy, contraceptive and abortion services have been supplied free of charge through an extensive public health network. Results from different studies in Vietnam show that the intrauterine device is the contraceptive method most widely used by women.20 21 22 The abortion rate has recently increased in Thai Binh province as well as in many other provinces.23 24 Shortcomings in contraceptive services, in combination with a strong demand for fertility control by the government or by the couples themselves, seem to have resulted in an increasing abortion rate.

Our study highlights central ethical and service issues related to the implementation of the family planning policy in Vietnam. Would stronger policy measures increase the demographic impact of the programme and would this be at all ethically justifiable? The results illustrate the needs to strengthen and diversify the family planning services, offering culturally acceptable advice and help.

We thank the participating women in this survey, our colleagues at Thai Binh Medical College, and the staff of Centre for Human Resources for Health.


  • Funding This study emanates from a bilateral Vietnamese-Swedish research collaboration, supported by Sida/SAREC (Swedish Agency for Research Collaboration with Developing Countries).

  • Conflict of interest None


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
View Abstract