In this chapter, an overview about population policies and family planning in Vietnam and some typical results will be examined. Then more details will be focused on current contraceptive choice in Vietnam.

1/ Population policies and family planning programs in Vietnam and some typical results

Rapid population growth has shown its adverse effects on socio-economic development. Therefore, since the early 1960s, Vietnamese government has formally enacted population policies addressing to birth control. During the 1961-1975 stage, population policies were only carried out in Northern Vietnam with the name “guided birth campaign” and then “planned birth campaign”. The target of the campaign was 3-child family. The campaign focused on women in reproductive age that already had large number of children, specifically female civil servants, women in army forces and women in populous plains. The program focused mainly on IUD insertion. Contraceptive choice was determined by managers of planned birth programs, not by users. As a result, the crude birth rate declined by 10.7%o from 43.9%o in the early 1960s to 33.2%o in 1975. Total fertility rate also went down from 6.39 to 5.25. In Southern Vietnam, the government in Saigon did not begin to implement a family planning program until it was advocated by the U.S. Agency for International Development in 1971. Due to fierce war, family planning program was not commonly implemented.

After reunification in the year 1975, population policies and family planning programs were extended to all over the nation. The program was the expanded to include all women in the reproductive age and men who had wives in reproductive age. On June 30, 1989, the National Assembly passed the National Health Law. This law established a legal basis for a policy of the people's freedom to choose whatever contraceptive means they desired to implement family planning. The tendency of population increase to offset the deaths in the war in the south, then the Southwest and the North border wars hindered the implementation of family planning program nationwide. Consequently, after 17 years from 1975 to 1992, the crude birth rate was reduced only by 3.16%o from 33.2%o in 1975 to 30.04%o in 1992. Total fertility rate (TFR) declined from 5.25 to 3.8 respectively.

In short, till the year 1992, family planning-population policies had shown to be unsuccessful and ineffective. Birthrate and TFR were high at 30.04%o and 3.8 respectively. Moreover, the target of reducing population growth rate to 1.7% by 1990 wasn’t attained. Rapid population growth was in fact a hindrance to Vietnam economy where people’s living standards were still very low compared to the world. Vietnam faced the danger of lagging behind if birthrates and population growth rate were not restrained. For that reason, two important documents related to PFP policies were introduced: Resolution No. 4 on Population Policies, Population and Family Planning Strategy to the year 2000.

The general goal of Resolution No. 4 is "Obtaining families with few children who are healthy to allow a good living conditions". The specific goal is “Every husband and wife should have only one or two children so that each family (each couple) will have two children on social average by 2015, leading to a stabilization in the size of the population in the middle of the 21^st century. All efforts should be made to have a clear change right in the 1990s.”

In order to implement the goal, viewpoints and measures of Resolution No. 4, the Prime Minister signed Decision 270/TTg on 03/06/1993 approving Population and Family Planning Strategy to 2000, institutionalizing guideline, policies, measures and developing effectively Resolution No. 4 in the stage 1993-2000. The Strategy set the goals for the stage 1993-2000: “Quickly reducing the 3^rd birth rate so that TFR will be 2.9 or lower and population size will be less than 82 million people by the year 2000”.

      Resolution No. 4 and Population and Family Planning Strategy to the year 2000 were successfully implemented. By the 1^st July, 2000, Vietnam’s population size was 77.68 million people, which was less than the set goal of 82 million people in Population and Family Planning Strategy to 2000 by 4.32 million people. This substantial reduction contributes much to socio-economic development. The average population growth rate between the 1989 census and the 1999 census was 1.7%, which was the lowest in the past 45 years since independence in 1954 in the North. Total fertility rate declined rapidly from 4.0 in 1988 to 3.2 in 1994 and 2.3 in 1997. This indicator shows once again the big success in the implementation of the Strategy to the year 2000, where the set goal of TFR by 2000 is 2.9.  The crude birth rates go down from 28.5%o in 1993 to 21.9%o in 1997 and 19.9%o in 1999.

Recently, the introduction of the two documents- Vietnam Population Strategy in the stage 2001-2010, and National Target Program on Population and Family Planning to the year 20050- show that the focus of the population program has changed from a strong focus on fertility control to more concern about quality of the family planning program and improving the quality of the population. This is consistent with the direction that was raised at the Cairo Conference on Population and Development in 1994.

      2/ Current contraceptive choice in Vietnam:

      Knowledge of at least one contraceptive method had increased continuously since 1988 from 94 percent to 97 percent by 1994 and to 99.48% by 1997-1998. Up to 1998, knowledge of at least one method had become almost universal. As far as specific methods are concerned, there were marked increases in familiarity with each method between the three surveys- DHS 1988, ICDS 1994 and VLSS97-98. For the IUD, there is relatively little difference in levels of knowledge between the three surveys. The high level of familiarity with the IUD reflects the historical dominance of that method in the government family planning program. Contraceptive awareness were rather equal by 1997-1998. More than 90% of women surveyed indicated awareness of IUD, condoms, male and female sterilization. Knowledge of traditional methods were also high, approximately 85%. Even vaginal methods (Diaphragm, jelly), that are not yet widely available in Vietnam, are familiar to nearly 15% of ever-married women. Knowledge of contraceptive methods of ever married women improved greatly during the period 1992-1993 and 1997-1998 by all characteristics. According to VLSS 1992-1993, there were obvious variations in contraceptive awareness by characteristics. Specifically, contraceptive awareness was obviously higher among urban women, women living in the Red River Delta, women with higher educational level, and women with 3-5 children. However, there were relatively small differences in level of contraception knowledge by all characteristics in 1997-1998.

      In term of source of information supply on contraceptive methods, family planning clinic played the most important role, then mass media, other medical facility and mass organization. This is in compliance with the general guideline of government programs, which underline information – education – communication (IEC) as the main tools. For traditional methods such as rhythm method, withdrawal and herbal medicine, relative or husband or friend also accounted for an important part in providing information on these methods.Surprisingly, school played very little role in supplying information on contraception. This shows that sexual education was still not introduced into school.

      Overall contraceptive prevalence in Vietnam is high compared to LDCs’ rate (75.82 percent as compared to 67 percent) and has increased in the period 1992-93 and 1997-98. However, the use of modern methods is much lower than that of LDCs. In Vietnam, 71.96 percent of currently married contraceptive users were practicing modern contraception whereas modern methods accounted for 91 percent of total contraceptive use in LDCs.

      As far as specific contraceptive methods are concerned, IUD was still the most widely used method of contraception by 1997-98. The second most commonly used method was traditional methods such as withdrawal and rhythm. The high proportion of Vietnam women using traditional methods means that there is considerable demand for family planning services among the Vietnamese population, which has not been fully met by the current family planning programs. Use of condoms, pill and sterilization, had increased between 1993 and 1998 but was still low, which was 6.08 percent, 3.69 percent and 6.6 percent respectively in VLSS97-98. Use of injection was extremely low, only 0.18 percent. The use of most methods increased in 1997-1998 as compared to 1992-1993 only in absolute terms. However, their use declined in relative terms. This reflects a more diverse use of contraception. The reasons for this method-mix inVietnam are considered in detail in the final part of this chapter.

      There existed gender inequality in contraceptive use. Men still had very little responsibility in the use of contraception. The vast majority of married contraceptive users were women, which was shown in very high rate of use of IUD, pill and female sterilization. These three main "female" methods accounted for nearly two thirds of all contraceptive practice in Vietnam, whereas "male" methods took a very small rate. Although sterilization procedures are available for both men and women and male sterilization is simpler, safer and usually less expensive procedure than female sterilization, female sterilization procedures far outnumbered their male counterparts by 7.93 percent to 0.8 percent. Contraceptive use was almost put on the shoulder of women.

      Chapter 3: Determinants of modern contraceptive use in Vietnam

  The main focus in this chapter is to study factors affecting whether a married woman aged 15-49 has access and desire to use a modern method of contraception. Firstly, the empirical model for the analysis will be presented; Then bivariate and multivariate analysis of modern contraceptive use will be examined.

      1/ Empirical model:

      In the study on “The Impact of women’s Schooling on Fertility and Contraceptive Use: A Study of Fourteen Sub-Saharan African Countries”, Ainsworth, Beegle and Nyamete used the following reduced-form model for analyzing determinants of modern contraceptive use: y = b­₀ + b₁x₁+b₂x₂ +b₃x₃ +b₄x₄ +b₅x₅ +b₆x₆+ m

      where y is a dichotomous dependent variable that takes on the value zero or one, indicating current use of modern contraception; b­_₀ is an intercept; x₁ is the woman’s schooling; x₂ indicates urban residence; x₃ is the woman’s ethnic group; x₄ is the woman’s region of residence; x₅ is the woman’s religion; and x₆ is a group of variables proxying the household’s income or wealth.

      Basing on theoretical studies presented in chapter I, the above-mentioned model, the particular purpose of the research and the specific characteristics in Vietnam, the author suggests the following model for analyzing the current use of modern contraception in Vietnam: FC = FC(x₁, x₂, x₃, x₄, x₅, x₆, x₇, x₇², x₈)        (3.1).

      Where FC is the current use of modern contraception; x₁ indicates urban residence; x₂ is woman’s region of residence; x₃ is household’s income; x₄ is the woman’s religion; x₅ is the woman’s ethnic group; x₆ is the woman’s education; x₇ is the total living children the woman has, entered in quadratic form; x₈ indicates whether the woman has a son.

      Since the dependent variable is dichotomous, taking a value of one or zero depending on whether a married woman aged 15-49 is currently using a modern method of contraception or not. Logistic model is used to estimate parameters of equation (3.1). As mentioned in the introduction part, the data the author has at hand at individual level cannot be used to estimate the logistic regression model by the standard OLS. Therefore, the author uses the maximum likelihood method to estimate the parameters.

      To examine in more detail the nonlinearity of the effect of number of living children on modern contraceptive use, specifically, to account for the fact that the use of modern contraception rises with the total living children the woman has, peaks, and then declines, the author controls for variables "Total living children" and  "number of living children squared" in the model.

      In model (3.1), the dependent variable is "Is a currently married woman in the reproductive age (15-49 years of age) using a modern method of contraception". The dummy dependent variable takes on the value 1 if the woman is using a modern method of contraception and 0 if she is not.

      In VLSS 1997-1998, 4411 ever-married women aged 15-49 were asked about their contraceptive use. However, only a sample of 4104 currently married women was asked about their current use of contraception. That is the reason why only currently married women aged 15-49 are included to study the situation of current contraceptive use in Vietnam.

      All the necessary independent variables have been specified in model (3.1). The independent variables’ definitions as well as characteristics are shown in table 3.1. It should be mentioned that:

      - In the thesis the author use expenditure quintiles as a proxy for income although data on income can be collected in VLSS97-98. The reason is: "First, consumption expenditure data are likely to be more accurate than income data, because questions on expenditure are often easier to answer (in particular, the self-employed have difficulty answering questions on income) and because some households are reluctant to reveal their true income. Second, income raises living standards only if it is consumed, and past income (savings) or borrowing can be used for consumption purposes. Thus consumption data are likely to reflect household’s welfare levels more accurately than would income data." (Glewwe, Zaman and Gragnolati, 1999, p.3)

Table 3.1: Variables’ definitions and characteristics

Source: Author’s calculation based on Vietnam Living Standards Survey 1997-98

      - In the model, family size is represented by the total living children a couple has. Here, the total living children not the total number of children is considered. This is because for parents the number of living children is their desired number of children. And it is this number of living children that affects their decision making on whether to have more children and whether to use contraception or not. “Number of children squared” is included in the model to explain its predicted nonlinear effect on modern contraceptive practice.

2/ Bivariate analysis of modern contraceptive use

Table 3.2 shows the effect of individual characteristics on the use of modern contraception. This is of interest to test the expectations discussed in chapter I about factors affecting modern contraceptive use.

As shown in the table, the prevalence of modern contraception is higher among urban than rural women. However, the difference is not large. One possible reason is that urban women have higher use of traditional methods. The other possible reason is that rural household finds the participation of children in the household economy less important. This is attributable to increasingly limited amount of agricultural land. Therefore, the demand for children to guarantee labor supply for farming decreases and the demand for contraceptive use increases in rural area. Better knowledge on contraception of rural women between VLSS92-93 and VLSS97-98 is another reason for high use rate of modern contraception in rural area.

Substantial regional variation in current use of modern contraception is evident. Use of modern contraception ranges from 30.6 percent of women in Central Highlands to 67.54 percent of currently married women in Red River Delta. In general, use of modern methods is noticeably lower in the three southernmost regions. It may be because use of traditional methods is higher in these three regions. Another reason for higher use of modern methods in the north than in the south is that family planning has been implemented in the north in the early 1960s while it didn’t begin in the south until it was advocated by the U.S. Agency for International Development in 1971. The distinctly lower level of modern method in the Central Highlands may result from socio-economic, geographic and cultural characteristics of this region. In developed region, women are generally characterized by lower demand for children and hence higher demand for modern contraception and vice verse in the less developed regions. The higher use rate in North Mountain and Midlands can be explained by the early implementation of family planning programs in the North in the early 1960s. The low level of modern contraception in Mekong Delta region may reflect better agricultural opportunities in this region, which can be seen from table 3.3. The table shows that the land area per capita under annual crop, one of the main sources of farming households’ income, in the Mekong River Delta is nearly three times as large as that in the Red River Delta and The North Central Coast. Therefore, the demand for children to guarantee an adequate labor supply for farming may be high, and hence modern contraceptive use are low.

Modern contraceptive use is higher for Chinese and Kinh group than for the minority. This is what we expected. On the one hand, Chinese and Kinh group may have better access to family planning. On the other hand, socioeconomic characteristics, customs of marriage and reproductive practices lead to lower use of contraception in minority group.

As far as religion is concerned, Buddhist and Catholic women have lower level of modern contraceptive use.

Table 3.2: Current use of a Modern Method of Contraception

by Selected Individual Characteristics by Currently Married Women Aged 15-49

Source: Author’s calculation based on Vietnam Living Standards Survey1997-98

Table 3.3: Agricultural and forestry land area per capita by type of land (m²)

Source: Based on GSO (1999) Vietnam Living Standards Survey 1997-98

The lowest quintile group has lowest proportion of modern contraceptive use of 56.62 percent. However, there is not much difference in the rest quintile groups. The effect of income on modern contraceptive use is not clear. This is what we expect from the discussion in chapter I.

There is a positive relation between women’s educational level and modern contraceptive use. This is also what we expected and the explanation is the same as in chapter I. After renovation policy, economic opportunities are broaden to women, especially to educated ones. Moreover, state supply of childcare service is shrunk. Women have to spend more time taking the children themselves. Therefore, opportunity cost of childbearing and rearing increased very much, especially for better-educated women. Income effects, higher demand for child schooling, lower child mortality and more effective use of contraception are the four paths by which women’s educational levels have a strong impact on modern contraceptive use.

The table indicates that a substantially higher percentage of women with at least a son were practicing contraception than those without a son (62.2 percent and 41.64 percent respectively). This is in compliance with the findings in Dominique Haughton and Jonathan Haughton’s study (1999) on son preference in Vietnam. According to them, son preference is one of the reasons influencing the fertility behavior of married couple. According to Vietnamese tradition, the desire for a son to continue the ancestral line is regarded as an important factor affecting fertility, and hence modern contraception. Under social pressure, women consider their reproductive behavior as the most important factor determining their status, relationship with their husband, husband's family, relatives, and neighbors. This may lead married couples who have an adequate number of children, but all of female gender, to desire more children and practice less contraception.    

As far as total living children is concerned, the relationship between family size and modern contraceptive use is nonlinear. Women with 3-5 children have the highest rate of modern contraceptive prevalence of 64.11 percent. Two-child policy in Vietnam has a strong impact on women’s desired number of children. Therefore, when they reach 2 children, the use of modern contraception increases sharply. The very low modern contraceptive prevalence among women with no number of living children (only 6.92 percent) emphasizes the immediate preference to give birth after marriage. Women with 6 children up have lower proportion of modern contraceptive use than do women with 3-5 children (only 47.37 percent). The reasons may be that women of higher parity tend to be older and could be of the generation where large families were desirable, or may have reached menopause or may shift to traditional methods as analyzed in the theoretical framework section.

      The factors affecting the use of modern contraception may be interrelated. For example, more educated women may be less influenced by son preference than others may. Therefore, we need a regression model in which all of the factors are included to see their partial effects on the use of contraception. That is the reason why we need to run logistic regression model which will be presented in the following section.

3/ Multivariate analysis of modern contraceptive use

The analyses presented in previous sections indicate that modern contraceptive use is associated with a variety of individual characteristics. In the following analyses, regression procedure will be used to identify the strength of association that each variable has on use of modern contraceptives when other factors are held constant. Comparison with the study by Pham Thanh Nhiem (1999) will be made but are only relative. This can be explained by the fact that Pham Thanh Nhiem estimated logistic models to examine the determinants of use of any contraceptive or of use of modern contraceptives for those women who have decided to use contraception only (see table 3.6 and table 3.7).

Region of residence shows to have the significantly strongest impact on modern contraceptive use, which is expressed by the highest levels of coefficients and odds ratios. The result of the regression indicates that the prevalence of modern methods between regions is in line with what we found in table 3.2 even when other variables are controlled for (see table 3.5 for comparison). The positive coefficients or the higher-than-one odds ratio for all regions listed in table 3.4 imply that the Central Highlands - the reference region - has the lowest probability of using modern contraception, even when controlling for other factors. Women in the Red River Delta are the most likely to use modern contraception. A woman in the Red River Delta has the highest odds of using modern contraception that are 3.550 times the odds for woman in the Central Highlands, other factors held constant. The significant strong impact of variable "region of residence" is persistent. In his study based on data from VLSS1992-93, Pham Thanh Nhiem also found the strongest effect of this variable on use of any contraception (see table 3.6). It should be noted that in his study on determinants of modern contraceptive use for women on contraceptives only, this variable is among the two variables that were significant (see table 3.7).

Rather high odds ratios induce education to rank second in its influence on modern contraceptive use. The effect is highly significant. The higher educational level a woman obtains, the higher the odds ratio are. This indicates that the probability of using modern contraception rises with female education. If other characteristics give a woman with no education 50 percent probability of using modern contraception, they will give a woman with primary education 59 percent probability, a woman with lower secondary education 65 percent probability, a woman with upper secondary education 68.8 percent probability, and woman with university or higher education 77.1 percent probability of so doing (see table 3.5). Women’s schooling has a consistently strong positive relation with contraceptive use. Pham Thanh  Nhiem also found in his study that women with more schooling were also increasingly likely to be practicing contraception (see table 3.6).

Pham Thanh Nhiem showed the significantly strong influence of variable "number of living children" on the use of any contraception method and on modern contraceptive use for women who were on contraceptives. The difference in family size also explains to a large extent the difference in modern contraceptive use among married women in the reproductive age with data from VLSS97-98. With the odds ratio in favor of using modern contraception of 2.221, "Total living children" almost ranks the third in affecting modern contraceptive use. Each additional living child increases the odds of modern contraceptive use by a factor of 2.221. This effect is translated into higher estimated probability, displayed in table 3.5, implying that when the number of living children rises, mothers are more likely to use modern contraception. However, the probability peaks at five children (see table 3.2), and after that point the likelihood of using modern contraception actually falls, reflecting the onset of menopause. This is evidenced by a significantly negative coefficient for variable "number of living children squared".

As predicted, the existence of a boy strongly affects whether a woman is using modern contraception or not. Pham Thanh Nhiem also found the differential in overall contraceptive use by son preference (see table 3.6). The higher-than-one odds ratio of 1.469 demonstrates that a woman with at least a son is significantly more likely to be using modern contraception than a woman with no son when other variables are controlled for. Son preference is the immediate reason for this difference. A woman who doesn’t have a son with a probability of using modern contraception at 70 percent, will have an estimated probability of 77.4 percent if she has at least a son.

Religion has slightly less strong influence on women’s behavior of modern contraceptive use than son preference (Both the significance of the coefficient and the size of odds ratio of "no religion" variable are smaller than those for "one or more son" variable). Some religions oppose the use of modern contraceptives to control birth. It can be seen that those women with no religion have the odds of using modern contraception that are 1.283 times the odds for those with a religion. It follows that women with no religion have more chance of using modern contraception compared to women who belong to a religion. When the prevalence of modern contraceptives among women with religion is 50 percent, the rate for women in the same circumstances with no religion is 56.2 percent (see table 3.5). This difference is statistically significant. The differential in modern contraceptive use by women’s religion is also a new finding as compared to Pham Thanh Nhiem’s study in examining contraceptive use in Vietnam. 

Table 3.4: Logistic model of current use of a modern method of contraception

related to selected individual characteristics for currently married women aged 15-49

Notes: - Number of observations = 4104. LR chi2(20)=444.20. Pseudo R²=0.0792

- The reference region is Central Highlands; the reference expenditure per capita group is the lowest quintile; the reference educational group is women who have no education.

Source: Author’s calculation based on Vietnam Living Standards Survey 1997-98

Table 3.5: Estimated probability that a currently married woman is using a modern method of contraception (percent)

Source: Author’s calculation

As far as the proxy for income - expenditure quintile - is concerned, most expenditure quintiles don’t have significant effect on modern contraceptive use except for the low-middle quintile, which is significant at 5 percent. On the one hand, the focus of family planning program in Vietnam on the poor, which has been specified in population policies in Vietnam mentioned in chapter II, mutes the differentials in modern contraceptive use by income. As analyzed in chapter I, on the other hand, a woman with higher income is more likely to use modern contraception. On the other hand, a woman will practice less contraception when she has higher income. This conflicting effect of income on modern contraceptive use is another possible reason why income plays no role in modern contraceptive use. Basing on data from VLSS1992-93, Pham Thanh Nhiem found that expenditure quintile was positively related to contraceptive use. In other words, contraceptive use increased as income rose and other variables are held constant (see table 3.6). This suggests that the focus of family planning in Vietnam on the poor has been successful in recent years.    

Like the study by Pham Thanh Nhiem, In this study, urban residence has no significant effect on modern contraceptive use. The focus of family planning program in Vietnam on rural area also mutes the differentials in modern contraceptive use by area of residence.

It is surprising that woman’s ethnicity has no impact on modern contraceptive use (p value is too large of 0.716) when other factors are controlled for. This unexpected result may be due to problems in data collection or processing or due to sampling.

Notes: Based on 2,769 observations. Pseudo R² = 0.12. The reference region is the Central Highlands; the reference expenditure per capita group is the poorest quintile; and the reference educational group is mothers who have not completed a primary education.

 Source: Pham Thanh Nhiem (1999), "Contraceptive Use" in Haughton et al. (Eds), Health and Wealth in Vietnam: An Analysis of Household Living Standards, Institute of Southeast Asian Studies, Singapore, p.35

Table 3.7: Logistic Model of Use of Modern Contraceptives

(for women on contraceptives only)

Notes: Based on 1,921 observations. Pseudo R² = 0.025. The reference regions are the Northern Uplands, Central Coast and Central Highlands; the reference expenditure per capita group is the poorest quintile; and the reference educational group is mothers who have not gone beyond a primary education.

Source: Pham Thanh Nhiem (1999), "Contraceptive Use" in Haughton et al. (Eds), Health and Wealth in Vietnam: An Analysis of Household Living Standards, Institute of Southeast Asian Studies, Singapore, p.36

Conclusions and Recommendations

The thesis can be concluded as follows:

An overview about contraceptive use in Vietnam shows that there has been a substantial improvement in the knowledge of contraception during the period 1993-1998 so that knowledge of contraception has become almost universal. There are relatively small variations in contraceptive awareness by woman’s characteristics in 1997-1998. Level of knowledge by all characteristics is very high, reaching almost 100 percent for all characteristics. This provides evidence that the population and family planning program has made progress in achieving the goals established by the government. 

Family planning clinic, mass media play important roles in providing information on contraceptive methods. It should be noted that school still plays very little role in information supply.

The overall contraceptive prevalence rate for currently married women of reproductive age was 75.82 percent in VLSS97-98. This represents a 6.42 percentage point increase over the rate found in the VLSS92-93 (69.4 percent). While the overall contraceptive prevalence in Vietnam was quite high, the use of modern, more effective methods was still low of 54.56 percent of currently married women and of 71.96 percent of married contraceptive users. The traditional methods still make up 28.04 % of contraceptive use. The large extent of the use of traditional methods will lead to high rate of unwanted pregnancy if these methods are not correctly used.

      Men still had very little responsibility in the use of contraception. Contraceptive use was almost put on the shoulder of women. The vast majority of married contraceptive users were women, which was shown in very high rate of use of IUD, pill and female sterilization. These three main "female" methods accounted for nearly two thirds of all contraceptive practice in Vietnam, whereas "male" methods took a very small rate (both condoms and male sterilization made up only 8.82% of contraceptive users).

Although contraceptive awareness by methods in Vietnam is rather high in 1997-98, their uses are still not diversified. IUD is still the most widely used among modern methods of contraception. According to VLSS97-98, more than 38 percent married women of reproductive age or 50.48 percent of married contraceptive users reported that they were currently using the IUD. The high level of familiarity with the IUD reflects the historical dominance of that method in the government program from which the vast majority of women obtain their contraceptives.

The practice of rhythm and withdrawal is substantial in Vietnam. According to VLSS97-98, around one-sixth of married women in the reproductive age said that they currently practice withdrawal and 15.33 percent reported rhythm as their method. The high level of the use of these less effective methods of contraception suggests that there is still a considerable demand for family planning services.

The still low prevalence of more effective, modern methods leads to the demand to study factors affecting its use so that appropriate policies will be desired to promote its use in the effort to reach the goals set by the government on fertility decline. Specifically, the prevalence of modern contraception should increase to 70% by the year 2010; the declining trend in fertility should be maintained so that replacement level of fertility will be reached on national average by the year 2005 at the latest; and above all, population stabilization to allow a good living standard should be attained. Logistic regression model is used as an instrument to analyze the affecting factors. In the model, only woman’s individual characteristics are considered. The reasons for this narrow focus are shown clearly in the introduction section.

      The regression results show that modern contraceptive use is responsive to woman’s characteristics. The responses are usually, but not always, expected. Specifically, woman’s region of residence has the significantly greatest impact on modern contraceptive use. Then are other characteristics of women such as education, family size, son preference, and religion. Women in the north, more educated women, women with at least a son and women with no religion are significantly more likely to use modern methods of contraception. Most of these results were found in Pham Thanh Nhiem’s model of overall contraceptive use in Vietnam using data from VLSS1992-93 except for "religion" variable. The differential in modern contraceptive use by women’s religion is also a new finding as compared to Pham Thanh Nhiem’s study in examining contraceptive use in Vietnam.

      However, income (as proxied by expenditure quintile), which had a significant effect on overall contraceptive use in 1992-93, becomes insignificant in explaining differences in modern contraceptive use in VLSS1997-98. This suggests that the focus of family planning in Vietnam on the poor has been successful in recent years. The insignificant impact of area of residence on modern contraceptive use also results from the focus of family planning program on rural area.

The finding that woman’s ethnicity has no impact on modern contraceptive use when other factors are controlled for is astonishing. This unexpected result may be due to problems in data collection or processing or due to sampling.

Policy recommendations:

So far we have identified five key factors which affect a woman’s decision to use modern contraception: woman’s region of residence, woman’s education, family size, son preference, and woman’s religion. Based on these findings, the author suggests the following policy measures to increase modern contraceptive use in Vietnam:  

Firstly, enhancing women’s status through the increased educational and employment opportunities for women. With more opportunities for them to become part of the paid labor force and to receive an education, they are much less likely than formerly to think of themselves primarily as mothers. Education of the mother is especially important in reducing the desire to bear children, and hence increasing the use of modern contraception by bringing better knowledge of birth control methods, delay of marriage, and greater opportunity costs of childbearing because the potential mother is more employable. Employability brings in turn less need for support from the children in old age. Educated mothers also tend to desire more education for their children, which raise the cost of procreation. Education of the mother also is correlated with reduced infant mortality, and hence the motive to bear more children to ensure the survival of some is reduced.

With better social, political and economic status, women will have more control over their lives within and outside the home including fertility and family planning decisions. This will to some extent help to reduce gender inequality in practicing contraception. Son preference can, to a certain extent, be alleviated by increasing women’s economic status and income earning potential, which is currently perceived by parents to be too low to count on as old age insurance.

Secondly, investment in female education should take gender inequality in education field into consideration since girl face limited education than boy. There are many reasons why parents might choose to invest more in educating boy than girls:

-         The financial returns from girls’ schooling are perceived to be lower than that for boys due to discrimination in the labor market. On an aggregate level, women tend to be concentrated in less lucrative sectors and occupations, and within those sectors in certain types of employment characterized by low pay, low skills or little opportunity for advancement.

-         Parents expect more direct benefit from investing in sons who typically provide for their parents in their old age, while daughters tend to leave upon marriage to joint a different household economic unit.

-         Girls are often pulled out of school earlier to help the family with productive and domestic work, for example, caring for younger siblings to free mother for work outside the home, tending farm animals, fetching fuel-wood, water and forage, and helping in the fieldwork boy.

Therefore, policies should be established to discourage this gender gap. propaganda is necessary so that everybody is well aware of the long term benefits and the multiplying effect of educating women. Moreover, female school attendance should be monitored in every school and bonuses provided to schools that maintain high performance in this respect. Incentives such as information, career orientation and financial assistance could be offered to girl students.   

Thirdly, Information-Education-Communication (IEC), Behavior Change Communication (BCC) activities and provision of family planning service should be further implemented to urge people to use modern contraception and to change people’s family planning behaviors. Sexual and reproductive health education should be introduced into school, especially lower and upper secondary school, extensively and intensively. This is a very effective channel of diffusing knowledge on family planning. Integrating the diffusion of family planning with other services, especially maternal and child health services, seems to help broaden knowledge on and promote the use of modern contraception. Since religion has a significant impact on modern contraceptive use, IEC and BCC activities should target religious groups to promote their use of modern contraception. The idea that use of modern contraceptive methods is immoral and unacceptable should be eliminated. The lower probability of modern contraceptive use in the south than in the north emphasizes the need to focus more attention on education and information activities and provision of modern contraceptives in the southern regions to change their behaviors from using traditional to more effective, modern methods of contraception. IEC and BCC activities should also target those regions with difficult socio-economic conditions such as the Central Highlands. The strong son preference in Vietnam has led to lower probability of modern contraceptive use for couples who haven’t got a son. In Vietnam son preference is related to the traditional role of son in ancestor worship and more pragmatically to parents’ reliance on sons to support them in old age. The first cause can only be addressed by changing deeply entrenched cultural values, countering the tradition of requiring a son to continue the lineage. Family planning activities should promote the idea that daughter can take care of parents as well as sons. Sexual and gender equality education for couple, especially young couples, is necessary to further promote modern contraceptive use.

Fourthly, family planning program in Vietnam has shown to be very successful in targeting the poor, remote and rural areas. Therefore, the main focus of population and family planning policies on rural, poor and remote areas should be maintained.

Finally, targeting the disadvantage areas through the provision of family planning services, buildup of education and health infrastructure. Priority should be given to the Central Highlands. Large family size in difficult socio-economic regions is aimed at taking advantage of child labor. Therefore, enhancement the enforcement of child labor law is a need to promote the use of modern contraception.

Limitation of the thesis and area for further research:

      In VLSS 1997-1998, 4411 ever-married women aged 15-49 were asked about their contraceptive use. However, only a sample of 4104 currently married women was asked about their current use of contraception. That is the reason why in the thesis, only currently married women aged 15-49 are included to study the situation of current contraceptive use in Vietnam. A problem arises from the choice of sample of currently married women in examining determinants of current use of modern contraception in Vietnam. By conditioning on marriage, this study has not captured the full effect of independent variables on current contraceptive use (before marriage). This is a limitation of the research.

In studying contraceptive use in Vietnam, an approach to promote the use of modern contraceptives is to study factors behind their use. The author uses this approach in the thesis. Other approaches are possible. They are to study determinants of non-use of modern methods of contraception or to study determinants of the use of traditional methods.

Another important area for research is looking at the determinants of men using contraception (i.e. male sterilization or condoms) by looking at the characteristics of the husbands of women interviewed about contraceptive use in the VLSS. This can help not only to promote the use of contraception in Vietnam but also to design policies to motivate men to share more responsibility in contraceptive use with women.